Abstract
Background
Doxycycline-based prevention of bacterial sexually transmitted infections (STIs) has been assessed in various studies and has been recommended by the European AIDS Clinical Society to be proposed to persons with repeated STIs on a case-by-case basis. However, while good preventive effects could be shown for Chlamydia trachomatis and Treponema pallidum in Europe, no reliable prevention against doxycycline resistance-affected bacterial causes of STIs like Neisseria gonorrhoeae and Mycoplasma genitalium was confirmed.
Methods
In a modelling-approach, we assessed potential beneficial effects even against the latter microorganisms in case of optimized adherence with doxycycline prevention. These effects were modelled for Germany in comparison to traditional prevention schemes like condom-based STI-prevention and testing-as-prevention.
Results
With estimated risk reduction in the ranges of 86% for N. gonorrhoeae and of 82% for Mycoplasma genitalium, expectable preventive efficacy similar to alternative preventive approaches could be calculated in case of optimized adherence with doxycycline prevention. In case of repeated risk exposure, the preventive potential of condom-based prevention was decreased compared to both optimized doxycycline prevention and testing-as-prevention.
Conclusions
As suggested by the applied modelling, the preventive effect of optimized doxycycline prevention against bacterial STIs is in a similar range, like other common prevention strategies.
Introduction
As a consequence of convincing results in well-designed studies like IPERGAY [1] and PROUD [2] from the middle of the recent decade, pre-exposure prophylaxis (PrEP) against sexual HIV (human immunodeficiency virus) infection has become a well-accepted preventive strategy. Nearly at the same time, another preventive approach was tested which, however, was even more controversially discussed. Published in 2015, the group by Bolan and colleagues had offered prophylactic use of doxycycline 100 mg once daily to prevent the acquisition of bacterial sexually transmitted infections (STI) by HIV-infected men-having-sex-with-men (MSM) still engaged in high-risk sexual activity. Non-significant risk reduction with odds ratios (95% confidence intervals (CI)) of 0.27 (0.04–1.73) for syphilis only, 0.42 (0.09–1.89) for syphilis or gonorrhea only as well as 0.30 (0.08–1.09) for any combination of gonorrhea, syphilis and/or chlamydial infections were recorded at the on-drug assessment at study week 36. At the follow-up assessment at study week 48, a weakly significant risk reduction was recorded for any combination of gonorrhea, syphilis and/or chlamydial infections with an odds ratio of 0.27 (0.09–0.83), while no significance was observed for syphilis alone and syphilis or gonorrhea with odds ratios of 0.24 (0.04–1.33) and 0.36 (0.08–1.56), respectively. A moderate adherence by at least a majority of study participants taking their drug was recorded in this study [3]. Not discouraged by these slightly disappointing results, the group by Molina and colleagues assessed the effects of doxycycline-based post-exposure prophylaxis (PEP) within the abovementioned IPERGAY study group. The PEP approach consisted of a 200 mg single dose doxycycline application within 24 h after a sexual risk contact. The approach was associated with a 47% risk reduction of acquiring any combination of gonorrhea, syphilis and/or chlamydial infections. Pathogen-specific risk reduction was 70% for chlamydial infections and 73% for syphilis while no measurable reduction was observed for gonorrhea. With an expected doxycycline resistance of 23.5% for Neisseria gonorrhoeae within the study setting, antimicrobial resistance alone might only partly explain the discrepancies. For syphilis and chlamydial infections, no antimicrobial resistance issues were expected. Of note, the study design with only risk-adapted drug intake did not allow for any reliable assessment of drug adherence [4].
Published in the first half of 2023, an American open-label randomized study on doxycycline PEP for the prevention of bacterial sexually transmitted infections confirmed a doxycycline-based risk reduction by factor three for users of HIV-PrEP and people-living-with-HIV (PLWH) regarding their risk of acquiring bacterial STIs [5]. In this assessment, individuals at risk in the study group were advised to take 200 mg doxycycline within 72 h after unprotected sexual exposure, an approach which was compared to standard care in the control group without doxycycline [5]. The 72 h interval was based upon the results by Zheng and colleagues [6], indicating that therapeutically active doxycycline concentrations can be expected for 72 h on human mucous membranes. In the American study [5], there were no severe adverse events attributed to doxycycline. Further, with 5 out of 13 N. gonorrhoeae isolates showing doxycycline resistance in the study group, 2 out of 16 resistant N. gonorrhoeae isolates in the control group and 4 out of 15 resistant N. gonorrhoeae isolates at baseline, resistance selection could not be statistically significantly demonstrated considering the low overall numbers of infections, while potential microbiome effects were not assessed [5]. Of note, in animal experiments, the latter were shown to be potentially relevant, associated with likely increased susceptibility to gastrointestinal infections [7]. Based on the available evidence, the European AIDS Clinical Society (EACS) recommended doxycycline PEP to “be proposed to persons with repeated STIs on a case by case basis” in the 12th version of their guidelines from October 2023 (URL: https://www.eacsociety.org/media/guidelines-12.0.pdf, last accessed on 18th December 2023). Although the EACS decision was controversially discussed by infectious disease experts, it is well documented that prophylactic doxycycline application is already a common phenomenon in European populations at high risk of acquiring bacterial STIs [8]. Studies in Sub-Saharan African high prevalence settings for bacterial STIs have been initiated as well [9, 10].
