To reduce HIV transmission, HIV prevention has been based on modifying individual behaviors, particularly with the wide use of condoms. Recently, other strategies of HIV transmission prevention have included “treatment as prevention” (TasP), male circumcision, postexposure prophylaxis, or topical or oral preexposure prophylaxis.1
Among these new strategies, TasP consists of the rapid detection of possible HIV-positive patients and prompt initiation of effective antiretroviral (ARV) therapy. ARV therapy decreases the viral load of the patient and reduces the risk of HIV transmission.2
The screening of the entire population has been recommended in the “2010–2014 French HIV Plan” to facilitate treatment and apply the TasP strategy.3 This strategy is now applied in many countries, and it is being implemented throughout France.
The efficacy of the TasP approach was demonstrated at the individual level, which indicates that a patient with an undetectable viral load is less at risk of transmitting HIV.4,5 At the community level, a new concept, the community viral load (CVL), has emerged. CVL has been defined as the mean of the most recent viral load test of HIV-infected individuals within a geographic area and within a specified period.6 Three recent North American studies concluded that there was an association between a decrease in the CVL and a reduction in the number of people who are newly infected with HIV in a population.6–8 Thus, CVL has been proposed as a marker of the risk of community HIV transmission and as a tool to monitor the HIV epidemic.6,7 In France, no study has examined CVL and its potential association with new HIV diagnoses.
The purpose of our study was to evaluate the value of CVL as a predictive marker of the efficacy of the TasP strategy on the rate of new HIV diagnoses in the northern and eastern regions of France.
MATERIALS AND METHODS
Since 2005, we have annually collected data on the number and epidemiological characteristics of new seropositive patients who initiate care in 5 voluntarily participating university hospitals from the northern and eastern regions of France. The source of data for these newly diagnosed cases was from hospitals themselves.
These 5 participating hospitals, which provide most care to HIV-infected persons in those areas, could be considered a sample, representing 60% of new HIV diagnoses in the 5 administrative regions in northern and eastern France every year.9 This result was estimated by comparing our data with national data from the National Institute for Statistics and economic studies. These hospitals serve clients who are referred from further away.
In the same university hospitals over the same period, we annually estimated the CVL and the CD4 level using the last recorded viral load and CD4 level of the year for every HIV-infected individual followed in each hospital, regardless of their treatment status. The data were collected using the software used in most of French hospitals for HIV patient follow-up (Nadis).10 The study received exemption from consent from appropriate French ethical bodies.
Concerning individual viral loads, the minimum value was set at 100 copies per milliliter, which corresponds to the highest threshold value used to describe undetectable patients in one of the 5 centers. Because more than 50% of the patients had viral loads below the threshold each year, we did not believe that it was relevant to determine the median. Consequently, we considered the annual total CVL, which was calculated as the sum of the values from the most recent viral load for all HIV-infected individuals for each year. Total CVL represented the absolute level of virus in a population; therefore, it is a measure of the potential infectiousness of this population. We modeled trends in the total CVL and new HIV diagnoses using linear regression models. A Poisson regression model was used to determine the relationship between the number of newly diagnosed cases and the total CVL for the same year and between the number of newly diagnosed cases and the total CVL for the previous year. viral load data were log transformed to approximate a normal distribution, and robust standard errors were used to account for overdispersion of the data. The statistical analyses were performed using Stata version 9.2 (StataCorp, Texas).
Between 2005 and 2010, our database included 24,747 patient years, with 17,411 patient years for men and 7335 patient years for women (the sex ratio = 2.37). The number of patients per year was 3573 in 2005, 3847 in 2006, 4034 in 2007, 4224 in 2008, 4431 in 2009, and 4631 in 2010. An individual patient may correspond to 6 different entries, one for each year of follow-up. The mean age was 43.7 ± 10.9 years, and the mean CD4 level was 559 ± 353 copies per milliliter. The proportion of men who have sex with men was 42.12% (n = 10,419), the proportion of heterosexual contamination was 42.14% (n = 10,425), and the proportion of intravenous drug users was 6.36% (n = 1574). The rate of the mean CD4 level at diagnosis increased from 395 per cubic millimeter in 2005 to 404 per cubic millimeter in 2009 (+2.3%, P < 0.001). The proportion of patients treated with ARV therapy for more than 3 months at the time of the measurement of their viral load increased from 69.6% in 2005 to 84.7% in 2010 (+17.9%, P < 0.001).
