Is antiretroviral therapy modifying the HIV epidemic? – Authors' reply

Abstract

We acknowledged in our paper that ours was an ecological study, and as such the results could not be taken as definitive proof of causality. We also indicated that the association between increasing coverage of highly active antiretroviral therapy (HAART) in HIV-infected individuals who met contemporary treatment guidelines and decreasing yearly new HIV diagnoses occurred against a background of increased yearly HIV testing, as well as improved risk ascertainment due to mandatory HIV reporting, and increased rates of sexually transmitted infections. Notably, our results were internally reproducible. We recorded decreases in yearly new HIV diagnoses during two distinct periods of HAART expansion, which were separated by a stable period of HAART use. The latter two periods were characterised on a prospective basis and were entirely consistent with the predictions of our previously published mathematical models.1Lima VD Johnston K Hogg RS et al.Expanded access to highly active antiretroviral therapy: a potentially powerful strategy to curb the growth of the HIV epidemic.J Infect Dis. 2008; 198: 59-67Crossref PubMed Scopus (163) Google Scholar Finally, we were able to relate the reductions in community plasma viral load during HAART expansion with the decrease in new HIV diagnoses, providing a plausible biological mechanism to account for the association, as proposed by others.2Das M Chu PL Santos G-M et al.Decreases in community viral load are accompanied by reductions in new HIV diagnoses in San Francisco.PLoS One. 2010; 5: e11068Crossref PubMed Scopus (624) Google Scholar We also acknowledged that the use of yearly new HIV diagnoses represented a limitation of the study; however, it is also clear that there is no widely accepted gold standard to estimate HIV incidence, particularly in population-wide studies. In this regard, the consistency of our results with our previous report looking at the correlation between community concentrations of HIV-1 RNA in plasma and HIV incidence in a well characterised cohort of injection drug users is highly reassuring.3Wood E Kerr T Marshall B et al.Longitudinal community plasma HIV-1 RNA concentrations and incidence of HIV-1 among injecting drug users: prospective cohort study.BMJ. 2009; 338: b1649Crossref PubMed Scopus (286) Google Scholar Andrew Grulich and David Wilson express concerns about our previous report of declines in unsafe injecting. However, that report was based on a small study within Vancouver's Downtown Eastside, and the present study reports on all of the province of British Columbia, where it has been estimated that the number of sterile needles and syringes distributed remains inadequate.4Harvard SS Hill WD Buxton JA Harm reduction product distribution in British Columbia.Can J Public Health. 2008; 99: 446-450PubMed Google Scholar Grulich and Wilson call for randomised controlled trials (RCTs) to further characterise the associations described in our paper. However, the feasibility of RCTs within the segment of HIV-infected individuals eligible for treatment could prove ethically difficult if this requires withholding HAART against current standard of care. RCTs might be more realistic within the segment of HIV-infected individuals not eligible for therapy, as proposed by the “test and treat” strategy, yet the most recent guidelines5Thompson M Aberg J Cahn P et al.Antiretroviral treatment of adult HIV infection—2010 recommendations of the International AIDS Society–USA Panel.JAMA. 2010; 304: 321-333Crossref PubMed Scopus (766) Google Scholar are so inclusive that this segment has become very small in our setting. Alternative approaches, such as a modified delayed-start or randomised-start design, might be successfully adapted to existing HAART rollout initiatives. In the absence of such data, the available evidence strongly points to a substantial preventive benefit that can be derived from aggressively rolling out HAART to all those in medical need—the actual focus of our report. We declare that we have no conflicts of interest. Is antiretroviral therapy modifying the HIV epidemic?Julio Montaner and colleagues (Aug 14, p 532)1 present data from British Columbia, Canada, on the association between rates of HIV diagnoses, coverage of antiretroviral therapy (ART), and average HIV viral load. They are incorrect in describing their study as a “population-based cohort study”. It is an ecological study, because the data on exposure (ART and viral load) and outcome (HIV diagnoses) do not come from the same couples. This is a crucial distinction because population associations found in ecological studies often fail to reflect individual-level biological effects. Full-Text PDF Is antiretroviral therapy modifying the HIV epidemic?Antiretroviral drugs have clear potential to reduce HIV transmissibility. However, the ecological association Julio Montaner and colleagues1 find between increased individuals on treatment, declines in viral loads, and declining new HIV diagnoses in British Columbia, Canada, does not provide good evidence that treatment reduces population-level incidence. The appropriate variables were not measured, timing of the associated changes lacks biological plausibility, and their findings have better alternative explanations. Full-Text PDF