INTRODUCTION ----------------------------------------------------------------------------2

The purpose of this meeting is: 1) to update the Antivirals Advisory Committee on the data supporting the safety and efficacy of PEGASYS (Pegylated Interferon alfa 2a) in combination with COPEGUS(ribavirin) for the treatment of adult patients with chronic hepatitis C infection and 2) to discuss questions about dosing that may result in request for postmarketing studies.

On -------------------, Hoffman-LaRoche Corporation submitted to FDA a License Application for PEGASYS (peginterferon alfa 2a) and COPEGUS

(ribavirin, USP) combination therapy for the treatment of chronic hepatitis C.

Peginterferon alfa-2a (PEG-IFN) 180 mg administered sc qw

Ribavirin 200 mg/tablet administered po daily in split doses (final dose 1000 or 1200 mg depending on body weight) and taken with food.

Comparator product: Interferon alfa-2b 3 (IFN) 3 MIU administered sc tiw

Ribavirin 200 mg/capsule (Schering/ICN brand) administered po

daily in split doses (final dose 1000 or 1200 mg depending on body

weight) and taken with food.

HCV causes 20% of all cases of acute viral hepatitis and 70-90% of all cases of chronic viral hepatitis. The acute infection is asymptomatic in the majority of cases. Recovery from acute infection occurs in approximately 15% of adults with the remainder developing chronic hepatitis C infection. Patients with chronic HCV infection typically have persistent viremia and in most cases abnormal ALT levels.

In the US, approximately 50% of children and 85% of adults infected with HCV develop chronic hepatitis leading to an overall prevalence of 2.7 million individuals in the US with chronic hepatitis C infection. After decades, liver cirrhosis develops in about 20% of adults with chronic HCV infection within 20 years and 30% within 30years. In children with chronic HCV the incidence of cirrhosis is lower compared to adults possibly due to the shorter duration of infection or a slower progression of disease. The development of cirrhosis is associated with liver failure, with portal hypertension with its related complications and with the development of hepatocellular carcinoma. Chronic HCV infection causes approximately12,000 deaths per year and is the primary reason for liver transplantation in the US. Genotype 1 is the most prevalent HCV genotype in the US (approximately 70% overall) and is less likely to respond to treatment compared with HCV genotypes 2 and 3. In addition to genotype 1, other factors associated with diminished response to treatment are: high circulating HCV viral load (>2x 106 copies/ml serum) and the presence of cirrhosis on histology.

Alpha interferons were the first antiviral agents to be licensed for the treatment of patients with chronic hepatitis C. Alpha interferons can induce loss of detection of circulating HCV RNA, normalization of serum transaminase levels, and a modest reduction in liver inflammation. At 6 months following completion of alpha interferon monotherapy, the proportion of patients with undetectable HCV RNA is low (<20%). The addition of ribavirin to alpha interferon treatment roughly doubles the proportion of responders to approximately 40%. At the time the safety and efficacy trials for this submission were undertaken, alpha interferon and ribavirin represented the most efficacious therapy available for chronic hepatitis C. Peginterferon alfa 2b, both as monotherapy and in combination with ribavirin, was approved for the treatment of chronic hepatitis C in 2001.

Arguably, the risk/benefit of interferon alfa monotherapy for chronic HCV is no longer acceptable especially for patients with poor prognostic factors such as genotype 1 or high viral titers. Fewer than 10% of patients with chronic hepatitis C with cirrhosis or genotype 1 achieve sustained viral response with interferon alfa monotherapy.

There is no definitive evidence that the loss of detection of circulating HCV RNA six months after the end of treatment (sustained viral response), which is the primary outcome measure of most studies, will decrease the incidence of serious long-term outcomes including progression of liver disease to cirrhosis and the development of hepatocellular carcinoma.

Alpha interferons adversely affect a number of organ systems and may induce a number of serious toxicities including neuropsychiatric disorders, autoimmune disorders, ischemic disorders and disorders of host defenses. The majority of patients on alpha interferons experience fatigue and flu-like symptoms; alpha interferons also commonly induce nausea, vomiting, anorexia and weight loss. They may induce liver decompensation in patients with chronic liver failure. Interferons are abortifacients.

Toxic effects of ribavirin include hemolytic anemia, teratogenicity, embryocidal activity and mutagenesis. The main dose-limiting toxicity of ribavirin is hemolytic anemia which if severe may induce ischemic cardiac or neurologic events. The anemia is usually reversible following discontinuation of treatment.

Study Hypothesis: The pegylation of interferon alfa 2a will increase the treatment response and tolerability of that agent against chronic hepatitis C due to the improved pharmacokinetic properties that result from pegylation with increased exposure to more consistent serum concentration for 7 days and the possibility of once a week dosing.

The clinical development program comprised 19 phase 1 clinical pharmacology studies (14 to support PEG-IFN alfa-2a and 5 to support Ribavirin). The 14 clinical pharmacology studies of PEG-IFN alfa-2a were part of the monotherapy development program. Table 1 lists the principal clinical safety and efficacy studies.

Including patients participating in the two pivotal studies to be discussed, a total of 1377 patients have received PEG-IFN alfa 2a alone in a range of doses but predominantly 180mg sc as part of the phase 1 clinical pharmacology studies; 357 patients have been exposed to Roche’s ribavirin alone and 1843 have been exposed to the combination of PEG-IFN-alfa 2a plus Roche’s ribavirin

The phase 2 study, NV15800 was a randomized, open label study in patients with chronic hepatitis C without cirrhosis in which all patients received PEG-IFN alfa-2a and ribavirin combination therapy. The major focus of this study was to ascertain the impact of food upon ribavirin absorption as well as the safety of the combination of peginterferon alfa 2a and ribavirin. Given the very few patients studied, the conclusions that can be drawn from this study are very limited. The study will not be considered further here. No dose-ranging studies of PEG-IFN in combination with ribavirin were carried out.

