An Interview with Frances Gotch
HIV/AIDS cohorts in Uganda
Frances Gotch, D.Phil has been working with cohorts of people affected by HIV/AIDS in the UK and in Entebbe, Uganda for more than 10 years. She was instrumental in setting up the immunology laboratories at the Uganda Virus Research Institute in Entebbe with funding from the Wellcome Trust, and in close collaboration with the MRC-funded Programme on AIDS in Uganda. Epidemiological studies have defined progression rates, risk factors, and virological, statistical and immunological studies within the cohorts have lead to a greater understanding of the pathology of HIV infection in developing countries. The AIDS Support Organisation (TASO) has provided a backdrop for all the work, providing care, support and counseling for people living with AIDS in Uganda.
Gotch started her working life as an immunologist with Professor John Humphrey—a founding figure in the immunological world—at the National Institute for Medical Research, Mill Hill, UK. She then spent 20 years in the ivory towers of Oxford University—most latterly in a hugely productive partnership with Andrew McMichael and Alain Townsend. The possibility of working at the Chelsea and Westminster Hospital, which houses the largest clinic for HIV-infected persons in Europe, brought her to London in 1996. She is now Professor of Immunology at Imperial College London, based at the Chelsea and Westminster Hospital, and a fellow of the Royal College of Pathologists. Her research interests center around cellular immunology, particularly the mechanisms of T cell activity in disease associated with HIV-1, HIV-2, SIV, influenza virus, cytomegalovirus, Epstein-Barr virus, human herpes virus 8 and malaria. Her current studies focus on identifying HIV-specific immunological functions in vaccinated, infected, and high-risk non-infected individuals, trying to tease out the immunological parameters that might confer benefit in HIV-associated disease. Gotch has been intimately involved in efforts to standardize assays across the various AIDS vaccine trial sponsor organizations and is the Principal Investigator at IAVI’s state of the art, GCLP-accredited core laboratory facilities at Imperial College London. She recently spoke with IAVI Report Editor Simon Noble about her Ugandan studies and efforts to define immunological parameters relevant to AIDS vaccine development.
Could you just talk a little bit about the kind of individuals who make up the Ugandan cohorts you've been working with?
I've been working in Uganda for 10 years now, primarily with two cohorts. One is what we call the Progression cohort, the other is the Exposed Seronegative (ESN) cohort. The Progression cohort is a group of HIV-infected individuals living in and around Entebbe who attend the local AIDS Support Organization (TASO), and we've been following these individuals every six months for about eight years now. The principle investigators in Uganda are Pontiano Kaleebu, Heiner Grosskurth, Tony Kebba and Jennifer Serwanga.
We originally tried to define people as rapid or slow progressors, simply on their CD4+ T cell count and the length of time that they had been infected, or at least length of time since diagnosis—we very rarely know the time of seroconversion.
Now we've been able to get the statisticians involved and they’ve carefully assigned these people as rapid or slow progressors, and everything in between. Of course there's only a small proportion of people who by these very stringent statistical criteria are true long term non-progressors (LTNPs), only about 3 or 4% of the total. We also now have our viral load assays in Uganda, and also the patients are staged according to the World Health Organization clinical staging of HIV disease; Stages One, Two, Three, Four.
Is this designed to get a clearer picture of the natural history of HIV infection in a developing country setting?
Yes, but personally I’m also interested in this group of patients in terms of potential immunotherapy, and the kind of responses that are made by the LTNPs in this group, compared to those individuals who are progressing extremely rapidly, so that we know what we're trying to reconstitute when we treat people with immunotherapy or with drugs or whatever.
Many people think that the kinds of responses in LTNPs are those that should be induced by a prophylactic vaccine, but it should be remembered that although these individuals are keeping the virus under control quite well, eventually most of them will get sick. So we're not certain that these responses, if they were there in the first place, would actually protect from infection.
We now have a very clearly defined group of LTNPs and we’re able to look at both CD4+ and CD8+ T cell responses, using conventional methods—ELISPOT assays and intracellular cytokine staining analysis. We've looked at responses to all HIV proteins—not just Gag and Pol but also Nef, Vif, Vpr and Vpu. Everybody has been HLA typed and we can sequence virus and see if escape has taken place. We've identified new epitopes because the HLA types are slightly unusual in some cases, and we've looked at other kinds of genetic polymorphisms, like the CCR5 polymorphisms. So this is a very well characterized progression cohort with a group of true LTNPs that we follow over time.
So what correlates of protection have you been able to identify in these LTNPs?
Probably the biggest hint is in the CD4+ T cell responses, which seem to be broader and more robust in the LTNPs. We see quite good CD8+ T cell responses even in some of the rapid progressors, so there’s not much difference there. Truthfully, I can't really say that we've had any very groundbreaking findings in Africa, similar to what we've seen in the UK.
One thing that always fascinates me is the fact that in Uganda the average time from seroconversion to AIDS is nine or ten years, which is not vastly different from the time in the UK or the US before HAART was introduced. If you consider all the other bombardments that people in Africa get—malaria, diarrhea, lack of clean water, typhoid, etc.—you would think that having a disease that makes you immunosuppressed would be very detrimental in such a setting. But it's almost as if there's something compensating.
