Despite many ongoing trials, a vaccine for HIV is still a distant goal, Dr Shahid Jameel of the International Centre for Genetic Engineering and Biotechnology tells Sandhya Srinivasan
Shahid Jameel is Group Leader of the Virology Group at the International Centre for Genetic Engineering and Biotechnology (ICGEB), New Delhi, India. He has served in this position since 1988.
Dr Jameel’s research focuses on viral biology and virus-host interaction. He uses the hepatitis E virus (HEV), human immunodeficiency virus (HIV) and the SARS virus as models to study how viral proteins modulate signalling and cell fate in the host. His interests also lie in understanding human immune responses to HEV and HIV, vaccine design and high throughput platforms for virus detection.
He received his BSc (1977) and MSc (1979) in chemistry from Aligarh Muslim University and the Indian Institute of Technology, Kanpur, respectively and his PhD (1984) in biochemistry from Washington State University. He carried out his postdoctoral work in molecular virology at the University of Colorado Medical School and has been a Visiting Fellow at the National Institute for Infectious Diseases, Tokyo, and Emory Vaccine Center, Atlanta. Dr Jameel has received the Rockefeller Foundation Biotechnology Career Development Award (1990), the Shanti Swarup Bhatnagar Award in Medical Sciences (2000) and the Wellcome Trust International Senior Research Fellowship (2001).
What are the issues in vaccine development, and what are the challenges?
To my mind there are two big challenges -- two big scientific challenges, because a vaccine is not just about the science of developing it but also the process of its manufacture.
The central scientific issue is that the virus mutates rapidly. So, do you make vaccines for specific regions and strains or those that are broadly applicable? This has plagued people since the very beginning.
To give you a sense of how much variation there is in HIV, consider the variation in influenza, for which a new vaccine needs to be developed every year to cover the flu outbreak for that year. The genetic variation in influenza is only 5-10% of what you see in HIV. So HIV genetic variation is huge. And what makes it further complicated, unlike viruses like influenza that cause an outbreak and go away, is that HIV causes a very persistent infection. It remains with the host, continually evolving with the host, so that change is constantly happening. The first scientific challenge is how to address this variation.
So far, since all the money in basic biology as well as in vaccine research is coming from western sources, they are obviously focusing on HIV strains that are dominant in the western world. Now if you look at the major group of HIV viruses, they can be divided into 10-11 sub-types. Sub-type C accounts for about 50% of global HIV infection whereas most of the effort in vaccine development research is in the sub-type B virus which is the dominant virus in the western world: the US and Western Europe. Africa is a melting pot of all kinds of strains. In India, 80% of infection is due to sub-type C. So far, a majority of the vaccine candidates that have advanced in the pipeline are basically sub-type B. How they will perform against sub-type C or other isolates, no one really knows.
Traditional scientific thinking would say that since there is quite a bit of variation between these two groups of viruses, a sub-type B vaccine is most likely not going to work against sub-type C infection. But the point that most people are after is: “At least let us set a proof of concept -- what will work and what will not work -- and once we have discovered what works, it can be engineered for sub-type C or other sub-types.”
The two vaccine candidates being tested in India, at the National AIDS Research Institute, Pune, and the Tuberculosis Research Centre, Chennai, are both against sub-type C viruses.
What is the other scientific challenge to vaccine development?
The other critical issue is: “What are the correlates of immune protection?” We really do not have a very clear idea.
What do you mean by “correlates of immune protection”?
What will protect? Is it antibodies alone, or do you only need to destroy virus-producing cells, or is it a combination of both? We do not have markers to predict whether, if an HIV vaccine candidate works in an animal model (monkeys), it will also work in humans.
When a pathogen infects our system, the host comes up with two kinds of immune responses. The first is based on the production of antibodies. Antibodies will neutralise viruses that are extracellular, that are present outside cells, because antibodies cannot enter cells. Once the HIV virus infects a cell, the virus’ genetic material integrates with the cellular genetic material and the cell becomes a factory for producing more viruses. So in addition to having antibodies to neutralise the virus, one must also have a mechanism to destroy cells that are already infected with HIV.
HIV is characterised by a drop in our CD4 cells, the very cells of the immune system that are critical for developing an effective immune response. CD4 cells are also those cells that are infected by HIV. Traditionally, one would think that once viruses infect cells, these cells die. But in HIV infection, the cells that are infected by the virus are actually preserved. It is the uninfected cells that are killed, by things that the virus produces. In an HIV-infected person only about one in 10,000 blood cells is actually infected by the virus. The numbers of cells that die during a viral infection are far greater. This is something that we call the bystander effect. That is the whole strategy of this virus, that it preserves infected cells and kills uninfected cells.
