rtf - FDA by 1G6xMwj5




                        + + + + +


                        + + + + +


                        + + + + +

                      68TH MEETING

                        + + + + +

                     MARCH 24, 2000

                        + + + + +

      The meeting was held at 8:30 a.m. in the Marriott
Washingtonian   Hotel,    9751   Washingtonian    Blvd,
Gaithersburg, Maryland, Dr. Barth Reller, Acting
Chairman, presiding.


L. BARTH RELLER, M.D.,      Acting Chairman
P. JOAN CHESNEY, M.D.,      Member
M.D., M.P.H.,               Member
ROBERT L. DANNER, M.D.,     Member
JAMES LEGGETT, JR., M.D.,   Temporary Voting Member
BARBARA E. MURRAY, M.D.     Member
CARL W. NORDEN, M.D.,       Temporary Voting Member
JUDITH R. O'FALLON, Ph.D,   Member
KEITH RODVOLD, Pharm.D,     Member
DAVID E. SOPER, M.D.,       Member
JOYCE DRAYTON, M.D.,        Guest Expert
JANET WITTES, Ph.D.,        Guest Expert
KIMBERLY TOPPER,            Executive Secretary

                       S A G CORP.
202/797-2525           Washington, D.C.       Fax: 202/797-2525


Call to Order ...................................... 3
      Barth L. Reller, M.D., Acting Chair

Conflict of Interest Statement ..................... 3
      Kimberly Topper, Executive Secretary

Opening Remarks .................................... 7
      Gary Chikami, M.D., Director
      Division of Anti-Infective Drug Products
      Office of Drug Evaluation IV, FDA

Applicant Presentation:

Pharmacia & Upjohn

Introduction and Microbiology ..................... 14
      Gary Tarpley, M.D.
      Vice President, Discovery Research

Clinical Pharmacology and
Clinical Trial Results ............................ 28
      Barry Hafkin, M.D.
      Director, Clinical Research

Early Pediatric Studies ........................... 62
      Donald Anderson, M.D.
      Vice President and Chief Scientific
      Research and Development

Clinical Pharmacology, Dr. Gail Jungbluth ......... 72

Toxicology, Dr. Greg Slatter ..................... 74

FDA Presentation:
      Dr. David Ross ............................. 108

Applicant's Response to Questions ............... 174

Presentation of Questions to the Committee ...... 252

                     S A G CORP.
202/797-2525         Washington, D.C.     Fax: 202/797-2525

 1                          P-R-O-C-E-E-D-I-N-G-S

 2                                                           (8:30 a.m.)

 3                    CHAIRMAN RELLER:              Good morning.      Good

 4   morning if everyone could take their seats, like an

 5   airplane the doors are closing for an on time departure.

 6                    Good morning again.             I'm Barth Reller,

 7   Acting Chairman for this meeting of the Anti-Infective

 8   Advisory Committee of the FDA.                  I'd like to welcome

 9   everyone to today's meeting and we'll begin with a

10   reading of the conflict of interest statement by

11   Kimberly Topper, our Executive Secretary for today's

12   meeting.       Kimberly.

13                    MS.     TOPPER:           The      following      the

14   announcement addresses the issue of conflict of interest

15   with regard to this meeting and is made part of the record

16   to preclude even the appearance of such at this meeting.

17    Based on the submitted agenda and information provided

18   by the participants the Agency is determined that all

19   recorded interests and firms regulated by the Center

20   for Drug Evaluation and Research present no potential

21   for conflict of interest at this meeting with the

22   following exceptions:

23                    In accordance with 18 U.S.C. 208 (b), a full

                                S A G CORP.
     202/797-2525                Washington, D.C.           Fax: 202/797-2525

 1   waiver has been granted to Dr. Keith Rodvold.                      A copy

 2   of this waiver statement may be obtained by submitting

 3   a written request to FDA's Freedom of Information Office

 4   located        in   room    12A-30,    located     in     the    Parklawn

 5   Building.

 6                       In addition one of our committee members

 7   has had a past interest relating to Zyvox, that we

 8   believe should be disclosed.              The FDA believes that it

 9   is important to acknowledge this involvement so that

10   his participant may be objectively evaluated.

11                       Dr.    James    Leggett      was    listed    as    an

12   investigator on the study for Zyvox, while Dr. Leggett

13   was listed an investigator on this study he did not

14   enroll patients and was not otherwise directly involved.

15                       I would like to remind the committe members

16   to please speak directly into the microphone this is

17   being recorded.

18                       Thank you.

19                       DR.    CHESNEY:         Joan       Chesney,        the

20   University of Tennessee in Memphis.

21                       DR.    SOPER:         David        Soper,     Medical

22   University of South Carolina.

23                       DR. KUEHNERT:      Matt Kuehnert, Center for

                                   S A G CORP.
     202/797-2525                  Washington, D.C.             Fax: 202/797-2525

 1   Disease Control.

 2                  DR. WITTES:     Janet Wittes.

 3                  CHAIRMAN RELLER:        Dr. Wittes could you

 4   introduce yourself again.       I don't think the mike was

 5   working properly.

 6                  DR. KUEHNERT:    Matt Kuehnert.     Center for

 7   Disease Control and Prevention.

 8                  DR. WITTES:      Janet Wittes.      Statistics

 9   Collaborative.

10                  DR. SORETH:    Janice Soreth, I'm a Medical

11   Officer in the Division of Anti-Infectives.

12                  DR. CHIKAMI:    I'm Gary Chikami.

13                  CHAIRMAN RELLER:       We need a little help

14   with the audio portion.

15                  DR. MURPHY:     Diane Murphy.

16                  CHAIRMAN RELLER:       Let's go to the other

17   side of the table and we can pick up when we get power.

18    Dr. O'Fallon.

19                  DR. O'FALLON:       Judith O'Fallon.         Mayo

20   Clinic, Cancer Center for Statistics.

21                  DR. MURRAY:    Barbara Murray, University of

22   Texas, Medical School at Houston, Infectious Diseases.

23                  DR. LEGGETT:    Jim Leggett, Medical Center,

                            S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1   Oregon Health Sciences University.

 2                    DR. DRAYTON:         Joyce Drayton, Morehouse

 3   School of Medicine, Division of Infectious Disease.

 4                    DR.     LOWY:         Frank       Lowy,      Columbia

 5   University,       College     of    Physicians      and     Surgeons.

 6   Infectious Diseases.

 7                    DR. CHRISTIE:      Celia Christie, University

 8   Hospital of the West Indies, Pediatrics, Infectious

 9   Diseases.

10                    DR. RODVOLD:       Keith Rodvold.         College of

11   Pharmacy and Medicine, University of Illinois, Chicago.

12                    DR. DANNER:       Bob Danner.          Critical Care

13   Medicine Department of NIH.

14                    DR. NORDEN:        Carl Norden.           Infectious

15   Diseases at University of New Jersey.                       School of

16   Medicine and Dentistry.

17                    CHAIRMAN RELLER:        Back to Dr. Murphy.

18                    DR. MURPHY:       Mine is working.         Dr. Diane

19   Murphy,        Office    Director      of      ODE-4,      which     has

20   anti-infectives, anti-viral, and special pathogens in

21   it at FDA.       Thank you.

22                    DR. CHIKAMI:       And I'm Gary Chikami.            I'm

23   the    Director     of   Division      of      Anti-Infective       Drug

                                 S A G CORP.
     202/797-2525                Washington, D.C.             Fax: 202/797-2525

 1   Products, FDA.

 2                  CHAIRMAN    RELLER:           For   today's    meeting

 3   we're especially pleased to have with us, Dr. Carl

 4   Norden, Dr. Leggett who will be voting members for

 5   today's session.     And a special welcome also to our

 6   guest experts who will be participating, but not voting

 7   on the questions that we'll address later.                And those

 8   individuals are Drs. Joyce Drayton, Matthew Kuehnert,

 9   Frank Lowy, and Dr. Wittes.

10                  Next we'll have opening remarks for today's

11   meeting by Dr. Gary Chikami, who's the Director of the

12   Division of Anti-Infective Drug Products of the Office

13   of Drug Evaluation for the FDA.                       Gary.

14                  DR. CHIKAMI:       Thank you Dr. Reller.            And

15   just a few organizational comments.            I'd like to welcome

16   --

17                  CHAIRMAN RELLER:        While Gary's doing that

18   I realized that I didn't introduce myself, fully.

19                  I'm in the Division of Infectious Diseases

20   in Direct Clinical Microbiology Laboratory at the Duke

21   University Medical Center.         And as noted earlier, will

22   be the Acting Chairman for today's meeting.                   Now Dr.

23   Chikami.

                               S A G CORP.
     202/797-2525              Washington, D.C.             Fax: 202/797-2525

 1                  DR. CHIKAMI:       Thank you Dr. Reller.           I'd

 2   like to first of all, welcome Dr. Reller as the new chair

 3   of the committee.        He's -- because of the paperwork,

 4   he will be acting as chair, but in future meetings he

 5   will be the permanent chair of our committee.

 6                  In addition, I'd like to welcome Dr. Leggett

 7   who also is joining the committee as a new member.

 8                  Today's    meeting      we'll    be    hearing     the

 9   presentation of the new drug application for Zyvox or

10   linezolid from Pharmacia and Upjohn.             And I would also

11   like to extend my welcome to the applicant this morning,

12   and also members in the audience who will be here for

13   what I think will be an important discussion of this

14   new drug product.

15                  We're having a little technical glitch with

16   the slides.     I just have a few general comments that

17   I want to make.

18                  This   meeting       today      will   discuss       an

19   application for a new drug product being developed for

20   a number of indications, but particularly for the

21   treatment of resistant gram positive infections.                 Over

22   the past several years there have been -- this committee

23   has met to discuss both specific and general issues

                               S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1   related to development of products in this area.

 2                   In July of 1998 and October of 1998 there

 3   were two general meetings that -- one with industry

 4   members of academia, the second in October, specifically

 5   with this committee, to discuss some issues related

 6   specifically to development of drug products in this

 7   area.    In March of '98 -- and there had been two product

 8   specific meetings.       One for Synocin (ph) in March of

 9   '98, the second most recently in November of '99 to

10   discuss supplemental application for Levaquin for the

11   treatment of penicillin-resistant strep pneumo.

12                   Now I think during the course of these

13   meetings a number of issues were raised and in two broad

14   areas.     One is what sort of evidence do we need to gather

15   in the course of drug development to support the -- to

16   support granting indications for resistant organisms?

17                   And I think the second area that I think

18   relates to the specific issues today is what are specific

19   trials designs that one:         may increase the experience

20   available for the treatment of resistant organisms.

21   Because as is often the case, in the course of the usual

22   clinical trial, it's difficult to gather sufficient

23   evidence       on   infections     with      specific   resistant

                              S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1   organisms.

 2                       And    the     second,     if     you're      developing

 3   products       in    an    area     where     there     is     no   approved

 4   comparator, what sorts of designs would be acceptable

 5   or provide us with control clinical trial information?

 6                       In regard to the first area I think some

 7   general issues or principles have come from the four

 8   meetings that I've talked about.                    One is that one would

 9   like to see a drug product being studied in a number

10   of areas to provide both in vitro evidence of activity

11   against both susceptible and resistant isolates.

12                       One    would     like    to     also    develop     other

13   preclinical sources of information such as animal model

14   data, which would speak to activity again, not only

15   against        suspectable        strains      of     the    organism      but

16   resistant        strains.            And     finally        the     important

17   underpinning          of   clinical         information       coming      from

18   controlled clinical trials.

19                       And then with regard to specifics -- how

20   would one apply those principles.                    For example, if one

21   is developing a product for resistant infections in

22   pneumonia.          It's important to understand how a product

23   works, not only for that side of infection that is

                                      S A G CORP.
     202/797-2525                     Washington, D.C.              Fax: 202/797-2525

 1   treatment of pneumonia in general, but treatment of

 2   pneumonia      for   susceptible       strains.          For       example,

 3   susceptible strains strep pneumo and finally gathering

 4   whatever information is available for treatment of

 5   pneumonia due to PRSP.

 6                    I think one can step through a development

 7   program and look for those themes in regard to the

 8   information that's gathered.

 9                    In regard to the second issue that --

10   specific designs which would one, gather information

11   or enrich clinical trial information for experience with

12   resistant organisms.           People have suggested a number

13   of approaches and one of them, I think, which is

14   exhibited in the application that will be discussed

15   today is to study or design pathogen driven studies as

16   opposed to the indication driven studies that we are

17   used    to     seeing   in    the    course       of    anti-infective

18   applications.

19                    And that is to look at design specific

20   trials for particular resistant organisms.                      We've seen

21   this     in     applications         in     the        past      for      VRE

22   vancomycin-resistant enterococci and we will see it

23   today it that setting and also in the setting of

                                  S A G CORP.
     202/797-2525                 Washington, D.C.                 Fax: 202/797-2525

 1   methicillin-resistant staph aureus.

 2                  How   one     integrates       that    information

 3   collected from those sorts of trials into the overall

 4   portfolio, I think, is important in our consideration

 5   of how we determine whether or not a product has been

 6   demonstrated to be a safe and effective for the treatment

 7   of a resistant infection.

 8                  Finally, I will touch briefly on the issue

 9   of the comparator.     There are clearly certain areas as

10   we   address   the   important        needs   of     treatment     of

11   infections for resistant organisms, where there may be

12   either no approved comparator agent or an acknowledged

13   standard of care.     This raises particular challenges

14   in an area where we are used to seeing active control

15   trials and moreover where the ethical imperative is that

16   one cannot run placebo control trials or -- there is

17   that real issue.

18                  And I think people have looked at various

19   approaches.     One approach may be to do a historical

20   controls or look for historical controls.                   That is

21   problematic, particularly in areas where patients have

22   multiple core morbidities.            And the second approach

23   that has been discussed in several meetings that we've

                                S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1   had with this committee is to think about alternative

 2   designs, such as a dose comparison.

 3                  And I think we'll see an example of that

 4   in today's development, the development program that

 5   Pharmacia and Upjohn has designed for linezolid.

 6                  So I think, as you consider the application

 7   before you -- a number of these themes are evident in

 8   the development of this product and we certainly look

 9   forward to the committee's discussion on the issues that

10   relate to the development of this product and certainly,

11   the important indications for which this sponsor is

12   requesting approval.

13                  Thank you very much.

14                  CHAIRMAN RELLER:       Thank you Dr. Chikami.

15    I'd like to next invite Dr. Gary Tarpley to step forward

16   and initiate the presentation by sponsor for linezolid.

17                  DR. TARPLEY:      Good morning.       I'm Gary

18   Tarpley from Discovery Research at Pharmacia and Upjohn.

19    And it's my privilege to begin our presentations today

20   on linezolid.    Linezolid is a new anti-bacterial from

21   a an entirely new structure of class.       The oxazolynons.

22                  We are here today seeking approval of

23   linezolid for the following indications:           nosocomial

                             S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1   pneumonia, community acquired pneumonia, complicated

 2   and uncomplicated skin and skin structure infections,

 3   and vancomycin-resistant enterococcus faecalis and e.

 4   faecium infections.

 5                   In our presentations today we will present

 6   substantial data that linezolid is effective and well

 7   tolerated in treating these gram positive bacterial

 8   infections.      I will begin with an introduction to

 9   linezolid and a summary of its microbiology and Dr.

10   Hafkin will present linezolid's clinical pharmacology,

11   pharmacokinetics, and the results of our clinical

12   program.       Next Dr. Anderson will provided a brief

13   presentation summarizing our early experiences treating

14   children with linezolid.        And then to conclude I will

15   make a few final remarks.

16                   I'd like to begin today by reminding us of

17   the serious clinical challenge we face treating gram

18   positive pathogens.      As you know the five pathogens

19   listed here are very commonly isolated and the top three

20   of these are gram positive bacteria.        And these bacteria

21   are an increasingly common cause of serious infections.

22    Infections such as pneumonias, skin and soft tissue

23   infections, and bacteremias.

                             S A G CORP.
     202/797-2525             Washington, D.C.        Fax: 202/797-2525

 1                   Now one striking example of the increased

 2   prevalence of these bacterias in U.S. hospitals come

 3   from the SCOPE Project.       And in SCOPE surveillance for

 4   blood stream infections was monitored at 49 U.S.

 5   hospitals over a three year period.            And greater than

 6   10,000 infection were detected and as shown here the

 7   gram positive staphylococci and enterococci species

 8   were found to have count for almost 60 percent of those

 9   infections.

10                   Now not only are these serious hospital

11   acquired infections increasing in frequency, but they

12   are frequently caused by drug resistant pathogens.

13   The percentages of drug-resistance in the U.S. are

14   already very significant for many of the gram positive

15   bacteria.      And many of the drugs historically used to

16   treat these pathogens are losing their efficacy.

17                   Drugs    such       as       methicillin          for

18   staph-epidermidis or staph aureus, penicillins for

19   strep pneumo, or vancomycin for the enterococcus.

20   And in 1997 we had the first the reports of glycopeptide

21   intermediate      resistant     staph       aureus.         Perhaps

22   signaling the future loss of vancomycin for treating

23   s. aureus infections.

                             S A G CORP.
     202/797-2525             Washington, D.C.             Fax: 202/797-2525

 1                  So clearly the human and economic burden

 2   associated with these infections are very significant

 3   and new antibiotics are required.              We need new drugs

 4   not only to help us preserve the efficacy of our current

 5   agents, but we also need to address some of the

 6   limitations of these drugs.                  Limitations such as

 7   tolerability,      limited    formulations        or    routes      of

 8   administrations, as well as drug resistant issues.

 9                  We need new anti-bacterials that have new

10   mechanisms of action and thus have broad bacterial

11   coverages that are well tolerated and flexibly dosed

12   by both the I.V. as well at the oral routes.

13                  Linezolid provides one important solution

14   toward meeting these goals and was discovered at

15   Pharmacia and Upjohn as a result of a massive medicinal

16   chemistry effort that has thoroughly investigated the

17   structure      activity    relationship         of     the     phenyl

18   substituted oxazolynon ring by designing, synthesizing

19   and   evaluating    literally      thousands      of   individuals

20   compounds.

21                  Linezolid is an entirely synthetic molecule

22   from the oxazolynon class.          Of course, a class that's

23   not previously been found in nature.           It has a very broad

                               S A G CORP.
     202/797-2525              Washington, D.C.             Fax: 202/797-2525

 1   gram positive anti-bacterial spectrum, which includes

 2   coverage of both drug sensitive bacteria as well as

 3   bacteria resistant to any other drug class.

 4                  Linezolid is an inhibitor of bacterial

 5   protein synthesis, that blocks synthesis at immediate

 6   site of action.      And of course an important consequence

 7   of this novel mechanism of action, is that there is no

 8   pre-existing cross resistance between linezolid and any

 9   other marketed antibiotic.

10                  Now    specifically,         linezolid    disrupts

11   bacterial protein synthesis by blocking the formation

12   of the essential initiation complex.            This slide is a

13   schematic of the ribosome cycle in bacterial protein

14   synthesis and as you know, in this process a variety

15   of ribosomal nucleic acids complexed with multiple

16   protein factors to form a functional 70-S ribosome.

17                  And this the site of peptide bond formation.

18    Linezolid disrupts the initiation of this process by

19   binding principally to the 50-S ribosome and thereby

20   interfering with the ribosome binding of the essential

21   fMET transfer of RNA.       And as a consequence of these

22   binding interactions disrupts the initiation of peptide

23   bond synthesis and actually prevents the formation of

                              S A G CORP.
     202/797-2525             Washington, D.C.           Fax: 202/797-2525

 1   the first peptide bond.

 2                   Now, of course, you are very familiar with

 3   a variety of other antibiotics that are clinically

 4   useful and also protein synthesis inhibitors.                       Drug

 5   such as the aminoglycosides, the macrolides, or the

 6   streptogramins.        All of these drugs also inhibit

 7   bacterial protein synthesis, but they do so much later

 8   in the cycle by blocking the elongation step.

 9                   These drugs have no inhibitory activity

10   disrupting initiation.         And in contrast to these drugs,

11   our experience have determined that linezolid has no

12   effect on blocking proteins synthesis elongation, but

13   rather all of its inhibitory is a consequence of blocking

14   the initiation of this process.

15                   Linezolid has an excellent pharmacokinetic

16   profile,       which   includes        a      100   percent         oral

17   bioavailability in multiple dosage forms.                      We have

18   studied three forms.           An isotonic solution for I.V.

19   infusion,       tablets,       and     suspension         for       oral

20   administration.         All    of    these      dosage     forms     are

21   equivalent, meaning that equal drug exposures are

22   obtained       after   equal    doses         independent      of    the

23   formulation or the route of administration.

                                S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1                   Now this property of linezolid will be

 2   clinically very useful, allowing a switch from an I.V.

 3   to an oral form without the burden of a dose adjustment.

 4    Thereby, potentially minimizing the length of I.V.

 5   therapy and thus, for some patients will offer the

 6   benefit of a more rapid hospital discharge.

 7                   We have extensively studied linezolid in

 8   a variety of gram positive bacteria infections.

 9                   In a few moments Dr. Hafkin will review the

10   results of seven phase III studies in adults.                       The

11   protocol numbers are illustrated here as well as the

12   types as infections that we have studied.               In all seven

13   of   these     phase   III     studies,         linezolid   has    been

14   demonstrated to be effective and well tolerated in this

15   patient population.

16                   I'd now like to change topics and talk about

17   linezolid's microbiology.             Of course, focused on the

18   key gram positive strains that are relevant to the

19   indications.

20                   The entire in vitro susceptibility database

21   for linezolid submitted in the NDA, consisted of three

22   major parts.

23                   There were numerous pre-clinical studies

                              S A G CORP.
     202/797-2525                 Washington, D.C.           Fax: 202/797-2525

 1   that studied a variety of different isolates that were

 2   conducted by both Pharmacia and Upjohn, as well as

 3   multiple       outside       laboratories.              Isolates         were

 4   collected       and   surveyed       as    part         as    the     Sentry

 5   Surveillance Study collected in 1998, from over 30

 6   medical centers.

 7                   And of course, we've also obtained and

 8   studied a variety different isolates as part of our phase

 9   III program.

10                   Taken together the entire susceptibility

11   database consists of more than 4,000 isolates of

12   streptococci,         greater      than     12,000           isolates      of

13   staphylococci,         and      nearly          4,000        isolates      of

14   enterococci.

15                   The next several slides will summarize

16   linezolid's MICs versus the particular gram positive

17   strains, as well as focused on the key resistance issues

18   within each group.        Against the streptococci, linezolid

19   was    deeply    active      against      penicillin-           sensitive,

20   intermediate resistant, and resistant strep pneumo

21   isolates.       With MIC 50s and 90 values that are about

22   two-fold different from one another, and MIC 90 values

23   that are consistently between one and two micrograms

                                  S A G CORP.
     202/797-2525                 Washington, D.C.                 Fax: 202/797-2525

 1   per mil.       Against a group A and B strep.     Strep pyogenes

 2   and agalactiae, there were similar levels of activity

 3   with MIC 90 values of about two micrograms per mil.

 4                     Comparisons     of     the    MIC     population

 5   distributions of the strep pneumo isolates that we've

 6   studied in our phase III program compared with those

 7   isolates collected in the Sentry Program, shown here

 8   in gray, reveal a very similar population distribution.

 9    These data allow us to conclude that the strep pneumo

10   isolates that we've studied in our phase III program

11   were a very relevant collection and very representative

12   of isolates obtained very broadly as part of the Sentry

13   Surveillance Program.

14                     Against the staphylococci, linezolid was

15   equally        active   against    methicillin-sensitive          and

16   resistant staph aureus and staph epidermidis with MIC

17   90 values between two and four micrograms per mil.

18                     We've also had the opportunity to evaluate

19   a limited a number of glycopeptide intermediate staph

20   aureus and staph epidermidis and again linezolid was

21   equally active against these bacteria.                Perhaps not

22   surprising for an agent that has a very different

23   mechanism of action compared to the glycopeptides.

                                S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1                   Comparison          of      the     MIC    populations

 2   distributions of the staph aureus isolates studied in

 3   phase III, with the isolates obtained from the Sentry

 4   Surveillance Program, again reveals a very similar

 5   population distribution.            Allowing us to conclude that

 6   the staph isolates studied in our clinical program were

 7   a very relevant clinical population.

 8                   And    similarly          for       the    enterococci,

 9   linezolid      had    equal     activity         against   vancomycin-

10   sensitive      and     vancomycin-resistant                enterococcus

11   faecalis and e. faecium.              With MIC 90 values between

12   two and four micrograms per mil.                 And as we've seen with

13   the other gram positive bacteria the distribution of

14   isolates that we've studied of the enterococcus species

15   that we've studied in our phase III program were very

16   representative of isolates very broadly as part of the

17   Sentry Program.

18                   Now we've generally described the in vitro

19   antibacterial        activity      of    linezolid        as   generally

20   bacteriocidal          versus         the         streptococci         and

21   bacteriostatic versus staphylococci and enterococci.

22                   Now, of course, we have been very interested

23   in studying the potential for linezolid resistance and

                               S A G CORP.
     202/797-2525                  Washington, D.C.             Fax: 202/797-2525

 1   we have investigated this thoroughly in the laboratory.

 2    First, it's important to note that we were unable to

 3   select for linezolid resistant bacteria via spontaneous

 4   mutation and thus at the limits of our detection of this

 5   experiment     were    able     to    conclude    that     resistance

 6   development via spontaneous mutation is very rare.

 7                  We estimate a frequency less than one in

 8   ten to the minus ninth.

 9                  We     were     similarly        unable    to     derive

10   resistant mutants by standard chemical mutagenesis or

11   serial passage experiments through two-fold direct

12   concentrations.       These are methods that we and others

13   have used in the field routinely to isolate bacteria

14   resistant through a variety of other antibiotic classes.

15    Because these procedures were unsuccessful, we relied

16   on a much more rigorous selection process, which

17   involves a spiral-gradient serial passage method.

18                  This is a method that allows you to capture

19   very subtle changes in antibiotic susceptibilities that

20   result from prolonged selected, drug pressure.                        In

21   using this more rigorous selected method we were able

22   to isolate two strains of resistant bacteria.                  A strain

23   of aureus and on of e. faecalis for our mechanistic work.

                                  S A G CORP.
     202/797-2525                 Washington, D.C.            Fax: 202/797-2525

 1                   We determined that the linezolid resistance

 2   determinance resided within the 50-S ribosome.                    And

 3   genomic sequencing of the 23-S ribosomal RNA genes

 4   revealed the presence of new mutations that had not been

 5   previously been described for any other antibiotic

 6   class.

 7                   These mutations correlate with the changes

 8   of MICs with linezolid and result in the presence of

 9   verticular transversions within Domain 5 of the 23-S

10   ribosomal RNAs, a domain known to be very centrally

11   involved in peptide bond formation.

12                   Now very interestingly, it's known that the

13   gram positive bacterias contain five to six copies of

14   the 23-S ribosomol RNA genes.           And we determined that

15   the    linezolid   MICs   correlated         with   the   ratio    of

16   wild-typed mutant genes and then in fact, a significant

17   increase in MIC required mutations in at least two of

18   the six genes.

19                   So over all, the results of our laboratory

20   work on linezolid resistance indicated that it occurred

21   only    after   prolonged     selected       drug   pressure      and

22   significant changes in MICs

23   were not the result of a single point mutation, but

                               S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1   rather required multiple mutations and a multi-gene copy

 2   family.

 3                   Now    we   have    investigated    the   in    vivo

 4   antibacterial activity of linezolid thoroughly and the

 5   animal models that are most relevant to the indications

 6   today are shown here.         Linezolid is very effective in

 7   mirroring models of systemic infection with the gram

 8   positive bacteria administered by a variety of different

 9   routes.

10                   It's also very active in mouse models of

11   soft tissue infection.          Active in a localized group A

12   streptylcoccal myonecrosis model and a model of severe

13   pneumococcal pneumonia.

14                   Linezolid was also evaluated in the mouse

15   thigh infection model, which indicated that a key

16   correlate of it's efficacy would be drug concentrations

17   exceeding the MICs for approximately 40 percent of the

18   dosing interval.        So in summary our data demonstrate

19   that linezolid is a new antibacterial that has a very

20   broad gram positive coverage, because of the novel

21   mechanism action of this agent there is a lack of

22   inherent       cross    resistance        with   other    marketed

23   antibiotics.

                                 S A G CORP.
     202/797-2525                Washington, D.C.         Fax: 202/797-2525

 1                  We expect that linezolid therapy will be

 2   initiated principally in a hospital or the institutional

 3   care setting.    And the multiple dosage forms, coupled

 4   with the equivalent I.V. oral dosing will provided

 5   treatment flexibility needed to manage these serious

 6   infections.

 7                  Thank you and now I would like to turn the

 8   podium over to Dr. Hafkin.

 9                  DR. HAFKIN:    Thank you.    It's a pleasure

10   to be here this morning.     I'd like to speak first about

11   the pharmacokinetic profile of linezolid and because

12   it's oral bioavailibility is the most remarkable part

13   of the story.

14                  Let's start with the time concentration

15   curve that we see with linezolid at steady state at 600

16   mgs, twice a day.

17                  Within about an hour the concentration

18   maximum is reached typically.        Peak concentrations are

19   on the average about 18 micagrams per ml, and at twelve

20   hours when at the nader of the dosing interval, our

21   average concentrations are still right at the MIC 90

22   for staph aureus.      Note the enterococcus and strep

23   species, in this case, strep pneumonia, MIC 90s are noted

                            S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   by the dotted lines.

 2                  When   we      compare         the   oral    and     IV

 3   preparations, we have in the orange color the I.V.

 4   preparation, you see a brief peak, but within a couple

 5   hours the concentrations, on average, are equal to the

 6   oral preparation and it troughed twelve hours, these

 7   really looked very much alike.            The yellow preparation

 8   is noted here, the oral linezolid is there, as you can

 9   see the AUCs are virtually identical with time.

10                  So drug exposure is equal whether the drug

11   is given intravenously or orally.                   This is not an

12   example of a typical step down therapy that we have used

13   to in medicine.

14                  Looking at the clinical pharmacology, as

15   I've already told you, bioavailbility is a 100 percent

16   by AUC, there is very little food effect.             With the Cmax

17   decreasing slightly, 18 percent, but the AUC being

18   equivalent whether the drug is given with food or without

19   food.    The volume of distribution of 15 liters is about

20   the volume of water in the body.               Protein binding at

21   31 percent is low, and the half life is five to seven

22   hours.

23                  The drug is a weak reversible inhibitor of

                            S A G CORP.
     202/797-2525               Washington, D.C.            Fax: 202/797-2525

 1   monoamine oxidase.

 2                    Now linezolid is not a substrate or an

 3   inhibitor or an inducer of p450 enzymes.                           It is

 4   metabolized by oxidation and there are two primary

 5   metabolites          of   linezolid    in    that.       The     drug   is

 6   eliminated       primarily       through         the   urinary    tract.

 7   Thirty-five percent of the drug is eliminated through

 8   the urine unchanged, 50 percent of the drug is eliminated

 9   through the urine as the primary metabolites, and ten

10   percent of the drug is eliminated through the feces.


12                    There is virtually no active drug in the

13   gut.    It's very well absorbed and there's very little

14   detectable intact drug in the feces.

15                    Now what we know now is that primary

16   metabolites can accumulate in patients with severe renal

17   insufficiency.            Creatinine clearance of less than 30

18   mls per minute.           Both linezolid and the metabolites are

19   dialyzable.

20                    In summary then, what we know is that there

21   will be no dose adjust recommended for -- dose or route

22   of administration so that -- whether the patient were

23   to   take      the    oral    suspension,        the   tablet    or     the

                                   S A G CORP.
     202/797-2525                  Washington, D.C.            Fax: 202/797-2525

 1   intravenous    preparation,     the        pharmacokinetics     are

 2   virtually identical.     There is no need to change the

 3   relationship of food and meals.

 4                  Indeed, gender and age doesn't effect AUC

 5   or exposure to the drug because the concentration of

 6   the active drug doesn't really change whether the

 7   patient has severe or minimal or no renal insufficiency,

 8   we can't recommend a reduction in the dose of linezolid.

 9                  And because the drug is almost eliminated

10   through the urinary tract, our studies have shown that

11   there is no change in AUC with hepatic insufficiency.

12                  I'd like next to talk about the efficacy

13   that we've seen in our phase III trials.            I'm going to

14   use this road map as a way to aggregate the studies and

15   actually to remind where I am.

16                  The first study I'm going to discus is

17   Protocol 55, which is a complex skin and soft tissue

18   trial.

19                  In this trial we compared in a double-blind,

20   randomized, equivalence trial, linezolid 600 mgs to

21   oxacillin, two grams every six hours.             The study was

22   set up the patient would be randomized to one of the

23   two   treatments.     They    were     treated     initially     in

                             S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   hospital with I.V. medication.        When the physician felt

 2   it was clinically appropriate they switched to oral

 3   therapy.

 4                  If the physician felt that gram negative

 5   coverage was necessary aztreonam would be added.            Very

 6   few patients had aztreonam, because this was a hospital

 7   based protocol, these patients had fairly deep infection

 8   associated with wounds and abscesses.            Some severe

 9   cellulitis was recruited to the trial.         There was some

10   post-operative wound infections as well.

11                  Typical treatment was ten to twelve days,

12   although by protocol physicians could choose bewteen

13   a ten to 21 day period of therapy.         A test of cure was

14   two to three weeks after antibiotics were stopped, in

15   follow-up and the population that we recruited to this

16   protocol, 819 patients.

17                  Now I have here a series of histograms for

18   clinical cure.     ITT population represents that group

19   of patients that got at least one does of medication.

20                  Clinically evaluable population meant that

21   the patients got at least five days of antibiotics and

22   were called a clinical cure, or at least two days of

23   antibiotics and then they could be an evaluable failure.

                            S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1                  The microvaluable population was based on

 2   the clinically evaluable population.          You had to have

 3   a   baseline   and   a   be   clinically    evaluable    to   be

 4   microbiologically evaluable in our analysis.          Clinical

 5   cure for linezolid in the orange and in the gray was

 6   the oxacillin group.

 7                  Note that here is the confidence interval.

 8    And in each case we are equivalent or better to the

 9   comparator.    The missing and indeterminate number for

10   each of the patient population is noted here.           So that

11   we have about equal missing and indeterminate patients

12   for each of the arms of the trial.

13                  I'm going to use this design to report the

14   results of all of our trials.        When we look at pathogen

15   eradication rates, whether we are talking about staph

16   aureus or the strep species you see very comparable

17   results in terms of eradication in these patients with

18   complex skin and soft tissue infections.

19                  I'd next like to tell you about our out

20   patient study called, Protocol 39.            This study was

21   carried out in North America.           We compared linezolid

22   400 mgs to clarithromycin, 250 mgs, twice daily.

23                  The treatment duration was a week to two.

                              S A G CORP.
     202/797-2525             Washington, D.C.        Fax: 202/797-2525

 1    The test of cure was seven to 14 days and the population

 2   was 753 patients.

 3                  Again, the same cure histograms and for each

 4   of these populations, whether we consider the ITT or

 5   the clinical evaluable or the microevaluable, we are

 6   equivalent to the comparator.              Again, these are the

 7   missing patients and are indeterminate in each of the

 8   arms.

 9                  When we look at the pathogen eradication

10   rates in this study you see the same percentage of

11   eradication very comparable outcomes.

12                  In   conclusion,    we      feel   that   we    have

13   demonstrated that linezolid is quite effective at both

14   complicated and uncomplicated skin and soft tissue

15   infection.     The drug is quite effective in treatment

16   of staphylococcal Group A strep, and Group B strep skin

17   infections.

18                  I'd like next to turn to pneumonia.              The

19   first trial I'm going to describe to you is a trial that

20   was designed to recruit patients with community acquired

21   pneumonia, but in a patient population sick enough to

22   require hospitalization.

23                  The patients were randomized either to

                             S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   linezolid, and if necessary concomitant aztreonam, or

 2   they were randomized to receive ceftriaxone.

 3                  When the patients were stabilized and the

 4   physicians felt appropriate they could switch to oral

 5   therapy in both of these arms.          Treatment duration was

 6   seven to 14 days, typically patients got eleven days

 7   of therapy for both drugs.        The test of cure again, was

 8   two to three weeks after the end of therapy and we

 9   recruited 747 patients into this trial.

10                  Again    using     the        same   clinical     cure

11   histograms you see the same pattern of equivalence for

12   every other populations, every population up here.

13   Again we've got the missing and indeterminate listed

14   below.     When we look at the pathogen eradication rates

15   for this trial, again, you see the same comparability

16   of eradication for staph and strep species.

17                  Turning to an outpatient pneumonia trial,

18   where we recruited patients and randomized half to

19   linezolid 600 mgs twice daily and half to cefpodoxime

20   to 200 mgs twice daily.         These patients got ten to 14

21   days of therapy.       The test of cure was the same as the

22   one we've discussed, two to three weeks after the end

23   of therapy.    And we recruited 540 patients and treated

                               S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1   them as out patients in this trial.

 2                   When we look at all of the three populations

 3   and   again,    we   see    the    same        consistent   sense    of

 4   equivalence, if you'll note the confidence intervals

 5   are here and the missing and indeterminate patients are

 6   there.

 7                   So we have an equivalence again in this

 8   trial when we look at the pathogen eradication rates,

 9   the same patients were randomized to 600 mgs of linezolid

10   twice daily or to vancomycin one gram, BID.

