HIV strategies in the West--and AIDS everywhere else

An interview with Kevin M. De Cock, MD, FRCP, DTM&H

When to Start Therapy - and What to Start With
Start with the Best, or Wait until Later?
Efficacy, Effectivness, and Adherence to Therapy
Best-case Scenario in Developing Countries
A "Package of Prophylaxis," and Perinatal Transmission
HIV and TB
An obscure Disease in a Faraway Place

"When the demons come at night," said Kevin De Cock at an HIV meeting last year, "you ask yourself what you're really doing on this planet." Referring specifically to five years spent heading the CDC's HIV project in Abidjan, C&/content/art12135.html/content/art12135.html#244;te d'Ivoire, he wondered whether the blizzard of publications, the constant presentations, are making a difference.

One might also wonder when, except at night, the demons could even ponder catching up with Kevin De Cock. With the stamina of a long-distance runner (an avocation), he has left trails through three continents in two decades, since taking his doctoral degree at the University of Bristol in 1984. Although he calls himself "an internist with an interest in infectious diseases," on any given day he might pass as a hepatologist, an epidemiologist, or an expert in tropical medicine, tuberculosis, HIV-1, or HIV-2. A citizen of the United States and Belgium who is now a senior lecturer at the London School of Hygiene and Tropical Medicine, he speaks English, French, and Dutch, has won research grants in US dollars, Swiss francs, and pounds sterling, and cites Alexander Pope in interviews.

De Cock first worked in Africa as a lecturer in medicine at the University of Nairobi from 1979 to 1982. He returned in 1986, as an epidemic intelligence service officer of the CDC, to track down five persons from the Yambuka region of Zaire whose mid-1970 blood samples had been positive for HIV. Three had already died of a disease whose symptoms matched those of AIDS. In 1988 he became director of Projet RETRO-CI in Abidjan, a cooperative effort between Côte d'Ivoire and the CDC to study HIV infection.

Today, besides teaching at the undergraduate and postgraduate levels and continuing his research on HIV and tuberculosis, De Cock works as consultant physician at the Middlesex Hospital's HIV unit and as an advisor to the World Health Organization. The Journal arranged to speak with him on November 3 at the Third International Congress on Drug Therapy for HIV Infection, thinking he might offer a different perspective on a few issues we have visited recently.

Journal: In the past year, physicians and people with HIV infection have been given a lot of new information about antiretroviral therapy, and a lot of opinions. How would you suggest that they sort through these data and opinions to make personal treatment decisions?

De Cock: It clearly is very difficult. For people with HIV infection, I think much depends on finding the right physician. Increasingly, I think the care of people with HIV disease is becoming more specialized. In the lecture we just listened to, Dr. [Robert] Schooley used the analogy of oncology versus cardiology to compare clinical endpoint trials with those using surrogate endpoints. What he didn't say is that both of those fields are extremely specialized areas in which the nonspecialist would not dabble. I think the modern management of HIV disease in the industrialized world is becoming a similarly specialized area, one in which physicians who don't do this for a substantial part of their clinical time probably shouldn't be involved. So people with HIV infection need a doctor in whom they're confident and who keeps up with the constant changes.

Physicians themselves must keep up by reading the literature, talking to colleagues, going to conferences. At the same time, one has to be very careful not simply to follow the latest fashion. I believe it was Alexander Pope who cautioned against being the first to try something new or the last to abandon something old.* And that's common sense in clinical medicine. It doesn't apply only to HIV.

Journal: Would you say, then, that HIV-positive people in developed countries, especially if they live in rural areas, should make a point of finding an experienced HIV physician, for example, at the closest university center?

De Cock: I think so, yes. The best analogy would be a malignant disease. I'm trained as an internist and I have an interest in infectious diseases. I would not be able to give sensible opinions about the treatment of even the commonest malignancies. I would never dream of trying. It's important for someone with my background to be able to diagnose these conditions and to discuss their implications. But today you'd never launch yourself and your patient into a complex cancer treatment unless you are specifically trained to do so. I think the same should be true with HIV.

