Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

20100301-CRDAC-B1-01-FDA

VIEWS: 4 PAGES: 135

									                                       Briefing Package

          Division of Special Pathogen and Transplant Products
                    Office of Antimicrobial Products
             Center for Drug Evaluation and Research, FDA

                                        BLA: 125,288
                                     Belatacept Injection

                    Applicant: Bristol-Myers Squibb Company


      Cardiovascular and Renal Drugs Advisory Committee Meeting
                            March 1, 2010




                                     Proposed Indication:
  Prophylaxis of organ rejection and preservation of a functioning allograft in adult patients
  receiving renal transplants. Belatacept has been used in combination with an interleukin-2
           receptor antagonist, mycophenolate mofetil (MMF), and corticosteroids.

Proposed Dosing Regimen:

Dosing for Initial Phase                                             Dose
      Day of transplantation, prior to implantation (Day 1)        10 mg/kg
      Day 5, Day 14, and Day 28 (1 month after transplantation)    10 mg/kg
      Month 2 and 3 after transplantation                          10 mg/kg
Dosing for Maintenance Phase                                         Dose
      Monthly, starting at Month 4 after transplantation            5 mg/kg




                                                                                                 1
                                            TABLE OF CONTENTS

1.      Background ............................................................................................ 7
2.      Overview of Clinical Development Program ...................................... 9
3.      Dose Selection ....................................................................................... 12
4.      Clinical Pharmacology ........................................................................ 13
     4.1.  Pharmacokinetics .............................................................................................. 13
     4.2.  Belatacept Trough Concentrations Observed in Phase 3 trials......................... 14
     4.3.  Exposure-Response Relationships .................................................................... 15
     4.4.  Mycophenolate Mofetil (MMF) Dosing and Mycophenolic Acid (MPA)
     Exposure Observed in Phase 3 Trials ........................................................................... 17
     4.5.  Trough Cyclosporine Concentrations of Observed in Phase 3 Trials............... 18
5. Overview of Efficacy Endpoints ......................................................... 19
     5.1.       Patient and Graft Survival at 12 Months........................................................... 20
     5.2.       Composite Renal Impairment and mean Glomerular Filtration Rate (GFR).... 21
     5.3.       Acute Rejection................................................................................................. 22
     5.4.       Acute Rejection – FDA analysis....................................................................... 23
     5.5.       Chronic Allograft Nephropathy ........................................................................ 23
6. Clinical Efficacy ................................................................................... 25
     6.1.    Patient Disposition ............................................................................................ 25
     6.2.    Primary Efficacy Results .................................................................................. 26
        6.2.1.    Patient and Graft Survival......................................................................... 26
        6.2.2.    Composite Renal Endpoint ....................................................................... 26
        6.2.3.    Acute Rejection......................................................................................... 27
     6.3.    Additional Efficacy Results .............................................................................. 31
        6.3.1.    GFR measured at Month 12...................................................................... 31
        6.3.2.    Chronic Allograft Nephropathy at Month 12............................................ 32
     6.4.    Summary of Efficacy ........................................................................................ 32
7. Clinical Safety....................................................................................... 34
     7.1.    Deaths ............................................................................................................... 35
        7.1.1.    Discussion of Death Narratives ................................................................ 36
     7.2.    Graft Loss (and Deaths) .................................................................................... 38
     7.3.    Post Transplant Lymphoproliferative Disorder (PTLD)................................... 39
        7.3.1.    Summary PTLD Cases Observed in the Trials 008, 027, and 100 by
        Treatment Group....................................................................................................... 41
        7.3.2.    Preponderance of CNS PTLD among Belatacept Treated Patients .......... 45
        7.3.3.    Incidence of PTLD in Relation to EBV Serostatus .................................. 46
        7.3.4.    PTLD rates among EBV+ Kidney Transplant Recipients Observed in the
        CsA Control Groups from Randomized Controlled Trials Previously Submitted to
        FDA       49
        7.3.5.    PTLD Rates Among EBV+ Kidney Transplant Recipients According to
        Registry and Claims Data ......................................................................................... 49
        7.3.6.    PTLD in the Belatacept Liver Transplant Program .................................. 52
     7.4.    Progressive Multifocal Leukoencephalopathy (PML)...................................... 53
     7.5.    Non-Fatal Serious Adverse Events (SAEs) ...................................................... 54


                                                                                                                                    2
  7.6.    Adverse Events Leading to Treatment Discontinuation ................................... 55
  7.7.    Common Adverse Events ................................................................................. 56
  7.8.    Acute Rejection................................................................................................. 57
  7.9.    Glomerular Filtration Rate (GFR) .................................................................... 59
  7.10.     Cardiovascular Risk Factors ......................................................................... 62
  7.11.     Adverse Events of Special Interest ............................................................... 65
     7.11.1. Infections................................................................................................... 65
     7.11.2. Thrombosis ............................................................................................... 67
     7.11.3. Proteinuria................................................................................................. 68
     7.11.4. Infusion-Related........................................................................................ 68
     7.11.5. Autoimmunity ........................................................................................... 69
     7.11.6. Immunogenicity ........................................................................................ 71
     7.11.7. Malignancies Other than PTLD ................................................................ 71
     7.11.8. Cardiac – QT Prolongation ....................................................................... 72
8.  Risk Evaluation and Mitigation Strategy (REMS)........................... 72
9.  Overall Risk Benefit............................................................................. 74
10. Draft Questions to the Committee...................................................... 78
11. Appendix A: NonInferiority (NI) Margin Justification .................. 79
12. Appendix B: PTLD Analysis............................................................... 83
13. Appendix C: Personal Communication from Dr. Gerhard Opelz.. 92
   14. Appendix D: Exploratory Analysis to Identify Additional Risk
Factors for Belatacept Associated PTLD ................................................. 93
15. Appendix E: REMS GUIDANCE ...................................................... 96




                                                                                                                           3
                                                       Table of Tables
Table 1: Selected Demographics in the Phase 3 Kidney Trials .................................. 12
Table 2: Dosing for Belatacept LI Regimen (Phase 3 trials)....................................... 12
Table 3: Pharmacokinetic parameters (Mean ± SD [Range]) in Healthy Subjects and
de novo Kidney Transplant Patients ............................................................................. 14
Table 4: Belatacept Threshold Trough Concentrations for Trials 008 and 027 ....... 15
Table 5: Relationship between Belatacept Trough Concentrations (Median; 10-90th
percentile) and Proportion of Patients with a Decrease in Measured GFR ≥ 10
mL/min/1.73 m2 from Month 3 to Month 12 ................................................................ 16
Table 6: Percent of Patients who Received MMF 2 g/day at Given Times in Phase 3
Trials 008 and 027........................................................................................................... 17
Table 7: Reasons for Treatment Discontinuations at 12 Months in IM103008 ........ 25
Table 8: Reasons for Treatment Discontinuations at 12 Months in IM103027 ........ 25
Table 9: Patient and Graft Survival at 12 months....................................................... 26
Table 10: Composite Measured GFR Endpoint........................................................... 27
Table 11: Acute Rejection (as Defined by the Applicant) at 12 months .................... 28
Table 12: Additional Patients Included in FDA’s Definition of BPAR in Study 008
........................................................................................................................................... 29
Table 13: Additional Patients Included in FDA’s Definition of BPAR in Study 027
........................................................................................................................................... 30
Table 14: FDA’s Analysis of Biopsy Proven Acute Rejection, Graft Loss, and Death
at 12 Months .................................................................................................................... 31
Table 15: Measured GFR at Month 12 ......................................................................... 31
Table 16: Prevalence of Chronic Allograft Nephropathy (CAN) at 12 Months ....... 32
Table 17: Safety Population from Pooled Studies....................................................... 34
Table 18: Deaths Occurring in the Phase 3 Trials ....................................................... 36
Table 19: Narratives of Deaths among CsA Patients in IM103008 at 24 Months .... 36
Table 20: Narratives of Deaths among Belatacept LI Patients in IM103008 at 24
Months ............................................................................................................................. 37
Table 21: Narratives of Deaths among Belatacept MI Patients in IM103008 at 24
Months ............................................................................................................................. 37
Table 22: Narratives of Deaths among CsA Patients in IM103027 at 24 Months .... 37
Table 23: Narratives of Deaths among Belatacept LI Patients in IM103027 at 24
Months ............................................................................................................................. 38
Table 24: Narratives of Deaths among Belatacept MI Patients in IM103027 at 24
Months ............................................................................................................................. 38
Table 25: Patient and Graft Status in the Phase 3 Trials............................................ 39
Table 26: Death and/or Graft Loss at 12 and 24 Months and Safety Update ........... 39
Table 27: All PTLD Cases for Belatacept Trials 008, 027 and 100 ........................... 40
Table 28: Summary of PTLD Cases Observed in Belatacept Trials 008, 027 and 100
........................................................................................................................................... 41
Table 29: Estimated Number Needed to Harm based on Incidence of PTLD:
Belatacept Trials 008, 027 and 100................................................................................ 49




                                                                                                                                            4
Table 30: 2-year PTLD incidence rates/100 Person-Years among EBV+ Belatacept
Patients in IM103100, IM103008, and IM103027 compared to EBV+ Patients
Maintained on CNI-based Regimens Followed in the UNOS and CTS Registries ... 50
Table 31: Estimated Number Needed to Harm (NNTH) based on registry data ..... 51
Table 32: Serious Adverse Events in at least 2% of Patients at Month 12................ 55
Table 33: Adverse Events Leading to Treatment Discontinuation in ≥1% of Patients
Up to Month 12 (Pooled Trials 008, 027, and 100)....................................................... 56
Table 34: Selected Adverse Events Associated with Class Effects Up to Month 12 . 57
Table 35: Acute Rejection (as Defined by the Applicant) at 12 Months.................... 58
Table 36: Use of Lymphocyte Depleting Therapy Through Month 24...................... 59
Table 37: Slope for Calculated GFR from Month 3 to 24........................................... 61
Table 38: Mean Blood Pressure at Month 12............................................................... 63
Table 39: Serum Lipids at Month 12 ............................................................................ 64
Table 40: NODAT at Month 12 ..................................................................................... 65
Table 41: Infections Reported as Adverse Events up to 12 Months in Trials 008, 027,
and 100 ............................................................................................................................. 66
Table 42: Cases of CNS Infections in Trials 008, 027, and 100 up to Database Lock
........................................................................................................................................... 67
Table 43: Cases of Tuberculosis in the Phase 3 Trials at Database Lock.................. 67
Table 44. 12-Month Acute Rejection and Efficacy Failure Rates from Literature .. 80
Table 45: PTLD Incidence at 24-Months: Belatacept Trials 008, 027 and 100......... 84
Table 46: Estimated Number Needed to Harm: Belatacept Studies 008, 027 and 100
........................................................................................................................................... 85
Table 47: Summary of PTLD from Previous Randomized Clinical Trials that
Included a CsA-based Regimen and the CsA-based Regimens from the Belatacept
Studies (008, 027 and 100) .............................................................................................. 86
Table 48: Estimated NNTH in EBV+ Patients Comparing Belatacept with CsA from
Prior RCTs ...................................................................................................................... 87
Table 49: Estimated Numbers Needed to Harm: Pooled Belatacept Compared to
FDA UNOS Estimates..................................................................................................... 87
Table 50: Estimated Numbers Needed to Harm: Pooled Belatacept Compared to
BMS and SRTR UNOS Estimates and Opelz Analysis ............................................... 88
Table 51: Comparison of PTLD Rates based on EBV status...................................... 89
Table 52: Comparison of Methods used in Analysis of the OPTN/UNOS Transplant
Registry Data................................................................................................................... 91
Table 53: Incidence of NHL (events per 100 patient years) among ........................... 92
Table 54: Results of the ROC Analyses to Identify Potential Factors Associated with
PTLD Risk in Belatacept-Treated Patients .................................................................. 94
Table 55: Belatacept PTLD Rates Among EBV+ Patients Based on CMV Serostatus
........................................................................................................................................... 94




                                                                                                                                            5
                                                 Table of Figures
Figure 1: Belatacept Clinical Development Program.................................................. 10
Figure 2: Belatacept Trough Concentrations in de novo Kidney Transplant Patients
Receiving the LI Regimen (left panel) and the MI Regimen (right panel) ................ 15
Figure 3: Whole Blood Trough Concentrations of CsA in Patients Randomized to
CsA in the Phase 3 trials 008 (red, left-side bars at each day) and 027 (blue, right-
side bars at each day)...................................................................................................... 19
Figure 4: Graft Survival After Biopsy for Recipients With and Without CAN ....... 24
Figure 5: Rates of PTLD Overall and by EBV Status by Treatment Arm ............... 47
Figure 6: NNTH Estimate from Belatacept Trials....................................................... 48
Figure 7: Cumulative incidence of non-Hodgkin’s Lymphoma according to EBV
status in kidney, heart, and liver transplant recipients in the Collaborative
Transplant Study registry data...................................................................................... 53
Figure 8: Mean Calculated GFR (mL/min) in Study 008 ........................................ 60
Figure 9: Mean Calculated GFR (mL/min) IM103027 .............................................. 61
Figure 10: Correlates of PTLD Risk in Belatacept-Treated Patients. ...................... 93




                                                                                                                             6
1. Background

Belatacept is a recombinant soluble fusion protein consisting of the extracellular domain
of human CTLA-4 and a fragment (hinge–CH2–CH3 domains) of a modified Fc domain
of human IgG1. CTLA-4 is a member of the immunoglobulin superfamily with structural
similarities to CD28. While both CTLA-4 and CD28 bind CD80/86 on antigen presenting
cells, CTLA-4 inhibits activation of T cells whereas CD28 augments activation of T cells.
As a soluble fusion protein, belatacept does not transmit inhibitory signals like CTLA-4,
but it interferes with CD28:CD80/CD86 interactions, key costimulatory signals required
for T cell activation. T cells require at least 2 signals for full activation. The first signal is
delivered by the T cell receptor and the second via costimulatory molecules like CD28.
By preventing signaling through CD28, belatacept modulates immune responses and acts
as an immunosuppressant.

Belatacept has been developed as a new therapeutic agent for immunosuppression in
kidney transplant recipients; it was studied as a replacement for cyclosporine in an
immunosuppressive regimen that also included an IL-2 antagonist, MMF and
corticosteroids .

Calcineurin inhibitors (CNIs), starting with the advent of cyclosporine (CsA), have been
largely responsible for the high rates of one-year patient and graft survival presently
observed among kidney transplant patients compared to the historically low survival rates
observed from those eras where transplantation was performed without
immunosuppression or with azathioprine and steroids only. While one-year survival rates
have substantially improved, two leading causes of long term adverse outcomes among
kidney transplant patients are death with a functioning graft and late graft loss due to
cumulative immunologic and non-immunologic injuries to the allograft. Despite their
beneficial effect on short-term survival, CNIs may contribute in part to poor long-term
outcomes due to their renal, cardiovascular, and metabolic toxicities, which can cause
renal impairment, hypertension, hypercholesterolemia, and diabetes mellitus. CNIs have
been associated with a direct nephrotoxic effect whose mechanisms have not been clearly
elucidated. This calcineurin inhibitor nephrotoxicity (CIN) may contribute to the
cumulative allograft injuries that result in late graft loss.

Transplantation has been established as the preferred therapeutic intervention in end stage
renal disease (ESRD). Beyond its impact on quality of life, transplantation clearly imparts
a significant survival benefit over the other renal replacement therapies (peritoneal
dialysis, hemodialysis). 1 While an ever increasing number of Americans (now over
500,000) meet the criteria for ESRD, the number of kidney transplants performed
annually has remained stable at around 17,000 for the past several years.2 Given the

1
  Wolfe RA, Ashby VB, Milford EL, et al. Comparison of mortality in all patients on dialysis, patients on
dialysis awaiting transplantation, and recipients of a first cadaveric transplant. NEJM 1999; 341(23): 1725-
30.
2
  http://www.usrds.org/


                                                                                                           7
shortage of available kidneys for transplantation (and the immunologic and non-
immunologic barriers to performing repeat transplants), increasing graft life is of critical
importance to improving patient outcomes.

Bristol Myers Squibb (BMS) has submitted a BLA requesting the approval of belatacept
for the indications of prophylaxis of organ rejection and preservation of a functioning
allograft in adult patients receiving kidney transplants. While a number of products have
been approved for the prophylaxis of organ rejection in patients who received an allograft
(e.g. kidney, heart, or liver), the indication of preservation of a functioning allograft has
not been previously considered or granted by the Agency. The applicant has suggested
that by avoiding the use of CNIs, patients treated with belatacept may avoid the
consequences of CNI-related adverse effects (including cardiovascular and metabolic
toxicities, as well as direct renal toxicities). It should be noted, however, that the absence
of a renal toxicity attributed to CNIs, while an important safety consideration, can not
constitute evidence of the efficacy of belatacept.

As part of the belatacept development program, the applicant conducted a number of
Phase 1 and Phase 2 studies, as well as two Phase 3 trials in de novo kidney transplant
recipients.
    • Study IM103008 was a randomized prospective clinical trial that examined the
        safety and efficacy of two different dose regimens of belatacept (designated as
        more intensive [MI] and less intensive [LI]) compared to a CsA control in
        recipients of living donor (LD) kidneys and standard criteria deceased donor
        (SCD) kidneys with cold ischemia time less than 24 hours.
    • Study IM103027 was a trial of similar design, included the same three treatment
        groups and was conducted in recipients of extended criteria donor (ECD) kidneys
        and also recipients of cadaveric donor kidneys with cold ischemia time greater
        than 24 hours and cadaveric donor kidneys retrieved from non-heart beating
        donors (also known as donor after cardiac death (DCD) kidneys). 3

The applicant also conducted a Phase 2 trial (IM103100), which used similar (but non-
identical) belatacept dosing regimens as those in the two Phase 3 trials. Patients in all
three of these trials received initial induction with basiliximab and concurrent treatment
with mycophenolate mofetil (MMF) and corticosteroids.

This document will summarize and discuss the efficacy and safety of belatacept for the
prophylaxis of organ rejection in patients undergoing kidney transplantation. The
efficacy discussion will cover the primary endpoints used in the Phase 3 clinical trials,
including those based on patient and graft survival, glomerular filtration rate (GFR), and
acute rejection (AR) as described in the protocol, as well as the FDA-defined analysis of
biopsy proved acute rejection (BPAR) which is the endpoint that FDA has used in the




3
 Metzger RA, Delmonico FL, Feng S, et al. Expanded criteria donors for kidney transplantation. Am J
Transplant 2003;3(Suppl 4):114-125


                                                                                                      8
analysis of all recent clinical trials of other immunosuppressive products. 4 The safety
discussion will include not only CNI-associated renal, cardiovascular, and metabolic
toxicities, but also other toxicities seen during the belatacept development program,
including increased rates of BPAR, post transplant lymphoproliferative disorder (PTLD),
and progressive multifocal leukoencephalopathy (PML).


2. Overview of Clinical Development Program

As mentioned above, the primary efficacy and safety data supporting the belatacept BLA
comes from three similarly designed trials in de novo renal transplant patients: a Phase 2
trial (IM103100) and two Phase 3 trials (IM103008 and IM103027). Additional trials
were conducted in transplant patients including a development program in liver
transplant. See Figure 1 below, as adapted from the applicant’s submission.




4
  In the FDA analysis of acute rejection, patients who experience graft loss, death, or patients who are lost
to follow up are imputed as cases of acute rejection, rather than either being considered “successes” (i.e., no
rejection) or being excluded from the analysis population, therefore this is a more inclusive analysis
compared to one that ignores or excludes with patients with these events.


                                                                                                             9
                   Figure 1: Belatacept Clinical Development Program


                                           All Belatacept
                                               studies



            Non-transplanted                                    Transplanted (IV)
           IV: IM103001 (NHV)
            IV: IM103002 (RA)
           IV: IM103024 (NHV)
          SC: IM103029 (NHV)
          SC: IM103038 (NHV)
                                                                            BMS Sponsored
         IV/SC: IM103046 (NHV)
                                                                              Studies



                                                                 Renal                   Liver
                Investigator Sponsored                                                 IM103045a
                        Studies                                                          N≈250
                    IM103030 (Renal)
                    IM103036 (Renal)
                   IM103058 (Islet cell)
                    Single Patient INDs             De novo                            PK
                                                                                    IM103047


                                                              Maintenance
                                                                 IM103010



                    Core Studies                    Steroid Avoidance
                  IM103008 N=666                         IM103034
                  IM103027 N=543
                  IM103100 N=218

a
 2 out of 146 EBV+ liver recipients on belatacept have developed PTLD to date
IV = intravenous, SC = subcutaneous, NHV = normal healthy volunteer, RA = rheumatoid arthritis
Source: Figure adapted from Figure 3 of BMS Briefing Document

Selected Demographics in the Phase 3 Kidney Transplant Trials

Reflective of the kidney transplant recipient population, approximately two-thirds of the
kidney transplant recipients enrolled in the Phase 3 trials were male (see Table 1). Most
patients in both trials were enrolled outside of the US: around 40% of the patients in
IM103008 and 25% of the patients in IM103027 were from North America. Compared to



                                                                                                   10
their prevalence in the US kidney transplant recipient population, Black patients (whether
enrolled in or outside the US) were under-represented in the trials: in 2009, Black
patients represented 25% of the kidney transplant recipients in the US, as opposed to 10%
of the patients in the belatacept Phase 3 trials. 5

Baseline disease characteristics of transplant donors were consistent with each study’s
inclusion/exclusion criteria. In IM103008, 58% of donors were living (42% related,
16% unrelated) and 42% were deceased. In IM103027, nearly all (99%) of the donors
met the ECD criteria defined in the protocol. Among these, 70% to 72% of donors met
the UNOS ECD criteria. 6

The proportion of donors and recipients with mismatched human leukocyte antigen
(HLA) serologies was balanced between treatment groups.

Subject EBV status was determined locally at the site and was reported on the case report
form. For 11 and 44 patients in Studies IM103008 and IM103027, respectively, whose
EBV status was unknown at baseline, retrospective blood sample analysis was performed
by the central laboratory. Approximately 85-90% of the patients enrolled in the Phase 3
trials were EBV positive (EBV+) at baseline.

In general, the various demographic characteristics of the patients appeared to be evenly
distributed across the treatment groups in both Phase 3 trials.




5
 http://optn.transplant.hrsa.gov/
6
 Metzger RA, Delmonico FL, Feng S, et al. Expanded criteria donors for kidney transplantation. Am J
Transplant 2003;3(Suppl 4):114-125


                                                                                                      11
              Table 1: Selected Demographics in the Phase 3 Kidney Trials

                                           IM103008                                   IM103027

                                  MI           LI           CsA             MI           LI          CsA
                                N=219        N=226         N=221          N=184        N=175        N=184
Mean Age (years)                 43.6         42.6          43.5           56.7         56.1         55.7
Gender, n (%)
            Male              151 (68.9)    146 (64.6)   165 (74.7)      119 (64.7)   129 (73.7)   116 (63.0)
           Female              68 (31.1)     80 (35.4)    56 (25.3)       65 (35.3)    46 (26.3)    68 (37.0)
Race, n (%)
           White              132 (60.3)    133 (58.8)   139 (62.9)      137 (74.5)   134 (76.6)   137 (74.5)
            Black              15 (6.8)      23 (10.2)    17 (7.7)        25 (13.6)    24 (13.7)   22 (12.0)
            Asian              27 (12.3)    29 (12.8)    27 (12.2)         7 (3.8)      3 (1.7)     4 (2.2)
            Other             45 (20.5)     41 (18.1)    38 (17.2)        15 (8.2)     14 (8.0)    21 (11.4)
Geographic region, n (%)
       North America           95 (43.4)     92 (40.7)    94 (42.5)       49 (26.6)    40 (22.9)    45 (24.5)
       Rest of World          124 (56.6)    134 (59.2)   129 (57.4)      135 (73.4)   135 (73.1)   139 (75.5)
Recipient EBV serology
          Positive            194 (88.6)    199 (88.1)   184 (83.3)      169 (91.8)   156 (89.1)   168 (91.3)
          Negative             25 (11.4)     27 (11.9)    37 (16.7)       14 (7.6)     19 (10.9)    16 (8.7)
         Unknown                   0             0            0           1 (0.5)          0           0

3. Dose Selection

Two dosage regimens were studied during the development of belatacept: a less
intensive (LI) regimen and a more intensive (MI) regimen. In the current BLA
application, BMS is seeking approval for the belatacept LI regimen studied in the Phase 3
trials for kidney transplantation.

As shown in Table 2, the belatacept LI regimen in the Phase 3 trials consisted of
belatacept (10 mg/kg IV) administration on Day 1 (the day of transplantation, prior to
implantation); and on Days 5, 14, and 28; then every 4 weeks through 3 months after
transplantation. Starting at Month 4 after transplantation, belatacept was administered at
the maintenance dose of 5 mg/kg every 4 weeks (± 5 days). This regimen is provided in
the table below.

                Table 2: Dosing for Belatacept LI Regimen (Phase 3 trials)
       Dosing for Initial Phase                                                     Dose
             Day of transplantation, prior to implantation (Day 1)                10 mg/kg
             Day 5, Day 14, and Day 28 (1 month after transplantation)            10 mg/kg
             Month 2 and 3 after transplantation                                  10 mg/kg
       Dosing for Maintenance Phase                                                 Dose
             Monthly, starting at Month 4 after transplantation                    5 mg/kg




                                                                                                    12
The LI regimen in the Phase 2 study (IM103100) was different from the regimen in Phase
3 in that there was no dose given on Day 5. The applicant added the Day 5 infusion of
belatacept to the LI regimen for Phase 3 in an attempt to “ensure that target drug
concentrations were achieved in the early period after transplantation and to reduce the
rates of subclinical rejection.” The Phase 2 LI regimen consisted of belatacept (10
mg/kg) administration on Days 1, 15, 29, 57, and 85; patients were then reallocated to a 5
mg/kg maintenance dose every 4 weeks or every 8 weeks. Therefore, patients in the
Phase 2 study LI regimen received less overall belatacept exposure relative to those in the
Phase 3 study LI regimen.

The MI regimen in the two Phase 3 trials consisted of belatacept (10 mg/kg)
administration on Day 1 (the day of transplantation, prior to implantation); and on Days
5, 14, 28, 42, 56, 70, and 84; then every 4 weeks through 6 months after transplantation.
Starting at Month 7 after transplantation, belatacept in the MI regimen was administered
at the maintenance dose of 5 mg/kg every 4 weeks (± 5 days). The MI regimen in the
Phase 2 study (IM103100) was essentially identical except that patients were reallocated
during the maintenance period to a 5 mg/kg dose every 4 weeks or every 8 weeks.
Consequently, approximately half of the patients received less frequent dosing after
Month 6 (every 8 weeks) in the Phase 2 study relative to those in the Phase 3 study (every
4 weeks).

4. Clinical Pharmacology

   4.1. Pharmacokinetics

The pharmacokinetics (PK) of belatacept are linear and plasma exposures (Cmax and
AUC) are increased dose proportionally in healthy subjects following single escalating
intravenous (IV) infusion doses from 1 to 20 mg/kg. The terminal half-life (T½) of
belatacept is approximately 8 to 10 days.

Table 3 summarizes the PK parameters of belatacept in healthy subjects after a single
10 mg/kg IV infusion and in de novo kidney transplant patients after multiple
5 mg/kg and 10 mg/kg IV infusions. Similar to healthy subjects, mean Cmax and AUC
increased dose-proportionally in de novo kidney transplant patients following multiple IV
doses of 5 mg/kg and 10 mg/kg every 4 weeks. Mean estimates of systemic clearance
(CL) and volume of distribution (Vss) of belatacept in kidney transplant patients were
also comparable to those in healthy subjects. Likewise, the mean estimates of belatacept
half-life (T½) were similar between kidney transplant patients (approximately 8 to 10
days) and healthy subjects.




                                                                                         13
 Table 3: Pharmacokinetic parameters (Mean ± SD [Range]) in Healthy Subjects and
                       de novo Kidney Transplant Patients
Pharmacokinetic Parameters        Healthy Subjects    Kidney Transplant        Kidney Transplant
                                   (After 10 mg/kg    Patients during the      Patients during the
                                 Single IV Infusions) Maintenance Phase           Initial Phase
                                                        (After 5 mg/kg          (After 10 mg/kg
                                                         Multiple IV              Multiple IV
                                                          Infusions)               Infusions)

                                         N=15                   N=14                   N=10
Peak concentration (Cmax)               300 ± 77              139 ± 28                247 ± 68
[µg/mL]                                (190-492)              (80-176)               (161-340)
AUCa                                 26398 ± 5175           14090 ± 3860           22252 ± 7868
[μg•h/mL]                           (18964-40684)           (7906-20510)          (13575-42144)
Terminal half-life (T½)                 9.8 ± 2.8             8.2 ± 2.4b              9.8 ± 3.2
[days]                                 (6.4-15.6)             (3.1-11.9)             (6.1-15.1)
Systemic clearance (CL)               0.39 ± 0.07               0.51b               0.49 ± 0.13
[mL/h/kg]                             (0.25-0.53)            (0.33-0.75)            (0.23-0.70)
Volume of distribution (Vss)          0.09 ± 0.02               0.12b               0.11 ± 0.03
[L/kg]                                (0.07-0.15)            (0.09-0.17)            (0.07-0.17)
 a
     AUC=AUC (INF) after single dose and AUC (TAU) after multiple dose, where TAU=4 weeks
 b
     TAU=8 weeks

 No formal PK studies were conducted in subjects with hepatic impairment, renal
 impairment, or in geriatric and pediatric populations. Population PK analysis showed that
 there was a trend toward higher CL of belatacept with increasing body weight, supporting
 a weight-based dose of belatacept. Age, gender, race, renal function (measured by
 calculated glomerular filtration rate [GFR]), hepatic function (measured by albumin),
 diabetes, and concomitant dialysis did not affect clearance of belatacept.

       4.2. Belatacept Trough Concentrations Observed in Phase 3 trials

 The Phase 3 trials IM103008 and IM103027 were designed to administer fixed doses of
 belatacept, with no target trough concentration range specified to be attained by
 therapeutic drug monitoring with dose adjustment. Table 4 summarizes the belatacept
 threshold trough concentration for Phase 3 trials IM103008 and 103027 that were shown
 to be effective in Phase 2 trial IM103100.




                                                                                                  14
                                             Table 4: Belatacept Threshold Trough Concentrations for Trials 008 and 027
                                             Threshold Trough            Less Intensive (LI)       More Intensive (MI)
                                              Concentrationa                  Regimen                    Regimen
                                                 20 μg/mL                     Month 1                   Months 1-3
                                                  5 μg/mL                    Months 2-4                 Months 4-6
                                                  2 μg/mL                   Months 5-12                Months 7-12
a
                     Based on Phase 2 trial IM103100

In the Phase 3 trials IM103008 and IM103027, belatacept trough concentrations were
higher than the threshold trough concentrations at each of the measured time points in
approximately 80% of the patients receiving the LI and MI regimens (see Figure 2). Data
from the Phase 3 trials were combined. The bottom and top of each box represent the
inter quartile range (i.e., 25th and 75th percentiles, respectively). The bar inside each box
represents the median trough concentration. The whiskers represent 5th and 95th
percentiles, respectively. The horizontal reference lines represent the threshold trough
concentration as outlined in Table 3. There was substantial between-patient variability in
trough concentrations in the belatacept LI and MI regimens; the percent coefficient of
variation (%CV) at Month 1 was lowest at < 50%, while %CV was > 60% after Month 1.

       Figure 2: Belatacept Trough Concentrations in de novo Kidney Transplant Patients
            Receiving the LI Regimen (left panel) and the MI Regimen (right panel)
                                        60




                                                                                                        60
Belatacept trough concentration [ug/mL]
                                 50




                                                                                                        50




                                                                                LI REGMEN                                               MI REGMEN
                        40




                                                                                                        40
                  30




                                                                                                        30
           20




                                                                                                        20
    10




                                                                                                        10
                        0




                                                                                                        0




                                                5   56   84   112   168   252   280   308   336   364        5   56   70   84   112 168 252 280 308 336 364


                                                                                        Time after transplantation [day]


                                             4.3. Exposure-Response Relationships


                                                4.3.1. Exposure-Response Relationships for Efficacy

The relationship between exposure (belatacept trough concentrations) and response (AR
and renal function) was assessed and the results are summarized below.
Acute Rejection (AR): Higher belatacept trough concentrations on Day 5 appeared to be
related to a lower incidence of AR during Month 1 post transplant. However, the


                                                                                                                                                              15
incidence of AR during Month 1 post transplant only accounts for ~1/3 of the total AR
episodes. Overall, no apparent relationship of belatacept trough concentrations with AR
was observed in the Phase 3 trials IM103008 and IM103027.

Renal Function: In the Phase 3 trials IM103008 and IM103027, renal impairment was
                                                        2
defined as a measured GFR (mGFR) < 60 mL/min/1.73 m or a decrease in measured
                          2
GFR ≥ 10 mL/min/1.73 m from Month 3 to Month 12 (mGFR10). The exposure-
response analysis was conducted with mGFR10 because it is not influenced by the
observed difference in baseline mGFR values in each patient. Overall, the proportion of
patients with mGFR10 decreased with increasing belatacept trough concentrations in
Studies IM103008 and IM103027 (see Table 5). Data from the LI and MI regimens in
IM103008 and IM103027 were combined for the quartile analysis.

 Table 5: Relationship between Belatacept Trough Concentrations (Median; 10-90th
   percentile) and Proportion of Patients with a Decrease in Measured GFR ≥ 10
                       mL/min/1.73 m2 from Month 3 to Month 12
                              Quartile 1       Quartile 2    Quartile 3    Quartile 4
                                                   a
                                          Month 1
Median (10-90th percentile)       25.5             32.9         38.7           49.9
belatacept Ctrough (μg/mL)   (18.9 – 28.8) (30.5 – 35.5) (36.5 – 41.8) (43.4 – 62.8)
Percentage of patients with       22.4             24.0         25.1           18.0
mGFR10 (%; n/N)                 (40/179)         (43/179)     (45/179)       (32/178)
                                       Months 2 to 4
Median (10-90th percentile)        6.1             12.1         22.3           34.4
belatacept Ctrough (μg/mL)     (3.6 – 8.1)     (9.7 – 16.0) (17.8 – 26.1) (28.5 – 49.5)
Percentage of patients with       31.3             23.1         21.4           19.4
mGFR10 (%; n/N)                 (57/182)         (42/182)     (39/182)       (35/180)
                                         ≥ Months 6
                th
Median (10-90 percentile)          2.4               4.0         5.7            8.5
belatacept Ctrough (μg/mL)     (1.1 – 3.1)      (3.4 – 4.7)  (5.0 – 6.7)   (7.3 – 17.0)
Percentage of patients with       28.4             28.4         23.5           20.5
mGFR10 (%; n/N)                 (46/162)         (46/162)     (38/162)       (33/161)
                                         All Months
Median (10-90th percentile)        7.5             11.3         15.6            27
belatacept Ctrough (μg/mL)     (5.1 – 9.3)     (9.9 – 12.8) (13.9 – 18.7) (20.9 – 42.7)
Percentage of patients with       31.4             21.7         22.7           14.1
mGFR10 (%; n/N)                 (61/194)         (42/194)     (44/194)       (27/192)
a
 taken on day 5 post-transplant
Ctrough = trough concentration
mGFR10 = decrease in measured GFR ≥ 10 mL/min/1.73 m2 from Month 3 to Month 12




                                                                                      16
       4.3.2. Exposure-Response Relationships for Safety

The relationship between exposure (belatacept trough concentrations) and response
(PTLD, infections, and other adverse events of interest) was assessed and the results are
summarized below.

Post transplant lymphoproliferative disorder (PTLD): Based on visual inspection of
graphical analyses comparing belatacept trough concentrations in patients with PTLD,
belatacept trough concentrations in patients with PTLD were not substantially different
from that in patients without PTLD.

Infections: Serious infections occurred more frequently in the first 6 months following
transplantation. From Month 2 to Month 6, belatacept exposure in subjects on the MI
regimen is generally higher than those on the LI regimen. However, the incidence rate of
serious infections does not appear to be substantially different between the MI regimen
and the LI regimen in that period. The incidence of some infections, such as BK virus and
herpes virus infections, tend to be higher in patients with higher belatacept trough
concentrations. However, overall incidence of these infections is too low to draw definite
conclusions.

Other Adverse Events: There was no apparent association between the incidence of new
onset diabetes mellitus after transplant (NODAT), hypertension, dyslipidemia, or
congestive heart failure and belatacept trough concentrations.

   4.4. Mycophenolate Mofetil (MMF) Dosing and Mycophenolic Acid (MPA) Exposure
        Observed in Phase 3 Trials

In the Phase 3 trials IM103008 and IM103027, the initial MMF dose was 2 g/day. However,
the MMF dose was allowed to be adjusted at the physician’s discretion based on clinical
signs of adverse events or efficacy failure. There was no substantial difference in percent of
patients who received MMF 2 g/day among the treatment groups in Studies IM103008 and
IM103027 (see Table 6). Approximately 60-75% of patients received 2g/day of MMF across
all treatment groups. The remainder of patients received less than 2 g/day of MMF.

Table 6: Percent of Patients who Received MMF 2 g/day at Given Times in Phase 3
                                Trials 008 and 027
                          Study 103008                    Study 103027
                Month 3     Month 6 Month 12 Month 3        Month 6 Month 12
Belatacept MI     74%         75%         68%      67%        66%        56%
Belatacept LI     71%         71%         67%      71%        66%        58%
CsA               70%         65%         64%      68%        60%        57%

Cyclosporine (CsA) inhibits enterohepatic recirculation of MPA, the active form of MMF,
and, consequently, lowers MPA exposure. Thus, systemic exposure to MPA would be higher
in the belatacept arms and exposure to the main metabolite, MPA glucuronide (MPAG),
would be lower, as compared to the CsA arm. In a subset of 21 kidney transplant patients
enrolled in the two Phase 3 trials, the mean dose-normalized MPA Cmax and AUC0-12 were


                                                                                           17
higher by 20% and 40%, respectively, when fixed dose MMF was co-administered with
belatacept than when co-administered with CsA. The mean dose-normalized MPAG Cmax and
AUC0-12 were lower by 25% and 30%, respectively, in those patients receiving belatacept and
MMF as compared to those receiving CsA and MMF.

