| |
FDA Panel Recommends Atazanavir Approval: Report 1 from NATAP
Written by Jules Levin
|
| |
| |
Today May 13 in beautiful downtown Gaithersburg, Maryland at the always well tuned Holiday Inn the Food and Drug Administration convened it's Antiviral Drugs Advisory Committee (AVAC) Meeting to discuss the approval of atazanavir, developed & manufactured by Bristol Myers Squibb. The panel makes a recommendation at the end of the day to the FDA on whether or not the panel recommends approval and the FDA has the freedom to make its own decision, but usually the FDA follows the panel's recommendation. In strong fashion, the FDA panel recommended atazanavir approval 15-0 after a day-long discussion. The discussion included review of the efficacy and safety data results from several key pivotol studies. This report contains information reported by the FDA, Bristol Myers Squibb (BMS) researchers, and the panelists.
I was impressed with the FDA's inclusion of a good cross-section of experts on the panel to address the key issues regarding this new protease inhibitor with several important characteristics, particularly it's once-a-day and has a very favorable lipid profile. The discussion by the panel, chaired by Trip Gulick, MD, from Cornell University's (NYC) HIV Clinic, addressed important issues for atazanavir including: the drug can lead to bilirubin elevations; the effect of atazanavir on PR and QT intervals (these are intervals between heart beats); the favorable lipid changes (cholesterol, triglycerides) associated with use of atazanavir & associated risk for heart disease; atazanavir resistance and it's signature mutation, I50L; the effect of atazanavir on liver function tests; and lipodystrophy.
The panel felt that atazanavir offers patients a significant benefit in terms of lipids compared to other treatment options. In study 034 ATV was compared to efavirenz and after 48 weeks patients receiving ATV had clearly better lipid profiles. Patients on efavirenz had elevations in total cholesterol (20% change from baseline), LDL (bad) cholesterol (15%), and triglycerides (20%); HDL (good cholesterol) also increased (20%) on efavirenz. Patients receiving ATV had no increases in total cholesterol, LDL cholesterol, had increases in HDL (10%), and decreases in triglycerides (10%). Study 043 compared ATV to Kaletra and saw similar findings. Unfortunately, lipodystrophy appears to occur on ATV despite the lack of lipid elevations. BMS called the data collection "passive" but patients receiving ATV had the same incidence of body changes as efavirenz in the 043 study and as nelfinavir in two studies comparing ATV to nelfinavir. The data appears relatively preliminary as I think its 48 week data. And the studies were not well designed to get great answers to this question.
BMS said antacids such as Tums should not be taken with atazanavir as it effects the PH in the stomach & ATV levels. Antacids should be taken 2 hours apart from atazanavir. The old buffered formulation of ddI should also be separated by two hours. But the new EC ddI formulation can be taken with atazanavir.Drugs like Tagamet are being studied to see if they have an interaction with atazanavir but so far this is unclear.
The FDA said that lipid effects of ATV appeared to persist through 108 weeks of ATV treatment, although data from phase 2 studies is limited by study design. Lipid effects do not appear to be associated with a reduced incidence of lipodystrophy. Cardiovascular benefit is unknown at this time. There was discussion by the panel on whether lipid abnormalities in HIV will lead to increased cardiovascular risk. Although we have not studied patients long enough to have clear certainty that lipid elevations in HIV will lead to similar risks as lipid elevations in HIV negative individuals, lipid elevations in HIV present a clear risk and should be prevented and addressed if they occur. Studies show that use of ATV results in less use of lipid lowering agents than used when other HIV drugs are used. The panelists clearly liked this, as it permits less use of additional drugs for patients already taking enough medications.
ATAZANAVIR RESISTANCE. The signature resistance mutation for ATV is I50L, and this appears to be a distinct resistance profile. The FDA reported 31% of 160 patient clinical isolates had greater than 2.5 fold ATV resistance. BMS said the phenotypic cutoff for ATV resistance will fall bwteen 2.5 and 6.0 and they are working with Virologic on this now. In the treatment-naive studies the FDA reported that 15% of 93 patient clinical isolates had greater than 2.5 fold resistance to ATV. BMS reports that 100% of the clinical isolates from patients failing ATV have the I50L mutation. When this mutation is present ATV resistance occurs. Development of this mutation occurred in a range of 2 to 80 weeks for patients, and BMS said it does nor develop easily. The FDA and BMS said ATV-resistant isolates with the I50L mutation have hypersusceptiblity to other protease inhibitors. In other words, they have increased sensitivity to other protease inhibitors. 9 patients from studies in treatment-experienced patients also developed the I50L. Clearly, this is a work in progress, as so far this work involved taking patient samples of blood and looking at this in a lab experiment. So far BMS has not actually taken these patients with the I50L mutation and placed them on new PI therapy to evaluate the response, but I'm sure they will do this. It appears that patients may have to remain on ATV to maintain the I50L mutation simultaneous to starting a new PI. BMS said resistance to ATV (2.4 fold change in ATV susceptibilty) usually occurs with the development of 4 mutations. This could be from any of these but should include at least I primary mutation: primary PI mutations 48V, 82 A/F/S/T; secondary PI mutations 10, 20, 24, 33, 36, 46, 54, 63, 71, 73. Resistance to ATV can develop with only the I50L mutation alone. However, lower level resistance to ATV (1.6 fold change) can occur with 2 (average 1 fold change) or 3 (average 1.6 fold change) key PI mutations. In testing ATV against a number of patient isolates BMS found that 25% of the isolates were resistant to ATV when these isolates were resistant to 3-4 proteases inhibitors. Drug resistance is not black and white; a patient is not either sensitive or resistant, there are gradients of resistant between full sensitivity and full resistance.