In words, as fashions, the same rule will hold;
Alike fantasic, if too new, or old:
Be not the first by whom the new are tried,
Nor yet the last to lay the old aside

--Alexander Pope, "An Essay on Criticism"

Return to Contents


When to Start Therapy - and What to Start With


Journal: One issue that even experienced HIV clinicians haven't resolved is when to begin antiretroviral therapy. To answer that question, do we need a long-term clinical endpoint trial of immediate therapy versus deferred therapy? Or can clinicians size up results from trials such as ACTG 175 and weigh surrogate marker data and so begin to figure out when to recommend treatment to individual patients?

De Cock: I think that actually remains the fundamental question of HIV therapy. Three common questions are when to start, when to switch, and when to stop. When to stop is a clinical decision that involves talking with your patient and exercising the art of medicine. This is true not only for patients with HIV, but for all patients with end-stage disease. There are times when patient and physician should be able to say, "Listen, we have come to the end of the road." Then symptom control and palliative care are really what's important. HIV is special in some ways, but in other ways it's just a disease, and the same clinical sense that is so fundamental to practicing good medicine applies to HIV just as it does to looking out for patients with cancer or heart disease. When to stop is a question that is no different today with better treatment for HIV than it was earlier.

I think when to switch is probably easier today in certain ways. We have such a better handle on whether a patient is responding or not because of the powerful tools at our disposal, particularly plasma HIV-1 RNA measurements. Using these tools certainly gives us laboratory values that indicate whether a patient's disease is escaping control, so we no longer wait for somebody to dwindle slowly downwards with the symptoms of uncontrolled HIV disease. Still, it's difficult to know what to switch to. Again, I appreciated what Dr. Schooley said in his lecture about always thinking ahead, making sure that you don't lose opportunities for future regimens by engendering cross resistance to drugs you haven't used yet.

When to start is, I think, as contentious an issue as it ever was. I personally think the guidelines published in July1 at the time of the International AIDS Conference are a quite sensible attempt to pull together reasonable conclusions from the data.** But we do not know that people who start antiretroviral treatment above a CD4+ count of 500 and with a high viral load are necessarily going to do better in the long term than those who delay treatment. It seems a plausible thing to do, but that does remain one of the big questions.

**The guidelines, proposed by the International AIDS Society-USA, suggested beginning antiretroviral therapy for

(1) all persons with symptomatic HIV disease,
(2) asymptomatic individuals with a CD4+ count of 500 cells/mm3 or less (some panel members favored 350 cells/mm3 or less),
(3) asymptomatic individuals with more than 500 cells/mm3 and a plasma viral load above 30,000 to 50,000 HIV RNA copies/mL or with a rapidly declining CD4+ count.

The panel also said therapy should be "considered" for anyone with a viral load above 5,000 to 10,000 copies/mL.

Return to Contents

How do you go about answering that question? I think that Dr. Schooley attempted to show the type of trials that we should be aiming for--without resorting to clinical endpoint trials--but I'm not sure I left that lecture entirely clear about answering this question about the best time to start.

Journal:So you're saying it's not certain how we would even produce the evidence of whether early therapy, at CD4+ counts higher than 500, is going to work better than later therapy?

De Cock: It's certainly difficult. We have very good data showing that high levels of HIV-1 RNA predict disease and death. We have good data showing that reducing plasma HIV-1 RNA results in short-term improvement. Now that still doesn't answer the question of whether we should be starting as early as possible.

Still, I think more and more we're coming to that conclusion. Certainly there's no infectious disease that we begin treating only when the infection's really got going. That doesn't make intuitive sense for any infectious disease, and it doesn't make sense in terms of our understanding of the pathogenesis of HIV disease. From short-term observations, we're beginning to get the data that now actually fit the intuitive understanding that we should be getting at the virus early. But it would be awfully nice to have the long-term clinical data that show it. Sometimes, though, the data you want doesn't come from drug trials. Sometimes you get data from observations, and I think that's OK, as long as you put it in the context of everything else we know. It's going to be very important to get good observational data on those people who, for one reason or another, have elected to go on early treatment. I personally would be persuaded by those kinds of data. So I'm not sure that we're ever going to have a trial. But we'd better make sure that we get the observational data on all of these different approaches.

Journal: In your summary lecture at the end of the Vancouver conference, you mentioned the ongoing Multicenter AIDS Cohort Study (MACS), saying you thought that the people who started that observational study should be commended for their foresight. Can MACS or other cohort studies provide the answers we need about starting therapy?