Nevertheless, and as already demanded in the early abovementioned proof-of-principle assessments [3, 4], potential undesired effects of doxycycline-based prevention like medical side effects, antimicrobial resistance selection and shifts in the microbiome composition need to be openly discussed. In combination with the strategy's observed imperfect preventive potency, initial reception of doxycycline prevention was mostly critical in the international literature and further research was demanded [8, 11]. In spite of such concerns, doxycycline-based prevention of bacterial STIs has been frequently and in a more or less uncontrolled manner applied, in particular by populations with sexual habits associated with a particularly high risk of STI acquisition [8, 12, 13]. Although resistance selection is even feared by users of doxycycline-based prevention themselves as well as by their prescribers [14], this abstract concern does not prevent a considerable interest in the individually protective potential of this preventive strategy by populations at risk in various areas of the globe [14–16].
Next to individual desire for protection in spite of sexual risk behavior, population medicine has also shown some interest in doxycycline-based prevention in the meantime. A study group from Philadelphia modeled its preventive effect on the spread of syphilis based on various levels of drug uptake and adherence with drug intake. They came to the conclusion that a modest preventive effect on population level could be achieved if reasonable adherence and drug uptake were guaranteed [17].
While the initial doxycycline PEP assessment by Molina and colleagues [4] had been focused on gonorrhea, syphilis and chlamydial infections, Berçot and colleagues also included Mycoplasma genitalium in their subgroup assessment of the IPERGAY study [18]. Similar as observed for gonococci [4], no decrease of the incidence was seen due to doxycycline PEP. Of note, the authors reported a mutation in the 16S rRNA gene of M. genitalium as potentially associated with doxycycline resistance in about 12.5% of the bacteria [4]. The representativeness of this molecular result targeting only one potential resistance mechanism is uncertain. According to the European guideline, therapeutic success in case of M. genitalium-induced infection can be expected in about 40% of the cases only [19].
The available data raise a number of intriguing questions. First, it is unclear whether and in how far optimization of drug adherence and application schemes with focus on pharmacokinetics might at least improve the so far nondetectable risk reduction for gonococcal and M. genitalium infections in spite of an unfavorable resistance situation. Second, how is the expected differential preventive effect of doxycycline-based prevention with adequate drug adherence in comparison to competitor strategies like condom- or test-based prevention in an exemplary European population? To address these questions, a modelling-based approach was chosen, primarily based on the European study results [4, 18], as detailed in the following.
Methods
Premises for the modellings
Premises for the modelling of the effects of optimized adherence
This assessment is based on the assumption that lacking adherence to doxycycline prevention and potentially also suboptimal application schemes at least partially contributed to imperfect protection rates. Although precise non-compliance rates are difficult to assess for the reasons stated above and in the original publications [3, 4], there are a number of reasons to assume that adherence issues may have been of considerable relevance. In the study by Bolan and colleagues [3], doxycycline blood levels suggestive of non-adherence even directly associated with assessment time points of the study were frequently observed. In the IPERGAY study [1], in the course of which Molina and colleagues performed their doxycycline-PEP assessment [4], measured non-adherence to HIV-PrEP with emtricitabine/tenofovir was shown to primarily account for the reduction of HIV-PrEP effectiveness to 86%. If the study participants were even willing to accept a non-compliance-associated increased risk of HIV infection, there is little reason to believe in higher compliance with PEP against bacterial STIs which can be therapeutically cured in case of occurrence. In addition, the observed protection rates in Molina's doxycycline PEP study and its subgroup analysis by Berçot and colleagues [4, 18] did not reflect expected failure rates considering the targeted pathogens' antimicrobial resistance rates within the time period of the assessment. As stated above, no doxycycline resistance in chlamydia and Treponema pallidum as well as 23.5% in gonococci were expected in Molina's study [4]. For mycoplasma, the resistance estimations were less reliable because a comprehensive molecular resistance assessment beyond individual mechanisms was not performed [18]. Next to adherence issues, suboptimal application schemes might also have contributed to poor preventive effects as discussed for gonococci [4], however, in case of optimization of both, better preventive effects should have been expected. For the modelling, it was assumed that such an optimization should allow a preventive effect of at least 95% for syphilis and chlamydial infections and based on this assumption, the proportional effect of suboptimal adherence and drug application was calculated. In spite of general optimization of drug adherence and intake, inappropriately low drug levels may still result from sporadic events like, e.g., drug interactions as well as medical or social conditions interfering with enteric resorption of doxycycline like diarrhea or the consumption of food or drinks containing relevant levels of divalent iron ions [20]. Accordingly, the expected risk reduction even in case of optimized doxycycline prevention is assumed to be less than 100% for syphilis and chlamydial infections in this modeling. Applying the abovementioned assumptions, it was calculated which effects might by achievable for gonococci and mycoplasma in case of optimization of doxycycline prevention. The detailed premises for the modelling are presented in Table 1.