   4.5. Trough Cyclosporine Concentrations of Observed in Phase 3 Trials

Patients randomized to the CsA arm in the Phase 3 trials IM103008 and IM103027
received CsA twice daily to achieve protocol-specified target trough CsA concentration
of 150 to 300 ng/mL during the first month post transplant and then 100 to 250 ng/mL
thereafter. The CsA troughs observed in the Phase 3 trials, however, tended to be higher
than the protocol-specified CsA targets during the first 2 to 3 months post transplant (see
Figure 3). For the first two months, trough CsA concentrations were attained within the
protocol-specified target range in approximately 50 to 60% of patients. In approximately
40% of patients, trough CsA concentrations attained for the first two months were higher
than the protocol-specified target range. After Month 3 post transplant, trough CsA
concentrations were attained within the protocol-specified target range in approximately
70 to 80% of patients. After the two week time point, the vast majority of trough CsA
concentrations outside of the protocol-specified target range were above the upper limit
of the target range.




                                                                                         18
    Figure 3: Whole Blood Trough Concentrations of CsA in Patients Randomized to
    CsA in the Phase 3 trials 008 (red, left-side bars at each day) and 027 (blue, right-
                                  side bars at each day).


                                             600
                                                                                         Red: IM103008
                                                                                         Blue: IM103027
          Whole Blood Cyclosporine (ng/mL)




                                             500



                                             400



                                             300



                                             200



                                             100



                                              0
                                                   0   25   50           75        100        200     300

                                                                 Time (Days Posttransplant)


5. Overview of Efficacy Endpoints

The Phase 3 trials (IM103008 and IM103027) were designed as three year studies, with
the primary analysis of efficacy at Month 12. The trials are still ongoing and all patients
have completed 2 years of follow-up.

The co-primary efficacy endpoints as pre-specified in the protocol were:

•     The composite of patient and graft survival at 12 months, and
•     The composite of renal impairment as assessed by measured glomerular filtration rate
      (GFR) < 60 mL/min/1.73m2 at Month 12 or a decrease in measured GFR ≥ 10
      mL/min/1.73m2 from Month 3 to Month 12, as measured by the cold-iothalamate
      method, and
•     The incidence of acute rejection (AR), defined as a biopsy confirmed histological
      evidence of rejection along with clinical evidence, at 12 months (IM103008 only)

The incidence of AR was a secondary endpoint in IM103027. Mean GFR at 12 months
and incidence of chronic allograft nephropathy (CAN) were also secondary endpoints in
both trials.

The protocol specified that the endpoints of patient and graft survival and of AR were to
be assessed for noninferiority (NI), whereas the endpoint of composite renal impairment



                                                                                                            19
was to be assessed for superiority. The endpoints were to be examined sequentially in the
order of hierarchy stated above. The Type I error rate was set at 2.7% using Dunnett’s
adjustment for each belatacept regimen versus CsA. Thus treatment differences between
each belatacept treatment groups and CsA were tested at the 0.027 significance level and
97.3% confidence intervals about differences between the treatment groups are reported.

The strengths and limitation of the endpoints of patient and graft survival, composite
renal impairment, mean GFR, AR, and CAN will each be discussed. Of particular
importance are the underlying limitations in the comparisons of GFR between the
belatacept and CsA groups due to the differential effects of the treatment regimens on
renal hemodynamics . However, because of the identified limitations with the evaluation
and interpretation of these endpoints, including the limitation in the ability to justify the
proposed NI margin for the patient and graft survival endpoint; the FDA also conducted
an additional analysis of acute rejection in the study populations. This analysis of acute
rejection by FDA has been used by FDA in the analysis of other kidney transplantation
trials, and the proposed NI margin for this analysis can be supported from previous
kidney transplantation studies (see Appendix A) and is described below. The results of
these endpoints are presented in Section 6.

      5.1. Patient and Graft Survival at 12 Months

The impact of CsA on patient outcomes, when compared to the conventional
immunosuppression of the era (azathioprine and steroids), was readily measureable by
evaluating patient and graft survival. The graft loss and death rate in the trials of
Sandimmune® were in the range of 10% to 20% in the Sandimmune arms compared to
approximately 35% to 45% in the control arms of azathioprine and corticosteroids (p-
value < 0.05). 7 In the modern transplant era, one can no longer depend on patient and
graft survival to demonstrate the efficacy of newer regimens by comparison to a regimen
reflecting standard of care, since the high rate of short-term (1 year) patient and graft
survival obtained using these regimens does not allow one to power a trial to demonstrate
the superiority of the newer regimen for an endpoint based on patient and graft survival at
12 months. In theory, a NI approach could be used to demonstrate that a newer regimen
had some efficacy for the indication of preventing patient death and graft loss, but the
size of the effect of CsA on patient and graft survival given the current standard regimen
including induction, corticosteroids, and MMF cannot be easily estimated and therefore
we cannot define a NI margin. However, given that the endpoint of patient and graft
survival is an important endpoint, we make sure there this is not an unacceptable loss of
efficacy.

For these reasons, the endpoint of patient and graft survival in the belatacept trials is not
viewed as a primary endpoint for the demonstration of efficacy; however, it is an
important endpoint for assessing safety, i.e., whether or not the trials merit further review.




7
    Sandimmune Summary Basis for Approval


                                                                                           20
    5.2. Composite Renal Impairment and mean Glomerular Filtration Rate (GFR)


Cyclosporine causes vasoconstriction of the afferent renal artery leading to a physiologic
decrease in GFR. 8 Such a physiologic effect on GFR could conceivably have deleterious
consequences, especially among renal transplant patients. However, the reliance on GFR
in the context of the Phase 3 trials as a measure of prophylaxis of organ rejection or
preservation of a functioning allograft in adult patients receiving renal transplants is
confounded by the hemodynamic effects of CsA, which was used only in the control
group. The difference in the hemodynamic effect between the belatacept groups and the
CsA groups manifested immediately in the trial: the data show that the belatacept groups
and the CsA groups had significantly different GFRs from the first measurable time
points (see Figure 8 and Figure 9 in Section 7.7). While the GFR for the belatacept
groups remained consistently higher than those in the CSA groups through 24 months,
the differences may simply reflect the hemodynamic differences rather than any
structural differences in the allografts between treatment groups. As GFR is an important
measure of renal function, the higher GFRs exhibited by patients maintained on
belatacept regimens may translate into important clinical benefits. However, one cannot
rely on an improvement in observed GFR to prove the efficacy of belatacept for
preserving GFR if the comparison is to CsA, which is known to decrease GFR through a
direct physiologic mechanism.

For purposes of assessing efficacy based on the renal composite endpoint, patients were
classified as successes or failures based on whether GFR was less than 60 mL/min at
Month 12 or whether GFR had declined more than 10 mL/min between Month 3 and
Month 12 (defined as failure). In the belatacept Phase 3 trials, most patients who met the
endpoint were classified as efficacy failures due to a GFR less than 60 mL/min, not due
to a decline more than 10 mL/min between Month 3 and Month 12. The selection of a
Month 12 GFR of 60 mL/min was based on a retrospective study of registry data which
suggested better outcomes among renal transplant patient whose serum creatinine at 1
year was less than 1.5 mg/dL.21 That literature analyzed populations of patients who were
almost universally maintained on CNI containing regimens and, therefore, had
comparable renal hemodynamics allowing meaningful comparisons of GFR as a
surrogate of kidney allograft function and structure. It is less clear; however, what a
comparison of GFRs between the belatacept groups and the CsA groups in the belatacept
Phase 3 trials implies about expected outcomes for the two groups. The imbalance, then,
in starting GFRs (post-transplantation once therapy was started) between the
investigational arms and the control arms confounds the analysis and interpretation of
these findings.



8
  Bobadilla NA, Gamba G. New insights into the pathophysiology of cyclosporine nephrotoxicity: a role of
aldosterone. Am J Physiol Renal Physiol. 2007; 293(1): F2-9.




                                                                                                     21
Study IM103008 exclusively enrolled recipients of LD and SCD kidneys and some of
these maintained GFRs greater than 60 mL/min, Study IM103027 enrolled patients with
ECD kidneys, and almost all patients in this trial (both in the belatacept and the CsA
groups) were treated as efficacy failures due to a Month 12 GFR less than 60 mL/min. In
retrospect, these results are not surprising, because the recipients of extended criteria
donor kidneys rarely achieve a GFR greater than 60 mL/min. In fact the majority of
patients in all arms in IM103027 started out with GFRs less than 60 mL/min. As a result,
the composite renal endpoint did not detect a significant difference between the
belatacept and the cyclosporine treatment groups despite a considerable difference in
mean GFRs observed between these same groups.

Given the limitations of the composite renal endpoint in the setting of incomparable renal
hemodynamics across treatment groups and given the limitations of the composite renal
endpoint in the setting of a trial of ECD kidneys, it is difficult to understand the
significance or relevance of these analyses as an efficacy endpoint. However, because
GFR remains an important measure of renal function, differences in mean GFR at Month
12 were given consideration both as a secondary efficacy endpoint and a safety endpoint.

   5.3. Acute Rejection

Study IM103008 specified the incidence of AR (defined as central biopsy proven
rejection that was either “clinically suspected by protocol defined reasons or clinically
suspected by other reasons and treated”) as one of its three co-primary endpoints while
Study IM103027 specified the incidence of AR as a secondary endpoint. However, given
the limitations of the patient and graft loss endpoint and the combined renal impairment
endpoint as described in the above paragraphs, it may be reasonable to examine the
outcome of AR in both trials as a primary efficacy endpoint. Acute rejection as an
endpoint has traditionally been used by the FDA to assess efficacy of other products for
the indication of prophylaxis of rejection in kidney transplant recipients.

In the modern transplant era, rates of acute rejection are low. Most transplant trials rely
on a non-inferiority approach to evaluate the efficacy of a new product for maintenance
immunosuppression. The belatacept trials also relied on a non-inferiority margin to
evaluate the efficacy of belatacept for the prophylaxis of acute rejection. In order to be
able to interpret a non-inferiority trial, a scientific justification of the non-inferiority
margin is needed. On the basis of data from a trial of an IL-2 receptor antagonist plus
MMF plus corticosteroids and an estimate that CsA-based regimens are expected to result
in a 15% rate of AR, the Division concluded that the data support a 20% non-inferiority
margin as the effect that CsA adds to this regimen on AR. Use of this margin then
allows for the conclusion of whether or not belatacept adds some contribution to the
regimen with respect to prevention of AR (see Appendix A for details).11,12 However,
while a 20% margin is sufficient to demonstrates an effect over placebo, it does not
demonstrate that the potential loss of efficacy compared to the standard of care is
clinically acceptable. Improved graft survival was observed in the period from 1988 to
1996 corresponding with the introduction of modern immunosuppression agents; an
analysis of registry data during that period showed that AR events during the first year of



                                                                                         22
transplantation had a detrimental effect on long-term graft survival. 9 Despite additional
incremental decreases in AR rates in the period from 1995 to 2000, however, no further
improvement in graft survival resulted. 10 The data suggest, then, that while effective
prophylaxis of AR is critical to patient and graft survival; the correlation between AR
rates and improved outcomes becomes less clear as AR rates approach those observed in
the modern era.

Therefore, the endpoint of AR as defined by the applicant and as analyzed by FDA in the
belatacept trials was viewed by FDA as the primary endpoint for the demonstration of
efficacy. Whether or not the potential loss of efficacy compared to the standard of care is
acceptable is a point for further discussion.

    5.4. Acute Rejection – FDA analysis

In the FDA analyses of AR, patients who experience death, graft loss, or loss to follow up
are not considered as having a positive outcome (i.e., no rejection). Instead, these
patients’ outcomes are imputed as efficacy failures from the point-of-view of the AR
endpoint. Patients with death, graft loss, or loss to follow up do not have complete data
at 12 months. Additionally, considering them as having a positive outcome for this
analysis does not seem appropriate given that the subjects who have obtained an endpoint
of death or graft loss have obtained an endpoint of even more clinical importance than
AR. 11 The FDA analysis of the AR endpoint has therefore often been described as the
“combined endpoint” of biopsy-proven acute rejection (BPAR), graft loss, death, and loss
to follow up. It is important that this endpoint be taken into consideration when assessing
the efficacy results of the Phase 3 studies in this application. The FDA also considers that
the definition of BPAR includes all events where the central review of the biopsy
indicated acute rejection, regardless of the reason for performing the biopsy. That
definition of BPAR differs slightly from that used by the applicant, who defined BPAR to
include only those events where central review of the biopsy indicated acute rejection and
AR was either clinically suspected by protocol defined reasons or clinically suspected by
other reasons and treated. Sensitivity analyses, however, confirmed that the conclusions
of the FDA analysis of BPAR remained unchanged regardless of which definition of
BPAR was used.

    5.5. Chronic Allograft Nephropathy

The applicant also assessed the presence of chronic allograft nephropathy (CAN) at
Month 12 as a secondary endpoint, using protocol biopsies in an attempt to assess
differences in structural changes across the treatment groups. CAN was developed as a
term to describe histopathologic findings on biopsy associated with progressive renal
dysfunction not attributable to a more specific diagnosis. In practice, its use has come to
encompass a heterogeneous group of findings that may develop from a variety of
9
  Hariharan S, Johnson CP, Bresnahan BA, et al. Improved graft survival after renal transplantation in the
United States, 1988-96. N Engl J Med 2000; 342: 605-12.
10
   Meier-Kriesche H-U, Schold JD, Srinivas TR, et al. Lack of improvement in renal allograft survival
despite a marked decrease in acute rejection rates over the most recent era. Am J Transplant 2004; 378-83.
11
   Lubsen J, Kirwan BA. Combined endpoints: can we use them? Statistics in Medicine 2002; 21:2959-70


                                                                                                        23
underlying processes. 12 Although the observation of CAN reflects structural findings, the
understanding of its clinical relevance remains rather uncertain and continues to evolve,
and recent publications suggest its prognostic value for long term graft survival is
unclear. For example, the Long Term Deterioration of Kidney Allograft Function
(DeKAF) multicenter consortium published data suggesting that CAN has no prognostic
value for the outcome of graft loss.13 Based on its study of 440 biopsies, the group
reported:
       “the most important finding of our study, to date, is the observation that, in the
       cross-sectional cohort, the local diagnosis of CAN was of no prognostic
       significance… Those with and without CAN had identical patterns of postbiopsy
       graft failure. There was also no functional difference (for those with and without
       CAN) in the postbiopsy slope of 1/creatinine versus time. Clearly, for recipients
       with new onset late graft dysfunction, the nonspecific diagnosis of CAN does not
       define a group of patients with a different outcome.” [See Figure 4] 14

     Figure 4: Graft Survival After Biopsy for Recipients With and Without CAN



     Figure taken from Gourishankar et al, Amer J Transplantation 2010; 10: 324-330




12
   Colvin, RB. Chronic Allograft Nephropathy. N Engl J Med 2003; 349: 2288-90.
13
   Gourishankar S, Leduc R, Connett J, et al. Pathological and Clinical Characterization of the ‘Troubled
Transplant’: Data from the DeKAF Study. Amer J of Transplantation 2010; 10: 324-330.
14
   Gourishankar S, Leduc R, Connett J, et al. Pathological and Clinical Characterization of the ‘Troubled
Transplant’: Data from the DeKAF Study. Amer J of Transplantation 2010; 10: 324-330.


                                                                                                            24
   6. Clinical Efficacy

       6.1. Patient Disposition

   Clinical trials in transplantation, even randomized controlled trials intended to support the
   approval of a new therapy, are rarely conducted in a double blind fashion, primarily due
   to logistical issues and patient safety concerns regarding therapeutic drug monitoring.
   The belatacept Phase 3 trials were also conducted overall as open-label trials although the
   two belatacept regimens were blinded. Discontinuations were roughly evenly distributed
   across the treatment groups in both IM103008 and IM103027 at 12 months (see Table 7
   and Table 8), but there were differences in the reasons for discontinuations between
   belatacept and CsA groups: more belatacept patients discontinued treatment due to lack
   of efficacy whereas more CsA patients discontinued treatment due to adverse events.

       Table 7: Reasons for Treatment Discontinuations at 12 Months in IM103008

                                                                 IM103008

                                            Belatacept MI       Belatacept LI           CsA
 Number randomized and transplanted            N=219               N=226               N=221
Number randomized, transplanted, and
treated                                         219                  226                 215

Number Discontinued Treatment, N (%)         46 (21.0)            45 (19.9)           42 (19.5)
Adverse Event                                  9 (4.1)             12 (5.3)            20 (9.3)
Lack of Efficacy                              26 (11.9)            24 (10.6)           10 (4.7)
Death                                          4 (1.8)              2 (0.9)             3 (1.4)
Lost to Follow Up                                 0                    0               1 (0.5)
Other                                          2 (0.9)              4 (1.8)             5 (2.3)
Non-Compliance                                    0                    0                2 (0.9)
Withdrew Consent                               5 (2.3)              3 (1.3)             1 (0.5)
Number Continuing Treatment                  173 (79.0)           181 (80.1)          173 (80.5)

       Table 8: Reasons for Treatment Discontinuations at 12 Months in IM103027

                                                                 IM103027

                                            Belatacept MI       Belatacept LI           CsA
 Number randomized and transplanted            N=184               N=175               N=184
Number randomized, transplanted, and
treated                                         183                  174                 179

Number Discontinued Treatment, N (%)         50 (27.3)            45 (25.9)           54 (30.2)
Adverse Event                                 22 (12.0)           27 (15.5)           31 (17.3)
Lack of Efficacy                              16 (8.7)             15 (8.6)            14 (7.8)
Death                                          5 (2.7)              2 (1.1)             2 (1.1)
Other                                          6 (3.3)              1 (0.6)             5 (2.8)
Withdrew Consent                               1 (0.5)                 0                2 (1.1)
Number Continuing Treatment                  133 (72.7)           129 (74.1)          125 (69.8)




                                                                                               25
      6.2. Primary Efficacy Results

         6.2.1. Patient and Graft Survival

Patient and graft survival at 12 months is presented in Table 9. In Study IM103008,
patient and graft survival at 12 months was 95%, 96.5%, and 93.2% in the belatacept MI,
belatacept LI and CsA groups, respectively. The patient and graft survival rates were
lower in Study IM103027 than those seen in IM103008; IM103027 evaluated recipients
of ECD kidneys and rates are 85.9%, 88.6%, and 84.8% in the belatacept MI, belatacept
LI and CsA groups, respectively. The lower bound of the 97.3% confidence interval
about the difference between belatacept regimen and CsA was greater than -5% for all
comparisons with the exception of the belatacept MI versus CsA comparison in Study
IM103027 which was -7.6%. It should be noted that these results are slightly different
than those presented by the Applicant since all patients with unknown survival status at
Month 12 are treated as failures.

                      Table 9: Patient and Graft Survival at 12 months
Study                                         Belatacept MI     Belatacept LI           CsA
008      Surviving with a functioning graft   208/219 (95.0)    218/226 (96.5)     206/221 (93.2)
            Graft Loss                          4 (0 died)        5 (1 died)         8 (1 died)
            Death w/ functioning graft               6                 3                 6
            Unknown status                           1                 0                 1
         Difference from CsA (97.3% CI)       1.8 (-3.6, 7.2)   3.3 (-1.8, 8.4)

027      Surviving with a functioning graft   158/184 (85.9)    155/175 (88.6)     156/184 (84.8)
            Graft Loss                          17 (2 died)       16 (1 died)        20 (3 died)
            Death w/ functioning graft               6                 4                  5
            Unknown status                           3                 0                  3
         Difference from CsA (97.3% CI)       1.1 (-7.6, 9.8)   3.8 (-4.7, 12.3)


         6.2.2. Composite Renal Endpoint

The proportions of subjects meeting the composite renal impairment endpoint as well as
the reasons for meeting the composite endpoint are presented in Table 10. The
difference in the proportion of subjects meeting the composite endpoint was statistically
significantly better for a belatacept versus CsA comparisons with the exception of the
belatacept LI versus CsA comparison in Study IM103027. It should be noted that the
analyses presented here are slightly different from that presented by the Applicant in that
subjects with missing measured GFR data for reasons other than death or graft loss are
also included as failures of the endpoint. The conclusions drawn, however, are not
different.




                                                                                                26
                       Table 10: Composite Measured GFR Endpoint
Study                                           Belatacept MI   Belatacept LI        CsA
008                                                (n=219)         (n=226)         (n=221)
          Composite endpoint                      125 (57.1)     128 (56.6)       174 (78.7)
          Reason for meeting composite:
          -M12 < 60 and Decrease ≥ 10 from M3        33               34             52
          to 12
          -M12 < 60 only                             58              58              92
          -Decrease ≥ 10 from M3 to M12 only         15              16               8
          -Imputed due to GL or death                 9               8              14
          -Missing                                   10              12               8
             p-value                               <0.0001         <0.0001
027                                                (n=184)         (n=175)         (n=184)
          Composite endpoint                      132 (71.7)      135 (77.1)      157 (85.3)
          Reason for meeting composite:
          -M12 < 60 and Decrease ≥ 10 from M3        27               41             37
          to 12
          -M12 < 60 only                              71              64             83
          -Decrease ≥ 10 from M3 to M12 only           4               6              7
          -Imputed due to GL or death                 22              19             24
          -Missing                                     8               5              6
             p-value                                .0022           .0575

          6.2.3. Acute Rejection

Acute rejection was defined in the Phase 3 protocols as a clinical/pathological event
requiring clinical evidence; and biopsy confirmation of histological evidence of rejection,
as determined by a blinded central pathologist. Clinical evidence of AR was defined by
one or more of the following conditions:

      •   An unexplained rise of serum creatinine (SCr) ≥ 25% from baseline;
      •   Occurrence of one or more of the following: an unexplained decreased urine
          output; fever and graft tenderness; a SCr that remained elevated within 14 days
          after transplantation and clinical suspicion of AR;
      •   A reason other than those listed above and the subject was treated for this episode.

The three most common reasons for clinical suspicion of AR in the Phase 3 studies
included an unexplained rise of SCr ≥ 25% from baseline, a SCr that remained elevated
within 14 days post-transplantation, and other reasons (related to renal dysfunction and
elevations of SCr that did not meet the protocol criteria).

Table 11 summarizes the incidence of AR, as defined by the applicant, at 12 months.
Noninferiority of all belatacept regimens to CsA based on a NI margin of 20% was
shown, as demonstrated by an upper bound of the 95% confidence interval (CI) of less
than 20%, with the exception of the belatacept MI regimen in Study IM103008.

However, it should also be noted that there were numerically more ARs in the belatacept-
treated subjects than in the CsA-treated subjects. In Study IM103008 for both the



                                                                                               27
belatacept regimens the lower bound of the 95% CI was above zero, indicating a
significant increase in the number of ARs compared to CsA. In addition, the severity of
rejection was greater in the belatacept groups, which will be discussed further in the
safety discussion (Section 7.6).

        Table 11: Acute Rejection (as Defined by the Applicant) at 12 months
Study                                    Belatacept MI      Belatacept LI          CsA
008     Acute Rejection                   49/219 (22.4)      39/226 (17.3)     16/221 (7.2)
         Mild IA                                7                  4                3
         Mild IB                                3                  8                5
         Moderate IIA                          16                 16                6
         Moderate IIB                          20                 10                2
         Severe III                             2                  1                0
        Difference from CsA (97.3% CI)   15.2 (7.4, 23.0)   10.1 (2.9, 17.3)

027     Acute Rejection                   33/184 (17.9)      31/175 (17.7)     26/184 (14.1)
         Mild IA                                 0                  4                2
         Mild IB                                 7                  2                2
         Moderate IIA                           11                 17               17
         Moderate IIB                           15                  8                5
         Severe III                              0                  0                0
        Difference from CsA (97.3% CI)   3.8 (-5.2, 12.8)   3.6 (-5.5, 12.7)

As discussed previously in the discussion on efficacy endpoints (Section 5), the FDA felt
there were limitations with the applicant’s evaluation of AR, specifically that events of
death, graft loss, or loss to follow up were imputed as successes. Therefore, the FDA
also conducted another analysis of AR through the traditional combined endpoint of
BPAR, graft loss, death or lost to follow-up.

In the FDA’s assessment, BPAR was also based upon central biopsy confirmed
histological evidence, but also included additional rejections not necessarily accompanied
by clinical signs and symptoms. Table 12 and Table 13 below describes the additional
patients included in the FDA definition of BPAR for Studies IM103008 and IM103027,
respectively.




                                                                                              28
Table 12: Additional Patients Included in FDA’s Definition of BPAR in Study 008
                  Patient   Biopsy Diagnosis       Date of   Final Clinical Diagnosis     Treatment
                                                   Biopsy
                                                   (Month)
Belatacept MI     70 M      Mild acute (IB)        12        None                         None
                  24 M      Mild acute (IB)        12        AR (non-steroid resistant)   Steroids
                  47 M      Mild acute (IB)        12        AR (non-steroid resistant)   Steroids
Belatacept LI     52 M      Mild acute (IB)        6         none                         None
                  50 M      Mild acute (IA)        6         AR (non-steroid resistant)   None*
                  55 M      Moderate acute (IIB)   6         none                         None
                  18 M      Mild acute (IB)        12        none                         None
                  45 M      Moderate acute (IIB)   12        Active Chronic Rejection     Steroids and
                                                                                          Lymphocyte
                                                                                          depleting
                                                                                          agent*
                  67 F      Moderate acute (IIB)   6         AR (non-steroid resistant)   Steroids
CsA               35 M      Mild acute (IB)        6         Focal involvement of graft   None
                                                             by PTLD
                  39 M      Mild acute (IB)        12        Histological rejection w/o   None
                                                             clinical findings
                  18 M      Moderate acute (IIB)   12        None                         None
                  56 F      Mild acute (IA)        12        AR (non-steroid resistant)   Other (?)
                  55 M      Moderate acute (IIB)   6         None                         None
                  29 M      Mild acute (IB)        12        None                         None
                  26 M      Mild acute (IA)        12        AR (steroid resistant)       Steroids and
                                                                                          OKT3*
* study drug discontinued




                                                                                                         29
Table 13: Additional Patients Included in FDA’s Definition of BPAR in Study 027
                  Patient   Biopsy Diagnosis       Date of   Final Clinical Diagnosis        Treatment
                                                   Biopsy
                                                   (Month)
Belatacept MI     39 F      Moderate acute (IIA)   6         Tubulopathy                     None
                  59 M      Moderate acute (IIA)   12        None                            None
                  69 M      Moderate acute (IIB)   6         AR (steroid resistant)          None*
                  49 M      Moderate acute (IIA)   12        None                            None
                  36 M      Moderate acute (IIA)   6         Acute tubular necrosis          None
                            Moderate acute (IIB)   6         AR (non-steroid resistant)      Steroids
                  56 M      Mild acute (IB)        12        None                            None
                  68 M      Moderate acute (IIA)   6         Hydrop tubli nephrosclerosis    None
Belatacept LI     44 M      Moderate acute (IIA)   12        AR (non-steroid resistant)      Steroids
                  60 F      Moderate acute (IIA)   6         Negative for cellular and       None
                                                             humoral rejection
                  62 M      Moderate acute (IIA)   12        None                            None
                  60 F      Moderate acute (IIA)   12        None                            None
                  53 F      Moderate acute (IIA)   6         Interstitial fibrosis and       None
                                                             hyalinization
                  50 M      Moderate acute (IIB)   6         None                            None
CsA               52 M      Moderate acute (IIA)   12        None                            None
                  48 M      Moderate acute ((B)    6         Histological rejection w/o      None
                                                             clinical findings
                  51 F      Moderate acute (IIA)   6         Glomerulosclerosis lesions      None
                  42 F      Mild acute (IB)        12        None                            None
                  73 M      Moderate acute (IIB)   6         None                            None
                  65 F      Moderate acute (IIA)   6         No clinical significant event   None
                            Moderate acute (IIA)   6         Histological rejection w/o      Steroids
                                                             clinical findings
                  66 M      Mild acute (IB)        12        AR (non-steroid resistant)      Steroids
                  72 M      Moderate acute (IIA)   6         nephrosclerosis                 none
* study drug discontinued

The results of the FDA’s composite endpoint analysis using the FDA definition of BPAR
are presented in Table 14. Of note, a similar analysis to that in Table 11 was also
conducted using the applicant’s definition of AR and similar results were obtained (data
not shown). Using a margin of 20% as defined in Appendix A, NI of all belatacept
regimens to CsA for the FDA endpoint was shown. While the upper limit of the 97.3%
CI for the LI arms is 13.7% and 11.9% for the two studies and thus each study meets the
20% margin, the upper bound of the CI for the belatacept MI arm in IM103008 is
essentially 20%.




                                                                                                        30
Table 14: FDA’s Analysis of Biopsy Proven Acute Rejection, Graft Loss, and Death
at 12 Months
Study                                                        Belatacept MI              Belatacept LI                     CsA
008        Met Endpoint                                       60/219 (27.4)              49/226 (21.7)                37/221 (16.7)
            Biopsy Proven Acute Rejection                          52                          45                          23
            Graft Loss                                              3                           3                           7
            Death                                                   4                           1                           6
            Unknown                                                 1                           0                           1
           Difference from CsA (97.3% CI)                    10.7 (1.6, 19.8)           5.5 (-2.7, 13.7)

027        Met Endpoint                                       62/184 (33.7)             51/175 (29.1)                 52/184 (28.3)
            Biopsy Proven Acute Rejection                           40                        37                           34
            Graft Loss                                              14                        11                           12
            Death                                                    6                         3                            5
            Unknown                                                  2                         0                            1
           Difference from CsA (97.3% CI)                    5.4 (-5.8, 16.6)          0.8 (-10.3, 11.9)
*First occurrence of biopsy proven acute rejection, graft loss, or death. Refer to subject and graft survival table for total number of
graft loss and death.


      6.3. Additional Efficacy Results

           6.3.1. GFR measured at Month 12

Measured GFR at Month 12 was significantly higher for both belatacept regimens
compared to CsA in Study IM103008 (see Table 15). In the study of ECD kidneys,
Study IM103027, measured GFR at Month 12 was higher for belatacept treated patients
compared to CsA but the finding was statistically significant only for the belatacept MI
versus CsA comparison.

                                      Table 15: Measured GFR at Month 12
Study                                                        Belatacept MI               Belatacept LI                    CsA
008        Mean (sd) at Month 12                               65.0 (30.0)                63.4 (27.7)                  50.4 (18.7)
           # in analysis                                           200                        206                         199
           Difference from CsA (97.3% CI)                    14.6 (8.9, 20.4)           13.0 (7.3, 18.7)
           p-value                                              <0.0001                    <0.0001

027        Mean (sd) at Month 12                              52.1 (21.9)                 49.5 (25.8)                  45.2 (21.1)
           # in analysis                                          154                         151                         154
           Difference from CsA (97.3% CI)                    6.9 (1.1, 12.7)            4.3 (-1.5, 10.2)
           p-value                                               0.0089                     0.0995

As noted previously in the discussion of efficacy endpoints (see Section 5), similar
differences were also detected as early as 1 month after transplantation. These early
differences may reflect hemodynamic (rather than structural) differences due to the
vasoconstrictive effects of CsA on the renal artery. The CsA group likely received less
renal blood flow, resulting in lower GFR.




                                                                                                                                          31
         6.3.2. Chronic Allograft Nephropathy at Month 12

In the Phase 3 trials, protocol biopsies were performed at Month 12 to assess structural
changes in the renal allografts. Direct examination of a protocol biopsy may provide
significant information to a local practitioner about the health of a transplant kidney. The
most appropriate methodology for quantifying protocol biopsy results for use in
transplant trials, however, remains undefined. In IM103008 and IM103027, biopsies were
classified as meeting or not meeting criteria for chronic allograft nephropathy (CAN). A
trend towards a higher prevalence of CAN was detected among the CsA groups (see
Table 16). As a histopathologic diagnosis, however, CAN has fallen into disfavor
specifically because it lacks a rigorous consensus definition and carries unclear
prognostic implications.


      Table 16: Prevalence of Chronic Allograft Nephropathy (CAN) at 12 Months
                                  Belatacept MI         Belatacept LI         CsA
    IM103008                          N=219                 N=226            N=219
    Prevalence (n, %)                40 (18.3)             54 (23.9)        71 (32.4)
    Difference of CsA 97.3% CI   -14.2 (-23.2, -5.0)   -8.5 (-17.9, 0.9)      NA
    IM103027                           N=184                N=175            N=184
    Prevalence (n, %)                 82 (44.8)            80 (46.0)        95 (51.6)
    Difference of CsA 97.3% CI    -6.8 (-18.2, 4.7)    -5.7 (-17.2, 6.0)      NA

     6.4. Summary of Efficacy

The co-primary outcomes of patient and graft survival, renal impairment, and AR (in
Study 103008) as defined in the protocol were examined and are reported in Section 6.2.
The other efficacy endpoints examined, GFR measured at 12 months and CAN at 12
months, were reported in Section 6.3. Briefly, these the results show that:

•    There were similar rates of patient and graft survival between the belatacept regimens
     and the CsA control in both studies.
•    The difference in the proportion of subjects meeting the composite renal endpoint was
     statistically significantly better for belatacept compared to CsA with the exception of
     the belatacept LI regimen in Study IM103027.
•    There were numerically more ARs in the belatacept-treated subjects than in the CsA-
     treated subjects; and in Study IM103008 the increase was significant. In this
     analysis, patients with graft loss, death, and loss to follow up were classified as
     successes. The severity of rejection was also greater in the belatacept groups, which
     will be discussed further in the safety discussion (Section 7.6).
•    The mean GFR at Month 12 in the belatacept treated patients was higher than in CsA
     treated patients (see additional discussion in Section 7.7).




                                                                                         32
•   The prevalence of CAN in belatacept treated patients was lower than in the CsA
    treated patients; however the prognostic value of CAN for long term survival is
    unclear.

The 10% NI margin proposed by the applicant for patient and graft survival could not be
justified given historical data. Though the effect of CsA on patients and graft survival
was large in the original studies versus azathioprine and corticosteroids, more recently
due to the introduction of new immunuosuppressants and newer treatment regimens, rates
of death and graft loss have been further reduced, and consequently a NI margin for this
endpoint can not be justified based on previously conducted studies. Nevertheless, the
rates of graft loss and death in studies of new product for kidney transplantation studies is
examined compared to approved therapies as an evaluation of safety.

The only endpoint studied by the applicant with a justified NI margin which could be
used to demonstrate efficacy was the AR endpoint, both as defined by the applicant and
as analyzed by FDA (See Appendix A for discussion and justification of the NI margin).
Using the 20% margin, NI of both belatacept LI regimens to CsA was demonstrated.
However, this margin was based only on historical data regarding the effect of CsA.
Whether or not the NI margin should be smaller based on clinical judgment (i.e, retain
more of the benefit of CsA) is a matter for discussion.

The endpoints related to renal function, though superiority were shown for some of them,
could not be used to demonstrate efficacy of belatacept since the absence of the known
toxicity of the control, CsA, does not prove the efficacy of belatacept.

The FDA conducted a further analysis of AR for Studies IM103008 and IM103027, the
results of which are presented in Table 14. The same analysis has been used for
essentially all recent clinical trials of kidney transplantation that evaluated other
immunuosuppressants. In the FDA-defined analysis of AR, patients with outcomes of
death, graft loss and loss to follow-up were counted as failures. The rationale for this
analysis is that these patients do not complete the 12 months of follow up and do not have
complete data on the incidence of AR. Therefore, these patients are imputed to have AR
for the purposes of the analysis and counted as failures. The FDA-defined analysis of AR
was performed for the two Phase 3 studies and showed that the belatacept LI regimen
was non-inferior to the CsA control based on a justified NI margin of 20% in both
studies. For the belatacept MI regimen in Study IM103008, however, the upper limit of
the 97.3% CI was nearly 20%, essentially the same as the NI margin, and 17% in
IM103027. Again, whether or not the NI margin should be smaller based on clinical
judgment (i.e, retain more of the benefit of CsA) is a matter for discussion.

The reason for justifying NI margins in active controlled studies is provided in the
regulations; specifically 21 CFR § 314.126, “Adequate and well-controlled studies,”
which state that when an active controlled study is conducted:

       “…the report of the study should assess the ability of the study to have detected a
       difference between treatments. Similarity of test drug and active control can mean



                                                                                          33
       either that both drugs were effective or that neither was effective. The analysis of
       the study should explain why the drugs should be considered effective in the
       study, for example, by reference to results in previous placebo-controlled studies
       of the active control drug.”

In the case of the current studies, this means that data from previously conducted studies
should be able to justify the use of the NI margins to assure that the study has assay
sensitivity and could have detected an effect between the control and a placebo, should a
placebo have been included. In the current trial, the background products used in both
arms (basiliximab, MMF and corticosteroids) are considered the putative placebo, and a
conservative effect of the amount that CsA contributes to these three products in the
regimen needs to be estimated so that a margin determined. Studies needed in order to
justify the margin for the endpoint of death and graft loss and for the composite renal
endpoint could not be identified. Only studies that allow the derivation of the margin for
the defined AR endpoints, both as defined by the applicant and by the FDA, were
available and support the conclusion that the belatacept LI regimen has been shown to
have efficacy based on the comparison to CsA, based on the justified margin of 20%.

7. Clinical Safety

Safety analyses were performed using the ITT populations for Studies IM103008 and
IM103027, and as-treated population in IM103100, as shown in Table 17 below, and
included 477 patients randomized to the belatacept MI regimen, 472 patients randomized
to the belatacept LI regimen, and 476 patients randomized to the CsA containing
regimen.

                  Table 17: Safety Population from Pooled Studies
  Study Number          Belatacept MI         Belatacept LI                    CsA
008                  219                   226                 221
027                  184                   175                 184
100                  74                    71                  71
TOTAL                477                   472                 476

One subject in the belatacept MI group (IM103027), 1 subject in the belatacept LI group
(IM103027) and 13 subjects in the CsA group (6 in IM103008, 5 in IM103027 and 2 in
IM103100) were included in the safety population, but did not receive study medication.

The number of subjects with at least 12 months of exposure to study medication was as
follows: 365/476 (77%) in the belatacept MI group, 369/471 (78%) in the belatacept LI
group, and 350/465 (75%) in the CsA group.

In the following discussion of safety, adverse events (AEs) are reported by Preferred
Terms (PT) and by System Organ Class (SOC) using the Medical Dictionary for
Regulatory Activities (MedDRA). AEs are reported for the first 12 months of the Phase
2 (IM103100) and Phase 3 studies (IM103008, IM103027), with the exception of deaths
and PTLD. Deaths are reported out to 24 months and also through the date of the Safety


                                                                                         34
Update Report (SUR) (covering events occurring between the March 2009 BLA database
lock and June/July 2009). In addition to reporting PTLD cases through database lock, the
applicant has also forwarded all new cases as they occur.