11            Patients were given concomitant aztreonam most

12   of the time.          Very few patients did not receive

13   something      for   gram   negative       coverage.        Treatment

14   duration was seven to 21 days, test of cure was the same

15   two to three weeks.         We recruited 396 patients. More

16   than half were on ventilators at baseline, when they

17   were recruited to the study.

18                   And again looking at the cure histograms,

19   we have the same confidence of equivalent performance

20   of these drugs for each of the populations.                   When we

21   look at the pathogen eradication rate, you see the same

22   comparable results for both strep pneumo and staph

23   aureus.

                                 S A G CORP.
     202/797-2525                Washington, D.C.            Fax: 202/797-2525

 1                  In conclusion, we feel that we've shown

 2   quite    conclusively   that    the    drug   works   well    for

 3   community acquired pneumonia and nosocomial pneumonia

 4   due to strepto pneumonia and staph aureus.

 5                  Now I would like to turn to our resistant

 6   pathogen studies and the first study I would like to

 7   discus is MRSA.

 8                  To come into this study the patient had to

 9   have the strong epidemiologic clinical story that

10   suggested a gram positive infection that was resistant

11   to routine battle actems.       The patients were admitted

12   empirically into this trial, randomized either to

13   linezolid 600 mgs, twice daily, or vancomycin one gram,

14   BID, on the basis of gram stain or a positive culture

15   that had MRSA in it.

16                  Concomitant     aztreonam       was     allowed.

17   Treatment duration was seven to 28 days.         We recruited

18   460 patients into this trial.          You could be admitted

19   into this trial if you had MRSA in any part of you body.

20    This wasn't a site-specific, but it was a bug- specific

21   protocol.

22                  As you know primary source of infections

23   are listed here and about half of the patients that we

                             S A G CORP.
     202/797-2525            Washington, D.C.          Fax: 202/797-2525

 1   recruited, 230, had skin and soft tissue infection.

 2   About 99 patients had pneumonia as the diagnosis and

 3   other diagnosis realized in this protocol are listed

 4   there.

 5                   The   clinical     cures       for    this      group    of

 6   patients,      whether    we   consider        ITT     or     clinically

 7   evaluable or microevaluable populations are the same

 8   throughout.     Missing and indeterminate are here.

 9                   I should mention one other point, a few

10   patients who were found after admission to the study

11   to   be   infected      concomitantly         with    resistant        gram

12   negative pathogens and could not be managed with

13   aztreonam did get aminogylcoside, but the number is

14   small.

15                   Clinical cure of the patients for skin and

16   soft tissue infection are shown here, and again you have

17   the same comparable outcome, no matter which of the three

18   patient populations we consider.                     When we look at

19   pneumonia,      again    you   have      the     same       pattern      of

20   equivalence or similarity.

21                   Here are the pathogen eradication rates for

22   those patients at the end of the day, proven to have

23   MRSA, due to skin and soft tissue infection, very

                                S A G CORP.
     202/797-2525               Washington, D.C.                 Fax: 202/797-2525

 1   comparable outcomes.        And again for those patients with

 2   hospital pneumonia -- hospital acquired pneumonia or

 3   nosocomial        pneumonia,     have    the   same     comparable

 4   outcomes.

 5                    If you take all of the patients with MRSA,

 6   treated with linezolid and vancomycin you have the same

 7   comparability.       Now the VRE study was similar in many

 8   ways.    The requirement was that you would have to have

 9   VRE in some site in the body.

10                    You could have pneumonia, skin and soft

11   tissue,        urinary   tract   infection,     intra-abdominal

12   abscess, what-have-you.          But we had no comparator at

13   that time, when we started this study there was no widely

14   held effective study for VRE infection, there was no

15   consistent choice of our investigators.

16                    It was very difficult to try and come up

17   with a comparator that our community, the infectious

18   disease community felt comfortable with.

19                    As a result of that we compared what we felt

20   to be the best dose, 600 mgs of linezolid, twice daily

21   to the lowest dose of linezolid based on our animal model

22   work and our in vitro model would work.               When we give

23   200 mgs linezolid twice daily we're above the MIC 90

                                S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1   for 50 percent of the time in the typical patient

 2   population.       So we chose to compare 600 mgs of linezolid

 3   to 200 mgs of linezolid.

 4                    It's important to know that we felt that

 5   200 mgs of linezolid would have efficacy, we thought

 6   that we would be able to find better clinical outcomes

 7   and maybe faster clinical outcomes.              So this was a

 8   randomized,      double-blind     superiority     trial    and    we

 9   recruited patients to the either 600 or 200 mgs of

10   linezolid, concomitant antibiotics were allowed.

11                    Treatment duration was seven to 28 days.

12    And I'm going to report to you about two populations

13   here.

14                    We have a completed study that we call 54A.

15    We recruited after more than a year, and a hundred

16   sites, 145 patients with VRE infection.           We closed that

17   study and did a full analysis and then started a

18   supportive study that we call 54.            It had very similar

19   designs.       Eighty-two of the 186 patients recruited into

20   that trial are available for us to discuss today.

21                    Now for both of these protocols, about 20

22   percent of our patients had bacteremia, about 20 percent

23   had     intra-abdominal     infection.         Primarily     those

                               S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1   patients were post-liver transplant.                  Urinary tract

 2   infection was common and we had a surprising number of

 3   skin and soft tissue infections in this protocol.                      A

 4   few pneumonia.

 5                       I'm going to report on the data in several

 6   fashions.           When you look at all patients with VRE

 7   infection, no matter what their site of infection is,

 8   and   look      at    the   ITT,    the    clinical   evaluable     and

 9   microevaluable population.                The study is underpowered

10   and did not reach statistical significance, but we have

11   a consistent pattern of improved performance with 600

12   versus 200.

13                       When we look at the interim result for the

14   smaller protocol of 82 patients, you have the same

15   pattern        in    the    ITT    and    the    clinical   evaluable

16   population.           But we have very few microbiological

17   evaluable patients recruited into this small interim

18   group.

19                       When we looked at clinical cure by site of

20   infection.          If you look at intra-abdominal infection

21   and typically these people had peritonitis, liver

22   abscess, they had infections in the wounds that were

23   persistent and recurrent, they were fairly sick people

                                   S A G CORP.
     202/797-2525                  Washington, D.C.          Fax: 202/797-2525

 1   and in fact most of them had bacteremia with VRE.                   And

 2   then we had the bacteremia of unknown origin.                Urinary

 3   tract infection, and of course, as I told you skin and

 4   soft tissue infection.         But the point of this slide is

 5   that we had that same pattern of generally better

 6   outcomes with 600 over 200.

 7                   And this is the data for our interims

 8   analysis of 82 patients, again in a way you see a pattern

 9   of better outcomes with the high dose versus the low

10   dose.      If you look at patients from microbiologic

11   outcome perspective and you compare linezolid 600 to

12   linezolid      200   mgs,    you'll      see   that   this     is    a

13   statistically significant difference with this p value

14   associated with this comparison.

15                   So you did have a better microbiologic

16   outcome in our study of 145 patients if you were

17   randomized to 600 mgs versus 200 mgs.             Now in support

18   of the observations we have here, you may have heard

19   about our Compassionate Use Program which did not

20   recruit patients, but physicians who had patients for

21   which there was not practical therapy would call PNU

22   and patients could be treated with linezolid, 600 mgs,

23   twice daily for up to three months.

                                 S A G CORP.
     202/797-2525                Washington, D.C.          Fax: 202/797-2525

 1                       The patients that I'm going to report to

 2   you today on -- will be 230 patients that we collected

 3   in this Compassionate Use experience by June of 1999.

 4   To   date      we    have   more   than     750   patients   in    the

 5   Compassionate Use Program.

 6                       When you look at Compassionate Use patients

 7   you won't be surprised that many of them were not culture

 8   positive or baseline, didn't have follow-up cultures,

 9   weren't worked up very completely, so the number of

10   wholly evaluable patients we actually have is small.

11    But if you look at the patients with intra-abdominal

12   infections like peritonitis, liver abscess, the cure

13   rate is noted here.

14                       Patients with bacteremia are noted here,

15   the patients with complicated skin and soft tissue here,

16   and in general good outcomes with all of the patients

17   that were clinical evaluable had received at least ten

18   days of therapy and had microbiologically proven VRE

19   infection.

20                       So in conclusion, we feel that linezolid

21   600 mgs, twice daily is effective in the treatment of

22   vancomycin resistant enterococcus.                And we feel that

23   out two comparative trials, which by the way are the

                                  S A G CORP.
     202/797-2525                 Washington, D.C.          Fax: 202/797-2525

 1   largest comparative trials of VRE infection to date.

 2                       We did see a persistent and consistent

 3   improved outcome in patients randomized to 600 versus

 4   200 mgs.        And we think our Compassionate Use Study

 5   supports the results of the dose comparative study very

 6   well.

 7                       Now in terms of efficacy, we've shown

 8   efficacy       in    community      acquired    pneumonia       and    in

 9   nosocomial pneumonia.            We've shown good efficacy in

10   skin and soft tissue, both complicated and uncomplicated

11   and we've shown efficacy in MRSA infection and VRE

12   infection.             I'd   like     to    turn     to     resistance

13   surveillance.

14                       The clinical trials the we've performed

15   were organized in a fairly traditional fashion.                       All

16   organisms isolated at baseline were sent to a central

17   lab, every failure that resulted in a positive culture

18   at follow-up was sent to that central lab.

19                       So we've had very good data concerning MIC

20   creep, resistance of development to linezolid in our

21   clinical       trials.       We've    treated      more   than     3,000

22   patients in the past few years with linezolid at full

23   therapeutic doses and we've identified no staph species,

                                  S A G CORP.
     202/797-2525                 Washington, D.C.             Fax: 202/797-2525

 1   whether we're talking about a coagulase positive or

 2   coagulase negative staph that has become resistant to

 3   linezolid.       There has been no four-fold change in MIC

 4   in any isolate.

 5                    We've      identified     15    patients      who    had

 6   isolates       with   the    four-fold      elevation     of    MIC    at

 7   follow-up out of 832 patients of enterococcal infection.


 9                    Where did the patients the patients come

10   from?     Well, we identified one patient in our first

11   study of a 145 patients and in the second study, of a

12   186 patients with VRE infection, we identified five

13   isolates       enterococci     that     became    more     resistant,

14   four-fold resistant at the end of therapy.                      And in

15   Protocol 25 our Compassionate Use Program, 501 patients

16   had been exposed and treated with linezolid for their

17   enterococcal infection and nine resistant isolates were

18   found.

19                    When we looked at all of the cases we had

20   three stories that kept coming again, and again, and

21   again.

22                    Number     one.      Patients     had   in-dwelling

23   prosthetic devices.           Left ventricular assist device,

                                  S A G CORP.
     202/797-2525                 Washington, D.C.            Fax: 202/797-2525

 1   in-dwelling     catheter   that     couldn't    be   removed,

 2   intra-abdominal devices that couldn't be removed.             Or

 3   we had undrained abscesses.         We had intra-abdominal

 4   abscesses, we had abscess that could not be removed

 5   because of the surrounding dwelling device, or we had

 6   also a fair number of patients randomized to this 200

 7   mgs of linezolid twice daily.

 8                  Now I would like to turn to the safety

 9   information that we have been able to collect in our

10   phase III trial.   Could I go back one slide?    Okay, thank

11   you.    I must have changes my lecture over night.       Next.

12    Go forward please.

13                  The safety data that I'm going to be

14   discussing is based on three ideas.       Number one.      That

15   it -- when we went into development, when we started

16   this program, we knew from our preclinical work that

17   the drug was a mild reversible inhibitor of monoamine

18   oxidase.

19                  We also knew that when we pushed the dose

20   of linezolid high enough we could get transaminase

21   abnormalities, when we push the dose of linezolid high

22   enough we could get trans-hematopoietic suppression.

23                  In every experiment when we did that, we

                            S A G CORP.
     202/797-2525           Washington, D.C.         Fax: 202/797-2525

 1   found rapid reversal of the abnormality when the drug

 2   was stopped.       So we went through our phase III program

 3   specifically looking for the signals that we saw in these

 4   early studies.

 5                    What is a monoamine oxidase?               How do you

 6   look for it?       Well there are two classical syndromes

 7   associated with potent irreversible MAOI drugs, like

 8   the classic anti-depressants,                   Nardil.       It's been

 9   associated with side-effects when a serotonergic agent

10   is given.      A serotonergic agent would be a common cough

11   suppressant like dextromethorphan.

12                    When these drugs are given together you can

13   get fever, confusion, hyperthermia, with flushing, you

14   can get hypertension.           You can get tachycardia.              And

15   that's called the serotonin syndrome.

16                    There's another classical syndrome that's

17   called     the    adrenergic      syndrome,       or    the     tyramine

18   syndrome.      Where you get hypertension and you can get

19   very accelerated hypertension.                  We've shown in the

20   experiments that follow that linezolid is a weak, is

21   not irreversible, very revisable MAOI inhibitor.

22                    So let me share with you some of the phase

23   I   trials       that   we    did.        We     used     the   classic

                                  S A G CORP.
     202/797-2525                 Washington, D.C.             Fax: 202/797-2525

 1   dextromethorphan as a serotonergic agent drug.                We

 2   treated patients with linezolid and dextromethorphan,

 3   20 mgs, every four hours.            We found no change in

 4   temperature, blood pressure, no cognitive difference.

 5    It was a negative study.

 6                  In another phase I trial we treated patients

 7   with linezolid in tyramine.        And we found that that it

 8   required more than a 100 mgs of tyramine to get a

 9   detectable blood pressure increase.         Now to give you

10   some reference, the typical glass of wine will have one

11   mg of tyramine.      The typical serving of blue cheese

12   might have two milligrams of tyramine, the typical

13   elaborate blood sausage might have ten mgs of tyramine.


15                  So we feel that no diet would give you a

16   100 mgs of tyramine and feel that no food restriction

17   would be necessary when using linezolid.

18                  Next we turn to another study of concomitant

19   medication where we treated patients with linezolid and

20   phenylpropanolamine, or pseudo-ephedra.         And what we

21   found in these studies is that we could show detectable

22   increases of blood pressure when used these two drugs

23   concomitantly.     When we took our patients in the phase

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   I unit and treated them with placebo, we could increase

 2   their blood pressure on the average of eight mm of

 3   mercury and we might have a range of seven mm of mercury.

 4    And    when   we   treated   them   with   linezolid    we    had

 5   essentially the same response that we got with placebo.

 6                  When we gave patients phenylpropanolamine,

 7   again we got an increase in response, it was just a little

 8   bit    more    than    the    placebo.        When   we       used

 9   phenylpropanolamine and linezolid we had a detectable

10   change from the placebo.       Note that this range of blood

11   pressures is still the range of blood pressures that

12   you get in daily living.       I would assume that my blood

13   pressure is at least that high at the present time.

14                  (Laughter.)

15                  DR. HAFKIN:      So -- it's not outside the

16   normal daily experience when you do get concomitant

17   phenylpropanolamine and linezolid.

18                  Now what happened in our trials?         When our

19   phase II trials -- we were cautious, we warned our

20   physicians participating in the phase II trials to watch

21   for the possibility of an interaction between MAOI

22   potentiator and MAOI drugs, and linezolid.

23                  So what did we find?         We found that our

                              S A G CORP.
     202/797-2525             Washington, D.C.        Fax: 202/797-2525

 1   investigators recruited 247 patients out of the 867

 2   patients that were actually recruited to our trials that

 3   had some exposure to these -- either potentiator of MAOI

 4   effect, at least sometime during the treatment interval

 5   and they -- the investigators were trained to look for

 6   trends of hypertension, arrhythmia, what-have-you.

 7                  When we looked at the data we found that

 8   there were no adverse events attributable to linezolid

 9   in the phase II trial.       Food restrictions were lifted

10   by us in our phase III trials as result of learning that.

11    We had been terribly harsh in our phase II trial,

12   warning people against people against American cheese.

13    Warning them against the most elaborate sort of diet,

14   I mean you had to stay on peanut butter and white bread

15   initially.

16                  And when we realized that the protocol had

17   recruited 247 patients that had potentially interacting

18   drugs and we had seen nothing, we were very relieved.

19                  So in our subsequent phase III trials, all

20   though we warned the physicians that the potential could

21   exist, if somebody drank a full bottle of soy sauce.

22                  (Laughter.)

23                  DR. HAFKIN:       The reality is that the

                            S A G CORP.
     202/797-2525            Washington, D.C.      Fax: 202/797-2525

 1   patients any problems in our phase II trials, so we

 2   lifted those restrictions.      There were no restrictions.

 3    And we lifted the restriction about MAOI, we said that

 4   if you have a patient that needs the therapy, you have

 5   to watch them.    So there was a warning not only in the

 6   protocol, but in the consent form for the patient.

 7                  Well, what was our experience in phase III?

 8    We identified 632 patients that had linezolid and

 9   concomitant MAOI potentiator.              Something that could

10   potentiate the effect of the MAOI drug.           And these are

11   the drug classes realized in our trials.

12                  Well, what did we find after we sliced and

13   diced the data?      We found, that well -- we had found

14   the 632 patients.      We looked for adverse events such

15   as hypertension, hypothermia, things like that.

16                  We found 13 patients that have hypertension

17   as an adverse event.    Twelve of the 13 patients who have

18   hypertension as an adverse event were felt by the

19   investigators to have nothing to do with linezolid.

20   One investigator felt his episode of hypertension was

21   related to linezolid.

22                  There was another experiment embedded in

23   our phase III trial.      We asked in Protocol 55 and 48

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   for investigators who were going to use MAOI drugs or

 2   potentiator of MAOI effect to take blood pressure before

 3   that drug was given and after that drug was given.             And

 4   that follow-up vital sign was supposed to be within two,

 5   two and half hours.

 6                  Now it's a limited experiment, because most

 7   of our phase III investigators really didn't do that.

 8    But we have observations on about 100 patients here.

 9                  When you look at the baseline blood pressure

10   for those patients who got linezolid, compared to the

11   comparator, look at the post-treatment with concomitant

12   medicine blood pressure or the potentiator and the

13   comparator.     You see there is no difference in the pre

14   and post-blood pressure results.

15                  Here is the range of blood pressures noted

16   in the experiment, here's the range of the comparator

17   blood pressure.      That's for systolic and the same is

18   true for diastolic blood pressure.          Again it was a small

19   study.     It was limited, but it was out there in the field

20   and I was just going to say almost 100 patients were

21   recruited into the trial and the data was collected in

22   this fashion.

23                  Now what are our conclusions in terms of

                             S A G CORP.
     202/797-2525             Washington, D.C.          Fax: 202/797-2525

 1   mono amine oxidase inhibition?             Well, we've proven in

 2   our preclinical and in our phase I unit that we do have

 3   a weak and reversible MAOI effect and in phase II and

 4   phase III we had 879 patients exposed to linezolid and

 5   a potentiator of MAOI effect.

 6                  We found one patient in whom blood pressure

 7   was attributed to the combination of drugs.               We feel

 8   that the risk of MAOI effect is small enough that

 9   benefit/risk relationship for linezolid in clinical use

10   is not effected.

11                  Next I would like to turn to the traditional

12   safety analysis that we do and for that I'm using every

13   phase III comparative trial observation that I have,

14   so we're including Protocol 55, the complicated skin

15   and soft tissue trial, 39A, which is the large North

16   American skin and soft tissue trial as an out patient,

17   and then there is a smaller study that was carried out

18   in Europe, Latin America, and Asia.

19                  Where we used the same dose, the same

20   protocol as the 39A along with our pneumonia trial and

21   our comparator trial for MRSA.        So all in all I'm taking

22   2,046 patients randomized and treated with linezolid

23   and comparing them to 2,000 patients treated with

                             S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   comparator.

 2                  On this slide, this is one of two slides

 3   that I have, every AE that was reported in our trials

 4   with or without attribution to drug.            And if you look

 5   at the most frequent adverse events reported they are

 6   typical for antibiotic trials.              The typical nausea,

 7   vomiting, and diarrhea.       The results are comparable for

 8   both of the populations.

 9                  Here's the continuation of the greater than

10   two percent.    AE presentation, we have essentially the

11   same rates for both of the populations, linezolid and

12   the comparator group.

13                  When we look at drug related adverse events,

14   these are adverse events attributed by the physician

15   to be due to the drug being used.             You see the same

16   pattern of diarrhea, nausea, and headache for both

17   linezolid and the comparators.               Taste alterations,

18   malaises are seen in comparable numbers, please note

19   that this abnormal LFT is slightly lopsided with an

20   increased number associated with linezolid.           I'm going

21   to share an analysis of more quantitative data in just

22   a moment.

23                  Looking   at   the    common    serious   adverse

                              S A G CORP.
     202/797-2525             Washington, D.C.          Fax: 202/797-2525

 1   events, these events are associated with the patients

 2   underlying       illness.        The    infection        that's        being

 3   treated, not drug related in any of these cases.

 4   Turning to the laboratory assessment.

 5                    What did we do with the safety labs that

 6   we collected on these patients?                 Well, as always we did

 7   the mean standard deviation and there are no difference

 8   between the experience linezolid treated patients and

 9   the comparator treated patients.

10                    We also used regression analysis to look

11   for differences between these populations and we found

12   that linezolid wasn't any different than the comparator.

13    The deed I'm going to show you is outlier analysis,

14   hazard     function       analysis,    and       an   extreme       outlier

15   analysis to assure you that there are really are no

16   significant          differences       between          the      linezolid

17   experience and the comparator experience.

18                    If we look at patients that significant

19   abnormalities        in   biochemistry,          with    typical       liver

20   function tests and the amylase, lipase, bilirubins,

21   creatine       and   kinase.       These        are   very      comparable

22   numbers, because we had the report of increased ALT in

23   our adverse event profile.

                                  S A G CORP.
     202/797-2525                 Washington, D.C.                 Fax: 202/797-2525

 1                  Let me next go to the ALT analysis and show

 2   you what we did.    This is called a hazard function, at

 3   least we call it a hazard function.         Where the risk of

 4   the abnormal result is here, and the time that that

 5   result occurs is here in terms of days.         Linezolid in

 6   the orange curve, the comparative it the gray curve and

 7   we see no difference between these two hazard function

 8   for linezolid treatment over time.

 9                  Now this is every patient that developed

10   a significantly abnormal liver function abnormality.

11    And in this case we're talking about ALT, patients

12   treated with linezolid.      As you can see the great body

13   of patients have low level ALT abnormalities, many of

14   them fall to normal within the treatment period.            This

15   is the extent of ALT abnormality, this the baseline,

16   this is the number of days of therapy, this -- from this

17   point on is greater than 13 days.          It's a complicated

18   curve.

19                  This green line is the switch from treatment

20   to post-treatment and this is the follow-up period.

21   As I was saying most every patient will have low level

22   ALT abnormality and it'll fall down within the normal

23   treatment period.     A few patients went wildly up high

                             S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1   -- up here.      And of these patients, all of them have

 2   pneumonia and all of them came down shortly after

 3   therapy, without adverse events.            None of them dropped

 4   out of the study.        Indeed they all had lower lobe

 5   pneumonia, all except for one, who had a right lower

 6   lobe pneumonia.     He had a left lower lobe pneumonia and

 7   had a history of hepatitis and that's this one right

 8   here.

 9                   Let's look at the comparator now.        It's the

10   same pattern.     There is no difference.       We have the same

11   story of transient increases in transaminase, wildly

12   high and then resolved.       And then these two patients,

13   typically have right lower lobe pneumonia.             We have the

14   same    dated    presentation     with      baseline    --   here.

15   Treatment -- here.      The green line demonstrating the

16   post-treatment phase.

17                   Lets's look at the hematologic indices.

18   If we look at the red cell series or the white cell

19   series, it's really no difference between linezolid and

20   the comparator.     There may be a difference here in the

21   platelet count, so let's investigate that with more

22   care.

23                   When we look at linezolid treated patients,

                              S A G CORP.
     202/797-2525             Washington, D.C.           Fax: 202/797-2525

 1   the orange line and compare it to the gray line, the

 2   comparator, you see that there's no difference until

 3   you get to 16 days when there seems to be a divergence.

 4    This divergence represents almost one percent of

 5   patients.      And it represents about 16 patients, so the

 6   difference between this line and this line is 16

 7   patients.

 8                   What's happening in the patients treated

 9   with linezolid?     This is a time of analyst result curve,

10    It's a bit complicated and I appreciate that this may

11   be the first time you've seen something quite like this.

12    What we did is, we plotted the patients individuals

13   analyst over time.      This is the baseline isolate and

14   these are results of the various platelet counts through

15   time and this is the first post-treatment day.

16                   And you can see, I think, that the drop in

17   platelet count in patients treated wit linezolid is

18   slow.     that most of the patients, about 50 percent,

19   although you can't see it in this slide, because of the

20   way we display it.     More than 50 percent of the patients

21   that actually had low platelet count on therapy,

22   actually had low platelet count at baseline.

23                   The patients rapidly increase in their

                              S A G CORP.
     202/797-2525             Washington, D.C.     Fax: 202/797-2525

 1   platelet count post-therapy and this lowest of the low,

 2   this 19,000 platelet count patient had no bleeding

 3   episodes and basically this patient's initial platelet

 4   count was here.

 5                  So they went from about 50,000 to about

 6   19,000 on therapy.   Let's look at the comparator.       This

 7   is what happens in those patients that have significant

 8   platelet count abnormalities with the comparator.        It's

 9   exactly the same curve.     It is qualitatively identical.


11                  So what do we conclude from our analysis

12   of the platelet data?    That we've found two risk factors

13   associated with decreased platelet counts.     One is that

14   if you have a low baseline value, you don't get better

15   with linezolid therapy.       The underlying illness that

16   caused the thrombocytopenia isn't affected by it.

17                  We saw a slight increase in the risk of

18   platelet counts dropping after more than two weeks of

19   therapy, we've found that the decrease in platelet

20   counts were mild, they weren't rapid, they weren't

21   precipitous, and they were reversible.      And we had no

22   clinical consequences in any of the patients who had

23   decreased platelets counts in our trial.

                             S A G CORP.
     202/797-2525            Washington, D.C.      Fax: 202/797-2525

 1                     And finally, what do we have to say in terms

 2   of the safety of linezolid therapy with 600 mgs for up

 3   21 days?        Well, the common side effects were the ones

 4   you've         associated    with     antibiotics           everywhere,

 5   diarrhea, nausea, and headache.                 There is no clear

 6   association between adverse events and the use of

 7   concomitant medication.

 8                     We didn't see a pattern that demonstrated

 9   a   monoamine        oxidase    inhibition          caused     clinical

10   detectable adverse events.           Changes in platelet events

11   were mild and transient and frankly we're not sure if

12   it's related to linezolid therapy.

13                     Well, I'd like to ask Dr. Don Anderson to

14   come up and report to you on our early pediatric

15   observations.        Thank you.

16                     DR. ANDERSON:     Good morning.       I'm proud to

17   have the opportunity to speak briefly to you today about

18   the development and our progress in the development of

19   linezolid, specifically for children.

20                     Pharmacia and Upjohn's commitment to the

21   earliest        possible    development        of    both    oral   and

22   intravenous formulations for pediatric use is certainly

23   justified and for several reasons.

                                 S A G CORP.
     202/797-2525                Washington, D.C.            Fax: 202/797-2525

 1                  It is self evident to the pediatricians here

 2   today.     An unmet medical need in children is clearly

 3   not less urgent than in adults, historically, that has

 4   always been true.     Gram positive bacteropathogens are

 5   of major importance in children in the emergence and

 6   continued emergence of PRSP, MRSA, VRE are of serious

 7   concern in the pediatric community.

 8                  Currently      few       safe       and      effective

 9   therapeutic options exist in the setting of infections

10   due to suspected or proven resistant gram positive

11   pathogens.     Some cases there are actually no options

12   as reflected by our experience in the Compassionate Use

13   Protocol 25.

14                  We believe that a critical need exists, even

15   now for alternative I.V. and oral agents for management

16   for serious infections in both healthy children in high

17   risk pediatric groups, including neonates.

18                  Linezolid is remarkably well positioned to

19   address these concerns.       This is true, not only because

20   of it's spectrum of anti-microbial activity but in

21   addition to other attributes.          Among these include its

22   potent     bactericidal    activity          for   pneumococci,        an

23   activity for virtually all isolates studied throughout

                               S A G CORP.
     202/797-2525              Washington, D.C.               Fax: 202/797-2525

 1   the world.

 2                  In addition, opportunities exist here for

 3   flexible dosing regimes, such as the I.V. to oral switch.

 4    So for these reasons it is appropriate for the committee

 5   to consider even for a few minutes our experience and

 6   progress in use of linezolid in children by this sponsor.

 7    We do not seek specific indications today.

 8                  You can bet we'll be back to do that, but

 9   we want to assure the advisory committee and the

10   pediatric healthcare community that we will carry out

11   the    requisite     clinical       trials      to   make     linezolid

12   available for children soon after it's registration.

13                  So our pediatrics program to date has

14   included phase I pharmacokinetic studies in patients

15   ages 3 months to 17 years.          Planned studies will include

16   pharmacokinetic       assessments          in     all   age    groups,

17   including      neonates,       to        define      optimal     dosing

18   requirements.

19                  Two     phase        II     studies      focused        on

20   pneumococcal     disease     have        been   completed.        These

21   included population pk, safety and efficacy assessments

22   in children with community acquired pneumonia and acute

23   middle ear disease.        Linezolid use in children in our

                              S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1   Compassionate       Use      Protocol           has     provided      some

 2   encouraging experience of systemic VRE infections.

 3                  Enrollment in this program, as Dr. Hafkin

 4   indicated continues to increase, to allow linezolid

 5   treatment in pediatric patients with essentially no

 6   other therapeutic options.

 7                  Finally,       a    phase        III    study    has   been

 8   carefully planned together with the Food and Drug

 9   Administration      and      is    very     near        implementation.

10   Comparative pharmacokinetic data from these pediatric

11   trials has revealed that the clearance of linezolid when

12   adjusted for body weight is inversely proportional to

13   age.

14                  Higher        clearance           in      this     inverse

15   relationship is especially apparent in patients less

16   than    five   years    of    age,    in    which       optimal    dosage

17   requirements will require further definition.

18   In children five years of age or older, receiving 10

19   mgs/kg oral doses, twice a day, the steady state values

20   for    clearance,      volume       for     of        distribution     and

21   elimination half-life are similar to those for adult

22   patients.

23                  Now the designs for two completes phase II

                                  S A G CORP.
     202/797-2525                 Washington, D.C.              Fax: 202/797-2525

 1   studies are shown here.       A dose of 10 mgs/kilo given

 2   BID was selected for both based on phase I data.           These

 3   were open label, uncontrolled studies.            The primary

 4   objectives     of    which    included      the    accrue      of

 5   pharmacokinetic and safety data in exposed children.

 6                  Populations were selected, however, in an

 7   attempt to target pneumococcal infections.                 These

 8   included seriously affected and hospitalized patients

 9   with the community acquired pneumonia, this was Protocol

10   45.     In patients with acute otitis media enrolled at

11   investigative sites with a high prevalence of antibiotic

12   resistant pneumococci and with an emphasis on previously

13   treated and refractory disease.

14                  Now since the clinical and microbiological

15   outcomes of these trials were not compared or controlled

16   and because the results are described in detail in the

17   brochure that you've already reviewed, I will only

18   indicate this morning that we are very encouraged with

19   the over all results especially with the management of

20   severe pneumococcal infections, including those due to

21   PRSP.

22                  The   safety   analysis     in   these   studies

23   included assessments of adverse events, chemistry, and

                             S A G CORP.
     202/797-2525            Washington, D.C.          Fax: 202/797-2525

 1   hematologic      safety      studies,          and    vital        sign

 2   determinations.        Overall,        both      I.V.       and    oral

 3   formulations of linezolid were well tolerated in these

 4   pediatric populations.          As shown, gastrointestinal

 5   symptoms and nonspecific skin eruptions accounted for

 6   the most common drug related adverse events reported

 7   in these studies.

 8                  Overall,    however,          these   were    of    mild

 9   intensity, transient and self limiting.               Only four of

10   143 patients in these two trials were discontinued from

11   the study because of adverse events considered to be

12   drug related by their investigator.                   All reported

13   serious adverse events are summarized on this slide.

14                  So this is the worst case scenario over the

15   entire safety profile.       These included one example each

16   of       bronchiolitis,         convulsion,           neutropenia,

17   pneumothorax, and vomiting.

18                  Among these the occurrence of neutropenia

19   and vomiting was the only basis for discontinuation of

20   linezolid treatment.       All serious adverse events were

21   self limiting and only one, the example of neutropenia,

22   was considered to be drug related by the enrolling

23   investigator.

                               S A G CORP.
     202/797-2525              Washington, D.C.              Fax: 202/797-2525

 1                     So   in    summary,       the      results    of    our

 2   preliminary studies in children are very encouraging

 3   to us.    They certainly justify our plans to conduct more

 4   definitive studies as part of a phase III pediatric

 5   program        which   is   near    implementation.            Completed

 6   pharmacokinetic studies to date suggest that a dosing

 7   regime of 10 mgs/kg BID in children five years of age

 8   or older approximates the pharmacokinetics and exposure

 9   in adults receiving 600 mgs BID.

10                     Further,      pharmacokinetic          studies      are

11   underway to define the appropriate dosing regimes under

12   five years of age including detailed studies in term

13   and pre-term neonates.

14                     In closing that Pharmacia and Upjohn will

15   certainly develop linezolid with the unmet needs of

16   children in mind, we don't have all of the answers, but

17   we will pursue these aggressively because children as

18   well other patient groups deserve the benefits of this

19   unique and very exciting agent.                  All of us at Pharmacia

20   and Upjohn look forward to working with the FDA and

21   committed         pediatric        investigators         towards      the

22   achievement of this goal.           Thank you for your attention.

23                     Now I would like to introduce Dr. Gary

                                  S A G CORP.
     202/797-2525                  Washington, D.C.            Fax: 202/797-2525

 1   Tarpley once again, who will make some concluding

 2   comments and restate the indications for which we seek

 3   approval today.

 4                     DR. TARPLEY:        Over the past hour you've

 5   heard quite a bit about linezolid.              In closing I'd like

 6   to briefly review a few of the salient facts.

 7                     There is an important unmet medical need

 8   treating Gram positive bacterial infections, and our

 9   current        antibiotics    have      significant    limitations.

10   Linezolid addresses many of these limitations.                  It has

11   a very broad Gram positive coverage and a unique

12   mechanism of action.

13                     Our   clinical        results      indicate      that

14   linezolid       is   effective     in    treating     Gram   positive

15   bacterial        infections     and     that    it   has     important

16   advantages, such as it's favorable PK profile and

17   multiple dosage forms.

18                     Linezolid was also well tolerated in this

19   patient population.            Overall linezolid has a very

20   promising safety profile.                The extensive clinical

21   results presented today support the use of linezolid

22   in patients with known or suspected Gram positive

23   infections.

                                  S A G CORP.
     202/797-2525                 Washington, D.C.           Fax: 202/797-2525

 1                  As I've indicated, we expect that linezolid

 2   therapy will be initiated principally in the hospital

 3   or institutional care setting, and that it will provide

 4   the needed flexibility and the clinical management of

 5   these serious infections.

 6                  Our studies have demonstrated significant

 7   clinical benefits of linezolid administered at doses

 8   of 400 or 600 milligrams twice a day to adult patients.

 9    The data presented today strongly support approval of

10   linezolid for the following indications shown here, and

11   we seek the committee's concurrence that linezolid is

12   effective and safe in the treatment of these infections.

13                  Thank    you    for    your     attention.       That

14   concludes our presentations, and we'd be happy to answer

15   your questions.

16                  CHAIRMAN RELLER:        Thank you, Dr. Tarpley,

17   and I'd also like to say that it's been most helpful

18   to have such a comprehensive, clear, sharply focused

19   and       superbly      organized            presentation       from

20   Pharmacia/Upjohn.

21                  The     data,    the     issues     presented       by

22   Pharmacia/Upjohn are now open for committee questions.

23    Dr. Norden.

                              S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1                    DR. NORDEN:     I'd also add my compliments

 2   on the presentation.         I have one concern, and that's

 3   the two metabolites that you have.               In terms of your

 4   recommendation for no addressment of dosage of patients

 5   with renal failure, one, what do we know about the

 6   toxicity of the two metabolites, and, two, what do we

 7   know about their potential at high levels to interact

 8   with MAO inhibitors.

 9                    DR. TARPLEY:     I'd like Dr. Jungbluth -- is

10   this on?       Thank you -- I'd like Dr. Jungbluth to address

11   that question from our clinical pharmacology group and

12   then Dr. Slatter from our toxicology group.

13                    DR.   JUNGBLUTH:         Gail    Jungbluth     from

14   Clinical Pharmacology.

15                    First, to address your question on the no

16   dose adjustment, I'd like to go back to the linezolid

17   pharmacokinetics in renal impairment to show why we feel

18   that is necessary.

19                    Could I have this slide on, please?

20                    This graph shows a single dose studies in

21   patients with varying degrees of renal impairment in

22   its linezolid plasma concentration versus time curve,

23   and you can see that regardless of renal function,

                               S A G CORP.
     202/797-2525               Washington, D.C.          Fax: 202/797-2525

 1   similar concentrations are achieved of linezolid, and

 2   this is why we feel that no dose adjustment is needed

 3   in order to maintain parent linezolid concentrations.