Premises for the modelling of the effects of optimized adherence and application schemes, based on the study by Molina and colleagues [4]
| STI/causative agent | Observed risk reduction in case of imperfect doxycycline-based prevention [4]# | Expected resistance proportion at the time of the study [4] | Expected risk reduction in case of optimized doxycycline prevention |
| Syphilis/Treponema pallidum | 73% | 0% | ≥95 |
| Chlamydial infection/Chlamydia trachomatis | 70% | 0% | ≥95 |
| Gonorrhea/Neisseria gonorrhoeae | No risk reduction | 23.5% | To be interpolated |
| Mycoplasmal infection/Mycoplasma genitalium | No risk reduction | >12.5%* | To be interpolated |
STI = sexually transmitted infection. *Resistance of Mycoplasma genitalium could not be reliably estimated in the study, because only a single associated molecular mechanism was assessed and so, resistance higher than the recorded 12.5% has to be expected. #The dimension of imperfect drug adherence and application in the underlying study [4] is unknown and needs to be interpolated for the modelling.
Premises for the comparative modelling of different preventive strategies in a European population at risk
For the comparison with other prevention strategies, it is necessary to consider the expectable preventive effects of alternative strategies. But even without any preventive approaches, the transmission risk of bacterial STIs is different from 100% with considerable pathogen-depending peculiarities. In line with previous estimations [5, 21], the basic transmission risk per heterosexual contact ranges from 60% for syphilis and 50% for gonorrhea to a range between 20% and 35% for chlamydial infections. For mycoplasma infections, the sexual transmission risk is unknown but due to lower pathogens concentrations compared to chlamydial infections, it has been proposed that it might be lower than for chlamydia [19]. For the modelling performed here, a 10% risk was assumed. Keeping in mind that all these values are approximations, it was assumed that they might apply to non-heterosexual contacts in a similar way as well, being aware of the associated uncertainty.
A reliable estimation of disease-specific condom-based protection is not easy. As discussed before [11] and as reported by others [21–24], condoms show efficient protection against body fluid-borne STI transmission but are considerably less efficient against smear-transmission. Best evidence is available for heterosexual HIV transmission, for which a Cochrane review suggested an averaged 80% risk reduction associated with condom use [25]. Recently, condom-based syphilis protection was estimated to be in the 30%–40% range [26], which is not surprising as syphilis is transmitted via smear infection. For gonococci, chlamydia and mycoplasma, available data are considerably less precise [21–24] and so, the following assumptions were chosen for the modelling. Although transmission of gonococci, chlamydia and mycoplasma is body fluid-borne similar to HIV transmission, HIV is poorly transmitted via oro-genital contacts during which condoms are less likely to be used, while this poor transmissibility cannot be expected for the bacterial agents. Accordingly, we assumed a protection rate for gonococci, chlamydia and mycoplasma in the range between condom-based HIV protection and HIV-based syphilis protection, resulting in an estimated 60% value.
Regarding the test-based prevention approach, test-characteristics of modern rapid molecular assays as introduced in a previous publication [27] were used as background for the calculations for gonococci, chlamydia and mycoplasma. For syphilis, PCR (polymerase chain reaction) is less reliable [28] but can support the diagnosis of early infection stages in case of positive results. For the modelling, we assumed a combination of syphilis PCR and stepwise serological syphilis testing as detailed elsewhere [29] and so, we felt justified to assume sensitivity of 98% in combination with specificity of 99% for this parameter (Table 2).
