Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 73
Review of the Fialuridine (FIAU) Clinical Trials 10 Eli Lilly Trial H3X-MC-PPPC (NIH Protocol 93-DK-0031) This study was conducted at the Liver Diseases Section of the National Institute of Diabetes and Digestive and Kidney Disease (NIDDK) at NIH. This open-label randomized study of FIAU used two treatment groups (0.1 and 0.25 mg/kg/day given orally in three divided doses). Duration of treatment was to be 6 months, in the hope of prolonging the temporary inhibition of HBV DNA seen in the previous NIH study (R91). The subjects in fact were largely drawn from patients who had participated in that earlier trial. Principal investigator was Jay Hoofnagle, a staff scientist in the section from 1978 to 1986. In 1986, Hoofnagle became Acting Clinical Director of NIDDK and subsequently Director of the Division of Digestive Diseases and Nutrition. Adrian DiBisceglie assumed the responsibilities of senior investigator in the Liver Diseases Section in 1988, and in 1991 he became Chief of the Hepatitis Study Section, supervising four medical staff fellows. The day-by-day conduct of clinical studies was the responsibility of DiBisceglie and the medical fellows. Michael Fried was the fellow primarily assigned to the PPPC trial and it was the responsibility of Fried and DiBisceglie to recruit patients for the study, discuss the protocol with the patients, obtain informed consent, and see patients during subsequent visits. NIH attracts first-rate fellows, the majority of whom have already completed internal medicine residencies and 2 years of gastroenterology training before coming to NIH. Graduates of this fellowship program rank among the finest hepatologists not only in the United States but in the world (D. Shafer, M. Peters, A. DiBisceglie, P. Martin, G. Davis, G. Dusheiko, to mention but a few). Hoofnagle is internationally recognized for his contribution to clinical research in the treatment of chronic viral hepatitis. Testimony to his stature in the field of hepatology is that he is recent Past-President of the American Association for the Study of Liver Diseases, the foremost hepatology association in the world. Rationale for the study came from the lack of available effective therapy for chronic HBV infection, the lack of major toxicity of FIAU in preclinical animal toxicology studies, and the promise of this compound as an anti-HBV drug in earlier human trials. Interferon has been approved by the FDA for treatment of this disease, but therapy inhibits viral replication in the minority of patients (approximately one-third), the drug must be given parenterally (subcutaneously), and side effects, although rarely leading to the discontinuation of treatment, are frequently significant. Phase I/II trials of FIAU (R90-001-01) by the AIDS Clinical Trials Group (ACTG) showed that FIAU has no significant anti-CMV activity, but showed a clear suggestion that it has significant anti-HBV activity. FIAU was readily bioavailable by the oral route of
OCR for page 74
Review of the Fialuridine (FIAU) Clinical Trials administration, and it appeared to be well-tolerated without major side effects. Marked reductions in serum HBV DNA levels were noted in the majority of patients in that trial, with clearance of HBeAg from two of the patients. In some this inhibition was sustained, but in others HBV DNA levels returned to the pretreatment baseline levels. Some of these patients had complex medical problems and were also receiving other antiviral drugs with potential side effects, complicating subsequent interpretation of possible FIAU-associated toxic effects. In addition, since HIV infected patients are immunocompromised and as such may be less likely to have a sustained response to therapy, treatment in nonimmunocompromised patients seemed the logical next step. Given the lack of effective orally available antiviral agents for the treatment of HBV infection and the promising nature of these early studies, a 1-month study was conducted at NIH to look at the safety and efficacy of this drug in patients with HBV infection alone. In NIH trial 91-DK-AI-123 (Lilly Trial R91-010), FIAU was shown to inhibit HBV replication effectively. However, this inhibition was transient, with viral reactivation in most patients. The rationale behind the final NIH FIAU study was that retreatment of approximately 20 patients with chronic HBV infection for 24 weeks would result in short-term inhibition of viral replication, normalization of serum aminotransferase levels, and potentially long-term virologic remission. The protocol was written in September 1992, before completion of the R91 study. Submission of the retreatment protocol to FDA was delayed because of the lack of long-term animal toxicity data. Six-month toxicity data from studies in animals or humans were not yet available. Appendix F summarizes the preclinical toxicity data available when the PPPC trial was finally begun in March 1993. These included a 6-month study in mice, and similar studies in rats, dogs, and monkeys were under way. AVAILABLE CLINICAL DATA REGARDING POTENTIAL TOXICITY At the time of initiation of the PPPC trial in March 1993, there was knowledge of eight deaths of patients who had participated in previous FIAC/FIAU trials: patients 107 (hepatic failure), 105 (bacterial pneumonia), 103 (AIDS), and 110 (AIDS) from R89; patients 101 (Pneumocystis pneumonia), 401R (hepatic failure), and 406R (pancreatitis) from R90; and patient 4D in protocol R91 (hepatic failure). Patient 408R in trial R90 also died of liver failure, but his death occurred after the follow-up period and investigators did not learn of the death until July 1993. These deaths have been reviewed in previous sections. Concern about patient 4D and recognition of a similar spectrum of signs and symptoms when it arose in patient 2 in the PPPC trial led to prompt and appropriate termination of this study in late June 1993. A variety of other potential toxic effects had been identified from earlier trials (R89 and R90) in HIV-positive patients. Nausea, vomiting, and abdominal discomfort were side effects of FIAC/FIAU and were dose limiting. Peripheral neuropathy was also considered to be a possible adverse event associated with FIAU therapy. One of the patients from trial R91 (patient 1A) had a clear-cut neuropathy. This 52-year-old man received 28 days of FIAU therapy in March and April 1992, but sensations of numbness and tingling were not noted until late September 1992 (5 months after receiving his last dose of FIAU). Because of progressive pain, he underwent nerve conduction studies at NIH in January 1993. These were consistent
OCR for page 75
