Association between Chlamydia trachomatis, Neisseria gonorrhea, Mycoplasma genitalium, and Trichomonas vaginalis and Secondary Infertility in Cameroon: A case-control study

Objective Data on the prevalence and etiology of infertility in Africa are limited. Secondary infertility is particularly common, defined as the inability of a woman to conceive for at least one year following a full-term pregnancy. We describe a prospective study conducted in Cameroon designed to test the hypothesis of an association between common treatable sexually transmitted infections (STI): Chlamydia trachomatis (CT), Neisseria gonorrhoeae (NG), Mycoplasma genitalium (MG), and Trichomonas vaginalis (TV) and secondary infertility in women. Methods In this case-control study, we enrolled women in Fako Division, Cameroon between November 2017 and December 2018 with secondary infertility (cases) or current pregnancy (controls). We conducted a baseline survey to collect sociodemographic, and sexual and medical history information. Nucleic acid amplification testing using Aptima (Hologic, San Diego, CA, US) was performed on endocervical swabs for CT, NG, MG, and TV. Multivariable logistic regression was used to assess the relationship between active STI and secondary infertility. Results A total of 416 women were enrolled: 151 cases and 265 controls. Compared to controls, cases were older (median age 32 vs 27 years) and had more lifetime sexual partners (median 4 vs 3) (p<0.001). Cases were more likely to report dyspareunia, abnormal menses, prior miscarriage, and ectopic pregnancy (all p<0.05). STI positivity was not significantly different among cases and controls (2.7% vs 5.4% for CT, 1.3% vs 2.9% for NG, 6.0% vs 7.0% for MG, respectively), with the exception of TV which was more common in pregnant controls (0.7% vs 5%; p = 0.02). Conclusion Study findings did not support an association between active STI and secondary infertility in Cameroon. Given high rates of pre-existing tubal damage, routine STI screening and treatment in younger women may be more impactful than costly STI testing during infertility assessments.


Introduction
Infertility affects 10-15% of couples, an estimated 80 million people globally [1]. Secondary infertility is defined as the inability to conceive for at least one year by a woman with a history of prior delivery. In Cameroon, a country in central Sub-Saharan Africa (SSA), the prevalence of infertility ranges from 15% to 30% and secondary infertility is twice as common as primary infertility [2,3]. Stigma associated with infertility is common in SSA and women in couples with infertility are more likely to experience domestic violence, divorce, and social marginalization compared to women without infertility [2,4]. Additional studies are needed to understand the etiology of infertility in SSA to inform prevention and treatment options in a region where access to assisted reproductive technology is limited [4,5].
Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) infections are common bacterial sexually transmitted infections (STI) that cause cervicitis in women and ascend to the upper reproductive tract in 10-20% of cases, leading to pelvic inflammatory disease (PID). PID can cause inflammation and damage to fallopian tubes that results in tubal factor infertility (TFI) [6,7]. CT infection is associated with a 3.4-fold increase in TFI [8]. Data on the association between Mycoplasma genitalium (MG) infection and infertility in women are limited and Trichomonas vaginalis (TV) infection is not known to impact fertility. Studies of MG, TV and infertility using highly sensitive PCR molecular diagnostic testing are ongoing but this STI testing is mostly unavailable in SSA [9,10]. Since women with secondary infertility have a prior history of delivery, they may be more likely to have a reversible cause of TFI, such as a recently acquired STI. If active infection is present, it is possible that STI treatment could reduce tubal inflammation and improve fertility. This case-control study was designed to test our hypothesis of an association between active common STIs and secondary infertility among women in Cameroon.

Study area and setting
The study was conducted in semi-urban areas of Fako Division (Buea, Limbe Tiko, and Mutengene) in the Southwest region of Cameroon in Central Africa. Participants were seen at the main reference hospital in the region. Cases with secondary infertility were enrolled at the gynecological unit and pregnant controls were enrolled at antenatal clinics in the same region.

Study design and participants
This unmatched case-control study enrolled women in the age range of 15-46 years. Cases were women with secondary infertility according to self-report or as identified by a gynecologist. Secondary infertility was defined as the inability to conceive by a woman despite at least one year of unprotected sexual intercourse following a prior delivery. Controls were randomly selected pregnant women seen in prenatal clinic who reported no history of infertility. Approximately two controls were selected for each case identified. Women were ineligible to participate if they refused to grant consent and if they had taken any antibiotics during the past 7 days.