Premises for comparative modelling of different preventive strategies
| STI/causative agent | Averaged transmission risk without any prevention | Assumed condom-based risk reduction | Assumed sensitivity and specificity to calculate the risk-reduction in case of test-based prevention (sensitivity, specificity) |
| Syphilis/Treponema pallidum | 60.0% | 30.0–40.0% | (98.0%, 99.0%) |
| Chlamydial infection/Chlamydia trachomatis | 20.0–35.0% | 60.0% | (97.0%, 99.4%) |
| Gonorrhea/Neisseria gonorrhoeae | 50.0% | 60.0% | (96.0%, 99.9%) |
| Mycoplasmal infection/Mycoplasma genitalium | 10.0%* | 60.0% | (96.0%, 99.3%) |
STI = sexually transmitted infection. *Estimated proportion as explained in the text.
For the modelling in a European population at risk, a number of additional premises was added. Regarding bacterial susceptibility, 0% doxycycline resistance in chlamydiae and T. pallidum like in the study by Molinia [4], 60% doxycycline resistance of mycoplasma like described for Europe elsewhere [19] and 27.8% doxycycline resistance of gonococci as suggested in a recent Belarussian study with pre-pandemic isolates [30] were postulated. As competitors to doxycycline prevention, preventive testing of potential sexual partners prior to unprotected sex (“testing-as-prevention”), condom use or no approach at all were chosen. Similar like in a previous model [28], low prevalence of all assessed bacterial STIs in the predominantly but not exclusively heterosexually active population was assumed. In detail, premises regarding the local prevalence of N. gonorrhoeae, Chlamydia trachomatis and M. genitalium in Germany within the heterosexual and MSM (men-having-sex-with-man) population were taken from a recent publication by Hahn and colleagues [27]. Estimation of syphilis prevalence was conducted based on the annual epidemiological report for 2019 (URL: https://www.ecdc.europa.eu/sites/default/files/documents/syphilis-annual-epidemiological-report-2019.pdf, last accessed on 18th December 2023) by the European Centre for Disease Prevention and Control (ECDC) and the prevalence assumptions as published by Smit and colleagues in 2022 [31]. As syphilis cases are usually early treated and syphilis late presenters have become rare in industrialized countries with well-equipped health systems, syphilis prevalence and incidence were used equivalent for this modelling. Based on these assumptions, the relative protection due to the different applied prevention strategies within this population was assessed as detailed in the following.
Modelling approach
Modelling of the effects of optimized adherence and application schemes
Modelling of different preventive strategies
Ethics
As neither human individuals nor their data or samples were included in the study but just aggregated data provided in openly accessible sources, no ethical clearance was required in line with German National laws.
Results
Modelling of the effects of optimized adherence and application schemes
For the modelling, it was assumed that the relative risk reduction between non-optimized and optimized prevention is constant across all infectious agents of interest. This results in the following preventive effects as indicated in Table 3, adjusted for the proportion of assumed resistance, for a modelling of potential efficacy in case of optimized drug adherence not observed in the underlying studies [4, 18] (Table 3).
Effects of adherence optimization on protective effects regarding infections with N. gonorrhoeae and M. genitalium
| Observed risk in case of imperfect doxycycline-based prevention [4, 18] | Observed risk in case of no doxycycline-based prevention [4, 18] | Modulated risk in case of optimized doxycycline-based prevention [4, 18] and with regard to assumed resistance rates | Modulated attributable risk (RR) compared to the situation without doxycycline-based prevention | |
| Gonorrhea (Neisseria gonorrhoeae) | 0.16 | 0.23 | 0.03 | 0.86 |
| Mycoplasma infection (Mycoplasma genitalium) | 0.10 | 0.10 | 0.02 | 0.82 |
This shows that the expectable preventive effect of doxycycline prevention under conditions of optimized adherence is still relatively high for N. gonorrhoeae (86%) and M. genitalium (82%), although the protection is well below a reduction of 95% as expected for C. trachomatis and T. pallidum in line with the assumptions above. In fact, this is due to the very low initial protection of non-optimized prevention (Table 3).
Comparative modelling of the effects of the alternative prevention strategies condom-based prevention and testing-as-prevention in a European population at risk
The modelling showed that the preventive potential of the test-based strategy is significantly more effective per sexual contact than the lack of prevention. The preventive effect of testing-as-prevention is also significantly higher than the effectiveness of condoms (Tables 4 and 5). Of note, the prevalence-dependent positive predictive value of positive test results in case of test-based prevention is higher in the MSM population than in the heterosexual population (Tables 4 and 5).