The dose of peginterferon was selected on the basis of analyses of data of peginterferon monotherapy. The efficacy of peginterferon alfa-2a monotherapy in adult patients not previously treated was evaluated in four clinical studies. These trials were comparative, open-label, randomized, with 48-week of therapy followed by 24 weeks of follow-up. A total of 1600 patients were enrolled, 996 patients were treated with PEGASYS.

The number of responders to peginterferon alfa-2a monotherapy was higher than the number of responders to unpegylated interferon alfa-2a monotherapy. The addition of ribavirin to unpegylated interferon alfa increases the number of responders. The sponsor hypothesized that the addition of ribavirin to peginterferon alfa-2a would further increase the number of responders.

The rationale for dose of peginterferon alfa-2a selected for the phase 3 studies was in part based on the analysis of treatment response and safety in peginterferon monotherapy study NV15496. In that study the sustained viral response achieved by both 135mg and 180mg doses of peginterferon once weekly SC were indistinguishable. The sponsor observed that the histologic responses in the peginterferon 135mg group and the interferon alfa-2a group were not different. However the histologic response of the peginterferon 180mg group appeared to be higher than that of the interferon alfa-2a group. The incidence of adverse events appeared to be similar in the 135mg and 180mg peginterferon dose groups. The sponsor concluded that the 180 mg peginterferon dose should be tested in the pivotal studies.

The dose of ribavirin chosen for study was 1000 or 1200 mg po daily based on body weight < or > 75 kg. This is the recommended dose for interferon alfa-2b and ribavirin combination therapy. Patients were instructed to take ribavirin with food because of evidence that food increases the bioavailability of ribavirin.

"A Phase III, Randomized, Multicenter Efficacy and Safety Study

Comparing the Combination of Pegylated-Interferon alfa 2a and Ribavirin

to REBETRON in the Treatment of Patients with Chronic HCV Infection"

Randomized, multicenter, international, partially blinded, active controlled (REBETRON) phase 3 study of two Peginterferon alfa 2a regimens in 1149 treatment naïve patients with histologically, serologically and virologically documented chronic hepatitis C. Subjects were randomly assigned to one of three treatment arms: Peginterferon alfa 2a monotherapy, peginterferon alfa 2a combined with ribavirin, or Intron A‚/Rebetol‚ on a 1:2:2 basis. Subjects were treated for 48 weeks and were followed for 24 weeks.

Ribavirin

The dose adjustment guidelines for ribavirin were the same across study groups. Ribavirin was to be reduced to 600 mg per day if (1) a patient without significant cardiovascular disease experienced a fall in hemoglobin to <10 g/dL and >8.5 g/dL or (2) a patient with stable cardiovascular disease experienced a fall in hemoglobin by >2 g/dL during any 4 weeks of treatment. Once a patient’s ribavirin dose was held due to either a lab abnormality or adverse event, the investigator could attempt to increase the dose of ribavirin to 600 mg daily, and further to 800 mg daily. However, it was not recommended that ribavirin be increased to its original assigned dose (1000-1200 mg).

Peginterferon and interferon

If a patient experienced a clinically significant adverse event or a laboratory abnormality, the dose of peginterferon or interferon was lowered to pre-specified levels

based on the degree of severity of the event as shown below.

The table below shows the guidelines for lowering the dose of peginterferon when patients experienced neutropenia or thrombocytopenia.

The dose of interferon was to be reduced to 1.5 MIU tiw if a patient experienced a fall in white blood count to <1.5 X 109/l ( but >1.0x109/l) or a fall in neutrophil count to <.75x109/l (but >.5x109/l) or a platelet count to <50x109/l (but >25x109/l)

If four or more consecutive doses of test drug were held or otherwise not administered (i.e. the patient had not received test medication for more than 28 days), the patient was considered intolerant of the test medication or non-compliant.

Men and women outpatients ≥ 18 years of age with serologically and histologically proven CHC, elevated ALT, and detectable HCV RNA. Excluded were patients with HIV, other forms of hepatitis, a history of severe psychiatric or cardiac disease, severe seizure disorders, severe retinopathy, or previous interferon or ribavirin therapy.

The primary outcome was a combined endpoint of non-detectable serum HCV-RNA (<100 copies/ml by a commercial PCR assay) and normal serum ALT activity after the 24-week treatment-free follow-up period (nominally 72 weeks).

The following were the principal secondary outcomes:

-End-of-treatment virologic and biochemical responses

-Sustained virological and biochemical responses

-End-of-follow-up histological changes from pretreatment.

In both studies, study subjects who did not demonstrate either a virologic response (HCV negative or ≥2 log10 decrease ) or a biochemical response(normal ALT) could be withdrawn from the study by 12 weeks of therapy and were to be withdrawn from study if unresponsiveness persisted by 24 weeks.

One certified laboratory performed all HCV RNA PCR determinations. Local laboratories could be used for repeat testing required for safety assessments.

Patients were assessed for safety at weeks 1, 2, 4, 6 and 8 and then every 4 weeks during treatment and follow-up. Patients who were prematurely withdrawn from treatment were followed for 12 weeks after treatment. Following the initiation of treatment, quantitative viral titers were also obtained at weeks 4, 12 and 24 only if the qualitative titer was positive. Qualitative titers (AMPLICOR) were obtained at weeks 4, 12, 24, 48, 60 and 72. Serum ALT activities were obtained by the same schedule for safety assessments. The liver biopsy at baseline and week 72 was reviewed by one pathologist.