De Cock: Perhaps epidemiologic surveillance, as conducted by CDC for example, can address some of these issues. I think that merits discussion. There are an awfully lot of patients in the United States on antiretroviral therapy, and there is a need to follow those people on a long-term observational basis. I'm sure that some of the data are at least being collected. Even when zidovudine monotherapy was a standard of care, lots of researchers published papers based on a retrospective cohort approach that attempted to show whether zidovudine had influenced survival. I think it's going to be important to do the same with the current combinations, but hopefully with an attempt to define when treatment should be started.

Return to Contents

Start with the Best, or Wait until Later?

Journal: When one does decide to begin therapy, there is another important question to answer: Is it better to begin with the most potent combination of drugs possible, or can one profitably start with a more modest regimen, holding some of the more powerful agents until later? What do you think?

De Cock: I certainly don't think there are any data on this point, so anything anybody says is pure conjecture. Everybody needs to remember that, even if whoever is saying it has a big name.

This is speculation purely based on what we know about other diseases like tuberculosis, but I think you clearly need to do everything you can to avoid the risk of drug resistance. So you need to be thinking not only about individual agents, but about combinations. That's the first point. Secondly, probably the worst thing to do is partially suppress viral replication. So, whatever the regimen you start with, the goal clearly must be to get replication as low as possible and keep it as low as possible. If you can do that, say, with a combination of drugs that does not include the most potent agents, which in a way you want to "protect," then it makes sense to start with combinations that do not include protease inhibitors. I guess my bias--and it is bias--would be to start with combinations that do not include protease inhibitors, but to start with two and possibly three other drugs and take it from there, reserving more potent combinations for later. Clearly, monitoring viral load is critical to that approach.

There is an interesting analogy from the treatment of tuberculosis. How valid these analogies are, I don't know--but when you practice medicine you try to practice it on a rational basis in a framework of science-based knowledge. The analogy is "protecting" rifampicin [rifampin] in tuberculosis treatment. If you get resistance to rifampicin--or, even worse, resistance to rifampicin and isoniazid--then you're in trouble. So, across the world, protecting rifampicin is a very important concept. I think one could carry that over to planning therapy for HIV infection.

Journal: Putting together two of the things you said--striving to lower viral replication as low as you possibly can, and not using protease inhibitors right away--today you would probably need a combination of two nucleosides and a non-nucleoside reverse transcriptase inhibitor. I believe that's the only combination of drugs not including a protease inhibitor that's been shown to suppress viral loads below detectable limits for months in a randomized trial.

De Cock: I think that's correct. In Vancouver, Julio Montaner reported that a combination of nevirapine, zidovudine, and didanosine managed to suppress viral load below detectable limits in a substantial proportion of patients for a considerable period of time.*** That's impressive. And many experts in Vancouver commented about achieving sustained suppression over 40 weeks in many patients with a combination that did not include a protease inhibitor. I think that was actually quite encouraging.

***Later, in Birmingham (abstract OP7.1), Montaner's group reported that nevirapine plus ZDV plus ddI drove levels of circulating virus below a 20-copy limit of detection in half of 51 individuals taking the combination for 40 to 52 weeks.

Journal: After seeing the early results of treatment of primary infection presented in Vancouver, clinicians in the United States who care for a lot of gay men who are not HIV positive are asking whether they should treat primary infection if they see it. Do you think they're jumping the gun, or do you think that's an option they owe their patients?

De Cock: You could actually even extend that question by asking whether they should be offering preventive therapy to people who think they've been sexually exposed. Let me answer your question first: It's very difficult. The rationale for offering treatment is that we know high viral replication in primary disease correlates with symptomatic disease and that both high initial viral load and symptomatic primary infection correlate with faster progression to AIDS. So it makes intuitive sense to try to suppress that replication, to try to alter the subsequent course. Now, we don't know whether the disease behaves like that; we don't know whether our intuitive understanding will prove true clinically. It's different from when we treat people with established HIV infection; there are data to show that if you lower viral load during chronic infection, there's a better subsequent clinical outcome.

Still, I would guess that in today's climate most clinicians and most informed patients would probably want to start treatment during primary infection. But there are two difficulties: How long do you treat for? And once you treat for a certain period, what do you do then? People are going to be forced to make those decisions with no data, which is always uncomfortable.