The primary safety endpoints of death, graft loss, and the combined endpoint of death and
graft loss were evaluated at 12 months and 24 months, as all patients have completed at
least 24 months of follow-up in Studies IM103008 and IM103027. Additional safety
endpoints include an assessment of cardiovascular risk factors known to be associated
with CsA and other CNIs. Endpoints related to hypertension, dyslipidemias, and new
onset diabetes after transplantation (NODAT) were therefore assessed prospectively.

The major safety signal identified in the belatacept arms was PTLD; this adverse reaction
is discussed in detail in Section 7.3.

   7.1. Deaths

Due to the differences in the LI regimen between the Phase 2 and Phase 3 trials, only the
deaths reported in the Phase 3 trials will be discussed in detail in this section.

The observed rates of death in both Phase 3 trials were similar across treatment groups.
While numerically fewer belatacept patients had died at 24 months in IM103008 and in
the belatacept LI group of IM103027, one must consider that several belatacept patients
in both trials alive at 24 months had already developed or would later develop PTLD (see
also Section 7.3). Given the significant mortality associated with PTLD – and
particularly associated with the PTLD in the central nervous system (CNS PTLD) seen in
the majority of the belatacept-associated cases – care should be taken when considering
the numeric advantage observed in the belatacept groups in terms of overall death at 24
months. Despite such caveats, however, the data from the Phase 3 trials support that
kidney transplant patients maintained on either belatacept regimen had similar mortality
outcomes at 24 months to the CsA control regimen, even among a patient population
which included EBV negative (EBV-) recipients and despite the impact of the imbalance
in PTLD presentations across treatment groups (see Table 18).




                                                                                       35
                      Table 18: Deaths Occurring in the Phase 3 Trials
                                     IM103008                                  IM103027
                       Belatacept    Belatacept     CsA        Belatacept     Belatacept         CsA
                          MI             LI                       MI              LI
                        (n=219)       (n=226)     (n=221)       (n=184)        (n=175)         (n=184)
  Death
   Up to Month 12        6 (2.7)      4 (1.8)      7 (3.2)       8 (4.3)        5 (2.9)         8 (4.3)

    Up to Month 24       7* (3.2)     8† (3.5)     13 (5.9)     13 (7.1)      11‡ (6.3)        12 (6.5)

    Through Safety       7* (3.2)     9† (4.0)     14 (6.3)     19 (10.3)     13‡ (7.4)        16 (8.7)
    Update Report
 PTLD patient = patient diagnosed with PTLD at any time during trial
 * = 3 PTLD patients alive at 24 months and at SUR
 † = 2 PTLD patients alive at 24 months and at SUR
 ‡ = 3 PTLD patients alive at 24 months and at SUR; Note: Two of the three patients in the belatacept LI
 treatment group of Study IM103027 with PTLD diagnoses who were alive at 24 months of follow up and at
 SUR have since died (PID 27-10-10029 and PID 27-138-10593, both in June 2009). All other patients in
 the Phase 3 studies who were alive at 24 months of follow up with PTLD diagnoses are still alive.


         7.1.1. Discussion of Death Narratives
 Patient narratives summarizing the circumstances of death were reviewed. As expected
 among kidney transplant recipients, most deaths appeared attributable to cardiovascular
 causes or infection. Several occurred in the immediate post-operative setting. The most
 notable difference between patient groups is the three deaths due to CNS lymphoma that
 occurred by the 24 month time point among belatacept-treated patients. Of these three
 deaths, two occurred in EBV+ patients and one occurred in an EBV- patient. Table 19-
 Table 24 below reflects the cause of death as inferred from the submitted death narrative
 by the reviewing FDA Medical Officer.

 IM103008
Table 19: Narratives of Deaths among CsA Patients in IM103008 at 24 Months
              Year
Patient       Died    Age      Sex     Cause of Death Inferred from Death Narrative
22-20679      1       38       M       Patient had "seizure" while driving, then died of MI in hospital
76-20171      1       55       M       Hemorrhagic shock immediately post-op, never received CsA
84-20700      1       48       F       Sepsis
89-20350      1       50       M       Cardiac arrest immediately post-operatively
98-20439      1       45       M       Suicide
118-20353     1       20       M       Pneumonia
123-20740     1       54       F       Sepsis, wound infection
10-20254      2       66       M       Cardiac arrest in setting of massive bleed on anticoagulation
41-20088      2       61       M       Cardiac arrest in setting of fever
48-20042      2       64       M       Pneumonia
84-20405      2       53       M       Pulmonary adenocarcinoma
91-20103      2       55       F       Hypoxia due to pulmonary edema
116-20453     2       37       F       Interstitial lung disease



                                                                                                          36
        Table 20: Narratives of Deaths among Belatacept LI Patients in IM103008 at 24
                                          Months
                  Year
   Patient        Died     Age         Sex     Cause of Death Inferred from Death Narrative
                                               Patient died on Day 358; distant history of wound infection, no other
   10-20037       1        55          M       information available
   46-20592       1        61          F       Sepsis immediately post-operatively
   123-20230      1        19          F       Generalized bleeding (not witnessed)
   124-20297      1        55          M       Myocardial infarction
   9-20120        2        47          M       Metastatic sarcoma
   25-20375       2        71          F       Non small cell lung cancer
   57-20645       2        37          M       Bowel ischemia due to severe mesenteric artery occlusion
   154-20618      2        26          M       No information available regarding circumstances of death

       Table 21: Narratives of Deaths among Belatacept MI Patients in IM103008 at 24
                                          Months
                  Year
   Patient        Died         Age     Sex               Cause of Death Inferred from Death Narrative
   46-20330        1           77       M      Hemorrhage from vascular anastamosis on Day 12
   46-20650        1           74       M      Urosepsis
   84-20669        1           41       F      Sudden cardiac death
   85-20526        1           71       F      CMV colitis, sepsis
   124-20561       1           72       M      Lung adenocarcinoma
   159-20686       1           53       M      Bronchopulmonary aspergillosis
   41-20344        2           46       M      Complications of West Nile Virus

     IM103027

Table 22: Narratives of Deaths among CsA Patients in IM103027 at 24 Months
               Year
Patient        Died      Age         Sex     Cause of Death Inferred from Death Narrative
2-10002        1         56          M       "Veered off road and collapsed"
32-10077       1         63          M       Rupture of mycotic aneurysm
57-10086       1         64          M       Pulmonary embolus
60-10260       1         73          M       Significant intragraft bleed immediately post-op, MI; never received CsA
85-10110       1         65          F       Sepsis
85-10531       1         72          M       Obstructive uropathy, sepsis
85-10572       1         70          F       Intestinal ischemia
105-10400      1         58          F       Renal vein thrombosis, GI hemorrhage, sepsis
10-10275       2         62          M       Sepsis (MRSA); also infected with pseudomonas and aspergillosis
17-10061       2         68          F       Fungal pneumonia, disseminated CMV disease
54-10502       2         79          M       Kaposi's Sarcoma
105-10573      2         68          F       Sepsis




                                                                                                               37
Table 23: Narratives of Deaths among Belatacept LI Patients in IM103027 at 24 Months
               Year
Patient        Died   Age      Sex    Cause of Death Inferred from Death Narrative
64-10372       1      65       M      Withdrew consent on Day 42; reported to have died on Day 353
66-10516       1      68       M      Sigmoid colon perforation; sepsis
85-10569       1      66       F      Necrotizing fascitis and sepsis on Day 3
91-10089       1      42       M      Disseminated tuberculosis, invasive CMV disease, sepsis
97-10452       1      75       M      Endocarditis, sepsis
26-10440       2      55       M      Complications related to transitional cell carcinoma
29-10427       2      68       M      Myocardial infarction
93-10037       2      49       M      Strongyloidiasis, CMV disease, pneumonia
101-10093      2      75       M      CNS lymphoma (PTLD); Pt was EBV+ at baseline
                                      Cryptococcosis, "cerebellar syndrome", seizures, coma; post-mortem
124-10466      2      55       M      cerebral cultures and biopsy were negative
124-10527      2      44       M      Patient found dead in house on Day 426; cause unknown

  Table 24: Narratives of Deaths among Belatacept MI Patients in IM103027 at 24 Months
               Year
Patient        Died   Age      Sex   Cause of Death Inferred from Death Narrative
22-10496       1      70       F     Cardiac arrest on Day 3
53-10399       1      63       F     Surgical wound infection
58-10390       1      68       F     Ischemic heart disease
70-10447       1      65       M     CNS lymphoma (PTLD); Pt was EBV- at baseline
85-10249       1      50       M     Myocardial infarction
                                     Severe post-operative MI and sepsis; Pt ultimately died on Day 96 of multi-
89-10114       1      79       M     organ failure and sepsis
93-10027       1      65       M     Myocardial infarction
102-10494      1      58       M     Post-operative renal hemorrhage; ischemic bowel
30-10184       2      72       M     Severe aortic valve disease
35-10185       2      71       M     CNS lymphoma (PTLD); Pt was EBV+ at baseline
97-10453       2      73       M     Withdrew consent Day 13; Pt reported to have died on Day 669
102-10164      2      61       M     Pneumonia
125-10433      2      62       M     Osteomyelitis, staph sepsis

           7.2. Graft Loss (and Deaths)

    Death, graft loss, and the combined endpoint of death and graft loss were evaluated in the
    Phase 3 trials IM103008 and IM103027 at 12 months and 24 months. Randomized
    clinical trials in transplantation lack the statistical power to rigorously evaluate the
    combined endpoint of patient death and graft loss. Numerically, however, more
    belatacept patients across the Phase 3 trials were alive with a functioning graft at both
    Month 12 and Month 24 (see Table 25 and Table 26). The numeric difference observed
    was driven by IM103008. Given the lack of statistical power, one cannot determine
    whether the numeric difference represents chance or true benefit with regard to GFR,
    blood pressure, diabetes, dyslipidemia on belatacept regimens. While one cannot
    conclude that belatacept patients had a more favorable outcome with respect to combined
    death and graft loss compared to CsA patients in the Phase 3 trial, it is somewhat
    reassuring that the numeric advantage persists despite the greater number of PTLD events


                                                                                                        38
        among belatacept patients (even if all belatacept patients alive and with a functioning
        graft at 24 months are imputed as failures for the endpoint).

                          Table 25: Patient and Graft Status in the Phase 3 Trials
                                               IM103008                                      IM103027
                                Belatacept     Belatacept        CsA         Belatacept      Belatacept           CsA
                                   MI              LI                           MI               LI
                                 (n=219)        (n=226)        (n=221)        (n=184)         (n=175)         (n=184)
At Month 12
   Alive and no GL                 208            218            206             158             155            156
   Graft Loss                       4          5 (1 died)     8 (1 died)     17 (2 died)     16 (1 died)    20 (3 died)
   Death w/ functioning graft       6              3              6               6               4              5
   Unknown                          1                             1               3                              3

At Month 24
   Alive and no GL                 204            212            198             151             146            148
   Graft Loss                   7 (1 died)     5 (1 died)     8 (1 died)     18 (2 died)     20 (3 died)    22 (4 died)
   Death w/ functioning graft       6              7              12              11              8              8
   Unknown                          2              2               3               4              1              6

Through Safety Update
  Alive and no GL                  204            208             195            147             143            144
  Graft Loss                    7 (1 died)     8 (1 died)     10 (1 died)    18 (4 died)     21 (3 died)    22 (4 died)
  Death w/ functioning graft        6              8               13             15              10            12
  Unknown                           2              2                3              4               1             6


              Table 26: Death and/or Graft Loss at 12 and 24 Months and Safety Update
                                                IM103008                                   IM103027
                                   Belatacep     Belatacept      CsA        Belatacept      Belatacept      CsA
                                      t MI           LI                        MI               LI
                                    (n=219)       (n=226)      (n=221)       (n=184)         (n=175)       (n=184)
    Death/Graft Loss
      Up to Month 12                   10             8            14           23            20             25
      Up to Month 24                  13*            12†           20           29            28‡            30
      Through Safety Update           13*            16†           24           33            31‡            34
    Death/Graft Loss/Unknown
      Up to Month 12                   11             8            15           26            20             28
      Up to Month 24                  15*            14†           23           33            29‡            36
      Through Safety Update           15*            18†           26           37            32‡            40
      PTLD patient = patient diagnosed with PTLD at any time during trial
      * = 2 PTLD patients alive with functioning graft at 24 months and at SUR
      † = 2 PTLD patients alive with functioning graft at 24 months and at SUR
      ‡ = 2 PTLD patients alive with functioning graft at 24 months and at SUR; Note: One of the two patients
      with PTLD alive with a functioning graft in the belatacept LI arm of Study IM103027 died after the Safety
      Update Report (PID 27-91-10044 died on June 27, 2009)

            7.3. Post Transplant Lymphoproliferative Disorder (PTLD)

        The applicant is seeking approval for the belatacept LI regimen studied in the Phase 3
        trials. That regimen included an extra 10 mg/kg infusion of belatacept on Day 5


                                                                                                             39
    compared to the belatacept LI regimen studied in the Phase 2 trial. As will be discussed,
    six cases of PTLD (including 3 cases of CNS PTLD) developed among recipients of the
    Phase 3 belatacept LI regimen (6/401 patients) and no cases of PTLD developed among
    recipients of the Phase 2 belatacept LI regimen (0/71 patients). PTLD represents the
    major safety signal detected in the belatacept trials and it appears biologically plausible
    that the extra exposure to belatacept in the early post-transplant period may explain the
    development of PTLD among Phase 3 recipients of the LI regimen.

    The numbers of patients across the Phase 2 and Phase 3 trials, who developed PTLD,
    while clinically significant, were too few to perform a meaningful exposure-response
    analysis. For that reason, many of the analyses of PTLD observed among belatacept-
    treated patients in comparison to CsA-treated patients pool patients across belatacept
    treatment groups and across the Phase 2 and Phase 3 trials. While the pooling represents a
    limitation of the analyses, the alternative of performing separate analyses for each
    belatacept regimen was judged to have greater limitations. Nonetheless, one should keep
    this limitation in mind when considering the number needed to harm (NNTH)
    calculations presented in Section 7.3.3.

    The safety signal of PTLD appeared consistently across all three trials in both belatacept
    treatment groups and within various patient subgroups. To date, there have been eight
    (1.7%), six (1.3%) and two (0.4%) reported cases of PTLD in the belatacept MI, LI, and
    CsA groups respectively (see Table 27 and Table 28). Please note that many of the
    analyses of PTLD which follow reference the incidence of PTLD at 2 years of follow-up.
    The 2 year time point was selected as complete PTLD data were available for all patients
    for 2 years of follow-up; the use of the 2 year time point has the effect of censoring a case
    of CNS PTLD that developed in a patient (PID 103027-91-1044) in the belatacept LI
    regimen in IM103027 after 24 months in those analyses.

                Table 27: All PTLD Cases for Belatacept Trials 008, 027 and 100
                       Bela MI        Bela LI       CsA         Diff (pooled     Diff (M1–CsA)       Diff (LI–CsA)
                       N = 477        N = 472      N = 478      LI/MI–CsA)
PTLD in all            8 (1.7)        6 (1.3)      2 (0.4)       (-0.1, 2.1)        (-0.1, 2.9)        (-0.4, 2.4)
patients              (0.7, 3.3)     (0.5, 2.7)   (0.1, 1.5)

EBV+                   N = 404        N = 401      N = 399       (-0.2, 1.6)        (-0.5, 1.8)         (0, 2.5)*
PTLD in EBV+           2 (0.5)        4 (1.0)       0 (0)
                      (0.1, 1.8)     (0.3, 2.5)    (0, 0.9)

EBV-                   N = 45         N = 51       N = 57       (-2.5, 13.1)        (-0.3, 22.3)       (-5.9, 11.9)
PTLD in EBV-          5 (11.1)        2 (3.9)      1 (1.8)
                     (3.7, 24.1)    (0.5, 13.5)    (0, 9.4)

EBV unknown           N = 28          N = 20       N = 22       (-20.3, 8.2)       (-19.4, 14.7)      (-23.0, 13.2)
 PTLD in EBV          1 (3.6)          0 (0)        1 (4.5)
unknown
   * exact p-value < 0.05
   All confidence intervals reported are exact confidence intervals due to the small number of events.




                                                                                                             40
        At 24 months of follow-up, three belatacept MI patients from the Phase 3 trials had
        already died from complications related to PTLD (and are therefore reflected in the total
        deaths presented in the 24 month efficacy and safety data). To date, two belatacept LI
        patients and one CsA patient from the Phase 3 trials have died from complications related
        to PTLD beyond the 24 month follow up period (and are therefore not reflected in the
        total deaths presented in the 24 month efficacy and safety data). Three additional
        belatacept MI patients and three additional belatacept LI patients from the Phase 3 trials
        who developed PTLD remain alive and at significant risk of death and other
        complications. A post hoc central pathology review of the PTLD cases suggests the
        possibility that two of the three belatacept LI patients who remain at risk may have been
        misdiagnosed with PTLD. One (PID 103008-113-20060) of those two patients, however,
        underwent six courses of rituximab, cyclophosphamide, adriamycin, vincristine, and
        prednisone (R-CHOP) based on the local pathology read; the other (PID 103008-122-
        20256) had PCR evidence of a clonal population with specific rearrangement of the
        heavy chain, consistent with a diagnosis of PTLD. No additional CsA patients from the
        Phase 3 trials, other than the single patient who has already died, have developed PTLD
        to date.

        Table 28: Summary of PTLD Cases Observed in Belatacept Trials 008, 027 and 100
                              Belatacept MI                     Belatacept LI*                         CsA
                                 (N=477)                           (N=472)                           (N=478)
                                             EBV                               EBV                               EBV
                     EBV+         EBV-     Unknown       EBV+       EBV-     Unknown       EBV+       EBV-     Unknown
     Trial          (n=404)      (n=45)     (n=28)      (n=401)    (n=51)     (n=20)      (n=399)    (n=57)     (n=22)
IM103100
CNS PTLD                      2Ω,†
Non-CNS PTLD                                1†                                                                 1‡
IM103008
CNS PTLD           1Ω         1Ω
Non-CNS PTLD                  1Ω                        2Ω,Ω                                         1‡
IM103027
CNS PTLD           1†         1†                        2Ω,†        1‡
Non-CNS PTLD                                                        1‡
           Total 2 (0.5)        5 (11.1)      1 (3.6)    4 (1.0)     2 (3.9)         0           0   1 (1.8)    1 (4.5)
      *The belatacept LI regimen studied in IM103100 is non-identical to that studied in IM103008 and
      IM103027
      Ω = Patient alive; † = Patient dead in first 24 months; ‡ = Patient died after first 24 months


                7.3.1. Summary PTLD Cases Observed in the Trials 008, 027, and 100 by
                     Treatment Group

        CsA:

        IM103008-101-20019: 35 year old EBV-, CMV+ male who received an EBV+, CMV+
        kidney. Patient experienced immediate Grade 1A AR requiring pulse steroids and
        increased CsA dosing. At Day 82, another for-cause biopsy showed borderline rejection



                                                                                                               41
requiring additional pulse steroids. On Day 163, the patient underwent re-implantation of
his ureter due to stricture. On Day 168, a liver ultrasound obtained in the setting of
elevated LFTs identified four liver nodules. Subsequent work up confirmed a diagnosis
of PTLD (large B cell lymphoma) involving liver, lung, and allograft kidney. The patient
underwent 6 courses of R-CHOP (rituxan, cyclophosphamide, doxorubicin, vincristine,
and prednisone) and appeared to have a good response: PET scan on Day 344 showed no
residual evidence of PTLD lesions. The patient began sirolimus on Day 337. The patient
appeared for a clinic visit on Day 803 where he appeared well. At home, however, he
developed vomiting and rigors and died on Day 804. Cause of death was reported as
lobar pneumonia.

IM103100-3-25: 53 year old EBV- male who received a kidney from a donor with
unknown EBV and CMV serologies. On Day 1417, the patient presented with fever,
chills, and nausea. Serum LDH was elevated and a CT scan of the abdomen revealed soft
tissue masses involving the liver and allograft kidney. Liver biopsy confirmed a diagnosis
of large B-cell PTLD. The patient was treated with rituximab from Days 1422-1437. The
patient developed a mucormycosis infection (sinus) on Day 1440 and died on Day 1473.

Belatacept MI:

IM103008-142-30548: 25 year old EBV-, CMV- male who received a kidney from an
EBV-, CMV- donor. The subject received thymoglobulin on Days 4-8 in the setting of
graft dysfunction. Renal doppler and angiography studies on Day 9 were consistent with
renal vein thrombosis; the RVT resolved with enoxaparin. On Day 498, the patient
presented with convulsions and tongue-biting. Head CT showed a lesion in the left frontal
lobe. On Day 506, a craniotomy and partial excision of tumor was performed – the tumor
was consistent with PTLD but stained negative for EBV. On Day 516, the patient began 4
courses of methotrexate, leucovorin, and procarbazine. MRI prior to the fourth cycle
showed a new nodule in the right occipital lobe and evidence of residual tumor. On Day
581, PET scan showed evidence of disease in the frontal lobe, L2 vertebra, and axillary
and pelvic nodes. The patient received additional treatment with rituximab and
temozolomide. As of         (b) (6) , the patient was still alive and with a functioning
allograft.

IM103008-27-20039: 22 year old EBV-, CMV- male who received a kidney from an
EBV+, CMV- donor. A for-cause biopsy on Day 76 showed Grade IIA AR. The patient
received 10 doses of thymoglobulin, stress dose steroids; belatacept was stopped and
tacrolimus was started. A follow-up biopsy on Day 90 showed PTLD (diffuse large B cell
lymphoma) and BK nephropathy. Repeat biopsy on Day 95 showed Grade III AR and
confirmed the PTLD diagnosis. Rituximab was started on Day 97; IVIg was also given
for rejection. PET scan on Day 125 (September 5, 2006) was compatible with PTLD
involving the neck, right axilla, and left inguinal region. As of  (b) (6) , the patient was
alive but had returned to dialysis.

IM103008-76-20734:38 year old EBV+, CMV- male who received a kidney from an
EBV+, CMV+ donor. A for-cause biopsy on Day 43 was consistent with Grade IA acute



                                                                                         42
rejection; the patient received 15 days of pulse steroids, followed by 5 days of
thymoglobulin and additional steroids. On Day 154, the patient was hospitalized with
CMV colitis. On Day 281, another for-cause biopsy was consistent with BK viral
associated nephropathy. On Day 427, the patient developed aphasia, vomiting, and
disorientation. An MRI showed 2 cerebral lesions; brain biopsy revealed PTLD (large B
cell lymphoma). The patient was treated with methotrexate, rituximab, and cytarabine.
On Day 499 (              (b) (6) ) follow up CT showed some improvement of the cerebral
lesions. Additional severe adverse events that have been subsequently reported include
renal impairment on Day 525, pulmonary embolism on Day 586, neutropenia on Day
712, and lung disorder on Day 712. As of          (b) (6) , the patient was reported as alive
with a functioning graft.

IM103027-35-10185: 71 year old EBV+, CMV+ male who received a kidney from an
EBV+, CMV+ donor. On Day 305, the patient developed CMV infection treated with
valgancyclovir. The patient was also treated with heparin for phlebitis. In the setting of
anticoagulation, the patient experienced a brain hemorrhage; cerebral MRI on Day 325
showed a bifrontal lesion. Brain biopsy on Day 367 confirmed PTLD (large B cell
lymphoma). The patient was treated with radiotherapy and rituximab. On Day 402, the
patient became comatose and developed a bronchial superinfection. The patient died on
Day 415.

IM103027-70-10447: 65 year old EBV-, CMV+ male who received a kidney from an
EBV+, CMV+ donor. On Day 256, the patient presented with left hemiparesis. Head CT
demonstrated a right frontal lesion. Patient underwent surgical excision on Day 272;
pathology was consistent with PTLD (large B cell lymphoma). Patient developed
convulsions of Day 296; methotrexate, vincristine, procarbazine, and dexamethasone
were initiated on Day 301. On Day 322, the patient developed E. coli sepsis and died on
Day 322.

IM103100-1-6: 36 year old EBV-, CMV- male who received a kidney from a donor with
unknown serologies. A for-cause biopsy on Day 27 was consistent with Grade III acute
rejection. The patient was treated with OKT3 for 10 days. Belatacept was discontinued
immediately; tacrolimus was started on Day 39. On Day 337, the patient developed left-
sided headaches in the setting of a pneumonia. MRI showed a lesion at the base of the
skull. PET scan was consistent with activity at the neck, esophagastric junction, and at
L2-L3 and L3-L4. Cervical node biopsy demonstrated PTLD (large B-cell lymphoma).
The patient received treatment with cyclophosphamide, hydroxydaunomycin, vincristine,
and prednisone (CHOP) and rituximab. On Day 477, whole body CT scans were
unremarkable. Clinically, the patient has been in remission since 2004; the patient was
reported alive with a functioning graft in       (b) (6) .


IM103100-9-4: 23 year old EBV-, CMV+ male who received a kidney from an EBV+,
CMV+ male. On Day 109, the patient was hospitalized with a CMV infection (fever,
neutropenia). On Day 262, the patient developed headache, nausea, vomiting, and
behavioral changes. MRI showed a lesion near the basal ganglia; stereotactic needle
biopsy confirmed PTLD (large B cell lymphoma). The patient was treated with high-dose



                                                                                            43
steroids and radiation, On Day 424, the patient was hospitalized with PCP pneumonia
and evidence of CMV disease. The subject died on Day 440 (autopsy indicated that
patient was in full remission with regards to lymphoma).

IM103100-3-31: 42 year old CMV+ female (EBV serostatus unknown) who received a
kidney from CMV+ donor (EBV serostatus unknown). A for-cause biopsy on Day 60
was consistent with Grade IA rejection; the patient was treated with pulse steroids. On
Day 93, the patient developed urethral obstruction requiring nephrostomy and ureteral
stent insertion. The patient required another for-cause renal biopsy on Day 111 which
demonstrated PTLD (large B cell lymphoma). The patient received treatment with R-
CHOP. On Day 122, the patient experienced graft loss. On Day 210, the lymphoma was
considered resolved. The patient died from septicemia in 2006 while awaiting re-
transplant.

Belatacept LI:

IM103008-113-20060: 30 year old EBV+, CMV- male who received a kidney from an
EBV+, CMV- donor. A for-cause renal biopsy on Day 73 showed a differentiated
monomorphic T cell infiltrate. Pt was asymptomatic; serum LDH was normal. Serum
creatinine was 1.5 mg/dL. On Day 78, head CT was unrevealing; chest and abdominal
CT showed two small lesions in the liver, but no lymphadenopathy. On Day 86, PET scan
showed no findings supportive of PTLD. On Day 93, follow-up renal biopsy showed no
evidence of acute rejection or PTLD. Based on the prior biopsy, however, the patient
received six cycles of rituximab, vincristine, and cyclophosphamide. Of note, a post hoc
central pathology reading disputed the diagnosis of PTLD on the original biopsy. As of
       (b) (6) , the patient was reported alive with a functioning graft.


IM103008-122-20256: 44 year old EBV+, CMV- male who received a kidney from an
EBV+, CMV- donor. A for-cause renal biopsy on Day 93 showed a B cell infiltrate
suggestive of PTLD. PCR provided evidence of a clonal population with specific
rearrangement of the heavy chain, consistent with a diagnosis of PTLD. Chest and
abdomen CT scans were unremarkable; serum LDH was normal. Patient was treated with
bolus prednisolone (20 mg/day for two weeks, then increased to 30 mg/day); belatacept
and MMF were stopped. No other treatment was provided. Serum creatinine of Day 135
was 0.95 mg/dL. Follow-up renal biopsy on Day 161 showed no evidence of lymphoma.
Patient converted treatment with tacrolimus and MMF on Day 220                (b) (6) ). Of
note, the post hoc central pathology reading also disputed this diagnosis of PTLD. As of
        (b) (6) the patient was alive with a functioning graft.


IM103027-101-10093: 75 year old EBV+, CMV- male who received a kidney from a
donor with unknown serologies. On Day 81, the patient developed fever and cough. He
began treatment for CMV disease with IV gancyclovir on Day 97; the CMV infection did
not resolve until Day 185 (despite discontinuing belatacept). The patient began treatment
with everolimus on Day 245. Due to failure to thrive, patient was hospitalized on Day
504. Bone marrow biopsy on Day 508 showed megaloblastic changes. Head CT on Day
509 suggested lymphoma. MRI on Day 511 showed two nodular lesions. Stereotactic



                                                                                         44
biopsy on Day 523 was consistent with PTLD (diffuse large B-cell lymphoma). On Day
594, the patient died.

IM103027-10-10029: 78 year old EBV-, CMV+ female who received a kidney from a
donor with unknown serologies. On Day 82, the patient underwent a for-cause biopsy
which was consistent with PTLD (large B-cell lymphoma). MRI of abdomen showed
multiple nodules in the renal allograft and enlarged retroperitoneal and pelvic lymph
nodes. Head CT did not show any intracranial disease. The patient received 3 doses of
rituximab. Belatacept and MMF were stopped; sirolimus was started. On Day 109, serum
creatinine increased to 5.3 mg/dL and hemodialysis was initiated. A transplant
nephrectomy was performed. Tissue obtained from the nephrectomy confirmed the
diagnosis of PTLD. The patient died in       (b) (6) secondary to a “lung problem” per
family report. No other information regarding circumstances of death is available.

IM103027-138-10593: 38 year old EBV-, CMV- male who received a kidney from an
EBV+, CMV- donor. On Day 221, the patient was hospitalized with fever, nausea,
vomiting, and hypoglycemia. On Day 231 (December 31, 2007), head CT showed
multiple enhancing lesions. Brain biopsy on Day 242 showed PTLD (large B-cell
lymphoma with T-cell reactive component). The patient was treated with dexamethasone,
methotrexate, prednisone, and radiotherapy. Follow up MRI on Day 324 showed some
improvement, but patient died on         (b) (6) secondary to “deterioration of cognitive
ability.”

IM103027-91-10044: 66 year old EBV+, CMV+ male who received a kidney from a
donor with unknown serologies. The patient initially enjoyed an unremarkable post-
transplant course. In              (b) (6) , however, he began to lose weight. In         (b) (6)
   (b) , he began to complain of malaise and anorexia. In                   (b) (6) he developed
   (6)
confusion and was hospitalized with a serum sodium of 116 mEq/L. Head CT revealed a
mass in the left anterior cerebral region. Cerebral stereotactic biopsy confirmed a
diagnosis of PTLD (diffuse large B cell lymphoma). The event remains unresolved. As of
             (b) (6) the patient is alive with a functioning graft.

        7.3.2. Preponderance of CNS PTLD among Belatacept Treated Patients

Other than perhaps the overall incidence, the preponderance of CNS involvement among
the kidney transplant recipients who received belatacept constitutes the most striking
aspect of the PTLD presentations. PTLD involving the CNS typically is a rare event
among kidney transplant patients. For example, a retrospective analysis of 1,094 post-
transplant lymphomas reported to the Collaborative Transplant Study registry, for
instance, reported involvement of the CNS in only 11.7% of cases. 15 In contrast, nine out
of fourteen (or 64.3%) of the PTLD cases observed in belatacept groups from the three
kidney trials involved the CNS. A post hoc central evaluation of the biopsy material casts
doubt on the accuracy of two of the diagnoses of PTLD, both of which represented non-
CNS cases. If one accepted this unplanned second opinion, the rate of CNS PTLD among

15
 Opelz G. Döhler, B. Lymphomas after solid organ transplantation: a collaborative transplant study report.
Amer J of Transplantation 2003; 4: 222-30.


                                                                                                       45
all PTLD cases would become nine out of twelve (or 75%) of the PTLD cases observed
in the context of kidney transplant patients exposed to belatacept presented in the CNS.

Inspection of the PTLD patient narratives underlines that the morbidity and mortality
associated with CNS presentations of PTLD differ from non-CNS presentation. Adverse
events associated with CNS presentation included seizures, hemiparesis, aphasia,
cognitive dysfunction, and behavioral changes. In a non-belatacept context, the literature
reports a significantly higher mortality rate among kidney transplant patients who
develop CNS PTLD compared to those who develop non-CNS PTLD. 16

The significance of the CNS involvement goes beyond any immediate implications with
respect to mortality and morbidity. The unusual pattern of PTLD presentations provides
additional context to other CNS-related signals detected in the belatacept clinical
development program, such as the two isolated cases of progressive multifocal
leukoencephalopathy (see Section 7.4) and the seven CNS infections distributed across
the belatacept MI groups in IM103100, IM103008, and IM103027 (including three cases
of cryptococcal meningitis, one case of West Nile Virus, one case of cerebral
aspergillosis, one case of neurological herpes zoster, and one of the PML cases)
compared to only one event among the belatacept LI groups (cryptococcal meningitis)
and one event among the CsA groups (also cryptococcal meningitis) (see Section
7.11.1.1). Even more importantly, however, the pattern of presentations implies that –
unless kidney transplant patients maintained on belatacept compensate by developing
fewer cases of non-CNS PTLD than those maintained on CsA – overall incidence of
PTLD is increased among patients with exposure to belatacept.

         7.3.3. Incidence of PTLD in Relation to EBV Serostatus

Known risk factors for the development of post-transplant lymphoproliferative disease
among recipients of kidney transplants include EBV seronegativity, the development of
CMV disease post-transplant, and exposure to lymphocyte depleting agents. 17,18,19 Not
surprisingly, then, the increased risk of PTLD was most evident among the EBV- patients
in the belatacept trials (see Figure 5). While the PTLD signal among EBV+ patients was
smaller across all treatment groups, however, the rates were still appreciably higher
among patients in the belatacept treatment groups than among patients in the CsA
treatment groups: two events out of 404 EBV+ patients in the belatacept MI groups and
four events out of 401 EBV+ patients in the belatacept LI groups, compared to zero
events out of 399 EBV+ patients in the CsA groups.

16
   Opelz G. Döhler, B. Lymphomas after solid organ transplantation: a collaborative transplant study report.
Amer J of Transplantation 2003; 4: 222-30.
17
   Opelz G. Döhler, B. Lymphomas after solid organ transplantation: a collaborative transplant study report.
Amer J of Transplantation 2003; 4: 222-30.
18
   Caillard S, Dharnidharka V, Agodoa L, et al. Posttransplant lymphoproliferative disorders after renal
transplantation in the United States in era of modern immunosuppression. Transplantation 2005; 80: 1233-
43.
19
   Shahinian V, Miurhead N, Jevnikar AM, et al. Epstein-Barr virus seronegativity is a risk factor for late
onset posttransplant lymphoproliferative disorder in adult renal allograft recipients. Transplantation. 2002;
75: 851-6.


                                                                                                          46
      Figure 5: Rates of PTLD Overall and by EBV Status by Treatment Arm


                12.00%

                10.00%

                 8.00%
                                                                       B-LI
                 6.00%                                                 B-MI
                                                                       CsA
                 4.00%

                 2.00%

                 0.00%
                             PTLD         in EBV+       in EBV-

The applicant suggests that the rate of PTLD in EBV+ patients observed in the CsA
control group across the trials was unexpectedly low and that rates observed in the
belatacept groups among EBV+ patients is consistent with estimates using registry data.
Findings from an FDA review of multiple clinical trials (submitted from other applicants
for non-related applications) that included similar control groups and analyses of registry
databases refute the applicant’s suggestion (see below and Appendix B).

Calculation of Number of Patients Needed to Harm (NNTH) Using Data from the
Belatacept Trials
A simple point estimate of the number of EBV+ kidney transplant recipients needed to
treat with a belatacept-based regimen rather than a CsA-based regimen in order to
observe one additional case of PLTD can be derived by taking the inverse of the
difference in incidence of PTLD among the belatacept treated patients and the incidence
of PTLD among the CsA treated patients in the clinical trials. This number will be
referred to as the number need to harm (NNTH). At 24-months post-transplant, there
were 5/805 and 0/399 PTLD cases in the pooled belatacept (MI and LI) and CsA control
respectively resulting in an estimated NNTH of 161 (i.e. [1/(5/805-0/399)]) EBV+ kidney
transplant recipients with exact 95% confidence interval of (-341, 68). This is interpreted
as ranging from as few as 68 belatacept subjects need to be treated for one case of PTLD
to as many as 341 CsA subjects for one case of PTLD at two years of follow up (see
Figure 6). As previously noted, the selection of the 2 year time point has the effect of
censoring a known case of CNS PTLD which occurred in an EBV+ belatacept LI patient.
The two year time point, however, reflects the latest time point at which complete PTLD
data exist for all patients.




                                                                                         47
                       Figure 6: NNTH Estimate from Belatacept Trials




Graph adapted from method used in Altman, D “Confidence intervals for the number needed to treat.” BMJ
1998;317:1309-1312 (7 November). Graphing method used to help interpret confidence intervals for number needed to
harm/benefit when CI contains infinity.


Incidence rate per person-years (PYs) was calculated to allow for comparison to registry
data. Additionally, given that not all randomized patients in the clinical trials had a full
two years of follow-up due to early discontinuation, it is useful to consider incidence per
person-years. The point estimate of the number needed to harm calculated based on the
difference between incidences per person-years is slightly higher. Please see Appendix B
for the derivation of the incidence of PTLD in person-years observed among EBV+
patients receiving belatacept in the clinical trials (the result is 0.35 per 100 patient years).
Given an incidence/PY of 0.0035 of the pooled data from the belatacept MI and LI
regimens and of 0.0 in the CsA regimen from studies 008, 027 and 100, we estimate a
number needed to harm (NNTH) of 145 (95% CI: -532, 64) patients (Table 29). This
suggests that for every 145 EBV+ patients treated with a belatacept-based regimen
instead of a CsA-based regimen for two years, one would expect at least one additional
case of PTLD. The range for this NNTH suggests that as few as 63 belatacept patients to
as many as 532 CsA patients need to be treated for two years to expect at least one case
of PTLD. Similar results are shown when comparing belatacept LI alone vs. CsA and MI
alone vs. CsA.