Primary efficacy analysis

Amendment D: Three hundred additional patients were added to protocol to ensure that the study could detect a statistically significant treatment difference of at least 10% improvement of PEG-IFN plus ribavirin over REBETRON. This decision was based on a blinded interim analysis of week 4 virological data. The total study population was raised to 1125 and the 1:2:2 randomization was retained. A liver biopsy at 48 weeks in a portion of patients was added (50 in each arm) to evaluate safety issues. The week 12 discontinuation policy for non-responders was made optional in light of literature reports of success with 24 weeks.

Amendment E, G: Previously, 150 patients (13% of total) were scheduled to have an end of treatment biopsy. To increase concordance with other studies, the numbers and timing of liver biopsies were changed to 225 patients or 20% at end of follow-up. Twenty patients treated with Rebetron‚ would have end-of-treatment liver biopsies.

Amendment H: The definition of sustained virologic response was two consecutive HCV RNA measurements ≥ 21 days apart measured ≥ 408 days with the lower limit of week 60. To be considered a responder, a patient had to have normal serum ALT activity and no detectable virus at the end of untreated-follow-up period

There were 81 participating study centers in US, Europe, Australia, Taiwan, Mexico, and Brazil. Differences in treatment response between geographic regions have been observed in studies of interferon therapies. These differences are attributable at least in part to differences in the distribution of baseline demographic and disease factors that predict response to treatment. For this reason the distribution of baseline variables, the treatment responses, and the safety data were analyzed by subdividing the study population into U.S., non-US, Europe, and "Other". The centers in the US contributed 41% of the patients participating in the study.

1149 patients were randomized; 28 patients were randomized but did not receive any treatment. The two comparison arms had equivalent numbers of individuals randomized but untreated. Table 3 shows that somewhat fewer subjects in the PEG-IFN monotherapy and IFN-alfa 2b+ ribavirin arms completed 48 weeks of treatment compared to the PEG-IFN +ribavirin arm (67% versus 76%). This difference was due to higher numbers of subjects withdrawn from the protocol for futility at 24 weeks in those two arms. The remainder of the withdrawals from the protocol were mostly for adverse events or laboratory abnormalities (See review of safety).

The mean age of study participants was approximately 43 years and this mean is within the age group (30-49 years of age) with the highest HCV prevalence in the US. The study protocol excluded pediatric patients. There was no upper age limit but very few elderly patients were enrolled. The mean weight of participants was approximately 79 kg in all three arms with the recorded range from 42 to 156 kg. The subjects enrolled in US centers were on average heavier than subjects enrolled in non-US centers (See Table 4). Peginterferon alfa-2a and interferon alfa-2b were administered as fixed doses and the ribavirin was crudely weight-adjusted based on body weight above or below 75 kg. Given the wide range in body weights of study subjects, the differences in body weight between subjects in US and Non-US centers, and the fixed or crude weight-adjusted dosing, the analysis of safety and efficacy by body weight would be an important review objective.

Approximately 2/3 of the participants were men. This gender distribution is consistent with the higher prevalence of HCV in men. Approximately 84% of participants were Caucasian. Overall 5% of participants were black, (11% of US participants). Approximately 5% of participants both in the total population and in the US sites were Hispanic. Since the prevalence of HCV in the US is higher in African Americans and in Hispanics compared to Caucasians, these two ethnic minorities were under-represented in the study. Response to interferon therapy is lower in African Americans and in Hispanics for reasons not fully understood. It should be noted that the sponsor has already undertaken studies to assess the safety and efficacy of peginterferon alfa-2a and ribavirin combination therapy in patients from these ethnic groups.

Table 4 summarizes the patient demographics in the treated population. US patients weigh more, have somewhat higher body mass index and are also older compared to Non-Us patients.

a Analysis group: Treated patients

The demographic characteristics were reasonably well distributed across the treatment groups (not shown).

The means of acquisition of HCV infection was similar across treatment groups (Table 5). The predominant mode of transmission was injection drug use. The second highest source of HCV in participants was sporadic or unknown and the third most common was transfused blood products.

Table 5. Infection Source and HCV Genotype PEG-IFN alfa-2a N = 227 N % PEG-IFN alfa-2a Ribavirin 1000-1200mg N = 465 N % IFN alfa-2b Ribavirin 1000-1200mg N = 457 N % Infection Source             Injection drug use 82 (36) 197 (42) 182 (40) Transfusion 47 (21) 87 (19) 101 (22) Unknown 72 (32) 134 (29) 127 (28)

Approximately two thirds of participants in the study were infected with genotype 1 (Table 6); the number was somewhat higher in US study patients compared to non-US patients (see Table 7). The distribution of genotype 1 observed in the study patients is consistent with the incidence of this genotype in HCV infected patients in the US population. Among the non-genotype 1, only types 2 and 3 were significantly represented in the study. The proportion of patients with high viral load was evenly distributed across the three study arms for the entire population and in geographic subgroups. No relationship has been established between the rate of disease progression or severity of liver disease and the HCV genotype or the number of circulating HCV-RNA copies. Both viral genotype and viral load predict response to interferon treatment, however, with high viral load genotype 1 having the least likelihood of response to interferon treatment.

It is also established that patients with cirrhosis have a lower likelihood of responding to interferon therapy. The proportion of patients who had cirrhosis at baseline was numerically similar across the study arms with 25 (17%) of PEG-IFN monotherapy, 40 (13%) of PEG-IFN/ribavirin and 32 or (11%) of the Interferon alfa-2b/ribavirin having histologic evidence of cirrhosis at baseline. The incidence of cirrhosis was numerically lower in Non-US centers (Europe in particular) compared to the US centers (Table7).

a Analysis group: Treated patients

The primary comparison in this study was between PEG-IFN alfa 2a plus ribavirin and IFN alfa 2b plus ribavirin. The purpose of the PEG-IFN monotherapy arm was to provide information on the impact of ribavirin on both efficacy and safety when used in combination with PEG-IFN. The intent to treat population was used for the efficacy analysis and was defined as all subjects randomized whether or not they received treatment.

