HIV postexposure prophylaxis (PEP) is a 4-week combination antiretroviral treatment recommended to persons exposed to human body fluids possibly infected with HIV to prevent HIV acquisition . The effectiveness of PEP to reduce the likelihood of HIV infection after high-risk exposure is estimated at 79–81% [2,3]. The majority of requests for nonoccupational PEP come from sexual exposure of MSM [4–6].
The annual number of PEP prescriptions to MSM in Amsterdam, the Netherlands has increased three-fold in the period 2000–2007 . A similar increase in PEP prescriptions was found in an Australian cohort of HIV-negative MSM . Knowledge of PEP in the Netherlands increased among MSM living in urban areas from 46% of men who were familiar with or had ever heard about PEP in 2006 to 73% in 2009 [9,10]. However, only a small minority reported to have ever used PEP [9,10].
A study conducted in the USA identified MSM who ever used PEP to be more likely to report high-risk sex compared with non-PEP users . Also, MSM in the Australian cohort who used PEP were found to be at continued risk of HIV infection . This could indicate that the rise in annual prescriptions of PEP in Amsterdam might have come, in large part, from a distinct high-risk group of MSM. If high-risk sexual behaviour after a PEP course is indeed continued by a select high-risk group of MSM, we would expect to find a relatively high incidence of HIV among this group. The goals of the present study were to determine trends in HIV incidence among MSM PEP users compared with that of the Amsterdam Cohort Studies (ACS) and to identify characteristics of MSM who were prescribed PEP over time.
Participants and procedures
MSM prescribed postexposure prophylaxis
We used data of MSM in Amsterdam who were prescribed a 28-day PEP course after a sexual-risk accident between 2000 and 2009. PEP prescription was done following European guidelines . The distribution of PEP in Amsterdam was centralized at the Public Health Service (PHS) Amsterdam in 2000. HIV-PEP starter kits are prescribed at the Department of Infectious Diseases of the PHS Amsterdam and at five hospital emergency departments. Those who initially received their kit at hospitals were referred to the PHS Amsterdam for follow-up . Further description of drug regimens used over time can be found in a study of Sonder et al..
Men were tested for infection with HIV at the start of their PEP course (baseline) and 3 and 6 months after PEP prescription. For the current study, we included MSM who tested HIV-negative at baseline and had follow-up HIV testing 3 and/or 6 months after PEP prescription.
From April 2000 to December 2009, a total of 506 PEP courses were prescribed to MSM in Amsterdam. Forty-four men were advised by physicians to stop their PEP course for various reasons: their sex partner tested HIV-negative (n = 27); the indication was reconsidered as low risk (n = 7); or they tested HIV-positive at baseline (n = 10). Another 28 had no indication for follow-up visits at the PHS Amsterdam, because they were nonresidents of the Netherlands, mainly tourists, who continued their treatment in their home country.
Of the 433 PEP prescriptions with an indication for follow-up after PEP at the PHS Amsterdam, 9.2% (n = 38) did not attend to appointments after baseline and could not be traced, resulting in a study population of 395 PEP prescriptions
Although the majority of MSM (87.6%, n = 346) had their follow-up HIV tests at the Department of Infectious Diseases of the PHS Amsterdam, a number of men chose to test elsewhere for various reasons. In order to be additionally tested for other sexually transmitted infections (STIs), 6.6% (n = 26) tested at the STI Outpatient Clinic of the PHS Amsterdam; some men were also participating in the ACS and preferred to have their HIV test at their routine ACS visit (2.3%, n = 9), and some tested at other places (e.g. general practitioner) (2.3%, n = 9). These men gave consent to obtain their test results. For five men, the follow-up HIV test was self-reported (all were HIV-negative). Although we were not able to confirm these results, we did use these results in our analysis (1.3%).