Preventive effect per unprotected contact in the heterosexual population
| Prevalence | Risk of infection without any prevention strategy | Risk of infection with condom-based prevention (risk reduction) | Risk of infection with test-based prevention (risk reduction), positive predictive value of the test result | |
| Syphilis (Treponema pallidum) | 0.00003 | 0.00002 | 0.00001 (0.40) | 0.0000004 (0.98), 0.003 |
| Chlamydial infection (Chlamydia trachomatis) | 0.01 | 0.004 | 0.001 (0.60) | 0.0001 (0.97), 0.62 |
| Gonorrhea (Neisseria gonorrhoeae) | 0.01 | 0.005 | 0.002 (0.60) | 0.0002 (0.96), 0.91 |
| Mycoplasma infection/(Mycoplasma genitalium) | 0.01 | 0.001 | 0.0004 (0.60) | 0.00004 (0.96), 0.58 |
| Overall risk of infection (risk reduction) | 0.001 | 0.004 (0.60) | 0.0004 (0.96) | |
Preventive effect per unprotected contact in the MSM (men-having-sex-with-men) population
| Prevalence | Risk of infection without any prevention strategy | Risk of infection with condom-based prevention (risk reduction) | Risk of infection with test-based prevention (risk reduction), positive predictive value of the test result | |
| Syphilis (Treponema pallidum) | 0.002 | 0.0009 | 0.0005 (0.40) | 0.00002 (0.98), 0.13 |
| Chlamydial infection (Chlamydia trachomatis) | 0.07 | 0.03 | 0.01 (0.60) | 0.0008 (0.97), 0.93 |
| Gonorrhea (Neisseria gonorrhoeae) | 0.07 | 0.04 | 0.01 (0.40) | 0.001 (0.96), 0.99 |
| Mycoplasma infection/(Mycoplasma genitalium) | 0.14 | 0.01 | 0.006 (0.40) | 0.0006 (0.96), 0.96 |
| Overall risk of infection (risk reduction) | 0.08 | 0.03 (0.59) | 0.003 (0.96) | |
Over time, as the number of sexual contacts increases, the protective potential of condoms is considerably reduced, especially in comparison to doxycycline prevention. In contrast, the test-based preventive approach still shows a protective potential comparable to doxycycline prevention (Tables 6 and 7).
Preventive effect in the heterosexual population after 100 unprotected contacts
| Prevalence | Risk of infection without any prevention strategy | Risk of infection with condom-based prevention (risk reduction) | Risk of infection with test-based prevention (risk reduction) | |
| Syphilis (Treponema pallidum) | 0.00003 | 0.002 | 0.001 (0.40) | 0.00004 (0.98) |
| Chlamydial infection (Chlamydia trachomatis) | 0.01 | 0.30 | 0.13 (0.56) | 0.01 (0.96) |
| Gonorrhea (Neisseria gonorrhoeae) | 0.01 | 0.39 | 0.18 (0.54) | 0.02 (0.95) |
| Mycoplasma infection/(Mycoplasma genitalium) | 0.01 | 0.10 | 0.04 (0.59) | 0.004 (0.96) |
| Overall risk of infection (risk reduction) | 0.61 | 0.32 (0.48) | 0.03 (0.94) | |
Preventive effect in the MSM (men-having-sex-with-men) population after 100 unprotected contacts
| Prevalence | Risk of infection without any prevention strategy | Risk of infection with condom-based prevention (risk reduction) | Risk of infection with test-based prevention (risk reduction) | |
| Syphilis (Treponema pallidum) | 0.002 | 0.09 | 0.05 (0.39) | 0.002 (0.98) |
| Chlamydial infection (Chlamydia trachomatis) | 0.07 | 0.92 | 0.64 (0.31) | 0.07 (0.92) |
| Gonorrhea (Neisseria gonorrhoeae) | 0.07 | 0.98 | 0.77 (0.21) | 0.14 (0.86) |
| Mycoplasma infection/(Mycoplasma genitalium) | 0.14 | 0.76 | 0.43 (0.43) | 0.06 (0.93) |
| Overall risk of infection (risk reduction) | 1.00 | 0.96 (0.04) | 0.25 (0.75) | |
Discussion
In response to the recently reported substantial increase in bacterial STIs over the past decade in Europe with particular focus on MSM [32], the study was conducted to assess the potential of doxycycline-based STI prevention in case of optimized adherence compared with alternative preventive strategies like condom-based prevention and testing-as-prevention. Doxycycline and test-based prevention were shown to have comparable performance, which is superior to condom-based prevention, especially in the long term. The test-based strategy is not recommended for syphilis due to the low positive predictive values. However, the latter ones are high in case of other infections, especially in the MSM population, and complementary use of testing-as-prevention could therefore be recommended, especially for the prevention of gonorrhea and mycoplasma infections, where doxycycline-based prevention shows weaknesses due to antimicrobial resistance.