The primary endpoint in this study was the combined sustained (24-weeks post-treatment) virologic and biochemical response. Table 8 shows a numerical difference (6%) that favors peginterferon alfa-2a and ribavirin over interferon alfa-2b and ribavirin. The p value for the difference was 0.057. Peginterferon alfa-2a and ribavirin combination therapy was superior to peginterferon alfa-2a monotherapy.

1Intent- to-treat population

2Global test of treatment difference: P<0.001

Table 9 shows a secondary efficacy analysis using sustained virologic response at 24 weeks post treatment. In this analysis, the treatment difference between PEG-IFN alfa 2a plus ribavirin and IFN-alfa-2b plus ribavirin is 8% and this difference does achieve statistical significance. Both regimens performed better than the PEG-IFN monotherapy which was consistent with expectations. There is a good concordance between biochemical and virologic response. However in a few patients who achieved a sustained virologic response, transaminase levels had not returned to normal by endpoint. The possibility exists that in some of these cases the ALT elevation might be unrelated to CHC.

1Intent- to-treat population

2Global test of treatment difference: P<0.001

Other secondary endpoint consider virologic and biochemical response at either the end of follow-up (72 wks) or at the end of therapy (48 wks) and are shown in Table 10. These endpoints provided evidence in support of the superiority of peginterferon alfa-2a and ribavirin, over interferon alfa-2b and ribavirin.

1Intent-to-treat population

Efficacy endpoints for pivotal trials of interferon-based therapies for CHC have evolved with the advent of validated assays for HCV RNA. The first interferon therapies were licensed based on demonstration of normalization of serum transaminase levels (biochemical response) supported by improvement in liver histology. Subsequently biochemical response combined with loss of detection of HCV RNA (virologic response) became a standard primary efficacy outcome. The agency’s current thinking is that virologic response alone is appropriate as a primary efficacy endpoint. The Anti-Viral Drugs Advisory Committee on December 12, 2001, endorsed the use of a sustained virologic response 24 weeks after the completion of treatment. While the p value for the pre-specified primary endpoint comparing PEGASYS and COPEGUS with REBETRON was 0.057 , a number of secondary endpoints including virologic response alone or biochemical response alone provided strong support for the superiority of peginterferon alfa-2a and ribavirin over interferon alfa-2b and ribavirin.

Finally, changes in liver histopathology in response to treatment were evaluated in a subset of patients. A liver biopsy with histology consistent with chronic hepatitis C within 12 months of enrollment was required for enrollment in the study. Paired liver biopsies were obtained after treatment in approximately 20% of patients. Response to treatment based on Histologic Activity Score was defined as >2 point decrease in score from baseline. Over 70% of patients in all treatment groups were observed to have a modest reduction in inflammation compared to baseline (Table 11). Approximately 20% of patients had no change in inflammation score and 3-5% had an increase in inflammation score.

When histologic response is compared to sustained viral response it would appear that those patients who had a sustained virologic response are more likely to have a histologic response. Overall, the number of subjects with paired biopsies is small, the improvement in histology from baseline is modest and there are no statistically significant differences between the treatment groups.

Patients with HCV RNA negative at the end of treatment but positive at the end of follow-up (24 weeks post-treatment) were considered to have relapsed. The proportion of patients who relapsed (positive or missing HCV-RNA) was highest in the PEG-IFN monotherapy group. As for the two combination therapy groups, the relapse rate in the PEG-IFN alfa 2a/ribavirin group was higher than in the IFN alfa 2b/ribavirin group (see Table 12). It should also be noted that the end of treatment response rate of the PEG-IFN alfa 2a/ ribavirin group was higher than that of the IFN-alfa 2b/ribavirin group.

The majority of individuals who ultimately achieve a sustained virologic response have negative HCV-RNA at 12 weeks of therapy. Those proportions were 75% in the PEG-IFN monotherapy group, 84% in the PEG-IFN/ribavirin group and 78% in the IFN/ribavirin group. By 24 weeks, those numbers had increased to 91% in the PEG-IFN monotherapy group, and 95% in the two combination therapy groups.

The attainment of a negative HCV-RNA at 12 weeks also has predictive value for the success of a subject in achieving a sustained viral response (see Table 13). In the PEG-IFN alfa 2a plus ribavirin group, 205 of the 279 patients who were HCV RNA negative at 12 weeks achieved sustained virologic response for a positive predictive value of 73%. On the other side, 134 of 174 patients who ultimately did not achieve sustained virologic response had a positive HCV-RNA at 12 weeks, a 77% negative predictive value. Those same positive and negative predictive values for the IFN-alfa 2b/ ribavirin were similar with a 74% positive predictive value and 82% negative predictive value. In contrast, the positive predictive value for the monotherapy arm was 54% with a negative predictive value of 88% presumably reflecting the lower efficacy of the monotherapy.

This study allowed for continuation of participation beyond 24 weeks for patients who had either evidence of virologic response (negative HCV RNA or ≥2 log 10 drop in viral titer) or evidence of a biochemical response (normalization of ALT). Expanding the definition of viral response to include subjects who have a positive HCV RNA but whose viral titer has decreased ≥2 log 10 permits even greater negative predictive value. As shown in Table 14, of the subjects receiving PEG-IFN alfa 2a/ribavirin only two not meeting this definition ultimately had a sustained viral response for a negative predictive value of 97%.

a2 consecutive HCV RNA negative at >12 wks post-treatment

Patients with viral genotype 1, regardless of viral load, had a lower response rate compared to patients with other genotypes (Table 15). Patients with high viral load (>2x106 copies of HCV RNA/ml serum) also had lower responses compared to patients with low viral load.

Patients with both genotype 1 and high viral load were least likely to respond both in the PEG-IFN alfa 2a/ribavirin group (39%) and in the IFN/ribavirin group (32%). The influence of viral load appeared to be less in non-genotype 1 patients in the PEG-IFN/ ribavirin group.

