CENTER FOR DRUG EVALUATION AND
`RESEARCH
`
`
`
`APPLICATION NUMBER:
`
`022406Orig1s000
`
`RISK ASSESSMENT and RISK MITIGATION
`REVIEW(S)
`
`
`
`
`
`

`

`
`
`Date:
`To:
`
`Through:
`
`From:
`
`Subject:
`
`Drug Name(s):
`Submission Number:
`Application
`Type/Number:
`Applicant/sponsor:
`OSE RCM #:
`
`Department of Health and Human Services
`Public Health Service
`Food and Drug Administration
`Center for Drug Evaluation and Research
`Office of Surveillance and Epidemiology
`
`
`
`February 13, 2009
`Rafel (Dwaine) Rieves, M.D., Director
`Division of Medical Imaging and Hematology Products (DMIHP)
`Solomon Iyasu, M.D., M.P.H. Director
`Division of Epidemiology (DEPI)
`Office of Surveillance and Epidemiology (OSE)
`Kate Gelperin, M.D., M.P.H.
`Medical Epidemiologist
`Division of Epidemiology
`Ongoing evaluation of potential severe liver injury signal in
`rivaroxaban clinical trials
`Rivaroxaban, BAY 59-7939
`NDA 22-406 submitted in July 2008
`NDA 22-406
`
`Bayer/Johnson & Johnson
`2008-2019
`
`
`
`
`
`

`

`
`
`1
`INTRODUCTION
`This review follows a request from the Division of Medical Imaging and Hematology Products
`(DMIHP) to review and comment on a potential signal for severe drug-induced liver injury
`identified by the OND medical reviewer during the mid-cycle review process, and to provide
`relevant background information regarding previous regulatory experience with hepatotoxicity
`signal detection, assessment, and subsequent considerations of the balance of potential
`therapeutic benefit(s) versus defined hepatotoxicity risk(s).
`Rivaroxaban (BAY 59-7939) is a highly selective direct factor Xa inhibitor with oral
`bioavailability. There are three active INDs for rivaroxaban: IND 64,892 (VTE); IND 75,931
`(ACS); and IND 75,238 (A Fib). The proposed indication for the current application is
`prophylaxis of deep vein thrombosis (DVT) and pulmonary embolism (PE) in patients
`undergoing hip replacement surgery or knee replacement surgery. The proposed dose is 10 mg
`once daily.
`
`2 MATERIAL REVIEWED
`
`The following materials were considered for this review:
`• Dr. Min Lu’s FDA mid-cycle clinical review slides dated December 2, 2008
`• Proposed package insert dated July 28, 2008
`• Sponsor’s laboratory datasets submitted to FDA January 22 and 30, 2009
`• Cases reviewed by Sponsor’s expert panel (LAP), Miami, February 17-18, 2008
`• Sponsor’s ISLS 6-month Safety Update dated February 2, 2008; Document No. EDMS-
`PSDB-9405338:2.0
`
`RESULTS OF REVIEW
`
` 3
`
`
`
`
`3.1 Overview of Clinical Program
`
`The rivaroxaban clinical program (excerpted from Dr. Min Lu’s mid-cycle review slides with
`cut-off date September 10, 2008) includes the following:
`• Completed studies: N=10,600 (Rivaroxaban exposure)
`o 4 phase 3 studies (RECORD 1-4): n=6183
`o 9 phase 2 studies: n=3300 (2 VTE Tx and 3 AF)
`o 51 phase 1 studies: n=1117
`• Ongoing studies: N=16,965 enrolled (as of September 10, 2008); N=34,236
`planned
`o 5 phase 3 studies:
`(cid:131) 2 VTE Tx: n=3160 enrolled, n=7500 planned
`(cid:131) 2 AF: n=10,008 enrolled, n=15,200 planned
`(cid:131) 1 Medically ill: n=316 enrolled, n=8000 planned
`o 1 phase 2 study: ACS n=3462 enrolled, n=3500 planned
`o 1 phase 1 study: CHF n=19 enrolled, n=36 planned
`
`
`
`1
`
`

