Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) infections are the most common bacterial sexually transmitted infections (STI) throughout the world.1 These infections may cause complications both in men and women, such as epididymitis, urethritis, cervicitis, pelvic inflammatory disease, and ectopic pregnancy.1,2 STI in extragenital sites, such as anus, rectum, and pharynx, are an increasing cause for concern. Recent studies show increasing reports of anal intercourse amongst heterosexuals and lower rates of condom use in anal intercourse compared to vaginal intercourse.3 Anorectal mucosa is vulnerable to HIV due to lack of appropriate protective humoral immune barrier and for being more susceptible to traumatic lesions then the vaginal mucosa.4 In addition, possible biological, behavioral, and social factors, such as insufficient knowledge regarding anorectal STI risks and anal intercourse to please the partner also contribute to STI infection.5 The low percentage of diagnostic screening in addition to inappropriate treatment maintain the bacterial STI chain-of-transmission, thus increasing STI and HIV transmission.6
The presence of CT and NG infection, especially in the anorectal region, is associated to the increased risk of HIV infection. In people living with HIV/AIDS (PLHA), CT and NG infections increase the genital HIV viral load (VL) and the possibility of sexual and vertical transmission of the virus.7 Despite the raise of the HIV epidemics amongst men who have sex with men (MSM) highlight the role of unprotected anal intercourse on the HIV transmission, the role of this practice in the heterosexual HIV transmission is still poorly understood.
Although studies show anorectal prevalence for CT in women (6.6-9.3%) to be similar to that of MSM (6.5-10.1%) and the therapeutic recommendations for infections on this site possibly differ from those in the urogenital sites,6,8,9 there is still no definition regarding the systematic investigation for CT and NG in extragenital sites in heterosexual women.
The aim of this study was to estimate the prevalence of anorectal and genital infection by C. trachomatis and Neisseria gonorrhea and the associated risk factors, such as lifestyle and sexual practices, in women and men living with HIV/AIDS receiving care in a reference center in Salvador, Brazil.
Material and methods
Patients and settings
This was a cross-sectional study conducted at Centro Especializado em Diagnóstico, Assistência e Pesquisa (CEDAP) from June 2013 to June 2015. CEDAP is the state reference center for STI and HIV in Salvador, Bahia, Northeast of Brazil, attending approximately 60% of PLHA in the state, with an average of 76 new cases of HIV/AIDS and 373 new cases of STI each month. The health center is staffed with infectious disease specialists, and other medical and paramedical professionals. Irrespective of their area of specialization physicians are trained to deliver care for patients with sexually transmitted diseases.
Confirmed HIV-infected patients undergoing treatment with the gynecologist and proctologist at the clinic were invited to participate in the study, regardless of their signs and symptoms of STI. Sexually active patients regardless of age were assessed. Patients who had used antibiotics 30 days before from the appointment, and women with genital bleeding at the exam were not included. Pregnant women with no recent obstetric complications were also included in this study.
The CT and NG screening was performed using qPCR in closed system - In vitro Diagnostic (IVD), COBAS 4800® Roche, using COBAS® PCR Media Female as transport for the endocervix and anorectal specimens, and COBAS® PCR Media Urine for male urine samples. The samples were collected according to the manufacturer instructions. The anorectal samples collection was adapted for COBAS® PCR Media Female, as it is not standardized for the IVD from COBAS 4800® Roche system. The anorectal samples were collected through the swab introduced 2-3 cm after anal margin and it was done in 360° turn; endocervix samples were collected by gynecologist during specular exam; and urine was collected by the patients in adequate recipients. The samples were collected during a medical appointment at CEDAP and were processed at the Professor Gonçalo Moniz Central Laboratory of Public Health of Bahia - LACEN-BA.