In spite of promising modelling results supporting the testing-as-prevention approach for causative agents of bacterial urethritis [27, 33], however, epidemiological evidence at least for screening-associated reduction of infection prevalence and associated disease is scarce [34]. Insofar, the epidemiological balancing of benefit and harm for causative agents other than T. pallidum and HIV is considered as uncertain and screening application for high-risk populations was recently discouraged by a working group unless future high-quality randomized controlled trials provide supportive evidence [34]. For testing-as-prevention strategies, i.e., testing of potential partners immediately before each intended risky sexual contact with mutual consent, no “real-world” study-based evidence is available as well.
The statement regarding the uncertain benefit at least of traditional screening strategies [34] is bad news for high risk-populations, especially for those in which condom use is unpopular. Against gonococci, partial protection due to meningococcal group B outer membrane vesicle-based vaccination has been demonstrated [35], but such effects in the 30% range are yet poor as a prevention approach on an individualized scale.
Insofar, it is reassuring that the here-provided modelling suggests at least partial effects of doxycycline prevention against resistance-affected bacterial agents like N. gonorrhoeae and M. genitalium. However, the remaining uncertainties regarding the quantitative dimension of doxycycline prevention-associated resistance selection may leave a feeling of uneasiness when filling-in the prescription in spite of the above-mentioned EACS recommendation. Evidence of doxycycline prevention-induced resistance selection has been shown in laboratory experiments [36], even selection-based cross-resistance spread was confirmed [37] and the loss of doxycycline as an option for the therapy of difficult-to-cure diseases like pelvic inflammatory disease would be challenging to compensate. Antibiotic therapy of traveler's diarrhea with partly doxycycline-containing schemes has been associated with increased selection of ESBL- (extended spectrum beta-lactamase-) expressing Enterobacterales in the travelers' gut [38]. On the other hand, the quantitative dimension of doxycycline-induced resistance selection in case of preventive application against the acquisition of bacterial STIs is so far unclear. In contrast, well-tolerated long-term doxycycline application for several months or even years has been described for antimalarial chemoprophylaxis as well as for the therapy of acne vulgaris and Q-fever endocarditis [39]. Also, the number of individuals in need for doxycycline prevention of STIs will most likely not be exceedingly high. As recently shown for Germany, where the here-presented modelling was set up, the coital frequency of the vast majority of the population is as low as 4–5 incidents per month [40] and thus similarly low as predicted by a recent STI-prevalence-based modelling approach [41], suggesting only a minor risk of STI acquisition for the most individuals.
Like all modelling-based studies, the main limitation of this work is the residual uncertainty regarding the premises chosen for the calculations. Insofar, errors regarding the underlying assumptions may negatively interfere with the reliability of the provided modelling results.
Conclusions
As shown by the modelling approach, doxycycline prevention against the acquisition of bacterial STI can show effects in the range of other well-established prevention strategies if optimized adherence can be guaranteed.
Conflicts of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Author contributions
Conceptualization, A.H., H.F., U.L.; methodology, A.H.; software, A.H.; validation, A.H.; formal analysis, A.H.; investigation, A.H., H.F., U.L.; resources, U.L.; data curation, A.H., H.F., U.L.; writing—original draft preparation, A.H., H.F.; writing—review and editing, A.H., H.F., U.L.; visualization, A.H.; supervision, H.F., U.L.; project administration, H.F.; funding acquisition, U.L. All authors have read and agreed to the published version of the manuscript.
Funding
No financial support was received for this study.
References
- 1.↑
Molina JM, Capitant C, Spire B, Pialoux G, Cotte L, Charreau I, et al. On-demand preexposure prophylaxis in men at high risk for HIV-1 infection. N Engl J Med 2015;373:2237–46.
- 2.↑
McCormack S, Dunn DT, Desai M, Dolling DI, Gafos M, Gilson R, et al. Pre-exposure prophylaxis to prevent the acquisition of HIV-1 infection (PROUD): effectiveness results from the pilot phase of a pragmatic open-label randomised trial. Lancet 2016;387:53–60.
- 3.↑
Bolan RK, Beymer MR, Weiss RE, Flynn RP, Leibowitz AA, Klausner JD. Doxycycline prophylaxis to reduce incident syphilis among HIV-infected men who have sex with men who continue to engage in high-risk sex: a randomized, controlled pilot study. Sex Transm Dis. 2015;42:98–103.
- 4.↑
Molina JM, Charreau I, Chidiac C, Pialoux G, Cua E, Delaugerre C, et al. Post-exposure prophylaxis with doxycycline to prevent sexually transmitted infections in men who have sex with men: an open-label randomised substudy of the ANRS IPERGAY trial. Lancet Infect Dis. 2018;18:308–17.
- 5.↑
Luetkemeyer AF, Donnell D, Dombrowski JC, Cohen S, Grabow C, Brown CE, et al. Postexposure doxycycline to prevent bacterial sexually transmitted infections. N Engl J Med. 2023;388: 1296–306.