As far as preventive therapy is concerned, again I think people are probably going to want it and clinicians are probably going to give it. Why? Because we know that the standard of practice when a healthcare worker gets stuck with a needle is to institute treatment. In fact, the standard of care is now combination therapy in that situation. A sexual exposure presumably should be handled in the same way. As far as I know there are no guidelines on this, but, yes, I think it's going to happen.

Return to Contents

Efficacy, Effectivness, and Adherence to Therapy

Journal: In your summary in Vancouver, you remarked on the studies showing that good adherence to therapy results in a better outcome, and you suggested the need for more work on the adherence question in collaboration with behavioral scientists. What can behavioral scientists bring to an understanding of that problem, and are there other clinical problems on which they could profitably collaborate with clinical investigators?

De Cock: There are several answers to those questions. Firstly, you must realize that we need to distinguish between efficacy and effectiveness. Efficacy is what you measure in a clinical trial. A clinical trial is the ideal situation; a clinical trial is not the real world. So you show in your phase III trial that, yes, penicillin does cure pneumococcal pneumonia. Then you go into the community and you give everyone who has pneumococcal pneumonia penicillin. But that's a different world, because some of these people with pneumococcal pneumonia are old, and some don't take their drug, and some are alcoholic, and some of them have sickle cell disease. And you find that the effectiveness of that drug is not what you thought it would be from your ideal-world efficacy trial.

>From a public health perspective, what really matters is effectiveness. Most people who come to conferences don't really think about effectiveness. Most of the people you'll hear talking about combination therapy here will be talking about efficacy. But it's effectiveness that actually matters. This is why collaboration with behavioral scientists--and not only behavioral scientists, but others as well--is so crucial. We need them to help us understand the best way to get the drugs to the people who need them and to make sure they actually take them.

Have behavioral scientists contributed greatly to the study of adherence? I would say yes. Has it been a useful contribution in terms of increasing adherence? If you look again at tuberculosis, where this has been studied the most, it's actually kind of discouraging. What has been the answer in treating tuberculosis, what have we come around to 25 years later? Directly observed therapy--watching the patient take the drug. The researchers who did these excellent trials on directly observed therapy in India decades ago, they're the ones who dreamt up the strategy. Since then, nothing behavioral scientists have come up with to improve adherence has been as effective.

So, getting back to drug-resistant HIV, I don't think we can expect one magic answer for every situation. There will probably be multiple answers, and it's going to be a bugbear.

The other thing about resistance we need to remember is that we study it in the laboratory, and what we find in the laboratory doesn't always correlate with what we find clinically. For example, you can sometimes have laboratory resistance to certain antimicrobials, and yet it doesn't really correlate with what happens in clinical practice, which is confusing to say the least. But obviously this whole issue is a major cause for concern, not least because an increased rate of resistance can lead to transmission of multiply resistant viruses. And we've seen what that has done as far as multidrug-resistant TB is concerned.

Journal:The prevalence of infection with non-B subtype virus is much greater here in the UK than in the US. But even in the US there is a trickle of reports of people with non-B virus. Where do you stand in the debate on whether non-B viruses could alter patterns of transmission or virulence here and in the States?

De Cock: As far as virulence is concerned, we reported in Vancouver that there is no evidence of a difference in the natural history of HIV disease among Africans living in London compared with non-Africans in London.2 Most of those Africans are infected with non-B subtypes. My conclusion from that retrospective cohort study is that, at least among people with CD4+ counts below 300, there does not appear to be a subtype-based difference in natural history if people have similar access to care.

On the question of transmission, the data are difficult and they are not conclusive. There are, first of all, laboratory data published by Max Essex's group suggesting that subtype E is more tropic to Langerhans' cells found in the foreskin of the male and the cervix of the female and that they grow better there than B subtypes.3 Essex argued that this perhaps explains the rapid heterosexual transmission of E-clade viruses. Also, a cross-sectional study from Thailand looked at the concordance of viral subtypes between men infected with subtype B or subtype E and their sex partners.4 That study suggested that more of these men had in fact transmitted E to their sexual partners than had transmitted B. But there are potential problems with cross-sectional observations of that kind. So those data are extremely interesting but do not resolve the issue.

In Belgium and in the United Kingdom, there are a lot of people with non-B subtypes. Most heterosexuals in Belgium and the United Kingdom are infected with non-B virus. Yet we have not seen an explosion of heterosexual transmission in either of these countries. And finally, don't forget that in countries like Honduras and Haiti you have heterosexual epidemics and the subtype there is predominantly B.

