Drug Induced Liver Injury: Implications in drug discovery and development
Paul B. Watkins University of North Carolina Chapel Hill, N.C.
Drug Induced Liver Injury (DILI) is Hot
FDA / Pharma steering committee Several Critical Path Initiatives $ millions spend in industry New Network (DILIN) SAE Consortium
Industry SAE Priorities 2006
Rank Order [1 highest to 5 lowest]
Your Company's Priority
Overall Priority
Variance
Variance
Hepatotoxicity QT Prolongation Rhabdomyolosis Serious Skin Rashes [SJS] Edema
1.1 2.6 3.3 3.5 4.4
low moderate moderate high high
1.2 2.5 3.5 3.4 4.5
low high mod high high
SAE Consortium Survey – courtesy of Arthur Holden
Troglitazone (Rezulin®)
1). PPARg agonist 2). Treats type 2 diabetes 3). Caused liver failure
Presentation by Dr. Mark Pierce (Parke-Davis Pharmaceutical Research)
Overall Post-Marketing Reporting Death/Transplant Rate March 1997 - March 1999
35 in 1.58 million = 1 in 45,098
Background incidence of liver failure with no known cause ~ 1 in 1 million
March 26th 1999 - Troglitazone - FDA Advisory Panel
Troglitazone (Rezulin®)
1). Acute liver failure reports continued despite warnings and monitoring recommendations 2). Second in class (2) came on the market and appeared to be safer
3). Withdrawn from the market
Troglitazone (Rezulin®)
The Rise and Fall of the Killer Drug Rezulin Los Angeles Times, June 4, 2000, p.1A.
“…a disparate collection of physicians inside the U.S. Food and Drug Administration waged a remarkable revolt that … combined meticulous research and bluntly worded e-mail messages to upbraid their government superiors for contributing to the needless deaths of patients.”
Why clinical drug development programs were terminated in 1991
% of total terminations
45 40 35 30 25 20 15 10 5 0
Un kn ow n/ ot he r Ef f ic ac y fg oo ds er cia l
Series1 Series1 Series2
PK /b io av ai la Sa bi lit fe y ty /T ox ic ol og y
Co m m
Nature Reviews: Drug Discovery, Aug, 2004
Co st o
Why clinical drug development programs were terminated in 2000
% of total terminations
45 40 35 30 25 20 15 10 5 0
Un kn ow n/ ot he r Ef f ic ac y fg oo ds er cia l
Series1 1991 Series2 2000
PK /b io av ai la Sa bi lit fe y ty /T ox ic ol og y
Co m m
Nature Reviews: Drug Discovery, Aug, 2004
Co st o
“Hepatotoxicity has been the most common single adverse effect causing major drug problems, including withdrawals and refusals to approve”
Bob Temple, M.D. FDA 2/15/01
2006 State of the Art How to avoid hepatotoxicity in drug development
1). Avoid certain molecular structures
Compound Pair
Ibuprofen
“Clean” Compound
HO O
Ibufenac
“Toxic” Compound*
HO
O
*withdrawn from the market in the 1960’s because of clinical liver toxicity
2006 State of the Art How to avoid hepatotoxicity in drug development
1). Avoid certain molecular structures 2). Target daily dose to < 10 mg/day
2006 State of the Art How to avoid hepatotoxicity in drug development
1). Avoid certain molecular structures 2). Target daily dose to < 10 mg/day 3). Low covalent binding in liver microsomes 4). Low production of glutathione conjugates
2006 State of the Art How to avoid hepatotoxicity in drug development
1). Avoid certain molecular structures 2). Target daily dose to < 10 mg/day 3). Low covalent binding in liver microsomes 4). Low production of glutathione conjugates 5). Low incidence (<5%) of ALT > 3 X ULN in clinical trials.
Example: Acetaminophen
1). Avoid certain molecular structures - NO 2). Target daily dose to < 10 mg/day – 4 grams/day 3). Low covalent binding in liver microsomes – NO 4). Low production of glutathione conjugates – NO
5). Low incidence (<5%) of ALT - NO
Watkins et al. ALT vs. time for APAP Group (n = 26), ULN = 40 Solid Line = on treatment, Dotted Line = off treatment
640 16 X
560
APAP Mean ALT 3 X ULN R eference
14 X
480
12 X
320
8X
240
6X
160
4X
80
2X
0 0 5 10 15 Time (days) 20 25 30
0X
JAMA, 392:87,2006
Fold ULN for ALT
400
10 X
ALT
Example: Acetaminophen
1). Avoid certain molecular structures - NO 2). Target daily dose to < 10 mg/day – 4 grams/day 3). Low covalent binding in liver microsomes – NO 4). Low production of glutathione conjugates – NO
5). Low incidence (<5%) of ALT - NO
Drug Induced Liver Injury (DILI)
1). Thorn in the side of drug development. 2). High priority to design out of drugs. 3). Little progress made to date.
Drug Induced Liver Injury (DILI) can mimick every known liver disease
Cholestasis (&vanishing bile duct syndrome) Steatosis (micro and macrovesicular) Phospholipidosis Veno-occlusive disease Occult fibrosis/ cirrhosis Liver cancer Acute hepatocellular injury – High ALT/AST
Regulatory actions due to DILI (1995-2006)
Withdrawals bromfenac troglitazone pemoline Second Line felbamate tolcapone trovafloxacin Warnings
acetaminophen leflunomide nefazodone nevirapine pyrazinamide/rifampin terbinafine valproic acid zifirlukast atomoxetine interferon 1b –1b and 1a saquinavir infliximab bosentan telithromycin (kava, lipokinex)
http://www.fda.gov/medwatch/safety.htm
Regulatory actions due to DILI (1995-2006)
Withdrawals bromfenac troglitazone pemoline Second Line felbamate tolcapone trovafloxacin Warnings
acetaminophen leflunomide nefazodone nevirapine pyrazinamide/rifampin terbinafine valproic acid zifirlukast atomoxetine interferon 1b –1b and 1a saquinavir infliximab bosentan telithromycin (kava, lipokinex)
http://www.fda.gov/medwatch/safety.htm
Of the 23 drugs/CAM that have undergone withdrawal, restriction or warnings
19/23 (82%) were associated with acute idiosyncratic hepatocellular injury
“idiosyncracy”
(Hippocrates, ~400 B.C.)
