HAART Drugs Are the New Prevention

David Crowe
January, 2012

When doctors and researchers who also happen to be some of Canada’s top AIDS drug pushers, such as Julio Montaner , say that non-disclosure of HIV status in a sexual relationship should be decriminalized – what’s not to like about that? Call me cynical, but I could not believe that AIDS experts, so tightly aligned with the drug companies, had decided that HIV is not a deadly sexually-transmitted disease or that HIV tests are not accurate.

I did not have to look far for my cynicism to be rewarded. The Canadian Globe and Mail newspaper reported researchers had found that, “HAART-treated patients become dramatically less likely to transmit the infection…Therefore, these people should not be found guilty for exposing sexual partners to HIV.” The country’s other national newspaper, the National Post was more specific, “Researchers reported earlier [in 2011] that early treatment with HAART reduces by 96% the chances that an infected person will transmit the virus to a partner”. Clearly, the get-out-of-jail-free card has some strings attached – lifetime adherence to the AIDS drug cocktails known as HAART – Highly Active Anti-Retroviral Therapy. If judges start to accept this reasoning they will almost certainly make regular testing for AIDS drugs a condition of freedom. Prison or Poison will be your choice.

One could argue against this research’s conclusions from basic principles, sight unseen. Given that HIV tests are not properly validated, given that the evidence for sexual transmission is contradictory, lacking the evidence that HIV-antibody-positive people will definitely or even probably get AIDS in the absence of other risk factors. the fact that people on AIDS drugs (HAART) have lower viral load or their partners are less likely to have HIV antibodies really does not mean anything.

But, given that the rest of the world, notably most doctors, researchers, politicians and judges, believe the HIV=AIDS dogmas, it is important to examine this research and find out if there are any flaws. It is, after all, hard to explain why there would be a difference in the development of HIV antibodies in one sexual partner based on the drug use of the other if there is nothing being sexually transmitted between them and if the data in this study are sound.

The HPTN 052 Clinical Trial

These newspaper articles stemmed from the HPTN 052 study, a comparison of HIV transmission in serodiscordant couples (i.e. one person was HIV-positive and one HIV-negative in each couple) observed for possible HIV seroconversion. [Cohen MS, 2011] Half of the HIV-positive people were put on HAART at the beginnning of the study and half had AIDS drugs delayed until they showed signs of AIDS or, more likely, their CD4 counts declined below 250. In almost exactly half of the couples the man was HIV-positive and in the other half the women. Most were stable, married, heterosexual relationships and the participants were recruited in India, Brazil, Thailand and several countries in southern Africa, most notably Malawi. A transmission was assumed to have occurred when the HIV-negative partner seroconverted to positive. The newly positive partner was then analyzed to see if the genetic materials believed to come from HIV were similar to that of their sexual partner.

It was easy to find articles about this study, all of them laudatory, particularly in the prestigious British medical journal Lancet (published in May and July), although it was difficult to find the actual research publication (not published until August). The adulation was kicked off by an editorial in the Lancet with a title that leaves no room for doubt, “HIV treatment as prevention – it works”.[Lancet, 2011] Canada’s Julio Montaner was invited to write another article, a sign of his high global status in the ‘Treat Everybody’ world, with the not very imaginative title, “Treatment as prevention--a double hat-trick”, but he also based his excitement on two NIH press releases in the rush to endorse the conclusions before the data were available.[Montaner, 2011] Epstein and Morris, both proponents of traditional HIV prevention (ABC=Abstinence, Be faithful, Condoms) were quite shocked that the drug pushers are trying to take over their traditional territory by claiming that drugs do a better job of HIV prevention than condoms but it was too late given the earlier editorial endorsement by the Lancet.[Epstein, 2011] The year ended with another prestigious journal, Science, declaring HPTN 052 the “Breakthrough of the Year”.[Cohen J, 2011]

If I can put myself inside the head of a mainstream AIDS researcher for a moment, I can understand the excitement. A 96% reduction in infectivity is simply stunning. In the ponderous world of scientism everyone is accustomed to fulsome praise for papers that claim an increase in lifespan of a few months or even just days for a cancer drug, an increase of a few percentage points in CD4 counts, or a few more percent with ‘undetectable’ viral load.