What I take away from all of this is that we need to keep an open mind. If you think epidemiologically, you start with the null hypothesis and you say there is no difference until there is proven to be a difference. The data I would want are epidemiologic data from populations, and those data are not there. So at the moment I see no evidence that there is differential transmission of these different subtypes or that they have a different natural history. I'm certainly not saying that there is not a difference, just that the data aren't there.

Return to Contents

Best-case Scenario in Developing Countries

Journal: We can all imagine the most pessimistic scenario for the HIV epidemic in the poor nations of the world. But let me ask you to look at it differently for a moment: If everything went as well as it possibly could over the next five years, what is the most optimistic scenario for poor countries?

De Cock: I'm not sure I can answer that question, but let me try to answer parts of it. In Vancouver there was a representative of ACT UP who spoke passionately in favor of cutting drug prices and giving drugs to the Third World. I don't believe that flooding Africa with ritonavir, for instance, would make a bit of difference.

We've talked about the complexity of HIV/AIDS care. People with HIV need a lot of care. They need a good physician who understands all this. If they go to the hospital, they need a laboratory that can do sophisticated hematology and microbiology, that can diagnose bacterial infections. They need to be able to get to the hospital. The patient needs food, needs nursing care. Drugs are only a very small part of the whole problem. One must go to the poorest of countries and see what healthcare is like there to understand what we're talking about. Clearly, a lot of people who talk at conferences, perhaps with the best of intentions, actually have no idea what this is all about.

To improve matters in the poorest of countries, we need to take a stepwise approach to dissect out all of the different things that are needed, and then ask where improvements can be made. I'm talking about the very poorest of countries. We have to be very careful: the developing world is not one place. Africa is not one place. We have a tendency to talk about Africa as if it's one large village, as if everybody is the same and they're all living at subsistence level. It's not like that.

You need to talk separately about a whole other category of countries, most of which are not in Africa, that are not industrialized but that certainly are not simply "developing countries"--countries like India, Thailand, Brazil. These countries are at intermediate levels between the poorest countries and industrialized countries. They have large pockets of extreme poverty that are not different from the poorest regions of Africa. But those intermediate countries actually have a whole lot of other problems that the poorest countries don't.

For example, there has been a nosocomial outbreak of HIV infection in a dialysis unit in Cairo. You won't get that happening in Tanzania: there are no dialysis units there. In India you've got particular problems with commercial blood transfusions. Well, if you don't have an industry of commercial blood transfusion, you won't have those kinds of problems. It's not a coincidence that the Middle East has such a high rate of hepatitis C infection; it's probably largely related to medical practice there. If you've seen some of the outbreaks of hemorrhagic fever, like Lassa fever in Nigeria, they were amplified by hospitals. Those steps along the road of development bring with them their own hazards.

But let's go back to really poor countries. What is the best-case scenario for them? They can do a lot. Let's take Uganda. In sentinel populations the incidence and prevalence of HIV infection have dropped. And I think we can safely say now these observations are real. Why it's happened is probably complicated. Some of it simply is that prevalence will have dropped because HIV-infected people have died and the incidence hasn't been high enough to replace them. But I think that behavior probably has changed. Now that's encouraging.

There are lot of things that could be done in such countries to make things better. What are they? In no particular order: better surveillance so that we better understand what's going on, both serosurveillance for HIV prevalence in sentinel populations, particularly pregnant women, and better AIDS case surveillance to assess the burden of disease and so inform decision making. Education, particularly of youth. STD control integrated with education about behavior change and integrated with condom promotion. Providing a safe blood supply, and part of ensuring a safe blood supply is, above all, getting rid of unnecessary transfusions. Identifying safe donors. Defining a minimum standard of care for HIV and making sure that at least what can be done is done in a consistent and sensible fashion. Investing more in tuberculosis control, and doing it right. And educating women and doing what one can for women's rights. That package, I think, would do a great deal, if there were political commitment and money for those things. Now that's not asking for the sky. All of that's doable. That's my best case scenario, but all of that


Journal: What do you mean by "defining a minimum standard of care"?