Study procedures
Women were enrolled in the study after providing informed consent. Gynecologists, nurses, and study staff surveyed women about socio-demographics, sexual history, obstetrical and gynecological history, including any infertility evaluations. Medical records were reviewed for STI history, pregnancy history, and other medical history, when available. Questionnaire and study procedure was formulated to follow hospital patient management protocols at the various study sites.

Biological specimen collection
Endocervical swabs were collected by a medical provider using Dacron swabs and placed directly into Hologic 1 Aptima 1 vaginal swab transport media (Hologic Inc San Diego, USA). Specimens were transported to the University of Buea Laboratory and stored at -80˚C until transport to the University of Alabama at Birmingham STI Laboratory (USA) for STI nucleic acid amplification testing (NAAT).

Ethical considerations
The study was approved by the Faculty of Health Sciences Institutional Review Board in Cameroon (No 2016/459/UB/FHS/IRB). Administrative clearance was also obtained from the Regional Delegation of Public Health for the Southwest Region and all the health facilities where study was conducted.

Statistical analysis
For a power calculation assuming two controls per case, a sample size of 408 (136 cases and 272 controls) would provide >80% power at an alpha level of 0.05 to detect an association between STI and secondary infertility if STI positivity in the control group was 19%.Chisquare or Fisher's exact testing was used for comparisons of categorical variables where appropriate; and, analysis of variance or the Kruskal-Wallis test was used for continuous variables. Univariate logistic regression was performed to identify predictors of secondary infertility. Variables of interest, specifically socio-demographics, sexual behaviors, and a composite STI measure were included in the full multivariable logistic regression based on published data to explore the association between STI positivity and secondary infertility. Unadjusted and adjusted ORs, 95% CI, and p-values were calculated with significance set at p<0.05. SAS version 9.4 (Cary, NC) was used for analyses.

Participant characteristics
A total of 416 women were enrolled: 151 cases and 265 controls (Table 1). Cases were older than controls, median (interquartile range) 32 years (27, 36) vs 27 years (24, 32), p<0.001. Sexual behavior differed somewhat between the groups: 26% of cases and 7% of controls had �5 lifetime partners (p<0.001) although the median number of lifetime partners was <5 in both groups. Two-thirds of women in both groups were married and cases were more likely to report multiple current sexual partners (4% cases vs 0.8% controls; p = 0.03).
Among cases, 46 had hysterosalpingography (HSG) results in their medical record. Of these 46, 31 women (67%) had documented evidence of tubal obstruction +/-hydrosalpinx (20 with bilateral obstruction, 11 with unilateral obstruction), 6 had uterine adhesions, and 1 had radiographic evidence of cervicitis. The STI positivity (CT/NG/MG/TV) amongst cases with HSG was 3/46 (6.5%). None of the controls had HSG results as a comparison group since they had no prior evidence of infertility according to inclusion criteria.

Factors associated with secondary infertility
In unadjusted models with secondary infertility as the outcome, older age, white collar occupation, higher number of lifetime sex partners and having more than one current sex partner were associated with infertility (Table 2). When compared to the youngest age category of 15-24 years, the crude odds ratios (OR) and 95% confidence intervals for secondary infertility increased in a stepwise fashion from 2.9 (CI 1.5-5.6) for women aged 25-29 years to 9.6 (95% CI 4.5-20.6) for women aged 40-46 years. Women with more than one current sex partner (OR 5.4; 95% CI 1.1-27.3) and �5 lifetime sex partners (OR 4.7; 95% CI 2.5-8.9) had greater odds of secondary infertility. Women with dyspareunia, abnormal periods, history of miscarriage and prior ectopic pregnancy also had greater odds of secondary infertility ( Table 2).