                                                                                                              48
       Table 29: Estimated Number Needed to Harm based on Incidence of PTLD:
                            Belatacept Trials 008, 027 and 100
EBV+             Belatacept LI and MI    Belatacept LI    Belatacept MI   CsA
Incidence/PY            0.0035               0.0041           0.0028     0.0000
ARRI (vs CsA)           0.0069              0.0082            0.0056       NA
2-YR NNTH                144.7                122.0            178.8       NA
95% CI NNTH           -531.6, 63.7        -726.0, 56.3      -431.7, 74.1   NA
  PY=person-year
  AARI: annualized absolute risk increase (events belatacept/(person-yrs belatacept/2)-events CsA /(person-yrs CsA /2)
  (assumes constant risk)
  NNTH: 1/AARI

           7.3.4. PTLD rates among EBV+ Kidney Transplant Recipients Observed in
                the CsA Control Groups from Randomized Controlled Trials Previously
                Submitted to FDA

  The applicant has suggested that the incidence of PTLD observed among EBV+ kidney
  transplant recipients in the CsA treated groups of the belatacept clinical trials
  underestimates the true rate of PTLD in this population, as the rate is known to be non-
  zero. A review of previous randomized clinical trials of de novo kidney transplantation
  submitted to the FDA, however, confirms that the rate may be close to zero. We
  identified three randomized trials that collected information on patients for at least 12-
  months post-transplant (including baseline EBV status) and that included a control group
  treated with CsA, MMF, corticosteroids, and IL-2 antagonist induction treatment group.
  Among 534 EBV+ kidney transplant recipients who received a CsA-based regimen, zero
  cases of PTLD were reported. Therefore, the estimated overall incidence of PTLD among
  EBV+ patients from these three trials previously submitted to FDA by other applicants
  was: 0% (0/534), 95% exact CI (0, 0.0069). Refer to Appendix B for additional analyses
  including estimates of NNTH using these data.

           7.3.5. PTLD Rates Among EBV+ Kidney Transplant Recipients According
                to Registry and Claims Data

  The applicant has previously suggested that analyses of registry data and Centers for
  Medicare and Medicaid Services (CMS) claims data estimate incidences of PTLD in the
  general kidney transplant EBV+ population that are similar to that seen in the EBV+
  patients maintained on belatacept in the clinical trials. The applicant has made this
  argument based on the United Network for Organ Sharing (UNOS) registry data – also
  known as the Organ Procurement and Transplantation Network (OPTN) registry data –
  and the Collaborative Transplant Study (CTS) registry data as well as the CMS claims
  data. The FDA performed separate and independent analyses of the UNOS registry data
  to estimate PTLD incidence by EBV serostatus. Results (described below) from these
  analyses also support the findings from the randomized controlled trials. Specifically,
  these analyses found that the incidence of PTLD among EBV+ patients on a comparable
  calcineurin inhibitor based regimen was indeed lower than the incidence of PTLD among
  EBV+ patients maintained on a belatacept regimen.


                                                                                                                   49
The Scientific Registry of Transplant Recipients (SRTR) is a non-profit private group
which holds the OPTN contract awarded by the Health Resources and Services
Administration (HRSA) for the statistical analyses of UNOS/OPTN data. The FDA and
SRTR independently performed analyses of the UNOS/OPTN registry data and produced
remarkably consistent results: both groups concluded that, at 2 years of follow-up, the
rate of PTLD observed among belatacept treated EBV+ patients in the clinical trials was
seven times higher than that observed in calcineurin inhibitor treated EBV+ patients in
the UNOS registry given similar induction therapy. The applicant performed its own
analysis on the registry data and found a slightly higher incidence than FDA and SRTR.
However, even the results produced by the applicant yield a rate more than 3-fold lower
that that observed among belatacept treated EBV+ patients in the clinical trials (see
Table 30).

The Collaborative Transplant Study is a registry database based out of the University of
Heidelberg in Germany. The CTS is maintained by Dr. Gerhard Opelz. While the primary
datasets of the CTS were not available for analysis by the FDA, Dr Opelz published data
in 2009 which suggests that the incidence of non-Hodgkin’s lymphoma among EBV+
kidney transplant recipients at 2 years of follow-up is approximately one-third of that
observed among the belatacept-treated EBV+ patients in the clinical trials.14 In a personal
communication to the FDA, Dr Opelz confirmed that his data supported a much lower
incidence of PTLD among EBV+ kidney transplant recipients than was observed among
the belatacept patients (see Table 30).

 Table 30: 2-year PTLD incidence rates/100 Person-Years among EBV+ Belatacept
   Patients in IM103100, IM103008, and IM103027 compared to EBV+ Patients
  Maintained on CNI-based Regimens Followed in the UNOS and CTS Registries
                 Belatacept      BMS        FDA UNOS*       SRTR      Opelz
                (pooled MI     UNOS*         N = 77,203    UNOS*      (CTS
                  and LI)      N = not                    N = 83,929 Analysis)
                      N = 949           specified
                   2-year Rate        2-year Rate       2-year Rate          2-year            2-year
                   (95% Exact          (95% CI)         (95% Exact            Rate              Rate
                       CI)                                  CI)
EBV Positive           0.346               0.09             0.051             0.050             0.12
                   (0.112, 0.807)      (0.06, 0.12)     (0.026, 0.089)
* UNOS analyses consider all recipients with cyclosporine or tacrolimus maintenance regimens

Calculation of Number of Patients Needed to Harm using Registry Data
The estimated numbers needed to harm based on the comparisons between the
incidence/person-years of 0.0035 of the pooled belatacept (LI and MI) to the estimated
incidences among de novo kidney transplant recipients on a CsA or tacrolimus regimen
from the three analyses of the UNOS/OPTN data (BMS, FDA and SRTR) and to the
estimated incidence from the analyses of the CTS data by Dr. Opelz are shown below in
Table 31.




                                                                                                        50
Table 31: Estimated Number Needed to Harm (NNTH) based on registry data
 EBV+ Kidney Transplant   FDA UNOS        BMS     SRTR UNOS          Opelz
        Recipients                       UNOS                         (CTS
                                                                    Analysis)
2-Year Incidence/PY         0.00051      0.0009      0.00050         0.0012
2-YR NNTH *                  169.6        192.3       168.9           217.4
95% CI NNTH              (115.5, 319.2)    **     (110.9, 354.3)       **
*NNTH (number need to harm) is based on the difference in the 2-yr incidence per person-years between
the pooled belatacept (MI and LI) incidence of 0.0035 and the respective registry incidence
**Unable to estimate confidence intervals given level of information provided
PY=person-year
NNTH: 1/AARI where AARI=annualized absolute risk increase=events belatacept/(person-yrs belatacept/2)-
events CsA /(person-yrs CsA /2) (assumes constant risk)

Limitations of Registry and Claims Data
UNOS/OPTN
The UNOS/OPTN registry is composed of data generated by every kidney transplant
program located in the United States; reporting to UNOS is mandatory and monitored.
The registry includes data regarding every organ donation and transplant event in the
U.S. since October 1, 1987. A required transplant recipient follow-up form 6-month post
transplant is generated six months post-transplant and on the treatment anniversary for
every living organ recipient. Included on this form is a section for post transplant
malignancy with a specific query regarding “De Novo Lymphoproliferative disease and
Lymphoma.” While UNOS/OPTN data are regarded as the ‘gold-standard’ source of
information for U.S. transplant patients, the data from any registry may under-estimate
the true rates of reported events. Additionally, while reporting to UNOS/OPTN is a
requirement for participating transplant programs, the quality of reporting can vary. One
must also recognize that the patient population followed by UNOS/OPTN is not identical
to that studied in the belatacept kidney trials; comparisons across these populations, then,
must be interpreted with caution.

CTS
The Collaborative Transplant Study (CTS) is a registry based out of the University of
Heidelberg in Germany; its member centers are predominantly located in Europe. Unlike
UNOS/OPTN, reporting to CTS is voluntary. The CTS recently published an article
regarding the epidemiology of non-Hodgkin’s lymphoma in its registry population. 20
While neither the applicant nor FDA obtained primary data from CTS to perform
independent analyses, Dr. Gerhard Opelz – the director of CTS and the primary author of
the publication – provided his result for incidence at PTLD among EBV+ kidney
recipients at 2 years follow-up. While the peer-reviewed publication did not provide that
specific numeric result, it did contain a figure which is consistent with that personal
communication (see also Appendix C and Figure 7 below). The CTS analysis included
only those patients whose transplant centers confirmed that complete malignancy
information had been reported. Therefore, the CTS registry analysis has potential for
over-reporting as well as under-reporting. One must also recognize that the patient
20
  Opelz G, Daniel V, Naujokat C, et al. Epidemiology of Pretransplant EBV and CMV Serostatus in
Relation to Posttransplant Non-Hodgkin Lymphoma. Transplantation 2009; 88: 962-7.


                                                                                                     51
population followed by CTS is not identical to that studied in the belatacept kidney trials;
comparisons across these populations, then, must be interpreted with caution.

CMS Claims Data
The sponsor provided estimates of PTLD incidence based on an analysis of CMS
inpatient and outpatient claims data where cases were identified via specific ICD-9 codes
(i.e. 200.x, 202.x and 204.x) which have not been validated to correspond to PTLD. This
analysis was limited to only kidney transplant recipients who had Medicare as primary
payer at the time of transplantation and therefore represents only a subset of the overall
kidney transplant population. Additionally, this analysis identified cases of PTLD only
using ICD-9 codes without supporting clinical information to confirm the presence or
absence of PTLD. A recent SRTR analysis of CMS data using the same ICD-9 codes,
found that the most commonly reported ICD-9 code (among the searched list) was 202.80
and that the number of events reported using this code from 2001 to 2008 increased in an
unusual fashion: the events reported in 2001-2008 were 67, 130, 145, 184, 241, 295, 289
and 322 respectively (each of these numbers represents events reported that calendar
year, not cumulatively to that point). The reason for the geometric progression of events
reported using this code is unclear but is not thought to correlate with a true increase in
the incidence of PTLD during the study time. The most likely explanation would appear
to relate to coding errors and differences in reporting practices across medical centers.
While FDA obtained the CMS datasets through USRDS to evaluate the applicant’s claim,
the FDA ultimately determined that the approach did not add value to its understanding
of PTLD in EBV+ patients and therefore has elected not to present any estimation of
NNTH based on the CMS data.

Given these limitations outlined above, it is inappropriate to rely on results from any one
analyses or comparison of registry data. Instead, results across all sources of data should
be considered in order from least potentially biased (i.e. comparing rates of PTLD
between treatment groups within the randomized clinical trials) to the more potentially
biased (i.e. comparing incidence from belatacept to outside sources such as other
randomized trials and registries).

       7.3.6. PTLD in the Belatacept Liver Transplant Program

Study IM103045 is a Phase 2 de novo liver transplant study of about 250 EBV+ patients;
146 patients have been enrolled into one of three belatacept treatment groups, while
around 100 patients have been enrolled into one of two control groups. To date, two
patients from the belatacept groups have developed PTLD within the allograft compared
to zero patients in the control groups; there have been no cases of CNS PTLD in any of
the patients.

The epidemiology of PTLD among liver transplant patients differs from kidney transplant
patients. PTLD is a more common event in liver transplantation. One explanation for this
may be that EBV+ serostatus does not appear to confer the protection against PTLD in
the liver transplant context that it affords in the kidney transplant context: while
incidence of PTLD in EBV- liver transplant recipients approximates that of EBV- kidney
transplant recipients, the incidence of PTLD in EBV+ liver transplant recipients is much


                                                                                         52
higher than that of EBV- kidney transplant recipients and is, in fact, similar to that of
EBV- liver and kidney recipients (see Error! Reference source not found.). 21

The observations in IM103045 and the known epidemiology of PTLD in liver transplant
patients may indicate a need for caution in the liver transplant program and measures to
mitigate the possibility of the off-label use of belatacept among liver transplant patients.

     Figure 7: Cumulative incidence of non-Hodgkin’s Lymphoma according to EBV
        status in kidney, heart, and liver transplant recipients in the Collaborative
                              Transplant Study registry data


              Figure taken from Opelz, G et al. Transplantation 2009; 88: 962-7




      7.4. Progressive Multifocal Leukoencephalopathy (PML)

A patient from the belatacept MI group in IM103027 was diagnosed with PML in Month
23; that patient died shortly after Month 24 (see summary of death narrative below). The
incidence of progressive multifocal leukoencephalopathy (PML) in the general kidney
transplant population has been estimated at 14 cases per 100,000 patient-years. 22 While
the observation of even that single case of PML in IM103027, then, is remarkable, it is
difficult to extrapolate too much from a single event. In addition, however, a second case
of PML was diagnosed in the belatacept liver transplant program (in Study IM103045) in
October of 2009. At the time of diagnosis, the patient was confused, had memory recall
problems, had significant motor impairments and gait abnormalities. Further follow up
has not yet been made available. Two occurrences of PML in the context of a drug
development program constitute a safety signal. While the impact of an increased risk of
PML on the overall risk-benefit profile of belatacept is potentially small, given that the
outcome remains rare, the signal is important because it supports other data suggesting
that belatacept may have specific effects on CNS immunity (e.g., the increased incidence


21
   Opelz G, Daniel V, Naujokat C, et al. Epidemiology of Pretransplant EBV and CMV Serostatus in
Relation to Posttransplant Non-Hodgkin Lymphoma. Transplantation 2009; 88: 962-7.
22
   Neff RT, Hurst FP, Falta EM, et al. Progressive Multifocal Leukoencephalopathy and Use of
Mycophenolate Mofetil after Kidney Transplantation. Transplantation 2008; 86(10): 1476-8.



                                                                                                   53
of CNS PTLD, the increased rates of CNS infections seen in the belatacept MI treatment
group).

       Death Narrative for IM103027-15-10045: 63 year old white female treated with
       belatacept MI regimen presented on Day 711 with homonymous hemianopia and
       recent history of falls. On Day 725, patient developed worsened vision and short-
       term memory loss. On that same day, MRI revealed a lesion in the right occipital
       lobe extending into the posterior right temporal lobe consistent with PML. She
       was hospitalized on Day 730 for progressive neurological deterioriation. Lumbar
       puncture and PCR confirmed the presence of JC virus. CMV viremia was also
       detected. On Day 742, patient was discharged to home hospice care with PML
       with symptoms of cortical blindness, memory disturbance, gait abnormalities, and
       headaches. The patient died on Day 754.

   7.5. Non-Fatal Serious Adverse Events (SAEs)

An analysis of serious adverse events (SAEs), defined as events that were either life-
threatening, required hospitalization, or prolonged hospitalization, reported in Studies
008, 027, and 100 by Month 12 was performed. The analysis, shown in Table 32,
suggested that SAEs were distributed evenly across treatment groups, though it did
provide some additional support that serious AR episodes were more common among
patients maintained on belatacept (please see section 6.5 for a discussion of safety issues
associated with the higher rate and grades of AR observed among belatacept patients).
The major adverse events associated with MPA exposure include leucopenia and
diarrhea. Despite the 25% to 40% higher MPA exposure among the belatacept treated
patients (see Section 4.3); the analysis did not detect any difference in serious adverse
events due to leucopenia.




                                                                                         54
              Table 32: Serious Adverse Events in at least 2% of Patients at Month 12

                                                    Pooled Trials (008, 027, and 100)

System Organ Class                  Belatacept MI             Belatacept LI               CsA
        Preferred Term                 N=477                     N=472                   N=476

Total Patients with SAE               289 (60.6)               263 (55.7)               298 (62.6)

Infections and Infestations           128 (26.8)               108 (22.9)               129 (27.1)
    Cytomegalovirus Infection          26 (6.5)                 28 (5.9)                 25 (5.5)
     Urinary Tract Infection           25 (5.2)                 24 (5.1)                 30 (6.3)
           Pneumonia                    9 (1.9)                  6 (1.3)                  6 (1.3)
         Pyelonephritis                10 (2.1)                  6 (1.3)                  6 (1.3)
Renal and Urinary Disorders           68 (14.3)                 64 (13.6)               83 (17.4)
      Renal Failure Acute              6 (1.3)                   7 (1.5)                 13 (2.7)
       Ureteric Stenosis               5 (1.0)                   7 (1.5)                 10 (2.1)
Injury, Poisoning, and
Procedural Complications              59 (12.4)                 59 (12.5)               80 (16.8)
        Graft Dysfunction             15 (3.1)                   13 (2.6)                21 (4.4)
   Therapeutic Agent Toxicity             0                         0                    10 (2.1)
Vascular Disorders                     24 (5.0)                 35 (7.4)                 47 (9.9)
         Lymphocele                     5 (1.0)                  9 (1.9)                 17 (3.6)
General Disorders and
Administration Site Conditions         35 (7.3)                 27 (5.7)                 34 (7.1)
            Pyrexia                    24 (5.0)                 22 (4.7)                 22 (4.6)
Investigations                         28 (5.9)                 24 (5.1)                 36 (7.8)
   Blood Creatinine Increased          19 (4.0)                 19 (4.0)                 28 (5.9)
Blood and Lymphatic System
Disorders                              28 (5.9)                 18 (3.8)                 24 (5.0)
           Leucopenia                  12 (2.5)                  3 (0.6)                 10 (2.1)
Immune System Disorders                19 (4.0)                 21 (4.4)                 10 (2.1)
      Transplant Rejection             16 (3.4)                 18 (3.8)                  8 (1.7)

          7.6. Adverse Events Leading to Treatment Discontinuation

      The proportion of subjects with AEs leading to study drug discontinuation up
      to Month 12 in Studies 008, 027, and 100 was highest in the CsA group (14%) and
      comparable in the belatacept LI (11%) and MI groups (10%), as shown in Table 33.

      As previously discussed in Section 5, numerically more patients receiving belatacept than
      CsA discontinued treatment in the Phase 3 trials due to efficacy failure (primarily AR).
      Given that efficacy failures received separate consideration already, such events were not
      included in the analysis of adverse events leading to treatment discontinuation. In
      contrast, numerically more patients receiving CsA than belatacept discontinued treatment
      in those trials due to adverse events. An inspection of the reported SOC and PTs
      associated with the responsible adverse events suggested that most classes of events were
      distributed proportionately among treatment groups. In each study, however, four CsA


                                                                                                55
patients discontinued due to hirsutism (compared to zero belatacept patients). In
IM103027 only, more belatacept patients discontinued due to renal vein thrombosis and
graft thrombosis (see also Section 7.9.3).

Table 33: Adverse Events Leading to Treatment Discontinuation in ≥1% of Patients
Up to Month 12 (Pooled Trials 008, 027, and 100)

System Organ Class                      Belatacept MI           Belatacept LI     Cyclosporine
Preferred Term                              N=477                  N=472            N=476
                                            n (%)                   n (%)            n (%)
Total Number of Patients                   47 (9.9)               54 (11.4)        66 (13.9)
with AE Leading to
Treatment Discontinuation
Injury, Poisoning and                       8 (1.7)                 14 (3.0)        22 (4.6)
Procedural Complications
                   Graft Loss               3 (0.6)                     2 (0.4)      4 (0.8)
            Chronic Allograft                  0                        2 (0.4)      5 (1.1)
                 Nephropathy
             Complications of                  0                        6 (1.3)      1 (0.2)
         Transplanted Kidney
Renal and Urinary                           8 (1.7)                     9 (1.9)     11 (2.3)
Disorders
       Renal Vein Thrombosis                 5 (1.0)                 2 (0.4)            0
Immune System Disorders                     10 (2.1)                11 92.3)         5 (1.1)
         Transplant Rejection               9 (1.9)                  9 (1.9)         3 (0.6)
Infections and Infestations                 8 91.7)                 12 92.5)         6 (1.3)
    Cytomegalovirus Infection                2 (0.4)                 5 (1.1)            0
Skin and Subcutaneous                           0                       0           10 (2.10
Tissue Disorders
                    Hirsutism                  0                          0          8 (1.7)
Source: Table 2.1.4 in the Summary of Clinical Safety, BLA submission

    7.7. Common Adverse Events

Other than the specific issues addressed above, the adverse event profile of the belatacept
regimen and the CsA regimen were similar as observed by the AEs reported up to Month
12 in Studies 008, 027, and 100. That observation is perhaps somewhat surprising as one
might have anticipated more of an imbalance reflecting known class effects of CsA and
mycophenolic acid (MPA). Known adverse events associated with CsA include
hyperkalemia, hypertension, dyslipidemias, hyperglycemia, and tremor. One would
expect the control arm to have experienced more of those adverse events. Known adverse
events associated with MPA include GI toxicities (diarrhea, nausea, abdominal pain,
vomiting) and bone marrow suppression (anemia and leucopenia). While the observed
dosages of MMF were similar across the belatacept groups and the CsA groups, a 25 to
40% greater MPA exposure was achieved among the belatacept groups (due to the impact
of CsA on the enterohepatic recirculation of MPA; see Section 4.3) – suggesting that the


                                                                                               56
     belatacept arms might have been expected to have experienced more adverse events
     related to MPA. Therefore, in addition to carefully examining AEs associated with
     belatacept, particular attention was also therefore afforded to reporting of those AEs
     known to be associated with CsA and MPA. An analysis of these selected AEs, however,
     suggests that the true differences across groups were mild (See Table 34): CsA patients
     only appeared to experience dyslipidemia (but not hypercholesterolemia) and tremor
     more frequently.

       Table 34: Selected Adverse Events Associated with Class Effects Up to Month 12

                                                 IM103008                               IM103027

                                      MI           LI           CsA          MI           LI           CsA
SOC/PT                               N=219        N=226        N=221        N=184        N=175        N=184
Metabolism and Nutritional
Disorders                           155 (70.8)   159 (70.4)   161 (72.9)   129 (70.1)   127 (72.6)   135 (73.4)
        Hypokalemia                  39 (17.8)    37 (16.4)    17 (7.7)    31 (16.8)    31 (17.7)    32 (17.4)
        Hyperkalemia                 14 (6.4)     38 (16.8)    32 (14.5)    38 (20.7)   42 (24.0)    35 (19.0)
        Dyslipidemia                 38 (17.4)    39 (17.3)    55 (24.9)    19 (10.3)    17 (9.7)    30 (16.3)
     Hypercholesterolemia            26 (11.9)    19 (8.4)     19 (8.6)     16 (8.7)     16 (9.1)     18 (9.8)
       Hyperglycemia                 30 (13.7)    32 (14.2)    30 (13.6)    32 (17.4)   25 (14.3)    28 (15.2)
Infections and Infestations         152 (69.4)   158 (69.9)   157 (71.0)   130 (70.7)   129 (73.7)   142 (77.2)
    Urinary tract infections         54 (24.7)    63 (27.9)    50 (22.6)    55 (29.9)    57 (32.6)    62 (33.7)
Upper respiratory tract infection    24 (11.0)    22 (9.7)     26 (11.8)    11 (6.0)     11 (6.3)     14 (7.6)
   Cytomegalovirus infection         13 (5.9)     17 (7.5)     19 (8.6)     21 (11.4)    24 (13.7)    24 (13.0)
Gastrointestinal System
Disorders                           139 (63.5)   157 (69.5)   154 (69.7)   122 (66.3)   130 (74.3)   136 (73.9)
         Constipation                59 (26.9)    72 (31.9)    60 (27.1)    52 (28.3)    57 (32.6)    73 (39.7)
           Diarrhea                  58 (26.5)    56 (24.8)    56 (25.3)    54 (29.3)    58 (33.1)   47 (25.5)
            Nausea                   47 (21.5)    48 (21.2)    56 (25.3)    42 (22.8)    37 (21.1)   41 (22.3)
           Vomiting                  30 (13.7)   34 (15.0)    35 (15.8)    25 (13.6)    38 (21.7)    31 (16.8)
       Abdominal Pain                24 (11.0)    27 (11.9)    25 (11.3)    19 (10.3)   28 (16.0)    25 (13.6)
Blood and Lymphatic Systems
Disorders                           103 (47.0)   124 (54.9)   101 (45.7)   114 (62.0)   103 (58.9)   122 (66.3)
            Anemia                   68 (31.1)    82 (36.3)   64 (29.0)    87 (47.3)    85 (48.6)    92 (50.0)
          Leucopenia                 34 (15.5)    36 (15.9)    34 (15.4)    44 (23.9)   30 (17.1)    49 (26.6)
Nervous System Disorders            66 (30.1)    75 (33.2)    75 (33.9)    53 (28.8)    55 (31.4)    66 (35.9)
           Headache                  39 (17.8)    37 (16.4)    21 (9.5)     21 (11.4)    22 (12.6)    21 (16.8)
            Tremor                    9 (4.1)     12 (5.3)     35 (15.8)    11 (6.0)     14 (8.0)    23 (12.5)

         7.8. Acute Rejection

     As discussed in Section 6.2.3, AR at Month 12 in the Phase 3 trials occurred more
     frequently among recipients of belatacept than recipients of CsA in IM103008: while the
     belatacept LI regimen met a 20% NI margin for the endpoint of AR (implying that
     belatacept has some efficacy for the indication of prophylaxis of AR), it appeared inferior
     to the CsA regimen in that trial for the endpoint of AR. In addition, the numbers of Banff




                                                                                                       57
grade II and III rejections were numerically greater among recipients of belatacept than
among recipients of CsA, especially in IM103008 (see Table 35).

         Table 35: Acute Rejection (as Defined by the Applicant) at 12 Months
Study                                           Belatacept MI         Belatacept LI               CsA
008      Acute Rejection                         49/219 (22.4)         39/226 (17.3)          16/221 (7.2)
          Mild IA                                      7                     4                     3
          Mild IB                                      3                     8                     5
          Moderate IIA                                16                    16                     6
          Moderate IIB                                20                    10                     2
          Severe III                                   2                     1                     0
         Difference from CsA (97.3% CI)         15.2 (7.4, 23.0)      10.1 (2.9, 17.3)

027      Acute Rejection                         33/184 (17.9)         31/175 (17.7)          26/184 (14.1)
          Mild IA                                       0                     4                     2
          Mild IB                                       7                     2                     2
          Moderate IIA                                 11                    17                    17
          Moderate IIB                                 15                     8                     5
          Severe III                                    0                     0                     0
         Difference from CsA (97.3% CI)         3.8 (-5.2, 12.8)      3.6 (-5.5, 12.7)

Prevention of AR episodes has represented a basis for drug approval in transplantation
over the past three decades. Acute rejection has been viewed as an important clinical
event largely because data suggesting that it is associated with decreased graft life. 23,24
Some epidemiologic data suggest, however, that decreases in early AR rates have not
been associated with improved long-term outcomes. 25,26 That literature may reflect,
however, decreases in low grade rejections (which are typically the drivers of overall
rejection rates).

The trend towards higher grades of rejection among belatacept patients, then, is
particularly concerning. The literature dissecting the clinical significance of higher grade
rejections, suggests that high grade Banff lesions are associated with worse outcomes,
including lower GFR and graft loss. 27,28




23
   Matas AJ, Humar A, Payne WD et al. Decreased acute rejection in kidney transplant recipients is
associated with decreased chronic rejection. Ann Surg 1999; 230: 493-8.
24
   Hariharan S, Johnson CP, Bresnahan BA, et al. Improved graft survival after renal transplantation in the
United States, 1988 to 1996. N Engl J Med 2000; 342: 605-12.
25
   Meier-Kriesche H-U, Schold JD, Kaplan B. Long-term renal allograft survival: Have we made
significant progress or is it time to rethink our analytic and therapeutic strategies? Am J Transplant 2004; 4:
1289-95.
26
   Meier-Kriesche H-U, Schold JD, Srinivas TR, et al. Lack of improvement in renal allograft survival
despite a marked decrease in acute rejection rates over the most recent era. Am J Transplant 2004; 378-83.
27
   Mengel M, Sis B, Halloran PF. SWOT analysis of Banff: strengths, weaknesses, opportunities, and
threats of the international Banff consensus process and classification system for renal allograft pathology.
Am J Transplant 2007; 7: 2221-6.
28
   Mueller A, Schneulle P, Waldherr R, et al. Impact of the Banff ’97 classification for histological
diagnosis of rejection on clinical outcome and renal function parameters after kidney transplantation.
Transplantation 2000; 69:1123-27.


                                                                                                             58
Beyond the implications to graft life, AR episodes have clinical significance due to the
morbidity associated with their reversal. High grade rejection episodes, in particular, may
be associated with significant morbidity as they often require treatment with lymphocyte
depleting therapies (e.g., thymoglobulin, rituximab). Such agents have been associated
with higher rates of several serious adverse events in transplant patients including PTLD,
CMV disease, and BK virus associated nephropathy. In keeping with the higher rates of
BPAR and higher grades of rejection observed among the belatacept patients, lymphocyte
depleting therapy (LDT) was used more often in the belatacept groups than the CsA
groups for treatment of AR in the Phase 3 trials (see Table 36). More patients in the CsA
groups received exposure to lymphocyte depleting agents in the context of anticipated
delayed graft function (DGF).

        Table 36: Use of Lymphocyte Depleting Therapy Through Month 24
                                     IM103008                           IM103027
                          Belatacept Belatacept    CsA       Belatacept Belatacept     CsA
                              MI         LI                     MI           LI
                           (n=219)     (n=226)    (n=221)     (n=184)     (n=175)    (n=184)
  For Treatment of         27 (12.3)   22 (9.7)    4 (1.8)    16 (8.7)     8 (4.6)    6 (3.3)
  AR
   Initial Treatment of    13 (5.9)   10 (4.4)    3 (1.4)     14 (7.6)    7 (4.0)     4 (2.2)
            AR
  Steroid Resistant AR     14 (6.4)    12 (5.3)   1 (0.5)     2 (1.1)     1 (0.6)     2 (1.1)
  For Anticipated          1 (0.5)     1 (0.4)    8 (3.6)        0           0       27 (14.7)
  DGF

Differences in the rates of PTLD observed between the belatacept and CsA groups are
discussed in Section 7.3; events of PTLD for all groups have been closely monitored and
carefully documented with complete data for all patients at 24 months of follow-up and
with partial data for many patients beyond 24 months. Despite the greater overall
exposure of belatacept patients to LDT agents in IM103008, rates of CMV disease and
BK virus associated nephropathy were similar across all treatment groups in this study
(and also in IM103027 and IM103100).

   7.9. Glomerular Filtration Rate (GFR)

Mean calculated GFRs were substantially higher throughout the 24 months of the Phase 3
trials among the belatacept treatment groups compared to the CsA treatment groups (see
Figure 8 and Figure 9). Due to the vasoconstrictive effects of CsA on the afferent renal
artery, it remains unclear whether the superior GFRs observed in belatacept patients
reflect a healthier transplant kidney, more favorable renal hemodynamics, or both.




                                                                                                 59
Figure 8: Mean Calculated GFR (mL/min) in Study 008




Error bars represent 95% confidence interval of the mean




                                                           60
                 Figure 9: Mean Calculated GFR (mL/min) IM103027




               Error bars represent 95% confidence interval of the mean



While a difference in measured GFRs in favor of patients in both belatacept groups
relative to the CsA control group is clearly evident in both Phase 3 trials, it is less
obvious that any difference exists between the groups with respect to their change in GFR
over time. A pre-specified calculation for slope of GFR based on a Month 3 starting time
numerically favors the belatacept groups (see Table 37), but visual inspection of Figure
8 and Figure 9 confirms what the confidence intervals pertaining to that calculation
suggest: the existence of a difference with regards to GFR progression over time between
the belatacept and CsA groups is uncertain.

            Table 37: Slope for Calculated GFR from Month 3 to 24
Study                               Belatacept   Belatacept LI         CsA
                                        MI
008   Slope* (standard error)       0.96 (0.71)    1.19 (0.70)     -1.73 (0.79)
      95% Confidence Interval      (-0.43, 2.36)  (-0.18, 2.57)   (-3.16, -0.30)
027   Slope* (standard error)      -0.74 (0.78)   -0.93 (0.77)     -1.88 (0.78)
      95% Confidence Interval      (-2.28, 0.79)  (-2.44, 0.58)   (-3.40, -0.34)
*mL/min/1.73 m2/year

Higher GFRs have been correlated with improved patient outcomes, both in the general
chronic kidney disease population and also specifically in the renal transplant patient
population. A recent multivariate analysis of Australia and New Zealand Dialysis and
Transplant (ANZDATA) registry data, for instance, showed that cardiovascular death



                                                                                      61
rates were significantly lower among patients with a GFR greater than 48 mL/min at one
month compared to those with GFR less than 38 mL/min at one month (risk ratio, 0.66;
95% confidence interval, 0.45-0.95). 29 A Cox regression analysis of OPTN registry data
showed that cardiovascular death rates were significantly higher among renal transplant
patients whose serum creatinine at 1 year was higher than 1.5 mg/dL compared to those
whose serum creatinine at 1 year was less than1.5 mg/dL. 30 As cardiovascular deaths are
the leading cause of deaths in the renal transplant population, reductions in
cardiovascular deaths would impact significantly on overall death rates. Higher GFRs at
one year have also been correlated with improved graft life: an analysis of OPTN registry
data also showed that serum creatinine at 1 year predicted long-term graft survival. 31
Whether the higher GFRs observed among the belatacept patients in the Phase 3 trials
reflected structural or hemodynamic differences, then, it may be reasonable to infer that
the improved GFRs afforded a clinical benefit.

     7.10.       Cardiovascular Risk Factors

Cardiovascular disease is the leading cause of mortality among renal transplant
patients. 32 Independent risk factors for cardiovascular death in the renal transplant patient
population include diabetes, hypertension, and dyslipidemias. 33,34 Poor renal function
represents another important cardiovascular risk factor (see Section 7.9). 35 Hypertension,
dyslipidemias, and new onset diabetes after transplantation (NODAT) are all class effects
of CNIs. The applicant has suggested that, as a biologic agent with a highly specific
target, belatacept may be less likely to induce similar effects and prospectively studied
endpoints related to hypertension, dyslipidemia, and diabetes. The applicant considered
the endpoints to represent secondary efficacy endpoints. These endpoints, however, do
not directly support the efficacy of belatacept for the proposed indications of either
prophylaxis of AR or preservation of a functioning allograft. While they may not inform
decisions about the effects of a drug, however, differences in hypertension, dyslipidemia,
and diabetes across approaches to kidney transplantation constitute an important safety
consideration. For that reason, the Division examined those prospectively measured
endpoints from the point-of-view of clinical safety.




29
   Pilmore H, Dent H, Chang S. Reduction in Cardiovascular Death after Kidney Transplantation.
Transplantation 2010; epublished ahead of print.
30
   Meier-Kriesche HU, Baliga, Kaplan B. Decreased renal function is a strong risk factor for cardiovascular
death after renal transplantation. Transplantation 2003; 75: 1291.
31
   Hariharan S, McBride MA, Cherikh WS, et al. Post-transplant renal function in the first year predicts
long-term kidney transplant survival. Kidney Int 2002; 62: 311-8.
32
   Aakhus S, Dahl K, Wideroe TE. Cardiovascular morbidity and risk factors in renal transplant patients.
Nephrol Dial Transplant 1999; 14: 648-54.
33
   Aakhus S, Dahl K, Wideroe TE. Cardiovascular morbidity and risk factors in renal transplant patients.
Nephrol Dial Transplant 1999; 14: 648-54.
34
   Kasiske BL, Chakkera HA, Roel J. Explained and unexplained ischemic heart disease risk after renal
transplantation. J Am Soc Nephrol 2000; 11: 1735-43.
35
   Pilmore H, Dent H, Chang S. Reduction in Cardiovascular Death after Kidney Transplantation.
Transplantation 2010; epublished ahead of print.


                                                                                                        62
Hypertension
Mean systolic and diastolic blood pressures were significantly lower at Month 12 for
patients in both belatacept groups relative to the CsA control group in both Phase 3 trials
(see Table 38).

                      Table 38: Mean Blood Pressure at Month 12
Study                                          Mean (Standard Deviation)
                                                  Number in Analysis
                                    Belatacept MI    Belatacept LI        CsA
008      Systolic Blood Pressure    132.7 (16.2)*     131.4 (16.5)*   138.7 (20.0)
                                        n=191             n=193          n=187
         Diastolic Blood Pressure    79.3 (11.5)*      78.7 (10.9)*    81.9 (11.1)
                                        n=191             n=193          n=187
027      Systolic Blood Pressure    141.4 (21.3)*     140.9 (21.1)*   149.5 (19.8)
                                        n=137             n=140          n=124
         Diastolic Blood Pressure    77.8 (13.8)*      78.3 (10.6)*    81.8 (11.7)
                                        n=137             n=140          n=124
*comparison of mean value of belatacept regimen to CsA p-value<0.027

The benefits of blood pressure control are well-established in the general population but
have been shown to be particularly important in the renal transplant population. For
instance, a retrospective analysis of the Collaborative Transplant Study registry data
showed that renal transplant recipients with a systolic blood pressure at 1 year less than
140 mmHg had significantly reduced rates of graft loss and cardiovascular death
compared with those with a systolic blood pressure at 1 year greater than 140 mmHg. 36
While it is difficult to quantify the clinical benefit in a renal transplant patient population
of the differences in systolic and diastolic blood pressures observed, it is reasonable to
infer that some benefit should accrue for differences of the magnitude measured in the
trial.

Dyslipidemia
Mean non-HDL cholesterol and mean triglyceride levels at Month 12 were significantly
lower in patients in both belatacept groups relative to the CsA control group in both
Phase 3 trials (see Table 39). The difference in non-HDL between the two groups
appeared to be driven by the contribution of triglycerides, as LDL levels at Month 12
were more similar across groups.




36
  Opelz G, Dohler B. Improved long-term outcomes after renal transplantation associated with blood
pressure control. Am J Transplant 2005; 5: 2725-2731.