MSM in the Amsterdam Cohort Studies
The comparison cohort consisted of MSM participating in the ACS. The ACS (www.amsterdamcohortstudies.org) is an open, prospective cohort of homosexual men living in the region of Amsterdam that was started in 1984 to investigate the epidemiology, pathogenesis, and prevention of HIV/AIDS. Men visit the PHS twice a year to complete a questionnaire on self-reported sexual behaviour and to test for HIV . For the current study, we included MSM who participated in the ACS between 2000 and 2009. The ACS is considered suitable as a reference cohort for estimating the HIV incidence among MSM who lived in the same region and in the same period as MSM who were prescribed PEP.
For MSM prescribed PEP, HIV testing was done at the Department of Infectious Diseases and the STI clinic by HIV antibody detection in serum using AxSym HIV1/2 G0 RGT (Abbott Laboratories, Abbott Park, Illinois, USA). Since March 2007, combined HIV antibody and HIV antigen detection was done using AxSym HIV COMBO RGT (Abbott Laboratories). Reactive samples were confirmed by immunoblot (Line Immuno Assay and Inno-Lia I/II Score; Innogenetics NV, Ghent, Belgium). For MSM participating in the ACS, similar HIV-testing procedures were used with antibody detection of consecutive generations of screening assays (obtained from Abbott Laboratories; Organon International, Oss, the Netherlands; and Bio Mérieux, Lyon, France), confirmed by western blot (Genelabs Diagnostics, Singapore, Singapore).
Variables and statistical analyses
For the present study, baseline characteristics, age, and country of birth (categorized into ‘the Netherlands’, ‘other western’, or ‘nonwestern country’, according to classifications of the Dutch central statistics office ) were collected using standardized questionnaires in both settings. First, the baseline characteristics of MSM prescribed PEP and MSM participating in the ACS were described and differences were tested using Student's t-tests for continuous variables and χ2-tests for independence for categorical variables.
The observed HIV incidence was calculated by dividing newly diagnosed HIV infections by total person-years under observation and presented per 100 person-years. The date of HIV seroconversion was defined as the midpoint between the last HIV-negative and first HIV-positive result. Person-time was calculated as time from study entry until the date of seroconversion or the end of follow-up. For MSM who were prescribed PEP, the maximum time under observation per PEP prescription was 6 months according to protocol. Any follow-up time per PEP prescription longer than 6 months, mainly from men tested outside the Department of Infectious Diseases of the PHS Amsterdam, was cut off at 6 months. For participants of the ACS, we included only those visits with a maximum of 1 year or less between visits in order to make an accurate estimation of the date of HIV infection.
To investigate trends in HIV incidence and to analyse differences in HIV incidence between both cohorts, incidence rate ratios (IRR), including 95% confidence interval (CI), were calculated using Poisson log-linear regression with ‘cohort’ (PEP versus ACS) as main risk factor and ‘calendar year’ as linear covariate, using the logarithm of person-years as offset value. In order to analyse trends in differences in HIV incidence between both cohorts over time, an interaction term between calendar year and cohort was included in all models. Socio-demographics (age at visit and country of birth) were only included in the final model if the model fit significantly increased using the likelihood ratio test.
Finally, to investigate whether the group of MSM who were prescribed PEP changed over time, differences in characteristics were analysed, comparing two equal periods (period 1: 2000–2004 versus period 2: 2005–2009) using t-tests for equality of means and χ2-tests. Characteristics also included specific questions related to the sexual accident PEP was prescribed for, such as type of sexual contact, HIV status of the sex partner (known HIV-positive or unknown HIV serostatus), and compliance and adverse effects due to PEP course. The type of sexual contact PEP was prescribed for included receptive unprotected anal intercourse (RUAI), insertive UAI, receptive oral sex with ejaculation, or other sexual contact containing blood and/or sperm. Because the per-contact probability of HIV transmission in MSM is highest when practising RUAI with a known HIV-positive sex partner [15,16], we created two variables related to the sex accident PEP was prescribed for: ‘type of sexual contact’ (RUAI versus non-RUAI) and ‘serostatus sex partner’ (known HIV-positive versus unknown HIV serostatus).