All patients had a blood sample drawn to assess HIV viral load and TCD4+/TCD8+ cells count at the time of the appointment. Lymphocyte TCD4+/TCD8+ count was performed by flow cytometry (Facscalibur, Becton and Dickinson, California, USA) and HIV viral load was quantified using PCR Real time (Abbot molecular, Illinois, USA)
Socio-demographic, behavioral, and clinical data were obtained through standardized medical interview. Patients were scheduled further medical appointment one month after sample collection for delivering tests results and prescribing treatment of the identified infections. All patients signed a written informed consent. This study was approved by the Ethics Committee of the Maternidade Climério de Oliveira/Universidade Federal da Bahia (process 292,413).
Data analysis were performed using SPSS 20.0 (SPSS Inc, Chicago, IL, USA). Chi-square test was used for univariate analysis of categorical variables like ethnicity (white × non-white), marital status (single × married/stable union), schooling (≤8 years of study × >8 years of study), alcohol intake, tobacco and drug use (yes × no). Continuous variables such as age, age at first intercourse, number of partners, and time since HIV diagnosis were analyzed by Student's t test. p-values lower than 0.05 were considered statistically significant; 95% confidence intervals (CI) were calculated for means and proportions. Variables with p ≤ 0.20 in univariate analysis were included in the logistic regression backward stepwise model for multivariate analysis. For women, the variables in the regression were: age, schooling, pregnancy, alcohol use, drugs use, pelvic pain, cervicitis, genital discharge, alcohol before sex, anal receptive intercourse, HAART use, HIV viral load. For men, those who referred no anal receptive intercourse were excluded from the logistic regression analysis because they did not have CT or NG anal infection. Therefore, the variables analyzed were: age, ethnicity, alcohol before sex, pain in anal intercourse, genital ulcer and painful anorectal exam.
A total of 521 PLHA, 208 men and 313 women were evaluated. Of those, 15 (2.9%) men who did not collect a urine sample and eight (2.6%) women who had inadequate anorectal samples were excluded. There was no statistically significant difference between the enrolled and excluded patients.
The final sample comprised 305 women and 193 men. The overall prevalence of any CT or NG infection was 12.3% (61/498), 9.2% (28/305) cases amongst women and 17.1% (33/193) amongst men. The overall mean age was 37.0 years (±10.5), 10.4% (52/498) self-reported to be white, 38.6% (192/498) were married or in a common-law marriage, and 73.5% (366/498) had more than eight years of regular education. A total of 83.1% (414/498) were on antiretroviral therapy, but 37.0% (165/446) had viral load above 40 copies/mL.
Some significant differences between genders were found in this study sample. Men were younger, mostly single, with higher education and family income. Alcohol, tobacco and drug use were more frequently declared by men, as well as higher number of sexual partners, receptive anal sex, and history of STI, as seen on Table 1.
|Characteristics||Men (n = 193)||Women (n = 305)||p-Value|
|Age (years), mean (SD)||35.8 (9.9)||37.7 (10.8)||0.042|
|Age at sexual debut (years), mean (SD)||14.9 (3.4)||16.4 (3.5)||<0.001|
|Lifetime number of sexual partners, median (IQR)||30 (12–200)||5 (3–10)||<0.001|
|White race, n (%)||30 (15.5)||22 (7.2)||0.003|
Marital status, n (%)
|46 (23.8)||159 (52.1)||<0.001|
|Educational level ≥8 years, n (%)||170 (88.1)||196 (64.3)||<0.001|
|Monthly household income ≤2 minimum wages, n (%)a||102 (53.4)||252 (82.6)||<0.001|
|Alcohol use, n (%)||144 (74.6)||166 (54.4)||<0.001|
|Tobacco use, n (%)||42 (21.9)||29 (9.6)||<0.001|
|Drug use, n (%)||52 (26.9)||43 (14.1)||<0.001|
|Alcohol use before sex, n (%)||91 (47.2)||123 (40.9)||0.169|
|Drug use before sex, n (%)||32 (16.6)||23 (7.5)||0.002|
|Transactional sex, n (%)||19 (9.8)||28 (9.2)||0.805|
|Anal receptive intercourse, n (%)||167 (86.5)||191 (62.6)||<0.001|
|Previous STI, n (%)||167 (86.5)||151 (49.5)||<0.001|
|Time from HIV diagnosis (months, median (IQR)||42.5 (10.5–121.7)||85.2 (32.8–136.8)||<0.001|
|ART in use, n (%)||159 (82.4)||255 (83.6)||0.722|
|Duration on ART (days), median (IQR)||24.3 (2.0–82.6)||48.7 (3.0–121.7)||0.004|
STI, sexually transmitted infections.The numbers do not always add up the total because of missing values.
aMinimum wage ≈$194.
