India recently completed its first clinical trial for an HIV candidate, a Phase 1 safety trial at the National AIDS Research Institute, Pune. A second Phase 1 trial is ongoing at the Tuberculosis Research Centre (TRC), Chennai. Sanjay Mehendale, Deputy Director of the National AIDS Research Institute (NARI), discusses the logistical and ethical issues involved in running a trial for an HIV vaccine with Sandhya Srinivasan

Sanjay Mehendale is Deputy Director and Head of the division of epidemiology and biostatistics at the National AIDS Research Institute, Pune. He was involved in conducting India’s first Phase 1 trial of an HIV vaccine candidate.

What are the challenges in conducting an HIV vaccine trial?

To do any clinical trial we need a strong foundation of four pillars: clinic facilities, a laboratory infrastructure with world-class laboratories, data management, and community linkages to identify populations and recruit participants. This is necessary for every study site.

Clinic facilities are where you interact with participants for information exchange, counselling sessions, physical exams, where the vaccine candidate can be injected. There are two kinds of laboratory facilities for trials. One is small and near the clinic, to collect samples for processing in order to establish the eligibility of participants -- pathology, haematology -- to confirm that they are healthy and able to participate. The other kind of laboratory facility is more sophisticated. For example, if we want to find out if a vaccine candidate produces an immune response we will have to do cytotoxic T lymphocyte assays which are markers of cell-mediated immunity. This needs sophisticated equipment that has to pass certification in, for example, proficiency in testing sera which are known positive, known negative, and so on.

Then you require data management. All world-class trials have sophisticated data management facilities to record data of ongoing trials, generate interim reports on study outcomes and, if necessary, track all past records for 10-15 years after completion of the trial depending on the requirements of good clinical practice guidelines.

Finally, you need the right population for the right study. For this, you have to look at what you want to achieve.

In Phase 1 trials we need healthy volunteers and you will not get them from the hospitals so you have to go to the community. For Phase 2 trials you can have both high- and low-risk populations.

In Phase 3 trials the route of administration is fixed, the dose is fixed, and you want to see if the vaccine works. For this you need people at risk of infection. So you need to go to high-risk populations. This is where we may not be successful in getting people in the community. It might be a good idea to work in clinics for STD treatment or other health problems. The presumption is that if they are high-risk they are more likely to come to clinics for treatment. Or you go to voluntary counselling and testing centres because people who come here perceive themselves to be high-risk. Or you might work with non-governmental organisations dealing with people with high-risk behaviour -- injecting drug users, commercial sex workers, men who have sex with men -- to look for people who are not clinically sick but at high risk.

The challenge is in developing multiple and large sites in different parts of the country where we can do such studies, to accrue the samples that we need.

What is the sample size?

For Phase 1 trials we need 30-50 participants but the number can go up to 100. Phase 2 trials generally have between 100 and 500. Phase 3 trials are large and the sample size is in the thousands -- 3,000, 5,000, 14,000. An ongoing HIV vaccine trial in Thailand has enrolled something like 16,000 people.

The sample size is determined by what you want to achieve. In a Phase 3 trial you want to find out the ability of the vaccine to prevent infection. So if the incidence -- the rate of new infections -- is low, you may need a huge sample.

Understand the difference between prevalence and incidence. Prevalence is the percentage of existing infections at any given time. Incidence is the rate of new infections over a specified period. When prevalence in a population is low -- for example in India today, we are looking at a prevalence in the general population of much lower than 1% -- then the incidence in the general population will be very, very low. Then, doing a Phase 3 in the general population will be practically very difficult because you will need a huge sample size.

I will give you two scenarios in India. If you have to choose between female sex workers and pregnant women attending a hospital, which of the two groups will give you more HIV-infected women by screening the smallest number of women? Naturally female sex workers, because they have high-risk behaviour and more are likely to be positive. If you screen 100 you will be able to pick up 30-40.

I want to clarify, the reason you need a huge (or particular) sample size in a Phase 3 vaccine trial is that you need to have a certain number of people exposed to risk and a minimum number of people who get infected in order to establish whether that vaccine is actually working or not. One group will end up with a higher number of positives, the other group will have a lower number of positives.

So you have to have that number for comparison. Essentially that’s why studies are done in high-risk populations. You can capture a greater number of people and observe the difference in the number of people who get infected by HIV, thereby indicating the efficacy of the candidate vaccine.

As part of vaccine trial protocol, investigators like us are expected to do extensive counselling on trial participants to ensure that risk behaviour is minimised during their participation in the vaccine trial.