De Cock: If you go and ask HIV-infected people what they want, or go and see the hospitals and define what you think they need, you can make a hierarchical list: People need shelter, they need clothes, they need food, they need water, they need nursing, they need to be kept clean. Now all of those things come far above ritonavir and indinavir. Because if you can't supply all that, what's the point of having a stack of protease inhibitors in the corner? And then you get to things like treatment for simple opportunistic infections, and symptom control. Actually, antiretroviral drugs come very, very, very low down on the list.

Having said all of that, we must remember that populations in these countries are not homogeneous. You can go to any Third-World city, even the poorest, and you will find a group of people who for one reason or another actually have access to tertiary medical care, medical care of North American or European standards, or who can fly out of the country. If I were working as a physician in some of the African countries I lived in, I would feel obliged to give some patients that kind of care, if they could afford it. A physician's job is to treat individual patients. If you're working as a clinician, it's a different hat from being a public-health person. So I'm sure that in every city there will be people struggling, with the resources they have, to get the sort of care we were talking about earlier in the interview.

Journal: What should the priorities be in those "intermediate" countries you talked about?

De Cock: I think the situation of countries like Brazil and Thailand is quite different from those poorest of the poor, although even in those countries you have pockets of rock-bottom poverty. In many ways it's more complicated for those countries to determine priorities, because they have more resources, and there may be more pressure to provide drugs. But I think people who work in public health in these countries are of a uniform mind that antiretroviral drugs and sophisticated care for people with advanced HIV disease is not a good investment, because you can get more health out of that money, including for HIV-infected people, by doing other things.

I think a clear priority for research in all developing countries is going to be developing a prophylaxis program (Table 1): What kind of package of prophylaxis should be offered, and to whom can it be offered? Even in the poorest of countries, there are probably groups to which such a package could be made available, because they are potentially easily identifiable. In Nairobi, for example, you're not going to be able to identify all HIV-infected people. But you know there's going to be a certain amount of HIV infection among the police, the military, students. You could be doing something for groups like that. It's inequitable and it's not going to do much for the epidemic. But care has to be provided where it can be.

Table 1. Research priorities for the prevention of
opportunistic infections in resource-poor countries.
  • Earlier identification of HIV infection
  • Documentation of spectrum of disease in different regions
  • Determination of environmental sources of opportunistic pathogens and ways to reduce exposure
  • Assessment of chemoprophylaxis and vaccination against specific OIs
  • Evaluation of inexpensive alternatives to CD4+ lymphocyte quantitation for triggering initiation of chemoprophylaxis
Adapted with permission from the American Journal of Tropical Medicine and Hygiene5

Return to Contents

A "Package of Prophylaxis," and Perinatal Transmission

Journal: What would be in such a package of prophylaxis for HIV-positive people?

De Cock: It would probably consist of preventive therapy for tuberculosis, co-trimoxazole, which would be for the rare cases of PCP but also to prevent toxoplasmosis and perhaps some bacterial infections including nocardiosis. It might include pneumococcal vaccine against pneumococcal pneumonia. Those would be the principal agents. In some countries you might consider including azoles because of the high incidence of cryptococcal disease, but that's expensive. Of course the merit of this approach is untested and speculative. [For a detailed analysis of opportunistic infections in the developing world, see the recent article by De Cock and colleagues from several US institutions.5]

Journal: Prevention of perinatal transmission appears to be one area in which antiretroviral drugs can play an immediate role in poor countries. Does that strategy have to be studied exhaustively from population to population, or should concerned parties simply be working on ways to implement these strategies as quickly as possible?

De Cock: The problem with the regimen recommended for the United States and Europe is that it requires identification of the woman by the third trimester. Treatment is started then and continued through delivery and is given to the infant in the first week of life. That is just not feasible in a Third-World context. In many developing countries most women have very little antenatal care, and in the poorest of countries they often present in labor. So the studies that are going on in developing countries are looking at a very different question: Would it be effective to give zidovudine late in pregnancy or during labor? Can you get away without treating the infant?**** I think those are valid research questions.

****One Thai study presented at Birmingham found that infusion of ZDV to the mother only during labor is not effective in preventing transmission.

But I think questions about preventing transmission are going to get more complicated now that we know other drugs are more potent than zidovudine. And now that we believe postexposure prophylaxis works, we have to think about postexposure prophylaxis for the infant. Breast feeding is another complication. If an HIV-infected woman breast-feeds, even if the infant escapes infection during birth, that child may remain exposed from breast- feeding afterwards. There are many unresolved issues. So I think continuing research is required. I think it's ethical, and I think it's needed. But I think it's complex.