All analyses were performed using the statistical packages SPSS version 17.0 (SPSS Inc., Chicago, Illinois, USA) and R version 2.13.0 (R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/).
Our study population of 395 PEP prescriptions to MSM between April 2000 and December 2009 came from 355 different individuals; 321 MSM had one PEP prescription and 34 men had multiple (two to four) PEP prescriptions. Overall, in 60.8% (n = 240) PEP was prescribed for RUAI and in 36.7% (n = 145) PEP prescriptions were done for a sex accident with a known HIV-positive sex partner. Although the majority reported adverse effects (86.3%, n = 341), the vast majority completed their PEP course (94.1%, n = 367).
In the ACS, 782 MSM with regular visits participated between 2000 and 2009, with a total of 8434 visits.
The mean age at entry of MSM who were prescribed PEP was 35.9 years (SD ± 8.5), and 62.1% (n = 244) were born in the Netherlands, 22.9% (n = 90) in another western country, and 15% (n = 59) in a nonwestern country. MSM participating in the ACS were younger at baseline [mean age 30.1 years (SD ± 8.2)] compared with MSM who were prescribed PEP (P < 0.01). Furthermore, those in the ACS group were more often born in the Netherlands (86.5%) compared with MSM who were prescribed PEP and less often born in another western or nonwestern country (respectively, 6.8 and 6.7%) (P < 0.01).
HIV incidence of MSM prescribed postexposure prophylaxis and MSM in the Amsterdam Cohort Studies
Total time under observation for MSM who were prescribed PEP was 171 person-years which increased from 4.9 person-years in 2000 to 37.1 person-years in 2009 (Fig. 1). Of the 11 MSM who seroconverted for HIV during follow-up, two had an HIV-positive test 3 months after PEP prescription, one had no HIV test at 3 months and tested HIV-positive at 6 months, and eight were HIV-negative at 3 months and HIV-positive 6 months after PEP prescription. This resulted in an overall HIV incidence between 2000 and 2009 of 6.4 per 100 person-years (95% CI 3.1–11.6). In the same period, total time under observation among MSM in the ACS was 4061 person-years. It increased from 362 person-years in 2000 to 423 person-years in 2009 (Fig. 1). In the ACS, 67 MSM with regular cohort visits seroconverted which equalled an overall HIV incidence of 1.6 (95% CI 1.3–2.1) per 100 person-years. Univariate poisson analysis comparing the HIV-incidence rate among MSM who were prescribed PEP and MSM in the ACS showed an IRR of 3.9 (95% CI 2.1–7.4) (Table 1), indicating that between 2000 and 2009 MSM prescribed PEP were almost four times more likely to seroconvert for HIV compared with MSM of the ACS.
However, there was an indication that incidence rates in both cohorts were not constant over time in the study period. Inclusion of calendar year revealed that both among MSM prescribed PEP as well as among MSM in the ACS, the trend in HIV incidence increased over time [IRRper calendar year (95% CI) 1.3 (0.9–1.7) and 1.1 (1.0–1.2), respectively] (Table 1 and Fig. 1, overall P value for calendar period was 0.06). Although the interaction term between calendar year and cohort was not significant (P = 0.33), there was a suggestion that the difference in HIV incidence increased over time: differences in HIV incidence between MSM who were prescribed PEP and MSM participating in the ACS were marginal in 2000 [IRR 1.1 (95% CI 0.1–12.9)], whereas the difference in HIV incidence was much larger in 2009 [IRR 4.8 (95% CI 2.1–11.5)] (Table 1 and Fig. 1). Age and country of birth were not associated with HIV seroconversion in univariate analyses, and inclusion in multivariate analyses did not improve our model.