1Intent-to-treat population

The effect of age, gender, ethnicity, cirrhosis on treatment response was examined in the three study groups. The effects were consistent across groups and were also consistent with previous observations (see Table 16). The effect of these factors on treatment response was also examined by geographic region and did not yield notable new hypotheses (not shown).

In general, younger patients had higher treatment responses with all three regimens compared to older patients (Table 16). The efficacy in US patients was lower than in non-US at most age categories.

atreated population

Overall there was little difference between the SVR rate of men and women in this study.

Response rates were lower in Black and Hispanic patients compared with the response rate in Caucasians. African Americans made up the vast majority of Black patients enrolled (48/53). The response rate in Asian patients on the other hand was higher than Caucasians. The validity of these observations is somewhat questionable due to the very low numbers of minority patients that were enrolled in this study.

The unfavorable effect of cirrhosis on treatment response was observed in all three treatment arms. In each arm, individuals with cirrhosis had a 9-14 point lower percentage of viral response. PEG-IFN alfa-2a and ribavirin combination therapy induced higher responses than IFN alfa 2b and ribavirin combination therapy (41% versus 35%). The US versus non-US trends favoring the latter are seen again in these subgroups.

Roughly 40% of study patients were enrolled in the US; 40% of the other patients were enrolled in Europe, and roughly 20% were enrolled in Australia, Taiwan, Mexico, Brazil. Table 16 shows that overall the proportion of virologic responders in each of the treatment arms was higher in patients enrolled at non-US centers compared to patients enrolled at US centers. Various prognostic factors were explored to examine this treatment difference. The difference in response was attributable to a higher proportion in US compared to non-US in poor prognostic baseline variables including older age, higher body weight, higher proportion of HCV1 or cirrhosis.

Exploratory analyses were conducted to analyze treatment effect in US and non-US populations in subgroups defined by body weight, genotype, and baseline viral titer. These analyses are provided in Appendix 1 (see Tables 17 and 18). A logistic regression analysis including genotype, viral titer, geographic region, and weight found that the geographic region was not a significant factor contributing to treatment response. Thus, the differences in sustained virological response for the overall population observed between US and non-US patients appear to be largely due to imbalance in other important prognostic factors for response.

Nearly all patients experienced one or more adverse events (Table 19). The incidence of severe adverse events was similar across groups. Serious adverse events were numerically higher in the peginterferon groups (12%) compared to the interferon/ribavirin group (9%). The incidence of treatment discontinuation for adverse events or laboratory abnormalities was 10-11% in the combination therapy groups and 7% in the peginterferon monotherapy group overall. The incidence of dose modification of the interferon therapy for adverse event or laboratory abnormality was higher for peginterferon (27-32%) compared to interferon (18%). The most common reason for dose modification was neutropenia and/or thrombocytopenia. The incidence of withdrawal from treatment was not different between the interferon combination arms. Also, the incidence of treatment discontinuation or modification was not different in the US centers compared to non-US centers. The higher rates of SAEs and dose modification suggest the potential for greater toxicity of PEG-IFN and ribavirin compared to IFN and ribavirin.

The incidence of new adverse events was summarized for the treatment period, the post-treatment period and the treatment and post-treatment periods combined

Table 20 shows that the incidence of the most common adverse events for the treatment plus post-treatment period was not higher in the peginterferon/ribavirin group compared to the interferon/ribavirin group.

The most common adverse event was a systemic reaction consisting of fatigue/weakness/asthenia with or without headache, fever, chills/rigors. The second most common group of adverse events was neuropsychiatric with irritability, insomnia, impaired concentration/memory impairment, depression, anxiety/nervousness. The musculoskeletal system was commonly affected with myalgias and arthralgia. GI manifestations were also prominent (nausea, vomiting, anorexia, weight loss, abdominal pain, diarrhea). Finally skin manifestations included alopecia, dermatitis, pruritus, and injection site reactions.

aincidence>5%, listed in discending order

There were 3 deaths occurring in patients participating in this study, none of which occurred during treatment. Two deaths occurred during the post treatment follow-up period. The third death occurred approximately a year after treatment was stopped, was caused by a hepatoma that first became manifest near the end of her treatment program. Screening CT scans at study enrollment and 6 months into treatment had failed to detect this mass. Of the other two deaths, one also in the monotherapy arm was a drowning associated with a motor vehicle accident. There was no concern that this represented a suicide. The final death occurred in an IFN combination arm patient who withdrew early from study because of irritability. His death occurred approximately 2 months after study withdrawal and was unobserved. Autopsy findings revealed cocaine cross reactives and 280 ng/mL of benzoylecognine but no free cocaine or cocaethylene in the blood. Autopsy findings of note were limited to the cardiovascular system. There was bilateral ventricular enlargement with left ventricular dilatation and hypertrophy. It was judged that his death was caused by hypertensive heart disease rather than cocaine overdose because of the lack of free cocaine in his system.

Individual serious adverse events were rare (<1%) and were primarily infectious, neuropsychiatric, gastrointestinal, inflammatory and hematologic disorders (see Table 21). Serious infections will be discussed separately. Serious neuropsychiatric disorders occurred in all three study arms and included depression, bipolar disorder, psychosis, suicide attempt, addiction relapse, fatigue, trigeminal neuralgia. Abdominal pain without identified cause occurred in all three arms as well as events suggestive of subjects’ underlying liver disease such as hemorrhoidal bleeding and hematemesis. Inflammatory or rheumatologic serious adverse events such as thyroiditis, pericarditis, arthralgias, new onset psoriasis, interstitial pneumonitis, polyarthritis, sarcoidosis occurred in all three arms but to a somewhat greater extent in the PEG-IFN arms. There were sporadic cases likely caused by ischemic or inflammatory mechanisms related to the interferons such as optic neuritis, retinal hemorrhage, and myocardial ischemia.