The question often asked of us is how do you measure efficacy (a vaccine’s usefulness in preventing actual infections) if you are putting so much emphasis on counselling, etc, as the participants will then reduce their risk behaviour and there will be no HIV infections. How will you know that the vaccine has worked? This is where the issue of a comparison group, the placebo control in a classical vaccine trial, comes in.

What is expected in such a case is that the impact of additional interventions like counselling and condom promotion will be there in both groups. What will be different is the administration of the vaccine in the vaccine group. That’s why the difference between the incidence among the vaccinated and the comparison group will be able to statistically indicate to what extent the difference is attributable to the vaccine. I hope I have made my point clear.

You have made your point clear but I still have a question. There will be an obligation to counsel in order to minimise risk. Despite the fact that in both these groups the risk will be minimised, you still have to arrive at a comparison and an understanding of how effective a vaccine is. Doesn’t the researcher have a conflict of interest? There is an interest in actually getting results, and there is also an obligation for the researcher to get that person to behave in such a way that there will be no results.

Of course as a researcher I will be interested in results. That does not mean that I will always be interested in positive results. I do not go with a preconceived notion as a researcher. If I am a principal investigator in a particular Phase 1 trial I do not go with a presumption that it is a safe vaccine or that it is able to produce an immune response. If I am a principal investigator of a Phase 3 trial I don’t go with the presumption that it is an efficacious vaccine. For that matter, for a conscientious researcher it is always better to accept the fact that this might be a problematic case. I have a greater obligation to my community, to my participants. Anything bad that happens to them is not acceptable to me. If anything happens to them I have to be able to categorically comment on whether it is vaccine-related or not. If it is vaccine-related, I have to be able to boldly say that it is not acceptable because safety concerns are there, or efficacy concerns are there.

At the same time that does not mean that all researchers are unethical.

I am not suggesting that researchers are behaving unethically. I want to put forward the proposition that there is a conflict if risk behaviour actually gives the best results.

I wouldn’t say this is a conflict. As a scientist and researcher I know that there are scientific ways in which you can analyse the data. You have to ensure that all proper practices are followed, the appropriate information is conveyed to the participants, so that risk gets minimised through counselling or any other intervention that is indicated in the research protocol. Even if this happens, the study is powered in such a way that in spite of this effect you eventually get analysable results. The study should be designed appropriately and have an adequate sample size, then no conflict exists.

We were talking about developing study sites…

One issue in selecting sites is an adequate incidence of infection. One of the important research agendas of this institute is vaginal microbicides and we had previously tested two products in Phase 1 trials -- pro 2000 and buffer gel. We opted out of a Phase 3 trial of a vaginal microbicide because we did not observe the desired incidence the study protocol required. Currently an international protocol, HPTN035, is looking at the efficacy of the same two products. It required that all sites included had an incidence -- rate of new infections -- of 3% per year. This is a large number. You start with a population of 100 negatives and you should be able to have at least three positives at the end of one year without any interventions being done, and so on and so forth. In the studies that we conducted we focused mostly on female partners of male STD clinic patients, and women coming to gynaecology clinics with symptoms of STDs. We were not able to document that high an incidence. So we decided to back out of this study.

The point I’m making is that getting sites with a high incidence of HIV is difficult. Sex worker sites can be one choice, but although sex workers will be the primary users they are not necessarily the only users of vaginal microbicides. We also talk of married monogamous women at risk of HIV infection from their husbands and they would need to use microbicides as well. Doing studies in high-risk populations like female sex workers has its own limitations.

There are also problems in doing studies among sex workers. Because of their risk many of them are already infected. Second, they are labile populations, they keep moving around. For any clinical trial that is another difficulty. Essentially for a Phase 3 trial, we need a population that we can keep track of with regular follow-up for three to five years. Female sex workers are also vulnerable as they are under someone’s control. Finally, other concurrent infections make them vulnerable to getting infected and that makes evaluation of endpoints difficult. So doing studies on FSWs is difficult.

So how to find populations where the incidence of HIV is high but the people are not sex workers? That is a real challenge as far as India is concerned, I would say.

Are you looking at such populations?

In a Phase 3 trial of an HIV vaccine -- or a microbicide -- we have to look at high-risk populations including sex workers.

These are the issues related to study sites.