Amongst women, there were seven cases of combined CT infection (endocervix and anus). Two men (6.1%) had both anorectal and urine positive for NG infection, and one (3.0%) for CT infection in both sites. Regarding the investigated site, there were 14.0% (27/193) of positive cases in anus and 3.1% (6/193) in genital region in men, while 5.6% (17/305) and 3.6% (11/305) cases in women, respectively (Table 2).
|Anus||18/193 (9.3)||11/193 (5.7)||27/193 (14.0)|
|Urine||3/193 (1.6)||3/193 (1.6)||6/193 (3.1)|
|Anus||16/305 (5.3)||2/305 (0.7)||17/305 (5.6)|
|Endocervix||11/305 (3.6)||0/305 (0.0)||11/305 (3.6)|
Clinical, behavioral, and epidemiological aspects associated with the presence of CT and NG infection in genital and anorectal areas in men and women are described in Table 3. Among men, anorectal infection was associated with age <29 years (p = 0.033), report of anal intercourse (p = 0.029), pain during anal intercourse (p = 0.028), and painful anorectal exam (p = 0.022). Only men who referred anal intercourse have CT and NG anal infection. After logistic regression, the variables that remained significantly associated with anorectal infection were painful anorectal exam (p = 0.014, OR-3.59, 95% CI 1.29-9.95), and white ethnicity (p = 0.018, OR-3.85, 95% CI 1.26-11.77). On the other hand, no association between genital infection and other variables were detected in univariate analysis. After logistic regression, only lifetime sexual partners >3 (p = 0.019, OR-22.41, 95% CI 1.67-305.94) was independently associated with genital infection.
|Urogenital CT/NG (n = 6)||Anorectal CT/NG ( n = 27)||Endocervical CT/NG ( n = 11)||Anorectal CT/NG ( n = 17)|
|n (%)||p||OR||95% CI||n (%)||p||OR||95% CI||n (%)||p||OR||95% CI||n (%)||p||OR||95% CI|
|Age ≤29 years old||2 (3.8)||.669||1.32||0.24–7.45||12 (22.6)||.023||2.61||1.12–6.10||8 (10.8)||.001||9.21||2.38–35.71||11 (14.9)||.000||6.55||6.33–18.40|
|Single||4 (2.7)||.652||0.62||0.12–3.31||21 (14.3)||.832||1.11||0.42–2.95||5 (3.1)||.630||0.76||0.23–2.54||11 (6.9)||.285||1.73||0.63–4.82|
|White ethnicity||2 (6.7)||.235||2.84||0.50–16.25||7 (23.3)||.108||2.18||0.82–5.72||1 (4.5)||.567||1.30||0.16–10.65||3 (13.6)||.114||3.03||0.80–11.48|
|Less than 8 schooling years||0 (0.0)||1.000||1.04||1.01–1.07||2 (8.7)||.748||0.55||0.12–2.50||8 (7.3)||.020||5.10||1.32–19.63||7 (6.4)||.630||1.27||0.47–3.46|
|Monthly income ≤2 MWa||3 (2.9)||1.000||0.87||0.17–4.42||15 (14.7)||.809||1.11||0.49–2.51||11 (4.4)||.222||0.96||0.93–0.98||14 (5.6)||1.000||0.98||0.27–3.54|
|Pregnant women||–||–||6 (18.8)||.000||12.37||3.53–43.31||5 (15.6)||.023||4.03||1.32–12.30|
|Alcohol use||5 (3.5)||1.000||1.73||0.20–15.15||22 (15.3)||.376||1.59||0.57–4.45||3 (1.8)||.120||0.30||0.08–1.16||8 (4.8)||.530||0.73||0.27–1.95|