Almost Perfect Condom Usage

The first striking claim in the study is the almost perfect rate of condom usage – only 5% or 6% of sexual contacts in the week before enrollment of each participant were reported as “unsafe” (i.e. without a condom). In other words, the researchers are claiming that at least 94% of the sex between the participants in the study was protected by condoms. The authors show no skepticism about these claims even though it is plausible that they might be driven by the desire of the participants to provide a morally correct answer to ensure they receive their compensation for participation (allowed by the study protocol with the amount determined locally) or simply to please the researchers. The researchers even go so far as to report the hazard ratio for transmission with 100% condom use versus less than 100% conformance. The precise number who claimed perfect condom adherence was not specified, but it must be a large percentage or it would have been impossible to determine a hazard ratio. Reporting a significant rate of transmissions in people using condoms religiously was perhaps intended to be another nail in the coffin of the condom.

It would seem pointless to arrange a study of HIV transmission in a population that seems so extraordinarily careful about protecting themselves against the virus – the number of transmissions should be very low. One reason for proceeding, assuming that the researchers were aware that this population would report near-perfect condom use, might have been to ensure ethical approval. By claiming that the majority of participants were protecting themselves they could avoid the accusation that the researchers were just going to observe people getting infected and dying. Whether the researchers truly believed the claims of the participants or not is irrelevant, if they had any doubts about what they were being told it would be in their interest to keep quiet.

A second reason for reporting a high rate of condom use could be political. By studying HIV transmission in an environment of extremely high condom use they could make the argument that condoms alone are not sufficient prevention and, if the reduction in seroconversion with drugs was good enough perhaps they could knock the condom off its throne as the king of protection against HIV. Again, skepticism about the reported rate of condom usage would not serve the interests of the researchers.

The claimed condom usage is, bluntly, impossible to believe. Other researchers reported that, in Malawi, condom usage had increased from 14% in 2000 to 20% in 2004 among 15-49 year old men in urban Malawi and from 27% to 35% among the younger group, 15-24 year olds.[Bello, 2011] Other researchers noted that, “Highly negative attitudes toward condoms in Malawi also remain a barrier to prevention” and found that a prevention intervention had only raised the percentage of sexually active adults who had used condoms at least once in the past two months from 12.3% to 24.5%. And, to emphasize, this is the percentage who had ever used condoms in a two month period so even the 24.5% might still have had sex without condoms some or most of the time.[Kaponda, 2011] Another study in two of the same cities as the HPTN 052 study (Lilongwe, Malawi and Johannesburg, South Africa) found that even among people who had recently been told that they had just become infected with HIV and even after two visits with researchers who were obviously intent on promoting the use of condoms, 11.8% still said that they had had unprotected sex in the previous week. And since well over 50% had not had sex at all, and since participants may well have started to give answers they believed to be morally correct, the percentage of sex acts without a condom could be considerably higher.[Pettifor, 2011]

If the condom data are wrong then all the data gathered from the participants are questionable and therefore the entire study and its dramatic conclusions. If the data are right it needs to be explained how these researchers stumbled across the nirvana that has eluded all HIV prevention researchers.