So we're very interested in concurrent infections—we've looked at helminth infection, and a little bit at malaria.
Could this be related to the ‘hygiene hypothesis’?
Well, it could be, who knows? We, along with Allison Elliott and Mike Brown, thought that intestinal helminth infection, which is known to induce Th2 type responses, would be rather detrimental if you were also HIV infected—we were proposing to treat all our vaccinees before vaccination.
But when you treat the helminth infection, if you're HIV infected your viral load actually goes up and your CD4 cell count drops. So it seems to be a balance between Th1 and Th2 that is so important.
One of the most interesting studies we're doing is looking at links between hepatitis G virus infection and HIV progression. There is no doubt that infection, or at least the presence of hepatitis G antigen, has been shown, to be associated with non-progression, so it looks as if there might be some kind of immune response to hepatitis G virus that is somehow keeping the HIV better under control. I think some very interesting data will come from these studies of intercurrent infections in our cohorts.
So do you see hepatitis G virus as a potential immune therapy?
My colleague in Uganda, Edward Wright, and I have discussed it. The data would suggest yes, but we actually don't know what this virus does, it hasn't ever been shown to have any pathogenic effect. One could imagine perhaps administering heat-killed virus. It depends on what data he gets in the next couple of years, whether we actually can see any kind of immune responses to hepatitis G virus that perhaps are cross-reactive with HIV, or maybe it could be some kind of innate immune response, who knows? The goal now is to identify a mechanism of this non-progression.
So what about the ESN cohort you work with in Uganda?
In my opinion, they must surely hold the key to the prophylactic vaccine, if you believe that such individuals exist. We really do think there are people who are in some way protected from infection. In Uganda we've chosen to go for a rather different type of cohort—discordant couples. These are heterosexual couples who've been living together in stable relationships for long periods of time, with documented frequent unprotected sex, with one partner HIV infected and the other not. We've recruited quite a large cohort, about 40 couples, and we're hoping to expand the cohort to 200 couples, because we think some of the trouble with research from ESNs, I think it's true to say, is that there are a lot of people who don't believe the data, so you have to have large numbers of these people because if you're looking at quite small responses you've got to see them in an awful lot of people.
This is largely the work of Tony Kebba, a Ugandan clinician, and he has a very stringent questionnaire to make sure, in his estimation, that these people genuinely are being exposed to virus—lack of condom use and so on. We have some evidence that they are exposed to virus because in about four years in the entire cohort we've had three seroconversions. We've also had several pregnancies, which also indicate that people are having unprotected sex.
A very important asset of this cohort where we know both partners is that we can look at the HIV-infected partner as well to characterize the quality and the quantity of the virus that the seronegative partner is exposed to. With a commercial sex worker, you really do not know that.
This is very much ongoing work but we've looked at lots of different things— humoral and cellular responses in the peripheral blood, as well as on mucosal surfaces, and in urine and semen. We also want to look at innate immune responses, natural killer cells in particular. We're beginning to get some kind of handle on what might be protecting these individuals. We do see HIV-specific IgA on mucosal surfaces in these ESNs, and we’re trying to map the epitopes on Env that these IgA’s actually bind to.
We also see CD4+ and CD8+ T cell responses, always of narrower breadth and quantitatively lower than in the seropositives. These responses would be defined as positive according to the most stringent criteria used in vaccine trials but they're not hugely impressive data, I would say. With the CD8+ T cell responses we tend to see responses to slightly different parts of HIV—we rarely see responses to Gag in the ESNs, which is contrary to what’s seen in the seropositives, and we see responses in the ESNs to some of the non-structural regulatory genes, which perhaps are seen less frequently in the seropositives.
What I must say is that although we have never been able to show any virus by sensitive PCR in these ESNs, that’s only sensitive down to a certain level and we can’t say that there is absolutely no virus, for instance in the lymph nodes, because biopsies wouldn't be acceptable.
Have you followed any of these ESNs over time and looked at the evolution of the immune response and compared that to the virus in their infected partner?
Yes. The three patients that we're following in greatest depth at the moment, the evolution of their virus, are the three seroconverters because if these people genuinely were ESNs, and they truly did have some response before they seroconverted, maybe their course of disease is going to be different because perhaps it will be as if they were vaccinated. We're also interested in transmission of escape mutants, and HLA sharing between the partners. Was Frank Plummer right, if people are very disparate in their HLA types, are they less likely to transmit than otherwise?
ARVs are now being rolled out in Uganda, which might mean that the non-infected partner will be exposed to very much less virus. If viral resistance then emerges in the HIV-infected partner, as it very often does, their viral load will increase again, and then the ESN may be more susceptible. Perhaps it will be similar to Sarah Rowland-Jones' sex worker cohort in Nairobi who stopped working for a while, then when they went back to sex work, some became infected—it was as if they had to keep on being exposed to the virus, keep on being vaccinated as it were. So we’re going to be looking out for that.