- 6.↑
Zheng H, Xue Y, Bai S, Qin X, Lu P, Yang B. Association of the in vitro susceptibility of clinical isolates of chlamydia trachomatis with serovar and duration of antibiotic exposure. Sex Transm Dis. 2015;42:115–9.
- 7.↑
Greene G, Koolman L, Whyte P, Burgess C, Lynch H, Coffey A, et al. Effect of doxycycline use in the early broiler production cycle on the microbiome. Front Microbiol. 2022;13:885862.
- 8.↑
Hornuss D, Mathé P, Usadel S, Zimmermann S, Müller M, Rieg S. Already current practice? A snapshot survey on doxycycline use for prevention of sexually transmitted infections in parts of the German MSM community. Infection 2023;51:1831–4.
- 9.↑
Stewart J, Bukusi E, Sesay FA, Oware K, Donnell D, Soge OO, et al. Doxycycline post-exposure prophylaxis for prevention of sexually transmitted infections among Kenyan women using HIV pre-exposure prophylaxis: study protocol for an open-label randomized trial. Trials. 2022;23:495.
- 10.↑
Oware K, Adiema L, Rono B, Violette LR, McClelland RS, Donnell D, et al. Characteristics of Kenyan women using HIV PrEP enrolled in a randomized trial on doxycycline postexposure prophylaxis for sexually transmitted infection prevention. BMC Womens Health. 2023;23:296.
- 11.↑
Frickmann H. Diversification of the prevention of sexually transmitted infections. Future Microbiol. 2019;14:1465–8.
- 13.↑
Tran J, Fairley CK, Bowesman H, Aung ET, Ong JJ, Chow EPF. Non-conventional interventions to prevent gonorrhea or syphilis among men who have sex with men: a scoping review. Front Med. (Lausanne) 2022;9:952476.
- 14.↑
Park JJ, Stafylis C, Pearce DD, Taylor J, Little SJ, Kojima N, et al. Interest, concerns, and attitudes among men who have sex with men and health care providers toward prophylactic use of doxycycline against Chlamydia trachomatis infections and syphilis. Sex Transm Dis. 2021;48:615–9.
- 15.
Fusca L, Hull M, Ross P, Grennan T, Burchell AN, Bayoumi AM, et al. High interest in syphilis pre-exposure and post-exposure prophylaxis among gay, bisexual and other men who have sex with men in Vancouver and Toronto. Sex Transm Dis. 2020;47:224–31.
- 16.
Zhang X, Qi SZ, Du FZ, Zheng ZJ, Cao NX, Zheng XL, et al. Awareness and willingness to accept syphilis chemoprophylaxis among men who have sex with men from three cities in China: a cross-sectional study. BMC Public Health 2022;22:1926.
- 17.↑
Tran NK, Goldstein ND, Welles SL. Countering the rise of syphilis: a role for doxycycline post-exposure prophylaxis? Int J STD AIDS 2022;33:18–30.
- 18.↑
Berçot B, Charreau I, Rousseau C, Delaugerre C, Chidiac C, Pialoux G, et al. High prevalence and high rate of antibiotic resistance of Mycoplasma genitalium infections in men who have sex with men: a substudy of the ANRS IPERGAY pre-exposure prophylaxis trial. Clin Infect Dis. 2021;73:e2127–33.
- 19.↑
Jensen JS, Cusini M, Gomberg M, Moi H, Wilson J, Unemo M. 2021 European guideline on the management of Mycoplasma genitalium infections. J Eur Acad Dermatol Venereol 2022;36:641–50.
- 20.↑
Jung H, Peregrina AA, Rodriguez JM, Moreno-Esparza R. The influence of coffee with milk and tea with milk on the bioavailability of tetracycline. Biopharm Drug Dispos. 1997;18:459–63.
- 21.↑
Sing A. Zur Epidemiologie von sexuell übertragbaren Erkrankungen: der Beitrag der Sozialen Netzwerk-Analyse zu einem komplexen Feld. Mikrobiologe 2011;21:158.
- 22.
Mindel A, Sawleshwarkar S. Condoms for sexually transmissible infection prevention: politics versus science. Sex Health 2008;5:1–8.
- 23.
Warner L, Stone KM, Macaluso M, Buehler JW, Austin HD. Condom use and risk of gonorrhea and Chlamydia: a systematic review of design and measurement factors assessed in epidemiologic studies. Sex Transm Dis. 2006;33:36–51.
- 24.
Vera EG, Orozco HH, Soto SS, Aburto EL. Efectividad del preservativo para prevenir el contagio de infecciones de trasmisión sexual. Ginecol Obstet Mex. 2008;76:88–96.