 4                     And the linezolid metabolites do accumulate

 5   in renal impairment.             We have found this in the single

 6   dose study.           Because of this finding we evaluated the

 7   two primary metabolites and have found that in patients

 8   in the severe impairment group                 and the anuric patients

 9   do accumulate, and the patients with moderate impairment

10   do   not       have    a    significant       accumulation     of   these

11   metabolites.

12                     So we evaluated these in multiple dose

13   setting using patients in our compassionate use study.

14                     The next slide, Dennis.

15                     As I said, our single dose data indicated

16   accumulation of these metabolites, and we then evaluated

17   multiple dose data in patients with severe impairment

18   with serum creatinines of over four or a creatinine

19   clearance estimate of under 30 mLs per minute.

20                     What we have found is that these levels

21   plateau during one week of dosing, and our data up to

22   four weeks of dosing shows no additional accumulation

23   of   these,      and       we   have   also     found   that   linezolid

                                     S A G CORP.
     202/797-2525                     Washington, D.C.          Fax: 202/797-2525

 1   metabolites are removed by dialysis.

 2                  I think another of your questions is what

 3   do we know about the linezolid safety and MAOI potential,

 4   and Dr. Greg Slatter will talk about the MAO.

 5                  DR. SLATTER:      Hello.    I'm Greg Slatter

 6   from Drug Metabolism Research.

 7                  We have investigated the MAO inhibition

 8   potential of linezolid and its primary metabolites.

 9                  Slide up please.

10                  We've used human MAO A and         MAO B in a

11   specific enzyme kinetic assay.        Here you see the results

12   for linezolid.     The KI for MAO A, which mediates the

13   pharmacokinetic drug interactions of hypertension, et

14   cetera, is 56 micromolardeaths.        The KI of the two major

15   human metabolites, first the minor metabolite 20-fold

16   higher at 1.1 millimolar, almost too slow to measure,

17   and this one about threefold higher at 147 or at 1.47

18   -- 147 nanamolar -- 147,000 nanamolar.         My apologies.

19                  So the MAOI potential of these two agents

20   against MAO A are significantly less than the parent

21   drug linezolid, it itself being a mild competitive

22   reversible inhibitor.

23                  DR. HAFKIN:     And finally, if you don't

                            S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1   mind, I'll share with you some of the clinical data that

 2   we have in patients that receive linezolid.                         In our

 3   Phase III trials, we did not exclude patients that had

 4   renal insufficiency.           We recruited patients of varying

 5   renal insufficiency, and we've taken and put to these

 6   tables     the    most   severe,       patients      that     had    serum

 7   creatinines of four or greater.                   The greatest serum

 8   creatinine in this small subgroup of patients was 12.

 9                    Could I have S-194, please?

10                    And if you'll look, we've identified 17

11   patients in the Phase III clinical database that

12   received linezolid and had very high serum creatinines

13   and 15 comparator agents.             As you can see, the typical

14   therapy was about ten days.                 The ITT population is

15   generally the same size.            The numbers of patients with

16   adverse        events    in    either       one     of   these       small

17   observational groups are small.                  I mean, we have, you

18   know, only a small number of patients this sick would

19   not have some adverse event.

20                    If you look at the number with drug related

21   adverse events, we've got comparable numbers.                           The

22   number of serious adverse events, they're comparable.

23    If you look at the number of deaths or adverse events

                                   S A G CORP.
     202/797-2525                  Washington, D.C.              Fax: 202/797-2525

 1   leading to discontinuation of the medicine, they're also

 2   very comparable.

 3                  Go to the next slide in this series.

 4                  Looking at -- this will be fine, 197, please

 5   -- if you look at the specific medical term for the

 6   adverse events, you see there is really, I mean, because

 7   the number of patients is so small; they are really very

 8   comparable.     The adverse events like infection and

 9   sepsis are the underlying illness.         The other adverse

10   events are rare and not properly related to anything

11   concerning the drug.

12                  Could I have the next slide, please, with

13   an extension of this slide?

14                  Again, the serious adverse events that were

15   seen for the linezolid and comparator are very similar

16   in the sense that there's no pattern.            There's no

17   signal.

18                  The only additional line of evidence I have

19   to answer the question how safe is linezolid in this

20   patient population comes from our compassionate use

21   trial.     If we look at the database of the first 230

22   patients we have reported to the FDA and FDA has had

23   a chance to look at, you'll find 34 patients with very

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   poor renal function, estimated creatinine clearance of

 2   less than 30 milliliter per minute.

 3                  When you look at those patients and you

 4   compare them to patients with renal insufficiency that

 5   is mild, there is no different -- or renal insufficiency

 6   that is normal, there is no difference in the pattern

 7   of serious adverse events for these populations.

 8                  And in the compassionate use trial, we have

 9   patients that have taken the drug for up to three months

10   who have nothing but an occasional dialysis.            So we're

11   not implying that this is an adequate safety database

12   to assure safety, but what we feel confident is that

13   there's no clear signal of increased toxicity.

14                  CHAIRMAN RELLER:      Dr. Murray.

15                  DR. MURRAY:    Barry, with some of the data

16   from the FDA, it would look like one of the pedalites

17   pedialites -- and with a three times less inhibition

18   of an MAO you still might expect that.         So how would

19   you address that?

20                  DR. HAFKIN:    Yes, we agree with you, Dr.

21   Murray, that the way to deal with this is to share with

22   the physician the lack of information that we have and

23   the fact that efficacy safety in this small population

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   group can't be spoken to as clearly as for those of the

 2   primary database.

 3                   So we agree with you totally that it should

 4   be a labeling issue.

 5                   CHAIRMAN RELLER:        Yes, Dr. Christie had a

 6   question.      Go ahead.

 7                   DR. CHRISTIE-SAMUELS:         I think it's more

 8   to that, when you have some details regarding antibiotic

 9   trials in children to look at.

10                   My question, however, I noticed that in one

11   of your slides you said about 12 to 80 percent of your

12   -- trials have -- cough suppressants but really I

13   wondered, bearing in mind that children -- from taking

14   over-the-counter drugs, cough medicines while being

15   treated for community acquired pneumonias and other

16   respiratory tract infections.

17                   I was wondering about the immune inhibitory

18   effect, and if this was of value to the prelim trials.

19    If so, what did you find?          What were your preliminary

20   findings?

21                   DR. TARPLEY:      So your question is the mono

22   amine oxidase inhibitory effects as they pertain to the

23   pediatric trials?

                                S A G CORP.
     202/797-2525               Washington, D.C.        Fax: 202/797-2525

 1                    Thank you.

 2                    DR. SLATTER:    Thank you for the question.

 3                    Could I have slide K-21, please?

 4                    In fact, many patients in the pediatric

 5   trials were on medications with MAOI potential effects

 6   and interactions.       The answer to your question -- I'm

 7   sorry.     Can you hear me?

 8                    The answer to your question, however, is

 9   that you've seen the safety profile.            There were no

10   adverse events that would reflect significant clinical

11   consequences       of   MAOI    inhibition.       Clearly     the

12   pediatricians enrolling were alerted, and of course,

13   discussion in the protocol indicated this potential

14   effect.        There were no examples of hypertension, no

15   examples of hyperthermia, no examples that would suggest

16   an adverse event related to an MAOI effect in the entire

17   study population.

18                    CHAIRMAN RELLER:      Dr. Leggett.

19                    DR. LEGGETT:     Back to the question about

20   the dialysis of --

21                    DR. TARPLEY:    Dr. Jungbluth.

22                    DR. LEGGETT:    --

23                    DR. TARPLEY:     I'm sorry.   I think we got

                               S A G CORP.
     202/797-2525              Washington, D.C.        Fax: 202/797-2525

 1   the first question, but we were unable to hear.              Maybe

 2   we could deal with the first question and you could

 3   repeat your second.

 4                  The first question, I believe, related to

 5   the method of dialysis used in the studies.

 6                  Thank you.

 7                  DR. JUNGBLUTH:        And I think there was an

 8   additional     part   of   your     question   on   whether     the

 9   elimination of the metabolites was similar.              From the

10   single dose data that we have in the renal impairment

11   study, the metabolites appear to be reduced to the same

12   extent as linezolid, and that's about 30 percent of the

13   dose.

14                  And we did not have specific information

15   on what dialysis membranes were used in that study.

16                  DR. LEGGETT:       And the second question was

17   did you see any more severe hepatic impairment patients

18   in the case of study?

19                  DR. TARPLEY:       Again, let me repeat it just

20   to make sure I'm answering your correct question.               Did

21   we treat any severe hepatic impairment patients in the

22   compassionate use program?

23                  DR. LEGGETT:       Or the entire program?

                                S A G CORP.
     202/797-2525               Washington, D.C.         Fax: 202/797-2525

 1                  DR. TARPLEY:      Or throughout the entire

 2   program.

 3                  DR. HAFKIN:    Unfortunately we did not have

 4   any of those patients recruited into any of our Phase

 5   III trials.    They weren't specifically excluded.

 6                  However, I should say since I think the

 7   audiovisual is off, I'll try to yell.

 8                  (Laughter.)

 9                  DR. HAFKIN:    The compassionate use, we had

10   some very, very sick patients who got linezolid for short

11   periods of time.     Those people with dreadful hepatic

12   function were typically patients who had very profound

13   infections in their interabdominal cavity, and they were

14   really going down fast.       They didn't even survive five

15   days.

16                  We have no reason to believe they died

17   because of the drug.         They had fulminant infection.

18   So there were one or two observations with people who

19   had just interabdominal catastrophes who had no renal

20   -- no hepatic function and no renal function, and they

21   received a couple of days.       We had one child, in fact,

22   like that.

23                  But there was no pattern of adverse event

                            S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   that would have suggested a signal.

 2                    CHAIRMAN RELLER:        Dr. Wittes.

 3                    DR. WITTES:     I actually have a series of --

 4   can you hear this?

 5                    DR. TARPLEY:     Yes.

 6                    DR. WITTES:     A series of questions related

 7   to design and statistics.          First, it wasn't only just

 8   factual.       The first is I don't understand.             The total

 9   number of patients who are listed in the slide for almost

10   every study was considerably larger than the number even

11   in the IIT, and I assume that means that there was a

12   group of people who didn't get any drug.            Is that right?

13                    And what was the mechanism by which somebody

14   gets randomized and not in the IIT?

15                    DR.   TARPLEY:       Dr.     Oliphant      from     our

16   Biostatistics Group will answer the question.

17                    Thank you.

18                    DR. OLIPHANT:      Dr. Wittes, Tom Oliphant,

19   PNU biostatistics.

20                    I     believe    your        question      is      what

21   differentiates the patients randomized from those who

22   were included in the ITT analysis populations.

23                    Slide on, please.

                                S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1                   Here we see for the Phase III studies for

 2   both linezolid and comparator the number of patients

 3   randomized, and those included in the ITT populations,

 4   and I believe the numbers, total numbers, range from

 5   about zero as you see in Study 54(a), all patients

 6   randomized were in the intent to treat population.                   For

 7   a couple of the studies it was as high as ten or 12

 8   patients who were randomized but were not included in

 9   ITT.

10                   The   ITT   population          is   basically    those

11   patients who were randomized and did receive at least

12   one dose of study medication.

13                   DR. WITTES:       Okay.        That's very helpful.

14    My calculation showed bigger differences.                  So that's

15   what I needed to see.

16                   Can I --

17                   CHAIRMAN RELLER:         Please continue.

18                   DR. WITTES:      Okay.         The next one has to do

19   -- the next question really has to do with historical

20   controls.      Are there any data that would give a sense

21   of sort of an anchor of what you would expect in an

22   untreated population for cure rates?

23                   And I know it would vary from indication

                                 S A G CORP.
     202/797-2525                Washington, D.C.             Fax: 202/797-2525

 1   to indication.

 2                  DR. HAFKIN:       I suspect you're interested

 3   in the VRE historical perspective.

 4                  DR. WITTES:       Well, no, actually I'm more

 5   interested in the others.

 6                  DR. HAFKIN:       Oh, well --

 7                  (Laughter.)

 8                  DR. HAFKIN:       -- in that case, I believe that

 9   the   performance    of    the    comparator        agents    in    our

10   comparative trials, our control trials are very similar

11   to the results that other companies have used in their

12   comparator trials.        So that if you look at --

13                  DR. WITTES:        No, I'm asking actually a

14   different question.        In untreated population, that's

15   the question I'm asking.

16                  DR. HAFKIN:       Okay.        I'm sorry.   Untreated

17   controls.

18                  DR. WITTES:       Yeah.

19                  DR.   HAFKIN:        If    you     go   back   to    the

20   pre-antibiotic era for diseases like skin, soft tissue,

21   and pneumonia, outcomes are really pretty good.                 It was

22   rare for a calamity to occur after a skin and soft tissue

23   infection, but it did occur.

                               S A G CORP.
     202/797-2525               Washington, D.C.             Fax: 202/797-2525

 1                  The difference between the pre-antibiotic

 2   era and the antibiotic era is the time at which patients

 3   are feeling better, and that's something that the

 4   displays that we share with you are not sensitive to.

 5                  You get better from very severe skin and

 6   soft tissue infection with enough time in the great

 7   majority of cases.        Osteomyelitis, life threatening

 8   sepsis did occur with regularity.            You know, we're in

 9   Washington.    I believe President Wilson's son died of

10   Staphylococcal bacteremia after stepping on a branch

11   in the White House lawn, but that was a rare event.

12                  And so that the great majority of patients

13   with skin and soft tissue infection would with time and

14   care    resolve.    The    likelihood        of   a   catastrophic

15   complication was very real.            It was relatively low.

16   Actually there are people here that are much more learned

17   in this area of the history of medicine than I, and if

18   we might, we can ask one of the real world's experts

19   to come up here and talk about it.

20                  But pneumonia II would with time resolve

21   in the great majority of healthy hosts.               It did kill

22   with great regularity elderly patient populations, and

23   if you look at our out-patient study, we have fairly

                               S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1   young people there, and so that the great majority of

 2   the patients randomized in 51, Protocol 51, would have

 3   been expected to get better with a long period of time.

 4                  However, if you look at 33 or 48, those

 5   patients would have very high mortality rates.

 6                  CHAIRMAN RELLER:      Dr. Chesney.

 7                  DR. CHESNEY:     My question has to do with

 8   the community acquired pneumonia and the penicillin

 9   intermediate and resistant strains, and on the materials

10   we have before we came, on page 45 if I added up right,

11   you have a total of 12 patients who did well with

12   linezolid -- excuse me -- 16 patients, 12 of 16 did well.

13    So four did not, and I was curious to know if this is

14   your      total   information        that   is      penicillin

15   nonsusceptible information or if you have additional

16   to what we have here.

17                  DR. HAFKIN:    Yes.     Let me show you Slide

18   189.

19                  This is an aggregate of all the data that

20   we have, looking at linezolid performance in Phase II

21   and linezolid performance in Phase III, and if you will

22   note here, we have really excellent results with

23   penicillin resistant Strep. pneumo. very comparable to

                            S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1   the performance of linezolid in the treatment of typical

 2   Strep.     pneumo.        There    really       is     essentially        no

 3   difference.

 4                  This is the result for Staph.                  aureus and

 5   MRSA.    Recall that the patient populations in the Staph.

 6   aureus group are very different than the patient

 7   populations in the Strep. pneumo. group.

 8                  Slide off.

 9                  So    we    feel    that        we've    got    excellent

10   activity, that the limited but real life experience of

11   treatment of resistant pathogens is pretty solid.

12                  DR. CHESNEY:        Can I ask one question?

13                  DR.    HAFKIN:         One       other      point    I   had

14   forgotten to mention.        There were five PRSP in pediatric

15   age group patients, and they all were cured.

16                  DR.    CHESNEY:            Why        did    you     choose

17   cefpodoxime as your comparator?

18                  DR. HAFKIN:        Well, it's a wonderful drug.

19                  (Laughter.)

20                  DR.    HAFKIN:           We      think       it's     under

21   appreciated, and we think that we've shown in many

22   studies that it's just a lovely drug.                      It's just not

23   loved enough.

                                S A G CORP.
     202/797-2525                Washington, D.C.                Fax: 202/797-2525

 1                   So in all honesty, it was available easily

 2   and quickly for us.      We feel that it is equivalent to

 3   all the second generation cephalosporins available.

 4                   CHAIRMAN RELLER:      Please continue the line

 5   of questions and we'll come to the others here, Dr.

 6   Chesney.

 7                   DR. CHESNEY:     This is my last.       I don't

 8   mean to -- Group A strep., are these the total numbers

 9   that you have for Group A strep. infections on page 60?

10    It shows us 23 of 29 for palpitated skin and soft tissue

11   was successful.

12                   DR. HAFKIN:    Yes.

13                   DR. CHESNEY:    Which is 79 percent.

14                   DR. HAFKIN:    That's correct.    That is our

15   entire experience.

16                   DR. CHESNEY:    Thank you.

17                   CHAIRMAN RELLER:      Dr. Rodvold.

18                   DR. RODVOLD:    I had a couple of questions.

19    Let     me    start   with    the    efficacy   question      on

20   Streptococcus pneumoniae.        Can you tell us more about

21   these patients in regards to the severity and/or their

22   pathogethic oral component areas, in particular, the

23   penicillin resistant bacteria?

                              S A G CORP.
     202/797-2525             Washington, D.C.         Fax: 202/797-2525

 1                   DR. HAFKIN:     The question, I think, is of

 2   these patients that we treated with Strep. pneumo.

 3   infection, we'll get the right slide.                I want to get

 4   that original slide that I showed.                 I think it was

 5   EP-138, was it?      The original one where we have Phase

 6   II and Phase III together.

 7                   We have 32 patients in the linezolid treated

 8   group that had pneumococcal bacteremia in all of our

 9   trials.        Protocol   51,   which        was   the   out-patient

10   pneumonia trial, had very few patients.                  Most of them

11   came from Protocol 33.

12                   Yes, if you could put this slide up, 189.

13                   So when we look at this, the population we

14   have here with more than 150 Strep. pneumo. infections,

15   only 32 of them on the linezolid arm were bacteremic.

16    Every one of the patients treated with linezolid had

17   resolution of the bacteremia.            Two recurred.          Let me

18   tell you about those patients.

19                   Patient number one was a patient with COL,

20   was on active immunosuppression, and I don't know why

21   that he stopped his therapy on day six.            He looked great.

22    He came back in two weeks with recurrent Strep. pneumo.

23   infection.      Unfortunately the second isolate never got

                               S A G CORP.
     202/797-2525              Washington, D.C.               Fax: 202/797-2525

 1   to our central lab.         So we were never able to understand

 2   whether        it     was   recurrent       infection,      recurrent

 3   bacteremia or whether it was a new infection.

 4                       The second story is the same story.          It was

 5   an out-patient pneumonia trial.                 Protocol 51, a patient

 6   with AIDS, with very low CD-4 counts.                    He took the

 7   medicine for five days and he died.                 We don't know why

 8   he died.

 9                       So we have in this population 32 blood

10   culture proven Strep. pneumo. infections.                  Both cases

11   were associated -- both failures associated with short

12   term therapy.

13                       The comparator actually is associated with

14   slightly more failures.           Frankly, I haven't looked at

15   them at the same level.          I mean I'm one of those people

16   that believes that you learn a lot from studying

17   failures.       The failures for the cephalosporin groups

18   are actually slightly greater in number, number one;

19   same number of bacteremias, about 30.                  They tended to

20   be more complicated.           The patients that failed with

21   cephalosporins tended to be more associated with either

22   short-term therapy as well, but also very resistant Gram

23   negatives.          So that these patients at baseline would

                                  S A G CORP.
     202/797-2525                 Washington, D.C.            Fax: 202/797-2525

 1   have Strep. pneumo. in the blood and then they would

 2   die with enterobacter or pseudomonas.

 3                  So there was one patient with recurrent

 4   bacteremia with Strep. pneumo., but so we don't have

 5   a clear picture of this.       At least I don't have a clear

 6   picture of the cephalosporin failures.

 7                  DR. RODVOLD:     But are the 12 isolates that

 8   are penicillin resistant for those patients?

 9                  DR.   HAFKIN:        We     had     no    recurrent

10   bacteremia.

11                  DR. RODVOLD:     And how many were bacteremia

12   and how many of those were considered severe infections?

13                  DR. HAFKIN:     Well, yes.    Five of them were

14   kids.     The additional population were all in elderly

15   or very sick people.          None of them came from our

16   out-patient trial at all in adults.         They all came from

17   33.    These were all people who had to be in the hospital

18   because their infection was severe.

19                  I honestly don't think we did a cut of the

20   analysis to see how many were bacteremic.                 I would

21   assume that we're talking about two or three.

22                  CHAIRMAN RELLER:      Please.

23                  DR.   MURRAY:        I'd     like    to    make     a

                             S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   clarification because I'm a little confused.                 The

 2   sponsor has not presented us with data in a form or in

 3   a written indication that they're looking for an

 4   indication for resistant organisms.

 5                   On the other hand, FDA has given us that

 6   in their question.      Do we think they're efficacious in

 7   each of these settings?

 8                   So I'm a little caught.     It does not appear

 9   that the sponsor is asking for specific labeling for

10   resistant pneumococcus or MRSA, and yet FDA is asking

11   us to evaluate that.     So I'd like clarification.       We're

12   not asking for that specific labeling.

13                   DR. RODVOLD:    Actually that's part of the

14   reason I'm asking the question, is that in other

15   presentations this committee has seen, some of the --

16   that was presented more clear, particularly the severe

17   pain in bacteremic patients with insulin resistant

18   isolates.      So that's easier to see, and maybe you could

19   pull that data together yet today and let us see that

20   in regards to helping us make judgment in the labeling.

21                   CHAIRMAN RELLER:       So, Dr. Chikami, could

22   you --

23                   DR. CHIKAMI:    Let me just clarify what Dr.

                              S A G CORP.
     202/797-2525             Washington, D.C.        Fax: 202/797-2525

 1   Hafkin     said.     What     he    showed     on    his     slide     for

 2   indications were the general sort of infection site

 3   indications.       In fact, within the labeling the company

 4   is requesting specific wording for penicillin resistant

 5   Strep. pneumo., that is, infection due to Strep. pneumo.

 6   including penicillin resistant strains and infections

 7   due to Staph. aureus, including MRSA.

 8                  DR. TARPLEY:         If I could just refer you also

 9   to page 6 of the brochure where the indications are

10   listed and the pathogens associated with each of those

11   indications are spelled out much more completely.

12                  DR. CHIKAMI:         Right.

13                  DR. SORETH:         It's on page 6 of your briefing

14   document from the sponsor.

15                  CHAIRMAN RELLER:            To summarize the thrust

16   of those comments, I think the committee must elicit

17   and ask all of the questions that the individuals would

18   need to be able to address specifically the questions

19   that we will vote on this afternoon.

20                  And    clearly,       the    issues    regarding        the

21   resistant components of pathogens within the different

22   indications is part of the task in which we will be asked

23   to render advice to the FDA.

                                 S A G CORP.
     202/797-2525                Washington, D.C.               Fax: 202/797-2525

 1                  Dr. Wittes.

 2                  DR. WITTES:        Yeah.        I had three more

 3   questions in my series.        Can I ask them?

 4                  CHAIRMAN RELLER:        Please proceed with your

 5   questions, and bundling them is helpful to keep the

 6   continuity of thought going.

 7                  DR. WITTES:       Yes.        That's what I'll do.

 8   I'll put them all together and you'll see.

 9                  CHAIRMAN    RELLER:           They   can   be   asked

10   sequentially, but the related questions, we'll stick

11   with the individual committee members so that we can

12   round out the issue.

13                  DR. WITTES:      Okay.

14                  CHAIRMAN RELLER:         Please, proceed.

15                  DR. WITTES:      Well, then let me just tell

16   you the series of questions and you can put them together

17   the way you want.

18                  The first has to do with definition of

19   equivalence, that in many of the trials that you showed,

20   it was pretty clear.      You look at the lower end of the

21   confidence limit.     It's pretty clear that, you know,

22   you wouldn't have -- whatever pre-defined definition

23   you had, it would have satisfied it.

                               S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1                   But in 48(a), it seemed to me that there

 2   was a very low lower bound, and I wonder whether -- how

 3   you, in fact -- whether you pre-define equivalence.

 4   How did it differ from indication to indication, and

 5   so forth?      So that's question number one.

 6                   Question number two and three has to do with

 7   54 and 54(a), and I actually was confused by the

 8   presentation today because it seemed different from what

 9   the briefing book said.           My understanding from the

10   briefing book was that there was a study whose name was

11   either 54 or 54(a).          You looked at the data early,

12   reported that as 54(a), and then continued with the study

13   calling it 54, or it may have the labels wrong, and what

14   we saw in the briefing book was, I think, only part of

15   the rest of this 54, and there was going to be more.

16                   That   was   my   understanding.        What    my

17   understanding today was that there was a preplanned

18   study, 54(a), quite distinct from 54, and that what we

19   see here is an interim analysis from 54, but I didn't

20   see any discussion of what that interim means and what

21   the alpha had and all of that sort of stuff.

22                   So I need to understand what the study

23   design is for the 54, 54(a) complex.

                              S A G CORP.
     202/797-2525             Washington, D.C.          Fax: 202/797-2525

 1                  And the final question, which is, again,

 2   an overall question related to the studies, has to do

 3   with blinding.     How much -- some of the studies are

 4   partially blind.     Some are unblind; some of them are

 5   not, and some are blind, but the ones that are not where

 6   there's clinical outcomes, how much of the clinical

 7   outcome is subjective enough to be affected by knowledge

 8   of treatment?

 9                  DR. TARPLEY:    Okay.       Thank you.

10                  Dr. Oliphant will start our responses on

11   the definitions of equivalency used, particularly in

12   Study 48(a), and I presume he can also address the

13   blinding issue.

14                  Then we'll come back and try and clarify

15   the Study 54.    Is that acceptable?

16                  DR. WITTES:    Good, fine.

17                  DR. OLIPHANT:     Dr. Wittes, your question

18   regarding our definition for equivalence, it was -- yes,

19   it was study specific.

20                  If I can have the slide on, please.

21                  Basically as you indicated for most of the

22   studies we had no problem meeting the requirement of

23   the lower limit of the confidence interval exceeding

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   minus ten percent.        That was based on an assumption of

 2   90 percent clinical cure rates in those studies.

 3                    The one exception was Study 48(a) where the

 4   assumption there was an expected clinical cure rate of

 5   70 percent.      So using sort of the guidelines in the FDA

 6   points to consider, their step function approach for

 7   what one should use for a delta based on expected cure

 8   rates, an expected cure rate of 70 percent translated

 9   to an equivalence margin of 20 percent.              So for that

10   study    the lower confidence limit needed to exceed minus

11   20 percent for a declaration of equivalence.

12                    Our next.      Well, I'll address the next

13   issue.     You had a question about blinding and whether

14   our outcomes were subjective enough to handle the fact

15   that the blinding did differ from study to study.                   I

16   guess I can best answer that by stating that all of the

17   clinical outcome results that you've seen presented

18   today are from a sponsor's clinical outcome, which was

19   a    generally      conservative        modification      of     the

20   investigator's clinical outcome.

21                    I can go through the various modifications

22   if you'd like, but basically that was what we used for

23   clinical       outcome,   was    the     sponsor's   assessment,

                                S A G CORP.
     202/797-2525               Washington, D.C.          Fax: 202/797-2525

 1   predetermined,      done    prior     to       breaking      any    blind;

 2   essentially     involved         sometimes            downgrading        an

 3   investigator's      assessment        of       cure    to    failure     or

 4   indeterminates, for instance, if not enough medication

 5   was received.

 6                  So that, I believe was our attempt to

 7   address any differences in blinding across studies.

 8                  Your third question regarding Study 54 and

 9   54(a), I believe I'll let Dr. Hafkin begin to address

10   that and may chime in if necessary.

11                  DR. HAFKIN:       The history of Protocol 54(a),

12   the    study   that    I    called     completed,           was    one   of

13   excruciating investment in time and effort.                   We had gone

14   to more than 100 sites and had had the study up for more

15   than a year, and our recruitment into the trial findings,

16   solid clinical observations for VRE infections, were

17   going badly.

18                  We     had    been     told       by     a    couple      of

19   investigators that they simply didn't feel comfortable

20   with the dose comparison design.                We made the decision

21   after more than a year in the field with this protocol

22   that it was based on our need to know.                      We needed to

23   know whether the design was working, whether the

                                 S A G CORP.
     202/797-2525                Washington, D.C.                Fax: 202/797-2525

 1   outcomes were going to be hopeful.         We needed to know

 2   what was happening with this protocol in terms of patient

 3   outcomes.

 4                  We talked about this not only inside the

 5   sponsor's organization, but shared our intention to stop

 6   the study with the agency.

 7                  We talked about it more, and we finally

 8   decided that the study had been ongoing for long enough.

 9    We needed to know and fully understand the nuances of

10   the study and its performance.

11                  So we closed the study, and we felt at that

12   time, well, that's it.      We've closed it.    We're going

13   to look at it.

14                  Then we got encouraging words from some

15   investigators that they were really wanting to continue

16   the study, the same general design.        The protocol was

17   modified somewhat, but it was continued in many of the

18   same sites, but we consider the, if you will, this 54

19   study to be a separate protocol.

20                  The protocol modifications included some

21   improvements that our investigators felt might suddenly

22   improve their ability to recruit patients, but we see

23   them as separate protocols.

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1                  DR. WITTES:     So what's the interim analysis

 2   in 54?

 3                  DR. OLIPHANT:         Dr. Wittes, the results

 4   you've seen presented for Study 54, the 82 patients out

 5   of the 186, those results were presented at the request

 6   of the agency to see results that we had available at

 7   the time.

 8                  We presented those results, but did not do

 9   any statistical testing or any calculation of confidence

10   intervals or anything of that nature.            So you did not

11   see any of those results in our presentation of results

12   for those 82 patients.

13                  CHAIRMAN RELLER:        Yes, Dr. Chesney.

14                  DR.    CHESNEY:          Just   two    points      of

15   clarification.       The 12 patients that had the resistant

16   pneumococci, I probably misheard.            Those don't include

17   the five children who had resistant pneumococci.                They

18   do?

19                  DR. TARPLEY:      They do.

20                  DR. CHESNEY:       And the children had what

21   infection?     They had the community acquired pneumonia

22   also?

23                  DR. TARPLEY:      Dr. Anderson can respond best

                              S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1   to that question

 2                  DR. ANDERSON:      Right.       Those included the

 3   studies reported in Protocol 45 and 49.                  There were

 4   three     bacteremic    Strep.     pneumoniae       isolate        from

 5   Protocol 45, all these from cures.               There were five

 6   overall Strep. pneumonia microbiologically evaluable

 7   in Protocol 45.        They were all cures, but these were

 8   always based on blood culture because there is no other

 9   way we can get microbiologic evaluability.

10                  Two cures came from Protocol 49.                    That

11   included two PRSP.

12                  There    were     other       isolates     that     were

13   intermediately susceptible, but in fact, those that I've

14   indicated as PRSP were clearly resistant.

15                  DR. CHESNEY:      Just one more clarification.

16    All of the complicated skin and soft tissue infections

17   in which methicillin resistant Staph. were a problem

18   are in our handbooks, which is, I think, two, three,

19   two successes; is that correct, on page 60?

20                  DR. TARPLEY:       Could you repeat the page,

21   please?

22                  DR. CHESNEY:      Page 60, the top of page --

23                  DR. TARPLEY:      Six, zero?

                               S A G CORP.
     202/797-2525              Washington, D.C.              Fax: 202/797-2525

 1                   DR. CHESNEY:       Six, zero, yeah, of what we

 2   got before.

 3                   DR. TARPLEY:       Thank you.

 4                   DR. CHESNEY:       Right.       I think it's three

 5   patients with methicillin resistant strains.                 Two were

 6   successful.

 7                   DR. HAFKIN:        Indeed, that is the results

 8   of our Protocol 55, which is the straightforward,

 9   traditional, complicated skin and soft tissue trial.

10    The reason we performed Protocol 31, the MRSA trial,

11   was just so we could augment that number.

12                   Speaking to MRSA, we had 15 patients with

13   bacteremia due to MRSA, and we had four recurrent MRSA

14   bacteremias.        Each    one     of    those      recurrent     MRSA

15   bacteremias were associated with short term therapy,

16   except for one patient that had osteomyelitis and

17   received       20   days    of     therapy.           Diagnosis      of

18   osteomyelitis       was    made,    and       they   had   recurrent

19   Staphylococcal infection.

20                   So we have substantial experience with MRSA

21   because of this additional clinical study.

22                   CHAIRMAN RELLER:         Dr. Leggett.

23                   DR. LEGGETT:       On the Slide 53, you showed

                                S A G CORP.
     202/797-2525               Washington, D.C.             Fax: 202/797-2525

 1   in Protocol 31 that most of your MRSAs had skin and soft

 2   tissue infections, which seems rather less challenging

 3   than the 50 cases of bacteremia.

 4                     Could you describe the causes of those 50

 5   cases of bacteremia due to MRSA and specifically how

 6   many cases of right side endocarditis, which I believe

 7   you mentioned in several different cases.

 8                     DR. HAFKIN:          Yes, there was only one

 9   patient        with   right     side     endocarditis     recruited,

10   Protocol 31, that they were recruited to vancomycin.

11    Indeed, they failed.

12                     DR. LEGGETT:         Do you recall bacteremia?

13   You list it?

14                     DR. HAFKIN:      I already shared with you the

15   15 cases of microbiological evaluable MRSA infection

16   with bacteremia.        The others that had bacteremia may

17   have had bacteremia, but they were not evaluable because

18   they didn't take medicine or they were randomized to

19   vancomycin.

20                     CHAIRMAN RELLER:         We are a little over the

21   time for our break.       We'll have ample time to come back

22   to these questions.

23                     Before breaking, however, for the benefit

                                   S A G CORP.
     202/797-2525                  Washington, D.C.          Fax: 202/797-2525

 1   of the sponsor, some of the things that I think we will

 2   be wanting to address subsequently, to give you time

 3   to pull together either out of a composite of what's

 4   there or if it's already packaged are some of the

 5   following.

 6                    The   questions      will      include     for    adult

 7   indications, but yet we've heard that some of the data

 8   presented on resistant organisms include pediatric

 9   data.     So I think we need to see what information we

10   have    on     those   organisms    from       adults   alone.        The

11   pediatric data are supportive and interesting, but we

12   have to deal with what we have today.

13                    So specifically, we know that there's a high

14   concordance of macrolide resistance and penicillin

15   resistance among Strep. pneumoniae, despite totally

16   different mechanisms of resistance.                   It would be of

17   interest if you have those data available what the

18   profile on the resistant organisms included in your

19   patients were for macrolides, as well as penicillin,

20   and what the response rates were.

21                    What proportion of patients with pneumonia

22   owning to Streptococcus pneumoniae, ideally by category

23   of   resistance,       in   fact,    did       have   positive     blood

                                 S A G CORP.
     202/797-2525                Washington, D.C.              Fax: 202/797-2525

 1   cultures?

 2                  And do we have any resistant organisms with

 3   bacteremia that responded to the compound?

 4                  And I think the same issues apply for

 5   infections with methicillin resistant Staphylococci,

 6   and ideally showing, including those with bacteremia,

 7   those with bacteremia with methicillin susceptible and

 8   resistance, and if there are any response differences,

 9   to see what those numbers end up being.

10                  So it's delving into more specific detail

11   having to do with the critical issues of resistant

12   organisms, confirmed with positive blood cultures and

13   without positive blood cultures in adults, and then what

14   supplemental information there may be from pediatrics,

15   but so that we see what there is currently with the adult

16   indications.

17                  At this time we'll take a 15 minute break

18   and begin promptly at five minutes of 11 to hear the

19   FDA's presentation.

20                  (Whereupon, the foregoing matter went off

21                  the record at 10:40 a.m. and went back on

22                  the record at 10:58 a.m.)

23                  CHAIRMAN RELLER:      We'd like to begin the

                             S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   session this morning, complete the session this morning.

 2                  For those who are recording this session,

 3   if any of the questions are not clear or the mechanics

 4   of things are not working, if you would raise your hand,

 5   I'll take a special effort to get the question or answer

 6   both repeated to capture the proceedings.

 7                  We now have Dr. David Ross, who will step

 8   to the podium and present the FDA's considerations on

 9   linezolid.

10                  David.

11                  DR. ROSS:      Thank you, Dr. Reller.

12                  I'm going to ask the obligatory first

13   question, which is:       can everybody hear me?     And I guess

14   the microphone is on.        So the answer is yes.

15                  I'm a medical reviewer in the Division of

16   Anti-infective Drug Products, and I'll be presenting

17   the agency's analysis of a new drug application for

18   linezolid.

19                  Could I have the next slide, please?

20                  What I'm going to do is briefly review some

21   issues     related   to    the     clinical     pharmacology     of

22   linezolid,     discuss       the     agency's     clinical      and

23   statistical analyses of efficacy data in this NDA, our

                               S A G CORP.
     202/797-2525               Washington, D.C.         Fax: 202/797-2525

 1   analyses of safety data in this NDA, and then discuss

 2   development of resistance to linezolid.

 3                  Could I have the next slide, please?

 4                  Just to review, after IV administration of

 5   linezolid, peak plasma concentrations are reached in

 6   about half an hour, in about an hour after oral

 7   administration.        Maximum     concentration   after     IV

 8   administration is 15 micrograms per mL, 21 after oral

 9   administration, and these refer to 600 milligram doses

10   given BID.

11                  Trough concentrations are     3.7 micrograms

12   per mL for IV dosing, 6.2 for oral, and the half-life

13   is about five hours for both IV and PO.

