Universal Voluntary Testing and Treatment for Prevention of HIV Transmission
Dieffenbach CW, Fauci AS. JAMA 2009;301:2380-2383. 

NIH leadership in HIV addressed the high priority research agenda of “test and treat” all persons with HIV infection as a public health strategy to control the HIV epidemic. The basis for the research is the modeling study of Granich, et al.. from WHO (Lancet 2009;373:48). This document predicts the pandemic could be ended by global HIV testing of all persons over 15 years annually with immediate institution of antiretroviral agents regardless of CD4 cell count. Several issues are raised by Drs. Dieffenbach and Fauci relevant to this hypothesis and strategy:

• Universal testing: The challenge here is great since the CDC attempt to substantially increase testing in the US is still confronting substantial barriers (MMWR 2006;55:RR14:1). There is the further issue of obstacles to treatment including health insurance, substance abuse, mental illness, and denial.
• State of infection: The model by Granich, et al.. assumes that 10% of transmissions will occur during acute HIV infection. Others have postulated a much larger number during the acute stage (Pinkerton SD. AIDS Behav 2008;12:677). This is highly relevant to this issue since annual testing will usually miss the acute stage of infection.
• Efficiency of ART for Prevention: The assumption in the model is that antiretroviral treatment will be 99% effective in blocking transmission. This is an assumption that is not well established in a treated population.
• Drug Resistance: The authors note two issues: First is the effect of resistance on clinical management and the second is the contribution of rebound viremia on transmission prevention.
• Behavioral disinhibition: One modeling study showed behavior disinhibition implemented with treatment led to loss of any prevention benefit (Velasco-Hernandez JX. Lancet Infect Dis 2002;2:487).
• Benefit to individual: The authors argue that there must be benefit to the individual with early treatment if the “test and treat” strategy is ethical. This benefit of early treatment is not clearly established.
• Cost-effectiveness for society: This can be measured based on economic modeling, but “focused changes could have significant impact.”

Conclusions: The authors note that the proposal of Granich, et al.. poses a “testable strategy that potentially could curtail the global HIV pandemic,” but a number of concerns could impact the model. These include the effects of behavioral disinhibition, drug resistance, and HIV transmission before the first test is done. The authors note that these variables need to be included in the mathematical model.

Comment: It is interesting to recount the opening address of the 2006 International AIDS Conference in Toronto with the statement on behalf of the WHO from Kevin DeCock that “we will never treat ourselves out of this epidemic.” The program proposed by Granich, et al. predicts that the “test and treat” strategy could reduce the incidence of HIV from 20 cases/1000 to 1 case/1000 in 10 years and could end the epidemic in 50 years. The review in JAMA identifies this issue as a major priority for NIAID. It is conceivably doable with adequate funding through the Global Fund and PEPFAR, but probably totally unrealistic in the current economic climate. It must be emphasized that less than half of adults in the US have ever had HIV testing and even the obstetrical experience which is considered so effective shows no test associated with 30% of deliveries. Then there is the issue of entry and retention in care. It seems very clear that successful treatment with no detectable virus nearly eliminates the risk of transmission on an individual basis, but application of this strategy at the population level is a huge unknown and logistic challenge. What the NIH is saying is that it is a thesis that must be tested, especially since other prevention strategies have largely failed.