So how do we identify infected cells? And how do we selectively kill infected cells? That is a critical issue. The host immune system has ways of dealing with this problem, but the virus seems to be one step ahead and it has already evolved mechanisms to evade all those steps.
Many of the vaccine candidates being developed perform very well in animal models. The closest animal model we have is the monkey model. But the monkey immune system does not respond in exactly the same way as our immune system does.
So those are the two core scientific challenges: how to address the genetic variation and continuous mutation in the virus, and what sorts of vaccines should be made to raise the right immune responses.
What are the vaccines being developed?
People are testing two types of vaccine candidates. One type is designed to raise antibody responses. Typically these are protein-based vaccines where you produce a recombinant protein that mimics what the immune system sees in the virus, and raises an antibody response. When a virus infection happens, the immune system sees certain viral surfaces. A vaccine candidate tries to mimic those viral surfaces with recombinant viral proteins.
The other type of vaccine candidate uses a virus vector. This strategy raises specific killer cells in the host that will target virus-infected cells. The adenovirus vectors, MVA, all these viral vectored vaccines raise a response in the host to destroy infected cells. Maybe a successful HIV vaccine has to be a mixture of both, but what sort, how much of each…
People are also trying a prime boost strategy: you prime the immune system by directly injecting naked DNA that can express one or more HIV proteins, and you boost with either one or more recombinant proteins or with a virus vector vaccine to get different responses. Modified Vaccinia Ankara (MVA) is usually used in this prime boost strategy: prime with DNA and boost with MVA. In this case, both naked DNA and MVA carry the same HIV genes.
In India, the Pune vaccine candidate uses an adeno-associated virus (AAV) as a vector. That vaccine candidate, produced by Targeted Genetics, a US-based company, is the one that IAVI tested first in India. IAVI was also testing the AAV vaccine candidate in Germany and Belgium. So far, the results have indicated “a modest immune response”.
The other vaccine candidate they are currently trying at the Tuberculosis Research Centre, Chennai, was originally made by Shekhar Chakravarty at the National Institute of Cholera and Enteric Diseases, Kolkata. Shekhar made the original construct and then collaborated with Therion (a biotechnology company) which had prior experience in MVA technology. There were lots of ups and downs trying to stabilise this construct, and finally the stabilised multi-gene MVA was tried on people in Chennai.
The first large HIV vaccine trial that took place, the Vaxgen trial, essentially used the surface protein of HIV, produced in a recombinant manner. It raised really nice antibodies. The problem was that those antibodies never protected. The vaccine was first tried in animals that had been infected with HIV, and it worked. Their mistake was that during animal testing they used lab-adapted strains of the virus to test the vaccine. These strains were neutralised very nicely, but the field isolates were not neutralised. When the Phase 3 trial happened, there was no difference in effect between placebo and vaccine.
But I still wouldn’t call it a failed trial. From the Vaxgen trial we did learn how to do HIV vaccine trials. We developed the in-vitro assays needed to see whether the antibodies are actually neutralising the virus. We arrived at the numbers required to get statistically significant answers.
What are the steps in HIV vaccine trials?
Vaccine trials in a human population are generally conducted in three phases. The first phase is a safety trial. You make sure the product is safe to administer to humans. Sometimes people also want to see whether the vaccine raises certain kinds of immune responses. So while normally Phase 1 is for safety, it can also include immunogenicity. Phase 1 is done on very small numbers, 30 individuals would be taken for a Phase 1. If a vaccine is proven safe, then it goes into Phase 2 which is really addressing the issue of immunogenicity and dosage -- how much and how frequently. So typically a Phase 2 is divided into different arms, where you test different dosages and combinations to work out what would be the best amount to use and the best schedule.
Once you have done that and come up with an optimal schedule and an optimal dose, you go to the large Phase 3 trial which is an efficacy trial. A Phase 3 trial essentially tells you whether this vaccine will work in the field or not. Now, in animals you can actually inject the virus and see whether the vaccine works. In humans you can’t do that. So in humans you have to depend on natural infection. Typically, Phase 3 trials are done in populations where the natural rate of infection is high. Phase 3 trials for an HIV vaccine will be done on sex workers, in men having sex with men, I don’t think they have done any trials in haemophiliacs. There are plans to do them on injecting drug users. You have a group that gets the vaccine and a control group that gets a placebo. And at the end of the trial, which typically lasts anywhere from three to five years depending on the numbers you want -- typically you are looking at large numbers, huge expenses, many ethical issues -- you come up with an answer to the question: did the vaccine population get fewer infections than the control population?