14                  I'd just like to remind you that the key

15   pharmacodynamic parameter for linezolid is time above

16   MIC in that as you've heard in the mouse thigh model,

17   the time above MIC is the most important parameter with

18   concentrations needing to be above the MIC for about

19   40 percent of dosing interval for efficacy against

20   Streptococcus pneumoniae.

21                  Okay.   The next slide.

22                  With 600 milligram oral BID dosing, there's

23   considerable variation in the exposure to linezolid with

                             S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   AUCs ranging from 68 to 209.           This variability remains

 2   after you normalize for body weight, with the exposure

 3   after normalization ranging from 11.2 to about 24.

 4                    Next slide.

 5                    As you've heard, linezolid has two major

 6   metabolites.       The toxicity of these metabolites has not

 7   been studied separately from the parent compound in

 8   animal or human studies.          The drug is excreted in urine

 9   and feces.      About 35 percent in the urine is parent drug;

10   50 percent in urine is metabolites; and ten percent in

11   the feces is metabolites.

12                    And, again, as you've heard and seen in the

13   briefing       package,   these    metabolites   accumulate     in

14   patients with renal impairment with the degree of

15   accumulation increasing with patients with more severe

16   renal impairment.

17                    Next slide.

18                    Let me move on to analyses of efficacy data

19   by the agency.

20                    Next slide.

21                    The clinical studies that I'm going to

22   discuss are those for community acquired pneumonia.

23   One    study      of   community      acquired   pneumonia      in

                                S A G CORP.
     202/797-2525               Washington, D.C.        Fax: 202/797-2525

 1   in-patients, one in out-patients.              Hospital acquired

 2   pneumonia for which a single study was submitted in the

 3   NDA.    Skin and skin structure infections, and here there

 4   were two studies for uncomplicated skin and skin

 5   structure      and   one   for    complicated    skin    and     skin

 6   structure.

 7                   A supporting study of methicillin resistant

 8   Staphylococcal species infections, and a study of VRE

 9   infection, and these are Studies 54(a) and supportive

10   data from Study 54.

11                   Next slide.

12                   What I'd like to do before discussing the

13   individual studies is highlight some differences in the

14   FDA's     method     for   assessing       outcomes   versus      the

15   sponsor's.

16                   For patients who did not have a post

17   baseline efficacy assessment, the sponsor considered

18   such patients to be failures, whereas the FDA considered

19   such patients to be missing unless they met certain

20   prespecified conditions for failure, such as receiving

21   another antibiotic for lack of efficacy.

22                   Deaths     were   considered    by    the   FDA    to

23   represent failure regardless of the cause in the ITT

                                S A G CORP.
     202/797-2525                Washington, D.C.          Fax: 202/797-2525

 1   analysis.       The sponsor did not directly consider death

 2   in assessing outcome.              Such patients were generally

 3   considered missing, again, unless they met certain

 4   prespecified criteria for being considered failures.

 5                    Finally, with respect to patients who were

 6   discontinued from study for lack of efficacy, these

 7   patients       were    generally,       although         not    invariably,

 8   considered to be failures in the sponsor's analysis.

 9    Such patients were considered by definition failures

10   in the FDA's analysis.

11                    Next slide.

12                    With respect to the analytic populations

13   studied in the FDA analysis, the all randomized patient

14   population was used to define the ITT all treated patient

15   population.       A modified intent to treat population was

16   identified, which consisted of all ITT patients who had

17   a pathogen isolated.

18                    The ITT patient population was also used

19   to   define      valuable       protocol          populations.           These

20   included clinically evaluable patients who met baseline

21   and post baseline criteria.

22                    A     microbiologically               evaluable      patient

23   population       was    also     defined          as    those     clinically

                                 S A G CORP.
     202/797-2525                   Washington, D.C.                 Fax: 202/797-2525

 1   evaluable patients who had a susceptible pathogen

 2   isolated within the baseline visit window.              Usually

 3   this represented 48 hours within -- patients who had

 4   pathogen isolated within 48 hours of study entry.

 5                   For specific studies that I will discuss

 6   later, particularly complicated skin and skin structure

 7   and VRE, the FDA analysis included examination of

 8   specific ITT patient populations that were predicated

 9   on important baseline characteristics.

10                   Next slide.

11                   Let me move to a discussion of specific

12   studies.       For community acquired pneumonia, as I've

13   said, the sponsor conducted two studies.

14                   Next slide.

15                   The first one that I'm going to discuss is

16   Study    33.     This   study   enrolled    and   treated     747

17   in-patients with community acquired pneumonia.               This

18   was a multi-center, multi-national, randomized, open

19   label trial.      The trial was initiated as an evaluated

20   blind study, and then was changed to open label during

21   the course of the study.

22                   Patients were randomized to linezolid or

23   to ceftriaxone given for seven to 14 days.               At the

                              S A G CORP.
     202/797-2525             Washington, D.C.         Fax: 202/797-2525

 1   discretion of the investigator could be switched to oral

 2   therapy, oral linezolid in the case of the linezolid

 3   arm, cephpodoxime in the case of the ceftriaxone arm.

 4                  Concomitant as aztreonam was allowed for

 5   Gram negative infections.

 6                  The primary endpoint in this study was

 7   microbiologic outcome.

 8                  Next slide.

 9                  This      slide    shows       a   summary    of    the

10   demographics for patients enrolled in the study.                    As

11   you can see, the arms were balanced with respect to age,

12   gender, and race.

13                  Next slide, please.

14                  Seven hundred and forty-seven patients were

15   enrolled and treated.            Of these, 254 had a pathogen

16   isolated.        There    were     559    clinically        evaluable

17   patients.      One hundred and ninety-one of these had a

18   susceptible pathogen isolated at baseline.

19                  Next slide.

20                  Response rates in the FDA analysis are shown

21   here    for    the    various     populations:         ITT,       MITT,

22   clinically evaluable, and microbiologically evaluable.

23                  Sizes of the populations are shown here.

                                S A G CORP.
     202/797-2525               Washington, D.C.            Fax: 202/797-2525

 1    These numbers exclude patients with missing outcomes,

 2   that is, those patients for whom there was no follow-up

 3   efficacy data.

 4                   As I indicated, patients who died before

 5   the test of cure assessment were considered failures

 6   in the ITT and MITT analyses.               Such patients were

 7   excluded       from   the     clinically       evaluable       and

 8   microbiologically evaluable analyses unless they were

 9   assessed as having died of their initial infection.

10                   As you can see, although the response rates

11   vary, the analyses are similar across the various

12   populations analyzed.

13                   Next slide.

14                   This slide shows the confidence interval

15   around the difference in response rates in the FDA

16   analysis and the corresponding confidence interval for

17   the sponsor's analysis.

18                   The dashed line indicates a difference of

19   zero so that values to the left favor comparator.

20   Values to the right favor linezolid.           The hashed marks

21   indicate the point estimate of the difference in

22   response rates.

23                   And as you can see, the confidence intervals

                              S A G CORP.
     202/797-2525             Washington, D.C.          Fax: 202/797-2525

 1   for the FDA and the sponsor and similar for the various

 2   analytic populations.

 3                  Next slide.

 4                  With respect to results by pathogen, and

 5   here we are discussing the microbiologically evaluable

 6   patient population, let me just focus on one line here.

 7    For patients with pneumococcal bacteremia, with 30

 8   patients in the ME patient population, the linezolid

 9   arm, 24 in the ceftriaxone arm, with response rates of

10   90 percent and 63 percent.

11                  Next slide.

12                  This analysis shows clinically relevant

13   subgroups      for    the    clinically        evaluable     patient

14   population for a number of factors that are predictors

15   of poor outcome, such as bacteremia, age greater than

16   50 years, and so on, and the response rates are as shown.

17                  For patients with tachypnea at baseline,

18   which has been identified in prospective studies as a

19   risk factor for poor outcome, the response rates were

20   79 percent for linezolid, 31 -- I'm sorry -- 74 percent

21   for ceftriaxone.

22                  It's    important        to     remember    that,    in

23   general, these are small numbers for these subgroups,

                                 S A G CORP.
     202/797-2525                Washington, D.C.          Fax: 202/797-2525

 1   and these were not respectively specified subgroup

 2   analyses.

 3                     Next slide, please.

 4                     Now, the results that I showed you before

 5   for   efficacy         rates   exclude      patients    with    missing

 6   outcomes.       To examine the effect of this missing data,

 7   we did one type of sensitivity analysis, which is to

 8   consider such patients to be failures, although it's

 9   important to remember that we don't really know the

10   outcome        since    we     don't    have      complete   follow-up

11   information on these patients.

12                     And the results are shown here.                    This

13   represents the patient population analyzed by excluding

14   such missing patients, and this represents the patient

15   population including such patients as failures.

16                     As you can see, the response rates fall as

17   you would expect by including such patients as failures,

18   but the pattern between the treatment arms is similar

19   to that of the primary ITT analyses.

20                     Next slide.

21                     Let me move on to Study 51.          This was a study

22   of linezolid in the out-patient treatment of community

23   acquired pneumonia.            This study enrolled and treated

                                    S A G CORP.
     202/797-2525                   Washington, D.C.           Fax: 202/797-2525

 1   540 patients with community acquired pneumonia.

 2                     This is a multi-center, multi-national,

 3   randomized, evaluated blind trial.                     Patients were

 4   randomized to linezolid or cefpodoxime; were given for

 5   ten to 14 days.          The primary endpoint was clinical

 6   outcome.

 7                     Could I have the next slide, please?

 8                     Demographics are shown here.               As you can

 9   see the treatment arms were balanced for age, gender,

10   and race.

11                     Next slide.

12                     Five   hundred    and       forty   patients         were

13   enrolled and treated.         One hundred and 20 of these had

14   a   pathogen      isolated.        There      were    421     clinically

15   evaluable patients.         Of these, 98 had a susceptible

16   pathogen isolated at the baseline.

17                     Next slide.

18                     Response rates are shown here.                    Again,

19   sizes of population are shown below.             The response rates

20   are comparable across the various analytic populations.

21    In contrast to Study 33, linezolid had higher response

22   rates.         Lower response rates were seen in the ITT

23   clinically evaluable and microbiologically evaluate

                                S A G CORP.
     202/797-2525               Washington, D.C.                 Fax: 202/797-2525

 1   analyses.      I'm sorry.     Just the ITT and CE.

 2                   Next slide.

 3                   Confidence intervals are shown here.         In

 4   general the FDA's and sponsor's confidence intervals

 5   were comparable with the exception of the clinically

 6   evaluable patient population.

 7                   Next slide.

 8                   With respect to results by pathogen, for

 9   pneumococcal pneumonia, response rates were 93 percent

10   for linezolid, 91 percent for cefpodoxime.           There were

11   a few bacteremic patients in the pneumococcal pneumonia

12   group.     Three out of three were cured in the linezolid

13   arm and three out of six in the cefpodoxime arm.

14                   Next slide.

15                   With respect to subgroup analyses, and

16   again, this is a clinically evaluable patient population

17   this time, and these were not respectively specified

18   in the protocol.      The results for predictors of core

19   outcome are shown.

20                   Next slide.

21                   We again examined the effectiveness in

22   data.     The sizes of the relevant patient populations

23   are shown below.     Again, as one would expect response

                             S A G CORP.
     202/797-2525                Washington, D.C.    Fax: 202/797-2525

 1   rates fall, but are similar to the primary ITT analysis.

 2                  Next slide.

 3                  Let   me    move    on    to     hospital       acquired

 4   pneumonia.     In this study the sponsor enrolled and

 5   treated 396 patients with hospital acquired pneumonia.

 6    This was a multi-center, multi-national randomized

 7   comparative    double      blind     trial.           Patients      were

 8   randomized to linezolid or to vancomycin.                  They could

 9   receive concomitant aztreonam.

10                  The primary endpoints were clinical and

11   microbiologic outcome.

12                  Next slide.

13                  Demographics        are        shown    here.         The

14   treatment arms were balanced with respect to age,

15   gender, and race.         In addition, APACHE II scores at

16   baseline were similar between the treatment arms.

17                  Next slide.

18                  Three hundred and ninety-six patients were

19   enrolled and treated.         One hundred and seventy-seven

20   of these had a pathogen isolated.                     There were 225

21   clinically evaluable patients.                 Of these, 95 had a

22   susceptible pathogen isolated at baseline.

23                  The    Applicant          also         identified        a

                               S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1   subpopulation of microbiologically evaluable patients

 2   who had a susceptible pathogen isolated through invasive

 3   respiratory procedures using quantitative criteria.

 4   This constituted 42 patients.

 5                     Next slide.

 6                     Response rates for the various populations

 7   are shown on this slide.           As you would expect, these

 8   vary from ITT to the per protocol patient populations.

 9    In general, there were higher response rates for

10   linezolid over vancomycin in these analyses.

11                     I think it's important to recognize for the

12   microbiologically evaluable patient populations these

13   are relatively small numbers.

14                     Next slide.

15                     Ninety-five percent confidence intervals

16   are shown here.        The width of the confidence interval

17   to   the       microbiologically    and       clinically    evaluable

18   patient populations reflect the sizes, the decreased

19   sample sizes, relative to the ITT analyses.

20                     Next slide.

21                     With respect to results by pathogen, and

22   again, this is for the microbiologically evaluable

23   patient population, these are as shown.                    For Staph.

                                S A G CORP.
     202/797-2525               Washington, D.C.            Fax: 202/797-2525

 1   aureus the response rates were 61 percent in both arms.

 2                    For MRSA, and these are small numbers, 59

 3   percent for linezolid versus 70 percent for vancomycin.

 4                    Next slide.

 5                    We also examined subgroups of interest, of

 6   clinical       interest.         With   respect        to     ventilator

 7   associated pneumonia, and this was defined here as

 8   patients who went on the ventilator at baseline, the

 9   ME patient population, the response rates were 61

10   percent for linezolid, 41 percent for vancomycin.

11                    If we did an analysis stratifying by APACHE

12   II score, for patients with the highest severity of

13   illness at baseline, and these are very small numbers

14   here, the response rates were 62 percent versus 25

15   percent.

16                    If we look at the same analysis in the MITT

17   patient population -- could I have the next slide,

18   please? -- because this analysis includes patients who

19   died before test of cure and considered such patients

20   to be failures, and there were a substantial number of

21   deaths in this study, response rates are lower.                         For

22   ventilator associated pneumonia, the response rates

23   were    54     percent     for    linezolid,      30    percent         for

                                 S A G CORP.
     202/797-2525                Washington, D.C.                Fax: 202/797-2525

 1   vancomycin, and for patients who were most ill at

 2   baseline, 46 percent for linezolid, 17 percent for

 3   vancomycin.

 4                   Again, these are small numbers in these

 5   groups.

 6                   Next slide.

 7                   Again, to examine the effectiveness in

 8   data, we considered such patients to be failures.               The

 9   results are shown here.

10                   Could I have the next slide, please?

11                   And   finally,    let    me   discuss   mortality

12   rates.     Again, let me just remind you this was one issue

13   where the FDA's analysis or analytic plan differed from

14   the sponsors, and the deaths were directly considered

15   by the FDA as to be failures.        In the sponsor's analysis

16   these    were   not   directly      considered    in    terms    of

17   assessment of cure failure.

18                   And for studies with a substantial amount

19   of deaths, this study and the MRSA study, which I will

20   discuss in a little bit, this could lead to a discrepancy

21   between the FDA's analysis, analytic results, and the

22   sponsor's.

23                   But at any rate, with respect to mortality

                              S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1   rates, for all cause mortality the rates were 18 percent

 2   in the linezolid arm and 25 percent in the vancomycin

 3   arm.     For patients whose death was assessed by the

 4   reviewer as being due to their initial infection, the

 5   mortality rates were five percent and nine percent.

 6                  Next slide, please.

 7                  Let me move on to uncomplicated skin and

 8   skin structure infections.

 9                  Next slide.

10                  The    sponsor     conducted         essentially       two

11   studies of this, 39(a) and 39.         Thirty-nine (a) enrolled

12   753 North American patients.               Study 39 enrolled 332

13   non-North American patients.           This was essentially one

14   study that was divided into two.

15                  This      was     multi-center,            randomized,

16   comparative     double    blind     trials.          Patients        were

17   randomized     to    linezolid    at   a     dose    of    either     400

18   milligrams -- I'm sorry -- at a dose of 400 milligrams

19   or they were randomized to clarithromycin at a dose of

20   250 milligrams.

21                  Patients were treated for seven to 14 days.

22    The primary endpoints were clinical and microbiologic

23   outcome.

                               S A G CORP.
     202/797-2525              Washington, D.C.                Fax: 202/797-2525

 1                   Next slide.

 2                   Demographic characteristics are shown on

 3   this slide.       As you can see, the treatment arms are

 4   balanced with respect to age, race and gender.

 5                   Next slide.

 6                   In Study 39(a), there were 753 patients

 7   enrolled and treated.      Six hundred and twenty-seven of

 8   these were clinically evaluable.            Of these, 210 had a

 9   susceptible pathogen isolated at baseline.

10                   Next slide.

11                   Response rates in the FDA analysis are shown

12   here for the ITT, CE and ME patient populations.               The

13   percentages here differ from those in the briefing

14   package.       This only reflects patients who had Staph.

15   aureus or Group A Strep.

16                   Next slide.

17                   And the confidence intervals are shown

18   here.     As you can see, the FDA's and the sponsor's

19   analyses of confidence intervals are similar.

20                   Next slide.

21                   With respect to specific pathogens, the

22   majority of isolates were Staph. aureus.             There were

23   no MRSA in this study.

                              S A G CORP.
     202/797-2525             Washington, D.C.          Fax: 202/797-2525

 1                  Response     rates      were   86   percent     for

 2   linezolid, 85 percent for clarithromycin.

 3                  Next slide.

 4                  This shows the effect of missing data.

 5   Again, if you consider patients with missing outcomes

 6   to be failures, response rates fall, but are similar,

 7   in general, to the primary ITT analysis.

 8                  Next slide.

 9                  Study 39, which had, as I've mentioned,

10   essentially the same design as 39(a), but were non-North

11   American patients.        The treatment arms were balanced

12   with respect to age, gender and race.

13                  Next slide.

14                  Three hundred and thirty-two patients were

15   enrolled.      Two hundred and fifty four of these were

16   clinically evaluable.          One hundred and one had a

17   susceptible pathogen isolated.

18                  Next slide.

19                  Clinical efficacy results are shown here.

20    Again, the ME patient population numbers are different

21   from those in your briefing package because here only

22   patients with Group A Strep. or Staph. aureus are

23   considered.

                               S A G CORP.
     202/797-2525              Washington, D.C.         Fax: 202/797-2525

 1                  Next slide.

 2                  And confidence intervals are shown here.

 3    The ME patient population, and this is true for 39(a),

 4   for the FDA is different than that for the sponsor.

 5   It's a smaller population because we only consider Group

 6   A Strep. and Staph. aureus, leading to wider confidence

 7   interval.

 8                  Next slide.

 9                  With respect to results by pathogen for

10   Staph.     aureus,   97   percent     versus    96   percent     for

11   clarithromycin.      There were a few MRSA isolates with

12   the results as shown.

13                  And this -- I'm sorry.          This is my chin on

14   the slide.     This is the microbiologically evaluable

15   patient population.

16                  Next slide.

17                  Let me move on to complicated skin and skin

18   structure infections.        The Applicant studied this in

19   Study 55 in which 819 patients with complicated skin

20   and skin structure infections were enrolled and treated.

21    This was a multi-center, multi-national, randomized,

22   comparative,     double    blind     trial.       Patients      were

23   randomized to linezolid or to oxacillin.               They could

                               S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1   be   switched      to     oral   therapy       with   linezolid     or

 2   dicloxacillin, depending on which arm they had been

 3   randomized to.

 4                    The primary endpoints were clinical and

 5   microbiologic outcome.

 6                    Next slide.

 7                    The treatment arms were generally balanced

 8   with respect to age, race, and gender.

 9                    Next slide.

10                    There    were   819     patients     enrolled     and

11   treated.       The FDA analysis focused on those ITT patients

12   who met inclusion criteria for complicated skin and skin

13   structure infections at baseline.               This population is

14   referred to here as the ITT prime patient population.

15    There were 629 patients in this population.

16                    Of these, 487 were clinically evaluable.

17    So these are patients who met baseline inclusion

18   criteria.         These    are   patients       who   met   baseline

19   inclusion criteria and post baseline criteria, such as

20   length of therapy.

21                    Finally, there were 209 microbiologically

22   evaluable patients who had a susceptible pathogen

23   isolated at baseline.

                                 S A G CORP.
     202/797-2525                Washington, D.C.           Fax: 202/797-2525

 1                    Next slide.

 2                    Response rates are as shown for the ITT,

 3   ITT prime, clinically evaluable and microevaluable

 4   patient populations.

 5                    In general response rates for linezolid

 6   were higher than those for oxacillin in all analyses.

 7                    Next slide.

 8                    The confidence intervals are shown here.

 9    The sponsor did not define an ITT prime population.

10    So only the FDA confidence interval is shown.

11                    As with the other studies, this is a

12   different ME population than the sponsor's.                 It does

13   not include many of the coagulase negative Staph.

14   species.       It really just includes Staph. epidermidis.

15    Therefore,       it's   a    smaller      sample   with   a   wider

16   confidence interval.

17                    Next slide.

18                    For specific pathogens, the rates were as

19   shown.     For Staph. aureus, 88 percent for linezolid

20   versus 86 percent for oxacillin.                There were two out

21   of three patients with MRSA in the linezolid arm who

22   were cured.

23                    For Group A Strep. the response rates were

                                  S A G CORP.
     202/797-2525                 Washington, D.C.         Fax: 202/797-2525

 1   69 percent versus 75 percent for oxacillin.

 2                    For the Enterococcus faecalis and faecium,

 3   I should mention that none of these isolates were

 4   vancomycin resistant.

 5                    Next slide.

 6                    With    respect    to   subgroups    of   clinical

 7   interest, for patients 65 or older, the response rates

 8   were    87     percent   versus    82    percent;    for   diabetic

 9   patients, 79 percent versus 68 percent; and for the

10   patients who were identified as having peripheral

11   vascular disease in the reviewer's analysis, 60 percent

12   versus 44 percent.

13                    Next slide.

14                    And, again, we examined the effect of

15   missing data through one type of sensitivity analysis

16   by considering such patients to be failures.                      The

17   response rates are as shown.

18                    Next slide.

19                    Let me move on to methicillin resistant

20   staphylococcal species infections, and this was a

21   supportive study, the idea being to garner data on

22   effectiveness of linezolid in the treatment of MRSA

23   infections at defined body sites, infections at defined

                                S A G CORP.
     202/797-2525                Washington, D.C.          Fax: 202/797-2525

 1   body sites.       So this is a pathogen driven study, but

 2   in the context of specific infections.

 3                    There were 460 patients with known or

 4   suspected methicillin resistant staphylococcal species

 5   infections, with pneumonia, skin and skin structure,

 6   urinary tract infection, bacteremia of unknown origin.

 7                    This was a multi-center, multi-national,

 8   randomized, comparative, open label trial.                    Patients

 9   were randomized to linezolid or to vancomycin for seven

10   to 28 days.

11                    As you heard from Dr. Hafkin, linezolid

12   could also be at the discretion of the investigator --

13   could be given PO after the IV course of therapy.

14   Patients       could   receive      concomitant        aztreonam      or

15   gentamicin, and the primary endpoints were clinical and

16   microbiologic outcome.

17                    I should just mention that the criteria used

18   to   define     pneumonia     and    skin      and    skin    structure

19   infections were consistent with those used for the

20   indication       specific      studies         that    I've    already

21   discussed.

22                    Next slide.

23                    Characteristics of the patients are shown

                                 S A G CORP.
     202/797-2525                Washington, D.C.             Fax: 202/797-2525

 1   on this slide.     As you can see, the mean age for both

 2   groups was 64 in the linezolid arm, 60 in the -- 64 years

 3   of age in the linezolid arm and 60 in the vancomycin

 4   arm.     The groups were balanced overall in terms of

 5   demographic characteristics.

 6                   Next slide.

 7                   Four   hundred     and        sixty   patients     were

 8   enrolled and treated.        Three hundred and one of these

 9   had a pathogen isolated.            There were 241 clinically

10   evaluable patients.        Of these, 126 had a susceptible

11   pathogen isolated at baseline.

12                   Next slide.

13                   Response rates are shown for the various

14   patient populations.        We focused on the MITT and the

15   ME patient populations since this was a pathogen driven

16   study.     The response rates for MITT were 59 percent

17   versus     66   percent,     whereas      for     the   ME    patient

18   population, the response rates were 76 percent versus

19   72 percent.

20                   One thing I want to remind you is that in

21   the FDA analysis, patients who died before test of cure

22   were considered failures in this analysis.                       Those

23   patients were excluded from this analysis unless they

                                S A G CORP.
     202/797-2525               Washington, D.C.             Fax: 202/797-2525

 1   died from their initial infection.

 2                     Next slide.

 3                     The    confidence     intervals   in    the    FDA's

 4   analysis and the sponsor's analysis are shown here.

 5   let me just mention two things about the FDA analysis.

 6    In the MITT analysis, as you've seen, the point estimate

 7   of the difference in response rates is negative.                  For

 8   the ME analysis it's positive.

 9                     A similar shift is seen for ITT to CE.

10                     Next slide.

11                     With respect to results by pathogen, the

12   vast majority of the isolates were methicillin resistant

13   Staph.         aureus.       The     response   rates       in    the

14   microbiologically evaluable patient population were 78

15   percent for linezolid versus 72 percent for vancomycin.

16                     Could I have the next slide?

17                     In the MITT analysis, the response rates

18   were 56 percent for linezolid versus 66 percent for

19   vancomycin.        So the ME analysis response rates were

20   higher for linezolid for MRSA patients.                  In the MITT

21   analysis they were lower.

22                     Next slide.

23                     When outcomes were broken down by site of

                                 S A G CORP.
     202/797-2525                 Washington, D.C.        Fax: 202/797-2525

 1   infection, pneumonia, skin and skin structure with their

 2   primary diagnoses, and this is the ME analysis, response

 3   rates for pneumonia were -- and these are small numbers

 4   -- 90 percent versus 71 percent; for skin and skin

 5   structure 79 percent versus 73 percent.

 6                  Could I have the next slide?

 7                  In the MITT analysis for pneumonia the

 8   response rate for linezolid was 43 percent versus 54

 9   percent in vancomycin; for skin and skin structure 69

10   percent versus 77 percent.

11                  So, again, in the ME analysis response rates

12   were higher for linezolid by site of infection for the

13   two   major    categories    of   infection.      In   the    MITT

14   analysis, they were lower.

15                  Next slide.

16                  The effect of missing data is shown here.

17    Again, response rates fall as one considers patients

18   with missing outcomes to be failures.

19                  Next slide.

20                  Let   me   move    on    to   studies   involving

21   vancomycin resistant enterococcal infections.

22                  Next slide.

23                  As you've heard, the sponsor had as its

                               S A G CORP.
     202/797-2525              Washington, D.C.         Fax: 202/797-2525

 1   pivotal study Study 54(a) which enrolled and treated

 2   145    adult    patients    with     known     or   suspected     VRE

 3   infection, which was defined in the context of infection

 4   at specific body sites.             This was a multi-center,

 5   randomized, dose comparison trial which was double

 6   blind, and unlike the equivalence trials that I've

 7   described before, this was a superiority trial.

 8                   Patients     were     randomized      to    receive

 9   linezolid 600 milligrams IV or to receive linezolid 200

10   milligrams IV, and the study hypothesis was that the

11   high dose arm was superior to the low dose arm.

12                   Patients      could       receive     concomitant

13   aztreonam or aminoglycosides, and the primary endpoint

14   was clinical outcome.

15                   Next slide.

16                   This shows the demographics of the patients

17   who were enrolled and treated.                As you can see, the

18   demographics were similar.          The Applicant also obtained

19   data on severity of illness at baseline using an MPM

20   II score.      The arms were balanced with respect to this

21   characteristic.

22                   Next slide.

23                   The primary patient population analyzed by

                                S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1   the FDA were those intent to treat patients who had VRE

 2   at baseline, which is referred to here as the MITT-VRE

 3   patient        population.      So    this      excluded     those     ITT

 4   patients who did not have VRE at baseline.

 5                     We also focused on the patients who had VRE

 6   bacteremia at baseline.          There were 117 in the MITT-VRE

 7   patient population.            Thirty-four of these had VRE

 8   bacteremia baseline.

 9                     Next slide.

10                     Response rates are shown here.                  In the

11   MITT-VRE patient population, and these exclude patients

12   with missing outcomes, the response rates were 67

13   percent versus 52 percent.                  The P value for the

14   difference was .16.

15                     For the bacteremic patient population, the

16   sizes of the population are shown here, 59 percent versus

17   29 percent, with a P value of .15.

18                     Next slide.

19                     With respect to results by pathogen, as you

20   would expect, most of the pathogens isolated were E.

21   faecium.        There were a handful of patients with E.

22   faecalis.        A few patients had both pathogens.             Response

23   rates in the high dose arms were 67 percent for E. faecium

                                  S A G CORP.
     202/797-2525                 Washington, D.C.              Fax: 202/797-2525

 1   and the low dose arm 53 percent.

 2                  For faecalis, three out of four patients

 3   in the high dose arm were cured; zero out of two in the

 4   low dose arm.

 5                  Next slide.

 6                  With   respect     to    outcome   by   site    of

 7   infection, the response rates in these obviously are

 8   small numbers.        Five out of ten for bacteremia of

 9   unknown origin were cured in the high dose arm; two out

10   of seven in the low dose arm.

11                  Skin structure infections, skin and skin

12   structure infections, 69.2 percent versus 100 percent.

13                  Urinary tract infection, 63 percent versus

14   60 percent.

15                  Pneumonia, two out of three high dose

16   patients were cured versus zero out of one.

17                  And for a category of other, which was

18   almost entirely complicated interabdominal infections,

19   the response rates were as shown.

20                  Next slide.

21                  Covariate analyses were performed.             It's

22   important to recognize that these were not prespecified

23   in the protocol.      The multivariate analysis performed

                              S A G CORP.
     202/797-2525             Washington, D.C.         Fax: 202/797-2525

 1   by the FDA incorporated risk of mortality at baseline,

 2   primary diagnosis in terms of site of infection, age,

 3   sex, weight, and presence of bacteremia at baseline,

 4   and the bottom line was that the adjusted and unadjusted

 5   analyses were consistent.

 6                  Next slide.

 7                  With respect to the effect of missing data,

 8   again, if one puts in patients with missing outcomes

 9   as failures, response rates fall, as shown here for the

10   MITT-VRE and VRE bacteremia patients.

11                  Next slide.

12                  With respect to mortality in this study,

13   all cause mortality in the MITT-VRE patient population

14   is as shown.        In the bacteremic population, four out

15   of 18 patients died in the high dose arm.            Nine out of

16   16 in the low dose arm.

17                  Next slide.

18                  We    looked   at   causes    of   death   in    the

19   bacteremic patients.       These are as shown.       In the high

20   dose arm, one patient was felt to have died by the

21   reviewer -- was felt by the reviewer to have died

22   definitively from VRE infection.             Two patients died

23   from sepsis.    The possibility that VRE contributed to

                               S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1   this    cannot    be     excluded.      One     patient    died     from

 2   respiratory failure.

 3                    For the low dose arm three patients were

 4   felt to have died from VRE infection, one from sepsis.

 5    FRE cannot be excluded as a cause of death in that

 6   patient, with the other causes as shown.

 7                    Next slide.

 8                    Covariate      analysis        of     mortality       in

 9   bacteremic patients was performed.              Again, this was not

10   prespecified.          It incorporated risk of mortality at

11   baseline, age and sex.            The adjusted and unadjusted

12   analyses were consistent.

13                    Next slide.

14                    Now, before presenting results from Study

15   54, let me just recapitulate some of the history of this

16   study.         Originally the Applicant planned a study

17   designated as 54 which would enroll 500 patients.                      In

18   June of 1999, a blinded decision was made to submit

19   patients       already    enrolled     as      Study   54(a),     which

20   constituted the 145 patients that you've just heard

21   about.

22                    This was submitted as a stand alone study,

23   and all alpha was considered to be spent on this trial.

                                 S A G CORP.
     202/797-2525                Washington, D.C.             Fax: 202/797-2525

 1                   Study 54 was continued as a support of

 2   trial.     Data on 82 patients was submitted to the FDA

 3   in December of '99.     As you heard from Dr. Hafkin, there

 4   was a total of 186 patients.            So we do not have data

 5   on 104 patients.

 6                   I think it's important to recognize that

 7   there bolstering the nonsignificant results of 54(a)

 8   with these results from Study 54 could correspond to

 9   multiple looks at the data without appropriate, that

10   is, prespecified statistical adjustment.

11                   Next slide.

12                   With that in mind, the efficacy results are

13   as follows.     For the MITT-VRE patient population, there

14   were a total of 71 patients.                 For patients with

15   non-missing outcomes, there were 28 in the high dose

16   arm, 35 in the low dose arm.            Response rates were 64

17   percent and 49 percent.        These are the response rates

18   if you add back in those patients with missing outcomes

19   as failures.

20                   Next slide.

21                   All right.     Let me change gears a little

22   bit here, a lot, I guess, and move on to safety.               I'm

23   going    to    be   discussing    clinical     adverse   events,

                              S A G CORP.
     202/797-2525              Washington, D.C.         Fax: 202/797-2525

 1   laboratory adverse events, and potential drug-drug

 2   interactions.

 3                  Next slide.       Next slide.

 4                  Adverse event rates for the various Phase

 5   III comparator controlled studies.             So I'm not going

 6   to show you any data from the dose comparison studies;

 7   just the comparator controlled studies are shown here.

 8                  As you can see, there were significant

 9   adverse event rates in both treatment arms across all

10   studies.

11                  For   all   studies      combined,   the   adverse

12   events rates were 56 percent versus 50 percent for

13   linezolid versus comparator.

14                  Next slide.

15                  If one looks at drug related adverse events,

16   in a number of the studies there were a higher rate of

17   drug related adverse events in linezolid arm than in

18   the comparator arm, although this was not invariably

19   true.    For example, in HAP there was a lower rate.

20                  Overall the rate of drug related adverse

21   events was 22 percent for linezolid, 16 percent for

22   comparator.

23                  Next slide.

                                S A G CORP.
     202/797-2525               Washington, D.C.         Fax: 202/797-2525

 1                  With respect to discontinuations related

 2   to adverse events, the rates are as shown.    These varied

 3   across studies for linezolid from three to ten percent.

 4    Overall six percent of linezolid treated patients were

 5   discontinued for an adverse event; five percent of

 6   comparator treated patients.

 7                  Next slide.

 8                  If one looks at discontinuations due to drug

 9   related adverse events, for some studies, particularly

10   the pneumonia studies, the rate of discontinuation due

11   to drug related adverse event was higher in the linezolid

12   arm, although, again, this was not invariably true for

13   HAP.    The rate was higher for the comparator arm.

14                  For all studies combined, 2.4 percent of

15   linezolid treated patients discontinued for a drug

16   related AE versus 1.9 percent of comparator treated

17   patients.

18                  Next slide.

19                  This shows discontinuations according to

20   specific adverse events, and one thing I want to be very

21   clear about is the percentages shown are relative to

22   the number of patients discontinued for any adverse

23   event, not the entire patient population.

                             S A G CORP.
     202/797-2525            Washington, D.C.      Fax: 202/797-2525

 1                        So   nine    percent      of    linezolid      treated

 2   patients who discontinued for any adverse event did so

 3   for nausea versus four percent for the comparator.                        The

 4   second most common cause was pneumonia for linezolid.

 5    The third most common was headache.

 6                        Other causes included diarrhea, dyspnea and

 7   vomiting, and again, these are just all adverse events

 8   whether drug related or not.

 9                        Next slide.

10                        If one now looks at drug related adverse

11   events,        and    again,      this     refers     to     patients     who

12   discontinued for any drug related adverse event, not

13   the entire patient population, 22 percent of linezolid

14   treated patients who discontinued for a drug related

15   adverse event did so for nausea versus eight percent

16   for comparator.

17                        For headache, the figures were 16 percent

18   versus three percent; vomiting, 12 percent versus eight

19   percent;       diarrhea,         12   percent       versus    11   percent;

20   thrombocytopenia, six percent versus zero percent.

21                        Next slide.

22                        Let me move on to a consideration of

23   laboratory findings, and then I'm going to focus on

                                      S A G CORP.
     202/797-2525                     Washington, D.C.             Fax: 202/797-2525

 1   thrombocytopenia.

 2                   Next slide.

 3                   This       shows        the        development            of

 4   thrombocytopenia in different studies in patients who

 5   had normal platelet counts at baseline.                       We do not

 6   consider in this analysis those patients who have

 7   abnormal platelet counts at baseline, and the sponsor

 8   has looked at this issue.

 9                   So   the     rate    varies       with      studies     for

10   linezolid ranging from two percent in the skin and skin

11   structure infection studies to 11 percent in the MRSA

12   studies.

13                   It's important to recognize that the MRSA

14   study involved sicker patients with a longer duration

15   of therapy.