Infections were more common in both PEG-IFN arms than in the IFN alfa 2b combination therapy arm. Bacterial causes, either proven or empirically treated predominated and where a causative organism was recovered it was likely part of the patient’s normal bacterial flora., it was likely part of the patient’s normal bacterial flora The most likely explanation for apparent impairment of normal host resistance in this setting would be the neutropenia commonly caused by interferon therapy.

As shown in Table 22, grade 4 neutropenia (<500 ANC) was detected in a case of amoebiasis in the PEG-IFN monotherapy arm and in a case of epiglottis caused by Staphyloccus aureus in the PEG-IFN combination arm. These may not be the only instances of a severe neutropenia playing a role in these serious infections since as is also seen in Table 22, there is often a significant time interval both before and after the onset of a serious infection when the neutrophil count is not recorded. Both of the two cases where grade 4 neutropenia is detected are unusual for the timeliness of the leukocyte sampling. Even when neutropenia is not documented, expected neutrophil response to serious infection often appears to be blunted. It is possible that sampling delays may be playing a role in this observation but this blunting is seen predominantly with the PEG-IFN arms and is not seen to the same extent in the IFN/ribavirin arm.

Significant and severe lymphopenia is much more common, especially in the 16 patients with serious infections in the PEG-IFN ribavirin combination arm. Of these 16 individuals, 10 had lymphocyte counts recorded below 1000. The significance of this lymphopenia to the subject’s serious infection is not known since most of the affected individuals appeared to be exhibiting either a bacterial infection or an inflammatory non-infectious entity.

A similar proportion of patients withdrew from treatment for clinical adverse events and

laboratory abnormalities in each of the combination therapy treatment groups (10% and

11%), whereas fewer patients withdrew from PEG-IFN alfa-2a monotherapy for these

reasons (7%). Most adverse events and laboratory abnormalities were successfully managed with dose reduction and therefore did not require withdrawal of the

patient .

The most common types of events leading to discontinuation were psychiatric disorders

(mainly depression-related events), especially in the combination therapy groups. The

highest proportion of patients discontinuing due to psychiatric events was in the

IFN alfa-2b plus ribavirin group (PEG-IFN alfa-2a monotherapy, 1%; PEG-IFN alfa-2a

plus ribavirin, 3%; IFN alfa-2b plus ribavirin, 4%). Other events leading to withdrawal

only in the combination groups and not in the monotherapy group include thyroid

disorders (8 patients), anemia, and dyspnea (both 3 patients), suggesting a possible causal

or aggravating role of ribavirin in these events. More patients in the PEG-IFN alfa-2a plus ribavirin group withdrew for hematologic abnormalities (2%) than in the other groups (both 1%), with thrombocytopenia and neutropenia being the most frequent abnormalities leading to withdrawal. All other types of events leading to discontinuation occurred in £1% of patients in each treatment arm.

In the peginterferon alfa-2a and interferon alfa-2b combination therapy groups the main adverse events that led to treatment discontinuation were neuropsychiatric disorders (3%) including suicide attempt, depression, fatigue, anxiety. The second most common causes were bone marrow toxicity/hemolysis (2%) with neutropenia, thrombocytopenia, pancytopenia, and anemia. Rarer causes (<1%) included the following: dermatitis, arthritis, colitis, and endocrine/metabolic disorders such as autoimmune thyroiditis (with hypo- or hyper-thyroidism), hypertriglyceridemia, exacerbation of diabetes mellitus (ketoacidosis, retinopathy).

The incidence of adverse events that led to modification of interferon dosage was higher in the combination therapy groups (11%) compared to the monotherapy group (6%). The incidence of all adverse events that led to modification of ribavirin dosage was similar (17-22%) across the three study groups (see Table 24).

Very few individuals were withdrawn for neutropenia, thrombocytopenia or anemia. The incidence of clinically significant laboratory abnormalities that led to modification of peginterferon dosage was 3-fold higher (24-25%) in the patients receiving PEG-IFN compared to patients receiving IFN (8%). The incidence of lab abnormalities leading to modification of ribavirin dosage was lowest in the peginterferon /placebo ribavirin group (4%), intermediate in the interferon/ribavirin group (19%) and highest in the peginterferon/ribavirin group (24%). Bone marrow toxicity toxicity (neutropenia, thrombocytopenia) was the principal cause for modification of interferon dose and was several-fold higher in patients receiving peginterferon compared to interferon (see Table 24). Anemia was the main cause of dose modification of ribavirin (approximately 20%) in patients receiving PEG-IFN or IFN plus ribavirin.

Analysis of the impact of dose modification of PEG-IFN upon treatment response using cumulative dose calculations demonstrates a proportional relationship with lower overall treatment response associated with lower cumulative doses of PEG-IFN in both the monotherapy and combination therapy arms.

Median neutrophil counts decreased from baseline values in all three treatment groups,

with the largest decreases occurring during the first 2 weeks of treatment. Median neutrophil counts then stabilized, remaining at around 45% of baseline in the two PEG-IFN alfa-2a groups and around 60% of baseline in the IFN alfa-2b group during the remainder of the 48-week treatment period. Median neutrophil counts increased following the end of treatment, returning to normal values by week 52 and reaching baseline levels by week 72 (see Figure 1).

Nearly 50% of subjects in the PEG-IFN arms experienced Grade 3 or higher neutropenia during treatment or in follow-up. By comparison, less than 25% of subjects in the IFN-alfa 2b combination arms experienced that level of neutropenia (see Table 26). The impact that this level of neutropenia had on infections appears to have been small in this study.