We also need researchers trained in this work. They should be educated in the capacities required to undertake this work, in the ethical issues, in the scientific methods, in good clinical practices, in that specific research protocol. In addition, there is other job-related training. For example, laboratory staff must be trained in that assay, data management staff must be taught the specific software, quality control, etc.

What is the difference between developing study sites and identifying groups for research?

Study sites are where you conduct the research, where you need to have the four components of a clinical trial: clinic facilities, a laboratory infrastructure with world-class laboratories, data management, and community linkages to identify populations and recruit participants. NARI is a study site for the Phase 1 HIV vaccine trial.

Identifying the population needs different considerations. In the NARI Phase 1 HIV vaccine trial we needed 14 visits in 12 months, so we didn’t expand the recruitment outside Pune. You have to decide what group will be considered for recruitment, where and how they will be approached: in clinics, through NGOs, or in the general community.

Are you talking about identifying groups or recruiting participants?

I’m talking about both, there is an overlap. Identifying groups is the first step, where you get information on prevalence and incidence. That is relevant for Phase 3, but for Phase 1 we don’t need this because we talk about healthy populations. Then the question is: where to go to find it? To social workers? Religious people? Women’s groups? Healthcare professionals? Scientists? We have to think of all the possibilities for approaching a large number of people and giving them systematic information to eventually complete adequate recruitment.

There is no direct contact with individuals in recruitment?

Not necessarily. It is a combination of both. For a learned population you give them information on the website and they will approach us, but not for others. The crux is how to reach the right kind of people for your study and then how to recruit them.

The whole success of recruitment is in getting the right groups. For example, we went to a number of government organisations anticipating that as we were a government organisation, it would be easy to organise programmes and give information. But we didn’t get a good response. Then we worked with anganwadi workers and got a good response. Then we went to individuals who had a primary level of motivation to participate…

I can’t imagine what that motivation could be.

You may not imagine yourself in that position, but don’t you agree that all the medicines we use have arrived after completion of clinical trials? Who participated in them? People like you and me, or else it would not have been possible.

We are new to clinical research, particularly Phase 1.

Most of the clinical trials in India are Phase 3, drugs that have been proved safe outside. We have few drug development trials. We did one of the first few Phase 1 trials in the country. It was challenging because it had the HIV tag. Perhaps if it had been a malaria or diabetes trial, people would have participated more easily.

We have not yet talked to our people systematically about clinical trials. The new contract research organisations, the way clinical trials are growing in India, our country is a growing site for clinical trials… There is a growing awareness among people, now people ask, why should we participate?

One reason for people not to participate is that there is no direct benefit, and that is as it should be. Those people I know who have participated have done so for the money. Clinical trials should not provide direct benefits, but then why should people participate? Going back to HIV vaccine trials, those who participate in Phase 1 trials will have to see some benefit. For example, someone they see who could have benefited, that is one way…

The single reason we identified, from among those whom we screened -- which is around 80 to finally enrol 30 in the actual trial -- was altruism: to do good for society. There was a variety of reasons, not necessarily that they had someone close who had HIV. They cared to participate for a social cause. There were a few who had someone in the family who was affected by HIV, who felt that this was an indirect way of helping them.

What is considered a benefit by one may not be seen as a benefit by others. For example, one of the comments made by an American when we talked about reimbursement of Rs 275 for women who participated in a vaginal microbicide trial was that this was an inducement. I tend to feel that he/she hadn’t lived in India for a long time. He/she said this was too much for a villager and this is how US-funded research gets a bad name. He/she did not know that before we arrived at this amount we talked to women’s organisations, to past trial participants and other advisory agencies to see what is and isn’t an inducement.

In our vaccine trial, we decided on Rs 500 for travel and loss of wages since there were 10 visits in one year and any particular visit entailed a commitment of three-four hours a day. This amount was not insignificant but it was not enough to induce a person into the study.

Another way we tackled this issue is that we employed a three-level programme. First, in community-based programmes we talked to groups of 20-200 and gave them a brochure or motivating flyer with contact information. With anyone who came back to us we had a second-level meeting where we talked to groups of five to seven and gave them more information. Then we did a risk assessment to see their motivation, whether they had misconceptions such as that the vaccine would give them protection, etc. This was for the Phase 1 trial. The third level was one-to-one contact with a further assessment of risk and motivation. Then we asked them to consult their significant others and come back to us. When they came back then we talked about reimbursement. This was to ensure that having passed through three stages they were interested, and now it was time for them to know also about reimbursement.

You have talked about various kinds of problems in getting participants: the need to identify high prevalence or rather high incidence. Then, how to recruit.