Review of the Fialuridine (FIAU) Clinical Trials with a peripheral neuropathy. Symptoms were similar to those that the patient had experienced 5 years previously. Because of this prior history and because the onset of symptoms was fully 5 months after FIAU treatment was stopped, the investigators were not convinced of a clear relationship between drug and toxicity. However, given the known neurotoxicities of other nucleoside analogs, the investigators were sufficiently concerned to include careful neurologic assessment and nerve conduction studies in the next protocol, H3X-MC-PPPC. In retrospect, it is almost certain that this neuropathy was FIAU related . A second patient in the R91 study (patient 4D) also complained of numbness and tingling in the lower extremities several months after stopping study drug. Peripheral neuropathy was confirmed by nerve conduction studies. This patient subsequently died following a cholecystectomy. Although both of these neuropathies were believed to be atypical for that seen with nucleoside analogs in that they occurred after the drug was stopped, and persisted (Ara-A induced neuropathies are usually transient), peripheral neuropathy was included as a potential complication of FIAU in the PPPC protocol, and nerve conduction studies were included in the study design. Pancreatitis was another potential side effect of FIAU. In the R90 study, patient 409R developed mild pancreatitis while on ddI, requiring hospitalization 3 months after discontinuation of FIAU (total dose received, 1,662.5 mg spread over 2 weeks). The pancreatitis resolved after the discontinuation of ddI treatment. Pancreatitis was therefore attributed to ddI, although ddI treatment was subsequently reinstituted at a lower dose. In retrospect, it is unclear even now whether this pancreatitis was FIAU related, given the known association of pancreatitis and ddI. The investigators cannot be criticized for failing to recognize a possible association at the time. The potential for myopathy was included in the consent form for the PPPC trial. Previous experience with nucleoside analogs had highlighted the difficulty in determining whether pain and numbness in the arms and legs was due to a neuropathy or a myopathy, or both. Myoclonic jerks, confusion, and/or seizures had been seen in three HIV-positive patients after they had received 5-10 doses of FIAC. These occurred only in critically ill patients with AIDS treated in the FIAC trial and were not seen in any patient on FIAU. Nevertheless, the possibility that seizures could be due to treatment was included in the informed consent document for the PPPC trial. In the R91 trial, a rise in CPK levels was noted in patient 2A as clinically significant. Hepatotoxicity was not believed to be a toxic effect of FIAU. In the R91 trial, a rise in liver enzyme levels associated with the loss of HBV DNA was interpreted as a positive event. Inhibition of HBV replication occurred in 9 of the 24 patients. This was associated with a rise in liver enzyme levels in eight of the nine patients. Subsequently, ALT levels returned to normal or near normal in all patients. The rise in liver enzyme levels associated with the loss of HBV DNA was interpreted as a flare associated with the resolution of infection. This was well documented in the patients charts at NIH. Patients 6A and 1B, for example, ALT levels rose to 722 and 733 IU/1 respectively. In contrast to earlier FIAC trials for CMV, an elevation in the ALT level was deliberately not included as an adverse event because of the expected flare seen in drug-induced inhibition of viral replication. This flare was unusual in these two patients in that it occurred after the completion of therapy, but in both patients it was associated with the loss of HBV DNA. Importantly, there was no other evidence of hepatic decompensation in association with this rise in ALT level (increased serum bilirubin or
OCR for page 76
Review of the Fialuridine (FIAU) Clinical Trials prolongation of prothrombin time). A liver biopsy performed during the flare showed acute or chronic hepatitis with less HBV core antigen (HBcAg) staining than in the pretreatment biopsy specimens. Neither liver biopsy was interpreted at the time as showing microvesicular steatosis (which was later identified as a hallmark of FIAU hepatotoxicity). Hyman Zimmermann of the Armed Forces Institute of Pathology subsequently reviewed the posttreatment liver biopsy in patient 006A and also found no evidence of microvesicular or macrovesicular steatosis. Patients 6A and 1B subsequently became HBeAg negative, with normalization or near normalization of ALT levels, suggesting that the investigators' interpretation was indeed correct and that the ALT elevations were indicative of a positive response to treatment rather than FIAU toxicity. Even with hindsight, these changes in ALT levels cannot be ascribed to FIAU toxicity, particularly in view of the minimal ALT elevations (less than two-fold elevations) characteristic of the fatal hepatic failure ultimately seen in the PPPC trial, and the recognized flare phenomenon that occurs with a positive response to antiviral therapy. AVAILABLE SAFETY DATA At the initiation of the PPPC trial there was considerable information regarding the safety of FIAU. In the R90 trial, the first patients coinfected with HIV and HBV treated at NIH received FIAU at 1 mg/kg/day and marked inhibition of viral replication was noted. The protocol was amended to evaluate the anti-HBV activity of lower doses of FIAU (0.5 and 0.1 mg/kg/day). Treatment duration was 14 days, with 28 days of follow-up. In the R91 trial, 24 patients with chronic HBV infection received four different doses of FIAU for 28 days (0.05, 0.1, 0.25, and 0.5 mg/kg/day), with six patients in each dosing group. Enrollment in this study was staggered to allow evaluation of the safety and efficacy of each dose before testing the next dose. All patients completed therapy without dose modification, and all patients completed follow-up. The side effects were mild and were seen mainly in those who had received the highest dose. It must be emphasized that although the planned duration of treatment in the PPPC trial was longer than those in previous trials, the doses used in the PPPC trial (0.1 and 0.25 mg/kg/day) were 4- to 10-fold lower than those used in early trials. DEVELOPMENT OF PROTOCOL AND FDA REVIEW On September 16, 1992, Hoofnagle submitted the PPPC protocol to the NIDDK IRB. The protocol was initially designed as a retreatment for patients who had not responded to therapy in the R91 study. The PPPC protocol was discussed extensively with the FDA's Division of the Antiviral Drug Products in late October 1992. Present at this meeting were the NIH investigators and representatives of both Lilly and Oclassen. Included in the protocol was discussion of peripheral neuropathy and the death of patient 4D from protocol R91. Other adverse events summarized included the elevations in ALT level seen in protocol R91 (patients 6A and 3B seen in association with the inhibition of vital replication and interpreted as a delayed flare of hepatitis). Concerns of hepatotoxicity were not raised by FDA either then or
OCR for page 77