Vaxgen went all the way to Phase 3; they failed in Phase 3.
There is also this new concept of Phase 2b, between 2 and 3, which does limited efficacy testing with fewer numbers. After Phase 2 gives you information on dose and scheduling, you increase numbers and test efficacy in a smaller population to give you a quick answer.
You may have read recently that Merck has stopped trials of its HIV vaccine. The Merck vaccine was farthest in the pipeline right now in the human population. Then in late-August 2007 it announced that it had stopped its Phase 2b trial because the vaccine was not working. The Merck vaccine was based on an adenovirus, the common cold virus, modified to carry HIV genes. The idea was that if you infected people with this modified virus it would also raise an immune response against HIV. That vaccine had done very well in animal trials and it looked safe in human trials, so that was far into the pipeline. Merck had recruited 3,000 volunteers, and midway, when they looked at the two populations, they found 24 infections out of 1,500 in the vaccine arm, and there were 21 out of 1,500 in the placebo arm, so there was no statistical difference. Their vaccine was not protecting. So they did the right thing and stopped the trial then and there.
Right now the situation is that your best vaccine candidate has failed.
How long ago did research start in vaccines?
HIV vaccine research started immediately after HIV was discovered as the virus that causes AIDS, around 1984. They developed the blood test very quickly. Remember that the mid-1980s was when a very successful hepatitis B vaccine was made and licensed. They used a very simple trick: they expressed a recombinant protein that makes up the surface of hepatitis B, produced this protein in yeast, purified it, put it into people, and they were protected. SmithKline Beecham and Merck did it simultaneously. That was the mid-1980s, when HIV was discovered. It also had a surface protein on it and they said, well that is simple.
If you go back and read some of the interviews of the time, the US secretary of health announced that a new virus had been identified for what they called the ‘gay disease’ at the time. She said, now that the virus has been identified we are going to get a vaccine in the next two years. We are talking 1984. That euphoria was based on the hepatitis B vaccine where a simple trick had worked and people thought an HIV vaccine too would be a piece of cake. That was not to be.
In light of that if you want to know when we will have a vaccine, I don’t think anyone will hazard a guess. Your best vaccine has just failed and there are many other candidates in the pipeline that work on principles that are similar to the failed Merck vaccine. So where does that place us?
Does the recent death in the US of a woman who was part of a trial using the same AAV used in the Pune trial have any relevance for us?
It is not right to make a blanket statement that since one person died following the use of an AAV vector in an arthritis trial, all trials with that vector should stop. It is all a matter of the dose used, the immune status of the individual, and so on.
Does that information tell you something?
Absolutely. One always has to evaluate as one goes along. If you knew everything it wouldn’t be science.
Neither Therion nor Targeted Genetics works in the area of HIV vaccines. What would these companies gain from such collaborations?
Well, if I had a company and I had a platform that could be used to produce a useful HIV vaccine candidate, I would try it because it is big business if it works.
Is there any scope for using the technology for other vaccines?
Of course, it would be a great ad for this platform. HIV is a tough virus, (a successful) vaccine technology could be used for other vaccines as well.
Where is the vaccine effort today?
For at least 20 years we have been hearing that an AIDS vaccine will be developed in the next 10 years. The point is that some of the best candidates have failed. Even IAVI has gone back in its thinking. Earlier, it focused on vaccines to raise responses to kill infected cells. It is now shifting research to vaccine candidates that raise neutralising antibody responses.
Antibodies will only kill viruses that are outside cells. HIV spends most of its life inside the cell. So you also have to destroy the infected cell. The adenovirus vectors, MVA, all these vaccines raise a response in the host to kill infected cells. Soluble proteins tend to raise antibody responses. Maybe it has to be a mixture of both, but what sort, how much of each…
Where are we today? Even with the best vaccine we are now up to maybe a Phase 2 trial. There are maybe 20 vaccines in Phase 1 and 2 right now, using various vectors. The candidate farthest down the pipeline failed at Phase 2b. So I guess one important thing is to learn from why the Merck vaccine failed. They are still analysing the data. It will be a year or two before they get some kind of answer. People have various theories but nothing is proven. It’s only a month-and-a-half. One theory is that the adenovirus backbone used to make the Merck vaccine is based on adenovirus serotype five, which is very common in humans, 70-80% of humans have already been exposed to it. So when you give a vaccine based on this backbone, the host will raise a quick response to neutralise the virus. So the search is on for an adenovirus with a very low seroprevalence so it won’t be recognised. But Merck had divided the study population into two: those with low titres against serotype five, and those with high titres. The vaccine failed in the low titre group...