Change in characteristics of MSM prescribed postexposure prophylaxis
The baseline characteristics ‘age at entry’ and ‘serostatus sex partner’ among MSM prescribed PEP changed over time. In the second period of PEP prescriptions (2005–2009), men were older compared with the first period (2000–2004) [mean age (±SD): 36.4 (±8.3) versus 34.2 (±9.0), P = 0.026]. Further, in the second period, PEP was more often prescribed for a sexual accident with a known HIV-positive sex partner (41.0 versus 24.5%, P < 0.01). Other baseline characteristics did not change significantly over time (Table 2).
Our study revealed a high HIV incidence of 6.4 per 100 person-years among MSM who were prescribed PEP between 2000 and 2009 in Amsterdam, almost four times higher compared with the HIV incidence among MSM participating in the ACS in the same period. This difference was particularly evident in the later part of the study period. These results might seem to contrast with those of some previous studies. In a study of Martin et al., an HIV incidence of 1.2 per 100 person-years was found among MSM PEP users in San Francisco, equivalent to that of the total MSM community in San Francisco. In the A Randomized Clinical Trial of the Efficacy of a Behavioral Intervention to Prevent Acquisition of HIV among Men who have Sex with Men trial in six cities in the USA, the HIV incidence among MSM who used PEP equalled that of MSM non-PEP users and was estimated at 1.6 per 100 person-years . However, taking into account that the period of these studies was early 2000, our results are not substantially different. Our results showed that in the early 2000s, the HIV incidence among MSM prescribed PEP was not higher compared with that of the ACS. The more recent Australian Health in Men Study, in which HIV-negative MSM were followed until the end of 2007, concluded that MSM who previously received PEP had an almost three-fold higher rate of HIV seroconversion compared with those who had not used PEP . This is in agreement with our results that, particularly in more recent years, MSM prescribed PEP are at continued risk for HIV seroconversion. Moreover, apart from a substantially higher HIV incidence among PEP users in more recent years, another 10 men who requested PEP unexpectedly tested HIV-positive at baseline, all after 2005, which also indicates that over time an increasing group of MSM with high-risk behaviour requested PEP. This also indicates that there is still a remarkable number of MSM who are not aware of their positive HIV status and who may be highly infectious and contributing to HIV transmission .
Although we cannot completely rule out PEP failure in our study, we assume PEP failure is not likely. First, our current study revealed that the majority of seroconverters tested HIV-negative 3 months after PEP prescription, indicating their HIV acquisition was most likely not related to the initial sexual contact for which PEP was prescribed. Also the study of Martin et al. showed that PEP failure was not likely, as MSM who seroconverted in this study all tested negative at the first HIV test after PEP prescription. Second, in a previous study of Sonder et al. none of the virus strains of the first five HIV seroconverters in this study group between 2000 and 2007 were resistant to PEP nor were the virus strains of the next six seroconverters in this study. Third, in this same study, in the (only) available sample of the source, the virus of the source did not appear to match the virus isolated from the patient . Fourth, the majority of the HIV seroconverters in the current study reported to the staff of the PHS that they had ongoing risk behaviour after PEP prescription.
As we found an increasing trend in HIV incidence, particularly among MSM PEP users, we analysed whether baseline characteristics of MSM prescribed PEP changed over time. We found that in more recent years PEP is more often prescribed to men who had a sexual accident with a known HIV-positive sex partner. However, we cannot conclude that the actual percentage of HIV-positive sex partners increased over time. Due to changing HIV testing policies in the Netherlands, more MSM nowadays are aware of their HIV status . In addition, we have to be somewhat vigilant interpreting these self-reported results because MSM peer groups were known to inform MSM that it was more likely to be prescribed PEP if you told the physician you had a sex accident with a known HIV-positive sex partner rather than with a sex partner with unknown HIV serostatus.