Review of the Fialuridine (FIAU) Clinical Trials at any later date until the toxicity of FIAU became readily apparent. The consent form for the PPPC trial was reviewed by FDA, which asked for the inclusion of the possible teratogenicity of FIAU in mice. After the review of the data, FDA approved the six month treatment study. On October 24, 1992, Hoofnagle responded to issues regarding the protocol raised by the chair of the NIDDK IRB, and the protocol was approved on October 30. In February 1993, two amendments to the protocol were submitted to the NIDDK IRB, to include (1) patients who had not previously received FIAU and (2) nerve conduction studies before and during therapy. The amended protocol was approved by the NIDDK IRB on March 5, 1993. PATIENT SELECTION AND ENROLLMENT All patients in the PPPC trial had chronic HBV infection with active viral replication (HBeAg and HBV DNA positive). Eleven patients (patients 1-10, and 15) had previously been treated with FIAU in trial R91, and four patients (patients 11-14) had never received FIAU. This was an open-label, randomized study of FIAU with two treatment groups (0.10 and 0.25 mg/kg/day given orally in three divided doses). Other patients from the R91 trial were ineligible for retreatment because of the loss of HBV DNA with initial therapy (n = 9), side effects from previous treatment (n = 1), need for interferon (n = 1), and inconvenience (n = 2). Thus, only 11 patients were suitable for retreatment, and the protocol was amended to include patients who had never received FIAU. Inclusion Criteria The following criteria were used for the inclusion of patients in the PPPC trial: (1) male or female patients age 18 years and older who had been included in 91-DK-AI-123 as well as additional new patients to bring the study total number of subjects to 24; (2) HBsAg positive with active viral replication (HBeAg positive and HBV DNA or HBV DNA polymerase positive for at least 6 months; (3) compensated liver disease (prothrombin time less than 3 seconds prolonged, serum albumin greater than 3 g/dl, serum bilirubin less than 4 mg/dl, no history of ascites, encephalopathy, or bleeding esophageal varices); (4) elevated serum ALT levels (average of three determinations of more than 50 IU/1); and (5) provision of written informed consent. Exclusion Criteria The following were evaluation criteria for the PPPC trial: (1) seropositivity for HIV, hepatitis C virus (HCV), or hepatitis D virus (HDV); (2) significant cytopenia (hematocrit less than 30 percent, white blood cell count less than 2,500/mm3, and platelet count less than 50,000/mm3); (3) compromised renal function; (4) substance abuse within the preceding 6 months; (5) pregnancy; (6) significant systemic illness; and (7) preexisting peripheral neuropathy.
OCR for page 78
Review of the Fialuridine (FIAU) Clinical Trials Re-enrollment The protocol stipulated that patients enrolled in the R91 study were eligible for retreatment 5 months after completion of a 28-day course of FIAU. Since the majority of patients from the previous FIAU study at NIH met this criterion and were already well known to the NIH investigators, enrollment of the first 10 patients was rapid (from March 23, 1993 to April 21, 1993). This resulted in a "bolus" of patients admitted into the study without the usual built-in delay for the observation of toxicity. Although the trial was halted after the recognition of major toxicity in only one patient in the study (patient 2), major toxicity in six other patients was unavoidable because they had already had prolonged exposure to the drug. Entry of the last retreatment patient (patient 15) was delayed 2 months while a skin lesion was investigated and a flare of hepatitis resolved. After the initial 10 patients had been treated for 2 months, four additional patients (patients 11-14) and the single delayed patent (patient 15) were begun on FIAU in June 1993. Possible Adverse Effects Side effects discussed in the protocol included (1) irritability; (2) gastrointestinal upset at the highest dose; (3) transient CPK elevations; (4) potential severe side effects that were not clearly related to FIAU, including one patient who died following laparoscopic cholecystectomy and two patients who developed a peripheral neuropathy after FIAU was stopped; mention was made of a single patient with an elevated CPK level that resolved on therapy and two additional patients with fatigue and irritability that resolved after FIAU was stopped; (5) Possible toxic effects, which included neuromuscular damage and bone marrow suppression; and (6) that no patient previously developed serious side effects while on treatment. Toxic events graded as mild, moderate, or severe for dose reduction or cessation included nausea, diarrhea, fatigue, muscle aches, fall in hematocrit, neutrophil and platelet counts, and rise in creatinine and CPK levels. The discussion of potential toxic effects of FIAU, the frequency of monitoring, the methods used for patient monitoring, and the parameters used for dose adjustment appeared to be appropriate for what was understood at the time about FIAU. Most notable was the lack of dosing guidelines for changes in ALT levels. This was specifically not included (after intense discussion between the investigators, sponsors and FDA), to prevent premature cessation of FIAU treatment in patients having a positive response to therapy. Informed Consent There were several submissions and modifications of the protocol and consent form, which took place in late 1992, and early 1993. The final consent form explains clearly to patients the rationale for the study (Appendix D). Each patient was briefed on the study by a medical staff fellow at NIH. The informed consent process included discussion of the possibility of peripheral neuropathy. After discussion
OCR for page 79
Review of the Fialuridine (FIAU) Clinical Trials with the NIDDK IRB, death was not included as a possible adverse event in the consent form itself since the death of patient 4D from the R91 trial was not attributed to FIAU. Nevertheless, patients were informed verbally of the possibility of death and that the death of patient 4D could have been FIAU related. The majority of the patients in the PPPC trial had previously been treated with FIAU in the R91 trial and had tolerated FIAU with minimal side effects. Risks and discomforts of FIAU included in the ICD were (1) Fatigue, irritability, and insomnia; (2) nausea and upset stomach; (3) skin rashes; (4) bone marrow suppression; (5) seizures; and (6) pains in the arms and legs. Review of the informed consent document by the IOM committee found that it was written clearly in language understandable to the layperson and was not misleading in minimizing the potential side effects of FIAU. CONDUCT OF STUDY Weekly staff meetings were held at NIH to discuss the status of research protocols. These meetings were regularly attended by both Hoofnagle and DiBisceglie. Present were the fellows and staff nurses. The original protocol had patient visits scheduled at weeks 1, 2, 4, 8, and 12, but NIH investigators asked patients to come in on two occasions between scheduled visits to monitor them more closely. In addition to these added clinic visits there was frequent telephone contact with the patients. Compliance with clinic visits was 100 percent for all 15 patients. The investigators kept flow sheets on all of the patients. These flow sheets were used for day-by-day monitoring and were separate from the charts which were kept in the medical records department at NIH. They were important for weekly meetings as well as for facilitating communication with patients. Given the attention to details, the close supervision of the junior medical staff, and the qualifications of these junior staff, the IOM committee finds no evidence to suggest that the monitoring of the patients by senior investigators was inadequate or that a different monitoring system would have averted the FIAU tragedy. Thorough review of the conduct of this study by the IOM committee has shown that the physician and nursing staff at the NIH were both compassionate and meticulous in the care of these patients. Documentation of the quality of care is apparent from the frequent patient monitoring visits and the frequent telephone contacts. There is no evidence of errors in clinical judgment related to interpretation of clinical or biochemical adverse events that would have predicted the fatal toxic events that occurred. Moreover, there is no evidence that the patients were coerced to remain within the trial or that the investigators overlooked significant adverse events that should have led to a dose reduction or dose cessation according to predetermined definitions of adverse events. In several instances the apparent lack of documentation of adverse events in case report forms (CRFs) resulted from a change in protocol to increase the frequency of visits without a concomitant change in the CRFs. If an event occurred at one of these additional visits, it was recorded in the clinic chart and the investigators and nurse coordinators were instructed to record the event on the page in the CRF at the subsequent clinic visit. In some instances, if the complaint had resolved before this subsequent visit, the event was recorded only in the clinic chart, not the CRF. Hence, the perception of protocol violations during the FDA audit.