|Tobacco use||2 (4.8)||.614||1.83||0.32–10.33||6 (14.3)||.962||1.02||0.39–2.73||0 (0.0)||.608||1.04||1.02–1.07||1 (3.4)||1.000||0.58||0.07–4.51|
|Drug use||2 (3.8)||.661||1.37||0.24–7.71||9 (17.3)||.420||1.43||0.60–3.42||3 (7.0)||.191||2.38||0.60–9.35||1 (2.3)||.483||0.37||0.05–2.83|
|Lifetime sexual partners >3||2 (66.6)||.092||18.20||1.41–235.34||2 (66.6)||.370||3.10||0.27–35-38||6 (3.0)||.409||1.66||0.50–5.57||9 (4.5)||.237||1.80||0.67–4.80|
|Irregular condom use||2 (4.8)||.614||1.83||0.32–10.32||7 (16.7)||.583||1.30||0.51–3.32||3 (3.0)||1.000||0.75||0.19–2.88||3 (3.0)||.194||0.41||0.12–1.47|
|Alcohol before sex||3 (3.3)||1.000||1.13||0.22–5.72||17 (18.7)||.076||2.11||0.91–4.89||1 (0.8)||.031||0.14||0.02–1.09||6 (4.9)||.631||0.78||0.28–2.16|
|Drug before sex||1 (3.1)||1.000||1.01||0.11–8.92||4 (12.5)||1.000||0.86||0.28–2.67||1 (4.3)||.584||1.24||0.15–10.11||1 (4.3)||1.000||0.76||0.10–5.97|
|Transactional sex||0 (0.0)||1.000||1.04||1.01–1.07||2 (10.5)||1.000||0.70||0.15–3.22||2 (7.2)||.267||2.29||0.47–11.17||1 (3.6)||1.000||0.60||0.08–4.73|
|Anal receptive intercourse||5 (3.0)||.585||0.77||0.09–6.88||27 (16.2)||.029||0.84||0.78–0.90||4 (2.1)||.108||0.33||0.09–1.14||12 (6.3)||.485||1.46||0.50–4.26|
|Previous STI||6 (3.6)||1.000||0.96||0.94–0.99||26 (15.6)||.136||4.61||0.60–35.53||6 (4.0)||.734||1.23||0.39–4.13||6 (4.0)||.228||0.54||0.19–1.49|
|Anal fissureb||2 (3.4)||1.000||1.12||0.20–6.31||10 (16.9)||.456||1.38||0.59–3.23||–||–|
|Pain in anal intercourse||2 (4.0)||.676||1.28||0.23–7.23||12 (24.0)||.028||2.55||1.09–5.99||–||–|
|Dyspareunia||–||–||3 (4.5)||.706||1.36||0.33–5.62||5 (7.6)||.696||1.24||0.42–3.73|
|Genital discharge||1 (16.7)||.178||7.12||0.70–72.72||1 (16.7)||.594||1.26||0.14–11.27||6 (5.5)||.188||2.21||0.66–7.43||9 (8.3)||.131||2.10||0.79–5.62|
|Genital ulcer||1 (4.0)||.570||1.36||0.15–12.13||6 (24.0)||.122||2.21||0.79–6.16||0 (0.0)||1.000||1.04||1.02–1.06||0 (0.0)||.610||1.06||1.03–1.09|
|Pelvic pain||0 (0.0)||1.000||0.58||0.07–4.66||3 (17.6)||.713||1.35||0.36–5.04||6 (5.5)||.188||0.58||0.18–1.85||4 (3.7)||.313||0.53||0.17–1.68|
|Anal pain||2 (2.6)||1.000||0.74||0.13–4.14||14 (18.2)||.179||1.74||0.77–3.95||0 (0.0)||1.000||1.04||1.02–1.07||0 (0.0)||.614||1.06||1.03–1.09|
|Anal fissureb||1 (1.6)||.666||0.38||0.04–3.36||9 (14.1)||.914||1.05||0.44–2.51||–||–|
|Genital discharge||1 (20.0)||.148||9.15||0.86–97.35||1 (20.0)||.533||1.56||0.17–14.49||5 (4.7)||.463||1.57||0.47–5.26||6 (5.6)||.985||1.01||0.36–2.81|
|Urethritis||1 (16.7)||.175||7.28||0.71–74.35||1 (16.7)||1.000||1.24||0.14–11.03||–||–|
|Cervicitis||–||–||3 (13.6)||.037||5.43||1.33–22.14||3 (13.6)||.114||3.03||0.80–11.48|
|Painful anorectal examb||0 (0.0)||.585||1.04||1.00–1.07||9 (25.0)||.022||2.88||1.13–7.34||–||–|
|Painful bimanual exam||–||–||2 (7.7)||.270||2.28||0.47–11.15||0 (0.0)||.382||1.07||1.04–1.11|