Conflicts of Interest

The interests of the researchers are the interests of the drug companies. Supplementary material contains lengthy financial disclosures revealing that one co-author, Joseph Eron, consults for Merck, Bristol Myers Squibb, Tibotec, Gilead and GlaxoSmithKline (all AIDS drug manufacturers) and received grants from Merck, BristolMyersSquibb and Roche. Susan Eshleman consults or has otherwise received payments from Roche, the Gates Foundation (which has extensive investments in pharmaceuticals), Abbott Diagnostics, Celera and Monogram Biosciences. Joel Gallant has received money from Abbott, BristolMyersSquibb, Gilead, Merck, Tibotec, VIIV Healthcare, GlaxoSmithKline, Pfizer, Monogram Biosciences, Sangamo Biosciences and Koronis. Susan Swindells received money from Gilead, Abbott, Pfizer, GlaxoSmithKline, BristolMyersSquibb. Given that disclosure forms are rarely audited, and often contain omissions, it is quite likely that conflicts exist with other researchers. Pharmaceutical support for the study (presumably free drugs) was provided by Boehringer-Ingelheim, Bristol-Myers Squibb, Gilead, GlaxoSmithKline and Merck, and the same companies participated in the protocol team. Although, as in most AIDS drug trials, specific drugs are not required, they are apparently all considered equivalent and interchangeable, almost all the participants in the trial were using drugs from names you will recognize – GlaxoSmithKline, BristolMyersSquibb, Gilead and Abbott. Over 90% were using regimen’s containing the first AIDS drug approved, perhaps the most toxic of them all – GlaxoSmithKline’s AZT.

Dramatic Differences in HIV Infectivity

The most interesting finding of the study is not found in the study publication at all. Buried in the supplementary data, which we can safely assume few will read, is evidence of dramatic differences in HIV “infectivity” between different geographical sites. The structure of the study provides an almost perfect analysis of HIV infectivity globally because 94% of serodiscordant couples were married (implying a stable sexual relationship) and the numbers of HIV-positive men, HIV-positive women, HIV-negative men and HIV-negative women were almost equal (i.e. in almost exactly half the couples the man was HIV-positive and in the other half the woman). This eliminates most differences in HIV prevalence in those geographical areas, it comes down to just the virus and the genitalia of their HIV-positive partner. This study can be predicted, based on its design and the standard HIV=AIDS dogma, to show similar rates of infection between geographical sites and a several times higher rate of transmission from man-to-woman than woman-to-man.

It does not.

Not only does the supplementary data shatter preconceptions about HIV infectivity, but there is no mention of these dramatic findings in the body of the paper. When I asked for some clarifications and additional data from the authors (such as the actual number of HIV seroconversions at each site), I was stonewalled.

The supplementary data from HPTN 052 compiled below reports the incidence of HIV seroconversion in the negative partners (the paper refers to it as “HIV Infection”) expressed in 100 person-years. An incidence of 2.0, for example, could mean 2 seroconversions in 100 people studied for 1 year, 1 seroconversion in 50 people studied for 1 year or 4 seroconversions in 100 people studied for 2 years. This measure is more useful than the actual number of seroconversions because it is adjusted for the number of local participants and the length of follow-up, both of which varied between sites in this study.

Table 1: HIV Incidence in HPTN 052

(Number of Couples)
HIV Incidence
(per 100 person-years)
Incidence relative
to India
Porto Alegre 90 0.5  
Rio de Janeiro 186 0.5
Brazil Subtotal 276 0.5 5
Boston, MA 2 0
USA Subtotal 2 0.0 n/a
Chennai 250 0.2  
Pune 175 0.0
India Subtotal 425 0.1 1
Chiang Mai 106 0.9  
Thailand Subtotal 106 0.9 9
Kisumu, Kenya 60 0.0  
Harare, Zimbabwe 240 1.0
Gaborone, Botswana 77 2.2
Johannesburg, South Africa 46 0.0
Soweto, South Africa 50 2.1
Blantyre, Malawi 230 3.2
Lilongwe, Malawi 251 3.4
All Africa Subtotal 954 2.3 23
Except Malawi Subtotal 473 1.1 11
Only Malawi Subtotal 481 3.3 33


If we accept that seroconversion is proof of infection, this table shows that infectivity was 33 times higher in Malawi than in India. Because all couples in the study were in approximately the same situation, a stable sexual relationship, mostly married, mostly heterosexual, it is difficult to explain this. Note that the study claimed that HIV incidence was 28 times higher without HAART than with HAART so the 33 times effect of being in Malawi rather than India is even more dramatic, yet not mentioned in the published article.