The FDA panel overall felt that the bilirubin elevations seen associated with atazanavir (ATV) have not shown a concern and appear benign, but recommended longer term follow-up for safety. Elevated bilirubin can cause jaundice (yellowing of skin). This was reported to occur in 9-10% of study participants. Scleral icterus, which I think is yellowing of the eyes, occurred in 6-11% of study participants. The panel felt that this has not been shown to be a safety issue but is a cosmetic issue. In speaking here privately with bilirubin experts they told me that sitiing under a sunlamp or in the sun can reduce bilirubin and improve jaundice.
BMS said the available data support the conclusion that hyperbilirubinemia associated with ATV is benign, not associated with hepatotoxicity, and readily manageable in individual patients. BMS recognizes the importance and value of instituting a risk management plan by which to build on the safety experience obtained from clinical trials. BMS plans to institute a risk management plan for hyperbilirubinemia that would include the following components:
--Post marketing surveillance of events that may be the consequence of long term
(> 2 years) albeit small increases in unconjugated bilirubin;
--Physician and patient education programs including patient information leaflet;
--Appropriate labeling language for the prescribing physician.
BMS reported that elevation in bilirubin is the most frequent laboratory abnormality observed in Phase II and III clinical studies. The majority of elevations in total bilirubin were isolated (ie, not associated with liver function test elevations) and reversible upon discontinuation or interruption of ATV. Greater than 500 HIV infected patients have received ATV for at least two years (range up to 3.5 years) and no long-term consequences of elevated total bilirubin have been identified. The elevations are primarily composed of indirect (unconjugated as opposed to conjugated which can be a concern) bilirubin.
Among ARV treatment-naive subjects, Grade 1 - 4 total bilirubin elevations were observed in 86% and 91% of subjects in Study AI424034 and the Phase II studies, respectively; Grade 3 - 4 elevations were observed in 35% and 47%, respectively. The incidence of elevated total bilirubin increased with higher ATV doses (Grade 1 - 4: 81% on ATV 200 mg increasing to 96% on ATV 600 mg; Grade 3 - 4: 31% on ATV 200 mg increasing to 66% on ATV 600 mg). Among ARV treatment-experienced subjects in Study AI424043, the incidence of total bilirubin elevations was lower than that observed in the ARV treatment-naive population (Grade 1 - 4, 76%; Grade 3 - 4, 22%). In Study AI424045, the incidence of total bilirubin elevations among subjects receiving
ATV 300/RTV (Grade 1 - 4, 88%; Grade 3 - 4, 45%) was consistent with those observed in ARV treatment-naive subjects treated with ATV 400 mg. The incidence of total bilirubin elevations among subjects receiving ATV 400/SQV was 68% and 17%, respectively. 10 patients across all study trials were reported to have total bilirubin elevations >10 mg/dL.
The FDA agreed with the mechanism identified by BMS of how the bilirubin elevations occur: inhibitionof UGT 1A1. It's primarily elevations in unconjugated bilirubin and based on data collected appears not to be a clinical safety concern. The elevations appear reversible upon discontinuation of ATV. Jaundice/scleral icterus are likely common adverse events in clinical practice resulting in more frequent discontinuations than seen in clinical trials. The risk for hepatotoxicity appears similar to other HIV antiretroviral drugs.
The issue of liver function tests were reviewed. ALT & AST were more elevated for patients on ATV compared to the comparator drugs in some studies but not in others. BMS reported data showing ALT elevations appeared no greater for ATV than for comparator drugs, nelfinavir and efavirenz. BMS showed data that ALT elevations were not associated with elevations in bilirubin. Still, Ken Sherman, the hepatologist on the panel, suggested that BMS and other HIV drug companies do studies to perform histologic (condition of the liver) assessment of hepatitis/HIV coinfected patients to evaluate the effect of HIV drugs and increases in ALT. The panel said regular monitoring of ALT on ATV therapy should be performed.
