Very Low Incidence of HIV-infection and Decreasing Incidence of STI Among PrEP Users in 2020 After Introduction of Health Insurance Coverage in Germany

Daniel Schmidt (  schmidtd@rki.de ) Robert Koch Institute Christian Kollan Robert Koch Institute Barbara Bartmeyer Robert Koch Institute Viviane Bremer Robert Koch Institute Tim Schikowski Robert Koch Institute Martin Friebe Robert Koch Institute Sven Schellberg Novopraxis Stefan Scholten Praxis Hohenstaufenring Markus Bickel Infektiologikum Frankfurt Nikola Hanhoff German association of physicians in HIV-care (dagnä e.V.) Robin Rüsenberg German association of physicians in HIV-care (dagnä e.V.) Knud Schewe German association of physicians in HIV-care (dagnä e.V.)


Introduction
Infection with the human immunode ciency virus (HIV) remains globally endemic with 1,5 million new infections annually despite continued advancements in its management, treatment, and prophylaxis (1). In Germany, HIV transmission remains an issue despite prophylactic measures, and a knowledgeable men who have sex with men (MSM) population which is the largest affected group (2). Whereas prevention historically focused on condom use and safe injection practices, the approval of the antiviral agents tenofovir disoproxil fumarate and emtricitabine (TDF/FTC) for HIV pre-exposure prophylaxis (PrEP) has added a of their PrEP using population in reference to that of all participating centers. Only persons with PrEP start prior to April 1st, 2020, could be recruited into the study.
Data were collected anonymously using an electronic reporting form in the rst quarter of 2021 retrospectively for the time-period of September 2019 through December 2020. PrEP start prior to September 2019 was also documented. Data included patient characteristics (age, gender), PrEP use (indication, starting date, duration, continuous vs. on-demand use as well as any interruptions or discontinuations and reasons), testing for and potential diagnosis of HIV, hepatitis and STI, as well as hepatitis antibody status. Aside from HIV-infection, STI included Chlamydia, Gonorrhea and Syphilis. Viral Hepatitis included A, B and C. The study duration was divided into ve time intervals: one September through December 2019, then four quarters in 2020. STI were documented for each time interval, HIV and hepatitis with a date of diagnosis.
In addition to recording PrEP use at the individual level, a separate study center survey provided information on PrEP use and delivery, including the total number of PrEP users and HIV-positive patients in care.
In accordance with other studies STI diagnosis within 90 days of PrEP initiation was de ned as baseline prevalence (11). Incidence rates were determined for STIs occurring while on PrEP and calculated by dividing the number of STI cases by the duration of PrEP exposure per time interval per 100 person years (py).
For incidence rates of infections with hepatitis and HIV, the duration of PrEP exposure in the observation period to the date of diagnosis was calculated as person-time. More than one positive test for a certain STI within one time interval was counted as one result. It was prede ned that PrEP was only considered interrupted or discontinued if stopped for longer than four weeks. PrEP pill coverage was calculated using the number of days prescribed divided by the number of days on PrEP.
Univariable and multivariable mixed effects logistic regression was used to analyze risk factors associated with infections with Chlamydia/Gonorrhea and Syphilis. The analysis was performed among all MSM with/without other risk factors being tested in the respective time interval adjusted for age, PrEP use (on-demand/daily), PrEP interruptions, PrEP discontinuations, PrEP duration, number of PrEP intervals under observation, PrEP start before statutory health insurance coverage, infections with Hepatitis or HIV, other STI than the one analyzed, STI history in the past six months before PrEP start, the number of STI tests, observation within or before the covid-19 pandemic period.
Data were collected and prepared using a MS-SQL server. Statistical analysis was performed using STATA Version 17.