Some of the possible explanations of the dramatic differences in incidence of seroconversion regionally are:

  • HIV is truly more infectious in Africa (although it is hard to see how something as simple as a virus could vary this dramatically in such a crticial characteristic).
  • Africans are many times more susceptible to the virus than Indians (although it is hard to see how there could be such dramatic differences in the permeability of human genital membranes).
  • HIV antibodies do not represent HIV infection but something else that differs between these countries, such as genetics or environment (although it is hard to explain why seroconversion would then be so much more frequent in the negative partner when the other person is taking drugs).
  • There is something terribly wrong with the study, particularly in the Malawi sites, such as bias, error or fraud.

Impact on the Conclusions

Even assuming that HIV seroconversion truly represents HIV infection, the study's conclusion that HAART reduces infectivity by 96% cannot be supported as a general statement. For one thing, there are far too few seroconversions in Brazil and India to draw any conclusions. Secondly, there must be something else causing the high rate of seroconversion in Malawi that must be analyzed. This could be an environmental factor causing susceptibility to HIV or false positive HIV tests, or some hidden bias or manipulation.

There is not a hint of interest in the article about what makes residents of India virtually immune from HIV compared to residents of Malawi. There is no consideration that excessive reliance on tobacco production or on vitamin-deficient maize , tragic agricultural policies originating with the dictatorship of Dr. Banda, that have produced so much malnutrition and stunting of children, could produce immune deficiency or false positive HIV tests to be misdiagnosed as AIDS. [USAID, 2005]

Comparison with Padian

Nancy Padian established, in 1997, generally accepted risks of HIV infection from sexual contact between an HIV-positive man and a woman (risk to woman 0.0009, just under 1/1000) and between an HIV-positive woman and a man (8 times lower, just over 1/10000).[Padian, 1997] It is important to compare the findings of HPTN 052 with these widely accepted estimates.

First we need to calculate, from Nancy Padian’s predicted rates of transmission, the expected number of seroconversions using some parameters provided by the study:

  • Median follow-up in this study was 1.7 years (i.e. half the participants were in the study for more than this time and half for less than this time).
  • The majority of couples reported that they had sex no more than twice per week, so we will estimate every couple having sex twice a week, likely a slight overestimate.
The probability of a seroconversion in the average couple over the duration of the study can then be calculated separately for men and women (using ^ to mean exponentiation):
  • M->F: 1 - (1-0.0009)^(1.7x52 (weeks) x 2 (sex/week) x 0.06 (frequency of sex without condom)) = 0.95%
  • F->M: 1 - (1-0.0009/8)^(1.7x52 (weeks) x 2 (sex/week) x 0.06 (frequency of sex without condom)) = 0.12%

In the study there were 418 couples with an HIV-positive man who did not initially get HAART and 441 with an HIV-positive woman, therefore we can expect about 4 women seroconverting from an HIV-positive man (0.95% x 418 = 3.97) and (being generous with the mathematics, as 0.12% x 441 = 0.53) one man from an HIV-positive woman, a total of 5. This is significantly less than the 35 seroconversions actually reported among the participants not taking HAART.

Even the difference between men and women predicted by Padian is not found. Men had a risk of seroconversion that was 73% that of women, but Padian predicted that a man's risk of seroconversion from sex with a woman should be 8 times lower. Men having sex with an HIV-positive woman should have had a risk of seroconversion that was only about 13% that of women having sex with an HIV-positive man not 73%.

Padian’s estimate of the number of seroconversions can be converted into incidence (cases per 100 person-years). Among the men a probability of 0.95% over 1.7 years for one man is equivalent to an incidence of 0.6 cases/100 person-years (0.0095/1.7 = x/100, solve for x). For women, the incidence is obviously much lower, 0.07. Averaging, since there are similar numbers of men and women in the study, Padian’s estimate of HIV infection risk predicts an incidence of about 0.3 cases per 100 person-years. This is a bit higher than India (0.1) and a bit lower than Brazil (0.5) but about 5 times lower than Africa in general (2.3) and about 11 times lower than HPTN 052 reported from Malawi.