I also think it's reasonable to do it for another reason: Many people say, "Look, this will never be practical in developing countries. It's not even ethical to try it." I think that's judging the future in a way one cannot do. One shouldn't forget there was a time when penicillin was so expensive, so valuable, that the urine of patients who took it was collected and distilled to recover the excreted drug. When the hepatitis B vaccine was studied in developing countries, people said, "This is not reasonable; it's too expensive." Well, in some countries today hepatitis B vaccine is integrated into the EPI [Expanded Program on Immunization]. The truth of today is not necessarily the truth of tomorrow. So I have no problem with trials of drugs to prevent perinatal transmission in developing countries, provided of course that they're done with all the ethical clearances and cautions that are necessary. I think that research is needed.

HIV and TB

Journal: Is it possible in non-Western countries even to think of stopping the HIV epidemic without simultaneously beginning to control tuberculosis?

De Cock: These two infections are very intimately associated and have very complex interactions (Table 2). Clearly, in developing countries tuberculosis is the most important opportunistic infection for people with HIV. People with HIV are highly at risk for TB, so it's very important for them that tuberculosis is as controlled as it could be. The other side of that coin is that, in areas with high rates of HIV infection, people with HIV contribute the majority of new cases of tuberculosis and they are infectious to the rest of the population, whether the rest of the population is HIV positive or HIV negative. So control of these two epidemics has to be better integrated than it is now.6

Table 2. Worldwide comparison of HIV and TB
Prevalence of infectionAbout 22 million people worldwide,
90% in developing countries

In Africa, the continent where the
prevalence of HIV infection is the
highest, 30% of people in some
cities are HIV positive
About 1.9 billion people worldwide
(one-third of the world population),
95% in developing countries

Up to 80% of people in the world's
poorest and most crowded cities
carry the TB bacillus
Morbidity and mortalityIn 1996, 3.1 million people
became infected with HIV and
1.5 million died of AIDS
Between 8 and 10 million people
become sick with TB every year
and 3 million die of the disease
Survival with infectionAbout 50% of HIV-positive people
have AIDS within 8 to 10 years of
infection; average survival with AIDS is
6 months in developing countries and
1 to 3 years in developed countries
About half of all TB
patients die within 5 years if
untreated; if they also have
AIDS they die within weeks
or months
CoinfectionApproximately 5.6 million people are coinfected with HIV and TB
From: Armstrong S. A Deadly Partnership: Tuberculosis in the Era of HIV. Geneva: World Health Organization, 1966. Adapted with permission.

Journal: You've observed the irony that for tuberculosis we now have a vaccine, preventive therapy, and curative therapy, yet it kills more than 3 million people a year. If we are ever lucky enough to have a vaccine and reasonable therapy for HIV disease, are we any more likely to control HIV globally?

De Cock: Based on our experience with other infectious diseases, one would have to be very skeptical. And there are examples of failure other than TB. Measles. We haven't done well enough with measles, yet measles can be prevented by one vaccine. It's a good vaccine. You give it to children and they don't die of measles. That's it--they're immune for life. So you certainly have to be skeptical about controlling HIV worldwide, in anything like a rapid fashion, even if our drugs get better and vaccines are developed.

Journal: The World Health Organization says that only 10 percent of the world's TB patients receive directly observed therapy, which is crucially important to preventing multidrug-resistant TB. How important is it to implement directly observed therapy globally, and how difficult is it?

De Cock: It's immensely difficult, and many consider it immensely important. No strategy has given better adherence and completion rates than directly observed therapy. I'm not sure that everybody believes that it's possible everywhere, but my own belief is that we aren't going to get any better strategy than that. The World Health Organization tuberculosis control program has adopted the slogan "DOTS"--directly observed therapy, short course. ThatÕs their marketing slogan and the term they've given to their control package. I don't think they're saying that every patient everywhere must be or can be treated with directly observed therapy throughout six or eight months, depending on which regimen is used. But they are recommending direct observation if rifampicin is used. As I mentioned earlier, a lot of people feel strongly about that. And there are ways to implement directly observed therapy without relying on healthcare workers. In some parts of the world, such as KwaZulu/Natal in rural South Africa, very innovative work has been done using community members to supervise treatment.