|≤40 cp/mL||4 (4.0)||12 (11.9)||3 (1.7)||6 (3.3)|
|>40 cp/mL||1 (1.6)||9 (14.5)||8 (7.8)||9 (7.8)|
|CD4+ ≤500 cells/µL||1 (1.7)||.653||0.42||0.05–3.84||10 (16.7)||.271||1.67||0.66–4.21||3 (3.6)||1.000||0.92||0.24–3.55||7 (8.4)||.178||1.99||0.72–5.55|
|No ART in use||1 (2.94)||1.000||1.07||0.12–9.48||5 (14.7)||.894||0.93||0.33–2.66||5 (10.0)||.008||0.22||0.06–0.74||5 (10.0)||.136||0.44||0.15–1.32|
CT/NG indicates chlamydia and/or gonorrhea infection.
aMW, minimum wage ≈$194.
bInvestigation on anal fissure and painful anorectal exam were only performed by the proctologist in male patients.
Among women, there was an association between CT genital infection and age less than 29 years old (p < 0.001), younger age at first sexual intercourse (p = 0.048), schooling less than eight years (p = 0.020), pregnancy (p < 0.001), viral load >50 copies/mL (p = 0.020), and no antiretroviral use (p = 0.008). After logistic regression, age less than 29 years old (p = 0.010, OR-8.25, 95% CI 1.67-40.76), schooling less than eight years (p = 0.012, OR-8.29, 95% CI 1.61-42.77), pregnancy (p = 0.002, OR-13.57, 95% CI 2.63-69.94), alcohol use before sexual act (p = 0.025, OR-14.54, 95% CI 1.39-151.61), and cervicitis (p = 0.056, OR-6.18, 95% CI 0.96-39.97) remained statistically significant.
Anorectal infection in women was associated with age less than 29 years old (p < 0.001) and pregnancy (p = 0.023), but it was neither associated with report of anal intercourse (p = 0.485) nor presence of symptoms. Only age less than 29 years old (p = 0.001, OR-10.54, 95% CI 3.23-34.41) remained associated with infection after logistic regression.
A high prevalence of CT and NG infection in PLHA (12.3%) was found in this study, which was higher among men (17.7%) than women (9.2%). Infections were more prevalent in the anorectal (8.8%) site than in the genital (3.8%) site. Nowadays, there is an increase in anorectal STI, even among heterosexuals.3,9,10 Some guidelines recommend CT and NG screening in extragenital sites on MSM, however it remains undefined for heterosexuals.11,12 Some authors suggest focusing the screening on women who report anal intercourse.10,13 The lack of association between women reporting anal intercourse and presence of anorectal infection in our study corroborate the findings of a study conducted in Baltimore, 2014.14 The presence of bacterial infection with or without symptoms in anus and rectum, the practice of anal intercourse in the general population, and the increasing risk of HIV transmission reinforce the need for appropriate STI diagnosis and treatment on anorectal site. In this study, if the screening had been only urogenital, CT and NG infection diagnoses would have been missed in 88.9% of cases in men and 58.8% in women.