Safety of HAART

Knowing that the researchers in this study are closely aligned with AIDS drug manufacturers it is not surprising that their own material would describe the fearsome nucleoside analog AZT as benign: “ZDV [Zidovudine aka AZT] is generally well tolerated, particularly in persons with CD4+ >200 cells/mm3. The major side effects include headache, fatigue, malaise, nausea, anemia, and neutropenia [deficiency of infection fighting white blood cells]. Long-term ZDV therapy is associated with myopathy [muscle degeneration] and rare cases of steatosis with hepatic [liver] failure and death.” It is not at all clear what “generally well tolerated” really means as it is often used to imply that a drug is safe and often, as here, close to a description of some of its possible fatal consequences.

There is not much in this study about safety, but there is an admission that, “Grade 3 or 4 laboratory abnormalities during study follow-up occurred in 242 participants (27%) in the early therapy group and 161 participants (18%) in the delayed-therapy group. The most frequent laboratory abnormalities included neutropenia [see above], abnormal phosphate level, and total bilirubin [liver enzyme] elevations [see above].”

An indirect indication of trouble was the fact that 66% of participants changed from their original drug combination to a so-called second-line regimen. The paper stated that, “For participants with virologic failure, specified second-line treatment regimens were provided” (and did not state any other reason for these regimens) but “Virologic failure was observed in…5% in the early-therapy group” (and 3% of the delayed therapy group who initiated treatment). No reason is given for the more than 60% of participants who changed drug regimens without virologic failure. It could be that they were searching for something less toxic and their only choice, apart from leaving the study, was the second-line regimen. It is hard to imagine any other reason.

One might think that the more vigorous virus in Malawi might cause more illness, but the opposite is true. The highest rate of clinical events (incidence 1.3/100 person-years) was in Chennai, India, the site with the lowest incidence of seroconversion except for those sites with no transmissions. Blantyre, Malawi, on the other hand, was one of the two sites in Malawi with the highest rate of HIV seroconversion but the third lowest rate of clinical events out of 12 sites reporting.

Transmission Difference Between Treated and Untreated Partners

It is not hard for AIDS rethinkers to explain geographical differences in the incidence of HIV seroconversion. Since these antibodies are known to be produced by a variety of conditions from prior pregnancy to vaccination to auto-immune illnesses, a higher rate in a country that has genetic and environmental differences from another is not surprising.

But it is not so easy to explain why only 4 of 893 participants in the early therapy group seroconverted and 35 in the delayed therapy group. If HIV antibodies are not indicative of HIV infection how can it be that a chemical given to one person results in a change in their sexual partner? Furthermore, it was claimed that only 1 of the 4 virus genomes in the early-therapy group was linked to the virus genome in the corresponding HIV-positive sexual partner, but 28 in the delayed-therapy group.

There are two alternatives to consider: That the HIV seroconversion and linkage results are valid, or that some problem in the study renders them invalid.

Validating the HIV seroconversion results is virtually impossible because essentially no information is provided. The article merely states, “Through viral genetic analysis, 28 transmissions were linked to the HIV-1–infected participant” without any description of the techniques used, how the cutoff between linked and unlinked was determined, how samples were blinded, how a chain of custody was maintained and so on. A handful of errors could have seriously distorted the results given the relatively small number of seroconversions. For example, if the blood of an HIV-positive partner was accidentally substituted for their HIV-negative partner you would immediately get a ‘seroconversion’ and a ‘matching’ genome.