Return to Contents

An obscure Disease in a Faraway Place

Journal: When I interview physicians and scientists about their prognosis for the HIV epidemic in poor countries, I always wonder what readers of the Journal think when they get to that part of the interview. Do they think, "I have enough problems with my own HIV patients here at Harlem Hospital, or even here in West Hollywood, to worry about the rest of the world"? Besides the moral reasons for concern about epidemic diseases in poor countries, are there also medical and economic reasons why people in the West should be concerned?

De Cock: I think so, and I think HIV is the best example: It didn't start in the United States. It was imported into the United States from somewhere else (although the United States has exported a fair amount of HIV as well). A separate topic of much discussion at the moment--one that is very important for the Centers for Disease Control and Prevention and a major interest of Vice President Gore--is the whole subject of emerging infectious diseases. That concern clearly demonstrates a growing awareness of the importance of medical events in other parts of the world. The world is so small today that you can travel from one extreme to another in considerably less time than the incubation period of most infectious diseases. So outbreaks of an apparently obscure disease in a faraway place are no longer irrelevant, as they may have been 50 years ago. If you look at events over the last 20 years--and HIV has to be the best example of an emerging infection--I think the lessons are there.

But, yes, in the introduction to that question I think you're right. People in developed countries who are not involved in international issues on a daily basis do turn away from it. I think you see that with current events in Zaire, for example. There is an impatience. People say, "Look, this is not my problem. There's nothing I can do about this." Some of it's understandable. Some of it is not.

Journal: When you look back on the years you spent in Kenya and Côte d'Ivoire, as a physician who was schooled and trained in the West, how did that experience most change the way you think about the practice of medicine?

De Cock: From a personal level, you go into that sort of work because you're interested in infectious diseases. That's why I went. And I think people who do it, on the whole, say that it changes them profoundly. It changes you both personally and professionally. I learned enormous amounts. Professionally you see things that you would never see. You get responsibilities that you would never have. It was enormously enriching and a true privilege to be able to do these things. It changes you for life.

That's speaking from a personal point of view. But, looking at it from a wider perspective, I think the international community should pay tribute to what's been learned in Africa about HIV. Much of what we now know about HIV disease was learned in Africa. Our whole understanding of the differences between HIV-1 and HIV-2 comes from Africa, as does our understanding of the relationship between HIV and SIV. The question of group O viruses was first raised in Africans. The development of diagnostic testing relied greatly on African specimens to make sure we could pick up the different subtypes of the virus. Heterosexual transmission. Mother-to-child transmission. HIV transmission through breast milk. The interaction between HIV and other STDs. Prevention of HIV transmission by control of STDs in the famous Lancet7,8 study. The list goes on. What we've learned about HIV from Africa is enormous. And when you ask how much this work cost, it was an extraordinarily good investment. If you compare that to what's spent in the United States on all kinds of stuff, you've got good return on your money.

Mark Mascolini writes about HIV (


1. Carpenter CCJ, Fischl MA, Hammer SM, et al. Antiretroviral therapy for HIV infection in 1996: Recommendations of an international panel. JAMA 1996;276:146-154.

2. Petruckevitch A, Del Amo J, Phillips A, et al. Survival and CD4+ lymphocyte decline in HIV-infected Africans and non-Africans in London. Abstract We.B.3337. XI International Conference on AIDS. Vancouver, July 7-12, 1996.

3. Soto-Ramirez LE, Renjifo B, McLane MF, et al. HIV-1 Langerhans' cell tropism associated with heterosexual transmission of HIV. Science 1996;271:1291-1293.

4. Kunanusont C, Foy HM, Kreiss JK, et al. HIV-1 subtypes and male-to-female transmission in Thailand. Lancet. 1995;345:1078-1083.

5. Kaplan JE, Hu DJ, Holmes KK, et al. Preventing opportunistic infections in human immunodeficiency virus-infected persons: implications for the developing world. Am J Trop Med Hyg 1996;55:1-11.

6. De Cock KM, Binkin NJ, Zuber PLF, et al. Research issues involving HIV-associated tuberculosis in resource-poor countries. JAMA 1996;276:1502-1507.

7. Grosskurth H, Mosha F, Todd J, et al. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial. Lancet 1995;346:530-536.

8. Laga M. STD control for HIV prevention--it works! Lancet 1995;346:518-519.

Return to Contents

©1997, Medical Publications Corporation