OCR for page 80
Review of the Fialuridine (FIAU) Clinical Trials STUDY OUTCOME In early June 1993, it became apparent to the investigators at NIH that the 10 patients who had been started on FIAU in late March and April were having increasing side effects (mainly gastrointestinal, with nausea, vomiting, diarrhea, and abdominal cramps). These were reviewed by Hoofnagle, DiBisceglie, Fried, Straus, and McKenzie, and although the doses were reduced for two patients, adverse events were not thought to be of sufficient severity to discontinue treatment. After a meeting on June 22 to review the status of the trial and to discuss these adverse events with representatives of Lilly (Jennifer Stotka), the NIH investigators elected to continue the trial for an additional week. HBV DNA levels had decreased in all patients, and 6 of 10 patients had become HBV DNA negative. Because of this marked therapeutic benefit, the trial was continued. At this point three patients were no longer receiving drug because of side effects. When one of these three developed lactic acidosis, hepatic failure, and renal failure on June 25, prior experience with a single puzzling patient (patient 4D) from the R91 trial made the investigators consider FIAU a likely cause. A decision was made to halt the trial on June 26. All 15 patients were seen at NIH within a few days. Seven of 10 patients who had received more than 2 months of FIAU treatment had evidence of FIAU toxicity (lactic acidosis, hepatic failure, pancreatitis, neuropathy, and/or myopathy). Despite stopping drug, seven patients had progressive hepatic failure and were transferred to other hospitals for possible liver transplantation. The other eight patients (three of whom had received FIAU for more than 2 months) had either minor toxicity or no adverse events. Deaths Patient 2 This 42-year-old man was diagnosed with HBV infection in 1981. A liver biopsy in 1984 revealed chronic active hepatitis. He was first evaluated at NIH in April 1985, when his only complaint was of fatigue. In March 1987 he was diagnosed with HBV-related membranoproliferative glomerulonephritis. A liver biopsy showed chronic persistent hepatitis. He was enrolled in an interferon protocol at the NIH in 1987 and then subsequently in the prednisolone/interferon protocol. Active HBV infection persisted. In June 1990 he was treated in the NIH ddI protocol, and a liver biopsy at that time showed chronic active hepatitis. Fatigue persisted on therapy and he continued to have active HBV replication. In April 1992 he began the 28-day FIAU protocol. A pretreatment liver biopsy again showed chronic active hepatitis. He received FIAU at 0.25 mg/kg/day (total dose, 476 mg) and tolerated treatment other than for mild and transient irritability. HBV DNA levels fell markedly with treatment, but he remained positive for HBsAg and HBeAg. In March 1993 he was retreated with FIAU (0.25 mg/kg/day; total dose, 1,218 mg; cumulative dose, 1,694 mg). He had side effects of fatigue and nausea, and after 4 weeks of therapy, he developed tingling and numbness of the feet. Nerve conduction studies were normal, but the dose of FIAU was decreased on June 2, 1993, and was stopped on June 8. He
OCR for page 81
Review of the Fialuridine (FIAU) Clinical Trials continued to complain of pain in the feet. The total bilirubin level remained within the normal range while he was on FIAU. The patient's pretreatment ALT level was approximately 140 IU/liter and rose to a peak of approximately 160 IU/liter 10 weeks after starting FIAU. A serum albumin level of 2.8 g/dl on June 14 was attributed to hepatic decompensation in the subsequent FDA task force report, but for several reasons we believe it is likely that this was not related to FIAU-induced liver damage because (1) hypoalbuminemia was present before the patient began FIAU; (2) the patient had membranoproliferative glomerulonephritis with significant proteinuria; and (3) a pretreatment liver biopsy did not show cirrhosis or even fibrosis, which would be necessary if the hypoalbuminemia were secondary to hepatic decompensation. On June 25 the patient contacted NIH complaining of nausea for 2 days. He did not wish to be seen at NIH, but an antiemetic was prescribed. Because of progressive nausea and weakness, he went to an emergency room in Virginia and was found to be markedly hypotensive and acidotic. The patient was seen by DiBisceglie that evening and was transferred to NIH. Evaluation at NIH revealed hepatic and renal failure, severe peripheral neuropathy, and rhabdomyolysis (CPK elevations). The CPK elevation was likely related to muscle necrosis associated with hypotension. On June 30, he was transferred to the University of Virginia for consideration of liver transplantation. He developed worsening liver failure, renal failure, and acidosis. He was treated with an experimental hepatic assist device and underwent emergency liver transplantation on July 4, 1993. There was primary nonfunction of the liver, and the patient died on July 6 of circulatory failure and progressive acidosis. Autopsy revealed hemorrhagic pancreatitis, glomerulonephritis, esophageal varices, and pneumonia. Histology of the liver explant showed micro-and macrovesicular steatosis, chronic active hepatitis with fibrosis, and cholestasis. Electron microscopy showed mitochondrial changes and fat accumulation. Patient 1 This 35-year-old man was found to have HBV infection in April 1988 when he was evaluated for jaundice. He was first seen at NIH in August 1988, when he was noted to have a predominantly indirect hyperbilirubinemia and was given the concomitant diagnosis of Gilbert's syndrome. A liver biopsy in December 1988 showed chronic active hepatitis. He was treated in a controlled trial of interferon at NIH in December 1988, initially receiving placebo and then 10 MU three times weekly. Although his HBV DNA levels fell, he remained HBsAg and HBeAg positive. In 1990, he was enrolled in a ddI trial at NIH. A repeat liver biopsy showed chronic active hepatitis. He was subsequently enrolled in the ribavirin trial at NIH in April 1991. In April 1992 he was started on FIAU (0.1 mg/kg/day) as part of the R91 study (total dose, 247 mg). While on drug therapy, his ALT level peaked at three times the baseline level, but his HBV DNA fell from 308 to 78 pg/ml. He remained HBsAg and HBeAg positive. His only side effects while on FIAU therapy were headaches.