Discussion
This is one of the first studies to focus on the association between active STI and secondary infertility in Africa using highly sensitive molecular diagnostic testing to detect CT, NG, MG, and TV in women of reproductive age in Cameroon. Our prospective case-control study did not support our hypothesis of a possible association between active STI and secondary infertility although classic STI risk factors (higher number of lifetime sexual partners, more than one current sexual partner) were more prevalent in cases compared to controls. STI prevalence among women in this study was similar to other studies in Africa [15][16][17][18][19][20]. Among the four STIs detected, only TV showed a significant difference between the two groups, with higher positivity in controls. Additional studies are needed to explore the association noted in unadjusted models between TV and pregnancy. Any assessment of the association between prior STI infection and secondary infertility in Cameroon is limited by the fact that, unlike high-income countries, routine STI testing among women is not recommended and sensitive diagnostic testing is mostly unavailable. The performance of syndromic STI management in SSA as recommended by the World Health Organization is demonstrably poor, particularly in women who normally have asymptomatic bacterial STI [21].
In terms of STI and TFI, tubal blockage has been detected in 50% of women with infertility (mostly secondary infertility) in urban Cameroon [22]. This is consistent but lower than the 67% of women with secondary infertility shown to have tubal blockage in the semi urban setting of this present study. The mechanism of pathogenesis of NG and CT fallopian tube infection and inflammation resulting in tubal infertility has been well-described [23]. In the literature, tubal obstruction is seen in 20-30% of cases of primary and secondary infertility [24]. Egbe et al. found that a history of CT was strongly associated with tubal factor infertility in Cameroon [25]. A meta-analysis demonstrated a higher prevalence of NG in infertile populations compared to the general population [16]. Although data on MG are fewer, a recent US study by Peipert et al found an association between serologic evidence of prior MG infection and infertility [26,27].
Our study did not show an association between active infection with CT, NG, MG, and secondary infertility. These findings do not rule out a possible association between bacterial STI and secondary infertility since HSG test results among participants with secondary infertility suggested that tubal damage had already occurred. One relevant implication of our study is that current infertility management protocols in Cameroon that rely on CT serology testing and antibiotic treatment, may be ineffective and costly for the patient if infection occurred in the past. Therefore, the WHO recommended syndromic STI management and treatment (which fails to detect asymptomatic infections) may not be an effective strategy to prevent tubal damage and adverse pregnancy outcomes. More frequent and earlier STI screening and treatment of women in SSA may be needed to prevent upper genital tract complications [28]. In addition to highly sensitive molecular diagnostic testing, it would be useful to measure STIspecific host responses to detect prior infection and further understand the role of each STI in terms of a causal pathway for TFI. A serologic study to compare immunoglobulin-specific responses to CT between cases and controls in our population will be informative about prior CT exposure and is currently ongoing. We have previously shown that CT antibody testing such as IgG1 to elementary bodies and CT OmcB can help elucidate the timing of prior infection [29].
Some factors associated with secondary infertility in our study align with other studies; age, adverse pregnancy outcomes (miscarriage and ectopic pregnancy) and dyspareunia [30,31]. Modifiable risk factors such as marital status, education, and age of sexual debut had no association with secondary infertility in our study. This differs from findings by Larsen et al. in Cameroon where secondary infertility was associated with early sexual debut (before the age of 15) and single marital status [32].
Our study had limitations. Importantly, cases and controls were not age-matched, duration of secondary infertility varied from 1-18years among cases and there was heterogeneity in lifetime number of sexual partners between groups (although a majority reported 1-5 partners). Since the risk of STI generally decreases with age and cases were older than pregnant controls, this may have biased our sample toward higher STI rates in pregnant controls (not observed for CT/NG/MG). Participant recall error and bias in STI history information are also possible since STI testing can be performed as part of the initial infertility evaluation. If cases received empiric STI treatment during infertility evaluation, they may have had more historical antibiotic exposure compared to controls. Finally, since the study sample was isolated to women in Southwest Cameroon, findings may not generalize to all women with secondary infertility.
Future studies should include detailed diagnostic evaluations to better understand the most common etiologies of primary and secondary infertility among women and men in sub-Saharan Africa as well as the frequency and etiology of tubal damage. Once the population attributable fraction due to STI is understood, prospective prevention studies can be designed.

Conclusion
This prospective case-control study of women with secondary infertility in Cameroon did not show an association with active STI (CT, NG, MG, or TV). Since rates of tubal damage were high, the benefit of routine CT/GC NAAT testing and prompt treatment in young sexually active women for the endpoint of infertility prevention should be investigated further.