16                   Next slide.

17                   If     one      looks       at     the      degree       of

18   thrombocytopenia and uses, for example, the NCI common

19   toxicity       criteria,     when      we       look   at    Grade      III

20   thrombocytopenia, which was the most severe grade that

21   developed, so we're looking at patients who develop a

22   platelet count of less than 50,000 during study.                        The

23   rates for linezolid range from zero percent to 2.5

                                  S A G CORP.
     202/797-2525                 Washington, D.C.               Fax: 202/797-2525

 1   percent.       Again, the most common -- this was most common

 2   in the MRSS study, Study 31.

 3                    Next slide.

 4                    If one looks at the effect of linezolid dose

 5   on the development of thrombocytopenia, there appears

 6   to be an effective dose.           If one looks at the dose

 7   comparison study, the rates were 13 percent for the high

 8   dose arm versus 11 percent for the low dose arm.

 9                    For all Phase III studies, and here we're

10   describing high dose as greater than a gram a day; low

11   dose is less than a gram a day of linezolid; five percent

12   versus three percent.

13                    For Study 11, this          was a Phase II study

14   of linezolid in bacteremic patients.              This only used

15   the high dose.         So there was no low dose arm for

16   comparison here, but the incidence of thrombocytopenia

17   in patients with normal platelet counts at baseline was

18   six percent.

19                    For all Phase II studies, three percent

20   versus two percent; and for all Phase II and III studies

21   combined, five percent versus three percent.

22                    Next slide.

23                    We also looked at the issue of resolution

                               S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1   of thrombocytopenia, and I just want to mention that

 2   this entire analysis for laboratory findings, and as

 3   I'll show you in a little bit for drug-drug interactions,

 4   was done with the assistance of Dr. Ana Scharffman, as

 5   well as Dr. Joyce Korvic, and I really want to thank

 6   them for their assistance with this, which allowed us

 7   to look at a variety of issues.

 8                  I also want to thank the Applicant for

 9   reorganizing the data sets to allow this analysis to

10   be done.

11                  But at any rate, this is linezolid.           This

12   is comparator, and this is just for Study 31, which as

13   you remember is the study in which the most pronounced

14   effect on thrombocytopenia was seen.

15                  Each red line or green star represents a

16   patient with thrombocytopenia.            The minimum value is

17   at the left.    The maximum value is at the right or --

18   I'm sorry -- value at follow-up is at the right.

19                  If the line continues off the graph, that

20   patient showed complete resolution of thrombocytopenia.

21    So for the majority of patients in the linezolid arm

22   who had thrombocytopenia, thrombocytopenia resolved or

23   it was going in the right direction.

                            S A G CORP.
     202/797-2525           Washington, D.C.           Fax: 202/797-2525

 1                  For these patients we do not have laboratory

 2   follow-up on these patients.            However, there were no

 3   clinical adverse events that were identified in relation

 4   to     thrombocytopenia,        such         as   gastrointestinal

 5   hemorrhage for these patients or a requirement for

 6   platelet transfusions.

 7                  Next slide.

 8                  So let me try and summarize this.                    The

 9   incidence of thrombocytopenia in the studies was one

10   percent of 13 percent; for Grade III, zero to 2.5

11   percent, depending on the patient population.                   Higher

12   doses appeared to be associated with an increased

13   incidence.

14                  Thrombocytopenia appeared to resolve in

15   linezolid      treated    patients       who      had     laboratory

16   follow-up.

17                  There     were   no     related     adverse      events

18   identified, and finally, I'll just mention that looking

19   at other cell lines, no parent effect was identified.

20                  Next slide.

21                  Let me move to drug-drug interactions.

22                  Next slide.

23                  Let me just step back for a minute, and

                               S A G CORP.
     202/797-2525              Washington, D.C.              Fax: 202/797-2525

 1   you've seen some data before from the Applicant about

 2   the relative MAO inhibition activity of linezolid.

 3   These are two classic MAO inhibitors, clorgyline and

 4   selegiline.           I    want    to   focus     on   the     inhibitory

 5   constants, the KIs.

 6                    For MAO A, and that's the activity that's

 7   associated with adrenergic hypertensive type crises,

 8   as well as MAO B, which is associated with serotonin

 9   syndromes.

10                    As you can see, the KI for linezolid is

11   considerably higher than for a drug such as selegiline.

12    However, I think it's important to recognize the peak

13   plasma concentrations of linezolid that are achieved

14   are in the neighborhood of the KI.

15                    Next slide.

16                    The sponsor was aware of this issue and has

17   examined this in their Phase I studies by conducting

18   a number of drug interaction studies, and I'll just

19   mention one here, and you've seen this data in another

20   form.          They       looked   at    both     interactions          with

21   sympathomimetic agents and serotonergic agents just to

22   look at the sympathomimetic amine interactions.

23                    A study was performed in which patients --

                                   S A G CORP.
     202/797-2525                   Washington, D.C.              Fax: 202/797-2525

 1   and these are -- I'm sorry -- not patients, but normal

 2   volunteers -- received placebo, phenylpropanolamine,

 3   linezolid           plus       placebo       or        linezolid       plus

 4   phenylpropanolamine.

 5                       The    maximum    change      in    systolic     blood

 6   pressure from baseline is shown here.

 7                       Next slide.

 8                       As you've heard, patients received a number

 9   of concomitant medications during the course of the

10   study.     One of the outcomes of the Phase I studies was

11   that the sponsor incorporated this issue into the study

12   design both with respect to cautioning physicians and

13   investigators          about    patients receiving concomitant

14   medications, as well as capturing data on the frequency

15   with which these medications were administered.

16                       These are some of the agents that we've

17   looked at.          As you can see, in general, for concomitant

18   medications the proportion of patients receiving these

19   were similar between treatment arms, generally five

20   percent        or     less,    except      for    some     pathomimetic

21   bronchodilators where it was 18 to 20 percent in the

22   two arms.

23                       Next slide.

                                    S A G CORP.
     202/797-2525                   Washington, D.C.             Fax: 202/797-2525

 1                    We examined the database in the NDA for

 2   potential        MAO        inhibitor        associated         drug-drug

 3   interaction events.           There were only small numbers of

 4   events found in patients who had received concomitant

 5   medications.

 6                    There was no clear association between

 7   adverse events examined and the use of concomitant

 8   medications,       and      classic    MAO       inhibitor     associated

 9   events were not seen.          There were no hypertensive crises

10   identified and no cases of serotonin syndrome.

11                    Next slide.

12                    Let me move on to linezolid resistance.

13   This has been induced in the laboratory.                 The mechanism

14   appears to be a GDU transversion on the 23S ribosomal

15   RNA.    The sponsor has found that the frequency is less

16   than one to ten to the ninth.                       It may result in

17   cross-resistance         to     lenclosomides         (phonetic)         and

18   chlorinfenacol.

19                    Next slide.

20                    With respect to development of resistance

21   in linezolid in clinical trials, as you've heard, this

22   has only been seen with enterococcal species.                         There

23   were    15     cases   in    the   NDA    database,      nine      in    the

                                   S A G CORP.
     202/797-2525                  Washington, D.C.               Fax: 202/797-2525

 1   compassionate use study, six in the dose comparison

 2   studies.

 3                     Mean duration of therapy in these patients

 4   was   32       days.   Almost   all    of     the    cases   involved

 5   enterococcus faecium.        There was only one that involved

 6   faecalis.

 7                     The increase in the MIC was to eight

 8   micrograms per mL for six isolates; 16 micrograms per

 9   mL for eight isolates; and 32 micrograms per mL for one

10   isolate, which was the Enterococcus faecalis isolate.

11                     Next slide.

12                     In the compassionate use trial, there were

13   nine cases of resistance developed.            Eight of these were

14   faecium.        One was Enterococcus faecalis.          Six of these

15   patients were considered therapeutic failures.                   Three

16   were considered cures.

17                     Next slide.

18                     In the dose comparison trials, there were

19   six cases of resistance development.                All of these were

20   Enterococcus faecium.        There were two in the low dose

21   group.     I'm sorry.     Four in the low dose group.            Three

22   of these four were considered failures.               There were two

23   in the high dose group.            One of these patients was

                                S A G CORP.
     202/797-2525               Washington, D.C.             Fax: 202/797-2525

 1   considered a failure.

 2                  So can I have the next slide?

 3                  This concludes the FDA's analysis.         I just

 4   would like the committee and the audience to be aware

 5   that this analysis was the result of a lot of hard work

 6   by a group of scientists of the agency who are shown

 7   here.    I want to thank all of them.

 8                  I'd also like to thank the Applicant for

 9   provision of data for this NDA.

10                  Thank you.      I'll be happy to answer any

11   questions.

12                  CHAIRMAN RELLER:        Thank you, Dr. Ross.

13                  Questions?

14                  Barbara.

15                  DR. MURRAY:      Dr. Ross, do you or perhaps

16   the sponsor over the lunch break would pull out in the

17   MRSA group, you mentioned some got aminoglycosides, and

18   I would be curious to know how many.               Were those

19   documented MRSA and what was the susceptibility of the

20    MRSA to getimicin if it was an MRSA that patients that

21   got that.

22                  DR. ROSS:    What I can tell you, and let me

23   just give you some numbers from the MITT analysis, and

                               S A G CORP.
     202/797-2525              Washington, D.C.        Fax: 202/797-2525

 1   this is not just MRSA.          This is just the entire MITT

 2   patient population.        So it does include some MRSE and

 3   it does include a handful, actually very few methicillin

 4   susceptible.

 5                  But   for     those     patients   who   received

 6   aminoglycosides, we identified -- let me just look at

 7   these numbers here -- in the MITT patient population

 8   there were 14 out of 30 cures in the linezolid arm.

 9   So that's 47 percent, versus 15 out of 27 in the

10   vancomycin arm.      That's 56 percent.

11                  We do not -- I don't have data for you on

12   susceptibility.

13                  If you look at patients who did not receive

14   aminoglycosides, there was a larger patient population.

15    That was 61 out of 98 in the linezolid arm.             So that

16   was 62 percent, versus 59 out of 85 in the vancomycin

17   arm, which is 69 percent.

18                  I have the corresponding figures for the

19   microbiological evaluable patient population if you

20   need those.

21                  CHAIRMAN RELLER:         Other questions?

22                  Yes, Dr. Danner.

23                  DR. DANNER:        Yeah, I have two questions

                                S A G CORP.
     202/797-2525               Washington, D.C.        Fax: 202/797-2525

 1   actually related to potential toxicity and one related

 2   to a question on metabolites.

 3                  In terms of potential toxicity, earlier in

 4   the day I think the numbers were that there were 13 of

 5   632 patients had episodes of hypertension, and these

 6   were patients on potentially interacting drug.               Of

 7   those patients, there were 13 of 632 had episodes of

 8   hypertension as an adverse effect, but only one of those

 9   were thought by the clinician on the scene to be related

10   to linezolid.

11                  What was the incidence of hypertension as

12   an adverse event in subjects not             on potentially

13   interacting drug?    Anybody know the answer?

14                  DR. HAFKIN:   If I could show I-98, please.

15                  CHAIRMAN RELLER:      Dr. Hafkin is answering

16   this question.

17                  DR. HAFKIN:   Now, it turns out that for the

18   great majority of both linezolid treated patients and

19   the comparator treated agents, the blood pressure

20   elevation was actually at baseline or was after the end

21   of therapy analysis.

22                  If you actually look at linezolid, of the

23   patients we have up there, the 13 with an interacting

                            S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   med., actually six patients had adverse event of

 2   hypertension reported at a time after the patient got

 3   linezolid and the comparator.           So it was actually within

 4   the -- made pharmacokinetic sense that he could have

 5   had a hypertensive response.

 6                    Only one of those patients was thought by

 7   the investigator to have been related.

 8                    Now,   it's   important        to   understand       the

 9   details of that one patient.            It was a 92 year old man

10   who    was     hypertensive     in    his      history,     had    acute

11   pneumonia, was admitted to the hospital, and was treated

12   simultaneously with salbutamol and linezolid.

13                    The    investigator           became     frightened,

14   stopped the treatment.             So we had no opportunity to

15   rechallenge the patient.

16                    DR. DANNER:       My other question with regards

17   to potential toxicity is regarding the effect on the

18   bone marrow effects and effects on platelets.                  From the

19   earlier presentation, it sounded to me like perhaps

20   Pharmacia/Upjohn found the problem might be more of a

21   problem with people who start out with lower platelet

22   counts, and that raises a question, is if you have

23   patients       who   have   bone     marrow     insufficiency        like

                                 S A G CORP.
     202/797-2525                Washington, D.C.              Fax: 202/797-2525

 1   somebody who has had a bone marrow transplant or has

 2   a hematologic malignancy or who has been heavily

 3   pretreated with myelosuppressive chemotherapy, is it

 4   conceivable that this problem with platelets, in fact,

 5   might be a bigger problem in that population?

 6                  Were there any patients like that in the

 7   compassionate use?

 8                  DR. HAFKIN:   I don't have slides prepared

 9   for the compassionate use trial, but what I can tell

10   you is that patients with terribly severe underlying

11   illness have taken linezolid for up to three months,

12   and even in that circumstance, our hematologic adverse

13   event rate is around three percent.

14                  If you have a minute, perhaps looking at

15   the worst case, which is the platelet count, we could

16   go through a couple of those slides that I showed you.

17    Would that be appropriate at this point?       Because we

18   actually go straight to the data.

19                  Let's start out with the hazard function

20   curve, which is on screen.        L51 would be fine.     This

21   is the analysis we did to detect the problem, and I might

22   note I think that you see no difference until 16 days,

23   and then you start seeing a divergence of the curve.

                            S A G CORP.
     202/797-2525           Washington, D.C.       Fax: 202/797-2525

 1    That divergence of the curve represents one percent

 2   of the population.

 3                  As I said, there are 16 patients that that

 4   increase in slope between 16 and 18 days represents.

 5                  Let's go to the next slide, which is the

 6   scattergram for the -- this is the distribution for the

 7   whole patient population.           I'd like the abnormal

 8   patient population, which should be the very next one.

 9    Yes, this is exactly what I want, 54.

10                  If you'll note here, about half -- you can't

11   see them carefully.     It turns out that if you look at

12   the relative risk, having a low platelet count at

13   baseline drives you to having a greater risk of reduction

14   of platelets at any time during the treatment period.


16                  But note that for those people that are down

17   here or below this blue dotted line, which is normal

18   limits, the change in platelets count over time is very

19   minimal, and it's our view that there's no increased

20   risk for this patient population.          There is an increased

21   risk for them to stay down here, but there's no increased

22   risk to get down lower.

23                  There is that one example where we have this

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   one patient who starts out at something like 75,000 and

 2   then goes to about 19,000.               Those numbers might be off

 3   a little bit, but they're based on my memory of the case,

 4   but even that worst case analysis where you saw the slow

 5   decrease in platelet count, because of the platelet

 6   count, suppression is the patient's underlying illness.

 7    This patient had malignancy and was on chemotherapy

 8   -- had a malignancy and was on chemotherapy.

 9                     DR. DANNER:         Did the metabolites have an

10   effect on this?

11                     DR. HAFKIN:        When we look at those people,

12   remember again we've got two lines of evidence.                       We've

13   got 34 patients in the compassionate use trial who have

14   received linezolid for up to 60 days.                   When we look at

15   hematologic           changes   in    that     group,      they   are    not

16   different from the main group of patients that are

17   treated        with    their    dreadful      underlying       illnesses.

18   These     are     very      sick     people,       often    transplants,

19   immunosuppressed.

20                     Perhaps a shift table would help you see

21   the extent of change.              Would you like to see a shift

22   table that's           --

23                     DR. DANNER:         Okay.

                                     S A G CORP.
     202/797-2525                    Washington, D.C.             Fax: 202/797-2525

 1                  DR. HAFKIN:    Yes?

 2                  CHAIRMAN RELLER:      Dr. Hafkin, let's see the

 3   shift table.     Dr. Hafkin was addressing the platelet

 4   --

 5                  DR. HAFKIN:    This is the worst -- another

 6   way of showing you the worst of the worst, and what I

 7   would point out to you is that the shift in platelet

 8   count is typically one box.        There isn't anybody that

 9   goes from this to this point, and here we've defined

10   the platelet number, and this is linezolid treatment

11   and this is comparator treatment.

12                  So you'll see if you look at the shift tables

13   the typical response if there's going to be one will

14   be from this box to that box or that box to that box.

15                  CHAIRMAN RELLER:      Dr. Rodvold.

16                  DR. RODVOLD:     Maybe we could just follow

17   up with this because I've done some platelet studies

18   in cats.

19                  Did you look at percentage?      I mean you're

20   trying to explain that to us, but if you look at

21   percentage drop from the baseline sometimes it gives

22   the clinician a better handle a little bit than if you

23   start at 100,000 and then drop by 50 percent, if at first

                             S A G CORP.
     202/797-2525            Washington, D.C.         Fax: 202/797-2525

 1   I start at 300,000 and I only drop it 25 percent.         Did

 2   you look at that?

 3                  DR. HAFKIN:    We only at one cut.        That

 4   initial table where we detected a signal in 2.4 percent

 5   of patients in the linezolid group versus 1.5 percent

 6   in the comparator group, we used a 75 percent reduction

 7   for that first cut.      We found that to be the most

 8   sensitive percent reduction.

 9                  DR. RODVOLD:   The other question I have on

10   the platelet count goes back to metabolites, but did

11   you look at -- you said you had a group of people that

12   had impaired renal function of platelets greater than

13   four.

14                  DR. HAFKIN:    Yes.

15                  DR. RODVOLD:    Did those patients have a

16   higher incidence of adverse events in the platelet count

17   and maybe even the hepatic test count in regards to the

18   renal impairment which indirectly may be telling you

19   the metabolites contributed to that profile if it was

20   higher?

21                  DR. HAFKIN:    Well, if you'll recall the

22   safety data I had to share with you, I only had 17

23   patients that fall into that area.        Let me pull up a

                            S A G CORP.
     202/797-2525           Washington, D.C.       Fax: 202/797-2525

 1   slide that I showed.     It was S-194, and we'll actually

 2   go from this.

 3                  We're looking at the number of deaths of

 4   these patients with serum creatinines greater than four,

 5   the number of patients that died.          No difference there.

 6    The number of patients with an adverse event leading

 7   to discontinuation.     Well, we've got one patient on the

 8   linezolid.

 9                  If I could go to the next slide in this

10   series, you'll see the reasons for or the adverse events

11   that are reported in the study for the small number of

12   patients, linezolid here, comparator here.          You'll find

13   anemia here.

14                  Let's go to the next slide.

15                  So there is no mention of thrombocytopenia

16   or anything that we can logically connect to hematologic

17   toxicity except for that anemia.

18                  DR.   RODVOLD:       But     did   you   look    at

19   percentage changes in the --

20                  DR. HAFKIN:   When we look at -- I can tell

21   you what we've done.         The average anilide result,

22   whether you are talking about hemoglobin, hematocrit,

23   white count, platelet count is terrible for both of these

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   groups, and it's really terrible throughout the period

 2   of treatment.      These are super sick people.

 3                    CHAIRMAN     RELLER:           Dr.   Ross,   in    your

 4   analysis of patients with uncomplicated skin and skin

 5   structure infections, one of 39 had infection owning

 6   to methicillin Staph. aureus, and in the complicated

 7   category of SSSI, three of 83 patients were infected

 8   with methicillin resistant Staph. aureus, and yet for

 9   the    studies     where     there      was     an    enrichment     for

10   methicillin resistant strains, we had 33 MRSA out of

11   51, 33 patients out of the 51 with MRSA had skin and

12   skin structure infection as the site involved.

13                    Do we know of those patients -- so now we

14   have the 33 with MRSA with SSSI -- how they broke down

15   in terms of uncomplicated and complicated infection?

16                    DR. ROSS:     Actually I'm going to once again

17   refer you to the Applicant to see if they can provide

18   that information.

19                    CHAIRMAN RELLER:         Seeking to see what our

20   numbers are having to do with the issue of MRSA in SSSI

21   uncomplicated and complicated in adults.

22                    DR. HAFKIN:      Yes.     If you use -- it depends

23   on our definition.          The number of people that had in

                                 S A G CORP.
     202/797-2525                 Washington, D.C.            Fax: 202/797-2525

 1   hospital infections, were severe enough in terms of

 2   comorbidity to require hospitalization, all of them,

 3   if you were to use that global diagnosis of an adverse

 4   event that was severe enough to keep you in the hospital,

 5   virtually everybody in Protocol 31 -- I mean the number

 6   of people who got to leave the hospital in that protocol

 7   with oral therapy was small because we had such a

 8   severely ill population of patients.

 9                   For those people that had the traditional

10   indicator of severity requiring surgery debridement

11   over the period of therapy, about one-third of patients

12   in   Protocol     31   required     at      least   one    surgical

13   intervention at baseline.

14                   Let me go to Slide E-32 or ER-32, which

15   breaks down the data a bit more by diagnosis.             This still

16   doesn't get to what you're trying to get to, which is

17   the severity of illness, but at least this gives you

18   specific diagnoses and outcome.             This is the clinical

19   care of the sponsor's group.         This is linezolid.         This

20   is vancomycin.     You'll see very comparable outcomes for

21   each diagnosis.

22                   As I say about a third of this group, maybe

23   as high as 38 percent of this group actually required

                              S A G CORP.
     202/797-2525             Washington, D.C.            Fax: 202/797-2525

 1   for both linezolid and vancomycin repeated surgical

 2   debridement because their infection was so extensive.

 3                       Perhaps after lunch, if there is more

 4   information that could be shared with us, whether these

 5   were infections that were complications of other things

 6   going on and not the primary reason for hospitalization,

 7   whether any of them were community acquired versus

 8   hospital       acquired     being      complications       of    surgical

 9   procedures, that sort of information, to get a better

10   feel for how this compound works in patients with

11   documented MRSA infections involving skin and skin

12   structure, and if we got bacteremia information on those

13   patients, in what setting it occurred.

14                       Dr.   Rodvold,      do    you   have   any    further

15   questions?          Dr. Lowy.

16                       DR. LOWY:     Regarding the hospital acquired

17   pneumonia, in the original study design was there any

18   consideration for switching patients to oxacillin, the

19   isolate        of    methicillin         susceptible       rather     than

20   vancomycin?

21                       CHAIRMAN RELLER:         So the question is those

22   patients who ended up having methicillin susceptible

23   strains; was there a revision or reversion to oxacillin

                                  S A G CORP.
     202/797-2525                    Washington, D.C.           Fax: 202/797-2525

 1   even though they may have been treated initially with

 2   vancomycin?

 3                       DR. HAFKIN:    For Protocol 31, that patient

 4   would not have been valuable if they had -- if they had

 5   a baseline isolate that -- perhaps I don't understand

 6   your question, sir.

 7                       DR. LOWY:     I'm just wondering in terms of

 8   the original design of the study if it would be possible

 9   rather than continuing on vancomycin for an individual

10   who has a methicillin susceptible Staph., whether that

11   individual after initiation of therapy could have been

12   switched       to    oxacillin,     which       might   have   been     a

13   preferable regimen.

14                       DR. HAFKIN:     There were several patients

15   who had methicillin resistant Staph. by the local lab,

16   but when we got that isolate to our central lab, they

17   were found to be methicillin susceptible.

18                       The physicians, of course, were always

19   capable of doing anything they wanted to do.                        They

20   always do, if you've ever participated in a study, on

21   the one hand.

22                       On the other hand, under the criteria of

23   the study, you had to be evaluable microbiologically

                                 S A G CORP.
     202/797-2525                 Washington, D.C.            Fax: 202/797-2525

 1   for this study.        You had to have a methicillin Staph.

 2   at baseline.

 3                    DR. LOWY:       Not for the hospital acquired

 4   pneumonias.

 5                    DR. HAFKIN:          No, no, that's true.          The

 6   complex skin and soft tissue trial 55, that patient

 7   population       did    get     oxacillin.        The     comparator

 8   population did get oxacillin.

 9                    DR. LOWY:        Let me ask another question

10   then.     Don't the individuals who had Staphylococcal

11   infections that were hospital acquired pneumonia cases

12   -- how many of them were actually mixed infections with

13   other organisms.        A great many of them.

14                    In    fact,     at   baseline    virtually      every

15   investigator used concomitant medications because they

16   were -- we actually have specific information that we

17   could show you about Gram stain results, culture results

18   for that patient population.

19                    CHAIRMAN RELLER:          Could we have that after

20   lunch as part of the follow-up to questions posed

21   earlier?

22                    The final question before lunch goes to Dr.

23   Leggett.       You had a question?

                                   S A G CORP.
     202/797-2525                  Washington, D.C.          Fax: 202/797-2525

 1                  DR. LEGGETT:    In a follow-up to this one

 2   in terms of trying to look at comparator arms, I think

 3   he was alluding to vancomycin as a lousy drug perhaps.

 4    Maybe we're going to use a better one.      So my question

 5   is in terms of the comparator arms, both with the

 6   oxacillin two grams Q6 and the clarithromycin, 250

 7   milligrams BID, did you do any calculations either

 8   before or during your studies about what the anticipated

 9   population time above MIC for your pathogens would be

10   to make it similar to the time greater than above the

11   MIC at 40 percent for the linezolid?

12                  DR. HAFKIN:    In Protocol 55, we felt that

13   we had -- at two grams every six hours, we felt we were

14   well above the MIC of our target pathogens throughout

15   the dosing interval.

16                  In   terms     of      clarithromycin,      the

17   pharmacokinetics of the drug are quite interesting.

18   It's well distributed, as you know.        We didn't try to

19   look at the activity of that drug from that perspective

20   because of its penetration into inflammatory cells.

21                  CHAIRMAN RELLER:       The sponsor has been

22   working on the questions that were posed before the

23   break, and they will be handled when we resume at 1:30

                            S A G CORP.
     202/797-2525           Washington, D.C.        Fax: 202/797-2525

 1   this afternoon.     At the moment we have no individual

 2   schedule for the open public hearing, but we will ask

 3   for three minute queries from the floor at that time.

 4                  There's a table reserved for members in the

 5   restaurant, and we will resume for follow-up of the

 6   questions addressed this morning promptly at 1:30, and

 7   then address the questions posed by the agency.

 8                  One, thirty reconvene.

 9                  (Whereupon, at 12:17 p.m., the meeting was

10   recessed for lunch, to reconvene at 1:30 p.m., the same

11   day.)










                             S A G CORP.
     202/797-2525            Washington, D.C.     Fax: 202/797-2525

 1                  A-F-T-E-R-N-O-O-N           S-E-S-S-I-O-N

 2                                                                (1:31 p.m.)

 3                    CHAIRMAN RELLER:          Good afternoon.          We have

 4   a new audio system for the members around the table.

 5    I'd like to demonstrate it.

 6                    There's an on/off button.              When it's on,

 7   the red light shows.           When it's off, the mic goes off

 8   and you don't hear me.            Simple as that.

 9                    (Applause.)

10                    CHAIRMAN RELLER:            I'd like to call the

11   meeting to order.        It is now time for the open public

12   hearing.       Are there any who would like to make comments

13   about the topics under discussion from the public?

14                    No response.)

15                    CHAIRMAN RELLER:          See and hearing none, the

16   open public hearing is closed.

17                    There were several questions that remained

18   at     the      close     of        this         morning's     session.

19   Pharmacia/Upjohn        has     organized        responses     to    those

20   questions, and I'd now like to ask Dr. Tarpley to

21   introduce the presentation of that responses, and this

22   does not mean that there can't be additional question.

23    Quite the contrary, but this is an organized forum from

                               S A G CORP.
     202/797-2525                  Washington, D.C.           Fax: 202/797-2525

 1   which to proceed to more detailed questions, anything

 2   that the committee wishes to discuss that would help

 3   us then vote on the questions asked of us by the FDA.

 4                  And when it comes time to voting, we will

 5   have lots of discussion, but when the votes come, it

 6   will be yes or no so that we have crisp replies, and

 7   then any additional recommendations over and above the

 8   questions asked we can forward on to the agency, but

 9   we will answer those questions yes or no at the end of

10   the discussion.

11                  Dr. Tarpley.

12                  DR. TARPLEY:      Yes, thank you, Dr. Reller.

13                  There were a number of questions that were

14   asked of us this morning.          What we'd like to do is I

15   will    introduce   Dr.    Hafkin      who   will   provide     the

16   responses, and in order to organize those questions,

17   we've agreed.    We've decided that we would repeat the

18   question so that everybody can remember what the

19   particular question was, and then we would provide the

20   response, and there are four or five of these that we

21   will proceed in that order.

22                  DR. HAFKIN:      Well, thank you.

23                  I heard several related questions, and so

                               S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1   I would propose that we try to answer questions about

 2   Strep. pneumoniae together.       I think that there was one

 3   very firm question about activity of linezolid against

 4   Strep. pneumoniae taking the pediatric data out, the

 5   activity of linezolid in Strep. pneumoniae resistant

 6   penicillin,    Strep.   species,     and   then   activity     of

 7   linezolid in bacteremia.

 8                  And so let me go first to all of our Phase

 9   III adult clinical trials, EB-1, and on this slide we

10   have -- it's a course slide that comes out of our

11   integrated summary of efficacy, but as you'll see, we

12   have the pathogen to the left, the treatment the patient

13   group     received,   the   clinical       outcome,   and     the

14   microbiologic outcome, and then finally the pathogen

15   outcome.

16                  And you'll see that for Strep. pneumoniae

17   we have about 90 percent, you know, pathogen outcome

18   cure.      If you look at Strep. pneumoniae that is

19   resistant, you see essentially the same number.               The

20   number of Strep. pneumo. species resistant to penicillin

21   in adults is only eight.

22                  Now, if I can go to the next slide, this

23   is that slide that I showed you just a few minutes ago

                             S A G CORP.
     202/797-2525            Washington, D.C.          Fax: 202/797-2525

 1   with linezolid Phase III and linezolid Phase II with

 2   the pediatric data removed.

 3                    And here you have for all patients with

 4   Strep. pneumo. -- we have a 92.7 percent cure rate, for

 5   intermediate Strep. pneumo. and for resistant Strep.

 6   pneumo.        This is bringing in all of our Phase II

 7   pneumonia trials.

 8                    So we have ten resistant Strep. pneumo.

 9   here.    We have 17 intermediate there, and we have 164

10   patients       with   Strep.   pneumo.        across   all   of    our

11   protocols.

12                    Now, if I can go to the next slide, now

13   looking at our bacteremia and our success rate, the most

14   fertile source of this data is in our in-patient

15   community acquired pneumonia trial.              We identified 29

16   patients with Strep. pneumo. with bacteremia, and we

17   have a 93.1 percent cure, and when you look at the

18   comparator, roseferin, you see a lower cure rate.

19                    And go to the next slide, please.

20                    We did have just a couple of people in our

21   out-patient pneumonia trial with Strep. pneumo., and

22   we had good cure rates with both agents.

23                    Now, if I can go to one more slide to give

                               S A G CORP.
     202/797-2525               Washington, D.C.            Fax: 202/797-2525

 1   you a little bit more feeling for the patients that

 2   failed     their   treatment    with     bacteremia,     only    one

 3   patient, and this is the 71 year old with chronic

 4   lymphocytic leukemia, profound immunosuppression and

 5   radiation.       That patient was treated for seven days with

 6   linezolid, was doing very, very well, went home, didn't

 7   complete his prescription, of course, went home, came

 8   back to the hospital in septic shock.

 9                    On reculturing his blood stream a couple

10   of weeks post discharge, he still had Strep. pneumo.,

11   and this is the patient I told you just a little bit

12   about.         We never got the second isolate to show

13   definitively that they are the same bug.            We don't know

14   whether this is a recurrence of the original Strep.

15   pneumo. infection or whether this is a new infection.

16                    So in conclusion, we have a solid database

17   with    Streptococcus     pneumoniae.        When   we    put    the

18   pediatric        experience    together      with     the     adult

19   experience, I think we have an excellent cure rate.

20   In the pediatric pneumonia trials, we had five patients

21   with Strep. pneumonia resistant to penicillin.                   Two

22   were bacteremic, and those in all the patients, whether

23   they were bacteremic or not, were clinical cures.

                               S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1                     The other point I'd make is that the biology

 2   and the natural history of pneumonia in children is

 3   equivalent to adults.               I mean obviously they have

 4   special        issues,     pharmacokinetic        issues    that    are

 5   critical, but we feel that the natural history of

 6   pneumonia in children is quite similar to adults and

 7   that you should consider the data that we're providing

 8   from the pediatric program and the adult program as well.

 9                     DR. MURRAY:        Would you show that first

10   slide again?

11                     DR. HAFKIN:       Yes, if we could go to the

12   first Strep. pneumoniae, yes, EB-1.

13                     CHAIRMAN RELLER:        Dr. Hafkin, are any of

14   the 32 patients with bacteremic pneumococcal pneumonia

15   --   were       any   of    those    with       resistant   isolates,

16   intermediate or highly resistant to penicillin?

17                     DR. HAFKIN:         There was no adult with

18   resistant Strep. pneumo. bacteremia.                 There were two

19   children that had excellent -- I mean they were cured.

20    We have two pediatric cures with resistant Strep.

21   pneumo.

22                     And to answer the question you asked us

23   earlier, both the children with bacteremia had Strep.

                                  S A G CORP.
     202/797-2525                 Washington, D.C.           Fax: 202/797-2525

 1   pneumo. resistant to penicillin and erythromycin in both

 2   cases.

 3                  CHAIRMAN RELLER:     Thank you.

 4                  Other questions related to Streptococcus

 5   pneumoniae     in   community     and     hospital      acquired

 6   pneumonia?

 7                  DR. KUEHNERT:    I just had    --

 8                  CHAIRMAN RELLER:     Dr. Kuehnert.

 9                  DR. KUEHNERT:    -- just a question, really

10   a clarification about your definition of hospital

11   acquired pneumonia versus community acquired pneumonia,

12   and just so that it has some bearing because I saw that

13   you had Strep. pneumo. as being a common pathogen, and

14   I don't really see that very often as a cause of hospital

15   acquired pneumonia.

16                  So if it was nursing homes that were

17   involved in that or hospitals?

18                  DR. HAFKIN:   It was a global study, and we

19   were quite surprised as well.             There were several

20   patients that came from an Eastern European site that

21   had very good evidence of Streptococcal pneumonia two

22   days after they had been in the hospital, three days

23   after they had been in the hospital for other diagnoses.

                            S A G CORP.
     202/797-2525           Washington, D.C.            Fax: 202/797-2525

 1    So it was interesting, and it's the first time I had

 2   seen that, as well.

 3                  Should    I     go    on     to    the    Staphylococcal

 4   questions you asked?

 5                  CHAIRMAN RELLER:              If there be no other

 6   questions for Streptococcus pneumoniae now, please go

 7   ahead.

 8                  DR. HAFKIN:          Again, these slides weren't

 9   prepared for this presentation.                  So they're a little

10   hard to read.    If we could go to EB-6, this has all of

11   our Staph. aureus that we identified and treated in skin

12   and soft tissue trials, all adult trials, and here we

13   have linezolid treatment outcomes across the top.                         We

14   have oxacillin outcomes there, and we have vancomycin

15   outcomes across the bottom.

16                  Recall    that         the        patients       who    were

17   randomized in the vancomycin arms were patients that

18   were     fundamentally        sicker        than        those    patients

19   identified and treated in our oxacillin treatment arm.

20                  Then if we look at methicillin resistant

21   Staph. aureus, you see a subset.                This group is a subset

22   of this group.      You'll see the outcomes here, the

23   clinical outcomes, the micro outcomes, the pathogen

                              S A G CORP.
     202/797-2525                 Washington, D.C.               Fax: 202/797-2525

 1   outcomes, and you'll have vancomycin down along the

 2   bottom.

 3                  Obviously we did identify two patients with

 4   MRSA randomized oxacillin, and one patient was a cure,

 5   at least clinically.

 6                  Now, if I could go after you've had a chance

 7   to study to the next slide, which is the analogous slide

 8   in pneumonias, if you look at Staph. aureus treated with

 9   linezolid in our pneumonias, this is all protocols, the

10   MRSA protocol, the nosocomial pneumonia protocol.           We

11   have the clinical outcomes here for all three agents.

12    Here we have the microbiologic outcome, and here we

13   have the pathogen outcome.

14                  Please, again, recall that the vancomycin

15   studies recruited much sicker patients.      So if we look

16   at linezolid versus vancomycin, here's their outcome.

17                  Now, if we could see the next slide, when

18   we look at Staphylococcal bacteremia, Staph. that came

19   from these trials, we did not identify very many people

20   with MRSA bacteremia.     We actually have an additional

21   story to share with you because there are in total,

22   taking this patient and putting them into our entire

23   database, we have 15 patients with MRSA bacteremia, but

                             S A G CORP.
     202/797-2525            Washington, D.C.      Fax: 202/797-2525

 1   the Staph. aureus sensitive to oxacillin are listed up

 2   here, and you can see very comparable results              for

 3   linezolid and oxacillin.

 4                  Now, if we consider the MRSA bacteremia that

 5   we identified across all of our protocols, we found 15

 6   patients with MRSA bacteremia either due to pneumonia

 7   or skin and soft tissue.

 8                  If I could have the next slide, this is the

 9   slide that gives you a flavor of each of the patients

10   who failed.     A 56 year old diabetic with an infected

11   line, never removed, on steroids, looked great after

12   15 days of therapy, had recurrent bacteremia 43 days

13   later.

14                  It was a bacteremia with Staph. aureus.

15   We don't know what the relationship of that recurrent

16   bacteremia to the original infection was.

17                  A 69 year old, profoundly immunosuppressed,

18   bacteremic.     They treated the patient only nine days

19   or, rather, they only treated the patient for five days,

20   and shortly after therapy, nine days after therapy, the

21   patient had recurrent bacteremia.          Why they treated

22   this patient for only five days is beyond me.

