There are a large number of experimental drugs now being developed to treat hepatitis C virus (HCV) infection. A race is on to find new HCV drugs because millions of people may one day require treatment, and the current generation of drugs are difficult to tolerate and do not reliably cure this potentially deadly liver infection (unlike HIV, HCV infection can sometimes be cured). The outcome for untreated HCV can be decompensated cirrhosis requiring transplantation-or, if no donor liver is available, death. The outlook for people infected with both HIV and HCV is worse, and in the US, blacks typically have poorer treatment outcomes than whites.
Because different patient groups have such varied outcomes from treatment, researchers are grappling with novel challenges over the best way to test these new drugs and bring them to market.
The US Food and Drug Administration (FDA) convened a two-day meeting of its antiviral advisory committee in October 2006 to discuss drug development issues for new HCV agents. The panel was composed of clinicians, researchers, industry representatives, and community advocates.
The following is an adaptation of the panel's deliberations on the question of how much data should be known about certain challenging populations-people with advanced liver disease, people with HIV, and African Americans-at the time the drug is approved for sale in the US. The original, full transcripts may be viewed at www.fda.gov/ohrms/dockets/ac/cder06.html#AntiviralDrugs.
Dr. Sherman: What patient populations should have been studied at the time of initial approval of a new HCV agent?
Dr. Seef: The question is: who should be treated in order to get approval as quickly as possible so we can get this drug into the market and, if possible, move on to secondary studies? My initial impression is that the groups that really warrant treatment are the non-responders, true non-responders. These are the people at highest imminent risk of ending up with serious disease.
I think we have to include African Americans in this. Almost a third of the people in this country who are infected with hepatitis C are African Americans and we know from studies that African Americans do not respond as well to treatment. Therefore to talk about an overall 40-50 percent response rate does not reflect reality because response rates are somewhere between 30 and 80 percent, depending on race, depending on genotype.
However, I would not involve people with decompensated liver disease in this first series of studies. I think their treatment is too complicated, and I think we need to first know whether these drugs are going to be effective in compensated patients.
Dr. Sherman: But would you start with compensated patients with cirrhosis?
Dr. Seef: Yes, I would include patients with cirrhosis. While they respond less frequently, they are appropriate to be treated.
Ms. Swan: I would reframe the question as: how much do we need to know about a drug's safety before it goes into a person with decompensated cirrhosis?
Dr. Vierling: I would like to see selected studies in decompensated patients who are listed for transplantation in specific regions of the country where, were they to have deterioration due to the natural process of their disease or unforeseen severe adverse events, they would have the rescue potential of transplantation. I think that we have a way to protect the patient, to do the study, and to obtain the evidence of potential benefit in those who are decompensated. There is no way, short of studying them, to know whether we are advancing a therapy that could be of benefit.
Dr. Haubrich: My bias is toward what is going to get the drug approved in the most efficient manner. With that in mind, for each of these categories of patient I would like to see safety data or at least pharmacokinetic data to some extent. But if involving a particular patient category could actually hinder the drug's development by introducing toxicity complications that delay studies in, say, a naïve population, then I would probably set the studies up but not necessarily require that they be done at the time of approval.
Dr. Chung: I think it has been an industry concern that an adverse event will arise and put the kibosh on a drug development program in treatment-naïve patients. I think that has created a concern with treating certain high-risk treatment populations. I think it is important for the FDA to perhaps allow a little bit of a leeway.
Dr. Birnkrant: We are in agreement with Dr. Chung. That is, if we did find a problem in a more advanced population, obviously it would raise concerns for us, but then we could take what we learned from that population and perhaps increase monitoring in a naïve study.
Dr. Alter: I am concerned that our assumption is that, we get the drugs to the market as soon as possible so that there is the greatest access for the most patients, but in fact they may not be appropriate for these groups. Yes, they are going to be licensed therapies, and therefore physicians can use them as they choose and with whoever they choose, but maybe they won't be useful in that group of patients. I honestly don't know what the forecast is and how generalizable these treatment regimens are going to be between these different patients. Certainly current therapies aren't very generalizable.
On the other hand, if you had a group of patients with decompensated cirrhosis who were going to die because they couldn't get a liver, would you offer them an experimental therapy that could be potentially dangerous? How did we do the first transplants? How did we do a lot of things that are actually life-saving? It is either that or death. So, maybe they are the group that should be right up front. How many people die every year waiting for a liver?
Dr. Sherman: Thousands.
Dr. Alter: With hepatitis C?
Dr. Sherman: Yes.
Dr. Alter: So, that is the group I am talking about, and therefore there may be an ethical obligation to initiate a study up front on those individuals.
Dr. Birnkrant: We do have means of making investigational therapies available to patients who desperately need them. So, if that were the situation, clearly they would be made available as long as the company agreed to provide it. If we received multiple requests for that type of population, at that point we would ask the company to develop some sort of protocol to actually actively collect the data.
Dr. Sherman: What data on people coinfected with HIV should be required at the time of initial approval of a new HCV drug?
Dr. Fish: We know that HIV is a factor for progression, so we would want to have early treatment data for this group. There is the concern that Ms. Swan raised earlier about drug-drug interactions and cytochrome P450 interactions, so we would have to be careful and thoughtful about the patients that we would have enter those trials. For those on antiretroviral therapy we would need pharmacokinetic data to look at drug-drug interactions and make sure that the nonnucleosides and the protease inhibitors for HIV maintain adequate blood levels, and that the hepatitis C therapy maintains adequate blood levels. Since those requirements would probably delay a trial, I would not see them as necessary for approval. So, we would like to have data on HIV but I don't know if it is realistic to expect that all of that information be available at the time of the approval process.