OCR for page 82
Review of the Fialuridine (FIAU) Clinical Trials He was enrolled in the final FIAU study in March 1993 (at a dose of 0.25 mg/kg/day, total dose, 1,366 mg; cumulative dose, 1,613 mg). He complained of nausea (rated as moderate on the CRF at week 1 and mild during subsequent visits). The patient also had some mild anorexia and moderate abdominal pain at week 4. Appropriate tests were scheduled to evaluate the pain. The dose of FIAU was reduced by the investigators on May 27, 1993. Abdominal pain had resolved at a visit 2 days after reducing the dose, but the patient eventually stopped taking FIAU on his own on June 10 because of abdominal cramps. HBV DNA had become undetectable, but he remained positive for HBsAg and HBeAg. Hyperbilirubinemia, which was predominantly indirect, was noted on several occasions while he was on FIAU therapy and was likely due to Gilbert's syndrome. His pretreatment serum bilirubin level was 3.0 mg/dl and it rose to 4.4 mg/dl. His ALT level rose from a pretreatment value of approximately 60 IU/liter to approximately 120 IU/liter. The IOM committee finds no clear-cut evidence of drug-related hepatotoxicity during FIAU treatment in this patient. Attribution of hyperbilirubinemia to Gilbert's syndrome appears reasonable even with current knowledge of the FIAU syndrome. There is documentation of frequent telephone contact as well as clinic visits throughout this period and no evidence that the patient was coerced into continuation of FIAU. Many attempts were made to try to reach him from June 26 to 28. Finally, he was contacted on the evening of June 28 and was admitted to NIH on the following day when he was noted to be jaundiced. Examination revealed a peripheral neuropathy. Over the ensuing days, he developed progressive hepatic failure and lactic acidosis. He was treated with glucose infusions. A computed tomography (CT) scan showed normal liver and pancreas, but ascites. Because of progressive liver failure, he was transferred to Pittsburgh for evaluation for liver transplantation. He became progressively acidotic and hypotensive and died on July 30, 1993. Autopsy revealed hemorrhagic pancreatitis, myocardial infarction, and pulmonary congestion. Liver histology revealed fibrosis, cholestasis, and severe micro-and macrovesicular steatosis. Patient 4 This 52-year-old man was diagnosed with chronic HBV infection in May 1989, at which time a liver biopsy showed chronic active hepatitis with cirrhosis. He was initially seen at NIH in December 1989 and was subsequently enrolled in an interferon protocol (randomly assigned to the control group), a ribavirin protocol, and then the R91 FIAU protocol. A pretreatment liver biopsy for the R91 trial showed chronic active hepatitis with cirrhosis. He received a total of 134 mg of FIAU in the initial trial, and his HBV DNA levels fell from 75 to 6 pg/ml. The levels remained low for 5 months following therapy, but then returned to pretreatment levels. He tolerated FIAU well, with minor symptoms of abdominal cramps, fatigue, and irritability. He was retreated with FIAU from April 1993 until he was told to stop taking the drug on June 26, 1993 (total dose, 628 mg; cumulative dose, 762 mg). During the second course of FIAU treatment, he developed cough, fever, muscle cramps, anorexia, fatigue, insomnia, and mild nausea. His dose of FIAU was reduced and he was treated with clarithromycin for a pneumonia documented on chest radiograph. Pretreatment serum bilirubin was 1.6 mg/dl and remained in the normal range while he was on FIAU therapy (1.7 mg/dl when FIAU therapy was stopped). His ALT level never rose
OCR for page 83
Review of the Fialuridine (FIAU) Clinical Trials above the pretreatment value of approximately 140 IU/liter and was 59 IU/liter when FIAU treatment was stopped. The patient was evaluated at NIH on June 26, 1993 at which time he had no new complaints. Hepatic synthetic function was well-preserved, and he was treated with oral carnitine, uridine, diuretics, and intravenous thymidine and glucose. However, over the ensuing 2 weeks he had progressive hepatic failure (serum bilirubin, 10.3 mg/dl, prothrombin time, 20 seconds) and lactic acidosis. His serum lipase level was elevated but his amylase level was normal. He was transferred to the University of Virginia on July 12, 1993, for liver transplant evaluation. Acidosis and hepatic failure worsened, and he became comatose and hypotensive and died on July 16 awaiting liver transplantation. Autopsy revealed active cirrhosis of the liver with marked macro- and microvesicular steatosis. Electron microscopy showed diffuse mitochondrial damage. There was hemorrhagic pancreatitis. Patient 6 This 37-year-old woman was diagnosed with HBV infection in March 1990, and a liver biopsy in October 1990 showed chronic active hepatitis. She was first seen at NIH in July 1991, at which time a repeat liver biopsy showed mild to moderate chronic active hepatitis. She was enrolled in the ribavirin protocol and then the R91 FIAU protocol, in which she received 115 mg of FIAU. With this first course she noted a mild increase in fatigue. HBV DNA levels fell but never became negative, and she also remained positive for HBsAg and HBeAg. Pretreatment serum ALT level was 66 IU/liter and rose to a peak value of three times the baseline level. These returned to baseline levels after stopping FIAU. In the PPPC protocol she received a total of 551 mg of FIAU (cumulative dose, 666 mg). She complained of right upper quadrant pain at the week 8 visit. She had tenderness but no complaints of pain at a previous visit (week 6). She had also had right upper quadrant pain before starting FIAU treatment, likely related to her hepatitis. The pain experienced while on FIAU occurred approximately three times per week, lasted only a few minutes, and was not of sufficient severity to require a dose reduction. Before retreatment her serum ALT level was 96 IU/liter, and this rose to three times the baseline level by week 8. She reported myalgias (rated as mild), which occurred on a single day and for which no medication was taken. These myalgias