The number is always a challenge. For Phase 3 trials we need thousands and thousands of participants. This means finding institutes that are interested and have the required competency -- trained researchers, clinic, laboratory, data management, community linkages for informing and recruiting adequate numbers of participants... And we need to find and prepare sufficient numbers of such sites with uniformity of procedures.

Do you need to try the vaccine on different populations besides female sex workers?

Of course. There are two or three ways in which HIV is transmitted: heterosexual, MSM, IDU, and mother to child. Which of these populations are ideal for preventive vaccines? You must do the study on all these groups because we need to know if the intervention will work in all of these settings.

The areas that you have described -- competence, training, screening, counselling, identifying populations -- are not really challenges but how to do things right. What are the challenges? Where do you think the problems will be?

I think these are big challenges because this is India. We are recruiting in a country where few have participated in clinical trials. We have to explain risk, etc. This is a challenge, I feel. There are so many technical terms. What approaches should we use when talking to the masses vs small groups, individuals? Should the mass media be involved? How do we ensure that informed consent is obtained? What procedure should we follow? Should it be textual, or audio, or both?

It is established that India has the technical competence. The challenge is to ensure that people understand what it means to participate.

What about the challenges of dealing with vulnerable populations?

I won’t say the issue of vulnerability is different in India. For example, women all over the world are vulnerable but the vulnerability in Indian women is higher. Indian women have less access to healthcare, lower social status. They are given low priority in family decisions compared to women in developed countries.

Should we think about a timeline for an HIV vaccine?

As a scientist I would say we should have a specific timeline. Considering the pace at which this research is going on in different parts of the world there is a hope and there is not. You must have heard of the closure of the Phase 2 Merck trial. That is not proceeding to Phase 3.

The only trial in Phase 3 is in Thailand and the results will be out in 2008. The final follow-up is going on. It’s a population-based Phase 3 envelop-based vaccine candidate. (Until it has cleared Phase 3 we describe it as a vaccine candidate. If it passes through a Phase 3 trial we will call it a vaccine.)

If the vaccine proves effective the issue is whether it will be applicable elsewhere where the sub-type is different. That’s something that scientists have to work on. Even if it works we still don’t know if it will work in other countries; we will still have to work on that.

What about the NARI vaccine trial?

The Phase 1 trial of the adeno associated virus (AAV) vaccine candidate tested at NARI had good safety but the immunogenecity was not as desirable.

Is it the same as the vaccine candidate tried in Belgium and Germany?

Yes.

Are the results out?

Yes. But it is not likely to proceed to Phase 2. But the MVA vaccine candidate in a Phase 1 trial at the Tuberculosis Research Centre (TRC) at Chennai -- the results are not officially out but it seems to be a promising vaccine and might go for Phase 2 trials.

You say it is unlikely that the AAV vaccine tried in NARI will proceed to Phase 2, but there are trials of this in Africa.

Phase 2 trials are ongoing in Africa but a higher dose than we used is being used there… But still, the preliminary results are not encouraging, so it might not go further.

If TRC looks promising, what is the timeline?

I would say in the best case scenario we are talking about six years. The worst case is 10 years.

What is the timeline for a vaccine candidate?

Look at the hepatitis B vaccine. It took 16 years. The polio and measles vaccines took more than 40 years. But two-and-a-half years after SARS first appeared, we had a vaccine.

What if you have a vaccine candidate, presuming that it gives good results at each stage, what is the possible time from Phase 1 to Phase 3?

Eight to 10 years. Phase 1 will take two to two-and-a-half years. Phase 2 will take three to four years. Phase 3 will take five to six years. The minimum is 12 years if you’re lucky…

When you talk about site preparation, what is the status?

As far as sites go, at NARI and TRC we have shown the ability to conduct AIDS vaccine trials by creating appropriate infrastructure and training personnel.

One more aspect, identifying high-risk groups … have you got that ready?

We started that in 1993 with an Indo-US collaboration, a study called PAVE (Preparation for AIDS Vaccine Evaluation). At the time the hope was that the vaccine would be ready in a year or two and the same cohort that was being prepared by us would be used in a vaccine study. But that was not to happen. So we have been doing long-term studies in high-risk populations that we could recruit.

Where are these populations?

In Kolkata, Pune, Chennai.

In Mumbai?

I don’t think so.

In Sangli?