Results
Forty-seven German HIV-specialty practices and ambulatory care centers participated in the study and provided representation of PrEP use as a whole from both a geographical and a diversity perspective ( gure 1). 4620 PrEP users were recruited into this study. Nearly all (99.2%) were male, median age was 38 years (IQR 32-45), PrEP indication was MSM for 98.6% of whom 10.6% had additional risk factors. Of the 39 PrEP users who were not male, 17 were female, 16 transgender and four non-binary gender. Characteristics are shown in Table 1. The majority of cases (n=1638, 35.5%) were documented in Berlin, which also re ects the pre-dominant location of current PrEP use (4775 PrEP users of the 22 366 cared for in all study centers).  Twelve percent (n=555) did not have any event related to PrEP use (testing, diagnosis, treatment of STI, HIV, Hepatitis, or PrEP prescriptions) for at least two time intervals before their last event, 84.5% (n=469) of whom did not indicate PrEP discontinuation or interruption.

HIV
During the observation period, four persons became diagnosed with HIV-infection while on PrEP (MSM, age 26-33), accounting for 0.087% and an incidence rate of 0.078/100py (95% CI 0.029-0.208). HIV-infections were diagnosed on day 32, 167, 270 and 295 after start of study observation, which corresponds to day 32, 167, 595 and 598 after PrEP start. For two individuals PrEP on demand was documented, for the other two continuous PrEP. The reason for infection in two cases was reported to be suboptimal adherence. In the third case, calculated PrEP pill coverage was 0.6 and resistance to FTC was observed. The fourth person, diagnosed on day 32 after PrEP initiation, reported condomless rectal intercourse a few days prior to PrEP start.
Three of these patients were also diagnosed with other STIs. Two of the individuals had presented with a Syphilis infection in the previous six months before PrEP start and three had other STIs during the observation period (Table 2).  Table 3 and were highest for all three infections in the rst time interval (September-December 2019). While the lowest incidence rates were seen in the second quarter of 2020 for both Chlamydia and Gonorrhea, it was lowest for Syphilis in the third quarter. A rise in the incidence of Gonorrhea and Syphilis was again seen towards the fourth quarter of 2020 despite fewer tests. Overall, a signi cant reduction of 34.8% was found for Chlamydia and of 26.1% for Gonorrhea over the course of the observation period, as well as a reduction of 38.4% for Syphilis from the rst time interval through the third quarter of 2020. The total incidence of any STI showed a signi cant reduction of 29.6% over the course of the observation period. in the second and fourth quarter of 2020 (n=2186 and n=2171, respectively) and for Syphilis in the fourth quarter of 2020 (n=3183).
The positivity rate remained relatively constant with a mean of 9.2%, 9.9% and 3.0% for Chlamydia, Gonorrhea and Syphilis respectively (Fig. 2). While 65.5% of Syphilis diagnosis were detected by screening of asymptomatic individuals, these were 55.6% for Chlamydia and 50.1% for Gonorrhea. During the period of observation, two infections with hepatitis B were diagnosed on days 197 and 229 after start of observation, accounting for 0.043% and an incidence rate of 0.04/100py (95% CI 0.01-0.16). Thirteen cases of hepatitis C were found at a median of 267 days (IQR 137-353) after start of observation in the study, and accounting for 0.28%, a positivity rate of 0.30% (4273 tests) and an incidence rate of 0.25/100py (95% CI 0.15-0.44).
Additionally, one case of each, hepatitis A and B infection, and hepatitis C were diagnosed prior to or at the start of observation in the study (on days -7, 0, and -32 respectively).