`

`
`
`FDA Safety Concerns – Potential Severe Liver Injury
`
`
`3.2
`
`3.2.1 Safety issue identified during rivaroxaban mid-cycle review
`
`The DMIHP medical officer’s mid-cycle review identified a major concern with potential
`severe and/or fatal drug-induced liver injury with rivaroxaban. In the completed studies, severe
`liver injury (defined as a concurrent increase of total bilirubin [TBL] >2x ULN and alanine
`aminotransferase [ALT] >3x ULN) was observed in 14/9310 (0.15%) rivaroxaban-treated
`patients, and 9/7001 (0.13%) patients in comparator groups, as described in Dr Lu’s review.
`Seven cases of severe liver injury in the RECORD studies were considered to be possibly
`related to rivaroxaban therapy by at least one member of the sponsor’s expert panel of
`hepatologists.
`Members of the sponsor’s expert panel of hepatologists considered that some cases of severe liver
`injury in completed and ongoing clinical trials, including at least two deaths, were possibly
`related, or of uncertain relationship to rivaroxaban. As presented in the mid-cycle clinical review,
`at least two cases of fatal liver injury for which a possible contribution of rivaroxaban has not
`been ruled out occurred after fewer than 30 days of drug exposure.
`
`3.2.2 Previous FDA experience with signal detection for severe liver injury with
`anticoagulant drug development for short-term versus long-term indications
`Previous FDA experience with assessment of severe drug-induced liver injury due to
`ximelagatran, an anticoagulant drug (direct thrombin inhibitor) developed for similar indications,
`found no cases of severe liver injury in the short-term (orthopedic) clinical trials; however, a
`strong signal was subsequently identified in long-term (atrial fibrillation) trials.
`After full evaluation of the signal, it was determined that 37/6948 (0.5%) ximelagatran-treated
`patients experienced severe liver injury versus 5/6230 (0.08%) patients randomized to warfarin
`(relative risk 6.6; 95% confidence interval 2.6 – 16.9). An expert causality assessment of severe
`liver injury cases was conducted by the sponsor, and determined that study drug may have caused
`or contributed to the severe liver injury in 19/6948 ximelagatran-treated patients compared to
`2/6230 patients in the comparator groups (relative risk 8.5; 95% confidence interval 2.0 – 36.6).
`Although a signal for severe liver injury was not detected in short-term orthopedic trials with
`ximelagatran, analysis of long-term data showed that initial signs of liver injury were observed
`within the first 30 days of ximelagatran administration for six study subjects who went on to
`develop severe liver injury, of which four cases were considered by the sponsor to be causally
`related to ximelagatran administration.
`A full consideration of the balance of drug benefit(s) versus risk of severe or fatal liver injury was
`conducted at a public Advisory Committee meeting, which determined that potential benefits of
`ximelagatran did not outweigh the risks. Based on this decision, the drug was not approved in the
`US, and subsequently the sponsor decided to withdraw ximelagatran from marketing worldwide.
`
`
`
`
`2
`
`

`

`
`
`3.3
`
`Laboratory Datasets from Ongoing Rivaroxaban Clinical Trials (blinded data)
`
`Initial inspection of clinical laboratory datasets by Dr. Ted Guo (biostatistician) from ongoing
`clinical trials received from the sponsor on January 30, 2009 show the following counts of
`cases (numbers of patients) of potential severe liver injury in ongoing clinical trials (defined as
`concurrent maximum ALT >3x ULN and maximum TBL >2x ULN). Please note that there
`were multiple measurements over the course of the trial for each patient. The greatest values of
`ALT and TBL of the patient were used to determine potential severe liver injury. Missing data
`in ALT and TBL did not affect the values of the maximum ALT and TBL. The effect of
`missing data has not been investigated. Therefore, these patient counts are preliminary, and
`current findings could potentially be somewhat biased.
`
`Open-label long-term EINSTEIN DVT/PE (Study #11702) – ongoing
`
`Treatment
`
`BAY 59-7939
`
`ENOXAPARIN 1 mg/kg s.c.
`/ Vitamin K-antagonist p.o.
`
`#Patients
`
`Mean Rx Duration
`in days
`
`# Cases potential
`severe liver injury in
`available data
`
`1682
`
`1673
`
`150
`
`154
`
`3
`
`1
`
`Blinded long-term ROCKET-AF (Study #11630; comparator warfarin) – ongoing
`
`Treatment labeled as
`
`#Patients
`
`Mean Rx Duration
`in days
`
`# Cases potential
`severe liver injury in
`available data
`
`Dummy A (BLINDED)
`
`Dummy B (BLINDED)
`
`5495
`
`5492
`
`233
`
`229
`
`8
`
`12
`
`Blinded long-term J-ROCKET-AF (Study #12620; comparator warfarin) – ongoing
`
`Treatment labeled as
`
`#Patients
`
`Mean Rx Duration
`in days
`
`# Cases potential
`severe liver injury in
`available data
`
`BLINDED
`
`
`
`
`
`1185
`
`201
`
`3
`
`3
`
`