23                  A 69 year old with diabetes, media stinitis,

                             S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   post CABG, was profoundly ill.                 This patient had, it

 2   turns out later on subsequent questioning, external

 3   osteomyelitis.         They treated the patient for 24 days,

 4   and the patient had recurrent bacteremia, and this is,

 5   I think, a real failure, but the patient had a diagnosis

 6   that made them ineligible for the study.

 7                   Then this last one, an 81 year old treated

 8   for    six     days.       He     had    an     abscessed       kidney,

 9   Staphylococcal         abscess   because       of   the   obstructive

10   uropathy.      The patient was treated for six days and then

11   had recurrent bacteremia.

12                   So these are our failures associated with

13   MRSA that had recurrent Staphylococcal bacteremia post

14   therapy.

15                   Now, if I could go on to deal with another

16   question if there are no questions following this.

17                   CHAIRMAN RELLER:         Dr. Lowy.

18                   DR. LOWY:        Regarding the information you

19   just provided and also the recommendations that you have

20   on page 6 for length of therapy for complicated skin

21   and soft tissue infections, the recommendation is ten

22   to 14 days.      Many clinicians would be concerned about

23   treating       bacteremic         Staphylococcal          infections,

                                S A G CORP.
     202/797-2525                Washington, D.C.              Fax: 202/797-2525

 1   particularly if the bacteremia was prolonged for 14 days

 2   alone because of the high risk of either recurrence or

 3   metastatic seeding.

 4                  I was wondering whether you had had any

 5   additional considerations about that.

 6                  DR. HAFKIN:      The data that we have suggests

 7   the failures -- in fact, all the failures whether we

 8   look at Strep. pneumo. or Staph. aureus or MRSA or even

 9   -- well, we don't have any Streptococcal bacteremias

10   that failed, but all those failures seem to be associated

11   with failures of treatment less than seven or eight days.

12                  So I can't answer the question.            We've all

13   seen    failures   with   the    beta        lactams   under    those

14   circumstances, and we don't for a minute believe that

15   we're better than oxacillin or naphcillin against a

16   sensitive Staphylococcal species.                 We think we're

17   equivalent to the beta lactams against beta lactam

18   sensitive organisms.

19                  CHAIRMAN RELLER:        Dr. Norden.

20                  DR. NORDEN:       Yeah, Hafkin, I'm a little

21   confused by numbers at this point, and I think this is

22   me, but the last slide you showed showed four failures.

23                  DR. HAFKIN:      Yes.

                               S A G CORP.
     202/797-2525              Washington, D.C.             Fax: 202/797-2525

 1                  DR. NORDEN:       The slide before that I

 2   thought I saw only one MRSA.

 3                  DR. HAFKIN:    That's true because it was a

 4   different protocol.     Let's go to the slide.

 5                  DR. NORDEN:    I see.

 6                  DR. HAFKIN:   If we can bring this slide up,

 7   I thought I had made that point.           Perhaps I -- this is

 8   one of the 15 MRSA patients that I told you about.                I

 9   stole the patient off this slide and put it into that

10   group of 15.     When I told you 15, I mean 15 patients

11   with MRSA bacteremia from any kind of diagnosis, any

12   treatment anyplace.

13                  DR. NORDEN:   In both the pneumonia and the

14   skin and soft tissue.

15                  DR. HAFKIN:    Skin and soft tissue, yes.

16                  DR. NORDEN:    Thank you.

17                  DR. HAFKIN:   Next I'd like to deal with the

18   amino glycoside questions, and I believe ER-30 is the

19   slide I believe I want.        Yes.        If I could have this

20   slide.

21                  We looked because a few patients as we told

22   you in our MRSA program, Protocol 33, looking for

23   resistant Staph.    We allowed patients who needed to have

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   an amino glycoside to have it, and so we looked at

 2   outcomes of patients who got amino glycoside and

 3   linezolid,     and    those   patients        that   got     no   amino

 4   glycoside.

 5                  And    so   here's    linezolid       outcomes,      and

 6   here's the vancomycin outcomes.               This is the success

 7   of patients that got linezolid and amino glycoside.

 8   Here is the percentage of patients that got vancomycin

 9   and amino glycoside.

10                  It certainly didn't have any effect on

11   linezolid, didn't seem to improve a thing, but the

12   numbers are small, as you can see.              We only -- in the

13   entire     protocol   only    identified       these    handful     of

14   patients that were clinically evaluable.

15                  CHAIRMAN RELLER:           Any further questions

16   about this component of the Staphylococcal infection

17   group?

18                  DR.    HAFKIN:        Now,     you    asked    a    very

19   important question, I thought, about the relationship

20   of linezolid sensitivity to Strep. pneumonia that might

21   be resistant, rather, to penicillin and erythromycin,

22   and although we don't have -- we didn't have enough in

23   the way of isolates in our clinical trial, we did have

                                S A G CORP.
     202/797-2525               Washington, D.C.            Fax: 202/797-2525

 1   -- we've taken the opportunity to look at that in the

 2   Sentry database isolates, and we could share that

 3   information    that   Dr.    Zurenko       has   done     if    you're

 4   interested in seeing the relationship of linezolid

 5   sensitivity to these organisms.

 6                  CHAIRMAN RELLER:      Anyone not want to see

 7   it?

 8                  (Laughter.)

 9                  CHAIRMAN RELLER:      Please.

10                  DR. HAFKIN:    Well, then I'd like to call

11   Gary Zurenko to the podium to show his work.

12                  DR. ZURENKO:     Thank you for the vote of

13   confidence.

14                  (Laughter.)

15                  DR. ZURENKO:    Slide Y-129, please.

16                  I'm Gary Zurenko from Discovery Research.

17                  And early in our evaluations of linezolid

18   we were very interested in, of course, the erm genes,

19   which are shown here for the MLS B phenotype.                  This is

20   a study using isogeneic strains in most cases, some

21   transconjugants, some transconductants, and as you can

22   see, for ermA, ermC, ermB, ermTR, most specific interest

23   here is the ermA.     The MIC of linezolid was unaffected

                             S A G CORP.
     202/797-2525            Washington, D.C.                Fax: 202/797-2525

 1   by these resistance genes.

 2                   Y-130, please.        Slide up.

 3                   Also       looking      at      macrolide         efflux

 4   specifically     here      in   estimoniae,       the     MEF   E    gene

 5   (phonetic), we saw virtually no difference in MIC.

 6   Therefore, the data predicted that we would not have,

 7   in effect, by the common erythromycin resistance genes

 8   on linezolid activity, which is compatible with the

 9   mechanism of action being distinct.

10                   Looking at the overall century database,

11   we did several correlations, one with erythromycin

12   versus linezolid, Y-270.             Slide up, please.          This is

13   just an X-Y scattergram of linezolid on this axis versus

14   erythromycin across the bottom, MIC versus MIC with

15   virtually no correlation.

16                   So hopefully that would convince all of us

17   that based on at least laboratory evaluations, we would

18   not    expect    any    cross     resistance       to     occur      with

19   erythromycin.

20                   CHAIRMAN RELLER:         Thanks, Gary.

21                   Do   you    have,     Dr.      Tarpley,     additional

22   information for us about the place of acquisition and

23   severity of skin and skin structure infections with the

                                 S A G CORP.
     202/797-2525                Washington, D.C.              Fax: 202/797-2525

 1   methicillin resistant Staphylococci?

 2                    And then secondly, further information

 3   about the species of enterococci in the vancomycin

 4   resistent enterococcus protocol 54 and 54(a)?

 5                    DR. HAFKIN:       We're pulling that slide up.

 6                    No, no, the complex skin soft tissue in 31.

 7                    What we're going to show is -- actually I

 8   didn't identify the slide very well.                  Yes, please, if

 9   you'd bring ER-32 up, this is the clinical outcome for

10   31, and these are the sickest of the patients.

11                    Recall that we identified in the 31 trial

12   230 patients with skin and soft tissue infection, and

13   then what we did in this analysis is we looked at everyone

14   for evidence of fever and high white count, and then

15   we   said      who   of   those   patients      had    a   significant

16   comorbidity, and you'll have that patient observation

17   here.

18                    So you see that in linezolid we were able

19   to identify 33 patients out of the 115, and then in the

20   vancomycin, we had about 30.             So these are the sickest

21   of our MRSA protocols by diagnosis.

22                    Most of these patients had very extreme

23   illness, but as you can see, many of them did have

                                 S A G CORP.
     202/797-2525                 Washington, D.C.              Fax: 202/797-2525

 1   infected surgical incisions in both arms.

 2                    This infected wound cellulitis, these could

 3   have been associated, of course, with central lines.

 4    If you look at the kind of illness that you collect

 5   in MRSA trials, looking at skin and soft tissue, it is

 6   quite often associated with an implanted device, but

 7   certain abscesses, cellulitis are rarely.

 8                    Now, this is the tip of the iceberg.       These

 9   are the sickest of the patients we have.           The rest of

10   the 230 patients -- I guess it would be about 180 patients

11   or something like that -- the rest of those patients

12   will have had less severity of their MRSA illness.

13                    Let's go to the next slide, which shows you

14   a slightly different cut of the data.            The other was

15   the ITT.       This is the clinical evaluable patients, and

16   we fall then into only 28 patients here and 27 patients

17   in vancomycin.

18                    But, again, give you the sense that no

19   matter what cut of the data we have, we have that same

20   assurance of comparable outcomes.

21                    CHAIRMAN RELLER:        You had a total of 15

22   patients       with   methicillin    resistant   Staph.    aureus

23   bacteremia.       Were most of them out of these patients?

                                S A G CORP.
     202/797-2525               Washington, D.C.        Fax: 202/797-2525

 1                    DR. HAFKIN:        They came primarily from

 2   Protocol 31, and they came primarily from the patients

 3   you're seeing here, yes.

 4                    CHAIRMAN RELLER:       Dr. Chesney.

 5                    DR. CHESNEY:      I wanted to go back to Group

 6   A strep. if I could, and looking at the information that

 7   the FDA presented this morning, on page 22, there were

 8   five    uncomplicated      strep.    infections      treated,     100

 9   percent cure.

10                    On page 25, again, uncomplicated, there

11   were seven with an 85 percent cure rate, and then on

12   page 29, there were 26 complicated skin and soft tissue

13   with only a 69 percent cure rate, and I guess my instincts

14   would have said that they should have been 100 percent

15   across the board, and I'm wondering why 69 percent, why

16   not 100 percent across the board.

17                    DR. HAFKIN:     Yes, yes.       As you'll recall,

18   our numbers are similar, but because the FDA uses

19   slightly       different   rules     for      evaluability,    their

20   numbers -- you know, they're very consistent.

21                    We've tried to understand what happened in

22   these Group A strep. infections, and unfortunately the

23   physicians that called these patients failures did not

                                S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1   give us enough information about the clinical basis of

 2   their failure.

 3                   They were not microbiologic failures in

 4   general.       In other words, they had Group A strep. at

 5   baseline.      So they had the bug there, and it was the

 6   pathogen that was important.           Unfortunately, because

 7   there were so few, we have only one with a positive

 8   culture at follow-up, and that patient was clinically

 9   cured, but was a microbiologic failure.

10                   So we really don't know what's happening.

11    Did we pick up a group of patients that failed

12   clinically whose Group A strep. was cured?         I'm afraid

13   that we're only talking about a handful of observations,

14   and I can't -- you know, I don't understand it.              You

15   know, I can't tell you a story that somehow puts it in

16   perspective.

17                   DR. CHESNEY:    Well, obviously if this were

18   to be approved for skin and soft tissue, people would

19   assume that it was effective for Group A strep, and this

20   is a little unnerving to --

21                   DR. HAFKIN:    Well, we have a substantial

22   animal model database, and if you'd like us to go into

23   that.

                              S A G CORP.
     202/797-2525             Washington, D.C.        Fax: 202/797-2525

 1                   DR. CHESNEY:      Thank you.         That would be

 2   good.

 3                   DR. HAFKIN:     Before we go too far, would

 4   you like me to give you the enterococcal data?               Because

 5   it involves another group of people.

 6                   CHAIRMAN RELLER:        Let's finish with the

 7   Group A Streptococcal question so that then we can keep

 8   these in categories, and then we'll come back to the

 9   enterococci.

10                   DR. HAFKIN:     Great.        I'll call one of my

11   preclinical colleagues.

12                   CHAIRMAN RELLER:         Gary Zurenko will be

13   presenting this data.

14                   DR. ZURENKO:     Yes.        Thank you.

15                   CHAIRMAN RELLER:       These data.

16                   DR. ZURENKO:        The in vivo activity of

17   linezolid was evaluated in the Myan Acrusis (phonetic)

18   model    by    Dennis   Stephens,     and      in   this   model    of

19   necrotizing fascitis mice were infected.                   Mice were

20   infected as shown.         Treatment with antibiotics was

21   initiated four hours after the challenge and then

22   continued every 12 hours BID for six doses.

23                   Next slide.

                               S A G CORP.
     202/797-2525              Washington, D.C.             Fax: 202/797-2525

 1                  Linezolid was administered in three doses,

 2   ten, 20, and 40 milligrams per kilo.             Clindamycin was

 3   administered at 86 milligrams per kilo and penicillin

 4   G at 98.

 5                  The animals were followed as described

 6   here, and the endpoints were considered as shown.

 7                  Next slide.

 8                  Let's go to -- let's jump right to the end.

 9    Y-198 please.     Slide up.

10                  In this figure we show that percent survival

11   post treatment at 12 days with the antibiotic dose across

12   the bottom, as you can see, very good survival in these

13   treatment groups.     This is a group treated with ten to

14   the seventh, ten to the eighth, and ten to the ninth

15   cells.     Obviously there's very severe challenge as is

16   shown here.     The drug is not as effective as with the

17   other two challenges, which are, in fact, still quite

18   large.

19                  The conclusion of the investigator was that

20   the activity was quite similar to that seen with

21   clindamycin, which is a very effective agent in this

22   model, but that he felt that a longer term of dosing

23   might     be   required   to    handle       these   very    severe

                               S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1   challenges.

 2                  CHAIRMAN RELLER:      Dr. Murray.

 3                  DR. MURRAY:     What's the half-life of the

 4   drug in mice?

 5                  DR. TARPLEY:    One moment please.

 6                  DR. LOWY:     I think it was between a half

 7   hour and an hour when I read this over before, as opposed

 8   to penicillin in 15 minutes and clindamycin about 20

 9   minutes in mice.

10                  DR. TARPLEY:    It's about an hour.

11                  CHAIRMAN RELLER:      Thank you.

12                  Dr. Rodvold.

13                  DR. RODVOLD:    I was wondering if you could

14   maybe clarify something with me for the methicillin

15   resistant Staph. aureus.      It seems in your protocol for

16   uncomplicated skin and skin structure there wasn't that

17   many isolates, if any, for MRSA, and then the other

18   isolate you've tied in from both your complicated

19   studies as well as just the MRSA directed pathogen study.

20                  My question gets to be that in uncomplicated

21   you studied a 400 milligram dose and in complicated you

22   studied a 600 milligram dose, but you want indication

23   for MRSA in both.    Is the 400 milligram dose in an MRSA

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   adequate enough to treat it?

 2                  DR. HAFKIN:    We agree that patients with

 3   MRSA, known MRSA infection should have 600 milligrams

 4   of linezolid twice daily.       We agree with you.

 5                  CHAIRMAN RELLER:      I think we're ready for

 6   the enterococcal discussion.

 7                  DR. HAFKIN:   You know, you make a wonderful

 8   slide and you don't bring it with you.        So we're going

 9   to have to live through a less than perfect collection.

10    If I could have ER-63 up, this is not the complete

11   database unfortunately, this is just that initial

12   database.

13                  ER-63, please.

14                  What I'm going to try and share with you

15   is the efficacy of linezolid against VRE by species.

16    If I could have this slide up, please.

17                  Unfortunately this does not include our

18   compassionate use and our 54 isolates, but this gives

19   you a sense of that first 145 patients where we have

20   E. faecium, E. faecalis, and E. avium, rather, faecalis

21   and faecium.

22                  And we have the microbiologic response here

23   and the microbiologic response here, this being the 200

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   milligram, this being the 600.

 2                    We've just found the other slides, but let

 3   me go through one more of these, and then we'll find

 4   the real slide.

 5                    Let me see it now and see if we have it.

 6    Yes, this is a great bottom line slide.                M-93, please.

 7                    Now, this is -- what we've done here is I've

 8   taken the 600 milligrams twice daily dose results for

 9   Study 54(a) and the results in our compassionate use

10   trial.         Recall   that   most    of      these   patients        have

11   interabdominal abscess.          More than 90 percent of those

12   patients have interabdominal abscess.                  This is the E.

13   faecalis result for the two.

14                    And then our conclusion, of course, is that

15   the drug is quite effective in the management of

16   vancomycin resistant enterococcus whether it's faecium

17   or faecalis.

18                    CHAIRMAN      RELLER:         Just    so    that    we're

19   perfectly clear, the number of vancomycin resistant

20   faecalis isolates treated was two?

21                    DR. HAFKIN:      Well, two, and this is half,

22   you know; this is half of our 50 -- the two represents

23   half of the experience I'm sharing with you.                  This comes

                                 S A G CORP.
     202/797-2525                Washington, D.C.                Fax: 202/797-2525

 1   from the 54(a), and this is from compassionate use.

 2   These     are     patients     that    are   microbiologically

 3   evaluable.       So we have nine from compassionate use, and

 4   we have two from our Study 54A.

 5                    CHAIRMAN RELLER:      So a total of 11 patients

 6   with vancomycin resistant E. faecalis?

 7                    DR. HAFKIN:    Yes.

 8                    CHAIRMAN RELLER:      Dr. Murray.

 9                    DR. MURRAY:    Just along those lines, it's

10   kind of interesting since there probably would have been

11   a comparator unless these were allergic patients for

12   the vancomycin resistent E. faecalis.              I'm kind of

13   surprised it was included in this protocol.

14                    Is that maybe they didn't know what it was

15   when they first started therapy?

16                    DR. HAFKIN:    Actually in Protocol 25, 70

17   percent of the patients roughly came to the protocol

18   because there was no option in terms of therapeutic

19   choices.       The patient had no other antibiotic that would

20   work, but the rest -- it is interesting -- came because

21   of intolerance, allergy to beta lactams, allergy to

22   vancomycin.       So these few were allergic patients.          They

23   couldn't take something else.

                               S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1                  CHAIRMAN RELLER:        Are there additional

 2   questions for the sponsor?

 3                  Dr. Murray.

 4                  DR. MURRAY:   A couple of questions sort of

 5   going back to early clinical data, and one relates to

 6   just partly curiosity, but it would, I'm sure, influence

 7   the labeling, is with the absorption of the drug orally,

 8   have very sick patients been looked at?

 9                  I mean the question for the clinician in

10   the hospital always is:         is your intubated patient

11   that's really, really sick able to absorb an oral drug?

12    And usually we just say no and give them intravenous,

13   but do you have data on that?

14                  I know data represented on the food, but

15   I don't remember if things like tube feedings were looked

16   at or antacids, calcium, the usual things that have

17   influenced other drugs, and how strong are the data in

18   those subsets?

19                  And then the other question relates to that

20   subset of individuals that seems to have low blood

21   levels, and I'm curious.       Do you have any information

22   on what's going on with those people?          Are they --

23   presumably they're different metabolically.          Is there

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   concern about patients?

 2                  How low can the levels be in somebody, and

 3   is there going to be a subset of patients that needs

 4   to be identified with MRSA bacteremia, for example, that

 5   would either need to have blood levels determined?            How

 6   confident are you that everybody that's sick enough is

 7   really going to have high enough levels?

 8                  Because according to the FDA analysis there

 9   are individuals that have quite low levels, relatively

10   low levels.

11                  DR. TARPLEY:     Dr. Jungbluth will --

12                  DR. MURRAY:    And any reason to think that

13   these, if they are hypermetabolizers, are they the

14   subset that either fail or that get into toxicity

15   problems, LFTs or whatever?

16                  DR. TARPLEY:     Let us organize a response

17   because there were at least three questions there.

18                  DR. MURRAY:    Yes.

19                  DR. TARPLEY:     Thank you.

20                  DR. JUNGBLUTH:    Okay.     I think part of your

21   first question was the oral absorption in sick patients,

22   and we don't have an exact bioavailability study

23   conducted in somebody like this, but we have a large

                             S A G CORP.
     202/797-2525            Washington, D.C.          Fax: 202/797-2525

 1   patient population.    We've done pharmacokinetics in our

 2   Phase II studies, and there's over 600 patients included

 3   in that, and they would have received IV and oral.

 4                  So we think that we have some pretty good

 5   data there that would suggest absorption is adequate,

 6   and the kinetics in these patients are similar to the

 7   profiles we have seen in the healthy volunteer data.

 8                  And did you have a second part to the

 9   absorption question?

10                  DR. MURRAY:      Any other specifics, two

11   feedings, antacids, acid, anything --

12                  DR. JUNGBLUTH:     No, nothing specific like

13   that.

14                  DR. HAFKIN:   To try and further answer that

15   question, we could go through our analysis of response

16   by APACHE score.       It's very similar to Dr. Ross'

17   analysis.      We also have the analysis of patients on

18   ventilators, and if you look at those lists of the

19   patients that got the drug under controlled Phase III

20   trial, there is not a fall off of efficacy.

21                  So although the clinical database is not

22   huge because we only have 2,000 patients in our database,

23   at least for that very sick subset of patients, we have

                            S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   good outcomes.

 2                  CHAIRMAN RELLER:      Dr. Leggett.

 3                  DR. LEGGETT:     Yeah, I think there was a

 4   third part to Dr. Murray's question, and then I had a

 5   follow-up.

 6                  DR. TARPLEY:    I'm sorry.      Could you repeat

 7   the third part, please?

 8                  DR. MURRAY:    The subset or the individuals

 9   that -- there seems to be a wide splay in the blood

10   levels, and there are some individuals that really fall

11   at a low range of blood levels.            Do you know anything

12   more about those?    Do you have reason for concern that

13   in very sick patients blood levels -- in some patients

14   blood levels should be obtained because there may be

15   an identifiable subset that has low levels that needs

16   more drug?

17                  DR. TARPLEY:   Yes.     The data that we showed

18   today were at a fixed dose of 400 or 600 milligrams twice

19   a day, and you saw that we were able to demonstrate very

20   good efficacy.    So we do not feel that it's necessary

21   to monitor blood levels.       You've seen the data in the

22   sickest patients, as well, and the efficacy is quite,

23   quite good.

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1                   CHAIRMAN    RELLER:           Dr.    Wittes   had     a

 2   question.

 3                   DR. WITTES:      Yeah, I want to circle back

 4   to 54 and 54(a), to the accounting.             This is a two-part

 5   question.

 6                   The first part of it is sort of a confirm

 7   and the second is a question.                 I count 331 patients

 8   randomized in all, of whom 145 were in 54(a), 82 in 54,

 9   and the remainder still on the study.               I assume the 145

10   in 54(a) were the first 145 randomized.                 I just want

11   to confirm that.

12                   And the second is the 82 in 54, are they

13   the next 82 randomized?         How did -- when the FDA asked

14   for them, how did they specify which ones were to be

15   looked at?

16                   And let me tell you the drift of the

17   question.      The question is that as we're comparing the

18   200 to 600, is there a potential for bias in the

19   information that we're seeing.

20                   DR. TARPLEY:      Dr. Oliphant will start the

21   response, and we'll see if we can satisfy your question

22                   DR. OLIPHANT:       Slide up, please.

23                   First of all, I should say that Dr. Ross'

                                S A G CORP.
     202/797-2525               Washington, D.C.            Fax: 202/797-2525

 1   slide which demonstrated the history of these studies

 2   was quite accurate, and      I think that told part of the

 3   story.     This slide here may tell the rest hopefully.

 4    This is the key activities time line for Study 54(a).

 5                  On June 20th, Study 54(a) was closed to

 6   enrollment, and then July 22nd was our cutoff date for

 7   the case report forms that we were able to retrieve for

 8   completed patients, and that resulted in 79 patients

 9   in the 600 milligram group and 66 patients in the 200

10   milligram group.

11                  DR. WITTES:    Can I clarify something?            Is

12   this -- are those the first 79 and first 66 randomized?

13                  DR. OLIPHANT:     Not necessarily, no.

14                  DR. WITTES:    Okay.

15                  DR. OLIPHANT:      And then on August 3rd,

16   1999, we had a teleconference with the FDA to discuss

17   the possibility of submitting the results of these 145

18   patients.

19                  Subsequently, on August 19th, the database

20   was closed for Study 54(a).                Treatment codes were

21   unblinded, and then on August 30th, we did generate and

22   present unblinded results.

23                  Now, may I have the slide for Study 54?

                            S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   Slide up, please.

 2                  Okay.    Now, Study 54 then, which was in

 3   effect a continuation, but was considered a separate

 4   study, as you see, it began enrollment officially on

 5   June 20th, which was the date of cutoff for Study 54(a).

 6                  Now, there were as you see in the footnotes

 7   some patients included in Study 54 who were enrolled

 8   prior to June 20th because their case report form data

 9   were not available at the time that we presented results

10   for Study 54(a).

11                  And December 27th, Study 54 was closed to

12   enrollment.    At that time we had the 186 patients, and

13   in late December actually it was that we presented the

14   data for those 82 patients to the FDA.

15                  DR. WITTES:        And then does that mean,

16   again, that those 82 are not necessarily the next

17   consecutive group?      Is that right?

18                  DR. OLIPHANT:      I believe that those 82 are

19   the next consecutive group, yes.

20                  DR. WITTES:       I guess I'm confused.        If

21   those are -- so that you determined by the 82 by date

22   of randomization, but the 54(a) by date of case report

23   form collection?       Is that right?

                              S A G CORP.
     202/797-2525              Washington, D.C.       Fax: 202/797-2525

 1                    DR. OLIPHANT:         Well, if we can go back to

 2   the previous slide, the 54(a) time line, slide on,

 3   please.        No, that's still 54.               One, twenty-seven, I

 4   think.     One, twenty-two?           Yeah, slide on.

 5                    Okay.    On June 20th, Study 54(a) was closed

 6   to enrollment.       So patients who were enrolled prior to

 7   June 20th and as the second bullet point indicates,

 8   patients for whom we were able to retrieve case report

 9   form    data     from    the     field      were     included   in    our

10   presentation of results for 54(a).                      So that's 145

11   patients.

12                    DR. WITTES:          Okay.       Again, you retrieve

13   when the entire -- that patient is finished, right?

14                    DR. OLIPHANT:          Correct.

15                    DR. WITTES:        So if somebody were recruited

16   sort of late, like June 15th, and were still being

17   treated in July, I mean, sort of what I'm trying to get

18   at, it seems to me there's a potential for a length bias

19   in the last group of patients captured by the case report

20   forms, and I think we need a handle of this in terms

21   of a handle for the magnitude of what this could be in

22   terms of looking at the comparison between the two doses.

23                    Because I mean, we were trying to look at

                                S A G CORP.
     202/797-2525                   Washington, D.C.           Fax: 202/797-2525

 1   superiority in that, and therefore, if there's a

 2   potential for a bias in ascertainment, that would be

 3   difficult.

 4                     And then the other question is if you

 5   couldn't get the sequential ones for 54(a), why could

 6   you get them for 54?

 7                     DR.   HAFKIN:         Can      I   make      a    clinical

 8   observation on the CRF availability?

 9                     The patients with the longest follow-up are

10   the patients with urinary tract infection, and because

11   we demand six weeks of follow-up data.                    So there was,

12   in essence, a bias against, in fact still going on in

13   a sense, against urinary tract infection experience

14   because the patient gets the same standard length of

15   therapy,        but   their    follow-up         goes    on      for    weeks

16   afterwards, whereas if you had the germ in the blood

17   stream     or    interabdominal        abscess,         your     acceptable

18   follow-up period would be two weeks to three weeks.

19                     The majority of patients that were pushed

20   from 54(a) into 54 were urinary tract infections, and

21   the majority of the patients that we have out in the

22   field that you haven't seen from 54, the little interim

23   analysis study, those will also be, in general -- there's

                                   S A G CORP.
     202/797-2525                  Washington, D.C.                 Fax: 202/797-2525

 1   a greater chance that they will be patients with urinary

 2   tract infection.

 3                  It is true that we wanted an experience with

 4   urinary tract infection, but at the present time, we're

 5   not actually asking the committee to judge where the

 6   drug is efficacious against urinary tract infection.

 7                  DR. OLIPHANT:      May I just add one more

 8   point, Dr. Wittes?     I believe I did misunderstand the

 9   second part of your question.        Those 82 patients in 54

10   are not necessarily sequential, as well.           We were asked

11   by the agency late last year for additional information

12   on patients we had available in that study.             So we had

13   to cut off the database at that time.          So those represent

14   the 82 patients for whom we were able to get information

15   at the time.

16                  DR. WITTES:    Thank you.

17                  CHAIRMAN RELLER:      Dr. Chesney.

18                  DR. CHESNEY:     This is totally switching

19   gears, and I hope I'm remembering this correctly.                 In

20   terms of adverse events, in the animal models there were

21   some    testicular   changes.       Am     I   remembering     that

22   correctly?     And we're not worried about that in humans?

23                  DR. TARPLEY:     No, we are not.        Dr. Petry

                             S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   can respond to that question from our Toxicology Group.

 2                  DR. PETRY:   Actually we had some fertility

 3   changes in the male reproductive tox. studies, but we

 4   do not have testicular lesions.            Rather what we have

 5   is a lesion in the epididymis, which as I'm sure you're

 6   aware is the site of sperm maturation.            That lesion

 7   requires a relatively long onset.

 8                  Actually I believe I've got a slide perhaps

 9   that would speak to that.       Could I see T-195, please?

10    Slide on, please.

11                  We, in fact, required seven to nine weeks

12   of administration, and it was completely reversible

13   following another seven to nine week period of recovery.

14    The time course of both the onset and the resolution

15   of the effect correlated very well with this epididymal

16   ductal epithelial hyperplasia.

17                  Again, as I indicated, this is a site of

18   sperm maturation.     There was, in fact, no testicular

19   lesion observed in these studies.            We looked to see

20   whether there might be some sort of a plasma testosterone

21   phenomenon going on here because the epididymis is a

22   testosterone dependent tissue, but we did not see that.

23                  Additionally, we had no lesions in the

                             S A G CORP.
     202/797-2525            Washington, D.C.          Fax: 202/797-2525

 1   canine reproductive tract.          So on balance, we don't feel

 2   that there's going to be a risk for male reproductive

 3   issues associated with the compound.

 4                  DR. CHESNEY:          You don't have any human

 5   studies though?      I mean, we're extrapolating from the

 6   rat, and that's probably fine.                  I just want to be

 7   reassured.

 8                  DR. PETRY:       I have no human data to offer.

 9    You're correct.

10                  CHAIRMAN RELLER:          Dr. Soper.

11                  DR. SOPER:      As a follow-up to that, I notice

12   that you relate mild fetal toxicity in rats at levels

13   that are consistent with what would be used in humans,

14   and that this was post implantation loss, and two out

15   of the three women that were pregnant and were exposed

16   to this agent suffered a miscarriage.

17                  You relate the similarities of some of these

18   findings to drugs like tetracyclines and quinalones,

19   which we don't really use during pregnancy.                    Is this

20   an agent that we need to avoid during pregnancy?

21                  And   what     category         drug   will   this    be,

22   pregnancy category?

23                  DR. HAFKIN:       I think I'll ask my colleagues

                             S A G CORP.
     202/797-2525                Washington, D.C.             Fax: 202/797-2525

 1   from tox. to come, but it's our feeling that our drug

 2   should be Category C.

 3                  In terms of the rate of miscarriage, you're

 4   quite right.     We had two miscarriages in very early

 5   pregnancy in our clinical trials, but that's consistent

 6   with spontaneous miscarriage in healthy young women.


 8                  We have literature.         We basically -- there

 9   have been a lot of reports of 50 percent, and that's

10   basically what we're carrying forward.

11                  DR. SOPER:    The miscarriage rate is about

12   20 percent of clinical pregnancies, which is what you

13   could relate to this observation.            So --

14                  DR. HAFKIN:    Well, let me make the point.

15    Everybody was pregnancy tested negative at baseline.

16    So these were not pregnancies diagnosable at baseline.

17                  DR. PETRY:     Could I actually see Slide

18   T-206, please?    Slide on, please.

19                  This does speak to what we've seen in our

20   peri and postnatal development studies.          There is a mild

21   and it's a reversible decrease in the offspring body

22   weight that classifies as fetal toxicity as these

23   animals continue that resolves.            There is an effect on

                             S A G CORP.
     202/797-2525            Washington, D.C.             Fax: 202/797-2525

 1   pre-implantation or peri-implantation loss.                   That's

 2   not necessarily -- we've looked at this now quite in

 3   depth, and we don't see that when we rebreed those males

 4   at 141 days.

 5                   It is a very mild effect in the fertility

 6   there.     In contrast or by comparison, in a number of

 7   labs it would be considered within the historical range

 8   of controls, but because we've got a new class of agent

 9   here, we look at these very carefully.

10                   Additionally      there      are     some     slight

11   reductions in the pregnancy rates coming out of these

12   animals.       About 12 percent of animals didn't sire a

13   litter in either of the breedings.

14                   CHAIRMAN RELLER:       Dr. Drayton.

15                   DR. DRAYTON:      I may be incorrect, but I

16   think     I    recall   reading      that    there     were      also

17   developmental problems in the offspring, and I wondered

18   whether they were physical developmental problems.                  Is

19   that of concern in pediatric population?

20                   DR. PETRY:    We don't think that it is, no.

21    The developmental delays that we see are sort of, again,

22   a mild decrease in initial body weight, which resolves

23   over time.      There was a slight delay in pena detachment,

                              S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1   which I don't see as a clinically significant issue.

 2                  CHAIRMAN RELLER:       Dr. Lowy.

 3                  DR.   LOWY:     Regarding    adverse    effects,

 4   diarrhea was one of the more common findings in the

 5   clinical cases.       Is there any information on the

 6   incidence of pseudomembranous colitis in the population

 7   of individuals who developed diarrhea?

 8                  DR. HAFKIN:    Yes, we've done analysis.         If

 9   I could ask you to look at S-169, yes, if you'd show

10   that slide.

11                  There was very little clostridium difficile

12   diarrhea that was actually diagnosed in the study.

13   Certainly there's no signal of increased risk with

14   linezolid therapy.

15                  CHAIRMAN RELLER:       For a drug that doesn't

16   get into the GI tract, why do you think this might be?

17    Is it a motility issue?        What's going on?

18                  DR. HAFKIN:     We really don't know.         We'd

19   have to speculate.

20                  CHAIRMAN RELLER:       Dr. Leggett.

21                  DR. LEGGETT:        Continuing with the VRE

22   question, given an underpowered study looking at 200

23   and 600 for most of the things that were looked at, I

                             S A G CORP.
     202/797-2525             Washington, D.C.         Fax: 202/797-2525

 1   have a hard time understanding if 600 equals 200 equals

 2   placebo or placebo is worse than 200 is worse than 600.

 3                     In looking at ways of trying to get around

 4   that, I guess Dr. Chikami in the morning talked about

 5   historical data, and I was wondering if anybody knew

 6   of   any       sort   of   maybe    comparable,   if   not   exactly

 7   equivalent patient populations with other drugs against

 8   VRE.

 9                     And the second comment, trying to get back

10   to Dr. Murray's question, in terms of responses, is it

11   possible that the people with the bad diarrhea did not

12   absorb drug as well?               Can you correlate folks with

13   increased incidence of diarrhea with those who failed?

14                     And do you have any drug levels in other

15   than volunteers?           Is there any way to look at kinetic

16   data combining 600 and 200 and trying to correlate that

17   with ACU above MIC or time above MIC so that the folks

18   with 200 who passed had better drug absorption, better

19   drug levels than the folks with 600 who had terrible

20   levels or the folks with 200 who had terrible levels?

21                     So of a lot of questions, but all trying

22   to get at another way to maybe convince us that this

23   drug is efficacious.

                                  S A G CORP.
     202/797-2525                  Washington, D.C.         Fax: 202/797-2525

 1                     DR. HAFKIN:       I hear the three questions,

 2   and let me try and answer the first, and then we'll try

 3   and approach the second.

 4                     The    estimates       for      mortality      in    VRE

 5   infection, enterococcal infection has been estimated

 6   by many people, and in fact, Barbara Murray just recently

 7   published in the New England Journal a review suggesting

 8   mortality of I believe it was 30 percent, but it may

 9   have been 40 percent in her review of the literature.

10                     Other authors have suggested that mortality

11   rates of 40 percent would commonly be seen, and indeed,

12   there is additional data in compassionate use trials,

13   such as ours which suggests that success rates of drugs

14   that have active efficacy or active activity against

15   VRE may actually be in the 60 percent success rate.

16                     So in terms of historical controls for all

17   comer kind of populations, another drug recently given

18   indication for VRE infection, I believe, had success

19   rates in similar protocols of 60, 65 percent.                    I'm not

20   an expert with that agent, but mortality rates have been

21   estimated to be 30 to 40 percent in multiple historical

22   papers.        So I think that's one way.

23                     Now,   in    terms     of      trying   to   help    you

                                   S A G CORP.
     202/797-2525                  Washington, D.C.             Fax: 202/797-2525

 1   understand the work we've done in terms of relationship

 2   between dosing and outcome, I'm going to ask my colleague

 3   from Pharmacokinetics, Ed Antal, to speak to that

 4   question.