                                                                                                     63
                      Table 39: Serum Lipids at Month 12
Study                               Belatacept   Belatacept LI                                    CsA
                                        MI
008   Non-HDL cholesterol          131.7 (36.8)* 131.5 (38.2)*                              144.1 (47.3)
                                      n=192         n=195                                      n=189
      LDL cholesterol              100.8 (29.5)   102.1 (33.4)                              107.3 (39.6)
                                      n=183         n=186                                      n=187
      Triglycerides                155.0 (85.1)* 149.4 (87.3)*                              184.6 (106.4)
                                      n=183         n=186                                      n=187
027   Non-HDL cholesterol          135.9 (46.3)* 134.4 (41.1)*                              153.0 (46.3)
                                      n=144         n=142                                      n=133
      LDL cholesterol              105.5 (38.8)   102.5 (37.2)                              109.8 (40.9)
                                      n=133         n=135                                      n=124
      Triglycerides                    169.6     152.9 (67.2)*                              211.3 (120.6)
                                     (120.8)*       n=135                                      n=124
                                      n=134
Mean (standard deviation)
Number in analysis
*comparison of mean value of belatacept regimen to CsA p-value<0.027

Lipid panels in ESRD patients awaiting transplantation are poorly understood: while
elevated LDL and total cholesterol levels are associated with cardiovascular disease in
the general population, the relationship of LDL cholesterol and cardiovascular death is
less clear in the dialysis population. 37,38 Low LDL levels in patients on dialysis are
common, but the significance of such LDL levels in that population is not entirely clear:
generalized inflammation and poor nutrition may drive down LDL levels. Serum LDL
and triglycerides typically increase after transplantation, reflecting both physiologic
changes after transplantation but also adverse effects of immunosuppressive drugs. 39
Treatment of dyslipidemias in the general CKD and ESRD population remain
controversial. After renal transplantation, however, management of lipids appears to
improve cardiovascular outcomes. 40,41 In the context of the controversies regarding
interpretation of lipid panels in these populations, it is difficult to quantify a clinical
benefit but the differences in effects on triglycerides appeared favorable to belatacept.




37
   Cheung AL. Is lipid control necessary in hemodialysis patients? Clin J Am Soc Nephrol 2009; S1: S95-
101.
38
   Kanbay M, Turgut F, Covic A, et al. Statin treatment for dyslipidemia in chronic kidney disease and
renal transplantation: a review of the evidence. J Nephrol 2009; 22(5): 598-609.
39
   Moore R, Hernandez D, Valantine H. Calcineurin inhibitors and post-transplant hyperlipidaemias. Drug
Saf 2001; 24 (10): 755-66.
40
   Holdaas H, Fellström B, Jardine AG, et al. Assessment of Lescol in Renal Transplantation (ALERT)
Study Investigators. Effect of fluvastatin on cardiac outcomes in renal transplant recipients: a multicentre,
randomized, placebo-controlled trial. Lancet. 2003; 361: 2024-31
41
   Cosio FG, Pesavento TE, Kim S, et al. Patient survival after renal transplantation: Part IV: impact of
post-transplant diabetes. Kidney Int 2002; 62: 1440-6.



                                                                                                           64
New Onset Diabetes After Transplantation (NODAT)

The trials defined NODAT as the use of an antidiabetic agent for more than 30 days or at
least two fasting plasma glucose values greater than 126 mg/dL in a subject who was not
diabetic at entry into the study. Across all treatment groups in the Phase 3 trials,
relatively few patients developed NODAT by Month 12. A trend towards fewer cases of
NODAT was apparent in both trials (see Table 40).

                                 Table 40: NODAT at Month 12
     Study           Belatacept MI           Belatacept LI                               CsA
      008             11/156 (7.1)             7/168 (4.2)                           16/162 (9.9)
      027              3/133 (2.3)            7/136 (5.1)                            11/118 (9.3)
     Pooled           14/289 (4.8)            14/304 (4.6)                           27/280 (9.6)
     Denominator is number of subjects with out diabetes at transplant

Although the incidence of NODAT in the patients in the belatacept groups was
approximately 50% less than the incidence of NODAT in the patients in the CsA groups,
the absolute reduction in the incidence was only about 5%. The impact of the measured
50% reduction of NODAT in the Phase 3 trials, then, may have been limited by the
overall low incidence of NODAT in the trials. The prevalence of NODAT in the general
US transplant population has been estimated at 15-20%, somewhat higher than observed
in the context of the belatacept Phase 3 trials.42,43,44 It is unclear whether the low
incidence of NODAT across the treatment groups reflected the demographics of the
patients studied or the use of a CsA, rather than tacrolimus, control arm. As the incidence
of cardiovascular death has been estimated to be 3-fold higher among transplant patients
with diabetes compared to transplant patients without diabetes 45 , the low rates of
NODAT observed in the context of belatacept would be clinically significant if
reproducible in the general US population.

       7.11.        Adverse Events of Special Interest

           7.11.1. Infections

Overall rates of infections at 12 months appeared similar across treatment arms of the
three clinical trials in an analysis of reported adverse events (see Table 41). An
imbalance, however, existed with regards to herpes infections: belatacept patients had
reported significantly more episodes of herpes infections at the time of the database lock.



42
   Bodziak KA, Hricik DE. New-onset diabetes mellitus after solid organ transplantation. Transplant Int
2009; 22: 519-30.
43
   Woodward RS, Schnitzler MA, Baty J, et al. Incidence and cost of new onset diabetes mellitus among
US wait-listed and transplanted renal allograft recipients. Am J Transplant 2003; 3: 590-8.
44
   Moore R, Ravindran V, Baboolal K. The burden of new-onset diabetes mellitus after transplantation.
Clin Transplant 2006; 20: 755-61.
45
   Lindholm A, Albrechtsen D, Frodin L, et al. Ischemic heart disease – major cause of death and graft loss
after renal transplantation in Scandinavia. Transplantation 1995; 60: 451.


                                                                                                         65
Table 41: Infections Reported as Adverse Events up to 12 Months in Trials 008, 027,
                                     and 100
                                   Through Month 12: n, (%)            Through Database Lock: n, (%)
                              Belatacept Belatacept      CsA         Belatacept Belatacept     CsA
                                 MI          LI                         MI          LI
                               (N=477)     (N=472)     (N=476)        (N=477)    (N=472)     (N=476)
 Infections reported as
                              128 (26.8)   109 (23.1)   130 (27.3)   169 (35.4)   142 (30.1)   166 (33.6)
 SAEs
 AEs in SOC of Infections
                              337 (70.7)   339 (71.8)   351 (73.7)   378 (79.2)   376 (79.7)   381 (80.0)
 and Infestations
 PTs Related to Viral
 Infections
            CMV               53 (11.1)    56 (11.9)    65 (13.7)    65 (13.6)    62 (13.1)    69 (14.5)
      Polyoma Virus            23 (4.8)      31 (6.6)    23 (4.8)     30 (6.3)     16 (3.4)     26 (5.5)
         Herpes                38 (8.0)      31 (6.6)    29 (6.1)    67 (14.0)    57 (12.1)     41 (8.6)
 PTs Related to Fungal
                            66 (13.8)     52 (11.0)   72 (15.1)     99 (20.8)   78 (16.5)      95 (20.0)
 Infections
SOC = System Organ Class
PT = Preferred Term
Source: Adapted from Table 2.1.5.3.A in Summary of Clinical Safety BLA submission

            7.11.1.1.       CNS infections

The high incidence of CNS PTLD and the single case of PML observed in the clinical
trials prompted a review of all CNS infections observed in these trials in order to
elucidate possible interactions between belatacept and CNS immunity. Similar rates of
CNS infections were seen among the belatacept LI patients and the CsA patients (one
each across the three trials through the time of database lock). Among the belatacept MI
patients, however, seven CNS infections were observed across the three trials through the
time of database lock (see Table 42).




                                                                                                   66
 Table 42: Cases of CNS Infections in Trials 008, 027, and 100 up to Database Lock
Study               Belatacept MI                   Belatacept LI                      CsA
100                  Herpes Zoster
                     Neurological                          --                              --


008             Cryptococcal Meningitis                                       Cryptococcal Meningitis

               Cryptococcal Meningitis/                    --
                 Chagas Encephalitis

                    West Nile Virus

                Cryptococcal Meningitis

027              Cerebral Aspergillosis        Cryptococcus Meningitis
                                                                                           --
                         PML

             7.11.1.2.   Tuberculosis

An imbalance in cases of tuberculosis was observed across treatment arms in the Phase 3
trials through the time of the database lock (see Table 43). Most of the cases of
tuberculosis involved extra-pulmonary sites: among the belatacept patients, tuberculosis
presented as disseminated disease twice, laryngeal tuberculosis once, tuberculosis
synovitis once, in a cervical ganglion once, and as gastrointestinal tuberculosis once. The
lone case of tuberculosis among the CsA patients presented within the intra-thoracic
lymph nodes. The vast majority of the patients lived in locations where tuberculosis is
endemic.

        Table 43: Cases of Tuberculosis in the Phase 3 Trials at Database Lock
                                      IM103008                               IM103027
                         Belatacept    Belatacept    CsA        Belatacept    Belatacept        CsA
                            MI             LI                      MI             LI
                          (n=219)       (n=226)     (n=221)      (n=184)       (n=175)      (n=184)
      Tuberculosis
          All Cases            2           0           1            2             3              0
       Extra-pulmonary         2           0           1            1             3              0
             Fatal             0           0           0            0             1              0



         7.11.2. Thrombosis

An excess number of graft losses due to primary graft thrombosis was observed among
patients receiving both the belatacept MI regimen (n=7) and the belatacept LI regimen
(n=8), compared to patients receiving CsA (n=2), in IM103027, a study which
exclusively enrolled recipients of ECD kidneys. As recipients of ECD kidneys are at
higher risk of primary graft thrombosis, one might postulate either that IM103027 might


                                                                                                      67
represent the study most likely to capture a safety signal related to graft thrombosis or
that the CsA group in IM103027 experienced surprisingly few events of primary graft
thrombosis. Despite the imbalance in primary graft thrombosis, fewer belatacept patients
(17 for belatacept MI, 16 for belatacept LI) in IM103027 experienced graft loss in the
first 12 months than CsA patients (n=20). Graft loss due to thrombosis was equivalent
across groups in IM103100 (one with belatacept MI, one with belatacept LI, and two with
CsA) and IM103008 (one with belatacept MI, one with belatacept LI, and three with
CsA). A post hoc analysis of other adverse events related to thrombosis (e.g., deep vein
thrombosis) did not suggest a clear correlation between belatacept exposure and
increased thromboses.

       7.11.3. Proteinuria

Proteinuria was first identified in the Phase 2 study, IM103100, as an adverse event more
frequently observed among patients receiving belatacept than CsA. The observation was
confirmed in both Phase 3 trials. The development of proteinuria is an especially
important clinical event in kidney transplant recipients where its presence correlates
strongly with decreased graft life. Unfortunately, however, the Phase 2 and Phase 3 trials
were poorly designed to capture data regarding proteinuria. Rather than measuring urine
protein to creatinine (UP/C) ratios, proteinuria was assessed by urine dipstick
assessments. According to the protocol, proteinuria was defined as 2+ on dipstick
assessment on 2 consecutive visits. Using that criteria, belatacept patients exhibited
proteinuria at 12 months more frequently than CsA patients. Reported rates in IM103008
were 6% for belatacept MI patients, 4% for belatacept LI patients, and 3.7% for CsA
patients; reported rates in IM103027 were 10.9% for belatacept MI patients, 10.1% for
belatacept LI patients, and 6.8% for CsA patients. It is difficult to assess the clinical
significance of this observation without more quantitative data (such as would be
provided by UP/C measurements). Like GFR, proteinuria is affected by renal
hemodynamics: in the presence of CsA, proteinuria should be reduced. Insufficient data
exist to conclude whether the differences in proteinuria derive from that physiologic
effect or from structural differences at the level of the glomerulus. While relatively few
patients on either CsA or belatacept met the protocol definition of proteinuria,
insufficient data exists to determine how many patients exhibited a clinically significant
degree of proteinuria.

       7.11.4. Infusion-Related

Adverse events that have been reported during or around the infusion of belatacept
include headache, flushing, hypotension, and hypertension. Approximately 5% of
patients receiving belatacept in the clinical trials reported an infusion related adverse
event during the course of the study. Anaphylaxis and/or hypersensitivity have not been
observed in the setting of a belatacept infusion.




                                                                                        68
       7.11.5. Autoimmunity

Although there was no evidence of autoimmunity in toxicology studies with belatacept in
rats and monkey, nonclinical studies of a related product, abatacept, showed evidence of
inflammation and autoimmunity in rats, one of three animal species tested. Abatacept
differs from belatacept by two amino acid substitutions and, in rodents, has greater in
vitro binding affinity and in vivo inhibition of a T-cell dependent antibody response than
belatacept. Belatacept has greater binding affinity and inhibition of T-cell proliferation
and activation than abatacept in human in vitro studies. Toxicology studies of abatacept
in rodents are therefore considered to be relevant to the belatacept safety assessment.

In studies with abatacept at clinically relevant doses, inflammation and autoimmunity
(insulitis and thyroiditis) were observed in juvenile and adult rats. The incidence and/or
severity of the autoimmune findings in rats increased during 2-3 month treatment-free
recovery periods. In a peri- post-natal study, thyroiditis was observed in one out of
twenty rat off-spring exposed in utero and during lactation, at the highest of three doses
(200 mg/kg, a maternal abatacept exposure ~11-times higher than the estimated
belatacept clinical exposure). The thyroids were not examined in maternal animals.
Autoimmunity was not observed in chronic toxicology studies with abatacept in mice and
monkeys; therefore, the autoimmunity described in rats may represent a rat-specific effect
of abatacept. While these findings may indicate a need for vigilance when abatacept is
used in patients including pregnant women, the relevance of these findings to patients or
pregnant women treated with belatacept remains unclear since no evidence of
autoimmunity has been observed in rat and monkey toxicology studies with belatacept.

Adverse events related to autoimmune disease appeared evenly distributed across all
treatment groups in the belatacept kidney trials: approximately 2% of patients in
IM103100, IM103008, and IM103027 reported autoimmune events. Psoriasis was seen in
three belatacept MI treated patients and three belatacept LI treated patients;
hyperthyroidism was reported in three CsA treated patients and no belatacept treated
patients. An episode of Guillain-Barré syndrome was reported in one patient in
IM103027 in the belatacept MI group. Diabetes mellitus was reported in 5% of
belatacept patients and 10% of CsA patients.

Patients who receive solid organ transplants generally require lifelong
immunosuppression, and immunosuppression could mask the appearance of
autoimmunity. In the Phase 3 trials of belatacept, 20% to 30 % of patients stopped
belatacept and the majority were switched to another form of immunosuppression.
Therefore, the lack of any significant autoimmune signal during the kidney transplant
trials may mean that (a) a signal was not seen because belatacept does not increase the
risk for autoimmunity in adult transplant patients, (b) systematic evaluation for the
development of autoimmunity was not done, (c) patients were receiving other
immunosuppressive products that masked/treated the autoimmune adverse reaction. It
remains unclear, however, whether the use of belatacept could result in autoimmunity in



                                                                                        69
cases where treatment with (exposure to) belatacept was stopped and no other
immunosuppression was instituted. Discontinuation of immunosuppression in the setting
of graft loss represents one scenario where a paucity of data on whether or not
autoimmunity would be seen exists. In utero exposure to belatacept represents another
important instance where belatacept exposure would be present during gestation but
would stop upon birth of the infant. The concern regarding autoimmunity in the offspring
of a woman receiving belatacept signal is supported by the finding of insulitis and
thyroiditis in one of twenty rat off-spring after exposure to abatacept in utero and during
lactation. Finally, the difference in maturation of the immune system between pediatric
patients and adult patients may also have implication on the differential risk for
immunity.

Abatacept is currently marketed as Orencia® for the treatment of adult rheumatoid
arthritis and juvenile idiopathic arthritis (pediatric patients 6 years or older). In the
currently approved Orencia® labeling, 46 the potential risk of autoimmunity is addressed in
the Pregnancy Section 8.1, the Pediatric Use Section 8.4, and the Animal Toxicology
and/or Pharmacology Section 13.2.

           8.1 Pregnancy

           Pregnancy Category C

           There are no adequate and well-controlled studies of ORENCIA use in pregnant women.
           Abatacept has been shown to cross the placenta in animals, and in animal reproduction studies
           alterations in immune function occurred. ORENCIA should be used during pregnancy only if the
           potential benefit to the mother justifies the potential risk to the fetus.

           Abatacept was not teratogenic when administered to pregnant mice at doses up to 300 mg/kg and
           in pregnant rats and rabbits at doses up to 200 mg/kg daily representing approximately 29 times
           the exposure associated with the maximum recommended human dose (MRHD) of 10 mg/kg
           based on AUC (area under the time-concentration curve).

           Abatacept administered to female rats every three days during early gestation and throughout the
           lactation period, produced no adverse effects in offspring at doses up to 45 mg/kg, representing 3
           times the exposure associated with the MRHD of 10 mg/kg based on AUC. However, at 200
           mg/kg, 11 times the MRHD exposure, alterations in immune function were observed consisting of
           a 9-fold increase in T-cell dependent antibody response in female pups and thyroid inflammation
           in one female pup. It is not known whether these findings indicate a risk for development of
           autoimmune diseases in humans exposed in utero to abatacept. However, exposure to abatacept in
           the juvenile rat, which may be more representative of the fetal immune system state in the human,
           resulted in immune system abnormalities including inflammation of the thyroid and pancreas [see
           Nonclinical Toxicology (13.2)].

           Pregnancy Registry: To monitor maternal-fetal outcomes of pregnant women exposed to
           ORENCIA, a pregnancy registry has been established. Healthcare professionals are encouraged to
           register patients and pregnant women are encouraged to enroll themselves by calling 1-877-311-
           8972.



46
     http://www.accessdata.fda.gov/drugsatfda docs/label/2009/125118s0086lbl.pdf



                                                                                                          70
       7.11.6. Immunogenicity

A minority of study patients developed antibodies against belatacept. During the Phase 3
trials, 25/796 patients (3.1%) developed anti-belatacept antibodies at some point during
treatment. Following discontinuation of treatment (i.e. ≥ 56 days post last dose or
discontinuation), 7 of the 88 patients evaluated who were previously seronegative became
seropositive (8%). For the long term extension phase (LTE) of the phase 2 trial, in both
the 4- and 8-week dosing regimens, 16/95 patients (16.8%) were seropositive at some
point during observation. An approximately equal number of patients were defined as
having transient positive anti-belatacept antibody titers as those defined as persistently
positive. Interestingly, a high proportion of the antibodies were directed against the
CTLA-4 domain of belatacept: fourteen of the twenty five patients (56%) in the phase 3
trials and 12 of the 16 patients (75%) in the phase 2 LTE who developed anti-belatacept
antibodies had antibodies which were reactive with the CTLA-4 domain of belatacept.
The sponsor has not examined whether such antibodies bind native CTLA4 on human T
cells, but if so, they could conceivably induce autoimmunity similar to that experienced
with the licensed anti-CTLA4 antagonist antibody ipilimumab. Conversely, such
antibodies could act as agonists and potentially trigger signaling through CD28.

There appears to be a trend towards higher rates of graft loss, death, and acute rejection in
those patients who were seropositive or indeterminate than in those who were
seronegative to anti-belatacept antibodies. However, the complexity and the small
numbers of clinical cases, and the limited immunogenicity and belatacept concentration
data taken around the time of the event of interest do not support a causal or temporal
association between the development of anti-belatacept antibodies and any of these
events.

In terms of mean and median belatacept clearance, there was no significant difference
among seropositive, seronegative and indeterminate patients, as well as among
neutralizing antibody (NAB)-positive and other patients.

       7.11.7. Malignancies Other than PTLD

Other than the excess events of PTLD associated with belatacept (see Section 7.3), the
incidence of malignancies appeared similar across treatment groups in IM103100,
IM103008, and IM103027. Non-melanoma skin cancer was the most common
malignancy detected in all groups (3.1% of belatacept MI patients, 1.3% of belatacept LI
patients, and 2.3% of CsA patients at time of database lock in a pooled analysis of all
three studies). Despite the excess of PTLD events among the belatacept group, the
number of patients with malignant neoplasms excluding non-melanoma skin cancers
remained similar across all patient groups (6.7% of belatacept MI patients, 4.9% of
belatacept LI patients, and 4.8% of CsA patients at time of database lock in a pooled
analysis of all three studies).




                                                                                          71
       7.11.8. Cardiac – QT Prolongation

Belatacept is a large fusion protein with a molecular weight of approximately 91 kDa and
with high specificity for its target; therefore a thorough clinical QT study was not
conducted. Non-clinical evaluations have suggested no evidence of cardiovascular or
hemodynamic abnormalities. In both Phase 3 studies, 12-lead ECGs were obtained at
baseline prior to transplant, prior to administration of study drug at week 12 and at week
52. Assessments were based upon local, non-standardized evaluations. Patients were not
excluded from the study based on any ECG criteria. Bazett’s correction for QT interval
was used.

Increases in the QTc interval > 60 msec compared with baseline were observed in ≤ 7%
of patients in all treatment groups at Month 3 and Month 12. No clinically relevant
effects on the PR or QRS intervals were apparent. A review of narratives relating events
of cardiac arrest, cardiopulmonary events, or sudden death did not identify any reports of
QT prolongation associated with those events. There were no reports of Torsade de
Pointes or other significant ventricular arrhythmias with temporal association to
belatacept administration.

In summary, based upon review of the ECG data and narratives for cardiac-related deaths
in clinical trials IM103008 and IM103027, there does not appear to be any large effect of
belatacept on the QT interval.

8. Risk Evaluation and Mitigation Strategy (REMS)

FDA considers risk management to be the continuing process of minimizing risks
throughout a product’s lifecycle to optimize its benefit-risk balance, as stated in the
Guidance for Industry: Development and Use of Risk Minimization Action Plans March,
2005 (and shown in Appendix E). 47

Typically, product safety issues are managed through a product’s package insert,
sponsor-provided communications, post-marketing studies, and routine post-marketing
safety surveillance. However, if the seriousness of the risks associated with the product
make it necessary to require and enforce risk management beyond these measures, the 2007
Food and Drug Administration Amendments Act (FDAAA) provides FDA authority to
require risk evaluation and mitigation strategies (REMS). Accordingly, REMS may be
required if FDA determines that a REMS is necessary to ensure that the benefits of the drug
outweigh the risks (Section 505-1(a)).



47

http://www fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM18412
8.pdf



                                                                                           72
REMS may include one or more of the following: A Medication Guide or patient package
insert for patients, a Communication Plan for healthcare providers, and Elements to
Assure Safe Use (ETASU), which often involve some form of restricted distribution and
or evidence of safe-use conditions. The statute [Section 505-1(d)] also requires that all
approved REMS for NDA and BLA products have a timetable for submission of
assessments of the REMS. These assessments are prepared by the sponsor and reviewed
by FDA.

A Medication Guide provides FDA approved patient labeling. A Medication Guide can
be required if FDA determines one or more of the following apply:

   •   Patient labeling could help prevent serious adverse events.
   •   The product has serious risks that could affect a patient’s decision to use or
       continue to use the drug.
   •   Patient adherence to directions is crucial to product effectiveness.

A Communication Plan consists of FDA approved materials used to aid a sponsor’s
implementation of the REMS and/or inform healthcare providers about serious risk(s) of
an approved product. For example, “Dear Healthcare Professional” letters, collaboration
with professional societies, brochures focusing on the important risk messages, and other
educational materials have been required to alert prescribers to serious risks associated
with the use of certain drugs and biologics.

ETASU can include one or more of the following requirements:

       •   Prescriber training or certification
       •   Certification of dispensers
       •   Drug administration restricted to certain health care settings
       •   Documentation of safe-use conditions prior to dispensing
       •   Monitoring of patients
       •   Enrollment of patients in a registry

Because ETASU can impose significant burdens on the healthcare system and reduce
patient access to treatment, ETASU are required only if FDA determines that the product
could be approved only if, or would be withdrawn unless, ETASU are required to
mitigate a specific serious risk listed in the labeling. Accordingly, the statute [FDCA 505-
1(f)(3)] specifies that ETASU:

       •   Must be commensurate with specific serious risk(s) listed in the labeling.
       •   Cannot be unduly burdensome on patient access to the drug.
       •   To minimize the burden on the healthcare delivery system, must, to the extent
           practicable, conform with REMS elements for other drugs with similar serious
           risks and be designed for compatibility with established distribution,
           procurement, and dispensing systems for drugs.




                                                                                         73
In summary, FDA has the authority to require REMS if additional measures are necessary
to ensure the benefits of a drug outweigh the risks. In considering a REMS for belatacept,
FDA must keep in mind the following factors: the estimated size of the population likely
to use the product, the seriousness of the disease or condition that is to be treated with the
product, the expected benefit of the product with respect to such disease or condition, the
expected or actual duration of treatment with the drug, the seriousness of any known or
potential adverse events related to the drug and the background incidence of such events
in the population likely to use the drug , and whether the drug is a new molecular entity.
FDA must also keep in mind the need to minimize the burden on the healthcare system
and the barriers to patient access while adequately managing the risks so that the benefits
of belatacept will outweigh the risks.

9. Overall Risk Benefit

Belatacept has been developed as a therapeutic agent for immunosuppression in de novo
kidney transplant recipients; in the Phase 2 and Phase 3 trials, it was compared to CsA as
the control, in an immunosuppression regimen that included basiliximab, MMF and
steroids in both the belatacept and the CsA arms. The belatacept development program
was designed to evaluate efficacy endpoints related to its effects on immunosuppression,
namely patient graft loss and death and AR. However, the trials also evaluated other
endpoints related to renal function (GFR), blood pressure, lipids, diabetes, and CAN;
which are known toxicities associated with other immunuosuppressants, specifically the
CNIs.

The differential effect of CsA and belatacept on renal hemodynamics (due to
vasoconstrictive effects of CsA) complicates the interpretation of the GFR results from
the Phase 3 trials for the evaluation of the efficacy of belatacept. While the composite
renal impairment endpoint as defined in the Phase 3 trials was not adequate to establish
efficacy, the belatacept treated groups also had significantly higher Month 12 GFRs.
Other CNI-free regimens have not consistently been shown to result in higher GFRs. In
other published randomized controlled clinical trials (RCTs), CNI-free
immunosuppressive regimens have been compared to CNI-based regimens and despite
the renal hemodynamic advantage, the results in terms of GFR have not been
consistent. 48,49,50,51 The literature also suggests that a higher GFR at Month 1 and Month
12 is associated with decreased cardiovascular death and improved graft survival;

48
   Ekberg H, Tedesco-Silva H, Demirbas A, et al. Reduced exposure to calcineurin inhibitors in renal
transplantation. N Eng J Med 2007; 357: 2562-75.
49
   Larson TS, Dean PG, Stegall MD, et al. Complete avoidance of calcineurin inhibitors in renal
transplantation: a randomized trial comparing sirolimus and tacrolimus. Am J Transplant 2006; 6(3): 514-
22.
50
   Hamdy AF, El-Agroudy AE, Bakr MA, et al. Comparison of sirolimus with low-dose tacrolimus versus
sirolimus-based calcineurin inhibitor-free regimen in live donor renal transplantation. Am J Transplant
2005; 5(10): 2531-8.
51
   Flechner SM, Goldfarb D, Solez, et al. Kidney transplantation with sirolimus and mycophenolate
mofetil-based immunosuppression: 5-year results of a randomized prospective trial compared to calcineurin
drugs. Transplantation 2007; 83(7): 883-92.



                                                                                                      74
however, these data were derived largely from analyses of groups of patients with more
comparable renal hemodynamics. 52,53,54

Another way in which to evaluate the issue of physiologic increases in GFR, not
necessarily related to underlying kidney structure, is by examining the effects of other
drugs used in transplant which alter GFR. Several classes of antihypertensives, including
calcium channel blockers, with varying effects on GFR are used in kidney transplant. 55
However, the National Kidney Foundation Kidney Disease Outcomes Quality Initiative
(NKF K/DOQI) guidelines do not recommend any one particular class of
antihypertensive drugs be used as treatment in kidney transplant patients and also state “it
cannot be determined whether the improved GFR with calcium-channel blockers reflects
the short-term hemodynamic effects of these agents or a long-term protective effect.” 56
Some literature, however, suggests that use of calcium channel blockers in renal
transplant patients may not only result in superior GFR but also in fewer graft losses. 57
Therefore, at the current time, the relative importance of optimizing GFR in the setting of
a kidney transplant remains controversial.

Given the uncertainties that exist about the importance of therapies that result in a direct
increase in GFR with an absent or unclear effect on kidney architecture, further insight
into the potential clinical impact of the GFR differences observed in the belatacept
clinical trials may be offered by a comparison of numbers of patients alive and with a
functioning graft at Month 24 in the Phase 3 studies combined. The results numerically
favor belatacept: 355 out of 403 (88.1%) belatacept MI patients and 358 out of 401
(89.3%) belatacept LI patients versus 346 out of 405 (85.4%) CsA patients. However, it
should also be noted, that these studies lack the power to make statistical conclusions
regarding outcomes of death and graft loss. Study IM103008 drives the observed
difference in numbers of patients alive with a functioning graft and also is the study
where the greatest difference in GFR between belatacept and CsA patients was seen,
strengthening the association between a higher GFR and an improved outcome.

Several other clinically important safety endpoints favor the belatacept regimen,
including those pertaining to cardiovascular risk factors (hypertension, dyslipidemias,
and NODAT). Of these, the differences in blood pressure detected between the belatacept
groups and the CsA groups appears the most likely to result in a clinical benefit. A very
low (4.6%) incidence of NODAT was observed among belatacept patients in the clinical
52
   Hariharan S, McBride MA, Cherikh WS, et al. Post-transplant renal function in the first year predicts
long-term kidney transplant survival. Kidney Int 2002; 62: 311-8.
53
   Meier-Kriesche HU, Baliga, Kaplan B. Decreased renal function is a strong risk factor for cardiovascular
death after renal transplantation. Transplantation 2003; 75: 1291.
54
   Pilmore H, Dent H, Chang S. Reduction in Cardiovascular Death after Kidney Transplantation.
Transplantation 2010; epublished ahead of print.
55
   Cross NB, Webster AC, Masson P, et al. Antihypertensives for kidney transplant recipients: systematic
review and meta-analysis of randomized control trials. Transplantation 2009; 88(1): 7-18.
56
   NKF K/DOQI Guideline 10: Pharmacological Therapy: Kidney Disease in the Kidney Transplant
Recipient (http://www kidney.org/PROFESSIONALS/kdoqi/guidelines_bp/guide_10.htm)
57
   Cross NB, Webster AC, Masson P, et al. Antihypertensives for kidney transplant recipients: systematic
review and meta-analysis of randomized control trials. Transplantation 2009; 88(1): 7-18.



                                                                                                        75
trials, though the implication of that observation is somewhat mitigated by the low rate of
NODAT observed in the CsA control group (9.6%). The difference in lipids observed
between the belatacept patients and the CsA patients was driven by triglycerides. While
the triglyceride profile associated with the belatacept patients seems more favorable, the
literature correlating outcomes with cholesterol largely relates to LDL and HDL.

The applicant has suggested that the data regarding the slightly lower incidence of CAN
in the belatacept treated groups demonstrates an improved underlying kidney architecture
compared to the CsA treated patients. However, the most current literature has
emphasized that a finding of CAN has little or no prognostic value in determining long-
term graft survival. 58

In terms of safety outcomes, the greatest risks associated with the use of belatacept in the
clinical trials were related to increased rates of AR (including higher grades of AR),
PTLD (predominantly CNS PTLD), PML, and CNS infection (only in the belatacept MI
regimen).

An increase in overall rates of AR was observed in Study IM103008 among patients in
the belatacept groups compared to the CsA control. In both IM103008 and IM103027, a
trend towards higher grade rejections was noted among the belatacept patients. Many of
the AR episodes in the belatacept arm were treated with lymphocyte depleting agents
(e.g., thymoglobulin) and the contribution of the agents to the subsequent development of
PTLD is unknown, but should be considered.

An imbalance in rates of PTLD was detected between the belatacept and CsA groups in
both Phase 3 studies as well as the Phase 2 study and a similar signal appears to be
present in a preliminary Phase 2 liver study. The finding that most of these cases
presented with CNS involvement was also distinguishing characteristic of these events.
The imbalance was greatest among EBV- patients, but was also detected among EBV+
patients. The applicant has concluded that the risk-benefit profile of belatacept justifies
its use among EBV+ kidney transplant recipients but not EBV- kidney transplant
recipients. The data support a point estimate that the use of a belatacept-based regimen
that also includes IL-2 induction, MMF, and corticosteroids instead of a CsA-based
regimen that also includes IL-2 induction, MMF, and corticosteroids will result in an
additional case of PTLD for every 150 EBV+ kidney transplant recipients treated. It is
expected that many of these additional cases would present in the CNS. The confidence
intervals around this point estimate, however, are rather wide. Although the applicant
suggested that the incidence of PTLD rates observed within the belatacept arms of the
study was consistent with rates of PTLD seen in registry and claims data of patients
maintained on CNI-based regimens. However, this interpretation was not supported by a
FDA review of previous NDA clinical trials or of registry data from UNOS, CTS, and
CMS.



58
  Gourishankar S, Leduc R, Connett J, et al. Pathological and Clinical Characterization of the ‘Troubled
Transplant’: Data from the DeKAF Study. Amer J of Transplantation 2010; 10: 324-330.


                                                                                                           76
The report of one case of PML in the belatacept MI treatment group of IM103027 and of
another PML case in a liver transplant patient exposed to a different belatacept MI
regimen is noteworthy. Together with the high rate of CNS PTLD observed among
recipients of both belatacept regimens and an elevated rate of other CNS infections
observed among recipients of the belatacept MI regimen, the observation of two cases of
PML in a clinical trial setting underline that CNS immunity is affected differently by
exposure to belatacept than by exposure to CsA. While concerning, the rate of PML
observed in the belatacept kidney program is unlikely to significantly change the estimate
of the NNTH based on the primary safety signal related to PTLD.

In summary, careful consideration should be given to whether the data regarding the
observed benefits associated with the belatacept regimen (improved GFR, improved
blood pressure, reduced rates of NODAT, numerically better outcomes with regard to
patients alive with a functioning graft at 24 months) outweigh the risks of higher
mortality and excess morbidity due to PTLD and PML and higher rates and higher grades
of AR.




                                                                                       77
10. Draft Questions to the Committee

Based on the information presented from Studies IM103008, IM103027, and IM103100,
please comment on the following questions regarding belatacept, in the context of the
proposed Less Intensive (LI) regimen in combination with an interleukin-2 (IL-2)
receptor antagonist (basiliximab), a mycophenolic acid (MPA) product (mycophenolate
mofetil), and corticosteroids:

   1. Has efficacy been demonstrated for the prophylaxis of acute rejection in de novo
      renal transplant recipients? (vote yes or no)
              a.    Please discuss your rationale for your vote.
              b.    If the answer is no, what additional information is needed or what
                    additional studies should be conducted?

   2. Has safety been demonstrated for the prophylaxis of acute rejection in de novo
      renal transplant recipients? (vote yes or no)
         a. Please discuss your rationale for your vote.
         b. If the answer is no, what additional information is needed or what
              additional studies should be conducted?

   3. Provide comments and discuss what risks should be addressed through the Risk
      Evaluation and Mitigation Strategy (REMS).

   4. Given the overall benefits and risks, do you recommend that belatacept be
      approved for the prophylaxis of acute rejection in de novo renal transplant
      recipients? (vote yes or no)
              a.    Please discuss your rationale for your vote.
              b.    If the answer is no, what additional studies would be needed to
                    demonstrate that everolimus is safe and effective?
              c.    If the answer is yes, please discuss whether additional information or
                    studies should be requested post-marketing.




                                                                                       78
11. Appendix A: NonInferiority (NI) Margin Justification

One objective of the belatacept Phase 3 program was to demonstrate that at least one of
the belatacept treatment regimens was non-inferior to the CsA regimen with respect to
the frequency of acute rejection at 12 months. The sponsor had specified a margin of
20% for both studies.

As with all NI studies, it is necessary to determine how the efficacy of the new drug can
be determined based on the results of the NI study. This is done by providing a
justification of the NI margin used to assess the results for the study. Regarding choosing
a margin, the ICH guidance document E10: Choice of control group and related issues in
clinical trials, states the margin:
    • “cannot be greater than the smallest effect size that the active drug would be
         reliably expected to have compared with placebo in the setting of the planned
         trial” (M1) and
    • “is based on both statistical reasoning and clinical judgment, should reflect
         uncertainties in the evidence on which the choice is based, and should be suitably
         conservative.” 59

In the setting of a multi drug regimen, the NI margin chosen should not be larger than the
amount of efficacy the control arm has over the putative placebo (i.e. the appropriate
placebo arm if one were included in the study). The control effect needs to be
determined by assessing the difference between the putative placebo and the control arm
using data from previously conducted clinical trials. For the current studies the treatment
arms are as follows:

Experimental:              belatacept + basiliximab (B) + MMF + corticosteroids (CS)
Control:                   CsA        + basiliximab + MMF + corticosteroids
Putative Placebo:                       basiliximab + MMF + corticosteroids

The ideal way to assess the effect of the control regimen over the putative placebo (i.e.,
estimate M1) would be based on obtaining a treatment effect from multiple randomized
studies that compared these two regimens. However, there are no studies available which
allow for the direct comparison of the control arm to the putative placebo. Since this
information is not available, the next step would be to estimate the efficacy of the
putative placebo and the efficacy of the control regimen from separate sources and
compare them.

There is only one published study (Vincenti, 2001) 60 available which studied a regimen
similar to the putative placebo. In this study subjects received daclizumab (D) instead of

59
   International Conference on Harmonisation. Guidance for Industry: E 10 Choice of control group and
related issues in clinical trials. Food and Drug Administration, DHHS, 2001.
60
   Vincenti et al. Multicenter Trial Exploring Calcineurin Inhibitors Avoidance in Renal Transplantation.
Transplantation 2001; 71:1282-7.


                                                                                                            79
basiliximab in addition to MMF and corticosteroids. The rate of biopsy proven acute
rejection at 12 months in this study was 53% (52/98) with a 95% confidence interval of
(42.7, 63.2). If graft losses, deaths, and subjects who are lost to follow-up are considered
events as well, defined as efficacy failure here, efficacy failure at 12 months in the
Vincenti study would be ~ 58.2% (57/98) with a 95% confidence interval of (47.9, 68.4).

Six studies were found that containing 7 randomized treatment arms of CsA +
basiliximab + MMF + corticosteroids similar to the control arm for the current study.
Table 44 contains the results of the acute rejection endpoint from these studies. A 95%
CI from pooling efficacy failure from these 6 studies is (13.8, 20.4). A conservative
estimate (high estimate) for this regimen is 20.4. Comparing this to a conservatively low
estimate from the Vincenti study 42.7, leads to a difference of 22.3 (42.7 – 20.4). Thus,
based on the limited amount of data available for assessing the effect of CsA in a regimen
that also contains basiliximab, MMF and corticosteroids with respect to the rate of acute
rejection, a NI margin no greater than approximately 20% would be acceptable from a
data driven point of view.

