Heuker et al.  found that men who have sex with men (MSM) who used antiretroviral postexposure prophylaxis (PEP) between 2000 and 2009 have a significantly higher HIV incidence than MSM participating in the Amsterdam Cohort Study (ACS) in the same period: 6.4 vs. 1.6 per 100 person-years, respectively. We argue that the different ‘time at risk’ used to calculate the incidence could confound comparisons. Indeed, PEP users follow-up was cut-off at 6 months, whereas that of ACS participants was not. This determined a very large difference in the total time under observation that could have diluted the incidence in the ACS: 171 person-years for the 355 PEP users (mean 0.48 person-years) and 4061 person-years for the 782 ACS (mean 5.19 person-years).
Moreover, the longer time of observation could have introduced a selection bias due to the heterogeneity in risk exposure to infection in the ACS: the highest risk individuals become infected earlier, leaving a pool of lower risk individuals at later time points. This could also explain in part the observed time trend.
To contribute to the issue, we reviewed data collected at our Counseling and Testing Site which identifies approximately 30% of the new HIV infections that occur annually in the Latium Region, and is in charge for providing PEP. HIV testing was performed by antibody detection using III generation enzyme immunoassay (EIA) from 2002 to 2006, and then by combined antibody and antigen detection IV generation EIA; positive tests were confirmed by Western blot. Since 2004, to estimate the time since infection, we calculated the Avidity Index of antibodies on the first positive sample, using an automated anti-HIV EIA; an Avidity Index less than 0.80 defines a recent infection acquired in the previous 180 days .
Data were collected from January 2002, when National Recommendations for the use of PEP after sexual exposure were issued, to December 2011 (Table 1).
In the present analysis, only MSM who had at least two HIV tests were included, and classified in two groups: 1182 not taking PEP (not-PEP MSM) and 142 taking PEP (PEP MSM) who received 157 PEP courses (nine had two, and three had three courses). In a subanalysis, MSM who had two negative HIV tests performed less than 3 months apart were excluded in order to have a higher confidence with a completed window period of seroconversion.
Incidence was calculated by dividing newly diagnosed HIV infections by total person-years under observation and presented as incidence rate (including 95% confidence intervals, CI) per 100 person-years. To compare HIV incidence between cohorts, incidence rate ratios (IRR) including 95% CI were computed using Poisson log-linear regression. All analyses were performed using the statistical packages SPSS, version 19.0 (SPSS, Inc., Chicago Illinois, USA) and STATA, version 11 (StataCorp LP, College Station Texas, USA).
For participants who remained HIV negative, person-years were calculated from the first HIV test to the last available one; for PEP MSM, a subanalysis considered as time-at-risk only the post-PEP period.
For seroconverters, person-years were calculated from the first HIV test until the seroconversion date.
This was calculated as the midpoint between the last HIV-negative and the HIV-positive result for those with a maximum of 1 year between the tests; in those with more than 1 year between the tests and an Avidity Index less than 0.80, the seroconversion date was estimated subtracting 180 days from the date of the positive test; finally, in the remaining cases, the seroconversion date was estimated according to the CASCADE collaboration on HIV seroconverters which assumed that infection occurred 1.19, 4.19 and 7.94 years before, respectively, for people with a CD4 cell count less than 500, less than 350 and less than 200 cells/μl at the time of the first HIV positive test .
A total of 102 seroconversions occurred, without a significant difference in incidence between the two groups. Although the seroconversion date was estimated by three different methods, using the actual positive test date increased negligibly the total person-years amount, without influencing the IRR.
Of the seven PEP MSM who seroconverted, six tested negative more than 6 months after PEP; the remaining one tested positive at 3 months; although he reported ongoing risk behaviour, failure cannot be completely ruled out.
Considering the total time at risk, we did not observe a higher HIV incidence in PEP MSM, differently from Heuker et al. . However, considering only the post-PEP period, the incidence among PEP MSM showed an increase, although it did not reach statistical significance probably because of the small numbers. Whether this is due to risk compensation cannot be completely elicited. In the literature, the studies that have evaluated whether the use of PEP has any effect on risk behaviours for HIV infection demonstrated conflicting results. Further studies are needed before assuming PEP per se as an appropriate indication for pre-exposure prophylaxis provision. Other risk factors should be investigated as potential identifiers of individuals who can benefit of pre-exposure prophylaxis provision.
Supported by Italian Ministry of Health AIDS Project 40H32 and Ricerca Corrente IRCCS.
Authors thank Nicoletta Orchi, Francesco Maria Fusco and Paola Scognamiglio who contributed to data collection, as well as Alessandro Agresta and Claudia Cimaglia for data analysis.
Conflicts of interest
There are no conflicts of interest.
1. Heuker J, Sonder GJB, Stolte I, Geskus R, van den Hoek A. High HIV incidence among MSM prescribed postexposure prophylaxis, 2000–2009: indications for ongoing sexual risk behavior
2. Alteri C, Svicher V, Gori C, D’Arrigo R, Ciccozzi M, Ceccherini-Silberstein F, et al. on behalf of SENDIH Study GroupCharacterization of the patterns of drug-resistance mutations in newly diagnosed HIV-1 infected patients naïve to the antiretroviral drugs
. BMC Infect Dis
3. Lodi S, Phillips A, Touloumi G, Geskus R, Meyer L, Thièbaut R, et al. Proportion of individuals likely to need treatment for CD4 thresholds <200, <350, and <500 cells
. Clin Infect Dis