`

`
`
`4
`
`DISCUSSION
`
`Instances in clinical trials (even very few of them) of transaminase elevation accompanied by
`elevated bilirubin (in the absence of biliary obstruction), have been associated with, and have
`often predicted, post-marketing serious liver injuries (fatal or requiring transplant). Drug-induced
`hepatocellular jaundice is considered a serious lesion, with an estimated mortality of at least 10%.
`The reason is that hepatocellular injury great enough to interfere with bilirubin excretion involves
`a large fraction of the liver cell mass.
`An FDA Office of Drug Safety review which was included in the background package for the
`ximelagatran Advisory Committee meeting in September 2004 is included for reference as an
`appendix to this memo.
`
` 5
`
`
`
` A
`
`RECOMMENDATIONS
`
` potential signal for severe liver injury associated with rivaroxaban therapy has not been fully
`characterized at this time. Complete risk assessment, fully evaluating safety data from long
`term clinical trials, should be undertaken in order to inform decisions about the balance of
`therapeutic benefit versus risk with rivaroxaban.
`
`
`
`
`
`4
`
`

`

`
`
`
`
`APPENDIX
`
`5
`
`24 Pages Withheld in Full as B4 (CCI/TS)
`Immediately Following this Page
`
`

`

`Brief regulatory history: withdrawals and risk management
`
` APPENDIX—DRUG-INDUCED LIVER INJURY
`
`A.
`
`During the past ten years, two drugs, DURACT (bromfenac) and REZULIN (troglitazone), have
`been withdrawn from marketing in the US because they were associated with an unacceptable
`risk of severe drug-induced liver injury (DILI) in the absence of a clear counter-balancing
`benefit. In both cases, attempts were made to manage the risk of hepatotoxicity while keeping
`the drug available for therapeutic use. In the case of bromfenac, approved by FDA in 1997 for
`use as a short-term analgesic (ten days or less), severe DILI was generally observed only in
`patients who took the drug for more than 30 days;47 however, despite attempts to regulate the
`duration of therapy by clear statements in product labeling, prescribers did not adequately heed
`this information and more than 50 cases of severe DILI were reported, prompting market
`withdrawal in 1998. In the case of troglitazone, approved by FDA in 1997 for glucose control in
`patients with type 2 diabetes, reports of fatal liver injury received by FDA shortly after
`marketing prompted a black box warning and a series of Dear Healthcare Professional letters
`recommending monthly transaminase monitoring. Despite these measures, transaminase
`monitoring was not regularly performed.48 Moreover, in some patients, liver injury still
`progressed to fatal liver failure despite stopping the drug in response to monthly transaminase
`monitoring due to rapid progression of liver injury to a state of irreversibility.49 Troglitazone was
`withdrawn from the US market in March 2000, after 94 cases of drug-induced liver failure had
`been reported, most of which were fatal. A more complete discussion of troglitazone is provided
`in Section D of this Appendix, under the heading Specific Examples.
`
`Also during the past ten years, there have been instances where regulatory action prompted by
`concern about severe DILI included risk management actions which stopped short of market
`withdrawal. Examples include CYLERT (pemoline) and TROVAN (trovafloxacin).
`
`Pemoline was approved by FDA in 1975 for ADHD with recommendations in the Precautions
`section to monitor transaminase levels periodically due to a 1% to 2% incidence of drug-induced
`liver injury. Reports of acute liver failure (ALF) led to a series of black box warnings and Dear
`Healthcare Professional letters in 1996 and 1999, shifting the drug to second line therapy and
`recommending baseline and bi-weekly transaminase monitoring. Although compliance with
`these recommendations was assessed to be poor,50 the use of pemoline dropped off substantially
`over the next five years,51 and no additional drug-related cases of liver failure were subsequently
`reported to FDA.52
`
`47 Fontana RJ, McCashland TM, Benner KG, et al. Acute liver failure associated with prolonged use of bromfenac
`leading to liver transplantation. Liver Transpl Surg 1995;5:480-4.
`48 Graham DJ, Drinkard CR, Shatin D, Tsong Y, Burgess M. Liver enzyme monitoring in patients treated with
`troglitazone. JAMA 2001;286:831-33.
`49 Graham DJ, Green L, Senior JR, Nourjah P. Troglitazone-induced liver failure: a case study. Am J Med
`2003;114:299-306.
`50 Willy ME, Manda B, Shatin D, Drinkard CR, Graham DJ. A study of compliance with FDA recommendations for
`pemoline (Cylert). J Am Acad Child Adolesc Psychiatry 2002, 41(7):785-790.
`51 FDA/CDER/ODS/DSRCS Review of the Proposed Risk Management Communication Plan for Cylert (pemoline)
`dated January 16, 2004.
`52 Racoosin JA. FDA/CDER/Division of Neuropharmacological Drug Products (HFD-120) memorandum to Patient
`Information Sub-Committee Members, dated February 6, 2004.
`
`
`
`25
`
`

`

`
`Trovafloxacin (a fluoroquinolone antibiotic) received FDA approval in 1997. During the first
`two years of marketing in the US, there were over 100 cases of clinically symptomatic liver
`toxicity, including 14 cases of ALF. An analysis of drug utilization based on data from IMS
`Health, National Disease and Therapeutic Index™ (NDTI™)53 showed that during the period
`from 1998 to 1999, approximately 91% of trovafloxacin prescriptions were for five days or
`longer, with only about 5% of prescriptions for 20 days or longer.54 A lag was noted between
`completion of trovafloxacin therapy and onset of liver symptoms in six of 14 probable ALF
`cases, which ranged from five to 20 days.55 Survival analysis was conducted on the spontaneous
`reports, and showed that the relative risk of ALF with trovafloxacin was elevated from the start
`of therapy, and increased with increasing duration of exposure.56 A Public Health Advisory in
`1999 warned about severe hepatotoxicity, restricted use to certain very severe infections, and
`announced that the manufacturer would restrict trovafloxacin distribution to inpatient facilities
`only.57
`
`Examples of drugs never marketed in the US because of hepatotoxicity include ibufenac,
`perhexilene, dilevalol (a beta blocker), tasosartan (an angiotensin II receptor antagonist), and
`Fialuridine (FIAU).58 In the case of dilevalol, the application was refused in 1990 based on
`findings of >3x ULN transaminase elevations and modest bilirubin elevation (>2 mg/dL) in a
`few patients, accompanied by an increased incidence of 3-fold transaminase elevation compared
`to placebo.59
`
`B.
`
`Drug-induced liver injury is an important cause of fulminant liver failure. The Acute Liver
`Failure Study Group found that, between 1998 and 2000, 52% of all cases of ALF in the United
`States were due to drug-induced liver injury.60
`
`Drug-induced liver disease can be predictable (dose-related, occurring at doses exceeding
`recommendations) or unpredictable (idiosyncratic, and occurring in susceptible individuals at
`usual therapeutic doses).61 Idiosyncratic liver injuries occur with a pattern that is consistent for
`each drug and for each drug class.62
`
`
`Range of issues: timing, tempo and reversibility of hepatotoxicity
`
`
`53 IMS Health, National Disease and Therapeutic Index, 1998-March 1999, extracted 6/99.
`54 FDA/CDER/OPDRA/DDRE Review of Trovan® (trovafloxacin and alatrofloxacin) and acute liver failure, dated
`July 12, 1999.
`55 ibid.
`56 ibid.
`57 Public Health Advisory (1999) Trovan (trovafloxacin / alatrofloxacin mesylate) and risk of liver failure. FDA
`June 9, 1999. Available from: http://www.fda.gov/cder/news/trovan/trovan-advisory htm
`58 Center for Drug Evaluation and Research (CDER). Drug-induced liver toxicity. Clinical White Paper. November
`2000. (Accessed June 1, 2004, at http://www fda.gov/cder/livertox/default htm.)
`59 ibid.
`60 Ramkumar D, LaBrecque DR. Drug-induced liver disease and environmental toxins. In: Hepatology A Textbook
`of Liver Disease. Fourth Edition. Zakim D and Boyer TD, (Eds.) Saunders, Philadelphia, 2003.
`61 ibid.
`62 Lee WM. Drug-induced hepatotoxicity. N Engl J Med 2003;349:474-85.
`
`
`
`26
`
`