The strength of our study is that we were able to make a reliable estimate of the HIV incidence among MSM who had been prescribed PEP in the region of Amsterdam. First, because the PHS Amsterdam has data of all PEP users since 2000 in Amsterdam and, second, because the majority of eligible PEP users had follow-up HIV testing data at 3 and 6 months after PEP prescriptions. This relatively high proportion of men with follow-up HIV tests is partly due to the organization of the PHS, in which not only the Department of Infectious Diseases is accommodated but also the STI outpatient clinic and the ACS. However, this could also have the drawback of overlap between men who were prescribed PEP and men participating at the ACS. Although we are not aware of the exact number of men who are participating in both cohorts, we know that nine men who were prescribed PEP had their follow-up HIV tests at the ACS and tested HIV-negative. Therefore, although the overlap seems to be limited, the ACS could not be interpreted as a cohort of MSM not having used PEP, but merely as a sample of MSM that occasionally uses PEP. Another limitation of this study is that we have to be cautious interpreting the ACS as a reflection of the overall MSM community in Amsterdam. MSM of Dutch descent appear to be overrepresented in the ACS and up to 2003, because of inclusion criteria, only young MSM (with a maximum age of 35) participated in the ACS. However, the socio-demographic variables of country of birth and age were not associated with HIV seroconversion in either cohort.
In conclusion, our study showed a high incidence of HIV among MSM who used PEP, an indication of ongoing sexual risk behaviour. This implies that PEP alone for this group is not sufficient to prevent HIV infection, and a combination of other more comprehensive preventive strategies is needed, as is shown in a study of Martin et al. in which 77% of MSM who were prescribed PEP and received counselling reported a significant decrease in risk behaviour 6 months after PEP prescription. Further, recent outcomes of a phase III trial indicated that oral pre-exposure prophylaxis (PrEP) used as part of a more comprehensive prevention strategy to be safe and effective in reducing HIV infection in MSM . However, an optimal regimen has not been identified , implying further research is needed. Assuming that conducting effective trials requires populations with a relatively high HIV incidence [21,22], MSM PEP users may well be suitable participants for such future PrEP intervention trials.
The authors thank R.M. Regez, K. Brinkman, J.M. Prins, J.W. Mulder, J. Veenstra, and F.A.P. Claessen of participating hospitals for their collabaration, and L. Bovee and M. van Rooijen for assistence with data management.
J.H. analysed and interpreted the data and wrote the draft manuscript; G.S. and A.V.D.H. designed and supervised the overall study and supervised data collection of PEP prescriptions; I.S. assisted with statistical analyses and supervised data collection of the ACS; and R.G. contributed to the analysis and design. All authors contributed to subsequent drafts and approved the final manuscript.
Conflicts of interest
The authors have no conflicts of interest.
1. Almeda J, Casabona J, Simon B, Gerard M, Rey D, Puro V, et al. Proposed recommendations for the management of HIV postexposure prophylaxis after sexual, injecting drug or other exposures in Europe. Euro Surveill 2004; 9:35–40.
2. Cardo DM, Culver DH, Ciesielski CA, Srivastava PU, Marcus R, Abiteboul D, et al. A case-control study of HIV seroconversion in healthcare workers after percutaneous exposure. Centers for Disease Control and Prevention Needlestick Surveillance Group. N Engl J Med 1997; 337:1485–1490.
3. Schechter M, do Lago RF, Mendelsohn AB, Moreira RI, Moulton LH, Harrison LH. Behavioral impact, acceptability, and HIV incidence among homosexual men with access to postexposure chemoprophylaxis for HIV. J Acquir Immune Defic Syndr 2004; 35:519–525.
4. Sonder GJ, van den Hoek A, Regez RM, Brinkman K, Prins JM, Mulder JW, et al. Trends in HIV postexposure prophylaxis prescription and compliance after sexual exposure in Amsterdam, 2000–2004. Sex Transm Dis 2007; 34:288–293.
5. Poynten IM, Smith DE, Cooper DA, Kaldor JM, Grulich AE. The public health impact of widespread availability of nonoccupational postexposure prophylaxis against HIV. HIV Med 2007; 8:374–381.