did not meet the protocol criterion for dose reduction. When the patient told an associate investigator of nausea, vomiting, headaches and dizziness on June 15, 1993, she was instructed to stop taking the drug. She was admitted to NIH Clinical Center on June 17 with nausea, vomiting, diarrhea, and fatigue. Hepatic synthetic function and amylase were normal; here ALT level was 221 IU/liter. Over the subsequent weeks she continued to have nausea and deteriorating hepatic function (ALT levels, 554 IU/liter, bilirubin levels, 8.2 mg/dl). The diagnosis of FIAU hepatotoxicity was made when patient 2 presented in liver failure. She was treated with thymidine and glucose, but on June 28 she developed severe abdominal pain and biochemical evidence of pancreatitis. On June 29 she was transferred to the University of Virginia for evaluation for liver transplantation. She had progressive lactic acidosis and hepatic failure. She was treated with
OCR for page 84
Review of the Fialuridine (FIAU) Clinical Trials an experimental hepatic assist device and underwent liver transplantation on July 9. Lactic acidosis and hepatic function improved but pancreatitis continued. A CT scan revealed infarction of the left lobe of the liver, and an angiogram diagnosed hepatic artery thrombosis. She underwent repeat liver transplantation on August 14, but died of complications of pancreatitis and renal and respiratory failure on August 31. The first liver explant showed chronic active hepatitis with fibrosis, microvesicular steatosis, and cholestasis; electron microscopy revealed mitochondrial degeneration. The second liver explant showed subtotal hepatic necrosis. Autopsy revealed multisystem failure with acute and chronic pancreatitis. Patient 7 This 44-year-old man with chronic HBV infection was first diagnosed in 1990 and was first seen at NIH for evaluation for possible therapy in March 1991. A liver biopsy in May 1991 showed mild chronic active hepatitis. He was enrolled in the ribavirin protocol in June 1991 without response and was therefore evaluated for and enrolled in the R91 FIAU protocol. He received 0.5 mg/kg/day for 28 days in August 1992 (total dose, 1,353 mg) which he tolerated well other than mild nausea and one episode of muscle cramps. His serum ALT level rose on therapy to three times the baseline levels, and HBV DNA levels decreased but did not become negative. He began the PPPC protocol in April 1993. His pretreatment ALT level was 58 IU/liter, serum bilirubin was 0.6 mg/dl, and albumin level and prothrombin time were normal. He received FIAU at a dose of 0.25 mg/kg/day for 75 days (total dose, 1,753 mg; cumulative dose, 3,106 mg). On therapy he developed increased fatigue, irritability, and mild nausea. He complained of slight bilateral abdominal pain, documented in the NIH chart on May 26, which had resolved by a follow-up visit on June 10. Given the mild and transient nature of the pain, there was no indication for a dose reduction since protocol criteria were not met. He also had occasional pains in the hands and feet, but nerve conduction studies were normal other than demonstrating carpal tunnel syndrome, which had been present on prior studies. On therapy his ALT level rose to approximately 140 IU/liter from a pretreatment level of 60 IU/liter and his HBV DNA level fell (from 249 to 4 pg/ml). His serum bilirubin level remained within the normal range on therapy. Echogenicity consistent with fatty infiltration of the liver was seen on an ultrasound performed on May 26 in a routine screening for hepatocellular carcinoma. These findings are nonspecific and poorly predictive of fatty liver, which is best documented by a liver biopsy. Repeated attempts to contact the patient were made on June 26 and 27. He was traveling and returned home late on the 28. He had noted increasing fatigue, anorexia, nausea, vomiting, abdominal pain, numbness, and tingling of the feet. Upon admission to NIH on June 29 examination revealed jaundice, hepatomegaly, and a peripheral neuropathy. His serum bilirubin level was 10.5 mg/dl, ALT was 121 IU/liter, prothrombin time was 15.1 seconds, and lactate was 8.3 mmol. Over the subsequent 2 days, he developed progressive liver failure, lactic acidosis, and renal failure. On June 30 he was transferred to Pittsburgh for liver transplant evaluation. Temporarily supported by an artificial hepatic assist device, he was HBsAg and HBeAg positive but HBV DNA negative. He underwent emergency liver transplantation on
OCR for page 85
Review of the Fialuridine (FIAU) Clinical Trials July 4, but was hypotensive and severely acidotic and the graft failed to function. The patient died of hepatic failure, acidosis, and circulatory collapse on July 5, 1993. Autopsy revealed hemorrhagic pancreatitis, severe pulmonary fibrosis, renal necrosis, and ischemic colitis. Liver explant showed chronic active hepatitis, marked cholestasis, and a mixed micro- and macrovesicular steatosis. Stains for HBsAg and HBcAg were positive. Liver Transplantation Of the five patients who underwent liver transplantation, only two patients survived (patient 3 at Emory University and patient 10 at the University of Virginia). Three patients who died following liver transplantation are described above (patients 6, 2, and 7). Patient 3 This 57-year-old man with chronic HBV infection had previously been enrolled in a ribavirin protocol at NIH. A liver biopsy before therapy reportedly showed chronic active hepatitis. A repeat biopsy prior to the R91 protocol showed chronic active hepatitis with fibrosis. As patient 1C in trial R91, he received 28 days of FIAU treatment in June 1992, with side effects of mild fatigue, irritability, and muscle cramps. During follow-up he had right upper quadrant pain. A laparoscopic cholecystectomy on October 29, 1992, revealed a single gallstone and gallbladder pathology showed cholecystitis. An intraoperative liver biopsy showed severe chronic active hepatitis with fibrosis. He was retreated with FIAU in the PPPC trial at a dose of 0.1 mg/kg/day from March 1993 until he was told to stop taking the drug on June 26, 1993 (doses of 121 mg