 5                  DR. LEGGETT:    Can I just jump in?         I'm not

 6   quite sure I understand.       If you said the traditional

 7   historical mortality rate, is that with treatment or

 8   without treatment of 30 percent?

 9                  DR. HAFKIN:    Well, it depends on what kind

10   of -- it is without treatment.        Those are the estimates

11   for without treatment.

12                  DR. LEGGETT:    Okay.       Because what I see is

13   a success rate of 67 percent, which sounds like 70

14   percent, which leaves 30 percent mortality, and that's

15   with treatment, which is the same as without treatment.

16                  That's the trouble I'm having.

17                  DR. HAFKIN:     Sure, and it depends on all

18   comers and population ranks.         Agreed.

19                  DR. ANTAL:     If I understood your question

20   correctly about the relationship between, say, response

21   or side effects in concentration or kinetics, we've

22   looked at the Phase II databases, the database that we

23   had     accessible   that     had     the     most     population

                            S A G CORP.
     202/797-2525            Washington, D.C.             Fax: 202/797-2525

 1   pharmacokinetic sense.            We use that as our basis.

 2                        We did see a relationship between AUC over

 3   MIC versus pharmacokinetic parameters.                  It was by no

 4   means 100 percent predictable.                   The relationship is

 5   there.      There higher doses do have the higher tendency.

 6                        We looked to see if there was a rationale

 7   that one could say, for example, you know, a certain

 8   cutoff point that you wanted to be above or things of

 9   that nature.          We didn't see it because I think that a

10   lot    of      the    failures    that      you're   looking   at    are

11   multi-factorial reasons.             It's just not a question of

12   a certain level of drug, but it's a lot of other things,

13   too.    So that's where we stand with our analysis right

14   now.

15                        So is there a specific other question you

16   have?

17                        DR. LEGGETT:    No, I was just thinking one

18   thing.      If you could get that subpopulation, maybe a

19   small group of folks where you actually had VRE grow

20   in the blood, you might be able to try to get a handle

21   on it with a very small, hopefully a much smaller number

22   of patient.s

23                        DR. ANTAL:     Sure.

                                  S A G CORP.
     202/797-2525                  Washington, D.C.           Fax: 202/797-2525

 1                     DR. HAFKIN:       Sure.

 2                     DR. ANTAL:      Your other question about the

 3   safety,        gastrointestinal,        we       saw   no   relationship

 4   between a low blood level and a higher incidence of

 5   gastric upset.

 6                     DR. HAFKIN:       If I can jump in in one more

 7   way,    there     is   a   difference            between    success     and

 8   mortality, and even in the low dose, we don't have

 9   mortality rates of 30 percent, and in fact, Dr. Ross'

10   mortality analysis is particularly potent in that

11   regard.

12                     I don't remember David what slide you had

13   when you looked at the VRE study from the point of view

14   of mortality.

15                     DR. LEGGETT:       It's page 14 of 18 of the FDA

16   packet that we got put out, Table 3-10.

17                     DR. ROSS:     If you'd like, we can bring that

18   slide up.

19                     DR. CHIKAMI:         Dr. Reller, may I make a

20   comment?

21                     CHAIRMAN RELLER:           Please.

22                     DR. CHIKAMI:        In regards to the issue of

23   trying to get a handle on historical information in this

                                S A G CORP.
     202/797-2525                  Washington, D.C.              Fax: 202/797-2525

 1   area, and I think this has come up in a previous meeting,

 2   and I am sure Dr. Murray can speak to this, but as we

 3   try   to       look   at,    one,      mortality      in   this   patient

 4   population, whether or not patients got treated, and

 5   if you look in the literature, most often patients got

 6   some sort of treatment, and it's very difficult to then

 7   sort of look at responses by whatever therapy was

 8   available.

 9                     But       if     one     looks      across      studies,

10   mortalities in the literature varies widely anywhere

11   from 30 percent up to 70 percent.                   So the estimates are

12   quite imprecise as one might expect, given the large

13   variability in the patients that are being reported in

14   these studies.

15                     To speak to the other issue in regard to

16   other experience in, for example, compassionate use

17   programs that have been used to treat                  VRE, there again

18   one runs into the problem of how comparable the patient

19   populations are.

20                     In that other experience, in fact, the

21   breakdown by site of infection was a little different,

22   and moreover, given that that was a compassionate use

23   program, not randomized controlled trials, the patients

                                      S A G CORP.
     202/797-2525                     Washington, D.C.           Fax: 202/797-2525

 1   tended to be a little sicker.

 2                       Having said all of that, if one looked at,

 3   again, in two of those protocols, the overall mortality

 4   rates were around 50 percent in those trials, and the

 5   overall success rates varied anywhere from 46 to 56

 6   percent.        Again, very difficult to try and use those

 7   as historical controls, given the differences in the

 8   patient populations.

 9                       DR. HAFKIN:        You know, I was wondering

10   whether        we   should   look      at    Dr.    Ross'   analysis      of

11   mortality.

12                       DR. ROSS:     So for the MITT-VRE population,

13   in the high dose arm, 16 out of 65 patients died, a

14   mortality rate of 24.6 percent.                    In the low dose arm,

15   18 out of 52, 34.6 percent.                 In the bacteremic patient

16   population -- so these are patients with VRE bacteremia

17   at baseline -- in the high dose arm, four out of 18,

18   that is, 22.2 percent and nine out of 16 in the low dose

19   arm.

20                       DR. HAFKIN:      And if I could make one point,

21   Dr. Ross' slides were prettier, but our analyses imply

22   the same pattern of increased success or differentiation

23   between the high dose versus the low dose, depending

                                  S A G CORP.
     202/797-2525                    Washington, D.C.            Fax: 202/797-2525

 1   on severity of illness.

 2                       So the higher the MPM-2 score, which is the

 3   score that's based on vital sign and underlining

 4   morbidity, the higher the chance of mortality, the more

 5   effective the high dose becomes in relationship to the

 6   low dose.

 7                       CHAIRMAN RELLER:        Dr. Murray.

 8                       DR. MURRAY:      Yeah, I was just going to point

 9   out that the figure, the 30 percent that's usually --

10   that figure that's quoted is from Edmond and Wincell

11   for an attributable mortality to VRE bacteremia, which

12   they defined, if I recall right, as either two positive

13   blood    cultures         or   one   plus     evidence     of   systemic

14   infection with high fevers or low blood pressure.                        So

15   that 30 percent figure is often quoted as an attributable

16   mortality, although not all studies have shown that,

17   but the definition of bacteremia has often varied from

18   even one or two or with or without signs of sepsis.

19                       So as Gary pointed out, it's extremely

20   difficult to, without matching case with case, to use

21   historical controls.

22                       CHAIRMAN     RELLER:          Drs.   O'Fallon      and

23   Wittes,        do   you   have    any   comments     you    might     make

                                    S A G CORP.
     202/797-2525                   Washington, D.C.            Fax: 202/797-2525

 1   addressing the numbers of patients with enterococcal

 2   bacteremia and differences in outcome with what we have

 3   here for 600 versus 200?

 4                  DR. WITTES:     The quick answer is no.            I

 5   mean, I've actually opened up the table of numbers to

 6   do exactly that.    It's very hard to answer without doing

 7   some calculations, but the numbers are small.            There's

 8   lots of different categories that are going on, and I'm

 9   still troubled by -- I'm still focused on the fact that

10   we don't have a consecutive sequence, and I don't know

11   how serious that is.

12                  So I'm punting to Judy.

13                  DR. O'FALLON:      Well, I, too, was having

14   trouble because of the enrichment study and trying to

15   figure that out, how those numbers fit into the rest

16   of them because, you see, the categories for me anyway,

17   we've been asked five questions.           When we look at these

18   numbers in the one with 104 MRSAs and so on, they're

19   grouped by disease.    I mean all of the diseases are stuck

20   together.

21                  So it's hard for me to try to separate them

22   out into the categories that we're being asked to do,

23   and without more time, I can't tell.

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1                  I went through and looked at the original

 2   studies, and they were asking do you have enough people,

 3   and the answer is no, no, no.

 4                  But then to get to the enrichment study,

 5   but then they're all bugged together.           All of the

 6   pneumonias are together and all the SSSIs are together,

 7   and I can't break them out.         So I don't know exactly

 8   what's going on here.     I couldn't -- I don't know what

 9   the numbers are.

10                  DR. CHIKAMI:      Dr. Reller, may I make a

11   comment?

12                  I understand the comment that's made, but

13   I think it relates to one of the issues that I raised

14   early on at the beginning of the morning.       That is, as

15   we, the agency, and at the two public meetings have

16   discussed with the advisory committee and members from

17   academia and the pharmaceutical industry, one of the

18   issues that is being faced is in doing a controlled trial

19   for a specific site of infection, for example, pneumonia

20   or skin, as they're traditionally done, one faces the

21   issue that one could do a study which would demonstrate

22   safety and effectiveness at that site of infection.

23                  If one then wants to also get in the course

                             S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   of that study sufficient experience with a resistant

 2   organism, for example, MRSA or PRSP, it becomes very

 3   difficult, one, because of the epidemiology of some of

 4   these resistant organisms.

 5                  The other issue is an unknowable because,

 6   for example, even when studies are done in community

 7   acquired pneumonia, for example, in South Africa where

 8   there are very high rates of penicillin resistant Strep.

 9   pneumo., one finds in doing the studies that in the

10   controlled trials you wind up with very low rates of

11   pen. resistant Strep. pneumo.          I can't explain that.

12    I don't know why that is.

13                  So that there is a problem in terms of for

14   some of these agents being able to do controlled trials

15   and get sufficient numbers of organisms to support

16   effectiveness or for us to conclude that there is

17   effectiveness.

18                  Having said that, various proposals have

19   been put on the table about how to develop that

20   experience, and one of them is what was done in this

21   development package, that is, to run a clinical trial

22   specifically    to   capture    those      patients   who    have

23   infections with resistant organisms.

                            S A G CORP.
     202/797-2525            Washington, D.C.          Fax: 202/797-2525

 1                   In   the   context     of     a   more   traditional

 2   development program where you have large, adequate

 3   clinical       trials   that     would        support    determining

 4   effectiveness at a specific site of infection and then

 5   using that pathogen driven study to supplement that,

 6   now we understand, and this was some of the discussion

 7   that we have those meetings, there is an issue of how

 8   does one then appropriately pool that information or

 9   consider that information in making the determination

10   that the overall development package demonstrates the

11   effectiveness of the product for the site of infection

12   and subsequently for the resistant organism.

13                   What we've said is that in the course of

14   those traditional studies, we would like to see a

15   substantial amount of information of activity for the

16   susceptible      organism,     for    example,      if     it's     pen.

17   resistant Strep. pneumo.         We would like to see in those

18   studies a substantial amount of evidence for penicillin

19   susceptible Strep. pneumo.

20                   As Dr. Hafkin pointed out, there is the

21   additional information both from in vitro studies and

22   animal studies that suggest that the drug is as active

23   against the resistant strains as it is against the

                                S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1   resistant strains.      Then one is left with a judgment

 2   about how much additional clinical information do you

 3   need to draw the conclusion that it's effective for the

 4   treatment of those resistant organisms.

 5                   So that's sort of the broad brush strokes

 6   of where we are in trying to assess an overall clinical

 7   development package, and for these anti-infectives it's

 8   never just a single study or a single site of infection.

 9    We have multiple studies that look at the activity of

10   the product at different sites of infections which may

11   have a different spectrum of etiologic agents and may

12   have different pharmacokinetic constraints in terms of

13   getting the drug to that site of infection.

14                   DR. O'FALLON:     I wasn't arguing with the

15   strategy.      What I was saying was that it's difficult

16   for me to use the data because the data from the

17   enrichment study was not broken out according to the

18   diseases that we were doing.       They were lumped together,

19   and the hospital based and the HAPs, the CAPs, the CAPs

20   whether they were hospitalized or not, that's all jumped

21   together.      We don't have any way of knowing how it all

22   goes.

23                   You know, you're asking us to say that those

                              S A G CORP.
     202/797-2525             Washington, D.C.       Fax: 202/797-2525

 1   are   all      alike   as far as the             -- as far as this

 2   effectiveness is concerned for these specific bugs, and

 3   I guess we weren't as clear as we would have liked.

 4   When we, the FDA, did these analyses, we first took the

 5   view that, yes, we will look at the studies as they're

 6   designed, that is, a randomized controlled trial for

 7   a specific pathogen at different sites of infection and

 8   perform the analysis in that manner because that's the

 9   way    the     study    was     designed,        understanding     that

10   eventually we would look at those subgroups of interest,

11   that is, at the specific sites of infection, and try

12   to make some assessment of the effectiveness for, for

13   example,       MRSA    in   the    complicated       skin   structure

14   infection.

15                    And so that speaks to why the presentation

16   was structured in the way it was, to look at the overall

17   result to see if, in fact, understanding that each of

18   these sites of infection may have a different prognosis

19   or potential for outcome, but as a first cut, look at

20   the overall result to see if the two agents are

21   reasonably comparable, and then to look at the specific

22   subgroups which were pre-specified, i.e., the site of

23   infection at the time of entry into the study, to see

                                   S A G CORP.
     202/797-2525                  Washington, D.C.          Fax: 202/797-2525

 1   if there's a consistency across those groups, that we

 2   can also attempt to estimate what the effectiveness was.

 3                     DR. O'FALLON:      But the numbers are so small

 4   that I can't tell you whether they're consistent.                     I

 5   don't see that there's enough evidence there to help

 6   me.    So that's my problem.

 7                     DR. MURPHY:     I guess one of the issues here

 8   we're going to just lay out is that we have found not

 9   just with this company, not just with this drug that

10   despite        what   we   hear   are    significant   numbers      of

11   resistant organisms out there, that if you expect to

12   get a study in each one of these indications with enough

13   resistant organisms, you're never going to reach that.

14                     Fortunately, at this point in a way that's

15   the bad news/good news at this point.           And we are asking

16   for you to then, as Gary said -- we know the numbers

17   are small, and so we had the choice of either saying

18   we are going to wait until we have enough in each category

19   which we think would -- we're balancing the risk of the

20   time that will take to the time it would take to try

21   to put pieces of information together.

22                     And what you're telling us, you just still

23   don't think the numbers are enough in the additional

                                  S A G CORP.
     202/797-2525                 Washington, D.C.          Fax: 202/797-2525

 1   analysis, the enrichment pool.

 2                   DR. O'FALLON:    I don't want to -- well, I

 3   don't feel like I'm an expert enough in this area to

 4   say that they are all alike, but you were just asking

 5   us one by one whether the efficacy is the same disease

 6   by disease.

 7                   And then we have 104 MRSAs in this class,

 8   and they're just thrown together and can't break them

 9   out.

10                   Now, I'm going on the assumption that if

11   you are asking for efficacy in five different diseases

12   with the expectation that they could be different, then

13   the effectiveness against the special bugs could be

14   different in the different diseases, and I would say

15   it would be very helpful for us to know what was the

16   particular mix that we had in there, that is, how many

17   HAPs were there, how many CAPs, how many SSIs.

18                   CHAIRMAN RELLER:      Dr. Murray.

19                   DR. MURRAY:   Yeah, and I just wanted to sort

20   of muse philosophically for a moment.         Again, part of

21   this, I believe, is to establish whether a specific

22   labeling can be added for a subgroup of a resistant type

23   organism.      We may not have any debate over whether the

                              S A G CORP.
     202/797-2525             Washington, D.C.           Fax: 202/797-2525

 1   organism class or organism species -- whether the drug

 2   shows efficacy, but whether the efficacy is against a

 3   particular subclass of that organism.

 4                    And if one were content perhaps to label

 5   as   Staph.     aureus   as    opposed       to   the   subgroup     of

 6   methicillin resistant Staph. aureus, that the numbers

 7   would be fine for one and perhaps --

 8                    DR. O'FALLON:     Nothing for another.

 9                    DR. MURRAY:     Or that may not be the best

10   example.       So some of this has to do with how fine the

11   subcategory needs to go, and I just wanted to go back

12   to the VRE issue.

13                    It is a tough one.          As Dr. Leggett said,

14   if we're not sure whether the 200 milligram dose is the

15   placebo dose or if that's better than placebo, and the

16   trend is all in the right direction for the higher dose

17   showing efficacy, and we would expect the 200 to have

18   some efficacy, so I congratulate the company for having

19   had the courage to do the study.             They may have chosen

20   a dose slightly too high at the 200 to be able to show

21   a statistical difference.         If they had chosen a lower

22   dose, it might have been unethical because it might have

23   been ineffective, but it's going to be tough to -- it's

                               S A G CORP.
     202/797-2525              Washington, D.C.              Fax: 202/797-2525

 1   tough for me to know what to do with the results.

 2                   The trend is right, and is there any hope

 3   for larger numbers in the future or combining numbers

 4   and applying statistics, which hasn't been done yet?

 5    Can somebody address the concerns about losing the

 6   alpha?

 7                   As I recall, the second set of numbers

 8   hasn't been added in and analyzed statistically, and

 9   can we get any help?      Are we going to get any?

10                   CHAIRMAN RELLER:      Dr. Wittes.

11                   DR. WITTES:   Well, I really think the first

12   step is to make sure that the sequential ones are there

13   in order.      I don't imagine that that's going to change

14   things very much, but I think in the absence of knowing

15   it, we've got to see that.

16                   DR. MURRAY:   Just for clarification for me

17   now, can I ask?     Now, this was a double blinded study.

18    So both 200 and 600 -- I mean they were blinded and

19   there was no placebo arm.      So what is the ascertainment?

20    I wasn't sure.     I could understand in general why there

21   might be a concern about ascertainment risk had there

22   been a placebo arm perhaps, but in a double blind study

23   are you still concerned about that?

                              S A G CORP.
     202/797-2525             Washington, D.C.           Fax: 202/797-2525

 1                  DR. WITTES:       Oh, absolutely, because I

 2   mean, I don't know the magnitude.           I don't know whether

 3   we're talking about two patients or 20 patients.            I just

 4   won't know.

 5                  But the concern is that you randomize, and

 6   what allows you to make the comparison is you're

 7   comparing like with like, and if you have one group that

 8   is the earlier randomized than the later, even if it's

 9   slight, or more importantly, I mean what I heard and

10   I didn't even know this when I asked the question, but

11   that it's the URIs that are -- the UTIs, sorry -- the

12   UTIs that are the longer -- the ones that are in longer.

13                  And if you look at the data, that's where

14   the big difference is.     Again, if you look at these very

15   small numbers and you look at the lines, the line that

16   has the biggest difference is that line.

17                  Now, is it possible that the group in one

18   of the arms -- that the group that is being plucked out

19   of 54(a) and plucked into 54 is that different in some

20   way because it took longer to cure them, because they

21   were so much more complicated that the case report forms

22   were harder to capture and so forth?

23                  So it has nothing to do with the double

                            S A G CORP.
     202/797-2525             Washington, D.C.           Fax: 202/797-2525

 1   blinding, and it has nothing to do with randomization.

 2    It has to do with essentially not following the

 3   randomization in a funny way, in an ascertainment way.

 4                  And then for 54, once more we don't know --

 5   and again, I'm sure you have the numbers in there.           It

 6   would be good to see them -- how many of the 54s are

 7   not yet analyzed because the data aren't in yet.

 8                  Once the data are all in, then it's easy.

 9    then we would know.

10                  In terms of the alpha, I guess I'm less

11   worried about that.     Sorry about that.

12                  DR. BRITTAIN:     This is Erica Brittain.

13                  There were, I think, 25 patients in 54 who

14   were really 54(a) patients, but when I looked at them,

15   put them into 54(a), it didn't really change the P value.

16                  DR. WITTES:      No, no, but that's right

17   because there's something --

18                  DR. BRITTAIN:    Yeah. I just saw where they

19   -- what their first dose day, June 20th, and when I threw

20   those 25 back into 54(a), the result didn't really change

21                  DR. WITTES:     But we don't know who in 54

22   are in the future that we haven't seen, right?

23                  DR. BRITTAIN:     Definitely.

                             S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1                  DR. WITTES:    But you know, right?

 2                  (Laughter.)

 3                  CHAIRMAN RELLER:      Dr. Hafkin, we'll wrap

 4   this tread up, and then Dr. Norden has another topic

 5   that he wants us to address.

 6                  Dr. Hafkin.

 7                  DR. HAFKIN:    I can't explain some of the

 8   issue.     The point I wanted to make was the point Dr.

 9   Britain made, which was we also got those additional

10   25 patients to put into our 54(a) analysis to answer

11   that question, but let me go back even farther.

12                  We're not, in a sense, happy either with

13   the fact that we studied this disease for more than a

14   year in more than 100 American referral centers, and

15   this is what you've got.     I mean this is an extraordinary

16   -- this represents an extraordinary effort.

17                  It is imperfect, but it is exciting to us

18   because we believe that the drug works not only on the

19   basis of these observations, which we think are very

20   encouraging    and certainly convincing to us, but also

21   that it works for similar germs in the same sites.

22                  So it's our feeling that this is not the

23   only study that we're providing to you.         We've shown

                            S A G CORP.
     202/797-2525           Washington, D.C.        Fax: 202/797-2525

 1   you, we believe, that the drug is equivalent to really

 2   very good drugs for Staph. aureus and for Strep.

 3   pneumonia.             We've   bracketed,         if   you     will,     the

 4   sensitivity of organisms.               The Staph. aureus germ is

 5   much more resistant to the effect of antibiotics in a

 6   general        sense    than    the     Strep.     pneumo.,       and    the

 7   enterococcus is right in the middle.

 8                     So that on the basis of the laboratory

 9   studies we've done, on the basis of the clinical

10   observations, we think we have a really good case to

11   be made when that information is added to the information

12   that comes.            Although it's not perfect, it is the

13   largest        randomized       blinded      study     of      vancomycin

14   resistant enterococcus done in the free world to date.

15                     Maybe someone has got one that I don't know

16   about, but it was an extraordinary effort by my team

17   to put this data together.

18                     Thank you.

19                     CHAIRMAN RELLER:          Dr. Norden.

20                     DR. NORDEN:         I just wanted to follow up on

21   one other point that Dr. O'Fallon was making.

22                     Dr. O'Fallon, it seems to me -- this is the

23   point that Dr. O'Fallon was making -- it seems to me

                                    S A G CORP.
     202/797-2525                   Washington, D.C.              Fax: 202/797-2525

 1   that we are given the information that you were asking

 2   about in terms of Study 31, for example.      The MRSA, if

 3   I heard you correctly, we do under -- we do know what

 4   diseases these patients had so that you can break them

 5   out separately.

 6                  What happens, of course, is the numbers just

 7   become smaller and smaller as you do it, but it's here.

 8    I mean, it's in Dr. Ross' analysis.

 9                  I'm sorry.     It's page 34 of Dr. Ross'

10   presentation.

11                  DR. SOPER:   Carl, correct me if I'm wrong,

12   but that table is -- the data in that table is in addition

13   -- the numbers are in addition to the Studies 55, 39(a),

14   33, 51 and 48(a).    In other words, there's more patients

15   added to that table than were studied in the comparative

16   trials.

17                  So when you go back and you actually find

18   out how many patients were studied in a comparative

19   community acquired pneumonia trial, there was one case

20   of MRSA in the comparative community acquired pneumonia

21   trial, five cases of PRSP, and the hospital acquired

22   pneumonia, there's 22 cases of MRSA.       So that might be

23   able to make some decisions about that, but only two

                             S A G CORP.
     202/797-2525            Washington, D.C.       Fax: 202/797-2525

 1   cases of PRSP.

 2                     In   the     uncomplicated     skin      and     skin

 3   structure infection, there's only one case of MRSA.

 4                     DR. NORDEN:     You're absolutely right.          No,

 5   I agree.       I just was say that in the study, the enrichment

 6   study, if you will, if that's the phrase that's being

 7   used, that the data is here.               I mean we have it, and

 8   I think the way the questions are phrased to us, I mean,

 9   the answer has to be, no, there aren't enough cases in

10   community acquired pneumonia.              No, there aren't enough

11   cases, but if you pool them all together, which is one

12   possible way to do it, the answer might be yes.

13                     DR. SOPER:     Actually as an adjunct to that,

14   what you might want to do is add a sixth question, and

15   that is that number five is infections due to vancomycin

16   resistant enterococci.             You might add number six,

17   infections due to MRSA.

18                     DR. MURRAY:     Well, and then we get into the

19   quibbles.       If something were good for MRSA and hospital

20   acquired pneumonia, would it not be okay for MRSA in

21   community acquired pneumonia, but I think that gets into

22   a final labeling.

23                     We can answer certain questions.          Are there

                                  S A G CORP.
     202/797-2525                 Washington, D.C.           Fax: 202/797-2525

 1   enough patients in a community acquired MRSA to say it's

 2   efficacious in that exact setting.        The answer would

 3   be no, but that's different.        To say that it would be

 4   approved or we would recommend it for nosocomial

 5   pneumonia due to MRSA and not community acquired

 6   pneumonia due to MRSA wouldn't be logical.

 7                  DR. CHIKAMI:   May I clarify the intent of

 8   the questions?      I think that we traditionally ask

 9   questions about each indication because, again, that's

10   the way these drugs are developed.        With the specific

11   resistant organisms, PRSP, MRSA, we understand as you've

12   all been discussing that we've had to rethink or relook

13   at how, in fact, we can get that additional experience.

14                  The intention of the questions was not to

15   say within community acquired pneumonia did the sponsor

16   collect sufficient evidence to support approval within

17   that study.     Our intention all along was to ask the

18   committee to answer the first part of that question,

19   okay, the general indication, and then looking at all

20   of the information that's been presented by the sponsor

21   by our analyses, which may include in addition to that

22   information collected within the specific indication

23   driven studies, to help us try and make that judgment

                            S A G CORP.
     202/797-2525           Washington, D.C.        Fax: 202/797-2525

 1   within this data set, within the overall development

 2   program, is there sufficient information to support

 3   effectiveness for the treatment of that resistant

 4   organism; sort of, again, speaking back to what I said

 5   at the outset of the morning.

 6                  These are difficult infections to study,

 7   and we need to be willing or I think we should be willing

 8   to look across all of the information to try and make

 9   a judgment.    Again, understanding that there are issues

10   related to pooling information, but I think given the

11   situation where in some instances we have infections

12   which are untreatable or may soon be untreatable, we

13   need to think about, again, that risk-benefit ratio that

14   we do as a regulatory agency in trying to certainly

15   insure that drugs are effective for their intended use

16   and they're safe for their intended use, but to also

17   try and get potentially important products to the market

18   as quickly as is prudently possible.

19                  CHAIRMAN RELLER:     Dr. O'Fallon and then Dr.

20   Ross.

21                  DR. O'FALLON:    Okay.      I wanted to point out

22   that the numbers that you guys mentioned are the ones

23   that I was working with, and what I'd like to point out,

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   two little pieces of information that are causing me

 2   trouble.

 3                  In the second group, hospital acquired

 4   pneumonia and you look at that's the one that has the

 5   22 with it, if you take a look at it, there the confidence

 6   interval is not -- there the efficacy is not there.

 7   The evidence is not good for that.            So that's just one

 8   cautionary issue.    They've been put together with the

 9   rest of the pneumonias.

10                  And in the skin infections, the vast bulk

11   of the ones in the VRE are -- the 66 -- are skin and

12   structure infections, and half of -- now, there there's

13   a slight preponderance of favor for the -- you know,

14   when you batch them together, the complicated and the

15   uncomplicated, there's a slight benefit to the new drug,

16   but the problem here is that we don't know what the

17   mixture of the complicated and the uncomplicated is.

18                  So those are some of the issues.

19                  CHAIRMAN RELLER:         Dr. Ross.

20                  DR. ROSS:     So in terms of one of the issues

21   is of those patients who carry a label of pneumonia in

22   Study 31, which of those are community acquired and which

23   are hospital acquired, and I think that that's an

                            S A G CORP.
     202/797-2525               Washington, D.C.         Fax: 202/797-2525

 1   important       point   because     those       are    obviously     two

 2   different entities.

 3                    And one point I would make is that the

 4   epidemiology of MRSA is such that that's much more likely

 5   to be a hospital acquired pathogen.               I mean, obviously

 6   as we've learned over the last year or so, unfortunately

 7   that can be a cause of outbreaks of community acquired

 8   infection, but it's much more likely to be a hospital

 9   acquired pathogen.

10                    DR. MURPHY:     Judith or Dr. O'Fallon, would

11   you go to the next page, to the MIT analysis on 48(a),

12   where we've got the --

13                    DR. O'FALLON:       What page is that?

14                    DR. MURPHY:     It's page 18 of the FDA.            And

15   this is the MITT analysis.            It's R-18 handout.

16                    Just trying to get the issue of severity

17   and MRSA.       Does that help you any where we've               got the

18   ventilated assisted pneumonia, and then you've got those

19   with MRSA and the MITT?          Okay.

20                    CHAIRMAN    RELLER:           Okay.       One    final

21   question for Dr. Hafkin, just to make sure that we're

22   clear on the numbers.         Of the patients with pneumonia,

23   mostly         community    acquired           with    Streptococcus

                                 S A G CORP.
     202/797-2525                Washington, D.C.             Fax: 202/797-2525

 1   pneumoniae, the 32 out of 169 patients had positive blood

 2   cultures.       So this was a substantial number and somewhat

 3   better than what most people had observed.

 4                    In that group of 32 though, there was but

 5   one penicillin resistant pneumococcus; is that correct?

 6                    DR. HAFKIN:      The only protocol that we have

 7   -- no, not additional.            Those patients with resistant

 8   Strep. pneumo. bacteremia were found in the pediatric

 9   trial.

10                    CHAIRMAN RELLER:         So that from the adults,

11   we have no data on bacteremic pneumococcal pneumonia

12   with     less    than    fully      susceptible     organisms      to

13   penicillin?

14                    DR. HAFKIN:      We had two with intermediate,

15   yes.    For all practical purposes, yes, yes.

16                    CHAIRMAN RELLER:         Right.   I know there is

17   not necessarily an answer, but I'd like to ask the

18   question anyway, and that is do you have any explanation?

19    Is it the distribution of where the studies are done?

20                    Why    is   it   so   difficult    when   in    many

21   hospitals in the United States 20, 30, 40, 50, 60 in

22   our place percent of strains of S. pneumoniae are less

23   than fully susceptible to penicillin; of why one would

                                  S A G CORP.
     202/797-2525                 Washington, D.C.         Fax: 202/797-2525

 1   have basically zero out of 32, but out of the others

 2   there were approximately 20 percent less than fully

 3   susceptible, ten percent at high level resistance, 15

 4   percent intermediate?

 5                        So that of the same groups of patients, the

 6   ones who had an organism in their blood and the ones

 7   who had an isolate recoverable differences in how the

 8   MICs were distributed to penicillin, what's going on

 9   do you think?

10                        DR. HAFKIN:    Well, of course, I think about

11   things like that.           Why don't you have protocols that

12   provide patients with resistance?                And I think it's

13   unfortunately an artifact of the protocol process.

14                        What happens is we strive so hard to have

15   evaluable patients, and so we put so many barriers

16   between        the    physicians,    identifying   patients     with

17   pneumonia and making sure that patient is likely to be

18   evaluable.           So we say, "Well, if the patient has been

19   on therapy for less than three days and is not a failure,

20   they don't fit.          If they've had more than 12 hours or

21   24 hours of antibiotics, they don't fit."

22                        So many of the patients, if you step back

23   and you think about the epidemiology of multiply drug

                                   S A G CORP.
     202/797-2525                  Washington, D.C.       Fax: 202/797-2525

 1   resistant organisms in general, they're patients with

 2   comorbidities      who    have    had    multiple    courses     of

 3   antibiotics, and in fact, they are probably empirically

 4   treated because they are sick.               They're at risk for

 5   having terrible outcomes, and so they come to the doctor.

 6    They're already on the antibiotic.             Whether they get

 7   better or not is another story.

 8                  If the physicians -- if we had a perfect

 9   system to capture the failures, then, indeed, we could

10   have trials.       But the problem is, as you know how

11   clinical trials are done in your institution, we have

12   one or two investigators who have an interest, and it's

13   just there are too many barriers to finding those

14   patients that are highly likely to be evaluable and

15   getting them into our protocol.

16                  So I think it's a protocol process itself.

17                  CHAIRMAN RELLER:         There's another reason

18   for asking the question.          I think some have observed

19   that    patients    who    have    unequivocally      documented

20   infection with Streptococcus pneumoniae tend to have

21   more susceptible strains as opposed to patients who are

22   colonized or who have respiratory isolates, that is,

23   respiratory isolates versus isolates from the blood.

                               S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1                   And it raises the question of the certainty

 2   with which one has the entity in hand, and some of these

 3   may     be     respiratory   isolates       in   patients      who

 4   unequivocally have pneumonia, but maybe not necessarily

 5   that that was the organism that was the etiology even

 6   though a pneumococcus was recovered, and that's another

 7   possible explanation.

 8                   One last comment, and from the committee

 9   we'll take additional ones, and then we need to get to

10   the voting.      Yes?

11                   DR. ANDERSON:       Your question was very

12   interesting.      So I appreciate a chance to answer it.

13                   We, in fact, specifically tried to identify

14   patients in Protocol 49 who had had the very experience

15   that Dr. Hafkin pointed out as limiting, and the results

16   were that five of 13, maybe it was 14 isolates, were

17   nonsusceptible or resistant.         So there you get your 35

18   to 40 percent.

19                   Because we were looking for patients who

20   had had previous experience, because we had patients

21   that had refractory exposure to antibiotics, we found

22   them, and that's the rate you'll find in the literature.

23                   CHAIRMAN RELLER:      Thank you.

                              S A G CORP.
     202/797-2525             Washington, D.C.          Fax: 202/797-2525

 1                  Now, we have a strategic issue for the

 2   committee.      We're actually a little bit ahead of

 3   schedule, but we also have pressing deadlines of airline

 4   flights so that we wouldn't have everybody with us at

 5   the end.

 6                  For the committee members, do we press ahead

 7   with the questions or do we take a very brief ten minute

 8   break and then press ahead with the questions?          What's

 9   your wish?

10                  All of those who wish to press on, raise

11   their hand.

12                  (Show of hands.)

13                  CHAIRMAN RELLER:       Those who would rather

14   take a break?

15                  Okay.   Anyone who needs to take a break,

16   do so quickly on the committee members, but come back

17   before the voting.

18                  We will now address the questions at hand.

19    I will read the question, and then we will entertain

20   any relevant discussion that any member of the committee

21   wants to make, and then we'll take a vote, sequentially

22   one by one.

23                  Community acquired pneumonia.     Do the data

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   support the efficacy and safety of linezolid in the

 2   treatment of adult patients with community acquired

 3   pneumonia?

 4                  Any comments before voting on this?

 5                  Dr. Christie.

 6                  DR.   CHRISTIE-SAMUELS:          I     may      have

 7   misunderstood, but I have one little thing that I want

 8   to have cleared up.     The thing is that if we approve

 9   this drug for adults, it will be used in children, too,

10   and the pediatric data is not yet in.

11                  I'm not sure if I heard earlier on that

12   somebody said that the clinical course in pneumococcal

13   pneumonia is similar to that in adults, and I wondered

14   if maybe there's somebody on your team who could tell

15   us, maybe Dr. Kaplan, if that is, indeed, true.

16                  DR. KAPLAN:       I'm       Shelly Kaplan from

17   Baylor College of Medicine in Houston.        I'm a consultant

18   for Pharmacia and Upjohn.

19                  I would say that pneumococcal pneumonia in

20   children, especially if it's bacteremic, is a very

21   serious infection, and I would not expect the child to

22   get better easily without antibiotics.          I can tell you

23   that in the pediatric pneumonia trial with linezolid

                            S A G CORP.
     202/797-2525            Washington, D.C.            Fax: 202/797-2525

 1   that in the six children who had bacteremic pneumococcal

 2   pneumonia, including two children with empyema, that

 3   the agent looked very, very effective, if that answers

 4   your question.

 5                    CHAIRMAN RELLER:         Dr. Murray.

 6                    DR. MURRAY:      If you look at the charts from

 7   I   think      the   Mandell     textbook,       the    mortality       in

 8   bacteremic pneumococcal disease for those less than two,

 9   after that age it goes down and it goes back up again.

10    I think the less than two is equal to about the over

11   65 or 70 in terms of mortality, whereas the older child

12   and the younger adult are quite equivalent also.

13                    DR.    CHRISTIE-SAMUELS:              That's    what    I

14   thought.

15                    CHAIRMAN RELLER:          In considering all of

16   these questions, before the voting I think that, you

17   know, an important consideration in trying to give

18   precise, specific recommendations to the agency on the

19   sense of the committee on these issues, that for the

20   resistant organisms there are a couple of ways to

21   approach       it,     and   considering        the    data     both    by

22   indication, as well as the pooled information, that a

23   recommendation for a recognition of safety and efficacy

                                  S A G CORP.
     202/797-2525                 Washington, D.C.             Fax: 202/797-2525

 1   for susceptible strains doesn't necessarily preclude

 2   the treatment of resistant strains, given that some of

 3   these    organisms      --    there     aren't         other    in    vitro

 4   susceptible strains at the outset.