(idios) - one’s own, self (syn) - together (crasis) - a mixing, mixture therefore
a person’s own mixture of characteristics, factors, nature and nurture, uniquely
John Senior - FDA
Liver injury ( ALT) safe
SAFE
Concept of idiosyncratic hepatocellular injury
death
jaundice
enceph Days on drug
Challenges in identifying factors underlying susceptibility to DILI
1). How to identify susceptible individuals. 2). What to do with them once you have them.
Liver injury ( ALT) safe
SAFE
Concept of idiosyncratic hepatocellular injury
Selection of patients based on serial ALT values in a clinical trial
1.34
C.V.
CV ALT
Non-susceptible
cases 0.34 0.23 0.16 0.03 0.18 0.58 0.77 1.10 controls 34.77
susceptible
Maximum ALT, fold Upper Limit of Normal (logarithmic scale)
ULN
A genetic test that predicts ALT elevations:
1). Would obviate need for ALT monitoring.
2). Would be useful in developing next in class drugs.
3). May provide only limited insight into mechanisms of idiosyncratic severe DILI.
Problem with ALT elevation as the endpoint
1). Occurs with drugs that do not have clinically important liver toxicity
2). Usually reverse with continued treatment even with drugs that can cause acute liver failure.
Incidence of ALT elevations (>3X ULN) and clinical hepatitis
ALT hepatitis
troglitazone 2% INH 15% diclofenac 3%
<0.1 <1% <0.01%
Treatment with tacrine through ALT elevations
unpublished
Reversed on treatment
Treatment stopped
unpublished
Reversal of rat liver necrosis with continued exposure to BDCM
1 week
3 weeks
Toxicol. Sci. 64:268 (2001)
Possible explanations for reversibility of ALT elevations
1). ALT elevations that reverse on treatment have no relationship to those that can progress to liver failure. 2). A subset of those with ALT elevations can progress on to liver failure (i.e. those who can not adapt).
Safe pathways Drug elimination
Reactive Metabolite
ALT elevations
Progressive injury
Adaptation
X
jaundice
Increased ALT
liver failure
safe
SAFE
Concept of idiosyncratic hepatocellular injury
APAP
safe elimination
NAPQI
Covalent binding/oxidative stress resolution progression
Effect of 8 days APAP pretreatment (---) on single dose toxicity in mice
Hepatology 29:436, 1999.
3-Cys-APAP adducts (brown) 2 hours after single toxic APAP dose
Saline pretreatment
APAP pretreatment
Changes in APAP metabolism that reduce toxicity APAP elimination
CYP2E1 CYP1A2 CYP2B
ROS (Nrf-2) Regeneration Acute phase (IL-6)
GST GSH
NAPQI
Transporters during recovery from APAP hepatotoxicity
TNFa
X
X
Transporters during recovery from APAP hepatotoxicity
Ntcp and Mrp4 expression 48 hours after APAP
Alkunes and Manatou, unpublished observations
MDR1 (P-glycoprotein) expression MDR1 In submassive necrosis (human) Normal liver
necrosis
J Pathol 200:553, 2003
MRP3 expression in submassive necrosis (human) Normal liver
MRP3
necrosis
J Pathol 200:553, 2003
Conclusion
Adaptation to liver toxicity can involve:
a). Down regulation of CYPs and uptake transporters
b). Upregulation of glutathione and efflux transporters
Serial ALT in healthy woman receiving APAP 1 g qid X 7 days
ALT (U/L)
140 120 100 80 60 40 20 0
ffo l o lllo w o w-up up 1 3 5 7 9 1 1 13 13
1st study 1st study 2nd study
Day
Unpublished data
Conclusions
• It is adapation to toxicity. • Arguably the most important issue in idiosyncrasy.
– a). Determines whether patient gets sick – b). Implications for monitoring – b). Susceptibilities may not be drug specific
• Current concepts do not account for the “memory”.
Where do we go from here?
The most appropriate model for studying idiosyncratic hepatotoxicity are the people who actually experienced this.
Selection of cases and controls from serial ALT values in a clinical trial
1.34
C.V.
CV ALT
Non-susceptible
cases 0.34 0.23 0.16 0.03 0.18 0.58 0.77 1.10 controls 34.77
susceptible
Maximum ALT, fold Upper Limit of Normal (logarithmic scale)
ULN
A cooperative agreement funded by the Division of Digestive Diseases and Nutrition,
National Institute of Diabetes and Digestive and Kidney Diseases
http://dilin.dcri.duke.edu/
Sphere of Influence
12.8 million lives
http://dilin.dcri.duke.edu/
Resources Created by DILIN
1). Genomic DNA bank. 2). Immortalized lymphocyte bank.
3). Registry of subjects.
Final take home points
1). The DILIN network represents the best opportunity to date to identify mechanisms underlying severe idiosyncratic DILI.
2). Research utilizing the resulting resources will be challenging.
Idiosyncratic hepatocellular injury due to drugs is a model for all environmental disease
1). Large population with known “exposure” to a defined xenobiotic (the drug). 2). Biomarkers that are cheap, safe, and sensitive.