in the R91 trial, and 678 mg in the PPPC trial, total dose, 799 mg). His serum bilirubin level remained within the normal range while he was on FIAU. His serum ALT level went from 163 IU/liter pretreatment to a maximum of 244 IU/liter. He developed symptoms of a peripheral neuropathy, but nerve conduction studies were inconclusive. He complained of cramps in his arms and legs on June 21, 1993. The patient underwent a full evaluation at the NIH Clinical Center on June 28. Initially, he had only mild hepatic impairment, but in the subsequent 2 weeks he had progressive jaundice and hypoprothrombinemia (serum bilirubin levels, 11.2 mg/dl; prothrombin time, 18.3 seconds) with lactic acidosis, although there were only minor changes in his serum ALT levels. His serum lipase level rose, but his amylase level remained normal. He was treated with oral vitamins, carnitine, and uridine, and intravenous thymidine and glucose. He was transferred to Emory University for liver transplant evaluation. His hepatic function continued to deteriorate and he underwent liver transplantation on August 4. Immediately pretransplant he was HBsAg and HBeAg positive but HBV DNA negative. He received perioperative and postoperative high dose hepatitis B immune globulin prophylaxis to prevent posttransplant HBV infection. He did well postoperatively and was discharged 17 days later. Pathology of the liver explant revealed chronic active hepatitis with cirrhosis, positive staining for HBsAg and HBcAg, cholangitis, cholestasis, and micro- and macrovesicular steatosis. He had persistent neuropathy following liver transplantation.
OCR for page 86
Review of the Fialuridine (FIAU) Clinical Trials Patient 10 This 63-year-old man with chronic HBV infection had been followed at NIH since 1988. A liver biopsy at that time showed mild chronic active hepatitis with fibrosis. He had been a subject in interferon, prednisone/interferon, and ribavirin trials at NIH and received a total of 154 mg of FIAU in the R91 study in June and July 1992. His only side effects were mild fatigue and irritability. Although his HBV DNA levels decreased with treatment, he continued to have active viral replication and was thus enrolled in the PPPC trial in April 1993. He received a total of 1,716 mg of FIAU between April 21 and June 26, when he was told to stop taking the drug (cumulative dose for both studies, 1,870 mg). Side effects during the second course of therapy included moderate fatigue, mild muscle weakness, and decreased libido. Before retreatment his serum ALT level was 42 IU/liter and serum bilirubin level was 1 mg/dl. He was admitted to NIH Clinical Center on June 28 complaining of fatigue, nausea, and weakness. His serum bilirubin level on admission was 3.9 mg/dl; his prothrombin time was normal, and his ALT level was 111 IU/liter. He was treated with oral carnitine and uridine, and intravenous glucose and thymidine. His hepatic function continued to deteriorate, with the development of jaundice (bilirubin, 35.7 mg/dl), ascites, hypoprothrombinemia (19.3 seconds), mild encephalopathy, and progressive lactic acidosis. On July 11 he was transferred to the University of Virginia for evaluation for liver transplantation. Because of deteriorating hepatic function and progressive encephalopathy, he required intracranial pressure monitoring and underwent successful liver transplantation on July 19. He had a complicated postoperative course but was eventually discharged 92 days following the transplantation. He received peri-and postoperative hepatitis B immunoprophylaxis and immunosuppressive therapy with cyclosporine and prednisone. The liver explant demonstrated cirrhosis with cholestasis, cholangitis, and marked microvesicular steatosis. He had persistent neuropathy following liver transplantation. Other Nonlethal Adverse Events In early June 1993 it became apparent that the 10 retreatment patients were having significant side effects, particularly nausea and peripheral neuropathy. Gastrointestinal side effects/pancreatitis The majority of patients (patients 1, 3, 4, 5, 6, and 7) had gastrointestinal complaints. Patient 3 complained of mild abdominal pain prior to retreatment with FIAU in March 1993. On therapy he complained of mild nausea and constipation. Patient 5 had abdominal discomfort and persistent hyperamylasemia. Hepatoxicity The mean rise in ALT levels in the seven patients who developed severe systemic toxicity was less than twice the pretreatment levels. In contrast, mean serum bilirubin levels
OCR for page 87
Review of the Fialuridine (FIAU) Clinical Trials in these same patients rose from within the normal range (less than 1.5 mg/dl) to greater than 15 mg/dl. Thus, hyperbilirubinemia, not marked aminotransferase elevation, is a feature of FIAU hepatotoxicity . Three surviving patients (patients 5, 8, and 9) who had received more than 2 months of FIAU treatment and who did not undergo liver transplantation had mild FIAU toxicity. ALT levels fell in patient 8, from a pretreatment level of 110 IU/liter to 80 IU/liter. The ALT level fell in patient 9, from a pretreatment level of approximately 360 IU/liter to less than 50 IU/liter. The ALT level remained less than 50 IU/liter throughout therapy in patient 5. Despite the mean fall in ALT levels in these patients while they were on therapy, microvesicular steatosis indicative of FIAU toxicity was present on liver biopsies performed subsequent to June 1993 in all three. Peripheral Neuropathy/Myopathy Several patients developed symptoms of a peripheral neuropathy while on FIAU therapy, but nerve conduction studies were normal or near normal. Fatigue Several patients (patients 6, 2, 3, 10, and 4) complained of fatigue, but given the nature of chronic HBV infection and the existence of fatigue before starting FIAU therapy in these patients, assignment of this symptom to FIAU therapy was and is problematic. The protocol specified a dose reduction for ''fatigue moderately affecting activity,'' but although fatigue in these patients was on occasion rated as moderate, there is no evidence that the fatigue "moderately affected activity." No Adverse Events All five patients who received FIAU for less than 1 month of FIAU (patients 11-15) were asymptomatic. RESPONSE TO THE EMERGENCY There is clear evidence that the investigators, medical staff fellows, and nurses at NIH were very involved in the care of the patients in this trial. Careful documentation of patient visits in CRFs, clinic charts, and flow sheets attests to the attention that was given to these patients. This was no more apparent than on June 25, 1993. After NIH was notified that patient 2 was hospitalized in Virginia with lactic acidosis, DiBisceglie drove to the hospital that evening (a trip of about 80 miles each way). Admittedly, this was an unusual, catastrophic event. Nevertheless, it would be erroneous to assume that all physicians, be they clinical researchers or community physicians, would respond in this manner. Hoofnagle was notified