     Table 44. 12-Month Acute Rejection and Efficacy Failure Rates from Literature
Study                    Treatment arm                 Acute Rejection             Efficacy Failure
Budde (2006) 61          CS + B + CsA + MMF            8/45 (17.8%)                9/45 (20%)
Silva (2007) 62          CS + B + CsA + MMF            29/212 (13.7%)              36/212 (17%)
Kamar (2005) 63          CS + B + CsA + MMF            26/100 (26.0%)              29/100 (29%)
                         CS + B + CsA + MMF            15/97 (15.5%)               23/97 (23.7%)
Sollinger (2001) 64      CS + B + CsA + MMF            13/70 (18.6%)               15/70 (21.4%)
Pescovitz (2003) 65      CS + D + CsA + MMF            7/50 (14%)
Lawen (2003) 66          CS + B + CsA + MMF            12/59 (20.3%)
                         Random Effect C.I*            (13.8, 20.4)                (17.0, 26.0)
* DerSimonian and Laird Method

The FDA’s preferred assessment of acute rejection is based upon the endpoint of efficacy
failure which is defined as the incidence of biopsy proven acute rejection, graft loss,
death or lost to follow-up. This analysis of acute rejection considered missing data due to
graft loss, death, or lost to follow-up as events in the analysis. Four studies were found
that considered efficacy failure at 12 month for the treatment regimen of CsA +
basiliximab + MMF + corticosteroids. A 95% CI from pooling efficacy failure from

61
   Budde K et al (2007) Reduced-exposure cyclosporine is safe and efficacious in de novo renal transplant
recipients treated with enteric-coated mycophenolic acid and basiliximab. Clin Nephrol;67(3):164-75.
62
   Silva, HT et al (2007) One-year results with extended-release tacrolimus/MMF, tacrolimus/MMF and
cyclosporine/MMF in de novo kidney transplant recipients. Am J Transplant; (7):595-608.
63
   Kamar N et al (2006) Impact of early or delayed cyclosporine on delayed graft fuctio in renal transplant
recipients: a randomized, multicenter study. Am J Transplant;6(5 Pt 1):1042-8.
64
   Sollinger H et al. Basiliximab versus antithymocyte globulin for prevention of acute renal allograft
rejection. Transplantation 2001; 72 (12) 1915-1919.
65
   Pescovitz MD et al (2003) Pharmacokinetics of daclizumab and mycophenolate modetil with
cyclosprorine and steroids in renal tranaplantation. Clin Transplant; 17(6):511-7.
66
   Lawen JG et al (2003) Randomized double-blind study of immunoprophylaxis with basiliximab, a
chimeric anti-interleukin-2 receptor monoclonal antibody, in combination with mycophenolate mofetil-
containing triple therapy in renal transplantation. Transplantation; 75(1):37-43.



                                                                                                         80
these 4 studies is (17.0, 26.0). A conservative estimate (high estimate) of this regimen is
26.0. Comparing this to a conservatively low estimate from the Vincenti study 47.9,
leads to a difference of 21.9 (47.9 – 26.0). Based on this data, with respect to the rate of
acute rejection, where graft losses, deaths and losses to follow-up are considered as
events, a NI margin no greater than approximately 20% would be acceptable from a data
driven point of view.

In efforts to find additional information to support the justification of the NI margin, the
results of a previously conducted literature search and mixed effects modeling approach
were used to estimate the effects of the control and putative placebo rates. The
previously conducted literature search identified all relevant randomized clinical trials
(RCTs) in de novo kidney transplantation, excluding trials conducted in special
populations (e.g. pediatrics, delayed graft function only, non-heart beating donor). The
search yielded 47 relevant clinical trials published between 1996 and 2008. No studies
were identified that evaluated the use of B+CS+MMF (the putative placebo) in renal
transplantation, many studies did not include a 12-month efficacy endpoint, and several
studies did not report the incidence of the composite endpoint used by the Division.
Therefore the BPAR endpoint was used as surrogate while assuming no major differences
in treatment effects between BPAR alone versus the composite endpoint. Lastly, there
was considerable variation in treatment regimens (i.e. drugs used and doses) used in the
identified studies. Therefore, for simplicity each calcineurin-inhibitor (CNI) was
categorized into ‘reduced (r)’ or ‘standard (s)’ drug exposure categories.

Given the lack of any RCTs evaluating the efficacy of the putative placebo in renal
transplant patients, a mixed effects model (MEM) was used to estimate the contribution
of each of the immunosuppressant drugs to the combination therapy event rate. These
models assume additive drug effects in a combination therapy on a log-odds scale. The
model was used to estimate the combined effect of the three immunosuppressant drugs
(B+CS+MMF) for the putative placebo group and of the four (B+CS+CsA(s)+MMF) for
the control group.

The mathematical model is as follows:

Let
  Yij ~ Binomial ( N ij ,π ij ) where Yij is the number of events in study i and treatment arm j
  N ij , is the number of patients and π ij is the event rate in study i and treatment arm j.

The log-odds of the event rate is related to the study effects based on a linear model with
the random study effect:

logit (π ij ) = μ + δ i + xij β where the intercept and effects of the immunosuppressant drugs and
covariates are μ and β respectively, the random effect δ i~ N (0,σ 2 ), vector xij includes indicators for
presence or absences of each of the immunosuppressant drugs and covarates

The model was fit using the maximum-likelihood method (PROC NLMIXED in SAS).


                                                                                                   81
Based on this modeling approach, the estimated failure rate of the control group
(B+CS+CsA(s)+MMF) is 19.3%, 95% CI (16.7, 21.8) and that for the putative placebo
(B+CS+MMF) is 63.4%, 95% CI (55.1, 71.7) and an estimated difference in event rates
(B+CS+MMF - B+CS+CsA(s)+MMF) of 44.1% (36.2, 52.1).

While this modeling approach includes several assumptions, this approach seems
reasonable given the absence of comparative (concurrent) clinical data. The lower bound
around the estimated difference in failure rates is 36.2% (e.g. the minimum (estimated)
amount that active control is better than putative placebo in terms of efficacy failure).
Thus providing additional support of a margin of 20%.




                                                                                       82
12. Appendix B: PTLD Analysis

At 24-month follow-up, there was a higher number of reported post-transplant
lymphoproliferative disorder (PTLD) cases (8/477 and 5/472 in the MI and LI groups
respectively) compared to the control (n=2/478) in the pooled Phase 2 and 3 trials. The
applicant notes that the rate of PTLD in belatacept-treated subjects was higher in EBV-
negative recipients (7.3%, 7 of 96) compared with EBV-positive recipients (0.6%; 5 of
805).

To further assess the rate of PTLD observed in the belatacept trials, we first compared the
incidence of PTLD in the belatacept groups to the CsA group using pooled data from all
three trials (Section A.1). Next, we compared the incidence of PTLD in the belatacept
regimen (LI and MI pooled and LI alone) to an estimated incidence of PTLD using data
from three previously reported randomized clinical trials that included the same CsA
regimen as was used in the three belatacept trials (Section A.2). Lastly, we compared he
incidence of the PTLD from the belatacept regimen against estimated incidence of PTLD
using data from the UNOS registry (Section A.3).


A.1 Comparing incidence of PTLD in the three Belatacept Clinical Trials

The calculated incidence per 100 person-years (PY) based on the reported cases of PTLD
by EBV status and treatment regimen across the three belatacept trials are shown in Table
43 below. Among EBV+ patients, the 24-month incidence of PTLD in the LI and MI
pooled is 0.346, 95% exact CI (0.112, 0.807). Among EBV+ patients receiving belatacept
LI, the incidence per 100 PY is 0.410, 95% CI (0.085, 1.198).




                                                                                          83
     Table 45: PTLD Incidence at 24-Months: Belatacept Trials 008, 027 and 100
                                      EBV Positive
                 N         PY        PTLD     Incidence per 100PY                      95% CI#
MI              404       715.1        2              0.280                           0.034, 1.011
LI              401       731.8        3              0.410                           0.085, 1.198
CSA             399       686.9        0              0.000                           0.000, 0.537
MI and LI       805       1446.9        5             0.346                           0.112, 0.807
                                      EBV Negative
                 N         PY        PTLD     Incidence per 100PY                      95% CI#
MI               45        77.6        5              6.441                          2.087, 15.034
LI               51        82.4        2              2.427                          0.294, 8.769
CSA              57       100.2        1              0.998                           0.025,5.561
MI and LI        96       160.0        7              4.374                          1.759, 9.015
                                  EBV Status Unknown
                 N         PY        PTLD     Incidence per 100PY                      95% CI#
MI               28        50.3        1              1.987                          0.050, 11.071
LI               20        35.8        0              0.000                          0.000, 10.304
CSA              22        29.5        0              0.000                          0.000, 12.505
MI and LI        48        86.1         1             1.161                          0.000, 6.4715
                                 All Randomized Patients
                 N         PY        PTLD     Incidence per 100PY                      95% CI#
MI              477       843.0        8              0.949                           0.409, 1.870
LI              472       850.0        5              0.588                           0.190, 1.373
CSA             478       816.6        1              0.122                           0.003, 0.682
MI and LI       949       1693.1       13             0.768                           0.408, 1.314
Person-years (PY) in patients not experiencing PTLD calculated as time from transplant to time to last
follow-up date or 2-years from time of transplant, whichever is smaller
Person-years in patients experiencing PTLD during first 2-yrs post-transplant is date from transplant to date
of PTLD
# 95% Confidence interval calculated using exact Poisson method

Number Needed to Harm
Given an incidence/PY of 0.0035 of the pooled data from the belatacept MI and LI
regimens and of 0.0 in the CsA regimens from Studies 008, 027 and 100, we estimate a
number needed to harm (NNTH) of 144.7 (95% CI: -531.6, 63.7) patients (see Table 46).
This suggests that for every 145 patients treated with belatacept for two years, we would
expect at least one case of PTLD. The range for this NNTH suggests that as few as 63
belatacept patients to as many as 532 CsA patients need to be treated for two years to
expect at least one case of PTLD. Similarly, the NNTH based on the comparison between
the belatacept LI regimen and CsA is 178.8 (95 % CI: -431.7, 74.1) suggesting that as
few as 74 belatacept patients to as many as 432 CsA patients need to be treated for two
years to expect at least one case of PTLD.




                                                                                                          84
 Table 46: Estimated Number Needed to Harm: Belatacept Studies 008, 027 and 100
EBV+             Belatacept LI and MI Belatacept LI  Belatacept MI         CsA
Incidence/PY            0.0035           0.0041          0.0028           0.0000
ARRI (vs CsA)           0.0069           0.0082          0.0056             NA
2-YR NNTH                144.7            122.0           178.8             NA
95% CI NNTH          -531.6, 63.7      -726.0, 56.3    -431.7, 74.1         NA
PY=person-years
AARI: annualized absolute risk increase (events belatacept/(person-yrs belatacept/2)-events CsA /(person-yrs CsA /2)
(assumes constant risk)
NNTH: 1/AARI



A.2 Comparison of Incidence of PTLD from Belatacept Trials with Previous
Randomized Clinical Trials

Information from previous randomized clinical trials of de novo kidney transplantation
were reviewed to identify cases of PTLD in patients randomized to a CsA-based
immunosuppressant regimen. We identified three randomized trials that collected
information on patients for at least 12-months post-transplant and that included a CsA,
MMF, corticosteroid with either basiliximab or daclizumab induction treatment group.
Out of 896 total patients (534 EBV+, 121 EBV- and 201 EBV unknown), we identified
one (1) case of PTLD, which was in a patient who was EBV- at baseline. Therefore, the
estimated overall incidence of PTLD among all three trials was 0.112% (1/896) and by
EBV status: 0% (0/534), 95% exact CI (0, 0.0069) among EBV+ patients, 0.621%
(1/161), 95% exact CI (0.0002, 0.0341) EBV-, and 0% (0/201), 95% exact CI (0, 0.0182)
EBV status unknown (Table 47).

Including data from the CsA groups in the belatacept trials (008, 027 and 100) along with
the data from the three previous RCTs, the estimated overall incidence of PTLD by EBV
status is 0% (0/933), 95% CI (0, 0.0039) among EBV+ patients (see Table 47).




                                                                                                                 85
     Table 47: Summary of PTLD from Previous Randomized Clinical Trials that
   Included a CsA-based Regimen and the CsA-based Regimens from the Belatacept
                            Studies (008, 027 and 100)
                  Previous RCTs (CsA Only)*      Belatacept RCTs (CsA Only)**                  Total
                    N          PTLD (%)            N             PTLD (%)            N         PTLD (%)
                               [95% CI]                          [95% CI]                       [95% CI]
        EBV+       534              0             399                 0             933              0
                               [0, 0.688]                        [0, 0.920]                     [0, 0.395]
         EBV-      161          1 (0.621)          57              1 (1.8)          218         2 (0.917)
                             [0.016, 3.412]                   [0.044, 9.392]                 [0.111, 3.275]
EBV Unknown        201        0 [0, 1.819]         22              1 (4.5)          223              1
                                                                [0.11, 22.8]
 *Total number of patients from three previous randomized clinical trials performed in de novo kidney
 transplant recipients who were randomized to a CsA, MMF, corticosteroid with either basiliximab or
 daclizumab induction treatment group
 **Total number of patients randomized to the CsA treatment groups across the three belatacept trials (008,
 027 and 100)

  Number Needed to Harm
  Comparing the pooled incidence in the CsA-based treatment groups from the prior RCTs
  and the belatacept trials (summarized above) with the rate of PTLD from both belatacept
  treatment groups (LI and MI pooled) we get an estimated NNTH 161[1/(5/805-0/933)],
  95 % CI (69.3, 697.8) patients among EBV+ patients (see Table 48). The range for this
  NNTH suggests that as few as 69 belatacept patients to as many as 698 belatacept
  patients need to be treated for two years to expect at least one case of PTLD.

  Comparing the pooled incidence in the CsA-based treatment groups from these three
  trials with the rate of PTLD in the belatacept LI treatment group only we get estimate the
  number needed to harm at 1-year post-transplant of 133.7 [1/(3/401-0/933)], 95% CI
  (46.1, 833.3) patients among EBV+ patients (see Table 48). The range for this NNTH
  suggests that as few as 46 belatacept patients to as many as 833 belatacept patients need
  to be treated for two years to expect at least one case of PTLD.

  Comparing the pooled incidence in the CsA-based treatment groups from these three
  trials with the rate of PTLD in the belatacept MI treatment group only, we estimate the
  number needed to harm at 1-year post-transplant of 202 [1/(2/404-0/933)], 95% CI (-
  3333.33, 56.2) patients among EBV+ patients (see Table 48). The range for this NNTH
  suggests that as few as 56 belatacept patients to as many as 3333 CsA patients need to be
  treated for two years to expect at least one case of PTLD.




                                                                                                        86
Table 48: Estimated NNTH in EBV+ Patients Comparing Belatacept with CsA from
                                   Prior RCTs
EBV+            Belatacept LI and MI     Belatacept LI   Belatacept MI
NNTH                   161.0                 133.7            202
95% CI NNTH         (69.3, 697.8)         (46.1, 833.3) (-3333.33, 56.2)


A.3 Comparison of Incidence of PTLD from Belatacept Trials with Estimated
Incidence from Registry Estimates

To estimate the rates the PTLD among US kidney transplant recipients on a CNI-based
immunosuppressant regimen, we calculated the 1, 2 and 3 yrs incidence per 100 person-
years using data provided by the United States Renal Data System (USRDS). Details of
these analyses and methods can be found in Table 51 and Table 52 respectively.

Number Needed to Harm
Based on an estimated 2-year incidence per person-year in patients on a CNI-based
regimen of 0.00051, we estimated that the NNTH in belatacept pooled LI and MI of 169
patients (see Table 49). That is, for every 169 patients treated with belatacept for two
years, we would expect at least one case of PTLD. The range for this NNTH suggests
that as few as 116 to as many as 319 belatacept patients need to be treated for two years
to expect at least one case of PTLD.

  Table 49: Estimated Numbers Needed to Harm: Pooled Belatacept Compared to
                             FDA UNOS Estimates
EBV+                     Belatacept LI   Belatacept LI   Belatacept MI  FDA UNOS-
                           and MI                                        CsA/Tac*
2-Year Incidence/PY         0.0035          0.0041          0.0028        0.00051
ARI (vs. UNOS-                                                              NA
CsA/Tac)                   0.00299         0.00359          0.00229
2-YR NNTH**                  169.6           139.2           218.4          NA
95% CI NNTH             (115.5, 319.2)   (92.3, 283.0)   (122.7, 989.1)     NA
ARI: absolute risk increase (Risk belatacept – Risk CsA)
NNTH: 1/ARI
*Based on 2-yr incidence reported in Table 51
PY=person-year
**NNTH: 1/AARI where AARI: annualized absolute risk increase (eventsbelatacept/(person-yrs belatacept/2)-
events CsA /(person-yrs CsA /2) (assumes constant risk)

Additionally, the estimated numbers needed to harm based on the comparisons between
the incidence/person-years of 0.0035 of the pooled belatacept (LI and MI) to the
estimated incidences among de novo kidney transplant recipients on a CsA or tacrolimus
regimen from the analyses of the UNOS/OPTN data (BMS and SRTR) and to the
estimated incidence from the analyses of the CTS data by Dr. Opelz are shown below in
Table 50.




                                                                                                            87
  Table 50: Estimated Numbers Needed to Harm: Pooled Belatacept Compared to
               BMS and SRTR UNOS Estimates and Opelz Analysis
   EBV+ Kidney Transplant       BMS UNOS       SRTR UNOS         Opelz (CTS
           Recipients                                             Analysis)
2-Year Incidence/PY                0.0009        0.00050           0.0012
2-YR NNTH *                         192.3          166.7            217.4
*NNTH (number need to harm) is based on the difference in the 2-yr incidence per person-years between
the pooled belatacept (MI and LI) incidence of 0.0035 and the respective registry incidence
Unable to estimate confidence intervals given level of information provided
PY=person-year
NNTH: 1/AARI where AARI=annualized absolute risk increase=events belatacept/(person-yrs belatacept/2)-
events CsA /(person-yrs CsA /2) (assumes constant risk)




                                                                                                     88
                                                  Table 51: Comparison of PTLD Rates based on EBV status
                        Belatacept (pooled MI and LI)               BMS UNOS#**                             FDA UNOS**                           SRTR UNOS**
                                   N = 949                          N = not specified                         N = 77,203                            N = 83,929
                                                                   No TCD Induction                        IL-2 Induction                        IL-2 Induction
                                                                                                             n = 24,775*                            n =18,409*
                      # cases/total PYs    Incidence per    # cases/total PYs   Incidence per   # cases/total PYs     Incidence per   # cases/total PYs     Incidence per
                                             100 PYs                              100 PYs                                100 PYs                               100 PYs
                                          (exact 95% CI)                          (95% CI)                           (exact 95% CI)
EBV Positive
               1-yr    Not Available      Not Available      Not Available      Not Available      11/13245             0.083            10/10396              0.096
                                                                                                                    (0.041, 0.149)
           1.5 yr      Not Available      Not Available       34/33792.9             0.10        Not Available      Not Available      Not Available       Not Available
                                                                                 (0.07, 0.14)
               2-yr     805/1446.9            0.346          Not Available           0.09          12/23640             0.051            10/20157              0.050
                                          (0.112, 0.807)                         (0.06, 0.12)                       (0.026, 0.089)
               3-yr    Not Available      Not Available       55/57805.9             0.10          15/31733             0.047            18/28485              0.063
                                                                                 (0.07, 0.12)                       (0.026, 0.078)
EBV Negative
               1-yr    Not Available      Not Available      Not Available      Not Available       12/2162             0.555            11/1519.5             0.724
                                                                                                                    (0.287,0.970)
           1.5 yr      Not Available      Not Available        72/7187.6             1.00        Not Available      Not Available      Not Available       Not Available
                                                                                 (0.80, 1.26)
               2-yr       96/160.0            4.374          Not Available           0.54           17/3937              0.432           18/2938.5             0.613
                                          (1.759, 9.015)                         (0.39, 0.73)                        (0.252, .691)
               3-yr    Not Available      Not Available       84/12338.3             0.66           19/5308              0.358           21/4109.5             0.511
                                                                                 (0.55, 0.84)                        (0.216, .559)
EBV Unknown
           1-yr        Not Available      Not Available      Not Available      Not Available      11/9328.5            0.118             7/6486.5             0.108
                                                                                                                    (0.059, 0.211)
           1.5 yr      Not Available      Not Available       59/30487.5             0.19        Not Available      Not Available      Not Available       Not Available
                                                                                 (0.15. 0.25)
               2-yr       48/86.1              1.161         Not Available      Not Available     15/17102.5             0.088           7/12542.5             0.056
                                          (0.000, 6.4715)                                                            (0.049, .145)
               3-yr    Not Available       Not Available      80/52908.6             0.15         21/23224.5             0.090          10/17725.5             0.056
                                                                                 (0.12, 0.19)                        (0.056,0.138)



        #Results obtained from Appendix B of the BMS IM103028 Addendum 3 report and BMS “Additional Responses to FDA Comments: January 25, 2010”
        *The “n” listed under the indicated induction regimen refers to the number of subjects in the subset.
        ** UNOS analyses consider all recipients with Cyclosporine or Tacrolimus maintenance regimens



                                                                                                                                                                89
Analysis of UNOS data: Methods

The USRDS is a national data system that collects, analyzes, and distributes information
regarding end-stage renal disease in the United States. The USRDS obtains data from the
Centers for Medicare and Medicaid Services (CMS), United Network for Organ Sharing
(UNOS), and the end-stage disease networks. Data requests can be made directly to
USRDS for which the USRDS provides standardized data files and user guides.
Transplant specific information is collected directly from transplant center and reported
to UNOS. We request data regarding all kidney transplants in the United States and
received multiple analysis files that were used to acquire the variables of interest. The
data provided by the USRDS came in three separate analysis file systems:
       1) Core Standard Analysis Files: includes files that provide basic demographic,
           claims and death data for patients, facility information, medical evidence,
           waitlist information, etc.
       2) Hospitalization Files: Includes files on all inpatient hospital visits of patients
           in the USRDS database, including ICD-9 codes
       3) Transplant Standard Analysis Files: Includes all transplant details collected by
           UNOS since 1988, separated by Kidney transplant only and Kidney and
           Pancreas transplant patients. Data from follow-up forms included.




                                                                                         90
The following table provides a summary of the inclusion and exclusion criteria used in
the analysis of UNOS data by FDA, BMS and SRTR.

Table 52: Comparison of Methods used in Analysis of the OPTN/UNOS Transplant
                                Registry Data
                                 FDA UNOS                  BMS UNOS                 SRTR UNOS
  Inclusion/Exclusion
  Age                               >=18                       >=18                      18-80
  Time frame                      2000-2006                 2000-2006                  3/01-7/07
  Required at least 1-yr             Yes                   Not specified                  Yes
  of follow-up
  Required Medicare as                 No                       No*                       No
  primary payer
  Excluded pts with                   Yes                       Yes                       No
  prior documented
  history of malignancy
  Included only 1st time              Yes                       Yes                       No
  recipients
  Induction Regimen                 IL-2                No TCD Induction                 IL-2
                             (Zenapax, Simulect)                                       (Zenapax,
                                                                                       Simulect)
  PTLD Case
  Identification
  Transplant recipient                Yes                       Yes                       Yes
  follow-up form
  Malignancy form                      No                       Yes                       No
 *Methods section of BMS states that only Medicare subjects were included, however BMS has stated in
 their reply dated January 13, 2010, that this analysis was not restricted to Medicare subjects as claims
 data was not used




                                                                                                       91
13. Appendix C: Personal Communication from Dr. Gerhard Opelz

In response to a submission from the applicant referencing personal communications
from Dr. Opelz regarding rates of PTLD observed among EBV+ kidney transplant
recipients followed in the Collaborative Transplant Study (CTS), FDA contacted Dr
Opelz to clarify the specific rate of non-Hodgkin’s lymphoma (NHL) among EBV+
kidney recipients at 2 years seen in a figure in an article published by Dr Opelz and
CTS. 67 Dr. Opelz clarified that the incidence in the CTS data was 0.12 events per 100
patient-years, as shown in Table 53. The following table is derived from an e-mail from
Dr. Opelz and presented here with his permission.

It should be noted, that the vast majority of the PTLD cases which developed in the
belatacept kidney trials were non-Hodgkin’s lymphomas.

           Table 53: Incidence of NHL (events per 100 patient years) among
          kidney transplant recipients in the CTS database according to EBV
                                         status:
                    Time period                     EBV–                   EBV+

                        Year 1                       1.73                   0.16

                Years 1-2 together                   1.23                   0.12

                 Years 1-3 together                  0.93                   0.12




67
  Opelz G, Daniel V, Naujokat C, et al. Epidemiology of Pretransplant EBV and CMV Serostatus in
Relation to Posttransplant Non-Hodgkin Lymphoma. Transplantation 2009; 88: 962-7.


                                                                                                  92
14. Appendix D: Exploratory Analysis to Identify Additional Risk
Factors for Belatacept Associated PTLD

Higher numbers of PTLD cases were observed in the belatacept-treated arms compared to
the CsA-treated arm in Studies IM103100, IM103008 and IM103027. These results led
to further analysis of the data to identify potential risk factors for belatacept-associated
PTLD using Receiver Operating Characteristic (ROC) analyses and chi-square analysis.

Data from Studies IM103100, IM103008 and IM103027 was used to generate ROC plots.
Figure 10 shows the results of the ROC analyses to identify potential factors associated
with PTLD risk in belatacept-treated patients. EBV serostatus was the strongest risk
factor, followed by CMV serostatus. A combination the two risk factors, EBV negative
serostatus and CMV negative serostatus, increased the predictive value of PTLD risk.
Interestingly, the predictive value of PTLD for the combination of these two factors was
very similar to that obtained with the combination of EBV serostatus and CMV infection
prior to detection of PTLD as well as to NRISK, a combination of potential risk factors,
age category, EBV status, use of T cell depleting agents and CMV infection prior to
detection of PTLD.

         Figure 10: Correlates of PTLD Risk in Belatacept-Treated Patients.




EBV: Epstein-Barr virus serostatus; CMV: Cytomegalovirus serostatus; CMVDIS: CMV infection prior to
PTLD diagnosis; AGECAT: Age category (below or above 60 years); SEX: Female/Male; TCELL:
Lymphocyte depletion prior to PTLD through treatment with T-cell depleting agents; NRISK: A
combination of risk factors; AGECAT, EBV, TCELL and CMV DIS.


                                                                                                 93
Table 54: Results of the ROC Analyses to Identify Potential Factors Associated with
                     PTLD Risk in Belatacept-Treated Patients

Test Result Variable(s)          Area         Std. Error            P-value

EBV                                   .719             .085                     .007
CMV                                   .666             .079                     .039
EBV+CMV                               .806             .065                   <.0001
CMVDIS                                .584             .087                     .296
EBV+CMVDIS                            .794             .070                   <.0001
AGECAT                                .428             .084                     .370
SEX                                   .378             .065                     .132
TCELL                                 .593             .087                     .247
NRISK                                 .799             .071                   <.0001

The results of the chi-square analysis are shown in Table 54. Joint consideration of EBV
and CMV serostatus provided additional information regarding PTLD. Specifically,
among EBV positive patients, those with CMV negative serology are more likely to
develop PTLD than those who are CMV positive (1.91% vs 0.34%), representing an
approximately 6-fold higher risk (OR 5.76 [1.05, 31.7]). Similar imbalance, although to
a lesser extent, was seen for PTLD manifested in the CNS (data not shown).

Table 55: Belatacept PTLD Rates Among EBV+ Patients Based on CMV Serostatus
           PTLD by CMV status in EBV positive belatacept-treated population
                                            PTLD                         Total
                                   No               Yes
CMV                     +          590           2 (0.34%)                592
                        -          205           4 (1.91%)                209
Total                              795           6 (0.75%)                801

The analysis found EBV serostatus (EBV negative recipients) to be the highest risk
predictor of PTLD in belatacept-treated patients. In addition, CMV (CMV negative
recipients) was also identified as a risk factor for PTLD. In combination with EBV,
CMV was identified as an important risk factor for PTLD in belatacept-treated patients
on the basis of ROC analyses and on the imbalance of PTLD in EBV positive, CMV
negative individuals. The results suggest that a combination of EBV and CMV serostatus
could be a better predictor of PTLD than EBV serostatus alone in belatacept-treated
patients.

While EBV status is a well appreciated risk factor for PTLD in patients receiving post-
transplant immunuosuppressants, the role of CMV serology to date has been equivocal.



                                                                                          94
Reports from two trials conducted outside of the United States (Germany and the UK) do
not substantiate the role of CMV in PTLD risk 68, 69,70 whereas data from three trials in the
United States support the hypothesis that CMV status (serostatus or mismatch) is a risk
factor for PTLD. 71,72,73




68
   Opelz G, Daniel V, Naujokat C, et al. Epidemiology of Pretransplant EBV and CMV Serostatus in
Relation to Posttransplant Non-Hodgkin Lymphoma. Transplantation 2009; 88: 962-7.
69
   Buyck et al. Prior immunosuppressive therapy with antithymocyte globulin increases the risk of EBV-
related lymphoproliferative disorder following allo-SCT for acquired aplastic anaemia. Bone Marrow
Transplantation 43:813-6(2009).
70
   Walker et al. Pretransplantation assessment of the risk of lymphoproliferative disorder. Clin Infect Dis
20:1346-53 (1995).
71
   Schubert et al. Diagnosis and treatment of post-transplantation lymphoproliferative disorder in pediatric
heart transplant patients. Pediatr.Transplant 13(1):54-62(2009
72
   Gao et al. Post-transplantation lymphoproliferative disease in heart and heart–lung transplant recipients:
30-Year experience at Stanford University. J Heart Lung Transplant 22(5):505-14(2003).
73
   Katz et al. Case-control study of risk factors for the development of post-transplant lymphoproliferative
disease in a pediatric heart transplant cohort. Pediatr Transplantation 11:58–65(2007).


                                                                                                           95
15. Appendix E: REMS GUIDANCE




                                96
    Guidance for Industry 

     Format and Content of 

 Proposed Risk Evaluation and 

 Mitigation Strategies (REMS), 

    REMS Assessments, and 

 Proposed REMS Modifications 

                                DRAFT GUIDANCE 

       This guidance document is being distributed for comment purposes only.

Comments and suggestions regarding this draft document should be submitted within 90 days of
publication in the Federal Register of the notice announcing the availability of the draft
guidance. Submit comments to the Division of Dockets Management (HFA-305), Food and
Drug Administration, 5630 Fishers Lane, rm. 1061, Rockville, MD 20852. All comments
should be identified with the docket number listed in the notice of availability that publishes in
the Federal Register.

For questions regarding this draft document contact (CDER) Kathleen Frost 301-796-2380, or
(CBER) the Office of Communication, Outreach, and Development (OCOD) at 301-827-1800 or
800-835-4709.


                      U.S. Department of Health and Human Services 

                               Food and Drug Administration 

                     Center for Drug Evaluation and Research (CDER) 

                    Center for Biologics Evaluation and Research (CBER) 


                                        September 2009 

                                          Drug Safety 

Guidance for Industry 

    Format and Content of 

Proposed Risk Evaluation and 

Mitigation Strategies (REMS), 

   REMS Assessments, and 

Proposed REMS Modifications 

                                Additional copies are available from: 

                                      Office of Communications

                          Division of Drug Information, WO51, Room 2201

                              Center for Drug Evaluation and Research

                                   Food and Drug Administration 

                                   10903 New Hampshire Avenue 

                                   Silver Spring, MD 20993-0002

                             Phone: 301-796-3400; Fax: 301-847-8714 

                                        druginfo@fda.hhs.gov 

    http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/default.htm 

                                                and/or

                   Office of Communication, Outreach, and Development (OCOD)

                            Center for Biologics Evaluation and Research 

                                    Food and Drug Administration 

                          1401 Rockville Pike, Rockville, MD 20852-1448 


                              Phone: 800-835-4709 or 301-827-1800

                                   E-mail: ocod@fda.hhs.gov
 http://www.fda.gov/BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/default.htm


                 U.S. Department of Health and Human Services 

                          Food and Drug Administration 

                Center for Drug Evaluation and Research (CDER) 

               Center for Biologics Evaluation and Research (CBER) 

                                  September 2009 

                                    Drug Safety 

                                                      TABLE OF CONTENTS 

I.	        INTRODUCTION............................................................................................................. 1

II.	       BACKGROUND ............................................................................................................... 2

      A.   FDAAA and REMS: Initial Approval and Postapproval Requirements................................. 2

      B.   Relationship Between REMS and RiskMAPs ............................................................................. 3

      C.   Products Deemed to Have in Effect an Approved REMS .......................................................... 4

      D.   Content of a REMS........................................................................................................................ 5

      E.   Assessments and Modifications of Approved REMS.................................................................. 6

      F.   REMS Are Enforceable................................................................................................................. 7

III.	      CONTENT OF A PROPOSED REMS SUBMISSION TO FDA................................. 7

      A.   Content of the Proposed REMS.................................................................................................... 7

      B.   Content of the REMS Supporting Document............................................................................ 16

      C.   Foreign Language REMS............................................................................................................ 21

IV.	       REMS ASSESSMENT AND PROPOSED REMS MODIFICATION 

           SUBMISSIONS TO FDA ............................................................................................... 22

V.	        COMMUNICATING WITH FDA REGARDING REMS.......................................... 23

      A.   Submission Type .......................................................................................................................... 23

      B.   Document Identification.............................................................................................................. 23

      C.   Questions about REMS ............................................................................................................... 26

GLOSSARY................................................................................................................................. 28

ATTACHMENT A: EXAMPLE OF A REMS DOCUMENT .............................................. 30

FOR A FICTITIOUS DRUG..................................................................................................... 30

                                    Contains Nonbinding Recommendations

                                          Draft – Not for Implementation
 1

 2                              Guidance for Industry1

 3               Format and Content of Proposed Risk Evaluation and 

 4                Mitigation Strategies (REMS), REMS Assessments, 

 5                        and Proposed REMS Modifications 

 6

 7

 8
   This draft guidance, when finalized, will represent the Food and Drug Administration's (FDA's) current
 9    thinking on this topic. It does not create or confer any rights for or on any person and does not operate to
10    bind FDA or the public. You can use an alternative approach if the approach satisfies the requirements of
11    the applicable statutes and regulations. If you want to discuss an alternative approach, contact the FDA
12    staff responsible for implementing this guidance. If you cannot identify the appropriate FDA staff, call
13    the appropriate number listed on the title page of this guidance.
14
15
16
17    I.      INTRODUCTION
18
19    This document provides guidance to industry on:
20       • The format and content of a proposed risk evaluation and mitigation strategy (REMS),
21           including REMS supporting documentation;
22       • The content of assessments and proposed modifications of approved REMS;
23       • What identifiers to use on REMS documents; and
24       • How to communicate with FDA about a REMS.
25
26    This guidance applies to certain drug and biological products submitted for approval or approved
27    under sections 505(b) or 505(j) of the Federal Food, Drug, and Cosmetic Act (FDCA), or section
28    351 of the Public Health Service Act (PHS Act), that are required by FDA to have a REMS. The
29    information on the content of a proposed REMS submission (section III of this document) also
30    applies to proposed REMS that are voluntarily submitted by applicants or holders of approved
31    applications (see section II.A of this document).
32
33    This guidance will address REMS elements and provisions that are broadly applicable to
34    proposed REMS and to assessments and modifications of approved REMS. Other provisions,
35    such as those that pertain only to abbreviated new drug applications (ANDAs), or expanded
36    information about REMS assessments and proposed modifications, will not be fully addressed,
37    but will be the subject of future guidance.
38
39    FDA’s guidance documents, including this guidance, do not establish legally enforceable
40    responsibilities. Instead, guidances describe the Agency’s current thinking on a topic and should
41    be viewed only as recommendations, unless specific regulatory or statutory requirements are
      1
       This guidance has been prepared by the FDAAA Title IX Working Group in the Center for Drug Evaluation and
      Research (CDER) and the Center for Biologics Evaluation and Research (CBER) at the Food and Drug
      Administration.
                                       Contains Nonbinding Recommendations
                                            Draft — Not for Implementation
42   cited. The use of the word should in Agency guidances means that something is suggested or
43   recommended, but not required.
44
45   II.        BACKGROUND
46
47   A.         FDAAA and REMS: Initial Approval and Postapproval Requirements
48
49   On September 27, 2007, the President signed into law the Food and Drug Administration
50   Amendments Act of 2007 (FDAAA) (Public Law 110-85).2 Title IX, Subtitle A, section 901 of
51   this statute created new section 505-1 of the FDCA, which authorizes FDA to require persons
52   submitting certain applications (applicants) to submit a proposed REMS as part of such
53   application if the FDA determines that a REMS is necessary to ensure that the benefits of a drug
54   outweigh the risks of the drug.3 Section 505-1 applies to applications for approval of
55   prescription drugs submitted under FDCA subsections 505(b) or (j) and applications submitted
56   under section 351 of the Public Health Service Act. These applications are termed covered
57   applications and refer to new drug applications (NDAs), abbreviated new drug applications
58   (ANDAs), and biologics license applications (BLAs). Please note that the term “drug” is used in
59   this guidance to refer to prescription drug and biologic products for which there are pending or
60   approved applications.
61
62   Section 505-1 also authorizes FDA to require holders of covered applications approved without a
63   REMS to submit a proposed REMS if the FDA becomes aware of new safety information as
64   defined in 505-1(b)(3) and determines that such a strategy is necessary to ensure that the benefits
65   of the drug outweigh the risks of the drug. Once the holder of an approved covered application
66   is notified by FDA that a REMS is necessary, the holder must submit a proposed REMS within
67   120 days, or within such other reasonable time as FDA requires to protect the public health
68   (section 505-1(a)(2)(B)).
69
70   In addition, persons with certain covered applications that were approved before the effective
71   date of Subtitle A, March 25, 2008, were deemed to have in effect an approved REMS and were
72   also required to submit a proposed REMS. See section II.C of this document, Products Deemed
73   to Have in Effect an Approved REMS.
74
75   An applicant may voluntarily submit a proposed REMS without having been required to do so by
76   FDA. For instance, without having been notified by FDA to submit a proposed REMS, an
77   applicant may include a proposed REMS in an original application or in a supplemental
78   application, or in an amendment to an existing original or supplemental application, if the
79   applicant believes a REMS would be necessary to ensure that the benefits of the drug outweigh
80   its risks and the other relevant statutory criteria in section 505-1 are met. Section V of this
81   document describes submission types and document identification. If FDA determines that a

     2
       See
     http://www fda.gov/RegulatoryInformation/Legislation/FederalFoodDrugandCosmeticActFDCAct/SignificantAmen
     dmentstotheFDCAct/FoodandDrugAdministrationAmendmentsActof2007/default.htm.
     3
         Subtitle A took effect on March 25, 2008, 180 days after enactment of FDAAA.