	Normal >2a N %	Grade 1 1.5-1.99 N %	Grade 2 1.0-1.49 N %	Grade 3 0.5-0.99 N %	Grade 4 <0.5 N %
PEG-IFN alfa-2a	13 (5.8)	25 (11.2)	96 (43.0)	81 (36.3)	8 (3.6)
PEG-IFN alfa-2a ribavirin	22 (4.9)	37 (8.2)	165 (36.7)	205 (45.6)	21 (4.7)
IFNalfa-2b	69 (15.6)	105 (23.8)	169 (38.2)	94 (21.3)	5 (1.1)

acounts x109/L

Median platelet counts remained close to baseline levels throughout treatment in the

IFN alfa-2b plus ribavirin arm. Median platelet counts decreased progressively during

the first 8 weeks of treatment in the PEG-IFN alfa-2a monotherapy and PEG-IFN alfa-2a

plus ribavirin treatment groups, stabilizing at about 55% of baseline levels in the

PEG-IFN alfa-2a plus monotherapy group and approximately 70% of baseline values in

the PEG-IFN alfa-2a plus ribavirin group. Median platelet counts normalized by 4 weeks

post-treatment and had recovered to baseline levels by week 72 (see Figure 2).

Patients receiving peginterferon with or without ribavirin had a two- to three-fold higher incidence of thrombocytopenia and were more likely (5% vs. 0.2%) to reach grade 3 thrombocytopenia defined as counts between 20,000 and 50,000/mm3 (see Table 27) There were no episodes of significant bleeding attributable to the thrombocytopenia associated with the PEG-IFN treatment.

	Normal >100a N %	Grade 1 75<100 N %	Grade 2 50 <75 N %	Grade 3 20<50 N %	Grade 4 <20 N %
PEG-IFN alfa-2a	108(48)b	65(29)	36(16)	14(6)	0
PEG-IFN alfa-2a ribavirin	303(67)	80(17)	45(10)	22(4)	0
IFNalfa-2b	388(88)	36(8)	17(4)	1(0.2)	0

acountsx109/L

The median hemoglobin concentration decreased in all treatment groups during the first 8

weeks of treatment. The decrease was similar in the PEG-IFN alfa-2a plus ribavirin and

IFN alfa-2b plus ribavirin groups, with the median hemoglobin concentration stabilizing

at approximately 12.5 g/dL from a baseline of approximately 15 g/dL. The decrease in

the PEG-IFN alfa-2a monotherapy arm was less marked (stabilizing at around 14 g/dL).

Median hemoglobin concentrations returned to baseline levels within 8 weeks post-treatment in all treatment groups (not shown).

Approximately 10% of patients receiving PEG-IFN or IFN plus ribavirin experienced drops in hemoglobin to levels below 10g/dl compared to 3% in the peginterferon monotherapy group (Table 28). Two percent of patients in the peginterferon ribavirin group reached hemoglobin levels < 8.5 g. The maximum drop was 9.5g.

	8.5 <10 g/dl N %	< 8.5 g/dl N %
PEG-IFN alfa-2a	7 (3.1)	1 (0.4)
PEG-IFN alfa-2a ribavirin	40 (9)	9 (2)
IFNalfa-2b	47 (11)	1 (0.2)

As has been previously mentioned, the interferons are administered as a fixed dosage in this study. The ribavirin is weight crudely weight adjusted above and below 75kg. There is a suggestion that efficacy is affected by body weight. An exploratory analysis was undertaken to examine the effect of Body Mass Index (BMI) upon the incidence of the four most frequently encountered hematologic effects adverse effects with interferon therapy: Hemoglobin < 10, Grade 4 Neutropenia, Grade 3 /4 Neutropenia, and Grade 3 /4 thrombocytopenia.

As demonstrated in Table 29, significant anemia (hemoglobin <10) was seen more frequently in subjects with BMI <25 than in those with BMI≥ 25 in all three study arms, approximately 4% greater incidence in the lower BMI. This same trend is seen in Grade 3/ 4 neutropenia although when Grade 4 Neutropenia is isolated, the greater incidence in the lower BMI is only seen in the PEG-IFN combination arm. With thrombocytopenia, the only significant difference is in the PEG-IFN monotherapy arm and then it occurs in the opposite direction, the greater incidence is in larger BMI patients.

The incidence of dose modification of ribavirin for severe anemia (not shown) were similar in the two combination therapy groups and were appropriately lower in the monotherapy group. It could not be determined how consistently protocol-required dose-modifications were applied across geographic regions. Median decreases in hemoglobin, were greater in the non-US centers versus the US centers.

The presence of antibodies to interferon was measured comparing baseline and week 56 testing. A total of 14 patients (4.8%) in the Peg-IFN combination arm developed neutralizing anti-interferon antibodies ranging in titer from 117 to 2600 interferon neutralizing units per milliliter (INU/mL) Anti-interferon neutralizing antibodies were detected in 3 patients(1.1%) of the IFN-alfa 2b plus ribavirin group and 3 patients(2.3%) in the monotherapy group. It should be noted that the assay for neutralizing antibodies to the IFN-alfa 2b is not validated. The immunogenicity of PEG-IFN in combination with ribavirin requires further study.

The development of neutralizing antibodies did not preclude achieving a sustained virologic response. Of the 14 individuals detected in the PEG-IFN combination arm to have interferon neutralizing antibodies, 9 achieved a sustained virologic response. In addition, 2/3 monotherapy patients and 1/3 IFN alfa 2b/ribavirin combination patients also achieved a sustained virological response (Table 30)

A higher proportion of cirrhotic patients withdrew from therapy in the combination

treatment arms (9% and 11%) than in the PEG-IFN alfa-2a monotherapy arm (6%)

The proportion was similar to that seen in the overall safety population. Reasons for withdrawal included hematologic abnormalities (anemia and thrombocytopenia) and psychiatric events.

