Predominance of high-level tetracycline-resistant Neisseria gonorrhoeae in Kenya: Implications for global implementation of doxycycline post-exposure prophylaxis for prevention of sexually transmitted infections

Abstract

Using archived Neisseria gonorrhoeae samples from 2008–2012, the prevalence of tet(M) gene mediating high-level tetracycline resistance in N. gonorrhoeae was 96% among 50 Kenyan women. Determining local and national prevalence of gonococcal tetracycline resistance and surveillance of gonococcal antimicrobial resistance can inform implementation of doxycycline post-exposure prophylaxis for STI prevention.

Short Summary

Most archived Neisseria gonorrhoeae samples collected from 2008–2012 from Kenyan women had tet(M)-encoding high-level tetracycline resistance; this could reduce the efficacy of doxycycline post-exposure prophylaxis against N. gonorrhoeae.

Introduction

Neisseria gonorrhoeae (NG) infection is the second most common bacterial sexually transmitted infection (STI) in the world, and it is associated with substantial morbidity and economic cost [1,2]. The World Health Organization (WHO) estimates that in 2020, 82.4 million new cases of NG occurred among adolescents and adults [1]. STIs, including NG infection, are often under-reported in many African countries, in part because syndromic management is used much more commonly than diagnostic testing in symptomatic individuals and screening is rarely available for persons who are asymptomatic [3]. Data from HIV PrEP trials among Eastern and Southern African women indicate high incidence of NG (up to 20 per 100 person-years) [4].

Biomedical interventions for STI prevention such as doxycycline prophylaxis have been proposed to be valuable for reducing the incidence of bacterial STIs among users of HIV PrEP, but antibiotic prophylaxis could both select for and be undermined by antimicrobial resistance (AMR).

Two clinical trials have demonstrated that doxycycline post-exposure prophylaxis (doxycycline PEP) reduces the incidences of bacterial STIs, NG, Chlamydia trachomatis (CT), and Treponema pallidum (TP) by 70% among men who have sex with men (MSM) in France and by 62% among both MSM and transgender women in the United States [5,6]. However, there are concerns that doxycycline PEP also could induce or promote AMR in STI and non-STI pathogens and commensal bacteria [7]. Antimicrobial-resistant NG is a public health issue globally, with a growing threat of untreatable multidrug resistant cases [8,9], but there are limited NG resistance data from Africa [10,11]. The community-level prevalence of tetracycline-resistant NG is an important clinical factor that may influence the effectiveness of doxycycline PEP for STI prevention in each setting. In addition, CT is more common than NG [1], and they often occur as co-infections. There are no reported cases of laboratory-confirmed tetracycline resistance in CT, although there is theoretical risk based on rare reports of resistance mediated by the tet(C) gene among Chlamydia suis in animals [12].

Using archived endocervical swab specimens collected from 2008 and 2012 for STI testing in Kenya [13,14], we evaluated the prevalence of tetracycline-resistant NG and CT among cisgender women prior to launching a study of doxycycline PEP among cisgender women in Kenya [15].

Methods

Between 2008 and 2012, 9,494 participants were enrolled and followed in the Partners PrEP Study, a randomized clinical trial of antiretroviral HIV pre-exposure prophylaxis (PrEP) among heterosexual HIV serodifferent couples, at nine sites in Kenya and Uganda [13,14]. The study protocol was approved by the University of Washington (UW) Human Subjects Review Committee and ethics review committees at each of the study sites, and all participants provided written informed consent in English or their primary language [13,14]. Endocervical swabs were collected from HIV seropositive female participants every 12 months for NG and CT testing by nucleic acid amplification test (NAAT) and duplicate endocervical swabs were stored at −80C at the UW for future STI research. All participants with an STI received treatment according to the national STI treatment guidelines. The laboratory work described in this paper was performed at the UW Neisseria Reference Laboratory. The DNA from 50 randomly selected swabs collected from heterogenous population from 4 sites in Kenya (Eldoret, Kisumu, Nairobi, and Thika) from visits at which NG had been detected by NAAT was purified and extracted (supplementary Table S1). DNA samples were used as templates for detection of the tet(M) genes of the American and Dutch plasmids using previously described PCR and cycling parameters, and standard positive and negative controls [16, 17]. Additionally, the DNA from 10 randomly-selected endocervical swabs, which had tested positive for CT by NAAT with no concurrent NG detected, were tested for tet(C) gene [12]. Descriptive data analyses were performed with Stata version 15.0 software (StataCorp).

Results

A total of 9,494 participants, including 4,747 cisgender women, were enrolled in the Partners PrEP Study, with a median age of 33 years. Among participants living with HIV (n=4,712), the prevalence of NG was 1.4% and CT 0.6%; NG incidence during follow-up was 4.9 per 100 person-years and CT 1.1 per 100 person-years. At enrollment, 9,415 participants were tested for STIs with 683 participants found to be positive (Trichomonas vaginalis, CT, and/or NG).