OCR for page 88
Review of the Fialuridine (FIAU) Clinical Trials at home at 1:30 a.m. on June 26, 1993. He had the patients' updated flow sheets on hand to try to understand what was happening. Again, the fact that he had this information with him at home attests to his close involvement with the study and the patients. Hoofnagle and Straus met at NIH on the morning of June 26 and attempted to call all of the patients to instruct them to stop taking FIAU and to return to NIH for evaluation. All 15 patients in the trial were seen within a few days. The IOM committee would like to emphasize that the investigators promptly recognized the syndrome of FIAU toxicity in the second patient (patient 2), with patient 4D from R91 being the index patient, and that their subsequent response to the crisis was exemplary. The committee has only minor criticisms of the conduct of the trial at NIH. These would include patient 1A from R91, in whom the peripheral neuropathy was ascribed to alcohol rather than FIAU. It is likely that the difficult nature of the patient-physician interactions had a negative impact on the care of this patient. Following the termination of the PPPC trial, ACTG investigators involved in prior FIAC/FIAU trials were informed, and they attempted to contact all previously treated patients. Only then was the death of patient 408R in the R90 trial discovered. Even with hindsight and full knowledge of the syndrome of FIAU toxicity, only one of the deaths that had occurred before the PPPC trial was clearly FIAU related (patient 4D from the R91 trial). Three deaths were possibly FIAC/FIAU related (patient 107 from the R89 trial, patients 406 and 409 from the R90 trial) and two deaths were probably not FIAU related (patients 401 and 408 from the R90 trial). The principal investigators at NIH, Hoofnagle and Straus, concluded that the deaths in the PPPC trial were due to previously undescribed and unique toxicity which was progressive and irreversible despite the discontinuation of therapy. A difficult feature of this syndrome, which included hepatic failure, lactic acidosis, pancreatitis, neuropathy and myopathy, was that it became manifest even after the drug treatment was stopped. Neither the nature nor the severity of the toxicity was predicted by preclinical toxicology studies in animals. Accurate association between the drug and the toxic events was confounded by liver abnormalities associated with chronic HBV infection, rises in liver enzyme levels that were interpreted as a positive response to the drug (the flare), and confounding variables including treatment in previous FIAU/FIAC trials and with other nucleoside analogs known to cause neuropathies and pancreatitis (ddC and ddI). LONG-TERM FOLLOW-UP The 10 surviving patients (including two liver transplant recipients) are still being followed at NIH every 1 to 3 months. Those patients who have not previously been treated with interferon have been offered this therapy. Ongoing studies of these patients with nuclear magnetic resonance spectroscopy are attempting to elucidate the mechanism of toxicity of FIAU. Two patients who survived after liver transplantation have evidence of peripheral neuropathy and one patient has muscle weakness (myopathy). The other eight patients have no clinical evidence of FIAU toxicity. One patient (patient 9) became HBV DNA and HBeAg negative while on FIAU therapy and has remained negative, with normal ALT levels in follow-up examinations. Patients
OCR for page 89
Review of the Fialuridine (FIAU) Clinical Trials 5, 8, and 11-15 have evidence of ongoing HBV infection (HBsAg, HBeAg, and HBV DNA positive, with elevated liver enzyme levels). Patient 3, who underwent liver transplantation, is doing well on cyclosporine, azathioprine, and prednisone. He is receiving long-term hepatitis B immunoprophylaxis and remains HBsAg and HBV DNA negative. He has physical signs of an ongoing peripheral neuropathy, confirmed by nerve conduction studies. He continues to be followed at Emory University and NIH every 3 to 6 months. Liver function is now normal, and a recent liver biopsy showed only nonspecific changes. Patient 10 continues to be followed closely at NIH following liver transplantation. He has both clinical and electrophysiologic evidence of a peripheral neuropathy. His liver function is normal, and he remains HBsAg and HBV DNA negative on hepatitis B immunoprophylaxis. His immunosuppression consists of cyclosporine and prednisone treatment. He continues to be weak and to have difficulty walking, although he can drive a car. A follow-up liver biopsy has shown only nonspecific changes. He will continue to be followed at NIH and the University of Virginia on a regular basis. CONCLUSIONS Although the outcome was tragic, the IOM committee believes that there was sufficient justification for initiating the PPPC trial, that the protocol and consent form were appropriate, and that the conduct of the trial was equal to or better than the prevailing standards of care. There is no evidence that there was information available at the initiation of the PPPC trial or at any time during the trial itself that would have or should have averted the ultimate disaster.
Representative terms from entire chapter: