[Congressional Record (Bound Edition), Volume 147 (2001), Part 17]
[Extensions of Remarks]
[Pages 24110-24112]
[From the U.S. Government Publishing Office, www.gpo.gov]



                        MEDICATIONS FOR DIABETES

                                 ______
                                 

                         HON. LINDSEY O. GRAHAM

                           of south carolina

                    in the house of representatives

                      Wednesday, December 5, 2001

  Mr. GRAHAM. Mr. Speaker, for years too many Americans have suffered 
the ravaging effects of Diabetes. While there have been many promising 
advancements in the diabetes research field, there have also been many 
disappointing setbacks.
  One key to proper treatment of Diabetes has been the development and 
the use of new medications. However, the Congress, questions have been 
raised about the safety of Rezulin and other medications approved by 
the Food and Drug Administration (FDA) for this use.
  In my home state of South Carolina, Mrs. Francis Geddings took 
Rezulin as a treatment from April 1997 to January 1998. She was 
hospitalized in 1999 and tragically passed away from liver failure last 
year. She left behind her husband, Eugene, and many questions about the 
safety of this drug.
  Rezulin was eventually removed from the market, but many questions 
remain. To avoid future tragedies like the one that visited the 
Geddings family, we must continually review how medication is made 
available for public use. Attached are documents that show only a small 
part of the Rezulin story. It is up to Congress to continue doing 
everything we can to make the FDA approval process as safe and open as 
possible.
  Americans need to know that according to an FDA document created by 
several of the FDAs premier scientists, 1 in 1,000 patients who took 
Rezulin for more than one year will die of fatal liver disease. 
Pharmaceuticals companies everywhere can learn from the tragic history 
of Rezulin.


[[Page 24111]]


         Department of Health and Human Services, Public Health 
           Service, Food and Drug Administration, Center for Drug 
           Evaluation and Research.
                                                December 19, 2000.
     From: David J. Graham, MD, MPH, Associate Director for 
       Science, Office of Postmarketing Drug Risk Assessment (HFD-
       400), Lanh Green, RPh. MPH, Safety Evaluator, Division of 
       Drug Risk Evaluation II (HFD-400).
     Through: Martin Himmel, MD, MPH, Deputy Director, Office of 
       Postmarketing Drug Risk Assessment (HFD-400).
     To: David G. Orloff, MD, Director, Division of Metabolic and 
       Endocrine Drug Products (HFD-510).
     Subject: Final Report: Liver Failure Risk with Troglitazone 
       (Rezulin'), NDA: 20-720.


                           Executive Summary

       The following report summarizes the activities of the 
     Office of Postmarketing Drug Risk Assessment and its 
     evaluation of the risk of acute liver failure (ALF) with the 
     use of troglitazone for the treatment of diabetes. The report 
     is divided into topical areas related to varying aspects of 
     the issue.
       We estimated the background rate of acute liver failure in 
     the general population to the about 1 case per million 
     persons per year (person-years). Using case reports data 
     supplemented by usage data from a large multi-slate managed 
     care organization, we estimated the rate of ALF with 
     troglitazone to be about 1 case per 1000 person-years 
     (accounting for underreporting). From three postmarketing 
     clinical studies, the incidence of ALF ranged from about 
     1,200 to 17,000 per million person-years. Survival analysis 
     suggested that the cumulative risk of ALF with troglitazone 
     increased with continuing use of the drug. The implications 
     of this for a product intended to bee used for decades should 
     not be overlooked.
       Based on a number of different analyses, underreporting of 
     ALF with troglitazone was extensive. This highlights the 
     limitations of voluntary (spontaneous) reporting systems. It 
     also illustrates the danger of using changes in reporting 
     over time as a message of success of an intervention. 
     Reporting naturally decreases quickly after the start of 
     marketing so that one cannot cite a decline in number of case 
     reports as evidence that a safety problem has been 
     successfully managed.
       Multiple labeling revisions and ``Dear Healthcare 
     Professional'' letters recommending monthly liver enzyme 
     monitoring did not improve the safety profile of 
     troglitazone. Enzyme monitoring was not performed regularly 
     or reliably even after the July 1998 relabeling. Analysis of 
     case reports suggested that even had monitoring been 
     performed, it probably would not have prevented many, or 
     perhaps any, cases of troglitazone-induced ALF. The ``point 
     of no return,'' that is, of irreversibility and inevitable 
     progression to liver failure appeared to be reached within 
     about a month or less of a time when liver enzymes were 
     normal.
       Troglitazone appeared to confer a substantially greater 
     risk of ALF than rosiglitazone. However, the risk of ALF with 
     rosiglitazone appeared to be higher than the expected 
     background rate.


                   Background on acute liver failure

       Acute liver failure is a rapidly progressive disorder 
     characterized by hepatic encephalopathy, and frequently, 
     coagulopathy (both platelets and clotting factors), 
     methobilic derangements (lactic acidosis, hypoglycemia, 
     electrolyte abnormalities), high output hypovolemic heart 
     failure, renal failure and sepsis. Survival without 
     transplant is below 25%.
       Drug-induced ALF is usually more aggressive than viral 
     forms, with survival rates around 10% without transplant. 
     There are several competing classification systems for ALF, 
     each relying on the length of time it takes for a patient to 
     progress from initial symptoms (US) or jaundice (UK, France) 
     to hepatic encephalopathy. The U.S. definition classifies ALF 
     as progressive from initial symptoms of liver dysfunction to 
     encephalopathy within 6 months. In Europe, progression from 
     jaundice to encephalopathy within 12 weeks is classified as 
     ALF. In subsequent work, we used the European criteria. We 
     choose the latter criteria because their shorter time-window 
     more closely reflected the fulminant nature of the cases we 
     were receiving. Also, the onset of jaundice is a clearer and 
     more definite time-point from which to begin counting 
     compared with initial symptoms, the onset of which might be 
     vague and hence unlikely to be reported accurately in case 
     reports.
       The etiology of ALF varies somewhat by country (slide 2). 
     Until recently, about 70% of ALF in the U.S. was due to viral 
     hepatitis (primarily hepatitis B), with 15% due to 
     acetaminophen and about 10% due to other drugs and toxins.

                           *   *   *   *   *

       The Diabetes Prevention Program (DPP) was a NIH-sponsored 
     clinical trial performed on patients with impaired glucose 
     tolerance (IGT), but not diabetes. Its purpose was to study 
     whether treatment of IGT with oral hypoglycemic agents could 
     prevent or delay the onset of diabetes. One arm of the trial 
     included 585 patients treated with troglitazone on average 
     for one year. From this group, one patient died of fulminant 
     ALF, for an incidence rate of 1,724 per 106 
     person-years (95% confidence interval 44-9,569).
       The REACH study was a Warner-Lambert/Parke-Davis sponsored 
     postmarketing study to collect additional information on 
     efficacy and safety of troglitazone. At the time when 2,433 
     patients were enrolled in the study, with an average duration 
     of treatment <4 months, one patient died of fulminant ALF, 
     for an incidence rate of 1,274 per 106 person-
     years (95% CI 32-7,077).
       Another Warner-Lambert/Parke-Davis postmarketing study, 
     Protocol II, was conducted to study the effect of 
     troglitazone use on the insulin does required by diabetic 
     patients enrolled in the study. There were 233 patients 
     enrolled in this randomized double-blind placebo-controlled 
     trial, each treated for a maximum of 6 months. Of this group, 
     one died of liver failure. Of note, this patient developed 
     liver enzyme abnormalities in November 1998 and was withdrawn 
     from the study. His liver enzymes did not normalize and in 
     early March 1999, the blind was broken for this patient to 
     see whether he had received troglitazone or placebo. He had 
     been treated with troglitazone. He was in hospital for 
     evaluation of his liver disease on the day of the March 1999 
     advisory meeting, and died of liver failure three days after 
     the meeting. Assuming that 50% of randomized patients were 
     treated with troglitazone for a maximum of 6 months, the 
     incidence rate in this study was about 16,949 per 
     106 person-years (95% CI 429-90,855).
       In each of these three studies, fatal liver failure was 
     observed at an extremely high rate, ranging from 1,274 to 
     16,949 per 106 person-years. Based on data from 
     the published literature discussed above, we would expect 
     about 1 case of ALF per 106 person-years meaning 
     that the occurrence of liver failure in these studies was 
     from about 1,300/ to 17,000/times greater than would be 
     expected by chance.
       In the original troglitazone NDA, there were 2 cases of 
     jaundice/hepatitis (one of which was hospitalized) and 1 
     other patient hospitalized with drug-induced hepatitis, but 
     no cases meeting our definition of ALF. This finding is still 
     compatible with an ALF incidence rate of 2,584 per 
     106 person-years.
       These studies demonstrate that liver enzyme monitoring on a 
     monthly basis does not prevent the occurrence of ALF with 
     troglitazone. Furthermore, they collectively support the 
     conclusion that the underlying incidence rate of ALF due to 
     troglitazone is extremely high, probably in the range of 
     1,000 to 2,000 per 106 person-years, representing 
     about a 1,000- to 2,000-fold increase in liver failure risk. 
     Another way of stating this is that 1-2 out of every 1,000 
     patients (1/500=-1/1,000) who use troglitazone for one year 
     will die of ALF.

                           *   *   *   *   *



                               Discussion

       The data presented here provide a comprehensive picture of 
     liver failure risk with troglitazone. Premarketing clinical 
     trial data from the company's NDA for troglitazone showed 
     that ALT elevation above 3 ULN occurred in 1.9% of treated 
     patients. More importantly, it provided an estimate of the 
     incidence of hospitalized drug-induced hepatitis that was 
     more than 50-fold greater than the background rate suggested 
     by the literature.
       Soon after US marketing began, FDA began receiving case 
     reports of ALF in patients who were using troglitazone. A 
     series of labeling revisions and ``Dear Healthcare 
     Professional'' letters followed, recommending increasing 
     performance of liver enzyme monitoring as a means of reducing 
     or eliminating risk of ALF. Despite those interventions, 
     cases continued to be steadily reported to FDA.
       Our analyses of the original 43 US cases found that there 
     were no apparent risk factors by which to identify patients 
     who might be at increased risk of developing ALF while using 
     troglitazone. Furthermore, the onset of liver disease was 
     usually heralded by the appearance of jaundice, by which 
     time, irreversibility had been passed in these cases who 
     usually progressed quickly to encephalopathy: Examination of 
     12 cases with adequate liver enzyme monitoring prior to onset 
     of liver disease showed that in 75%, patients went from 
     having normal liver enzymes to irreversible progression 
     towards liver failure within the recommended monitoring 
     interval. In the three other cases, the patients remained on 
     troglitazone after the first recorded enzyme abnormally so 
     that it was not possible to identify when the point of 
     irreversibility was passed. Of note, there were no 
     differences between the 12 ``rapid risers'' and the remaining 
     31 cases for whom we lacked data on the time-course of their 
     liver enzyme elevations. From these data, we concluded that 
     it was not possible to prevent ALF by patient selection or to 
     predict who was at risk. Also, monthly liver enzyme 
     monitoring would probably fail to prevent at least 75% and 
     perhaps 100% of cases.
       The cases reported to FDA were also used to estimate the 
     pattern of ALF risk over time of continued use of 
     troglitazone. This too was presented at the March 1999 
     advisory meeting. Analysis showed a marked rise in risk 
     beginning with the first month of

[[Page 24112]]

     troglitazone use. With continued follow-up after the advisory 
     meeting, our expectation was confirmed that heightened ALF 
     risk continued for as long as troglitazone was used. In other 
     words, the risk of ALF did not disappear after the first few 
     months or even first 18 months of use. The pattern suggested 
     that cumulative risk of ALF would continue to rise for as 
     long as troglitazone was used, having important implications 
     for a drug intended to be used for 20, 30 or 40 years or 
     longer.
       Against this backdrop of case reports, epidemiologic data 
     suggested that the expected incidence rate of ALF in the 
     general population was about 1 case per million per year. The 
     data from case reports were markedly higher than this. At the 
     March 1999 advisory meeting, we presented data showing that 
     if we assumed there was no underreporting, the cumulative 
     risk of ALF was about 1 case per 15,000 patients who used 
     troglitazone for at least 8 months. If we factored into the 
     analysis that only 10% of cases had been reported, the 
     cumulative risk became 1 case per 1,500 at 8 months (about 1 
     case per 1,000 per year). With an additional year's worth of 
     case reports (through December 1999), the cumulative risk was 
     1 case per 7,000 patients after 18 months of troglitazone 
     use, assuming no underreporting. With 10% reporting, this 
     would be 1 case per 700 patients at 18 months (about 1 case 
     per 1000 per year). The first analysis through 8 months of 
     use led us to conclude prior to the March 1999 advisory 
     meeting that the risk of ALF with troglitazone was probably 
     increased at least 1000-fold over the expected background 
     rate.
       Independent population-based data prior to the March 1999 
     advisory meeting supported this. In two separate 
     postmarketing clinical studies, one conducted by the National 
     Institutes of Health and one conducted by the company, a case 
     of fatal ALF occurred among small numbers of patients treated 
     with troglitazone. This was highly statistically significant, 
     and suggested that the incidence rate of ALF with 
     troglitazone could range from 1,200 to 1,700 per million per 
     year, with upper bounds approaching 10,000 cases per million 
     per year. These data, in combination with case reports data, 
     formed the basis for this medical officer's recommendation 
     prior to the March 1999 advisory meeting that troglitazone be 
     removed from the market. Subsequent to the advisory meeting, 
     FDA learned of a third postmarketing study, this one 
     randomized and double blinded, in which a patient treated 
     with troglitazone died of ALF just three days after the 
     advisory meeting. The incidence rate of ALF in this study was 
     over 17,000 per million per year.
       An important component in the troglitazone analysis was an 
     assessment of the effect of FDA interventions in the form of 
     labeling changes recommending periodic liver enzyme 
     monitoring as a means of managing the ALF risk of 
     troglitazone. The FDA study from UnitedHealth Group found 
     that monitoring was not regularly or reliably performed and 
     that repeated labeling revisions had not meaningfully 
     improved the performance of monthly liver enzyme testing. 
     Based on the data at hand prior to the March 1999 advisory 
     meeting, we concluded that FDA labeling had not had a 
     clinically important effect on medical practice and that 
     monthly enzyme testing was largely not being performed. From 
     our case analysis, we concluded that monitoring, were it 
     performed, would fail to prevent most or all cases of 
     troglitazone ALF.

                           *   *   *   *   *




                          ____________________