Amongst the women evaluated on this study, the association between anorectal and genital site of infection and younger age (p < 0.001 on both cases) and pregnancy (p < 0.001 and p = 0.023) highlights the need for screening that population during routine follow-up. These associations have been found in studies with women living with HIV/AIDS, and in the general population,15-17 but it is not yet a routine in most services in Latin America. Recent studies detected a higher chance of HIV vertical transmission on women co-infected with NG or CT.18,19 A systematic CT and NG investigation for young women, pregnant women, and women living with HIV/AIDS can contribute to the reduction of HIV transmission, and be an effective prevention measure.
In our study, the prevalence found on women's anorectal region was 5.3% for CT and 0.7% for NG; 29.4% of women with infection in the anorectal region denied practice of anal intercourse. There are few reports regarding the investigation of this site in women who do not admit anal intercourse or symptoms on the anal region.10 The prevalence rates described in women with this practice vary between 8.6%-12.7% for CT and 1.0%-2.9% for NG on anorectal sites.6,14,20 We found an association between presence of infection in cervix and in anorectal site, similar to other authors.13 Some studies highlight the possibility of self-inoculation or "translocation" of genital site infection.14,21 Our study underscores the need for investigating the anorectal site in women living with HIV/AIDS, regardless of anal intercourse practice.
We found a high rate of anorectal site infection among men (14.0% positive cases, 9.3% for CT and 5.7% for NG), similar to the CT and NG prevalence found in other studies in Brazil (10.0% and 2.5%), USA (7.9% and 6.9%), Netherlands (10.1% and 5.5%), and Russia (7.3% and 2.0%).6,22-24 Only men who referred anal intercourse have CT and NG anal infection. The characteristics of this population are similar to those studied in other countries, with multiple sexual partners, receptive and insertive anal intercourse, higher education/socioeconomic situation. Despite the access to information on STI prevention, the high prevalence indicates a high exposure to NG and CT. These points toward an urgent need for reinforcing preventive measures, such as condom use, education on sexual transmission risk, and prompt access to bacterial and HIV infections post-exposure prophylaxis.
Antiretroviral therapy (ART), HIV viral suppression and improved immune status did not protect against STI infection in urogenital or anal sites in male patients of our study. The available evidence on these factors are controversial: some authors detected more infections in patients with higher TCD4 + cell counts,25 but Cunha et al. did not find such association.22 Some studies have shown an increased transmission of HIV in men or women infected by a bacterial STI, possibly due to the immune activation that increases expression of CCR5 receptors on mucosa and genital viral load.7,26,27 Receptive unprotected anal sex represents risk of HIV transmission of 5-18 times higher on each sexual act.28,29 The sense of protection currently publicized by the use of ART cannot be extended to the STIs on the population of men living with HIV.
This study has limitations: an interview was conducted face to face, leading to some degree of inhibition by the respondents, especially as to sexual practices. The population included in the study attends the service and, therefore, has access to care and guidance. There is no such data on HIV-negative people, or on those who do not attend specialized health centers. PLHA, is a population with increased prevalence of other STIs, and we have to be cautions when extrapolating these findings to the general population. Defining who is the heterosexual population who needs screening for infection in extragenital sites still requires more epidemiological and cost-effectiveness studies.
The "Treat as Prevention" strategy adopted in 2013 by Health Ministry of Brazil30 and in 2015 by the WHO, intends to reduce HIV sexual transmission through increasing virological suppression. The impact of this policy on bacterial STIs transmission, such as syphilis, CT, and NG in PLHA is unknown. The increase of unprotected sex due to the possibility of not occurring sexual transmission of HIV could lead to an increase of these STIs. This may be a limitation to the future success of the "Treat as Prevention" strategy. Training of health professionals, access to early diagnosis and treatment of STIs, educational interventions on sexual practices, risk management, and prevention along with the population, especially PLHA and key populations, are essential measures to end the HIV epidemic, the global goal for 2030, and break bacterial STI chain-of-transmission.