`

`As Lee has proposed in a recent review of drug-induced liver injury,63 most idiosyncratic drug
`reactions result from a succession of unlikely events, a “multihit” process. This implies that a
`“series of events that first involve intracellular disruption, cell necrosis, or apoptosis, followed by
`activation of the immune sequence, might explain the features of idiosyncratic drugs reactions:
`their rarity, their severity, and their resolution despite continued use of the drugs by patients with
`phenotypes that appear to be adaptive.”64
`
`Timing: Risk vs. duration of treatment (hazard rate over time)
`
`Idiosyncratic reactions are characterized by a variable delay or latency period, typically ranging
`from 5 to 90 days from the initial ingestion of the drug, and are frequently fatal if the drug is
`continued once the reaction has begun.65 The relationship of onset of liver injury with duration
`of drug exposure is not predictable. An increased risk of severe DILI has been found to be
`positively associated with increasing duration of therapy for several drugs including
`trovafloxacin,66 troglitazone,67 pemoline,68 and bromfenac.69
`
`Tempo and reversibility of injury
`
`The range of tempos of injury is a characteristic both of individual drugs and patients. Rapid
`acceleration of liver injury in some individuals may preclude an absolute protective value of
`standardized periodic transaminase monitoring.70
`
` key issue in effective intervention to prevent fatal liver injury is “recoverability” at time of
`sign or symptom onset. This refers to a “point of irreversibility”, after which there is an
`inexorable progression to liver failure and often death. The contrast between isoniazid liver
`injury (chronic parenchymal injury)71 and that characteristic of troglitazone72 demonstrates the
`contrast between a situation where stopping the drug at the time of symptom onset most often
`prevents progression to irreversible injury, and one where it does not in many cases. Drugs that
`can cause severe DILI generally demonstrate a range of responses, with varying proportions of
`patients who recover whether or not the drug is stopped, versus the proportion of patients who go
`on to develop irreversible injury.
`
`
`
`63 ibid.
`64 ibid.
`65 ibid.
`66 Graham DJ, Ahmad SR, Piazza-Hepp T. (2002) Spontaneous reporting – USA. In: Mann RD and Andrews EB,
`(eds), Pharmacovigilance, John Wiley and Sons, Ltd.
`67 Graham DJ, Green L, Senior JR, Nourjah P. Troglitazone-induced liver failure: a case study. Am J Med
`2003;114:299-306.
`68 Safer DJ, Zito JM, Gardner JE. Pemoline hepatotoxicity and postmarketing surveillance. J Am Acad Child
`Adolesc Psychiatry. 2001 Jun;40(6):622-9.
`69 Fontana RJ, McCashland TM, Benner KG, et al. Acute liver failure associated with prolonged use of bromfenac
`leading to liver transplantation. Liver Transpl Surg 1995;5:480-4.
`70 Avigan M. Responses to a signal of drug-induced hepatotoxicity. FDA/CDER/ODS/DDRE presentation at Drug-
`Induced Hepatotoxicity Workshop, January 28, 2003, Washington, DC.
`71 Ramkumar D, LaBrecque DR. 2003. op cit.
`72 Graham DJ, Green L, Senior JR, Nourjah P. 2003. op cit.
`
` A
`
`
`
`27
`
`