                                                             2

                                     Contains Nonbinding Recommendations
                                          Draft — Not for Implementation
 82   REMS is necessary to ensure that the benefits of the drug outweigh the risks, FDA will
 83   determine which elements of a REMS are necessary and will approve the REMS once the
 84   Agency has determined that the proposed REMS will ensure that the benefits of the drug
 85   outweigh the risks, and the other relevant statutory criteria in section 505-1 are met. An
 86   approved REMS that was voluntarily submitted is subject to the same requirements and
 87   enforcement as a REMS that was originally submitted as a required proposed REMS. If an
 88   applicant voluntarily submits a proposed REMS, it will not be approved as a REMS unless and
 89   until the FDA determines that it is required to ensure that the benefits of the drug outweigh the
 90   risks and that it meets the FDAAA criteria. Proposed REMS that are not approved are not
 91   subject to the requirements and enforcement of an approved REMS. FDA will notify applicants
 92   who voluntarily submit a proposed REMS whether the REMS will be required. If the FDA
 93   determines that a REMS is not required, an applicant may undertake voluntary risk management
 94   measures that would be performed outside of a REMS.
 95
 96   B.         Relationship Between REMS and RiskMAPs
 97
 98   Before FDAAA was enacted, FDA approved a small number of drug and biological products
 99   with risk minimization action plans (RiskMAPs). A RiskMAP is a strategic safety program
100   designed to meet specific goals and objectives in minimizing known risks of a product while
101   preserving its benefits. RiskMAPs were developed for products that had risks that required
102   additional risk management strategies beyond describing the risks and benefits of the product in
103   labeling and performing required safety reporting. For the majority of approved products,
104   labeling and routine reporting requirements are sufficient to mitigate risks and preserve benefits.
105   In a small number of cases, when additional measures were needed to ensure that the benefits of
106   a drug outweigh the risks of the drug, FDA approved the drug with a RiskMAP. In 2005, FDA
107   issued a guidance for industry on Development and Use of Risk Minimization Action Plans4 (the
108   RiskMAP guidance), that described how to develop RiskMAPs, select tools to minimize risks,
109   evaluate and monitor RiskMAPs and monitoring tools, and communicate with FDA about
110   RiskMAPs.
111
112   Now that FDAAA has given FDA the authority to require REMS when necessary to ensure that
113   the benefits of a drug outweigh the risks, FDA anticipates that:
114
115          •    A product that would previously have been approved with a RiskMAP will, instead, be
116               approved with a REMS if statutory requirements for a REMS are met.5
117          •    Products that would previously have been approved with a Medication Guide or patient
118               package insert that meet the statutory requirements for a REMS will now be required to
119               have a REMS.
120          •    While certain products approved with RiskMAPs that included certain types of risk
121               management tools have been deemed to have in effect an approved REMS (see section
122               II.C of this document), all other approved RiskMAPs and approved Medication Guides
123               and patient package inserts that were in place when Subtitle A took effect will continue
124               to be in effect, unless they are replaced by or included in a REMS. They will be

      4
          http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM071616.pdf
      5
          Unless it is an ANDA based on a reference listed drug with an approved RiskMAP.


                                                          3

                                     Contains Nonbinding Recommendations
                                          Draft — Not for Implementation
125                replaced by or included in a REMS if FDA determines, based on new safety
126                information identified after approval of the product, that a REMS is necessary to ensure
127                that the benefits of the drug outweigh the risks.
128          •     ANDAs for which the reference listed drug has an approved RiskMAP will be approved
129                with a comparable RiskMAP that includes the same essential elements.
130          •     ANDAs for which the reference listed drug has a REMS will be approved with the
131                elements of that REMS applicable to ANDAs.
132          •     Revisions of existing Medication Guides or patient package inserts that meet REMS
133                requirements will be approved as part of a REMS.
134
135   Many of the principles that were included in the RiskMAP guidance are embodied in the
136   FDAAA REMS provisions as implemented by FDA. Many of those principles pertaining to
137   REMS are included in this guidance, and others will be included in future guidance documents
138   related to REMS. The RiskMAP guidance continues to apply to products with existing
139   RiskMAPs (e.g., products with RiskMAPs that were not deemed to have in effect an approved
140   REMS) and to products with new RiskMAPs (e.g., ANDAs for which the reference listed drug
141   has a RiskMAP).
142
143   C.         Products Deemed to Have in Effect an Approved REMS
144
145   Section 909(b)(1) of FDAAA addresses products approved before the effective date of Subtitle A
146   that have been deemed to have in effect an approved REMS.
147
148              A drug that was approved before the effective date of this Act is . . . deemed to
149              have in effect an approved risk evaluation and mitigation strategy under section
150              505-1 of the Federal Food, Drug, and Cosmetic Act . . . if there are in effect on
151              the effective date of this Act elements to assure safe use—
152                      (A) required under section 314.520 or section 601.42 of title 21, Code of
153                      Federal Regulations; or
154                      (B) otherwise agreed to by the applicant and the Secretary for such drug.
155
156   Section 909(b)(2) states that the REMS for a drug deemed to have an approved REMS consists
157   of the timetable required under section 505-1(d) and any additional elements under subsections
158   505-1(e) and (f) in effect for the drug on the effective date of FDAAA.
159
160   Section 909(b)(3) of FDAAA states:

161              Not later than 180 days after the effective date of this Act, the holder of an
162              approved application for which a risk evaluation and mitigation strategy is
163              deemed to be in effect . . . shall submit to the Secretary a proposed risk
164              evaluation and mitigation strategy. Such proposed strategy is subject to section
165              505-1 of the Act as if included in such application at the time of submission of
166              the application to the Secretary.6
167

      6
          121 Stat. 951.


                                                         4

                                     Contains Nonbinding Recommendations
                                           Draft — Not for Implementation
168   On March 27, 2008, FDA published in the Federal Register a list of drugs that were identified as
169   deemed to have an approved REMS, and directed holders of approved applications for those
170   products to submit a proposed REMS by September 21, 2008.7 For most of these drugs, the
171   elements of the existing RiskMAPs or restricted distribution and risk management programs
172   were or will be simply converted to the new content and format of a REMS in the proposed
173   REMS. FDA generally does not intend to make substantial changes to these programs during
174   this conversion unless new safety or effectiveness information identified since the drug was
175   approved (including an evaluation of the program identifying deficiencies) suggests that the
176   existing REMS should be modified to ensure that the benefits of the product outweigh the risks.
177   In those cases, FDA has or will require modifications to the REMS.
178
179   D.      Content of a REMS
180
181   A REMS for an NDA or BLA product must have a timetable for submission of assessments of
182   the REMS (505-1(d)). In addition, a REMS may include any or all of the other REMS elements,
183   if specified criteria are met. These additional elements are listed below and described in more
184   detail in section III of this document:
185
186         1. Timetable for Submission of Assessments
187
188         Section 505-1(d) requires that all approved REMS for NDA and BLA products have a
189         timetable for submission of assessments of the REMS. FDAAA specifies that the timetable
190         for submission of assessments of the REMS must include an assessment by the dates that
191         are 18 months and 3 years after the strategy is approved, and an assessment in the 7th year
192         after the strategy is approved, or at another frequency specified in the strategy (see section
193         III.A.6 of this document for additional information).
194
195         2. Additional Potential Elements
196
197         Section 505-1(e) lists “Additional Potential Elements” of a REMS that may include the
198         following (see section III.A.3 of this document for additional information):
199
200               •    A Medication Guide as provided for under part 208 of title 21, Code of Federal
201                    Regulations
202               •    A patient package insert if such insert may help mitigate a serious risk of the drug
203               •    A communication plan to health care providers if the plan may support
204                    implementation of an element of the strategy
205
206         3. Elements to Ensure Safe Use (ETASU)
207




      7
       See Federal Register Notice “Identification of Drugs and Biological Products Deemed to Have Risk Evaluation
      and Mitigation Strategies (REMS) for Purposes of the Food and Drug Administration Amendments Act of 2007”
      (73 FR 16313, March 27, 2008).


                                                            5

                                      Contains Nonbinding Recommendations
                                            Draft — Not for Implementation
208         Section 505-1(f)8 lists certain Elements to Assure Safe Use that may be required if the drug
209         has been shown to be effective, but is associated with a serious adverse event and can be
210         approved only if, or would be withdrawn unless, such elements are required as part of a
211         strategy to mitigate the specific serious risk(s) listed in the labeling of the product.
212         Elements to assure safe use may be required for approved products when an assessment
213         and Medication Guide, patient package insert, or communication plan are not sufficient to
214         mitigate these risks. The elements to assure safe use must include one or more goals to
215         mitigate the specific serious risk(s). If a REMS includes certain elements to assure safe
216         use, the REMS may also include required implementation systems to enable the applicant
217         to monitor, evaluate, and improve the implementation of the elements (see section III.A.4
218         of this document for additional information).
219
220   This guidance document uses the word tool to describe a process or system designed to
221   implement one or more REMS elements. In some cases, an element itself, such as a Medication
222   Guide, may be viewed as a tool. In other cases, such as for an ETASU that requires that a drug
223   be dispensed to patients with evidence or other documentation of safe-use conditions (505-
224   1(f)(3)(D)), specific tools are used to implement a REMS element; for example, systems to
225   ensure that certain laboratory test result outcomes are obtained before a drug may be dispensed.
226
227   E.      Assessments and Modifications of Approved REMS
228
229   FDAAA includes provisions for the assessment and modification of an approved REMS in
230   section 505-1(g). Additional information on assessments and modifications is included in
231   sections III.B.4 and IV of this document.
232
233           1. Voluntary Assessments and Proposed Modifications (505-1(g)(1) and (4))
234
235                In addition to required assessments of an approved REMS described below, an
236                applicant may voluntarily submit an assessment of, and propose modifications to, an
237                approved REMS at any time. Proposed modifications may enhance or reduce the
238                approved REMS, and may include additions to or modifications of the timetable for
239                submission of assessments, including a proposal to eliminate assessments, and/or the
240                addition, modification, or removal of a Medication Guide, patient package insert,
241                communication plan or ETASUs.
242
243           2. Required assessments (505-1(g)(2))
244
245                REMS assessments are required under the following circumstances:
246
247                • When submitting a supplemental application for a new indication for use, unless
248                  the approved REMS for the drug includes only a timetable for submission of
249                  assessments. FDA anticipates rarely requiring a REMS that includes only a
250                  timetable for submission of assessments.
      8
       FDA is considering the implications of section 505-1(f) on the restricted distribution provisions under 21 CFR 314
      Subpart H (drugs) – 314.520, and 21 CFR 601 Subpart E (biologics) – 601.42 and will address this in a future
      guidance.


                                                              6

                                      Contains Nonbinding Recommendations
                                            Draft — Not for Implementation
251                • When required by the approved REMS, as provided for in the timetable for
252                  submission of assessments
253                • When required by the FDA, within a time period to be determined by the FDA, if
254                  the FDA determines that new safety or effectiveness information indicates that the
255                  timetable for submission of assessments should be modified and/or that a
256                  Medication Guide, patient package insert, communication plan, or ETASUs should
257                  be added, modified, or removed
258                • Within 15 days when ordered by the FDA, if the FDA determines that there may
259                  be a cause for withdrawal or suspension of approval under section 505(e) of the
260                  FDCA
261
262       F.   REMS Are Enforceable
263
264   REMS required under section 505-1 are subject to inspection and are enforceable under the
265   FDCA as amended by FDAAA.9 A drug is misbranded under section 502(y) if the responsible
266   person for that drug10 fails to comply with a requirement of the approved strategy. Also, under
267   section 303(f)(4)(A) of the FDCA, a responsible person who violates a REMS requirement is
268   subject to civil monetary penalties of up to $250,000 per violation, not to exceed $1 million in a
269   single proceeding. These penalties increase if the violation continues more than 30 days after
270   FDA notifies the responsible person of the violation. The penalties double for the second 30-day
271   period, and continue to double for subsequent 30-day periods, up to $1 million per period and
272   $10 million per proceeding. In imposing a monetary penalty, FDA will consider the responsible
273   person’s efforts to correct the violation. In addition, under 505(p), a person may not introduce or
274   deliver for introduction into interstate commerce an approved drug that is the subject of a
275   covered application, if a REMS is required with respect to that drug, and the person fails to
276   maintain compliance with the requirements of the approved REMS or with other requirements
277   under 505-1, such as requirements regarding assessments of approved REMS.
278
279
280   III.     CONTENT OF A PROPOSED REMS SUBMISSION TO FDA
281
282   A proposed REMS submission to FDA should include two parts: a proposed REMS, which is a
283   concise document that describes the proposed goals and elements of the REMS and, once
284   approved, will be the basis for enforcement; and a REMS supporting document, that expands on
285   information included in the proposed REMS and provides additional information not included in
286   the proposed REMS, including a thorough explanation of the rationale for, and supporting
287   information about, the content of the proposed REMS. These two parts of a proposed REMS
288   submission are described below.
289
290   A.       Content of the Proposed REMS
291
292   The proposed REMS should include concise information describing the goal(s) of the REMS and
293   the REMS element(s) proposed for inclusion in the approved REMS for the specified product.
      9
       See FDAAA Title IX, section 902.

      10
        The term ‘responsible person’ means the person submitting a covered application or the holder of the approved

      such application. Section 505-1(b)(7). 



                                                              7

                                 Contains Nonbinding Recommendations
                                     Draft — Not for Implementation
294   All proposed materials that are included as part of the REMS (e.g., proposed communication and
295   education materials, Medication Guide, patient package insert, enrollment forms, prescriber and
296   patient agreements) should be appended to the proposed REMS. The proposed REMS should be
297   written to clearly describe the responsibilities of the applicant in implementing the REMS; for
298   example, statements will generally begin with, “[Name of the applicant] will…” The proposed
299   REMS should include the date by which each of the REMS elements will be implemented.
300
301   A template for the proposed REMS is available on the FDA’s “Postmarket Drug Safety
302   Information for Patients and Providers” Web site, at
303   http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProvider
304   s/default.htm. Attachment A provides an example of a completed proposed REMS for a
305   fictitious product that an applicant would submit to FDA for review. The preferred template may
306   be periodically updated as we gain more experience with REMS; therefore, applicants should
307   check the Web site for the latest version. Questions should be directed to the FDA contacts
308   described in section V.C of this document.
309
310   Prior to approving a REMS, FDA may require applicants to revise a proposed REMS to ensure
311   that the benefits of the drug will outweigh the risks.
312
313   FDA will append any REMS materials that will be included in the approved REMS, as described
314   above, to the final REMS. The final REMS and appended documents will be referenced in and
315   appended to the approval letter for the application or supplement that contains the proposed
316   REMS, and the approval letter and appended documents will be posted on the following FDA
317   Web sites:
318
319   For products regulated by CDER:
320
321      •   The Drugs@FDA Web site at http://www.accessdata.fda.gov/scripts/cder/drugsatfda/.
322      •   The Postmarket Drug Safety Information for Patients and Providers Web site
323          (http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsand
324          Providers/default.htm). This Web site also includes a list of approved REMS
325          (http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsand
326          Providers/ucm111350.htm). The list of approved REMS includes links to the REMS
327          document and REMS materials, excluding Medication Guides.
328      •   Medication Guides can be accessed on the Drugs@FDA Web site and on the Postmarket
329          Drug Safety Information for Patients and Providers Web site through the link to approved
330          Medication Guides (http://www.fda.gov/Drugs/DrugSafety/ucm085729.htm).
331
332   For products regulated by CBER:
333
334      •   The Biologics Products and Establishments Web site at
335          http://www.fda.gov/BiologicsBloodVaccines/ucm121134.htm
336      •   The Postmarket Drug Safety Information for Patients and Providers Web site (see link
337          above)
338



                                                    8

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
339   The elements of an approved REMS are enforceable under FDAAA, Title IX, section 902 (see
340   section II.F of this document), and any changes to the REMS, including to the appended
341   documents, must be submitted as a proposed modification of an approved REMS and approved
342   by FDA before being implemented (see section IV).
343
344   The proposed REMS should contain the following sections as appropriate to manage the risks of
345   the particular product; if an applicant is not proposing one of the elements, the proposed REMS
346   should include a statement that the element is not necessary.
347
348          1.    Product and Contact Information
349
350   The proposed REMS should include the application number, proprietary and established names,
351   dosage form of the product, the drug class as described in the product’s label, and the applicant’s
352   name and address. The proposed REMS should also include contact information, including
353   position titles, for those responsible for the REMS policy, management, and implementation.
354
355          2.    Goals
356
357   All REMS should include a statement of one or more overall goals. In addition, if the REMS has
358   one or more elements to assure safe use (505-1(f)), the REMS must include one or more goals to
359   mitigate a serious risk listed in the labeling of the drug for which the ETASUs are required.
360   Even when ETASUs are not part of a REMS (e.g., a REMS with a Medication Guide or
361   communication plan only), the goals of the REMS should be identified. Assessments of
362   approved REMS should measure whether the goals are being met.
363
364   As used in this document, a proposed REMS goal is the desired safety-related health outcome or
365   the understanding by patients and/or health care providers of the serious risks targeted by the use
366   of specified REMS elements. REMS goals should target the achievement of particular health
367   outcomes or knowledge related to known safety risks and should be stated in a way that aims to
368   achieve maximum risk reduction. The following are examples of REMS goals: “Patients taking
369   W drug should be aware of the serious risks relative to the potential benefits,” “Patients on X
370   drug should not also be prescribed Y drug,” or “Fetal exposures to Z drug should not occur.”
371   Goals should be stated in absolute terms. Although it might not be possible to ensure that the
372   goal can be met for every patient (i.e., no one on X drug receives Y drug), FDA believes that a
373   goal, as the term implies, is a statement of the ideal outcome of a REMS.
374
375   REMS goals should be associated with pragmatic, specific, and measurable program objectives
376   that result in processes or behaviors leading to achievement of the REMS goals. Objectives can
377   be thought of as intermediate steps to achieving the overall REMS goal. A REMS goal can be
378   associated with more than one objective, depending upon the frequency, type, and severity of the
379   specific risk or risks being minimized. For example, a goal may be the elimination of
380   occurrences of a serious adverse event caused by an interaction of the drug with another drug.
381   The objectives could include lowering physician co-prescribing rates and/or pharmacist co-
382   dispensing rates for the specific drugs.




                                                      9

                            Contains Nonbinding Recommendations
                                 Draft — Not for Implementation
383
384   3.     Additional Potential REMS Elements
385
386         (a) Medication Guide and/or Patient Package Insert
387
388   As one element of a REMS, the FDA may require the development of a Medication
389   Guide, as provided for under 21 CFR part 208, which sets forth requirements for patient
390   labeling for human prescription drug products, including biological products, that the FDA
391   determines pose a serious and significant public health concern requiring the distribution
392   of FDA-approved patient information. Medication Guides will be required if the FDA
393   determines that one or more of the following circumstances exist:

394   (1)     The drug product is one for which patient labeling could help prevent serious
395           adverse effects.

396   (2)     The drug product is one that has serious risks (relative to benefits) of which patients
397           should be made aware because information concerning the risks could affect
398           patients’ decision to use, or to continue to use, the product.

399   (3)     The drug product is important to health and patient adherence to directions for use
400           is crucial to the drug's effectiveness.

401   Under 21 CFR part 208 and in accordance with 505-1 of the FDCA, the applicant is
402   responsible for ensuring that the Medication Guide is available for distribution to patients
403   who are dispensed the drug. This section of the REMS should describe the mechanisms
404   the applicant intends to use for distribution of the Medication Guide.

405   In addition, FDA may require a patient package insert as part of a REMS if the FDA
406   determines that the patient package insert may help mitigate a serious risk of the drug.
407   Having both a required patient package insert and a Medication Guide for the same drug
408   is not expected to occur frequently. In most instances, FDA anticipates requiring a
409   Medication Guide (or requiring conversion of an existing PPI to a Medication Guide) if FDA is
410   requiring patient labeling that meets Medication Guide requirements.
411
412   The following types of changes to a PPI would not ordinarily trigger the need to convert
413   a PPI to a Medication Guide:
414
415   •     Editorial changes
416   •     Changes related to how to use a product (e.g., how to inject the product
417         subcutaneously) unless these changes have the potential to mitigate a serious risk,
418         such as overdose
419
420   Copies of Medication Guides and patient package inserts that are part of a REMS should
421   be appended to the proposed REMS.
422
423         (b) Communication Plan
424


                                                10

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
425   FDA may determine that a communication plan targeted at health care providers is a necessary
426   element of a REMS if it may support implementation of the REMS. The communication plan
427   may include sending letters to health care providers; disseminating information about REMS
428   elements to encourage implementation by health care providers or to explain certain safety
429   protocols, such as medical monitoring by periodic laboratory tests; or disseminating information
430   to health care providers through professional societies about any serious risks of the drug and
431   any protocol to assure safe use (section 505-1(e)(3)).
432
433   Copies of communication plan materials should be appended to the proposed REMS.
434
435   If an NDA has been approved with a REMS with a communication plan, and subsequently an
436   abbreviated new drug application (ANDA) is approved with that NDA product as the reference
437   listed drug, then FDA must undertake the communication plan (section 505-1(i)(2)(A)). Neither
438   the holder of the NDA that is the reference listed drug nor the ANDA holder has to undertake a
439   communication plan once an ANDA is approved. However, many tools that have previously
440   been considered part of a communication plan, such as training materials, specified procedures,
441   patient/physician agreements or other informed consent, patient educational materials, safety
442   protocols, medical monitoring procedures, and data collection forms may fit under one or more
443   elements to assure safe use (ETASU) if specified criteria are met. Both NDA holders and
444   ANDA holders are required to implement ETASUs.
445
446          4.    Elements to Assure Safe Use
447
448   Elements to assure safe use are intended to provide safe access for patients to drugs with known
449   serious risks that would otherwise be unavailable. Required ETASUs are put in place to mitigate
450   a specific serious risk listed in the labeling of a drug. Before requiring one or more ETASUs, the
451   FDA must make the following determinations (505-1(f)(1)):
452
453          •    That the drug, which has been shown to be effective but is associated with a serious
454               adverse drug experience, can be approved only if, or would be withdrawn unless,
455               such elements were required; and
456          •    That for a drug initially approved without ETASUs, other possible elements of a
457               REMS are not sufficient to mitigate such serious risk.
458
459   This subsection of the proposed REMS should describe the ETASUs included in the proposed
460   REMS and any tools designed to implement one or more elements to assure safe use. Copies of
461   all relevant materials should be appended to the proposed REMS. Examples of relevant
462   materials include health care provider attestations; pharmacy, practitioner, health care setting,
463   and patient enrollment forms; training materials; specified procedures; patient/physician
464   agreements or other informed consent; patient educational materials; safety protocols; medical
465   monitoring procedures; and data collection forms.
466
467   The following lists the elements to assure safe use that may be included in the REMS. Note that
468   some of the tools designed to implement the elements to assure safe use may appear in more than
469   one category:
470


                                                      11

                              Contains Nonbinding Recommendations
                                   Draft — Not for Implementation
471   A. Health care providers who prescribe the drug have particular training or experience, or
472      are specially certified.
473
474      In general, section 505-1(f)(3)(A) pertains to prescribers of the drug. Elements under this
475      category might require certification of training, or attestation of specific experience or
476      knowledge, before the health care provider is enrolled in a program that allows that
477      provider to prescribe the product.
478
479      For example, in order to be certified, a health care provider may be required to
480      demonstrate that he or she:
481
482              •   Can diagnose the condition for which the product is indicated
483              •   Understands the risks and benefits of the product and has read the educational
484                  materials for prescribers
485              •   Can diagnose and treat potential adverse reactions associated with the product
486
487      The program may require periodic recertification and reenrollment.
488
489      The opportunity to obtain this training or certification must be available to any willing
490      provider, for example through an on-line or mail course, at reasonable cost to the
491      provider (505-1(f)(3)(A)).
492
493   B. Pharmacies, practitioners, or health care settings that dispense the drug are specially
494      certified.
495
496      In general, section 505-1(f)(3)(B) pertains to how the drug is dispensed. Elements under
497      this category might require certification of training or attestation of specific experience or
498      knowledge before the pharmacy, practitioner, or health care setting is enrolled in a
499      program that allows the practitioner or staff at the pharmacy or health care setting to
500      dispense the product.
501
502      For example, to be certified, practitioners and staff at pharmacies, hospitals, and infusion
503      sites may be required to demonstrate that they:
504
505              •   Understand the risks and benefits of the product and have read the educational
506                  materials before the drug is dispensed
507              •   Agree to fill a prescription and dispense the drug only after receiving prior
508                  authorization
509              •   Agree to check laboratory values, or check for the presence of stickers that
510                  providers affix to prescriptions for specified products to indicate that the
511                  patient has met all criteria for receiving the product (“qualification stickers”),
512                  before dispensing a drug
513              •   Agree to fill a prescription and dispense the drug only within a specified
514                  period of time after the prescription is written
515              •   Agree to fill prescriptions only from enrolled prescribers
516


                                                   12

                              Contains Nonbinding Recommendations
                                   Draft — Not for Implementation
517      The program may require periodic recertification and reenrollment.
518
519      The opportunity to obtain this certification must be available to any willing provider
520      (505-1(f)(3)(B)).
521
522   C. The drug be dispensed to patients only in certain health care settings, such as hospitals.
523
524      In general, section 505-1(f)(3)(C) pertains to restrictions on dispensing the product to
525      patients in specific health care settings.
526
527      For example, the applicant may be required to
528
529              •   Ensure the drug is dispensed only to patients in hospitals that have met
530                  certain conditions
531              •   Ensure the drug is dispensed only to physicians’ offices equipped to treat the
532                  potential risks associated with the drug following administration of the drug
533                  (e.g., access to medication and equipment necessary to treat a serious allergic
534                  reaction)
535
536   D. The drug be dispensed only to patients with evidence or other documentation of safe-use
537      conditions, such as laboratory test results.
538
539      In general, section 505-1(f)(3)(D) pertains to ensuring that patients meet specified criteria
540      before drug exposure.
541
542      For example, evidence or other documentation of safe use conditions may include the
543      following:
544
545              •   Patients have been counseled about the risks and benefits of the product and
546                  have signed an acknowledgment that they understand the risks and benefits of
547                  the product
548              •   Patients have been provided a copy of patient educational materials and
549                  demonstrated that they understand the risks and benefits of the product
550              •   Patients receive drug only after specified authorization is obtained and
551                  verified by the pharmacy. Examples of authorizations include checking
552                  laboratory values and checking for physician qualification (stickers) on the
553                  prescription
554
555   E. Each patient using the drug be subject to certain monitoring.
556
557      Elements under 505-1(f)(3)(E) might require that patients be monitored or that specific
558      follow-up should occur at specific time points.
559
560      Examples include the following:
561



                                                   13

                                      Contains Nonbinding Recommendations
                                            Draft — Not for Implementation
562                      •   Patients’ laboratory tests are monitored on a specified periodic basis to
563                          prevent the serious risk
564                      •   Patients are required to contact the prescriber within a specified period of time
565                          after beginning treatment with the drug to ensure they are still appropriate
566                          candidates for treatment
567                      •   Patients are required to contact their prescriber periodically during and
568                          following treatment to ensure they did not experience the serious risk
569                          associated with the use of the drug
570
571        F. Each patient using the drug be enrolled in a registry.
572
573           In general, section 505-1(f)(3)(F) pertains to enrolling patients into a program as part of
574           an overall strategy to mitigate a specific serious risk listed in the labeling of the drug.
575           The use of a registry may be combined with other ETASUs, such as when a registry is
576           used to document that the drug is dispensed to patients with evidence or other
577           documentation of safe-use conditions; or to document that each patient using the drug is
578           subject to certain monitoring.
579
580           Drug access may be contingent on patient enrollment. The types of information that may
581           be collected on enrolled patients include:
582
583                •     Information on clinical outcomes
584                •     Clinical and laboratory data
585                •     Safety information
586                •     Data on compliance with prescribed management and prescribing protocols
587                •     Data on the impact of tools on ensuring compliance and outcomes
588
589           Registries that are established with the primary purpose of enrolling patients to mitigate a
590           serious risk associated with a drug would be required under a REMS. Registries may
591           also serve as a repository for clinical data and allow for case finding and follow-up.
592           These registries are not considered PMRs, but studies conducted using the data may be.11
593
594           5.       Implementation System
595
596   Section 505-1(f)(4) of the FDCA gives the FDA authority to require an implementation system
597   for a REMS that includes the ETASUs described under 505-1(f)(3)(B), (C), and (D). Through
598   the implementation system, the applicant may be expected to take reasonable steps to monitor
599   and evaluate implementation by health care providers, pharmacists, and other parties in the
600   health care system who are responsible for implementing those elements, and to work to improve
601   their implementation.
602


      11
        See the draft guidance for industry on Postmarketing Studies and Clinical Trials — Implementation of Section
      505(o) of the Federal Food, Drug, and Cosmetic Act, available on the Internet at
      http://www fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/default htm.


                                                              14

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
603   FDA may require the implementation system to include a description of how applicable products
604   will be distributed. In addition, as part of the implementation system, FDA may require the
605   certification of wholesalers and/or distributors who distribute the product to ensure that the
606   product is distributed only to certified or otherwise specified pharmacies, practitioners, or health
607   care settings that dispense the drug, or only to patients who meet the requirements of the REMS.
608
609   Other examples of methods used to monitor and evaluate implementation of REMS with
610   ETASUs described under 505-1(f)(3)(B), (C), and (D) include the following:
611
612   •   The applicant maintains a validated and secure database of all certified entities (pharmacies,
613       practitioners, and health care settings) to ensure any certification requirements or other
614       requirements for pharmacies, practitioners, or health care settings are met
615   •   The applicant conducts periodic audits of pharmacies, practitioners, and health care settings
616       to ensure compliance with ETASUs (e.g., documentation of safe-use conditions prior to
617       dispensing drug)
618   •   If the ETASUs include limits on where and how a drug may be dispensed, the applicant
619       conducts periodic audits of wholesale shipment or distribution systems to determine that the
620       drug is only being distributed to authorized entities
621
622          6.    Timetable for Submission of Assessment of the REMS
623
624   This subsection of the proposed REMS should describe the proposed timetable for submission of
625   assessments of the REMS as required by section 505-1(d) of the FDCA. REMS for NDAs and
626   BLAs must include a timetable for submission of assessments of the REMS. REMS for ANDAs
627   do not include a timetable for submission of assessments. Additional information on REMS and
628   ANDAs will be included in future guidance.
629
630   Under section 505-1(d), each timetable for submission of assessments of a REMS must at a
631   minimum include assessments submitted by 18 months and by 3 years after the REMS is initially
632   approved, and in the 7th year after the REMS is initially approved, with additional dates if more
633   frequent assessments are necessary to ensure that the benefits of the drug continue to outweigh
634   the risks. Factors that may influence the need for more frequent assessments of the REMS
635   include, among others, the estimated size of the population likely to use the drug, the seriousness
636   of known or potential risks that may be related to the drug, and knowledge about the
637   effectiveness of REMS elements to mitigate the risks. The requirements for the assessments
638   submitted by 18 months and by 3 years may be met through assessments submitted at specified
639   earlier dates; for example, assessments required in an approved REMS to be submitted at 12
640   months and 24 months would meet the requirements for the assessments submitted by 18 months
641   and 3 years.
642
643   The timetable specifies when the assessment will be submitted to FDA, not when the assessment
644   will be performed. This subsection should specify the interval that each assessment will cover
645   and the planned date of submission to the FDA of the assessment. To facilitate inclusion of as
646   much information as possible while allowing reasonable time to prepare the submission, the
647   reporting interval covered by each assessment should conclude no earlier than 60 days before the
648   submission date for that assessment. For example, the reporting interval covered by an


                                                       15

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
649   assessment that is to be submitted by July 31 should conclude no earlier than June 1. The
650   assessment is to be received by the FDA on or before the due date.
651
652   Requests for modification of the timetable for submission of assessments, including eliminating
653   assessments, may be made after approval of the REMS (see 505-1(g)(4)). After the assessment
654   due by 3 years after the REMS is initially approved is submitted, all further assessments,
655   including the 7th-year assessment, may be eliminated if the FDA determines that serious risks of
656   the drug have been adequately identified and assessed and are being adequately managed.
657
658   B.    Content of the REMS Supporting Document
659
660   The REMS supporting document should provide a thorough explanation of the rationale for and
661   supporting information about the content of the proposed REMS. A template for the REMS
662   supporting document is available on the FDA’s “Postmarket Drug Safety Information for
663   Patients and Providers” Web site, at
664   http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProvider
665   s/default.htm. The REMS supporting document should include the sections listed in the template
666   for the applicable proposed REMS elements for the specified product, as well as a table of
667   contents. The REMS supporting document should include a description of how and when each
668   REMS element will be implemented and should specify the rationale for the overall timelines
669   and milestones. If any REMS activity will not be implemented at the time of REMS approval,
670   the REMS supporting document should include the rationale for the implementation schedule.
671   For example, the document should address the rationale for whether a communication plan
672   would be implemented before, or concurrently with, other elements. Additional information on
673   each section of the REMS supporting document is described below.
674
675          1.    Background
676
677   The Background section of the REMS supporting document should explain why a REMS is
678   necessary and provide a concise summary of how the proposed REMS would ensure that the
679   benefits of the drug outweigh the risks. For a new REMS that is proposed for an already-
680   approved product, the Background section should also include the description of the new safety
681   information that suggests a REMS is necessary.
682
683   The Background section should describe what is known about the risk to be minimized by the
684   REMS, including the magnitude, severity, and frequency of the adverse events, whether there are
685   particular populations at risk, the background incidence of the risk in the population likely to use
686   the product, whether the adverse event can be prevented or is reversible, and the benefits that
687   would be preserved by the implementation of the REMS. It should also describe the factors that
688   FDA considers when determining whether a REMS is necessary to ensure that the benefits of the
689   drug outweigh the risks: the estimated size of the population likely to use the product, the
690   seriousness of the disease or condition that is to be treated with the product, the expected benefit
691   of the product with respect to such disease or condition, the expected or actual duration of
692   treatment with the drug, the risks and benefits of alternative therapies, and whether the drug is a
693   new molecular entity. The statute specifically requires these factors to be considered for REMS




                                                       16

                                   Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
694   required at initial approval (505-1(a)(1)), but FDA will also consider these factors in making
695   determinations about postapproval REMS.
696
697   The Background section of the REMS supporting document should include a discussion, if
698   pertinent, about the successes and failures of actions by regulatory authorities, systems of health
699   care, or applicants in mitigating the risks of concern for this product or similar products.
700   Information on risk management plans submitted to other regulators, such as the European
701   Union’s EU Risk Management Plan,12 should be included, with a clear description of how that
702   information supports the proposed REMS, along with reasons for any differences between the
703   proposed REMS and other risk management plans for the product.
704
705   Information provided by the applicant regarding relevant past experiences, domestically or in
706   other countries, will assist in the development of REMS that are compatible with established
707   distribution, procurement, and dispensing systems within the health care delivery system, and
708   that avoid the cost of implementing REMS tools already determined to be unsuccessful. In
709   addition, we encourage applicants to provide applicable information or evaluations from past
710   experiences with products or programs that are similar to the proposed REMS. Brief
711   descriptions of the available evidence regarding the effectiveness of each element and tool
712   included in the proposed REMS may be mentioned in the Background section. Thorough
713   descriptions should be included in the “Supporting Information on Proposed REMS Elements”
714   section.
715
716          2.    Goals Section
717
718   This section of the REMS supporting document should describe the rationale for the proposed
719   goals of the REMS and summarize how each proposed element and stated objectives will
720   individually and collectively contribute to achieving the goals. All REMS should include a
721   statement of one or more overall goals. In addition, if the REMS has one or more elements to
722   assure safe use (505-1(f)), the REMS must include one or more goals to mitigate a serious risk
723   listed in the labeling of the drug for which the elements to assure safe use are required. Even if a
724   REMS does not contain elements to assure safe use (e.g., a REMS that includes a Medication
725   Guide or communication plan only), the goals of the REMS should be identified. Additional
726   information about how each particular element and tool will contribute to achieving the goals of
727   the REMS should be included in the “Supporting Information About Proposed REMS Elements”
728   section described immediately below. REMS goals are described in more detail in section
729   III.A.2 of this document.
730
731          3.    Supporting Information About Proposed REMS Elements
732
733   This section should include a description of why particular elements and tools were chosen for
734   the proposed REMS and how each particular element and tool will contribute to achieving the
735   goals of the REMS. Each subsection about elements included in the proposed REMS should
736   include a thorough description of the element(s) proposed for mitigating the risk or risks targeted
737   by the proposed REMS; any tools proposed to be implemented under each element; how the
      12
       GUIDELINE ON RISK MANAGEMENT SYSTEMS FOR MEDICINAL PRODUCTS FOR HUMAN USE,
      Doc. Ref. EMEA/CHMP/96268/2005 http://www.emea.europa.eu/pdfs/human/euleg/9626805en.pdf.