 5                   So that there are lots of ways of being able

 6   to handle the questions, and I think it's important as

 7   we vote to take them one by one and our best judgment

 8   on the specific question asked, and at the conclusion

 9   if there be additional studies or things that the

10   committee      thinks    would     be    very     important          in   the

11   continued development and deployment of this agent,

12   since we're talking about it today, but it would apply

13   to other things that we address in the future, is that

14   we can make those recommendations as well, but we want

15   to answer these questions as they are written.

16                   The     voting.         We      will    start    on       the

17   right-hand side for those who are voting, and I think

18   from the outset we delineated who were voting members.

19    In addition to the current members of the committee,

20   Drs. Norden and Leggett will be voting this afternoon.

21                   Dr. Soper.

22                   DR. SOPER:       Yes.

23                   DR. CHESNEY:         We are answering 1(a); is

                                  S A G CORP.
     202/797-2525                 Washington, D.C.              Fax: 202/797-2525

 1   that correct?

 2                  CHAIRMAN RELLER:      We are answering 1(a) and

 3   only 1(a).

 4                  DR. CHESNEY:    Yes.

 5                  CHAIRMAN RELLER:      Yes.

 6                  DR. NORDEN:    Yes.

 7                  DR. DANNER:    Yes.

 8                  DR. RODVOLD:    Yes.

 9                  DR. CHRISTIE-SAMUELS:        Yes.

10                  DR. LOWY:   Yes.

11                  DR. MURRAY:    Yes.

12                  DR. O'FALLON:     Yes.

13                  CHAIRMAN RELLER:         Section 1(b).       Given

14   that we unanimously believe that this agent has been

15   shown to be effective and safe for the treatment of

16   adults with community acquired pneumonia, do the data

17   support the inclusion of a specific wording to the effect

18   that there are sufficient data to support safety and

19   efficacy in the treatment of methicillin resistant

20   Staph. aureus as an etiology of community acquired

21   pneumonia in adults?

22                  And we'll take these individually.               So

23   we're talking now 1(b)(I).

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1                  Dr. Soper.

 2                  DR. SOPER:    No.

 3                  DR. CHESNEY:    No.

 4                  CHAIRMAN RELLER:      No.

 5                  DR. NORDEN:    Yes.

 6                  DR. DANNER:    Yes.

 7                  DR. RODVOLD:    Yes.

 8                  DR. CHRISTIE-SAMUELS:       Yes, based on AGP

 9   data.

10                  DR. LOWY:    No.

11                  DR. MURRAY:     Yes, based on the aggregate

12   data in other sites and in hospital acquired pneumonia.

13                  DR. O'FALLON:      No.

14                  CHAIRMAN RELLER:      We have a split decision

15   of five-five.

16                  (Laughter.)

17                  CHAIRMAN RELLER:      One (b)(2).

18                  DR. SOPER:    No.

19                  DR. CHESNEY:    No.

20                  CHAIRMAN RELLER:      No.

21                  DR. NORDEN:    Yes.

22                  DR. DANNER:    Yes.

23                  DR. RODVOLD:    Yes.

                            S A G CORP.
     202/797-2525           Washington, D.C.            Fax: 202/797-2525

 1                  DR. CHRISTIE-SAMUELS:           No.

 2                  DR. LOWY:     No.

 3                  DR. MURRAY:      No.

 4                  DR. O'FALLON:       I vote no, but this one had

 5   more data than the other one did.

 6                  (Laughter.)

 7                  DR. MURRAY:      But less data for resistant

 8   organisms.

 9                  CHAIRMAN    RELLER:           All   right.      Having

10   completed that question, since the vote here is three

11   yeses and seven noes, would anyone like to amplify for

12   the agency, you know, post voting discussion?                Because

13   Dr. O'Fallon has pointed out that the numbers -- and

14   I think it has to do with the inclusion of ancillary

15   data from the specific pathogen studies that were

16   incorporated into the decision.

17                  Dr. Murray, comment?

18                  DR. MURRAY:     For the penicillin resistant

19   pneumococcus, I thought there was, despite the sponsor's

20   efforts, just not enough cases overall of penicillin

21   resistant disease, although biologically there is no

22   earthly reason to think that a penicillin resistant

23   pneumococcus is going to respond any different from a

                               S A G CORP.
     202/797-2525              Washington, D.C.             Fax: 202/797-2525

 1   penicillin susceptible pneumococcus.         We've had that

 2   discussion before with the floroquinalones.

 3                  So biologically I think there is no reason

 4   to expect a difference.       However, we have had a -- we

 5   have, in essence, set a standard and asked for very

 6   rigorous data for that labeling and gotten it.                  My

 7   contrary vote on the methicillin resistant Staph. aureus

 8   relates to the what I viewed as efficacy and hospital

 9   acquired pneumonia with the organism and have even less

10   reason then to think that the community acquired

11   pneumonia with an MRSA would be any different.

12                  CHAIRMAN RELLER:      Dr. Chesney.

13                  DR. CHESNEY:    After lunch, Dr. Hafkin gave

14   us numbers of 23 of 27 penicillin nonsusceptible

15   pneumococci responded, which is a good number, but if

16   this is approved, then that's what people will use first,

17   and I think those numbers are just too small, and that's

18   why I voted against it, even if we take everything that's

19   available.

20                  CHAIRMAN RELLER:      Dr. Norden.

21                  DR. NORDEN:    I voted yes because I think

22   in the aggregate -- I'm not sure I know what the right

23   number is, but I remember, and we talked about this at

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1   lunchtime, we did approve levoquin for PRSP, PRSP, yeah,

 2   something like that, and I don't think the numbers were

 3   much greater, and I have no reason to doubt that this

 4   data is good.

 5                   So it seems to me that I agree with Barbara.

 6    It should work, and therefore, I'm not clear why we

 7   would    vote   not   to   recommend          it   for   pneumococcal

 8   pneumonia.

 9                   CHAIRMAN RELLER:        Dr. Leggett.

10                   DR. LEGGETT:         I was not here for the

11   levofloxiquin vote, but I can say that since it's now

12   being used for everything, we are seeing a markedly

13   increased rate of floroquinalone resistance, and the

14   timing is pretty much the same.

15                   CHAIRMAN RELLER:        Dr. Rodvold, do you have

16   any comment you want to make?

17                   DR. RODVOLD:       The other comment is that

18   this question is a little bit different, I think, than

19   in some of the levoquin discussions, that we're dealing

20   with just the penicillin resistant pathogen, and we

21   didn't get into so much of the bacteremia issue, which

22   I think we dealt a little bit more with levoquin, and

23   I think we got more information.

                                S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1                    I think from a post marketing point, okay,

 2   or from what the agency maybe should require after

 3   whatever you approve is that it would be very helpful

 4   to   have      experience   in    CAP    with   bacteremia,       with

 5   penicillin        resistant        isolates      for      everyone,

 6   particularly as they're going to proceed on to try to

 7   get a pediatric indication because I think that will

 8   be even more important up there.

 9                    So I voted yes a little bit just based off

10   of a pathogen issue, but I'm not -- if you had said,

11   "And bacteremic," you'd have changed my vote instantly

12   back out to a no.

13                    So I am staying only pathogen, without going

14   into -- going past the lung, so to speak, and I think

15   it would be very helpful for everyone in multiple of

16   these other indications that they seek those type of

17   patients later on.

18                    CHAIRMAN RELLER:          Actually, Dr. Rodvold

19   touched on one of my concerns, and in the discussion

20   earlier, it was noted that many of these patients had

21   multiple organisms isolated, and with all due -- I mean,

22   doing clinical research is an exceedingly difficult

23   task, and getting a precise etiology is also tough, and

                                 S A G CORP.
     202/797-2525                Washington, D.C.           Fax: 202/797-2525

 1   it's getting tougher, given the microbiological support

 2   that is available to some investigators, you know, at

 3   the front lines.

 4                  But I, frankly, am a bit concerned about

 5   Staphylococcus aureus, be it methicillin resistant or

 6   susceptible.     I think the numbers of patients with

 7   pneumonia who have the organism isolated versus the

 8   number who have pneumonia caused by the organism may

 9   be   considerably   different,       especially    in   hospital

10   acquired pneumonia.

11                  We need to move on to question number two.

12                  Two (a), do the data support the efficacy

13   and safety of linezolid in the treatment of adult

14   patients with hospital acquired pneumonia?

15                  Dr. Sober.

16                  DR. SOPER:    Yes.

17                  DR. CHESNEY:    Yes.

18                  CHAIRMAN RELLER:      Yes.

19                  DR. NORDEN:    Yes.

20                  DR. DANNER:    Yes.

21                  DR. RODVOLD:    Yes.

22                  DR. CHRISTIE-SAMUELS:        Yes.

23                  DR. LOWY:    Yes.

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1                  DR. MURRAY:    Yes.

 2                  DR. O'FALLON:        No, and the reason I'm

 3   saying that is because the confidence intervals on this

 4   are getting down into the dangerous area.            That's why

 5   I'm saying no.

 6                  CHAIRMAN RELLER:      Thank you, Dr. O'Fallon.

 7                  I think we have nine-one supporting safety

 8   and efficacy for HAP.

 9                  Two (b), are there sufficient data to

10   support the claim for efficacy with hospital acquired

11   pneumonia owning specifically to MRSA?

12                  Dr. Sober?

13                  DR. SOPER:    Yes.

14                  DR. CHESNEY:    Yes.

15                  CHAIRMAN RELLER:      Yes.

16                  DR. NORDEN:    Yes.

17                  DR. DANNER:    Yes.

18                  DR. RODVOLD:    Yes.

19                  DR. CHRISTIE-SAMUELS:        Yes.

20                  DR. LOWY:    Yes.

21                  DR. MURRAY:    Yes.

22                  DR. O'FALLON:        Mixed message.       I'll go

23   with yes.

                            S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1                   (Laughter.)

 2                   CHAIRMAN RELLER:          Two (b)(2), penicillin

 3   resistant Streptococcus pneumoniae.

 4                   Soper?

 5                   DR. SOPER:     No.

 6                   DR. CHESNEY:    No.

 7                   CHAIRMAN RELLER:      No.

 8                   DR. NORDEN:    Yes.

 9                   DR. DANNER:    Yes.

10                   DR. RODVOLD:    No.

11                   DR. CHRISTIE-SAMUELS:         No.

12                   DR. LOWY:     No.

13                   DR. MURRAY:    No.

14                   DR. O'FALLON:       No.

15                   CHAIRMAN RELLER:      Eight to two not thinking

16   that that supports that specific claim.

17                   Dr. Chikami, do you want to make a comment?

18                   DR. CHIKAMI:    I guess as you discuss this

19   issue also with the community acquired pneumonia, if

20   the committee had any other comments or reasons for

21   their,     in   particular,    for    their     no   votes,     with

22   additional information.             For example, Dr. Rodvold

23   suggested, for example, more experience in patients with

                              S A G CORP.
     202/797-2525             Washington, D.C.            Fax: 202/797-2525

 1   bacteremia.        Is that sort of a similar concern in this

 2   case?

 3                     CHAIRMAN RELLER:       I think I could probably

 4   speak for the committee.           I mean, the reality is that

 5   we have 32 patients who unequivocally have pneumonia

 6   caused by Streptococcus pneumoniae, and most of them

 7   are as appropriate in the community.               We know that

 8   people     can    acquire   pneumococcal       pneumonia   in    the

 9   hospital.        It can be transmitted in the hospital, but

10   the numbers are simply not there with resistant strains.

11                     I don't think anyone has concerns that a

12   linezolid         susceptible      organism      would     respond

13   appropriately to treatment, but that's a different issue

14   from having the data in hand for a specific indication

15   for that organism.       It doesn't in any way preclude using

16   it for strains that are less than fully susceptible to

17   any other antimicrobial if they're susceptible in that

18   institution to linezolid.

19                     Question three, uncomplicated skin and skin

20   structure infections.           Are the data there to support

21   safety and efficacy of linezolid for this indication

22   in adults?

23                     Dr. Soper.

                                 S A G CORP.
     202/797-2525                Washington, D.C.         Fax: 202/797-2525

 1                  DR. SOPER:      Yes.

 2                  DR. CHESNEY:       Yes.

 3                  CHAIRMAN RELLER:         Yes.

 4                  DR. NORDEN:       Yes.

 5                  DR. DANNER:       Yes.

 6                  DR. RODVOLD:       Yes.

 7                  DR. CHRISTIE-SAMUELS:           Yes.

 8                  DR. LOWY:      Yes.

 9                  DR. MURPHY:       Yes.

10                  DR. O'FALLON:        Yes.

11                  CHAIRMAN RELLER:          Are the data there to

12   support a specific subsidiary recommendation for skin

13   and    skin    structure      infections       with   methicillin

14   resistant Staph. aureus?

15                  DR. SOPER:      No.

16                  DR. CHESNEY:       No.

17                  CHAIRMAN RELLER:         No.

18                  DR. NORDEN:       No.

19                  DR. DANNER:       Yes.

20                  DR. RODVOLD:       No.

21                  DR. CHRISTIE-SAMUELS:           No.

22                  DR. LOWY:      No.

23                  DR. MURRAY:       Yes.

                            S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1                  DR. O'FALLON:     No.

 2                  CHAIRMAN RELLER:        Eight to two thinking

 3   that there are not sufficient data at present.

 4                  Question 4(a) -- comments, sure.

 5                  DR. MURRAY:    My commentary on the positive

 6   side is based on the overall gestalt and with the other

 7   supporting data across pathogen site.          If it's going

 8   to work in community in nosocomial pneumonia, I think

 9   that with supporting data it will work in complicated

10   skin and soft tissue infections.

11                  So I let the across the site organism drive

12   me even with the small numbers for this particular site

13   pathogen.

14                  DR. SOPER:     And what I suggest that we do

15   to try to remedy this at the end of the day is add question

16   number six about MRSA.

17                  DR. RODVOLD:    I brought up a point earlier

18   and the company answered it, is that in the uncomplicated

19   trial, they actually used 400 milligram dosages.             So

20   if you are going to give an MRSA indication, they

21   themselves said that 600 was what they felt comfortable

22   with.    So to me if you do come around to an indication,

23   I think you've got to deal with a dosage recommendation

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   here.

 2                  And your MICs had a tendency and a means

 3   and some of the global data raised a little bit.                 So

 4   I think you may have a labeling dosage issue as well

 5   to sort out or it needs to be at least discussed.

 6                  DR. CHIKAMI:       May I ask other members of

 7   the committee if, in fact, that was an issue that they

 8   took into consideration?

 9                  Because, as Dr. Rodvold pointed out, if --

10   that can certainly be handled in labeling, given the

11   fact that the preponderance of the evidence for -- excuse

12   me -- preponderance for data for MRSA comes from Study

13   31, which a 600 milligram BID dose was used, so, in fact,

14   that, in fact, could be linked to that particular --

15   treatment of that particular organism in any of the skin

16   indications.

17                  So I just sort of wanted to get the

18   committee's read on that.

19                  CHAIRMAN     RELLER:          Comments   for     Dr.

20   Chikami?

21                  DR. NORDEN:     Yeah, I agree with Dr. Rodvold

22   on that.

23                  CHAIRMAN     RELLER:          It   influenced      my

                               S A G CORP.
     202/797-2525              Washington, D.C.          Fax: 202/797-2525

 1   consideration that, you know if there is any singling

 2   out of this organism, that it would be important to have

 3   the higher dose until we had sufficient data that the

 4   other dose was comparable effective for this particular

 5   pathogen.

 6                  The other concern that I had was you're

 7   exactly right that most of the data that would support

 8   this, since there was only one patient with MRSA in the

 9   uncomplicated     category     from        the   trials    is    the

10   supplemental data was important and useful, but I'm not

11   absolutely clear.

12                  We saw the depiction of the categories of

13   those patients.      Many of them were surgical wound

14   infections,    and   what    role,    removing     the    sutures,

15   draining the stitch abscess, all of those things had,

16   and admittedly those patients were very sick, but I got

17   the sense that they were very sick not necessarily

18   because of the Staphylococcal infection, but because

19   they were very sick patients who had a staphylococcal

20   infection.     So I think it's -- you know, I'm not exactly

21   sure of the comparability, not that it isn't important

22   information, but that it's to me a bit different from

23   an enrollment specifically for a skin and skin structure

                             S A G CORP.
     202/797-2525            Washington, D.C.             Fax: 202/797-2525

 1   infection.

 2                  DR. LOWY:     And that was the major reason

 3   for my no, because I'm not sure that the complicated

 4   skin    and    cell   tissue    people      were    really     that

 5   complicated.     It sounds like you could just open up the

 6   wound, and you didn't really need antibiotics, and we

 7   are then going to launch this into diabetic feet, osteo

 8   that's not recognized, a whole can of worms that has

 9   not even been studied, and the MICs are near the floor

10   level, which is with a bacteriostatic drug makes me very

11   worried.

12                  CHAIRMAN RELLER:       Dr. Chesney.

13                  DR. CHESNEY:     I just wanted to endorse Dr.

14   Soper's comment that we vote -- make number six be

15   basically study number 30 -- 31.

16                  CHAIRMAN RELLER:        We'll get there after

17   five.

18                  Dr. Soreth.

19                  DR. SORETH:     Two comments.       With regard to

20   complicated skin and skin structure infections with some

21   notation as to how many patients require surgical

22   adjunctive intervention at baseline, that's actually

23   part of the division's guidance document for                    the

                             S A G CORP.
     202/797-2525             Washington, D.C.           Fax: 202/797-2525

 1   definition of complicated skin and skin structure

 2   infections.

 3                   So we ask that sponsors study patients in

 4   complicated skin and skin structure infections for whom

 5   it is assumed and shown that a certain proportion of

 6   them require surgical intervention.               That's what helps

 7   us   to    deduce      that     those     patients    are,    indeed,

 8   complicated in addition to having co-morbid conditions,

 9   perhaps immunodeficiency, and so forth.

10                   A second comment unrelated to that, but

11   related to this proposed sixth question has to do with

12   the way the division grants indications, and this

13   application is no different in that respect.                         We

14   typically      grant    indications         for   a   site   specific

15   infection, pneumonia or hospital acquired pneumonia,

16   interabdominal infections, urinary tract, and so forth.

17                   VRE represents an exception,a nd I think

18   to date the only exception.                 We have not granted a

19   methicillin resistant staph. infection indication at

20   all body sites.        The sponsor has not requested it, and

21   in developing and co-developing this program of drug

22   development, we thought that the approach of having this

23   catch-all Protocol 31 that would develop and enrich four

                                   S A G CORP.
     202/797-2525                  Washington, D.C.          Fax: 202/797-2525

 1   methicillin resistant staph. infections would then, in

 2   fact, support, give numbers to the site specific

 3   indications, knowing from hard experience for however

 4   many reasons there are that we don't get large numbers

 5   of such resistant organisms within that site specific

 6   indication.

 7                  So that's why we have not drafted that sixth

 8   question, because we don't grant the indication that

 9   way.

10                  CHAIRMAN RELLER:      I understand.

11                  Question 4(a), do the data support efficacy

12   and safety of linezolid in the treatment of adult

13   patients with complicated skin and skin structure

14   infections?

15                  DR. SOPER:    Yes.

16                  DR. CHESNEY:    No.

17                  CHAIRMAN RELLER:      Yes.

18                  DR. NORDEN:    Yes.

19                  DR. DANNER:    Yes.

20                  DR. RODVOLD:    Yes.

21                  DR. CHRISTIE-SAMUELS:        Yes.

22                  DR. LOWY:    Yes.

23                  DR. MURRAY:    Yes.

                             S A G CORP.
     202/797-2525            Washington, D.C.           Fax: 202/797-2525

 1                  DR. O'FALLON:      Yes.

 2                  CHAIRMAN RELLER:          Are there sufficient

 3   data to support efficacy and safety of linezolid for

 4   the treatment of complicated SSSI caused by methicillin

 5   resistant Staph. aureus?

 6                  DR. SOPER:    No.

 7                  DR. CHESNEY:    No.

 8                  CHAIRMAN RELLER:       No.

 9                  DR. NORDEN:     Yes.

10                  DR. DANNER:     Yes.

11                  DR. RODVOLD:    Yes.

12                  DR. CHRISTIE-SAMUELS:         No.

13                  DR. LOWY:    No.

14                  DR. MURRAY:     Yes.

15                  DR. O'FALLON:      This is crazy.      You have in

16   this -- in this study there are 66 in this area that

17   have, you know, information on this, and it's a 70 --

18   what is it nine versus 73 percent in favor?            So there's

19   more evidence here, but I don't know whether it's

20   complicated or uncomplicated.               It's just the skin

21   structures are just stuck together.                So I would say

22   there's evidence of some efficacy.            Yes.

23                  (Laughter.)

                            S A G CORP.
     202/797-2525           Washington, D.C.              Fax: 202/797-2525

 1                   CHAIRMAN RELLER:       Yes.

 2                   DR. STEVENS:      Dr. Reller, it seems like

 3   there's some confusion about what a complicated soft

 4   tissue infection is.

 5                   My name is Dennis Stevens.        I have had a

 6   major interest in soft tissue infections for a number

 7   of years.      I participated in a number of the linezolid

 8   clinical       trials,   including      the   complicated     and

 9   uncomplicated soft tissue infections.

10                   These patients were very sick patients.

11   They were not folks that just had a simple surgical wound

12   infection that may be manipulating the wound a little

13   bit out, taking a suture out, would go away.

14                   I could give you an example of the type of

15   patient that we admitted into the complicated skin

16   infection.      A 65 year old diabetic patient with a huge

17   staph. and Group A strep. abscess on his back that had

18   bacteremia, had a huge debridement, area about this big,

19   and had a necrotizing superficial infection, the skin

20   and the fat, but not the fascia because that was an

21   exclusion criteria.

22                   I think these were very ill patients.        They

23   weren't minor at all, and I think that's kind of been

                              S A G CORP.
     202/797-2525              Washington, D.C.        Fax: 202/797-2525

 1   a misconception around the room.

 2                    CHAIRMAN RELLER:       Thank you, Dr. Stevens.

 3                    Any other post vote commentary on question

 4   four?

 5                    Question number five, do the data support

 6   the efficacy and safety of linezolid in the treatment

 7   of adult patients with infections caused by vancomycin

 8   resistant enterococci?

 9                    DR. SOPER:      Yes.

10                    DR. CHESNEY:     Yes.

11                    CHAIRMAN RELLER:       Yes.

12                    DR. NORDEN:     Yes.

13                    DR. DANNER:     Yes.

14                    DR. RODVOLD:     Yes.

15                    DR. CHRISTIE-SAMUELS:         Yes.

16                    DR. LOWY:       Yes, with a caveat that we

17   continue to look for data to firm up whether it really

18   works.

19                    DR.   MURRAY:      I    ditto   those     comments

20   exactly.       Yes.

21                    DR. O'FALLON:     I'm just going to reinforce

22   Dr. Wittes' concern.         These data could be biased, and

23   we don't have any way of knowing, and so the results

                               S A G CORP.
     202/797-2525              Washington, D.C.            Fax: 202/797-2525

 1   we have look like it's effective, but we aren't sure

 2   what kind of data we have here.

 3                  CHAIRMAN    RELLER:           Is   that    a   no,    Dr.

 4   O'Fallon?

 5                  DR. O'FALLON:       Make it a no.

 6                  (Laughter.)

 7                  CHAIRMAN RELLER:        For the record.

 8                  Now, 5(b) is a very important question

 9   because it gives us the opportunity to tell the agency

10   what we would ideally like to see in this whole arena

11   of treating these emerging, increasing infections, but

12   that take place in patients who are so critically ill

13   that     sorting   out     efficacy          becomes     exceedingly

14   difficult.

15                  Recommended      additional         studies?          Dr.

16   Soper, Dr. Chesney?

17                  We're contemplating.

18                  Dr. Norden.

19                  DR. NORDEN:      No, I don't have anything to

20   add.

21                  DR. DANNER:      I'm not sure this is exactly

22   part of five, but I would like to see more information

23   on the metabolism of this drug.                The fact that it's

                               S A G CORP.
     202/797-2525              Washington, D.C.               Fax: 202/797-2525

 1   broken down by oxidants yet to be elucidated, I'm not

 2   sure I know what that means, and it makes me concerned

 3   that there might be interactions we don't know about

 4   with drugs that have oxidant or anti-oxidant activity,

 5   and whether there might be interactions with things like

 6   inhaled nitric oxide; that, you know, can things like

 7   that potentially attack these rings and break down the

 8   drug so that if you have somebody in the ICU on inhaled

 9   nitric oxide and this drug, they're essentially not

10   really getting the drug.

11                  I don't know if you know the answer to that,

12   but I think it's knowing more about how it's metabolized

13   and potentially interacts with things that affect

14   oxidant pathways, might help us, you know, know how to

15   use the drug in those unusual settings.

16                  DR. RODVOLD:    Well, the other things that

17   you're asking about, one of those things Dr. Murray

18   brought up before about absorption in the ICU.       I mean,

19   VRE people are going to be -- a lot of them are going

20   to be in the intensive care unit, and it would be good

21   to know whether or not you can dump this down an NG tube,

22   you know, enteral feedings, et cetera, et cetera, all

23   these other issues about viability and sorting out are

                             S A G CORP.
     202/797-2525            Washington, D.C.      Fax: 202/797-2525

 1   these the patients with the low concentrations on the

 2   end or not.

 3                   And so there's almost a little bit of need

 4   of   more      population     pharmacokinetics     and/or        some

 5   interaction studies, and I was a little surprised there

 6   was no drug interaction study that's ever been done with

 7   antacids, either out-patient or in-patient.                At least

 8   that's the impression I got.            So I mean, that's in all

 9   these indications.

10                   The other issue though is that this is VRE,

11   and it doesn't split out whether or not this is faecalis

12   versus faecium, and they had to really grasp the faecalis

13   data and pull it out as we asked for that, and it makes

14   me a little bit -- you know, the numbers are pretty low

15   on the faecalis side here.

16                   And so I think at minimum you need some more

17   post marketing faecalis data in this, and secondly, it

18   would be helpful for the prescriber to have some of that

19   put in the labeling either in the back of the studies

20   or something so that they actually know how many isolates

21   of    faecalis     have     been     treating   and      what     the

22   comfortability factor is on that because otherwise

23   you're blanketing this across VRE.

                                 S A G CORP.
     202/797-2525                Washington, D.C.          Fax: 202/797-2525

 1                  CHAIRMAN RELLER:         Dr. Chesney.

 2                  DR. CHESNEY:        I think maybe it would be

 3   helpful to break out the bacteremia data because we kind

 4   of had to find it and it was in 31 where we found the

 5   staph. bacteremia data, and I think that's always the

 6   most convincing for something new.

 7                  So I think somehow to present that up front

 8   or make that as a separate entity.

 9                  CHAIRMAN RELLER:           Other recommendations

10   from the committee members?           Dr. Murray?

11                  Oh,   one      is    not       compelled      to     have

12   recommendations, but there be any, we want to hear them.

13    Dr. Christie.

14                  DR. CHRISTIE-SAMUELS:            Thank you.

15                  I just felt we should collect as much data

16   as    possible   using      compassionate          use     protocols,

17   gathering data through historical controls, although

18   this is very difficult.

19                  In the pediatric population, I think in

20   newborns we can collect a lot of information, and in

21   the    transplant    population       and     in   those      who    are

22   bacteremic.

23                  And I don't know if it's possible to look

                                S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1   at a comparator trials with synercid and other agents

 2   as they become available.

 3                  CHAIRMAN RELLER:         Dr. Lowy?

 4                  DR. LOWY:     I guess I still worry about the

 5   hospital acquired pneumonias and the MRSAs and whether

 6   there was a way of getting cleaner data on MRSA

 7   pneumonias.

 8                  DR. LEGGETT:       I guess back to the point I

 9   made earlier in the day.         This is a drug that is going

10   to be used in the hospitals for nosocomial infections,

11   and yet I am not sure I was convinced about the severity

12   of the cases that were presented dealing with a static

13   drug when we may be dealing with line infections and

14   endocarditis that is not recognized, and I would like

15   to see more data and more serious patients, discitis

16   (phonetic), osteomyelitis, that kind of stuff.

17                  DR. MURRAY:        Just to point out that our

18   question asked for VR enterococci, whereas on page 6

19   in the sponsor's document they were separated out E.

20   faecalis, E. faecium, and I don't think in my opinion

21   that the data were sufficient for VRE faecalis.                So I

22   don't know if you wanted any comment on whether we would

23   split them out or not.

                            S A G CORP.
     202/797-2525               Washington, D.C.         Fax: 202/797-2525

 1                  And the other thing I would encourage the

 2   sponsor to do again relates to the wide diversity of

 3   levels that can be obtained because there may be some

 4   patients out there whereas in a population they're

 5   responding, but as an individual might fail because they

 6   are one of those that ends up with a lower blood level.

 7    So I would encourage some sort of follow-up study on

 8   that because it may benefit that one patient out of --

 9   I don't know -- 20 or 50 that would perhaps benefit from

10   a higher drug level.

11                  And then to go back to points we made about

12   earlier about metabolism, I assume there will be

13   follow-up studies on, again, looking more at patients

14   who are building up metabolites for their potential

15   clinical adverse events.

16                  CHAIRMAN    RELLER:           Dr.   O'Fallon,     any


18                  DR. O'FALLON:      I do like the idea of having

19   the so-called enrichment studies.            I think that's a very

20   good idea.     I just think that for any future studies

21   it would be helpful to break out that data by the

22   indications that we're being asked to evaluate.                 That

23   would have saved a lot of problems on my part anyway.

                               S A G CORP.
     202/797-2525              Washington, D.C.           Fax: 202/797-2525

 1                  In terms of did anyone ever look at the

 2   efficacy related to -- are they looking at -- I was

 3   interested in the fact that some of them were spiked

 4   and some of them were the long, sloping curve, area under

 5   the curve.     Has anyone looked at the efficacy, you know,

 6   by max. level as versus the area under the curve?

 7                  That was just thrown out there, but I

 8   wondered if it made a difference.

 9                  CHAIRMAN     RELLER:           When   the    clinical

10   studies were initiated for linezolid, there was not a

11   comparator available, and hence the clinical trial with

12   200 versus 600 milligrams to show a difference.

13                  I would very much like to see as part of

14   the plans for this compound in what the agency does in

15   its   final    assessment     is   to    see   future      data   with

16   comparative trials with the options that are now

17   available.

18                  The numbers of patients with vancomycin

19   resistant enterococci, albeit predominantly but not

20   exclusively     faecium,     are     increasing,      and    they're

21   substantial, and it looks like even with the trials that

22   were done, the numbers are almost as many as with less

23   than fully susceptible pneumococci.

                               S A G CORP.
     202/797-2525               Washington, D.C.           Fax: 202/797-2525

 1                  So this is, I think, an important issue to

 2   really get comparative clinical trials prospectively,

 3   you know, underway so that down the line we'll know.

 4                  And given the difference that looked like

 5   it was there with 600 versus 200, even the possibility

 6   of having a comparator and 200 and 600 to show that 200

 7   might be like the comparator, but the 600 is better than

 8   either or them if the numbers were larger, that's a study

 9   that I'd very much like to see.

10                  The difficulty with a straight comparison

11   of this agent versus another is always the nagging

12   suspicion that what would it be if you didn't do

13   anything, but if you can demonstrate 600 versus 200 and

14   the 200 looks more like a comparator, that would be very

15   telling information for the future, I think, especially

16   with patients with positive blood cultures, and with

17   the caveats that Dr. Leggett pointed out.

18                  Now,   we    probably          have   a   little     more

19   discussion that people would like to have.               Whether it's

20   in the form of a specific question or simply the kinds

21   of additional information or studies that the committee

22   members would like to see if there were for this or any

23   other agent to have specific indications for methicillin

                                S A G CORP.
     202/797-2525               Washington, D.C.              Fax: 202/797-2525

 1   resistant staphylococci.

 2                    Dr. Soper?     Chesney?

 3                    I'd like to see some additional clinical

 4   trials.        I mean I think everybody on this committee,

 5   all the clinicians on the committee have a sense that

 6   vancomycin       is    a   tolerable      drug       for    methicillin

 7   staphylococci         because   we     heretofore          didn't     have

 8   anything       else   really    to   use,      but    that    it's     not

 9   necessarily a very good drug for methicillin resistant

10   or    staphylococci,        period,      if    you     have      another

11   alternative and would prefer, as many have pointed out,

12   to use, for example, naphcillin for susceptible strain.

13                    So that comparative clinical trials with

14   the options for serious Staphylococcal infections, both

15   staph. susceptible and the increasing proportion that's

16   more than 50 percent in many hospitals, to really put

17   this and other agents to the test prospectively, all

18   at the same time in randomized trial, for example, three

19   currently available agents and do a blinded prospective

20   trial would be exceedingly good to know.

21                    Dr. Norden.

22                    DR. NORDEN:     Yeah, I would like to see this

23   drug tested in osteomyelitis.           I think it's in many ways

                                 S A G CORP.
     202/797-2525                Washington, D.C.               Fax: 202/797-2525

 1   an interesting agent.         It's been given for 28 days

 2   already, and with the only caveat being the platelet

 3   count, which you'd have to watch, but it's available

 4   orally.

 5                  For MRSA vancomycin is not a great drug

 6   osteomyelitis at all.       We don't know anything about

 7   synercid for osteomyelitis.        So I think this would be

 8   an inappropriate indication for the company to consider

 9   trials in.

10                  CHAIRMAN RELLER:     Any other comments?     Dr.

11   O'Fallon.

12                  DR. O'FALLON:    In looking at this data, we

13   don't know exactly.    Have you ever considered measuring

14   the time to failure, say, of these drugs as opposed to

15   just -- what would you call it? -- cure, yes/no?

16                  Well, that's fine, but have they ever looked

17   at time to cure or time to failure in these as a

18   measurement of the efficacy of these drugs comparing

19   drugs?

20                  DR. CHIKAMI:    In the usual studies that we

21   see, that we have reviewed for drug approval, we've not

22   seen those sorts of design, but I think you raise a good

23   point, particularly for and speaking generally for those

                             S A G CORP.
     202/797-2525            Washington, D.C.        Fax: 202/797-2525

 1   infections in which the real impact of antimicrobial

 2   therapy is to shorten the course of disease.

 3                   Then, in fact, a relevant measure may be

 4   time to resolution or cure.           Now, of course, that adds

 5   a technical complexity to the trials.            One needs to then

 6   define appropriately what that event is, what cure is,

 7   and then design the trial so that you capture the

 8   information in a relevant way so that, in fact, you can

 9   make reasonable comparisons in regard to the time to

10   event sort of analyses.

11                   But, I mean, I think it's a reasonable

12   suggestion.

13                   I think in those diseases where there's a

14   substantial mortality, for example, one could still make

15   those measurements, but I think ultimately the relevant

16   impact we want to have on those sorts of diseases is

17   the overall impact on mortality, for example, in sepsis

18   trials where the primary thing we're looking at there

19   is the 28 day, all cause mortality.

20                   Well,      for   example,      serious   and     life

21   threatening infections, such as meningitis, where we

22   not   only     look   at   survival     and    resolution   of   the

23   infection, but also the incidence of serious neurologic

                                 S A G CORP.
     202/797-2525                Washington, D.C.         Fax: 202/797-2525

 1   sequelae at six months, for example.

 2                   CHAIRMAN RELLER:      Dr. Drusano.

 3                   DR. DRUSANO:     Just as a commentary to Dr.

 4   O'Fallon,      I'm   currently    involved   in   advising       a

 5   particular sponsor looking at bacteremia, getting blood

 6   cultures every day and looking at the time to clearance

 7   of the blood culture and then looking at covariates like

 8   measure of drug exposure relative to the MIC, AUC-MIC,

 9   peak to MIC, time above MIC as covariates in a Cox

10   proportional hazards analysis.

11                   We've also done this and published this in

12   the journal AIDS.        There was an analysis that was

13   requested by the FDA for the time to progression of CMV

14   retinitis with the drug phoscarnate.

15                   So this is a very powerful way of being able

16   to get more information about how well a drug actually

17   does its job, and it's much more sensitive than using

18   Lodgets (phonetic) or just any other kind of yes/no kind

19   of analysis.

20                   CHAIRMAN RELLER:     We've just heard from Dr.

21   George Drusano who's a professor of pharmacology and

22   infectious disease clinician in Albany Medical College.

23                   Are there other comments or questions?

                              S A G CORP.
     202/797-2525             Washington, D.C.         Fax: 202/797-2525

 1                        Dr. Chikami, have we done our job and is

 2   there any other thing that you want to ask of us?

 3                        DR. CHIKAMI:    No.

 4                        (Laughter.)

 5                        DR. MURPHY:    Barth, but we do want to say

 6   thank you.

 7                        (Laughter.)

 8                        DR. MURPHY:    And that we do think we are

 9   moving this field forward as we develop these products.

10    It has been a very difficult row to hoe here with not

11   having standards as comparators and having to do

12   standards of cares, dose response type of studies, and

13   we think that we are continuing to move.

14                        Some of the suggestions were very helpful,

15   and we do thank you for your input.

16                        CHAIRMAN RELLER:      In closing, I'd like to

17   thank          all      of    the       presenters,     especially

18   Pharmacia/Upjohn for their collegial addressing all of

19   the questions posed and then some into those from the

20   FDA in their presentations to try to do justice to the

21   importance of the topic at hand.

22                        Thank you and have a safe journey home.

23                        (Whereupon, at 4:01 p.m., the meeting was

                                  S A G CORP.
     202/797-2525                  Washington, D.C.        Fax: 202/797-2525

 1   concluded.)
















                    S A G CORP.
     202/797-2525   Washington, D.C.   Fax: 202/797-2525

To top