Discussion
The signi cantly higher risk of infection with Chlamydia/Gonorrhea for MSM aged 16 to 29 years and the signi cantly lower risk for MSM aged 40 to 59 years when compared to 30 to 39 years may re ect higher sexual activity and more partners in younger age groups. Unexpectedly, age 50 to 59 years was associated with an increased risk for syphilis. This nding is not supported by the national mandatory syphilis reports, which show the highest incidence for men between 30 to 39 years of age (31). Further analysis in terms of study center location indicated that there may be an interrelated infection pattern that possibly explains this discrepancy.
A higher risk for Chlamydia/Gonorrhea was found in MSM with PrEP start before its coverage by health care in September 2019. It likely re ects these individuals' involvement in the sexually active MSM community. In addition, this indicates persistent HIV risk behavior, which goes along with infection risks of other STIs. The lower risk for Chlamydia/Gonorrhea seen with on-demand PrEP use, on the other hand, may in turn re ect less frequent high-risk sexual contacts.
As expected, symptom triggered testing and a history of STI was associated with a higher risk for infections with Chlamydia/Gonorrhea and Syphilis. The latter supports the importance of a comprehensive anamnestic before PrEP initiation.
We found a high baseline protection against hepatitis A and B in approximately three-quarters of the cohort. Of note, immunity against hepatitis A may be in part self-reported and the actual percentage therefore lower. Antibodies against hepatitis B on the other hand should routinely be determined prior to commencing PrEP and every six months while on PrEP due to the interaction between tenofovir and hepatitis B (32). Accordingly, no infection with hepatitis A and only two with hepatitis B were diagnosed within the course of the study. However, thirteen cases of hepatitis C occurred, re ecting an ongoing epidemic of sexually transmitted hepatitis C among MSM, which in the past was almost exclusively seen among HIV infected MSM (33). Additionally, at screening before PrEP initiation one case of each hepatitis A, B, and C were detected underlining the importance of a medically guided PrEP start.
A limitation of this study is its evaluation period of sixteen months. Also, the SARS-CoV-2 pandemic drastically affected social interaction and probably in uenced the outcomes found in this study. Further investigation in a regular non-pandemic setting with a longer follow-up period would be valuable. Partner numbers or sexual behavior were not documented. Our population may not represent all PrEP-users in Germany, nor all PrEP prescribing centers. However, since PrEP care is provided mostly by HIV-specialty practices in Germany, we assume that we were able to obtain a representative sample of PrEP users and study centers. With limiting the number of PrEP users that could be documented within each center, bias in patient selection may have taken place; a randomized approach was encouraged to avoid selection bias.

Conclusions
In Germany HIV-PrEP with TDF/FTC is almost exclusively taken by MSM. We found a very low incidence of HIV-infection in this cohort of PrEP users. The main reason for PrEP failure was suboptimal adherence. We did not see an increase in infections with Chlamydia, Gonorrhea and Syphilis; in fact, we saw a partial decrease in their incidence rates, very likely in part attributable to the SARS-CoV-2 pandemic. Our results support the coverage of  Availability of data and materials The datasets generated and/or analyzed during the current study are not publicly available due to data protection and con dentiality but are available from the corresponding author on reasonable request.
Competing interests DS, MF, BB, VB, TS, MB, RR, KS, and NH declare that they have no competing interests. CK is small shareholder in companies manufacturing antiretroviral drugs. SS has received honoraria for lectures, conference sponsorship and participation on advisory boards from Abbvie, Gilead, Janssen, GSK, MSD, ViiV Healthcare, Theratechnologies. SVS has received honoraria and study compensation from companies developing and marketing antiretroviral drugs and is shareholder of various pharmaceutical companies including ones developing and marketing antiretroviral drugs.

Funding
The PrEP evaluation (EvE-PrEP) was funded by the Federal Ministry of Health, Germany. We acknowledge the nancial support of the German Federal Ministry of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Authors' contributions DS was responsible for the study design, performed the data analysis and statistical analysis, interpreted the data and drafted the manuscript. CK contributed to the conception of the study and interpretation of the data and was responsible for the data management. KS, RR and NH contributed to the conception and coordination of the study and interpretation of the data and to draft the manuscript. MF and TS supported the management and coordination of the study and contributed to improving the data quality and coverage. SS, SVS, MB are site principal investigators, contributed data and supported the interpretation of the data. VB and BB supported the management and coordination of the study and the overall study design and analysis approach. All authors participated in the critical discussion of the results, and all read and approved the nal manuscript. Primary reasons for discontinuing or interrupting PrEP