6. Kahn JO, Martin JN, Roland ME, Bamberger JD, Chesney M, Chambers D, et al. Feasibility of postexposure prophylaxis (PEP) against human immunodeficiency virus infection after sexual or injection drug use exposure: the San Francisco PEP Study. J Infect Dis 2001; 183:707–714.
7. Sonder GJ, Prins JM, Regez RM, Brinkman K, Mulder JW, Veenstra J, et al. Comparison of two HIV postexposure prophylaxis regimens among men who have sex with men in Amsterdam: adverse effects do not influence compliance. Sex Transm Dis 2010; 37:681–686.
8. Poynten IM, Jin F, Mao L, Prestage GP, Kippax SC, Kaldor JM, et al. Nonoccupational postexposure prophylaxis, subsequent risk behaviour and HIV incidence in a cohort of Australian homosexual men. AIDS 2009; 23:1119–1126.
9. Hospers HJ, Dörfler TT, Zuilhof W. Schorer Monitor 2006 [in Dutch]. Schorerstichting. 2006.
10. Hospers HJ, Roos E, Zuilhof W. Schorer Monitor 2009 [in Dutch]. Schorerstichting. 2009.
11. Donnell D, Mimiaga MJ, Mayer K, Chesney M, Koblin B, Coates T. Use of nonoccupational postexposure prophylaxis does not lead to an increase in high risk sex behaviors in men who have sex with men participating in the EXPLORE trial. AIDS Behav 2010; 14:1182–1189.
12. Sonder GJ, Regez RM, Brinkman K, Prins JM, Mulder JW, Spaargaren J, et al. Prophylaxis and follow-up after possible exposure to HIV, hepatitis B virus, and hepatitis C virus outside hospital: evaluation of policy 2000–3. BMJ 2005; 330:825–829.
13. van Griensven GJ, de Vroome EM, Goudsmit J, Coutinho RA. Changes in sexual behaviour and the fall in incidence of HIV infection among homosexual men. BMJ 1989; 298:218–221.
15. Jin F, Jansson J, Law M, Prestage GP, Zablotska I, Imrie JC, et al. Per-contact probability of HIV transmission in homosexual men in Sydney in the era of HAART. AIDS 2010; 24:907–913.
16. Wilson DP, Jin F, Jansson J, Zablotska I, Grulich AE. Infectiousness of HIV-infected men who have sex with men in the era of highly active antiretroviral therapy. AIDS 2010; 24:2420–2421.
17. Martin JN, Roland ME, Neilands TB, Krone MR, Bamberger JD, Kohn RP, et al. Use of postexposure prophylaxis against HIV infection following sexual exposure does not lead to increases in high-risk behavior. AIDS 2004; 18:787–792.
18. Williamson LM, Dodds JP, Mercey DE, Hart GJ, Johnson AM. Sexual risk behaviour and knowledge of HIV status among community samples of gay men in the UK. AIDS 2008; 22:1063–1070.
19. Heijman RL, Stolte IG, Thiesbrummel HF, van Leent E, Coutinho RA, Fennema JS, Prins M. Opting out increases HIV testing in a large sexually transmitted infections outpatient clinic. Sex Transm Infect 2009; 85:249–255.
20. Grant RM, Lama JR, Anderson PL, McMahan V, Liu AY, Vargas L, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med 2010; 363:2587–2599.
21. Poynten IM, Jin F, Prestage GP, Kaldor JM, Kippax S, Grulich AE. Defining high HIV incidence subgroups of Australian homosexual men: implications for conducting HIV prevention trials in low HIV prevalence settings. HIV Med 2010; 11:635–641.
22. Seage GR III, Holte SE, Metzger D, Koblin BA, Gross M, Celum C, et al. Are US populations appropriate for trials of human immunodeficiency virus vaccine? The HIVNET Vaccine Preparedness Study. Am J Epidemiol 2001; 153:619–627.
© 2012 Lippincott Williams & Wilkins, Inc.