`

`Experience with clinical trial data
`
`Dose-related hepatotoxicity
`
`Acetaminophen is an example of a drug with predictable dose-related toxic effects. At higher
`doses, acetaminophen can rapidly cause hepatocyte injury. Acetaminophen toxicity produces the
`most common form or cause of ALF in the US, accounting for 39% of cases in a recent survey of
`tertiary care centers,73 both after attempted suicide by acetaminophen overdose and after
`unintentional overdose, in which use of the drug for pain relief in excess of the recommended
`dose has occurred over a period of days.74
`
`C.
`
`Possible “signals” for severe DILI are abnormalities (signs or symptoms) that reflect ongoing
`liver injury 1) in the same individual if drug is continued, and 2) in other drug-treated individuals
`due to a common mechanism of toxicity.75 Signals can be generated in clinical trials by subjects
`with clinically mild reversible drug-induced liver injury.
`
`The observation that “instances (even very few of them) of transaminase elevation accompanied
`by elevated bilirubin (even if obvious jaundice was not present) have been associated with, and
`have often predicted, post-marketing serious liver injuries (fatal or requiring transplant)” was
`first made by Dr. Hyman Zimmerman,76 and has been dubbed “Hy’s Law”.77 The ominous
`implications of Hy’s Law proved to be true for bromfenac, dilevalol, troglitazone, and
`trovafloxacin, even though no cases of life-threatening serious injury were seen for any of these
`drugs pre-marketing.78
`
`Zimmerman noted that drug-induced hepatocellular jaundice is a serious lesion, with mortality
`ranging from 10 to 50 percent.79 More recent mortality estimates continue to regard the
`combination of pure hepatocellular injury and jaundice as ominous, with about 10-15% of
`patients who show such findings as a result of drug-induced injury going on to die80. The
`explanation for this outcome is that hepatocellular injury great enough to interfere with bilirubin
`excretion must involve a large fraction of the liver cell mass.81
`
`Increased transaminases alone – examples
`
`Clinical trials with statins have generally shown an imbalance in transaminase elevations (ALT
`>3x ULN) between active drug and placebo. However, extensive marketed experience with the
`
`
`73 Ostapowicz G, Fontana RJ, Schiodt FV, et al. Results of a prospective study of acute liver failure at 17 tertiary
`care center in the United States. Ann Intern Med 2002;137:947-54.
`74 Lee WM. 2003. op cit.
`75 Avigan M. 2003. op cit.
`76 Center for Drug Evaluation and Research (CDER). Drug-induced liver toxicity. Clinical White Paper. November
`2000. (Accessed June 1, 2004, at http://www fda.gov/cder/livertox/default htm.)
`77 Reuben A. Hy’s Law. Hepatology 2004 Feb;39(2):574-8.
`78CDER. Drug-induced liver toxicity. 2000. op cit..
`79 Zimmerman HJ. Drug-induced liver disease. In: Hepatotoxicity The Adverse Effects of Drugs and Other
`Chemicals on the Liver. Appleton-Century-Crofts, New York, 1978, 1999.
`80 CDER. Drug-induced liver toxicity. 2000. op cit.
`81 ibid.
`
`
`
`28
`
`