Another concern is the effect of ATV on heart function. The normal heart beats have intervals between beats. I hope I get this correct as it is a difficult subject to understand & explain. If the intervals are prolonged heart problems can occur. There are two types of intervals, the PR interval and the QT interval. Both can be prolonged by ATV, but it appears that all protease inhibitors can prolong QT intervals. ATV, however, can prlolong PR intervals. Two cardiovascular experts were on the panel and an expert from Massachusettes General Hospital was brought in by BMS. The overall panel opinion was that ATV does not appear to be a concern regarding this issue. And no more than any other drug, including HIV protease inhibitors. A panel heart expert and the BMS expert felt convinced by the data that there is no concern regarding this issue and ATV. A concern was raised by an expert for patients who may be among patient populations with greater risk prior to starting ATV, which includes people with prior heart disease and patients on other medications that can prolong the intervals. Perhaps these patients should receive an EKG an proper assessment before starting ATV therapy. There was general agreement that additional longer-term follow-up for safety should be conducted, and very importantly there was general agreement that HIV treaters and care providers need to be educated about this issue as well as bilirubin.
The FDA said the data from BMS study 076, which explores the concern of potential cardiac effects, is limited by lack of a positive control (eg, moxifloxacin). But the current data indicates that ATV has little or no effect on the QT interval, however, additional studies can assess the overall risk. No signal for increased risk relative to comparators was identified in clinical trials. The incidence of prolonged QT intervals were similar between ATV and efavirenz regimens (2%). In study -043 9 patients experienced a post-baseline prolongation (ATV 2 subjects, Kaletra 7 subjects). ATV can lead to dose dependent prolongation of the PR interval. The incidence appears to be similar to that seen for other protease inhibitors.
ANTIVIRAL EFFICACY of ATAZANAVIR
FDA and BMS researchers agreed the atazanavir and efavirenz showed comparable efficacy in phase III study A1424-034, which had 810 patient participants; each drug was given with AZT/3TC. Using the most recent ITT analysis adopted by FDA, ATV/ZDV+3TC was similar to EFV/ZDV+3TC: 70% of subjects on ATV 400 mg had HIV RNA levels < 400 copies/mL at week 48 compared to 64% on EFV. Results for the proportion of subjects with HIV RNA levels < 50 copies/mL were 32% on ATV vs 37% on EFV. When looking at whether patients had under or above 100,000 copies/ml of viral load the response rates for achieving less than 400 copies/ml were similar between efavirenz and ATV.
In phase II study A1424-007 with 420 patient participants comparing atazanavir to nelfinavir (Viracept), where each was given with d4T/ddI, atazanavir and nelfinavir showed comparable antiviral efficacy. Using the ITT analysis 62% of patients on ATV 400 mg once daily and 61% of patients on NFV had less than 400 copies/ml (undetectable viral load); when looking at achieving less than 50 copies/ml (undetectable viral load by sensitive test) 33% of patients receiving ATV 400 mg once daily and 28% receiving NFV sustained this.
In a study (A1424-043) in 300 patients who failed 1 protease inhibitor and had limited PI resistance atazanavir was compared to Kaletra, and Kaletra had better antiviral efficacy: after 24 weeks of treatment in the study, the mean HIV RNA viral load change from baseline was -1.73 log on the ATV regimen and -2.16 log on the LPV/RTV (Kaletra) regimen. The proportion of subjects with HIV RNA < 400 copies/mL at Week 24 was 61% on the ATV regimen and 81% on the LPV/RTV treatment regimen. The proportion of subjects on the ATV regimen compared to the LPV/RTV regimen with HIV RNA < 50 c/mL was 41% vs 52%.
In another study (A1424-045) 347 patients who failed at least 2 regimens received ATV 300 mg/ritonavir 100 mg once daily, ATV 400 mg/saquinavir 1200 mg once daily or Kaletra twice daily. The majority of randomized subjects had recently taken a NRTI (96%) or NNRTI (60%), whereas only 34% had taken a PI.
Sensitivity to a specific PI ( 2.5 x IC50 of control strain) ranged from 56% to 83%, with 23% of the treated subjects highly resistant (> 10 x IC50 of control strain) to NFV and 21% of the subjects highly resistant to RTV. Seventy-four percent
(74%) and 75% of subjects were susceptible to ATV and LPV, respectively.
Susceptibility to ATV or LPV was comparable across the treatment regimens. An important caveat regarding interruption of baseline PI susceptibility data is that only 34% of randomized subjects were taking a PI at study entry. This would lead to PI susceptibility measurements that are overstated (ie, resistance, therefore, being understated).
Only 16 week data was reviewed by the FDA and 24 week data has been submitted. So the FDA felt it was premature to comment much on this study. BMS reported results after 24 weeks: the proportion of patients with less than 400 copies/ml (ITT analysis) was 64% for ATV 300 mg/RTV 100mg, 44% for ATV 400 mg/saquinavir 1200 mg, and 62% for Kaletra. The proportion of patients with less than 50 copies/ml was 39% for ATV 400 mg/RTV 100 mg, 23% for ATV 400 mg/saquinavir 1200 mg, and 42% for Kaletra. Kaletra and ATV 300mg/RTV appeared equivalent and ATV 400/SQV was inferior. Bilirubin was reported not to be a concern in the RTV boosted ATV regimen. The lipid benefit of ATV was not lost for patients receiving the RTV boosted regimen. Jaundice occurred in 6% of patients receiving ATV 300mg/RTV.
|
|
| |
| |
|
|
|