- 25.↑
Weller S, Davis K. Condom effectiveness in reducing heterosexual HIV transmission. Cochrane Database Syst Rev. 2002;1:CD003255.
- 26.↑
Han L, Xiong W, Li M, Li R, Wu J, Tang X, et al. Couple-level determinants of syphilis infection among heterosexual married couples of reproductive age in Guangdong Province, China: a population-based cross-sectional study. Front Public Health 2022;10:1004246.
- 27.↑
Hahn A, Frickmann H, Loderstädt U. Testing as prevention of resistance in bacteria causing sexually transmitted infections-A population-based model for Germany. Antibiotics (Basel) 2021;10:929.
- 28.↑
Leslie DE, Azzato F, Karapanagiotidis T, Leydon J, Fyfe J. Development of a real-time PCR assay to detect Treponema pallidum in clinical specimens and assessment of the assay's performance by comparison with serological testing. J Clin Microbiol. 2007;45:93–6.
- 29.↑
Hahn A, Podbielski A, Meyer T, Zautner AE, Loderstädt U, Schwarz NG, et al. On detection thresholds-a review on diagnostic approaches in the infectious disease laboratory and the interpretation of their results. Acta Trop. 2020;205:105377.
- 30.↑
Aniskevich A, Shimanskaya I, Boiko I, Golubovskaya T, Golparian D, Stanislavova I, et al. Antimicrobial resistance in Neisseria gonorrhoeae isolates and gonorrhoea treatment in the Republic of Belarus, Eastern Europe, 2009–2019. BMC Infect Dis. 2021;21:520.
- 31.↑
Šmit R, Wojtalewicz N, Vierbaum L, Nourbakhsh F, Schellenberg I, Hunfeld KP, et al. Epidemiology, management, quality of testing and cost of syphilis in Germany: a retrospective model analysis. Front Public Health. 2022;10:883564.
- 32.↑
Mitjà O, Padovese V, Folch C, Rossoni I, Marks M, Rodríguez I, et al. Epidemiology and determinants of reemerging bacterial sexually transmitted infections (STIs) and emerging STIs in Europe. Lancet Reg Health Eur. 2023;34:100742.
- 33.↑
Hahn A, Schwarz NG, Meyer T, Frickmann H. PCR-based rapid diagnostic tests as a strategy for preventing infections with sexually transmitted diseases-a ‘diagnostics-as-prevention' modelling approach. Lett Appl Microbiol. 2018;67:420–4.
- 34.↑
Kenyon C, Herrmann B, Hughes G, de Vries HJC. Management of asymptomatic sexually transmitted infections in Europe: towards a differentiated, evidence-based approach. Lancet Reg Health Eur. 2023;34:100743.
- 35.↑
Petousis-Harris H. Impact of meningococcal group B OMV vaccines, beyond their brief. Hum Vaccin Immunother 2018;14:1058–63.
- 36.↑
Kenyon C, Gestels Z, Vanbaelen T, Abdellati S, Van Den Bossche D, De Baetselier I, et al. Doxycycline PEP can induce doxycycline resistance in Klebsiella pneumoniae in a Galleria mellonella model of PEP. Front Microbiol. 2023;14:1208014.
- 37.↑
Gestels Z, Manoharan-Basil SS, Kenyon C. Doxycycline post exposure prophylaxis could select for cross-resistance to other antimicrobials in various pathogens: an in silico analysis. Int J STD AIDS. 2023;34:962–8.
- 38.↑
Kantele A, Lääveri T, Mero S, Vilkman K, Pakkanen SH, Ollgren J, et al. Antimicrobials increase travelers' risk of colonization by extended-spectrum betalactamase-producing Enterobacteriaceae. Clin Infect Dis. 2015;60:837–46.
- 39.↑
Rothe C, Veit O, Bühler S, Feldt T, Grobusch MP, Kapaun A, et al. Empfehlungen zur Malariaprophylaxe. [Recommendations for antimalarial prophylaxis.] Flugmedizin – Tropenmedizin – Reisemedizin 2023;4:168–208.
- 40.↑
Briken P, Dekker A, Cerwenka S, Pietras L, Wiessner C, von Rüden U, et al. Die GeSiD-Studie „Gesundheit und Sexualität in Deutschland“ – eine kurze Einführung. Bundesgesundheitsbl 2021;64:1334–8.
- 41.↑
Hahn A, Kröger C, Meyer CG, Loderstädt U, Meyer T, Frickmann H, et al. Comparison of self-reported sexual activity among heterosexuals with sexual spread of poorly transmittable agents: a minimalistic approach to estimating sexual activity based on HIV incidence. Int J Environ Res Public Health. 2020;17:5504.