                                                       17

                                 Contains Nonbinding Recommendations
                                      Draft — Not for Implementation
738   elements or tools will mitigate the risk; how the elements or tools conform with elements or tools
739   for other products with similar risks; and whether the elements or tools are compatible with
740   established distribution, procurement, and dispensing systems.
741
742   A thorough description of the available evidence regarding the effectiveness of each element or
743   tool should be provided, including, where applicable, results from pretesting of proposed
744   elements or tools or a time frame for when these will be submitted. These subsections should
745   also note whether the applicant sought input from patient or health care interests, and if so, a
746   description of the feedback received regarding the feasibility of its REMS.
747
748   Elements to Assure Safe Use. Section 505-1(f)(2) requires that FDA consider how to ensure
749   access and minimize the burden of a REMS that includes ETASUs. Therefore, for a proposed
750   REMS that includes ETASUs, the Elements to Assure Safe Use subsection of the REMS
751   supporting document should include the following:
752
753       •    An explanation of how the proposed ETASUs correspond to the specific serious risks
754            listed in the labeling
755       •    An explanation of how the proposed ETASUs will mitigate the observed serious risk
756       •    Verification that the proposed elements are not unduly burdensome on patient access to
757            the drug considering the risk being mitigated. Include particular consideration of
758            patients with serious or life-threatening diseases or conditions and patients who have
759            difficulty accessing health care.
760       •    A description of how, to the extent practicable, the proposed ETASUs will minimize the
761            burden on the health care delivery system: how the proposed ETASUs conform to
762            those required for other drugs with similar serious risks, and how the proposed elements
763            are designed to be compatible with established distribution, procurement, and
764            dispensing systems for drugs.
765
766   Implementation System. This subsection should include the rationale and supporting information
767   for the proposed implementation system, including each method used to monitor and evaluate
768   implementation of the REMS and any planned ways to improve its implementation.
769
770   Timetable for Submission of Assessments of the REMS. This subsection should include the
771   rationale and supporting information for the proposed timetable for submission of assessments of
772   the REMS. This subsection should also include the rationale for the interval that each
773   assessment will cover and for the planned date the assessment will be submitted to the FDA.
774
775           4.   REMS Assessment Plan
776
777   This section should describe the rationale and supporting information for the proposed plan to
778   assess the REMS. Section 505-1(g) of the FDCA describes the requirements for REMS
779   assessments. REMS assessments should include an evaluation of the extent to which each of the
780   REMS elements are meeting the goals and objectives of the REMS, and whether or not the goals,
781   objectives, or REMS elements should be modified. Plans to obtain this information should be
782   included in the REMS supporting document to ensure that sufficient information will be
783   collected to do a valid assessment of the REMS.


                                                      18

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
784
785   In accordance with section 505-1(g)(3)(A), for a REMS that includes one or more ETASUs, the
786   REMS assessment shall include an assessment of the extent to which the ETASUs are meeting
787   the goal (see section III.A.2), or whether the goal or such elements should be modified.
788
789   This subsection should describe the proposed REMS assessment plan, including the following:
790
791      •   The proposed evaluation methods (including measurements or measures) for assessing
792          the overall effectiveness of the REMS and the effectiveness of each of the REMS
793          elements and tools (e.g., claims-based data systems, surveys, registries) and the rationales
794          for the chosen measures.
795      •   Targeted values for each measure and the timeframe for achieving them. Include
796          interpretations of expected results under best- and worst-case scenarios. In addition, this
797          section should specify what values of measures at specific time points will trigger
798          consideration of REMS modification.
799      •   The type of data that will be collected, and the nature and timing of data collection,
800          analyses, audits, or monitoring that will be used to assess the performance of each
801          individual REMS element or tool in achieving the REMS’s objectives and goals.
802      •   Where applicable and possible, this section should discuss plans to assess unintended
803          and/or unfavorable consequences of the REMS following implementation.
804
805   For example, a REMS may indicate that the following data will be collected to support an
806   assessment:
807
808      •   A survey to evaluate knowledge of a labeled serious adverse event to determine whether
809          patients are using the product correctly to prevent the adverse event, or to evaluate use of
810          the product as labeled, particularly when the indicated use is for a restricted population or
811          when numerous contraindications exist.
812
813      •   Information about use patterns of the drug including:
814              o Use by prescriber specialty
815              o Patient-level data (age, gender, race)
816              o Length of therapy
817              o Indication
818
819      •   Population-based administrative or claims-based data that capture service or payment
820          claims to measure rates of specified serious adverse events.
821
822      •   Active surveillance using sentinel reporting sites to determine rates of specified serious
823          adverse events.
824
825   Whenever possible, specific assessment instruments (e.g., surveys) and methodology should be
826   included in the REMS supporting document. If the assessment instruments and methodology are
827   not available when the proposed REMS is submitted to FDA, at least 90 days before the
828   assessments will be conducted, the applicant should update the REMS supporting document to
829   include specific assessment instrument and methodology information. Updates to the REMS


                                                      19

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
830   supporting document may be included in a new document that references previous REMS
831   supporting document submission(s) for unchanged portions of the REMS, or updates may be
832   made by modifying the complete previous REMS supporting document, with all changes marked
833   and highlighted. See section V.B.3 for information on how to identify the submission that
834   includes specific assessment instruments when they are submitted after the REMS is approved.
835
836   For a REMS that includes a Medication Guide, information needed for assessment of the REMS
837   should include but may not be limited to the following:
838
839          (a) Survey of patients’ understanding of the serious risks of the drug
840          (b) Report on periodic assessments of the distribution and dispensing of the Medication
841               Guide in accordance with 21 CFR 208.24
842          (c) Report on failures to adhere to distribution and dispensing requirements, and
843               corrective actions taken to address noncompliance
844
845   If a product is distributed in unit-of-use packaging that includes a Medication Guide with a
846   quantity of product dispensed to a single patient and not divided, the reports in (b) and (c) above
847   would not be necessary.
848
849   This subsection of the REMS supporting document might also include information describing the
850   rationale for, and a description of, all elements proposed to be included in the assessments of the
851   REMS, such as the following:
852
853                  •   Narrative summary and analysis of serious adverse events of interest
854                  •   Summary of data that will be tracked in a REMS-related database
855                  •   Summary of wholesaler shipment data
856                  •   Summary of surveys conducted
857                  •   Summary of data on drug use
858                  •   Summary of registry data
859                  •   Refill frequency and amount
860
861   The assessment should include sufficient detail to identify the need for changes to the REMS.
862   For example, an applicant may be required to assess reports of adverse events associated with the
863   effectiveness of the REMS, each known occurrence of prescriptions written by health care
864   providers who do not have required certification, or dispensing of the product by a pharmacy,
865   practitioner, or health care setting that does not have the required certification. The assessment
866   should also describe any corrective actions taken for these occurrences.
867
868                  Requirements for Information on the Status of Any Postapproval Study or
869                  Clinical Trial Required Under Section 505(o) or Otherwise Undertaken to
870                  Investigate a Safety Issue
871
872   In accordance with section 505-1(g)(3)(B) and (C), all REMS assessments shall include certain
873   information about any postapproval study or clinical trial required under section 505(o) or
874   otherwise undertaken by the applicant to investigate a safety issue.



                                                       20

                                       Contains Nonbinding Recommendations
                                             Draft — Not for Implementation
875        •   For postapproval studies, the REMS assessment shall include the status of each study,
876            including whether any difficulties completing the study have been encountered.
877        •   For postapproval clinical trials, the REMS assessment shall include
878                (a) The status of each clinical trial, including whether enrollment has begun,
879                (b) The number of participants enrolled,
880                (c) The expected completion date,
881                (d) Whether any difficulties completing the clinical trial have been encountered, and
882                (e) Registration information with respect to registry and results databank
883                     requirements under subsections (i) and (j) of section 402 of the Public Health
884                     Service Act. This includes information on whether the data have been
885                     submitted to clinicaltrials.gov, and proper certifications have been submitted to
886                     the FDA.
887
888   The REMS assessment can satisfy the requirements in section 505-1(g)(3)(B) and (C), for
889   information on the status of any postapproval study or clinical trial required under section 505(o)
890   or otherwise undertaken to investigate a safety issue, by referring to relevant information
891   included in the most recent annual report required under section 506B of the FDCA and 21 CFR
892   314.81(b)(2)(vii) or 21 CFR 601.70, and including any updates to the status information since
893   the annual report was prepared, as long as the information required about postapproval studies
894   and clinical trials described above was provided in the annual report. Failure to submit a
895   complete REMS assessment under 505-1(g)(3) could result in enforcement action.
896
897            5.    Other Relevant Information
898
899   This subsection should include information on the positions within the applicant’s company
900   responsible for REMS policy, management, and implementation, including organizational
901   chart(s) that include these REMS-related positions.
902
903   In addition, this subsection should include any other information relevant to the proposed REMS
904   not included elsewhere.
905
906   C.       Foreign Language REMS
907
908   Foreign-language versions of REMS, including any materials appended to the REMS such as
909   Medication Guides, patient package inserts, communication and education materials, enrollment
910   forms, prescriber and patient agreements, and others, are not considered part of the approved
911   REMS. FDA will not review foreign-language versions of REMS.
912
913   Consistent with CDER’s approach to foreign-language labeling, when applicants distribute
914   foreign-language versions of a currently approved REMS, they are responsible for ensuring that
915   such materials are complete and accurate.13 Supplemental applications for foreign-language
916   REMS are not required and should not be submitted.
917

      13
        Note that applicants are required to comply with the requirements regarding distribution of labels and labeling
      under 21 CFR 201.15(c).


                                                               21

                                 Contains Nonbinding Recommendations
                                      Draft — Not for Implementation
918   IV.    REMS ASSESSMENT AND PROPOSED REMS MODIFICATION
919          SUBMISSIONS TO FDA
920
921   REMS assessments must be submitted according to the timetable for submission of assessments
922   included in the REMS, and as otherwise required (see section II.E of this document and 505-
923   1(g)). Applicants may also voluntarily submit an assessment of, and propose a modification to,
924   an approved REMS at any time. An applicant’s proposal for modification of an approved REMS
925   must include an assessment of the REMS.
926
927   Under section 505-1(g)(2)(C), when FDA determines that new safety information indicates that
928   an element of the REMS, such as a Medication Guide, should be modified, the application holder
929   is required to assess the REMS. Where the application holder agrees with the Agency's proposed
930   modification to a REMS that consists solely of a Medication Guide and/or a communication
931   plan, that assessment may consist of a statement that the Medication Guide and/or
932   communication plan would be adequate with the proposed modifications to achieve its/their
933   purpose.
934
935   Proposed modifications may include an enhancement or reduction to the approved REMS, and
936   may include additions or modifications to the timetable for submission of assessments, including
937   a proposal to eliminate assessments (after the 3-year period described in 505-1(d)), and/or the
938   addition, modification, or removal of a Medication Guide, patient package insert, communication
939   plan, or ETASU.
940
941   A proposed modification of an approved REMS that is not associated with an existing
942   supplemental application should be submitted as a new prior-approval supplemental application
943   as described in section V of this document.
944
945   Any proposed modification to the approved REMS, including any proposed changes to materials
946   that are included as part of the REMS (e.g., communication and education materials, enrollment
947   forms, prescriber and patient agreements), must be submitted as a proposed modification to an
948   approved REMS in a new prior-approval supplemental application, as described in section V of
949   this document, and must not be implemented until the modified REMS is approved by FDA.
950
951   Each proposed modification submission should include a new proposed REMS (based on the
952   proposed REMS template described in section III.A) that shows the complete previously
953   approved REMS with all proposed modifications highlighted. In addition, the submission should
954   include an update to the REMS supporting document that includes the rationale for and
955   description of all proposed modifications and any impact the proposed modifications would have
956   on other REMS elements. Updates to the REMS supporting document may be included in a new
957   document that references previous REMS supporting document submission(s) for unchanged
958   portions of the REMS, or updates may be made by modifying the complete previous REMS
959   supporting document, with all changes marked and highlighted. The content of the proposed
960   REMS and REMS supporting document are described in section III of this document.
961   Additional information on assessments and modifications to approved REMS is included in
962   section II.E of this document. More complete information on assessments and modifications of
963   approved REMS will be the subject of future guidance.



                                                     22

                                   Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
 964
 965
 966   V.     COMMUNICATING WITH FDA REGARDING REMS
 967
 968   A.    Submission Type
 969
 970   A proposed REMS may be included in the initial submission of an original or supplemental
 971   application, or may be submitted as an amendment to an existing original or supplemental
 972   application. All supplemental applications that include a proposed REMS or proposed
 973   modifications to an approved REMS should be submitted as prior-approval supplements, not as
 974   changes being effected supplements (see 21 CFR 314.70 and 601.12).
 975
 976   A proposed REMS submitted after approval and not associated with an existing supplemental
 977   application should be submitted as a new supplemental application.
 978
 979   Assessments of approved REMS may be submitted voluntarily at any time and must be
 980   submitted as required in the timetable for submission of assessments of the REMS and as
 981   otherwise required (see sections II.E and IV of this document). A REMS assessment alone (i.e.,
 982   not proposing a modification) is not considered a supplemental application.
 983
 984   REMS assessments that include a proposed modification to the approved REMS should be
 985   submitted either as a new supplemental application or included in a related supplemental
 986   application. They can be included in a related supplemental application either at the time of
 987   submission or as an amendment to the supplemental application.
 988
 989   A supplemental application for a new indication for use for a product with an approved REMS
 990   must include a REMS assessment unless the drug is not subject to section 503(b) and the REMS
 991   for the drug includes only the timetable for submission of assessments (505-1(g)(2)(A)). The
 992   supplemental application for the new indication should include the required REMS assessment
 993   and may propose modifications to the REMS.
 994
 995   A proposed REMS and proposed modifications to an approved REMS should be submitted using
 996   the format in the template for a proposed REMS described in section III.A, and, to facilitate the
 997   review process, the submission should include electronic versions of the proposed REMS or
 998   proposed modifications to an approved REMS as an Adobe Acrobat pdf document and in a
 999   document generated using a word processing program.
1000
1001   As described in section III.C, supplements for foreign-language REMS are not required and
1002   should not be submitted.
1003
1004   Send requests for current information on where REMS-related documents should be included
1005   when submitted as part of an electronic common technical document (eCTD) and questions
1006   about electronic submissions to FDA to the following email address: esub@fda.hhs.gov.
1007
1008   B.    Document Identification
1009



                                                       23

                            Contains Nonbinding Recommendations
                                 Draft — Not for Implementation
1010   1.    Proposed REMS
1011
1012   Regardless of when or how a proposed REMS is submitted, it is critical to provide
1013   identifying information on the submitted REMS document so that it can be tracked,
1014   routed, and reviewed appropriately. In each case, the first page of the submission should
1015   prominently identify the submission as providing a PROPOSED REMS in bold capital
1016   letters at the top of the page. This wording on the first page of the submission should be
1017   combined with any other applicable content identification, for example:
1018
1019   When the proposed REMS is submitted as part of an original application:
1020
1021   NEW ORIGINAL APPLICATION FOR <name of drug>
1022   PROPOSED REMS
1023
1024   When the original proposed REMS is submitted as an amendment to an existing original
1025   or supplemental application:
1026
1027   NDA/BLA/ANDA [assigned #]
1028   PROPOSED REMS
1029
1030   NDA/BLA/ANDA [assigned #] SUPPLEMENT [assigned #]
1031   PROPOSED REMS
1032
1033   When the original proposed REMS is submitted postapproval as a new supplemental
1034   application:
1035
1036   NEW SUPPLEMENT FOR NDA/BLA/ANDA [assigned #]
1037   PROPOSED REMS
1038
1039   When the original proposed REMS is submitted postapproval with a new supplemental
1040   application:
1041
1042   NEW SUPPLEMENT FOR NDA/BLA/ANDA [assigned #]
1043   < other applicable content identification >
1044   PROPOSED REMS
1045
1046   On the first page of subsequent submissions related to an already-submitted proposed
1047   REMS, prominently identify the submission by including this wording in bold capital
1048   letters at the top of the letter:
1049
1050   NDA/BLA/ANDA [assigned #]
1051   PROPOSED REMS-AMENDMENT
1052
1053   NDA/BLA/ANDA [assigned #] SUPPLEMENT [assigned #]
1054   PROPOSED REMS-AMENDMENT
1055



                                                24

                           Contains Nonbinding Recommendations
                                Draft — Not for Implementation
1056   2.   Assessments and Modifications of Approved REMS
1057
1058   On the first page of the submission of an assessment of an approved REMS, prominently
1059   identify its content in bold capital letters at the top of the page:
1060
1061   NDA/BLA/ANDA [assigned #]
1062   REMS ASSESSMENT
1063
1064   If a REMS assessment is submitted as a part of another submission, it is critical to
1065   provide complete identifying information on the submission so that it can be tracked,
1066   routed, and reviewed appropriately. In each case, the first page of the submission should
1067   prominently identify the submission as providing a REMS ASSESSMENT in bold
1068   capital letters at the top of the page. This wording on the first page of the submission
1069   should be combined with any other applicable content identification.
1070
1071   The first page of the submission of an assessment of an approved REMS submitted with a
1072   supplemental application for a new indication for use should prominently identify the
1073   content in bold capital letters at the top of the page. The submission may include
1074   proposed modifications to the approved REMS. This wording on the first page of the
1075   submission should be combined with any other applicable content identification, for
1076   example:
1077
1078   NEW SUPPLEMENT FOR NDA/BLA/ANDA [assigned #]
1079   < other supplement identification >
1080   REMS ASSESSMENT
1081   PROPOSED REMS MODIFICATION (if included)
1082
1083   The first page of the submission of proposed modifications to an approved REMS
1084   submitted as a stand-alone new supplemental application or included with another new
1085   supplemental application should prominently identify the content in bold capital letters at
1086   the top of the page. This wording on the first page of the submission should be combined
1087   with any other applicable content identification, for example:
1088
1089   NEW SUPPLEMENT FOR NDA/BLA/ANDA [assigned #]
1090   < other supplement identification >
1091   PROPOSED REMS MODIFICATION
1092   REMS ASSESSMENT
1093
1094   The first page of the submission of proposed modifications to an approved REMS
1095   submitted as an amendment to a pending supplemental application should prominently
1096   identify the content in bold capital letters at the top of the page:
1097
1098   NDA/BLA/ANDA [assigned #] SUPPLEMENT [assigned #]
1099   PROPOSED REMS MODIFICATION
1100   REMS ASSESSMENT
1101



                                               25

                                 Contains Nonbinding Recommendations
                                      Draft — Not for Implementation
1102          The first page of subsequent submissions related to a proposed modification to an
1103          approved REMS should prominently identify the submission by including this wording in
1104          bold capital letters at the top of the page:
1105
1106          NDA/BLA/ANDA [assigned #] SUPPLEMENT [assigned #]
1107          PROPOSED REMS MODIFICATION -AMENDMENT
1108
1109          3.   Other REMS Submissions
1110
1111         An applicant may submit REMS submissions that are not proposed REMS, proposed
1112         modifications to an approved REMS, amendments to proposed REMS, proposed
1113         modifications to an approved REMS, or REMS assessments. Such submissions may
1114         include a request for information about what to include in a proposed REMS, information
1115         about the REMS assessment plan for an approved REMS (e.g., assessment instruments
1116         and methodology), general correspondence about an approved REMS that does not
1117         include a proposed modification, amendment to a proposed modification, or a REMS
1118         assessment, or other submissions that do not fall into the categories described above. On
1119         the first page of such submissions, prominently identify its content with the words,
1120         “REMS - OTHER” followed by a concise description of the content in bold capital letters
1121         at the top of the page. For example:
1122
1123          NDA/BLA/ANDA [assigned #]
1124          REMS-OTHER
1125          SURVEY METHODOLOGY
1126
1127          The first page of a submission requesting Agency input on the content of a proposed
1128          REMS that has not yet been submitted should include the following wording in bold
1129          capital letters at the top of the page:
1130
1131          NDA/BLA/ANDA [assigned #]
1132          REMS-OTHER
1133          REQUEST FOR GUIDANCE ON CONTENT OF PROPOSED REMS
1134
1135         If the proposed REMS has already been submitted, such a request should be identified as a
1136         proposed REMS amendment – see section V.B.1.
1137
1138   C.    Questions about REMS
1139
1140   In the Center for Drug Evaluation and Research (CDER), the primary contact about a proposed
1141   REMS for a product under an NDA or BLA is the regulatory project manager in the Office of
1142   New Drugs (OND) review division assigned to that product. The primary contact about a
1143   proposed REMS for a product under an ANDA is the Director of the Division of Labeling and
1144   Program Support in the Office of Generic Drugs (OGD). The Office of Surveillance and
1145   Epidemiology, and other program offices as needed, will work with OND and OGD in the
1146   review and development of a proposed REMS.
1147



                                                     26

                                  Contains Nonbinding Recommendations
                                      Draft — Not for Implementation
1148   In the Center for Biologics Evaluation and Research (CBER), the primary contact about a
1149   proposed REMS is the regulatory project manager in the office with product responsibility. The
1150   Office of Biostatistics and Epidemiology, and other program offices as needed, will work with
1151   the product office in the review and development of a proposed REMS.
1152
1153




                                                     27

                                   Contains Nonbinding Recommendations
                                        Draft — Not for Implementation
1154   GLOSSARY – applicable to terms as used in this document
1155
1156   Assessment: An assessment of the approved REMS as described in section II.E and III.B.4 of
1157   this document.
1158
1159   Changes Being Effected Supplement: Also called a “changes being effected supplemental
1160   application.” A supplement that includes changes that do not require supplement submission and
1161   approval prior to the changes being implemented; the application holder may commence
1162   distribution of the drug product involved upon receipt by the agency of a supplement for these
1163   changes. A “Changes Being Effected in 30 days” supplement includes changes that do not
1164   require approval prior to the changes being implemented, but requires supplement submission at
1165   least 30 days prior to distribution of the drug product made using the change. If, after review,
1166   FDA disapproves a changes being effected supplement or a changes being effected in 30 days
1167   supplement, FDA may order the manufacturer to cease distribution of the drug products made
1168   using the disapproved change (21 CFR 314.70(c) and 601.12(c)). See section V.A of this
1169   document.
1170
1171   Goal: The desired safety-related health outcome or the understanding of serious risks targeted
1172   by the use of specified REMS elements. See section III.A.2 of this document.
1173
1174   Objective: An intermediate step to achieving the overall goals of the REMS. Objectives should
1175   be pragmatic, specific, and measurable. Objectives may use one or more elements or tools that
1176   result in processes or behaviors leading to achievement of the REMS goals. A REMS goal can
1177   be translated into different objectives, depending upon the frequency, type, and severity of the
1178   specific risk or risks being minimized. See section III.A.2 of this document.
1179
1180   Prior-approval Supplement: Also called a “prior-approval supplemental application.” A
1181   supplemental application that includes changes requiring supplement submission and approval
1182   prior to the distribution of the product made using the change. (21 CFR 314.70(b) and
1183   601.12(c)). See section V.A of this document.
1184
1185   Qualification Stickers: Stickers given by the applicant to providers to affix to prescriptions for
1186   specified products to indicate that the patient has met all criteria for receiving the product.
1187
1188   REMS: Stands for “Risk Evaluation and Mitigation Strategy,” and is the enforceable document
1189   that describes the elements that an applicant is required to implement. See section III.A of this
1190   document.
1191
1192   REMS Supporting Document: A document that includes a thorough explanation of the
1193   rationale and supporting information for the content of the proposed REMS. See section III.B of
1194   this document.
1195
1196   Tool: A process or system designed to implement one or more REMS elements. In some cases
1197   an element itself, such as a Medication Guide, may be viewed as a tool. In other cases, such as
1198   for an ETASU that requires that a drug be dispensed to patients with evidence or other
1199   documentation of safe-use conditions (505-1(f)(3)(C)), specific tools are used to implement a


                                                       28

                                  Contains Nonbinding Recommendations
                                      Draft — Not for Implementation
1200   REMS element. Examples of such tools include systems that ensure certain laboratory test result
1201   outcomes are obtained before a drug may be dispensed.
1202




                                                     29

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
1203   ATTACHMENT A: EXAMPLE OF A REMS DOCUMENT FOR A FICTITIOUS DRUG
1204
1205                                      NDA ##-### Drug X

1206              RISK EVALUATION AND MITIGATION STRATEGY (REMS)

1207
1208                                     Class of Product as per label
1209                                     ABCD Pharmaceuticals

1210                                          123 Fake Street
1211                                          City, State Zip
1212                            Contact Information for those responsible for
1213                           REMS policy, management, and implementation
1214
1215                                          (555)-xxx-xxxx
1216                                        www.emailaddress.xxx
1217
1218   I.     GOAL

1219   To minimize the risk of drug exposure during pregnancy in women of child-bearing potential
1220   taking Drug X. Because Drug X is teratogenic, ABCD Pharmaceuticals (ABCD) will mitigate
1221   this risk by:
1222
1223          •   Ensuring that only females of childbearing potential with a negative pregnancy test
1224              begin therapy with Drug X and only females of childbearing potential with a monthly
1225              negative pregnancy test continue therapy with Drug X.
1226          •   Ensuring that females of childbearing potential understand the risks to the fetus and
1227              know what precautions are necessary to prevent pregnancy.
1228          •   Ensuring that all patients and health care providers understand the risks associated
1229              with Drug X.
1230   This drug is contraindicated in female patients who are or may become pregnant.

1231   II.    REMS ELEMENTS
1232
1233          A. Medication Guide (FDCA Section 505-1(e)(2))
1234
1235   A Medication Guide will be dispensed with each Drug X prescription. To ensure compliance
1236   with 21 CFR 208.24, ABCD will attach a Drug X Medication Guide to each unit-of-use
1237   package of Drug X to ensure that the Medication Guide is given to each patient with each new
1238   prescription and refill. A copy of the Medication Guide is appended to the REMS Document.
1239   The Medication Guide will be available on the ABCD Web site within 10 days of approval of
1240   the Medication Guide.


                                                       30

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
1241          B. Communication Plan (FDCA Section 505-1(e)(3))
1242
1243   ABCD will implement a communication plan to health care providers to support implementation
1244   of this REMS:
1245
1246   1.     The audience for this communication plan is health care professionals (HCPs)—
1247          especially neurologists, endocrinologists, and pharmacists.
1248
1249   2.     ABCD will provide physicians and pharmacists with educational materials listed below
1250          that describe the key risks and benefits of Drug X:
1251
1252          a. Prescriber Materials — Dear Health Care Professional Letter
1253          b. Pharmacist Materials — Dear Pharmacist Letter
1254          c. Additional Resources — Drug X REMS Program Internet Site
1255
1256          The printed communication and educational materials listed above are appended.
1257
1258   3.     Distribution of materials: Communication plan materials will be distributed within 60
1259          days of approval of the Drug X REMS.
1260
1261          a. At the time the Drug X REMS elements to assure safe use are implemented, ABCD
1262             will send the Dear Health Care Professional Letter by mass mailing to targeted Drug
1263             X prescribers to announce the REMS program and the requirements of the program.
1264             The mailing will include the materials listed in 2a above. Copies of these materials
1265             will be available through the product Web site.
1266
1267          b. At the time the Drug X REMS elements to assure safe use are implemented, ABCD
1268             will send the Dear Pharmacist Letter by mass mailing to targeted pharmacies who
1269             currently order Drug X, to announce the REMS program and the requirements of the
1270             program. The mailing will include the materials listed in 2b above. Copies of these
1271             materials will be available through the product Web site.
1272
1273          C. Elements To Assure Safe Use (FDCA Section 505-1(f)(3))
1274
1275   ABCD will implement the following elements to ensure safe use to mitigate the risk of drug
1276   exposure during pregnancy by women of child-bearing potential. The elements to assure safe
1277   use will be implemented within 60 days of approval of the Drug X REMS.
1278
1279   1.     Drug X will be prescribed only by prescribers who are specially certified under
1280          505-1(f)(3)(A) by enrollment in the Drug X REMS program.
1281
1282          a. ABCD will ensure that physicians and other appropriately licensed health care
1283             providers who prescribe Drug X are specially certified. ABCD will ensure that, to
1284             become certified, each prescriber, on the prescriber enrollment form, attests to the
1285             following:
1286


                                                      31

                                 Contains Nonbinding Recommendations
                                      Draft — Not for Implementation
1287             •     To have read and understood the communication and educational materials for
1288                   prescribers regarding the risks and benefits of Drug X, including the Drug X
1289                   Prescriber Guide and the Prescriber Contraception Counseling Guide
1290             •     To have knowledge of the high risk of severe birth defects associated with
1291                   Drug X
1292             •     To know the risk factors for unplanned pregnancy and the effective measures to
1293                   avoid pregnancy
1294             •     To prescribe Drug X after ensuring documentation of safe use conditions
1295                   described below
1296             •     To submit information about any pregnancy they learn about to the pregnancy
1297                   registry
1298             •     To monitor patients treated with Drug X as described below
1299
1300         b. ABCD will maintain a list of all certified prescribers and will provide the list to those
1301            needing to verify that a prescriber has obtained the required certification.
1302
1303         c. ABCD will ensure that prescribers will be recertified in the Drug X REMS program
1304            annually.
1305
1306         The following materials are part of the REMS and are appended:
1307
1308             •      Prescriber enrollment form,
1309             •      Prescriber Guide
1310             •      Prescriber Contraception Counseling Guide
1311
1312   2.    Drug X will be dispensed only by pharmacies that are specially certified under
1313         505-1(f)(3)(B) by enrollment in the Drug X REMS program.
1314
1315        a. ABCD will ensure that responsible pharmacy personnel from pharmacies that dispense
1316           Drug X are specially certified. ABCD will ensure that, to be certified, responsible
1317           pharmacy personnel will attest to the following:
1318
1319             •     To have read and understood the communication and educational materials for
1320                   pharmacists regarding the risks and benefits of Drug X, including the Drug X
1321                   Pharmacist Guide
1322             •     To have knowledge of the high risk of severe birth defects associated with
1323                   Drug X
1324             •     To train all pharmacists to fill and dispense Drug X only after ensuring
1325                   documentation of safe-use conditions described below
1326             •     To ensure that all pharmacists who fill and dispense Drug X comply with
1327                   required documentation of safe-use conditions described below
1328             •     To agree not to sell, borrow, lend, or otherwise transfer Drug X to or from
1329                   another pharmacy
1330




                                                      32

                                    Contains Nonbinding Recommendations
                                         Draft — Not for Implementation
1331           b. ABCD maintains a list of all certified pharmacies and will provide the list to those
1332              needing to verify that a pharmacy has obtained the required certification.
1333
1334           c. Drug X will be distributed to certified pharmacies.
1335
1336           d. Pharmacies will be recertified in the Drug X REMS program annually.
1337
1338        The pharmacy enrollment form and Pharmacist Guide are part of the REMS and are
1339        appended.
1340
1341   3.      Drug X will only be dispensed to patients with documentation of safe-use conditions
1342           under 505-1(f)(3)(D)) described below:
1343
1344           a.   ABCD will ensure that prescribers of Drug X will:
1345
1346                •       Register each patient in the Drug X REMS program (patient enrollment form
1347                         is appended)
1348                •       Determine the childbearing status of all female patients
1349                •       Counsel each female of childbearing potential (FCBP) before beginning
1350                         therapy with Drug X and on a monthly basis to avoid pregnancy by using
1351                         effective contraceptive forms or refer the patient for contraception
1352                         counseling
1353                    o    Provide them with the following educational materials: Guide for Patients
1354                         Who Can Become Pregnant (appended)
1355                    o    Confirm that FCBP have signed the appropriate informed consents —
1356                         Informed consent for Patients Who Can Become Pregnant (appended)
1357                •       Counsel males and females not of child bearing potential about the risks and
1358                         benefits of Drug X before beginning therapy with Drug X.
1359                    o    Provide them with the following educational materials: Guide for Patients
1360                         Who Cannot Become Pregnant (appended)
1361                    o    Confirm that males and females not of childbearing potential have signed the
1362                         appropriate informed consents — Informed consent for Patients Who Cannot
1363                         Become Pregnant (appended)
1364                •        Complete for each patient either the Drug X Prescriber Checklist for Patients
1365                         Who Can Become Pregnant, or the Drug X Prescriber Checklist for Patients
1366                         Who Cannot Become Pregnant (appended)
1367                •        For female patients of childbearing potential prior to each prescription:
1368                        o Indicate patient’s chosen contraceptive forms each month by telephone or
1369                            secure Internet Web site
1370                        o Order CLIA-certified pregnancy test for each patient prior to each
1371                            prescription and enter results of pregnancy test each month by telephone
1372                            or secure Internet Web site
1373
1374           b.   ABCD will ensure that pharmacies that dispense Drug X will:
1375



                                                        33

                                   Contains Nonbinding Recommendations
                                        Draft — Not for Implementation
1376                 •      Obtain authorization from the Drug X REMS program by telephone or
1377                        secure Internet Web site for every Drug X prescription and write the
1378                        authorization number on each prescription
1379                 •      Dispense only a 30-day supply
1380                 •      Dispense within 7 days of a last negative pregnancy test
1381                 •      Dispense the Drug X Medication Guide with each prescription
1382
1383        c.       ABCD will ensure that Drug X is dispensed only to patients who have met the
1384                 following conditions:
1385
1386                 •     All patients have:
1387                      o Signed the informed consent prior to beginning therapy with Drug X
1388                 •     Females of childbearing potential (before each prescription) have:
1389                      o Obtained a CLIA-certified pregnancy test
1390                      o Indicated chosen contraceptive forms each month by telephone or secure
1391                         Internet Web site
1392                      o Completed a questionnaire each month through a secure Internet Web site
1393
1394   4.   ABCD will ensure that patients who are treated with Drug X are monitored by their
1395        prescribers monthly for the duration of Drug X therapy and for 1 month following Drug
1396        X discontinuation under section 505-1(f)(3)(E). Monitoring will include the following
1397        elements:
1398
1399                 •      Re-counseling all patients about the risks and benefits of Drug X therapy and
1400                        determining whether they are still appropriate for Drug X therapy
1401                 •      Determining whether the childbearing status of female patients has changed
1402                 •      Obtaining a CLIA-certified pregnancy test prior to each Drug X prescription
1403                 •      Ensuring FCBP are still on appropriate contraception and re-counseling
1404                        FCBP of the importance of complying with contraceptive methods during
1405                        and for 1 month following therapy with Drug X
1406
1407   5.   ABCD will ensure that Drug X will only be dispensed to patients who are enrolled in the
1408        REMS program registry under 505-1(f)(3)(F) and who meet the following conditions:
1409
1410             •        Patient must understand that severe birth defects can occur with the use of
1411                      Drug X by female patients.
1412             •        Patient must be reliable in understanding and carrying out instructions.
1413             •        Patient must agree to not share Drug X with anyone.
1414             •        Patient must agree to not donate blood while on Drug X and for 1 month after
1415                      Drug X discontinuation.
1416             •        Females of child-bearing potential (FCBP) must:
1417                       o Not be pregnant and understand the importance of avoidance of
1418                           pregnancy
1419                       o Be capable of following mandatory contraceptive measures
1420



                                                       34

                                Contains Nonbinding Recommendations
                                     Draft — Not for Implementation
1421        The following information will be collected on enrolled patients:
1422
1423           •       Age, gender, and childbearing status
1424           •       Documentation of counseling
1425           •       Prescription data (e.g., dates RX filled, quantity dispensed)
1426           •       For FCBP:
1427                      o       Baseline and monthly pregnancy test (dates and results)
1428                      o       Chosen methods of contraception
1429           •       For females who become pregnant
1430                      o       Maternal and fetal outcomes
1431                      o       Information on circumstances that led to failure to prevent
1432                              pregnancy
1433
1434        D. Implementation System (FDCA Section 505-1(f)(4))
1435
1436   The implementation system will include the following components:
1437
1438   1.      ABCD will maintain a validated and secure database of all entities enrolled under
1439           505-1(f)(3)(B) and (D) and 505-1(f)(4), including wholesalers/distributers,
1440           pharmacies and patients.
1441   2.      ABCD will ensure that wholesalers/distributers who distribute Drug X are specially
1442           certified. To become certified, wholesalers/distributers will be enrolled in the Drug X
1443           REMS program.
1444
1445               a. The Drug X REMS Program wholesaler/distributor enrollment process is
1446                  composed of the following three steps that must be completed prior to
1447                  receiving Drug X inventory for distribution:
1448                   i. The Distributor’s Authorized Representative reviews the
1449                      Wholesaler/Distributor Program Materials.
1450                   ii. Prior to receiving Drug X, the Distributor’s Authorized Representative
1451                       completes and signs the Distributor Enrollment Form and faxes it to the
1452                       Drug X REMS Program. In signing the Enrollment Form, the
1453                       Representative is required to indicate they understand that Drug X is
1454                       available only through the Drug X REMS Program, agree to comply with
1455                       program requirements, and acknowledge that:
1456                          A. I will ensure that relevant staff are trained about the Drug X REMS
1457                             Program for Drug X procedures.
1458                          B. I will ensure that relevant staff distribute Drug X only to Drug X
1459                             REMS pharmacies that are active in the database.
1460                          C. I will provide monthly records of Drug X shipments to each Drug
1461                             X REMS pharmacy.




                                                    35

                                  Contains Nonbinding Recommendations
                                       Draft — Not for Implementation
1462                            D. I will permit a program-related audit of our shipping records to
1463                               corroborate that we are shipping Drug X only to Drug X REMS
1464                               pharmacies.
1465                     iii. A Drug X REMS Program professional reviews the form, requests any
1466                          missing or illegible information, and, when the form has been verified to
1467                          be accurate and successfully completed, the distributor is notified of
1468                          activation.

1469                 b. Upon initial activation, wholesalers/distributors remain active until a
1470                    corrective action of inactivation occurs or expiration of the enrollment period.
1471                 c. If a previously active wholesaler becomes inactive, the wholesaler/distributor
1472                    can become active again by completing the standard wholesaler enrollment
1473                    process in its entirety.
1474                 d. Wholesalers/distributors are re-educated and re-enrolled following substantial
1475                    changes to the program or at least every 2 years. Substantial changes to the
1476                    Drug X REMS Program are defined as changes that modify the operation of
1477                    the Drug X REMS Program in a way that changes Drug X REMS Program
1478                    procedures for distributors.
1479                 e. The Distributor Enrollment Form is part of the REMS and is appended.
1480
1481      3.      ABCD will monitor wholesaler distribution data to ensure that only registered entities
1482              are dispensing Drug X.
1483      4.      ABCD will monitor pharmacies to ensure these entities are dispensing Drug X to
1484              patients only after receiving authorization.
1485      5.      ABCD will correct pharmacy noncompliance with program requirements.
1486      6.      ABCD will conduct periodic audits of registered pharmacies to determine whether the
1487              data collected is in the manner and frequency agreed upon with FDA.
1488      7.      ABCD will maintain a Call Center (1-800-ABCD411) to respond to questions from
1489              practitioners, pharmacists, and patients (FDAAA Section 505-1(f)(3)(B), and (D)).
1490
1491           E. Timetable for Submission of Assessments
1492
1493   ABCD will submit REMS Assessments to FDA every 6 months from the date of the approval of
1494   the REMS. To facilitate inclusion of as much information as possible while allowing reasonable
1495   time to prepare the submission, the reporting interval covered by each assessment should
1496   conclude no earlier than 60 days before the submission date for that assessment. ABCD will
1497   submit each assessment so that it will be received by the FDA on or before the due date.
1498
1499   [Attachments are not included in this example.]




                                                         36


								
To top