There are other products already in the pipeline and we have no option but to test them. All of these have worked on animals. Even the Merck vaccine worked fantastically in animals but failed in humans. Working in animals is not (a sufficient) measure. We don’t have a surrogate marker for vaccine efficacy. The nature of the beast is such that you don’t know if it will work until you test it. So to say that because the Merck vaccine failed we should stop testing would be wrong. You have to test. It is a gamble, a shot in the dark.
There are also the manufacturing challenges. Let’s say tomorrow we have a situation where a DNA prime vaccine with an MVA boost strategy works. I saw an estimate that if, in 2007, it is shown that DNA plus MVA works, then your total global requirement of DNA in 2017 (10 years after the vaccine is shown to work) would be 600 kg. To put things in perspective, as of today, the entire world has produced 1 kg of DNA of that quality, safely injectable in humans. To reach the capacity you need in 2017 you need to plan 15 years ahead of time, or you won’t have plants, regulatory requirements to make that product. In terms of that planning we have already missed the bus.
And obviously we can’t plan unless we know what works and what doesn’t. Even if we get something that works, it will take another 15 years to manufacture it and bring it to the market.
Is it really possible at all? Why are we looking for a preventive vaccine rather than a therapeutic vaccine?
A lot of people are realising that research in preventive vaccines can be used to look at therapeutic vaccine potential. Yes, absolutely, therapeutic vaccines should be tried. They are easier than preventive vaccine trials. You are going to try it on people who are already infected with HIV and you don’t have to wait as long to see whether they work.
But there are problems even in therapeutic trials. Ethically you cannot give just the vaccine; you also have to give antiretroviral therapy. So you have to compare ART vs ART plus the vaccine. And ART in its initial phases shows a dramatic reduction in virus titres, an improvement in quality of life, etc. So even a therapeutic vaccine trial must go on for a fairly long time to address the question of whether the vaccine is actually providing an added benefit.
Why not just concentrate on drugs?
Drugs have improved the quality of life for people living with AIDS. But they will always be expensive. Second, the virus will always mutate, and drugs will never clear the virus, only suppress it. The nature of the virus is such that it cannot be eliminated with drugs. So, drugs are a tool, but to control HIV, to prevent its spread in the population, we have to introduce a preventive or therapeutic vaccine. A therapeutic vaccine would ideally reduce the concentration of virus and virus-infected cells in the body fluids. Transmission is based on numbers: the higher your titres the more frequently you transmit. Nevirapine reduces the chances of perinatal transmission because it reduces the viral titres in the infected mother.
What are the collaborations you see today?
There are 20 vaccine candidates, most for sub-type B, some for sub-type C. Some are in Phase 2. There are no longer any in Phase 3. They are being developed by various academic groups and companies. Merck was the single large company into HIV vaccine development. I don’t know if Merck will continue in it or not. Large pharma is generally not interested in the HIV vaccine, in vaccines in general.
You also need people who can bring other people together. Academics are not necessarily on the same wavelength as industry. IAVI’s contribution is to tie up research, manufacturing, clinical trials, governments.
We (ICGEB) don’t work on an HIV vaccine. We study the basic biology of HIV. Two years ago, DBT funded us to set up a core lab to develop assays so that if they started HIV vaccine research they would have a place for quality control.
Is HIV vaccine research a priority for India?
I don’t see a single serious group in India trying to develop an HIV vaccine. The vaccine candidate tested in Pune was not developed here at all. With the MVA vaccine, the original constructs were developed here by Shekhar but the larger scientific community was never involved. So as of today if you ask me whether there is a serious effort to develop an HIV vaccine in the country I would say no.
Are you asking me is it a priority, or should it be a priority? Well, should it be a priority, I would say yes, why not, if you have the funds. Because you may never develop a vaccine but you will learn so much as you go on. I may be biased; I also work on HIV.
I have often heard the argument that more children die of malnutrition and diarrhoea than HIV. My answer is I’m not saying don’t reform your public distribution system. Reform it. Children die of diarrhoea because we don’t get them oral rehydration solution in time. But HIV will not go away with something like ORS.
Having said that, I don’t see any serious effort in India towards an HIV vaccine. You are trying something but money is not going towards basic research. I don’t fault government agencies. The fault lies with the biomedical scientific community not coming together to think about novel ways of doing things.
We have to look at the systems too. The department of biotechnology put about Rs 5 crore into the vaccine effort at the All India Institute of Medical Sciences over the last 10 years. Then the person in charge of the programme, Pradeep Seth, retired, and the whole effort died. We knew he was going to retire. Why didn’t the funding agency ensure that succession was in place? A body of work had been put together. We have systems problems that we need to address as well
InfoChange News & Features, January 2008