`

`Specific Examples – long-term indications (chronic therapy)
`
`older statins (e.g., simvastatin), as well as several large morbidity and mortality trials82, have
`shown that serious liver injury occurs rarely, not exceeding background, with several of these
`drugs. For instance, during clinical trials with lovastatin, ALT > 3x ULN occurred in 2.6% and
`5.0% of patients on doses of 20 mg and 80 mg/day, respectively. The elevations are reversible
`with continuing therapy and are dose related. Postmarketing, lovastatin exposure is estimated
`worldwide to be 24 million patient-years. Rare cases of liver failure have been reported at a rate
`of approximately 1/1.14 million patient years, which is approximately equal to the background
`rate of idiopathic ALF.83
`
`Increased Hy’s cases – examples
`
`Troglitazone is an example where “Hy’s cases” observed during clinical trials portended a
`significant postmarketing issue with severe DILI and fatal liver failure. Troglitazone is discussed
`below in Section D.
`
`D.
`
`Troglitazone
`
`In the clinical trials which led to troglitazone’s approval by the FDA in 1997,84 there were no
`cases of liver failure in 2510 patients. In the NDA database (N=2510), 1.9% of troglitazone-
`treated patients had ALT >3x ULN, 1.7% had ALT >5x ULN, and 0.2% (5 patients) had ALT
`>30x ULN (two of whom also experienced jaundice). The median duration of troglitazone
`therapy before peak ALT elevation was 121 days. In 1997, NIH sponsored a large Diabetes
`Prevention Program85 designed to determine whether non-insulin-dependent diabetes mellitus
`can be prevented or delayed in persons with impaired glucose tolerance. Study groups included
`intensive lifestyle intervention with diet and exercise, metformin or troglitazone with standard
`diet and exercise, and a control group. The troglitazone arm was discontinued in 1998 due to
`reports of severe hepatotoxicity.86 In the NIH Diabetes Prevention Trial (N=585), 3.0% of
`troglitazone-treated subjects had ALT >3x ULN, 1.5% had ALT >8x ULN, and two patients had
`ALT >30x ULN. One of these patients developed liver failure and died, despite receiving a liver
`transplant. The second patient recovered. The median duration of troglitazone therapy before
`initial ALT elevation was 126 days, and to peak elevation was 143 days.87
`
`In response to worrisome and continuing reports of ALF associated with troglitazone use, a
`series of “Dear Healthcare Professional” letters were sent to practicing physicians between 1997
`
`82 Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with
`simvastatin in 20,536 high risk individuals: a randomized placebo-controlled trial. Lancet 2002;360:7-22.
`83 Tolman K. The liver and lovastatin. Am J Cardiol 2002;89:1374-1380.
`84 FDA Center for Drug Evaluation and Research. Medical review of troglitazone – efficacy supplement, NDA 20-
`720, Dr. Robert Misbin, March 12, 1999. www fda.gov/cder/foi/nda/99/20720S12S14 Rezulin htm (accessed July
`12, 2004.
`85 Muniyappa R, El-Atat F, Aneja A, McFarlane SI. The diabetes prevention program. Current Diabetes Reports
`2003 Jun; 3(3):221-2.
`86 ibid.
`87FDA Center for Drug Evaluation and Research. Medical review of troglitazone – efficacy supplement, NDA 20-
`720, Dr. Robert Misbin, March 12, 1999. www fda.gov/cder/foi/nda/99/20720S12S14 Rezulin.htm (accessed July
`12, 2004.
`
`
`
`29
`
`

`

`and 1999, warning about severe liver injury and recommending monthly transaminase
`monitoring. Unfortunately, transaminase monitoring was not regularly performed.88 Moreover,
`an analysis of 94 cases of liver failure which were reported spontaneously to the FDA showed
`that the progression from normal hepatic functioning to irreversible liver injury occurred within
`one month in 19 patients who were indistinguishable clinically from the 70 patients who had an
`unknown time course to irreversibility. Of the 89 cases of ALF, only 11 (13%) recovered without
`liver transplantation. The onset of injury began from three days to after more than two years of
`troglitazone use. Progression from jaundice to hepatic encephalopathy, liver transplantation, or
`death was rapid, averaging 24 days. The authors concluded that “progression to irreversible liver
`injury probably occurred within a one-month interval in most patients, casting doubt on the value
`of monthly monitoring” of serum transaminase levels as a means of preventing severe DILI.89
`
` marked increase in risk with each month of troglitazone use was demonstrated by Graham90 in
`an analysis of interval-specific hazard rates (per million person-years) for each month of
`continued troglitazone use, based on ALF cases reported to the FDA (numerator) and the
`estimated person-years of troglitazone exposure for that corresponding month of use
`(denominator). A table in that report is reproduced below,91 and shows the cumulative risk of
`ALF calculated as “1-survival probability” for each month of continued use, derived from the
`life-table analysis, and expressed in the form of “1 case per x persons treated” for each month of
`continued use (slide 29 in the original document).
`
`This analysis of troglitazone data through the close of 1999 showed that the interval-specific
`hazard rate was substantially elevated above the expected background rate of one per million
`person-years beginning with the first month of troglitazone use. The cumulative risk of ALF
`increased from one case per 131,000 users at one month of use to one case per 7,000 users with
`18 months of continued troglitazone use.92
`
` A
`
`
`88 Graham DJ, Drinkard CR, Shatin D, Tsong Y, Burgess M. Liver enzyme monitoring in patients treated with
`troglitazone. JAMA 2001;286:831-833.
`89 Graham DJ, Green L, Senior JR, Nourjah P. Troglitazone-induced liver failure: a case study. Am J Med
`2003;114:299-306.
`90 Graham DJ, Green L. Final Report: Liver Failure Risk with Troglitazone (Rezulin). FDA/CDER/ODS/DDRE
`consult, dated December 19, 2000.
`91 ibid, page 20.
`92 ibid, page 20.
`
`
`
`30
`
`