Sangli is a place with a known high-risk population but cohort studies are not being done there. A cohort study will be necessary to identify the incidence of HIV in that population and to maintain a population for a future trial. PATH along with the ICMR is working on site preparation of Phase 3 studies on vaginal microbicides. They could be used for vaccines as well. I don’t know where…

Overall, there are sites and cohorts being followed. Preparatory exploratory studies are being undertaken to identify populations where trials can be undertaken in other parts of the country.

Tomorrow if you had to start a Phase 2 trial, for example if TRC’s results come out and are good, are you ready?

Yes, at TRC, NARI, at Kolkata…

Can you talk about a therapeutic vaccine?

As a public health man, I am interested in prevention. A therapeutic vaccine would be a kind of immunomodulator that will possibly change the course of the disease, improve the quality of life, increase CD4, decrease viral loads, reduce opportunistic infections, lower death rates. Drugs are doing the same things. I don’t know how this is different. Moreover, we don’t know how frequently the vaccine will have to be given, the costs, dangerous reactions and side-effects, immune phenomena… moreover we will need to test them… the major difference is that they will be in HIV-positive individuals.

Can you talk about what it means if we have a vaccine that is not 100%
successful?

Hardly any vaccine is 100% successful. I would be happy even with a 30-50% efficacy, because to start with such a vaccine with extensive coverage could significantly avert new infections. You would have to talk to people with a good math modelling background for a more elaborate understanding.

I want a rough picture...

A modelling has been done that if a vaginal microbicide came into the market with 30% efficacy and coverage was extensive -- that is, if it reached 70% of the women needing it -- we could bring HIV incidence down by 50%. That is a sizeable impact even with a low level of efficacy. Efficacy and coverage are two components of impact.

If there is a high incidence of a disease, a relatively ineffective vaccine will still be of value. But with a low incidence of disease, you need to have a high efficacy vaccine. As we now know that the prevalence of HIV is very low in India, the incidence is also expected to be low. That has a bearing on whether you will use the vaccine in India.

I think this is a good point. What you are trying to say is: to see a good impact on prevalence and incidence in India, where the rates are pretty low, we would have to have a vaccine with much greater efficacy. Unlike in Africa where a vaccine with lower efficacy would have a better effect.

This is an important consideration given that the prevalence and incidence were earlier believed to be higher. Now we really do have to have a vaccine of much greater efficacy.

Yes, possibly.

What do you think are the ethical challenges in all this?

We touched upon some of them. The major challenge is to ensure that people participate with full knowledge, to give information, to ensure that they understand, the success of the informed consent process, the spirit in which it is done.

Any trial is likely to fail if it does not ensure retention. Retention becomes an ethical challenge. If, in a trial, people have been enrolled but retention rates are poor, you have lost a lot of your participants, the results will not be good. What about those who participated? This is because the trial was started without ensuring that people would be retained. If science is defeated, ethics will be defeated.

What about care and treatment to vaccine trial participants?

As far as IAVI is concerned, they have a policy to give ARVs to all who become positive in the trial. Other institutes may not have this policy. However, as an institute we cannot differentiate between IAVI-sponsored trial participants and others. Suppose we have a Phase 3 trial, and we are not able to give the benefit of the vaccine to the population after the trial completion, it will be an ethical issue. So how do I link up with other programmes?

What is the current status?

We have been given an ARV centre by NACO so we can cover all our trial participants.

Post-trial access to the vaccine, if it is proved effective, is important. One issue is the ability to have an affordable vaccine. Before the trial is started the parties signing the agreement discuss this. The IAVI-ICMR-NARI memorandum of understanding addressed this. Two-three companies in India were inspected and an agreement was signed to transfer technology to them to manufacture the vaccine.

There have been discussions on the issue of negotiations between the different collaborating organisations in such research, where the local organisations may not be able to have equal negotiating power. Does this come up?

This is an issue of training. My job is to get the best deal for my participants. There are instances where the local site-level investigators do not have the skills or feel empowered, but we are changing, we are getting more experienced. That again is to some extent an issue of familiarity. Earlier we used to just accept clinical trials that came our way. We have more training in these areas, including good clinical practices. Also there is an awakening in the general population.

Do you think the ethics review at different levels is able to identify issues and to resolve them?

I think there is confidence in both levels. It is an evolving process. Maybe a raw person does not have the insight of an experienced person, but as you participate you get better or improve. We encourage our ethics committee members to participate in international programmes for training. If five years ago we did not have the confidence about a sound ethical review, we certainly have it now.

InfoChange News & Features, January 2008