Do you think there's any mileage in mimicking this continuous exposure idea using a microbicide containing antigen, as a kind of topical vaccine?
It seems clear that providing continual exposure to a vaccine antigen may be useful, and that immune responses on mucosal surfaces may be very important. Sometimes I find it a little bit hard to think about true vaccine protection on mucosal surfaces. Whether or not this could be provided by a microbicide containing antigen remains to be seen, but a microbicide with an ARV and a monoclonal neutralizing antibody could, I think, have a huge and profound effect upon the epidemic. I'm very, very keen on microbicides in principal. They empower women to protect themselves, which is sometimes very difficult, and studies in Uganda suggest that they will be acceptable. They will have to be relatively easily applied, but it makes absolute sense.
To switch to Uganda more generally, it's been hailed as the success story of Africa time and time again.
I do think Uganda has been a success story, although I think we mustn't get overwhelmed, there's still a higher than acceptable prevalence of HIV in some parts of Uganda. I think above all it’s been the openness in Uganda that has helped to combat the epidemic to an extent. It’s not a huge country and it has had a relatively stable government ever since Moseveni has been in power. They have a fairly open press, including television, and HIV comes up in almost all of the soap operas and consequently people are not scared. Prostitution is illegal in Uganda, as it is in many places in the world, but they don't throw anybody in prison, they try treating STIs.
Almost everybody in Uganda, I would think, has been touched by HIV in some way. There's a Member of Parliament in Uganda who has stood up and said "I am HIV positive", and education has been taken extremely seriously, public health messages on hoardings. People in developed countries sometimes don't realize what profound effects such public health measures can have.
There's always been the admission in Uganda that HIV causes AIDS. There's also been the admission that hunger and poverty may have exacerbated the epidemic, and I think nobody would argue with that, but they do not say that hunger and poverty causes AIDS. They have a serious wish to understand what HIV is all about.
I don't think the prevalence is dropping, not vastly, it seems to be have plateaud. Now ARVs are being introduced, so we might even see an apparent increased prevalence. Imagine 10 years ago, if you were a Ugandan, perhaps you might have put yourself at risk a couple of times, you might think to yourself, “I wonder?” What really would have been the point of going to be tested? Because there were no drugs. But now there is a point because now the drugs are beginning to be available. I can imagine now, if you were that Ugandan, you might think, “Well, there's a clinic just up the road, I've got a couple of friends who are HIV infected, they're getting these drugs, I'll go and be tested, and if I’m infected then maybe I'll be eligible for these drugs.” So I think we might see an increase, but it won't be a real increase.
Recently there's been good progress on standardizing assays, but the field now seems to be taking a step back and looking at it again as a real research endeavor, using more sophisticated, multi-parameter flow cytometry to identify a marker that better reflects an effective immune response. What do you think are currently the most promising markers that will better inform us about the functionality of any induced response?
Flow cytometry is extremely powerful technology and I think we need combinations of techniques to get complementary information. Really, what we're interested in are effector functions, the way these cells might actually work, not just the markers they have—it's now accepted that mere detection of virus-specific T cells is not enough. We could measure CSFE staining for cell division and simultaneously characterize the phenotype and proliferative ability of T cells—it won’t necessarily tell you their function, but it's this simultaneous analysis of different characteristics that flow cytometry allows.
I think the Holy Grail would be the ability to measure two functions, to measure recognition of a virally-infected cell and then destruction of that cell. Or with humoral immune responses, neutralization of the virus. We've got to have assays that really do reflect functionality. You can measure cytokines, such as TNFalpha or interferon gamma, but you can almost as easily in the same kind of assay measure secretion of cytotoxic granules, the perforins and the granzymes, for example. With perforins in a CD8+ T cell you are actually looking at its cytolytic potential, but you might also want to look at the suppression of HIV replication by those cells, which is, ultimately, what you want to measure.
But it's very, very important that, whatever we decide, we've got to have standard operating procedures and work according to good laboratory practice, and we've got to then use the same reagents and equipment. The culture of working to good laboratory practice with standard operating procedures is spreading.
What's your current picture of the immune response against HIV? What do you think are the relative contributions of neutralizing antibody and cellular immunity, and within that, CD8 versus CD4?
I think that high titer neutralizing antibodies are of extreme importance. Just because I describe myself as a cellular immunologist doesn't mean I don’t recognize the importance of antibodies. I think that the AIDS vaccine field, and the HIV field generally, has become too compartmentalized—you're either an antibody person, or you're a T cell person, or sometimes you're just a CTL person—and we're not thinking in a broad enough fashion. The immune response is a concerted, beautifully designed series of interdependent events, and you shouldn’t study one part of this response divorced from another because the interplay between them is obvious to anybody who learns about immunology.
However, having said that, I think central to all this is the CD4+ T cell, which, by its very name, the helper T cell, is central to the successful immune response to HIV. That the loss of function of HIV-specific CD4+ T cells happens from such an early stage of infection sets the scene for what is to come. That’s what we've got to induce, to de-anergize or reconstitute in the HIV-infected individual.