Dr. Sherman: Tracy Swan?
Ms. Swan: I would like to say first that drug-drug interaction studies with antiretrovirals and also other drugs commonly used by people who are living with HIV need to be done very early in the drug development process so that the lack of data can't be used as a rationale for not using the drugs in co-infected people who are taking antiretroviral agents and other drugs. I can't stress the importance of interaction studies enough. There was a life-threatening interaction between an antiretroviral drug, didanosine and the HCV drug, ribavirin. I don't know off the top of my head how many deaths resulted but they were all unnecessary, and if better studies had been done to characterize that interaction those lives would have been saved. If we can bring these treatments into a population with such urgent need we are going to save more lives. So, I would say at the barest minimum what I would find an acceptable amount of data would be the interaction studies and at least 12-week efficacy data in co-infected people.
Dr. Chung: I would amplify on both what Dr. Fish and Ms. Swan said and say that now is the time to start those PK studies and cytochrome P450 studies so that the groundwork can be laid to do parallel trials in both mono-infection and coinfection. One plausible scenario could be an initiation of a naïve trial in HCV/HIV co-infection at the same time you are doing a naïve trial in HCV mono-infection. That would be a treatment group that had a reasonable likelihood of success, of superior responsiveness to the add-on therapy to the standard of care, and could allow licensing and immediate implementation within the HIV co-infected population and likely extension into more difficult-to-treat populations within the HIV co-infected group. So, I would argue for parallel trials in both mono-infection and co-infection. But that requires, as Tracy suggested, early up-front work on the part of pharma to do the interaction studies.
Dr. Haubrich: I will take an intermediate stance. The expectation of having PK studies completed with 22 approved antiretroviral drugs is probably not realistic. So, that work has to be targeted. And exactly what data is needed to have a full parallel registrational trial in HIV? I think that is also unrealistic and would probably slow down the process. So, I would be satisfied with pilot data over 12 weeks.
Ms. Swan: From my understanding, there is a large group of co-infected people who have both advanced HIV and advanced hepatitis C. That is where I would see the greatest clinical need and the greatest urgency to move these therapies forward, although I also think stratification by HAART or no HAART or other parameters is a very good idea. The other thing is that some of the new drugs in development might be good candidates for pharmacokinetic boosting with ritonavir, which is given with a lot of other HIV protease inhibitors. So, it sort of begs the question: if you have a three-times-a-day regimen (which are notorious for poor adherence, risk of resistance, etc.), and it could be improved by boosting with a commonly used HIV drug, wouldn't you want to examine that scenario and make sure we are getting the data we need?
Dr. Sherman: Remember, one of the features of the population of HCV/HIV co-infected is a tendency towards very high viral loads, which is probably one of the factors that affects efficacy but, again, may be an issue in terms of resistance. So, I would argue that some understanding of resistance emerging in the setting of very high viral load be evaluated before a drug is released and used in that population.
I think the feeling here is that prior to initial approval efforts should be made to initiate early stage studies at least in co-infected patients; that those studies should include analysis of major drug interactions and pharmacokinetics.
This is very similar to what we agreed upon earlier for decompensated cirrhotic patients, that there should be studies initiated and under way. They don't have to be pivotal trials taken to completion, but we shouldn't wait until Phase IV.
Dr. Sherman: Should groups, such as African Americans where response rates are lower, be included in the main efficacy studies or would there any reason for separate trials? Should their inclusion be required?
Dr. Alter: In my opinion we at least have to include the two major racial ethnic groups in the U.S., if not three. It should be a requirement that there be a sufficient sample size to address efficacy in the three major racial ethnic groups in the United States.
I just want to make sure this isn't one of these situations where we say, "Yeah, we can have some PK data in these groups when we go for approval" but that there would actually be sufficient data for approval in these populations.
Ms. Swan: We really need to get population-specific PK data during Phase II to see if there is any signal of difference before we move into Phase III with diversely populated studies. Also, it is not a question of whether you can enroll diverse populations, it is a question of how. There are studies that have done it. Many of the sponsors of these products have done studies in HIV that have enrolled people of color without a problem, so it can definitely be done.
Dr. Seef: I cannot believe that we are even thinking about this. This has to be a reflex I believe in doing this. We must have whites and blacks in the study, absolutely.
Dr. Chung: When you are planning the studies, especially in African Americans, given what we know about their high frequency of null responses or at least non-responses from VIRAHEP-C and other studies, we ought to plan to look carefully at biologic endpoints including resistance.
Dr. Sherman: I think the issue of the barriers to enrollment in clinical trials needs to be raised. In the major pivotal trials ongoing today, African American have exceedingly low enrollment relative to their risk and prevalence in the population. I think this committee should encourage the FDA to look at barriers that appear in trial designs that then lead to enrollment of primarily upper middle class white populations that are not representative of the disease as a whole in this country.
Dr. Alter: These trials take a lot of work to begin with, granted. And it takes a lot more work to get difficult-to-reach populations. But that doesn't mean it can't be done. It just takes more work. I think that there are a lot of people experienced in getting to hard-to-reach populations and there are a lot of ways to do it.
Dr. Sherman: I would point out two salient points in ACTG 5071, which was a co-infection study. There was 33 percent African American enrollment, and the overall dropout rate in the study was 13 percent, which was no different than what was seen in the pivotal trials in HCV mono-infection.
Dr. Seef: Also, compliance was not an issue. There was no less compliance among the African Americans, at least in the VIRAHEP-C trial, than there was among the whites. So, I don't think that is an issue either.