As in the overall safety population, a higher proportion of cirrhotic patients required

modification of their interferon dose for clinical adverse events and laboratory

abnormalities in the PEG-IFN alfa-2a monotherapy and combination therapy arms than in

the IFN alfa-2b plus ribavirin arm. Cirrhotic patients were more likely than noncirrhotic patients to require modification of their interferon dose in all treatment groups. In the PEG-IFN alfa-2a groups, this difference was mainly due to dose modifications for laboratory abnormalities, in particular, dose reductions for thrombocytopenia (21% for cirrhotics vs. 6% for the overall population in the monotherapy group and 16% for cirrhotics vs. 4% for the overall population in the combination group). Dose reductions for neutropenia in the two PEG-IFN alfa-2a groups were similar in cirrhotic patients and in the overall population. In contrast, in the IFN alfa-2b plus ribavirin group, the difference between cirrhotic patients and the overall population was mainly the result of dose reductions for clinical adverse events (19% for cirrhotics vs 11% for the overall population).

Peginterferon alfa-2a (180 ìg once weekly SC) in combination with ribavirin (1000/1200 mg po daily in split doses taken with food) is superior to interferon alfa-2b (3x106 three times weekly SC) in combination with ribavirin (1000/12000 mg po daily in split doses taken with food).

The treatment difference is an absolute 6-8% as shown by sustained viral response, biochemical response, and combined viral and biochemical response at 12-24 weeks after the end of 48 weeks of therapy.

HCV genotype 1, high (>2 x106/ml) circulating HCV RNA titers, presence of cirrhosis, older age, African American origin, higher body weight are associated with lower response. The effect of body weight on treatment response may be confounded by other prognostic factors.

Response rates were lower in patients enrolled in US centers compared to non-US centers. Lower response was attributable to higher incidence in US compared to non-US in baseline variables (older age, higher body weight, higher proportion with HCV1 or cirrhosis) that predict poorer response to treatment. Response rates appeared to be lower in patients with higher body weights.

Peginterferon alfa-2a and ribavirin combination therapy is superior to peginterferon alfa-2a monotherapy. The treatment difference is approximately 20% absolute.

Compared to interferon alfa-2b combination therapy, peginterferon alfa-2a combination therapy is associated with slightly higher incidence of overall serious adverse events

(12% vs. 9%) serious infections (4% vs. 2%), grade 4 neutropenia (5% vs. 1%) and grade 3 thrombocytopenia (5% vs. 0.2) and higher requirement of dose modification (32% vs. 18%). There was no difference in premature withdrawals between combination treatment groups (10% vs. 11%). There was a suggestion of an effect of body weight on toxicity.

The 6-8% higher response rate with peginterferon combination therapy compared to interferon combination therapy should be weighed against the risk of higher toxicity (in particular bone marrow) of peginterferon. Given the effect of body weight on treatment response and toxicity, and the lack of dose ranging studies of combination therapy, studies of the safety and efficacy of lower doses of peginterferon may be needed. Given the effect of body weight on treatment response, studies of efficacy and safety of weight-based dosing compared to fixed dosing should be considered.

"A phase III, randomized, multicenter, efficacy and safety study examining the effects of the duration of treatment and the daily dose of ribavirin in patients with chronic hepatitis C virus infection treated with the combination of peginterferon alfa-2a and ribavirin."

The goal was to determine the optimal duration of combination therapy as well as the optimal dose of ribavirin for the combination.

Randomized, multicenter, international, double blinded phase 3 study evaluating the safety and efficacy of four PEG-IFN alfa 2a and ribavirin regimens. Subjects were randomly assigned to one of the four treatment arms. All participants were given the same dose of PEG-IFN-180mg sc q week. In the first two arms, PEG-IFN alfa 2a was administered with either 800mg or 1000-1200mg of ribavirin for 24 weeks. In the remaining two arms PEG-IFN alfa 2a was administered with either 800mg or 1000-1200mg of ribavirin for 48 weeks. A 24- week treatment free follow-up was part of the design for all 4 study arms. Patients in the first two study arms were not informed of the 24 week duration of therapy prior to reaching that point to minimize bias.

Patients were randomized and assigned to each of the study arms on the basis of HCV viral genotype and viral titer. In the original protocol it was stated that within each geographic region, patients with genotype non-1 and low viral load, genotype non-1 and high viral load and genotype 1 and low viral load would be randomized in a ratio of 1:2:1:2 to treatment groups A:B:C:D. Patients with genotype 1 and high viral load would be randomized in a ratio of 1:1:3:3 to treatment groups A:B:C:D This was subsequently changed to 1:1:5:5 because of inability to recruit low viral titer genotype non-1.

Peg-IFN alfa 2a: 180mg in 1 ml solution administered sc once weekly for either 24 weeks or 48 weeks.

Ribavirin: 200mg tablets, between 800 and 1200 mg administered daily in split doses. The ribavirin was administered with an identical appearing placebo tablet as specified by study arm to make up the appropriate daily dosage taken with food

Placebo: Identical appearing to the ribavirin tablet, administered as specified in the study arm in split dose.

Ribavirin would be reduced to 600 mg per day the following occurred: (1) a patient without significant cardiovascular disease experienced a fall in hemoglobin to <10 g/dL and >8.5 g/dL or (2) a patient with stable cardiovascular disease experienced a fall in hemoglobin by >2 g/dL during any 4 weeks of treatment.

Moreover, ribavirin would be discontinued under the following circumstances:

a patient without significant cardiovascular disease experienced a fall in Hgb to

less that 8.5g/dl; a patient with stable cardiovascular disease maintains a Hgb <12g/dl despite 4 weeks on reduced dose. If ribavirin was discontinued, it could be reintroduced at the investigators’ discretion at a daily dose of 600mg

The general dose guidelines and dose adjustments for low absolute neutrophil and platelet counts were the same for this study as the previous one.