The most important criticism of HIV genome analysis is that the PCR primers being used have never been proven to come from HIV (as it has never been purified) let alone that they are unique to HIV. Furthermore, in any serodiscordant couple in countries with many poor, probably not very mobile, people, there should, in theory, be the person who caused the initial infection nearby, and that person also should have a similar HIV genome. That person is then another possible source of infection or, more likely, some of their other contacts (assuming heterosexual contact is the cause of these seroconversions). Therefore, the use of local controls is critical to make sure that the genomes of the two partners are not just similar according to some arbitrary cutoff point, but that they are more similar than the genomes found in other members of the community.[Abecasis, 2011]


The impact on the world of the uncritical acceptance of HPTN 052 will be significant. In poor countries, particularly in Africa, it will result in more pressure for healthy HIV-positive people to go on drugs. This pressure will be external, from social workers and medical personnel, but also from within the relationship with spouses telling their HIV-positive partner to take AIDS drugs not just to protect their own health but also that of their partner.

In richer countries it is not clear that there are many more willing consumers of AIDS drugs but the impact will be on sexually active HIV-positive people, particularly heterosexual men. Judges who consider themselves progressive may now give HIV-positive people a choice between a life sentence, 10 or 20 years in jail, or a death sentence, a few years on HAART before dying of ‘AIDS’.


The HPTN 052 study is a genuine hat trick although it is not clear whether this refers to three amazing things or a magician pulling rabbits out of a hat.

The first amazing thing is the claim that HAART drugs reduce infectivity by 96%, a claim that made headlines in the popular and scientific press around the world.

The second amazing thing is that the researchers found several populations in Asia, Africa and South America that are near-perfect condom users. This stunning finding was hidden in plain sight in the article, completely ignored by other media that perhaps did not read past the press releases.

The third amazing thing is that living in India relative to Malawi does more to reduce your risk of HIV seroconversion than taking toxic HAART medications (assuming that HIV seroconversion is a risk). This finding was buried more deeply, in the supplementary files on the journal website, not linked directly to the article, and not mentioned in any way in the article.

The magicians at the major drug companies focussed our attention on the first amazing thing because that is very much in their own financial interest. At they same time, like any good magician, they distracted our attention from the two other amazing things in HPTN 052, that call into question the integrity of the study.

The discrepancies in HPTN 052 are so great and the inherent biases of the authors so obvious that the study cannot be relied upon. Without protest, however, this study will be used to coerce more HIV-positive people onto medication.


[Abecasis, 2011] Abecasis AB et al. Science in court: the myth of HIV fingerprinting. Lancet Infect Dis. 2011 Feb; 11(2): 78–9.
[Bello, 2011] Bello G et al. Evidence for changes in behaviour leading to reductions in HIV prevalence in urban Malawi. Sex Transm Infect. 2011 Jun; 87(4): 296–300.
[Cohen J, 2011] Cohen J. Breakthrough of the year. HIV treatment as prevention. Science. 2011 Dec 23; 334(6063): 1628.
[Cohen MS, 2011] Cohen MS et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011 Aug 11; 365(6): 493–505.
[Epstein, 2011] Epstein H et al. HPTN 052 and the future of HIV treatment and prevention. Lancet. 2011 Jul 16; 378(9787): 225.
[Kaponda, 2011] Kaponda CP et al. Outcomes of an HIV prevention peer group intervention for rural adults in Malawi. Health Educ Behav. 2011 Apr; 38(2): 159–70.
[Lancet, 2011] HIV treatment as prevention – it works. Lancet. 2011 May 21; 377: 1719.
[Montaner, 2011] Montaner JS. Treatment as prevention--a double hat-trick. Lancet. 2011 Jul 16; 378(9787): 208–9.
[Padian, 1997] Padian NS et al. Heterosexual Transmission of Human Immunodeficiency Virus (HIV) in Northern California: Results from a Ten-Year Study. Am J Epidemiol. 1997 Aug; 146(4): 350–7.
[Pettifor, 2011] Pettifor A et al. Continued high risk sexual behavior following diagnosis with acute HIV infection in South Africa and Malawi: implications for prevention. AIDS Behav. 2011 Aug; 15(6): 1243–50.
[USAID, 2005] The government dimensions of food security in Malawi. USAID. 2005 Jun.