Of the fifty archived endocervical swabs with NG detected by NAAT, 48 (96%) carried the American type plasmid-mediated tet(M) gene conferring high-level tetracycline resistance. None of the 50 endocervical swabs harbored the Dutch type tet(M). Sixteen (32%) participants received treatment as syndromic management or directed treatment for a documented STI prior to swab collection; however, there was no doxycycline given to participants prior to collection of swabs tested for tet(M) [Table 1]. Prophylactic use of trimethoprim-sulfamethoxazole as part of HIV care to reduce bacterial infections was common among the 49 participants, and 7 (14%) had a prior diagnosis of pelvic inflammatory disease [Table 1]. None of the 10 additional endocervical swabs with CT tested positive for tet(C).

Table 1.

Recent antibiotic exposures and risk factors for tetracycline antimicrobial resistance among 50 female participants in Kenya found to have N. gonorrhoeae infection in the Partners PrEP Study, 2008–2012

Characteristics	tet(M) Negative N = 2	tet(M) Positive N = 48*
Prior diagnosis of vaginal discharge	1 (50%)	11 (23%)
Prior diagnosis of Cervicitis	2 (100%)	3 (6.3%)
Prior diagnosis of Pelvic Inflammatory Disease	2 (100%)	5 (10%)
Prior diagnosis of Neisseria gonorrhoeae	0 (0%)	14 (29%)
Prior diagnosis of Chlamydia trachomatis	0 (0%)	1 (2.1%)
Taking prophylactic trimethoprim-sulfamethoxazole	2 (100%)	47 (98%)
Recent treatment with doxycycline (<1 month)	0 (0%)	0 (0%)

^*All 48 tet(M) Positive samples were American variant of the tet(M) gene

Discussion

We observed a high prevalence of high-level tetracycline resistance mediated by the American variant tet(M) gene among non-cultured NG samples from 50 Kenyan women. Therefore, doxycycline PEP for STI prevention may have limited efficacy against gonorrhea in sub-Saharan Africa, although use of doxycycline is also unlikely to select for additional resistance to tetracyclines in NG given the near-universal prevalence of tet(M). The impact of trimethoprim-sulfamethoxazole prophylaxis among persons with advanced HIV immunosuppression on cross-resistance selection to other categories of antibiotics such as tetracyclines is yet to be fully elucidated. We did not detect the tet(C) from the CT samples confirming no hypothetical resistance, which is consistent with results to date worldwide [12].

Recently, analyses of whole genome sequence (WGS) data have shown that high percentages (86–97%) of NG isolates collected in Kenya between 2010 and 2018 carried the plasmid-mediated tetracycline resistance tet(M) gene [10,18]. Moreover, high percentages of carriage of the plasmid-mediated tet(M) gene, conferring high-level tetracycline resistance determined by PCR and/or by antimicrobial susceptibility testing, have been reported from cultured NG isolates from other sub-Saharan African countries including South Africa, 2008–2012 (73–92%) [19, 20], Mozambique, 2005 (77%) [21], and Uganda, 2008–2009 (90%, TET MIC₉₀ 16µg/mL) [22]. Consistent with our finding of the American variant tet(M) gene from all 48 non-cultured NG samples, a previous study from South Africa found all 136 NG isolates with high-level tetracycline resistance (MIC≥16µg/mL) to carry the American variant of the tet(M) gene [23], and a recent study from Kenya found that 30 (97%) of 31 NG tet(M)-encoding plasmids were American type [18]. Epidemic spread of plasmid-mediated tetracycline-resistant NG has been previously reported in sex workers in Kinshasa, Zaire [24], and gonorrhea treatment failures have been associated with high-level tetracycline resistant NG strains [25]. Of note, data from Africa are in contrast to those from the United States, where circulating plasmid-mediated tetracycline resistant NG are not universal and where recent data revealed NG prevention in a doxycycline PEP trial [6].

We acknowledge the small sample size and the use of endocervical specimens collected approximately 10 years ago. However, our culture-independent molecular approach for detecting plasmid-mediated tetracycline resistance from non-cultured clinical specimens is a strength and has been recommended by WHO for enhancing gonococcal AMR surveillance [26].

Public health and provider messaging about doxycycline PEP efficacy against NG will vary by NG tetracycline resistance rate, which is dynamic in the setting of increasingly interconnected sexual networks; but it could still be efficacious for CT and TP. We recommend that integral components of the implementation of doxycycline PEP for STI prevention should include establishing the local prevalence of NG tetracycline resistance and ongoing robust surveillance of gonococcal antimicrobial resistance.