The trend analysis showed a decrease in the number of new HIV infections (from 256 in 2005 to 246 in 2010; P = 0.067) and a decrease in the total CVL (from 7.9 log in 2005 to 7.76 log10 in 2010; P = 0.005) (Fig. 1). There was a significant correlation between the decrease in the total CVL and the decrease in the number of new diagnoses for the same year (P = 0.01).
Similar to the 3 previous studies,6–8 we found a statistically significant but moderate association between a decrease in the total CVL and a decrease in the number of new HIV diagnoses for the same year in our community.
The moderate association between a decrease in the CVL and a decrease in the number of new HIV diagnoses in our community, representing 5 of the 22 French regions, could be explained by the large, diverse, and open population of our community, which indicates that infections arise from individuals from outside the considered perimeter.
Our approach concerning CVL as a predictive marker of the efficacy of the TasP strategy on the rate of new HIV diagnoses has 2 limitations. First, the approach was based on the viral loads of known seropositive patients. However, in France, 29,000 seropositive people were unaware of their condition, and 22,000 patients who were aware of their seropositive status were not followed up.3 These patients thus constitute a hidden reservoir that could help maintain a higher CVL despite the increase in the number of patients who are treated. These undiagnosed patients were, therefore, at a greater risk of transmitting the virus because they likely had viral loads greater than treated patients did. One may even postulate that the moderate association between CVL and the decrease in the newly diagnosed HIV patients is related to this hidden reservoir of undiagnosed HIV patients, who should be tested earlier and then administered care and treatment for a test-to-treat strategy to be effective.
Moreover, the decrease in the number of new HIV diagnoses does not necessarily imply a reduction in the number of new HIV infections. These results can also be explained by a reduction in new infections in the past or a change in screening practices.
The use of the total CVL as a marker of the risk of community HIV transmission and a tool to monitor the HIV epidemic remains attractive. However, to use this tool, the viral load of the persons who are unaware of their seropositive status should be taken into account.
The clinical stage of HIV-positive patients should also be considered. In fact, HIV transmission risk is higher during acute and early HIV infection than during chronic infection. The viral load of early HIV-infected patients is high, and they may have more unprotected relations because they do not know their seropositive status.11
The decrease in the total CVL was associated with a mild reduction in the number of new HIV diagnoses in a large population. This low association is most likely related to the large number of undiagnosed patients and, to a lesser extent, the size, extent, and dispersal of our community's population.
These results have prompted us to extend screening, as suggested by the most recent French recommendations,3 which is fundamental to implementing effective strategies, such as TasP. The assessment of CVL as a robust predictive marker of the efficacy of the TasP strategy on the rate of new HIV diagnoses is interesting. However, CVL as a marker should be improved to account for the large number of undiagnosed patients and those with early HIV infection.
1. Cohen J. HIV prevention. Halting HIV/AIDS epidemics. Science. 2011;334:1338–1340.
2. Cohen MS, Gay CL. Treatment to prevent transmission of HIV-1. Clin Infect Dis. 2010;50:S85–S95.
4. Quinn TC, Wawer MJ, Sewankambo N, et al.. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med. 2000;342:921–929.
5. Attia S, Egger M, Müller M, et al.. Sexual transmission of HIV according to viral load and antiretroviral therapy: systematic review and meta-analysis. AIDS. 2009;23:1397–1404.
6. Wood E, Kerr T, Marshall BDL, et al.. Longitudinal community plasma HIV-1 RNA concentrations and incidence of HIV-1 among injecting drug users: prospective cohort study. BMJ. 2009;338:b1649.
7. Das M, Chu PL, Santos G-M, et al.. Decreases in community viral load are accompanied by reductions in new HIV infections in San Francisco. PLoS One. 2010;5:e11068.
8. Montaner JSG, Lima VD, Barrios R, et al.. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet. 2010;376:532–539.
9. Henard S, Letranchant L, Borel A, et al.. Study of new HIV infection between 2000 and 2007 in the north and east of France. Med Mal Infect. 2010;40:517–523.
10. Pugliese P, Cuzin L, Enel P, et al.. NADIS 2000, development of an electronic medical record for patients infected by HIV, HBV and HCV. Presse Med. 2003;32:299–303.
11. Powers KA, Ghani AC, Miller WC, et al.. The role of acute and early HIV infection in the spread of HIV and implications for transmission prevention strategies in Lilongwe, Malawi: a modelling study. Lancet. 2011;378:256–268.