`

`Interval-Specific Hazard Rates (x10-6 pyrs) and Cumulative Risk of
`Liver Failure with Rezulin, by Duration of Use*
`
`Mon hs Use Cases
`
`Int Hazard
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`
`9
`5
`9
`14
`13
`8
`3
`10
`2
`2
`2
`2
`1
`
`2
`1
`
`89
`58
`117
`206
`216
`149
`62
`230
`52
`57
`64
`72
`40
`
`135
`79
`
`Cum Risk
`(1 per "x")
`131K
`79K
`44K
`25K
`17K
`14K
`13K
`10K
`10K
`9K
`9K
`9K
`8K
`
`8K
`7K
`
`*Duration missing
`for 11 cases
`
`29
`
`
`
`
`Table reproduced from Graham DJ, Green L. Final Report: Liver Failure Risk with Troglitazone
`(Rezulin). FDA/CDER/ODS/DDRE consult, dated December 19, 2000.
`
`More recently, the incidence of hospitalized idiopathic acute liver injury and ALF in
`troglitazone-treated patients was estimated in an observational retrospective cohort study using
`claims data from a large multistate health care organization.93 The inception cohort included
`7568 patients who began troglitazone during the study period. A total of 4020 person-years of
`exposure were observed. The incidence rates per million person-years of acute idiopathic liver
`injury (95% CI) were as follows: hospitalization with jaundice (n=4), 995 per million person-
`years (271, 2546); ALF (n=1), 240 per million person-years (6.3, 1385). This represents a
`marked increase in risk compared to estimated background rates of hospitalization for idiopathic
`acute liver injury (22 per million person-years) and ALF (1 per million person-years).94
`
`Although the pathogenesis of troglitazone toxicity is not understood,95 experience with
`troglitazone provides a clear example of a situation where “Hy’s Law” cases observed during
`clinical trials prior to approval were predictive of a high risk of severe DILI and ALF post
`marketing. Troglitazone was withdrawn from the US market in March 2000, after 94 cases of
`drug-induced liver failure had been reported.96
`
`93 Graham DJ, Drinkard CR, Shatin D. Incidence of idiopathic acute liver failure and hospitalized liver injury in
`patients treated with troglitazone. Am J Gastroenterol 2003;98(1):175-9.
`94 ibid.
`95 Lee WM. 2003. op cit.
`96 Graham DJ, Green L, Senior JR, Nourjah P. Troglitazone-induced liver failure: a case study. Am J Med
`2003;114:299-306.
`
`
`
`31
`
`

`

`Isoniazid
`
`Isoniazid remains a first-line agent against tuberculosis, even though increased levels of
`aminotransferase are observed in 15 to 30 percent of patients who take the medication and one in
`1000 patients will have severe hepatic necrosis.97 98 Recent experience in public health clinics
`has shown that risk of severe hepatotoxic reactions to isoniazid can be effectively minimized by
`instructing patients to stop drug and immediately report symptoms of liver injury as soon as they
`occur.99 In a recent 7-year survey from a public health tuberculosis clinic in Seattle, WA, a total
`of 11,141 consecutive patients who started a regimen of isoniazid preventive therapy for latent
`TB infection were followed to determine the rate of developing signs and symptoms of
`hepatotoxicity during clinically monitored therapy.100 Monitoring for the safety of isoniazid was
`done by a clinical evaluation for symptoms rather than by transaminase monitoring because
`many patients experience a transient rise in serum transaminase levels during isoniazid therapy.
`During the 7-year study period, eleven patients (0.1%) experienced hepatotoxic reactions while
`receiving isoniazid. All eleven patients had highly elevated serum transaminase levels and nine
`(82%) patients were hyperbilirubinemic. Only one patient was hospitalized because of
`hepatotoxicity. All eleven patients recovered without sequelae.
`
`Similar experience was reported from a tuberculosis clinic in California, with outcomes ava