Rectal Chlamydia trachomatis infection in women. Is it overlooked?

INTRODUCTION

Chlamydia trachomatis (CT) is the most common sexually transmitted bacterial infection in the UK. ¹ The D-K serovars are the most prevalent and can cause ocular, genital, pharyngeal and rectal infection. Patients with rectal infection with serovars D-K are often asymptomatic but these serovars can cause anorectal pain, tenesmus or rectal discharge. ² More recently in Europe there has been an outbreak in men who have sex with men (MSM) of lymphogranuloma venereum (LGV), which is caused by Chlamydia trachomatis serovars L 1–3. Current European cases of LGV have been due to the L2 genotype, which can cause painful inguinal lymphadenopathy (inguinal syndrome) but has manifested primarily an acute haemorrhagic proctitis (anorectal syndrome) in the MSM outbreak. ³

With the steady increase in the diagnosis of LGV in Europe, since 2003, screening for and serotyping of rectal chlamydia infection has gained importance. Whilst not adopted universally, many clinics offer routine testing for rectal CT in MSM but such screening is not routine in women, even in those who admit to receptive anal intercourse. In various studies the prevalence of rectal chlamydia in women is estimated at between 5% and 21%. ⁴ However, these estimates were made before more sensitive nucleic acid amplification tests (NAATs) were available and may be falsely low.

Following an outbreak of LGV cases in Swansea in 2004/2005, at Singleton hospital genitourinary (GU) medicine clinic, we have routinely offered screening for rectal CT in all MSM in accordance with British Association for Sexual Health and HIV (BASHH) guidelines. Given that many of our female patients admit to receptive anal sex, we subsequently decided to offer screening for rectal as well as cervical CT infection in this group. In addition, because gonococcal and chlamydial infections are epidemiologically linked, contacts of Neisseria gonorrhoeae infection were also screened for both cervical and rectal chlamydial infection. Patients reporting sexual assault, sex workers and those with symptoms of proctitis were also offered rectal CT tests in addition to routine screening. Given the increased costs involved, we undertook a study to determine the prevalence of rectal chlamydia in the women tested and also assessed the degree of concordance between cervical and rectal CT results.

In addition, within this study cohort, we analysed results of gonorrhoea and chlamydia tests in known and alleged contacts of gonococcal infection to assess the concordance between these infections. This paper is a summary of our findings.

METHODS

We performed a retrospective study analysing case-notes of all women who underwent rectal CT testing between September 2006 and August 2008. The additional rectal swabs were taken sporadically for women reporting anal intercourse as not all doctors/nurse practitioners enquired about this type of intercourse and not all women who admitted to anal intercourse had rectal swabs taken. Rectal swabs were obtained either using a blind technique or proctoscopically. Rectal chlamydia swabs were tested using a BD Probe Tec FT strand displacement amplification (SDA) assay (Becton Dickinson Diagnostic Systems, Sparks, MD) until April 2008 and subsequently with the Abbott RealTime™ CT/NG polymerase chain reaction (PCR) assay (Abbott Molecular Diagnostics, Des Plaines, IL, USA). Neisseria gonorrhoeae infection was diagnosed by standard culture methods.

The patient's age and indication for testing was noted. These indications were patients who admitted to anal intercourse/sexual abuse, proven and alleged contacts of gonorrhoea, patients with rectal symptoms and ‘other’ (commercial sex workers, intravenous drug users and patients with bisexual partners). The CT detection rate in each group was recorded and we also noted the degree of concordance between cervical and rectal swabs. The presence of gonococcal infection in contacts of known cases of gonorrhoea and alleged contacts was also noted and related to the presence or absence of ano-genital CT infection.

RESULTS

One hundred and seventy-five rectal CT swabs were taken from 154 women in the two-year study period. Thirteen rectal swabs were tests of cure (all of which were negative) and have been eliminated from our data. Clinical records were not available from two patients leaving a study group of 152 patients from whom 160 rectal swabs were taken for chlamydia prior to treatment.

Sixteen patients were aged 40 or over and one chlamydia detected from a rectal swab of these had. Twenty-four patients were aged 30–39 and two (8.3%) had rectal CT, 83 patients were in the 20–29 age group of whom 10 (12%) had positive swabs for rectal CT and 31 patients were aged less than 20 of whom seven (22.6%) had rectal CT infection. Indications for rectal CT testing and detection rates were as in Table 1.

Fifteen patients with pertinent anorectal symptoms were screened. These symptoms included perianal itching, rectal discharge, lower abdominal pain with negative cervical swabs and other non-specific gastrointestinal symptoms.

In the study period, 156 paired rectal and cervical swabs for chlamydia were taken simultaneously and a further four patients had only rectal swabs taken during their clinic visit (all four tests were negative). Among the 156 patient episodes in which paired rectal and cervical swabs were taken, 19 tests gave concordant positive results for both rectal and cervical chlamydia, two tested positive from the cervix only, one had a positive cervical swab and an inhibitory result in the rectum and one tested positive from the rectum only.

Thus, overall, of 160 rectal swabs taken in the two-year study period, 20 were positive (12.5%), all specimens were confirmed to be non-LGV serovars by the Sexually Transmitted Bacteria Reference Laboratory (STBRL), Colindale, UK. We also had two rectal swabs reported as inhibitory and one swab was reported as equivocal. Of the 20 patients with rectal CT, only six specimens were obtained by proctoscopically.

Of the 24 women who either were known or had alleged contacts of gonorrhoea, 10 had ano-genital gonococcal infection detected. The sites of the gonococcal infection in these 10 women were as follows: cervix and rectum, six; cervix, only two; rectum, only one. For unknown reasons one gonorrhoea contact had rectal CT swabs taken but not rectal gonococcal swabs. Nine of the 10 women also had urethral swabs taken, of which eight were positive for gonorrhoea. Similarly of the 10 women who actually had gonorrhoea, four had concomitant chlamydial infection. Three of the four had both cervical and rectal CT (30%) and one had cervical infection only.

DISCUSSION

The most alarming finding of our study is the very high (12.5%) prevalence of rectal CT in our high-risk female group. This is an even higher prevalence than in our high-risk men who have sex with men (MSM). An unpublished departmental audit showed that between May 2005 and August 2008, we took 551 rectal CT specimens from MSM and of these, 57 were positive (10.3%) including six LGV cases.

The age-related prevalence of rectal CT in our female study group is also of note. An astounding 22.6% of our high-risk women who were aged less than 20 had rectal CT while the prevalence of rectal CT in our older women with similar risk factors was much less. This is possibly due to the fact that they have older partners who are less likely to have CT infection or that these women may have acquired immunity to chlamydial infection. ⁵

In female patients, BASHH guidelines advise screening for rectal CT only in those who complain of an abnormal genital discharge and whose sexual history or local symptoms/signs may indicate rectal infection. However BASHH guidelines are confusing since it also mentions that most rectal CT infections are asymptomatic while paradoxically recommending rectal screening only for symptomatic female patients. ⁶ Since the prevalence of rectal chlamydia in women varies in different studies, we urgently need further research to assess the actual prevalence of rectal CT in both high risk and routine clinic patients. ⁷

BASHH guidelines also recommend that rectal swabs are taken via proctoscopy but anecdotally, practice varies in different clinics and little is published to see if BASHH guidelines are followed. ⁸ Thus, although the high prevalence of rectal CT in our group of female patients is alarming, our study showed that most positive specimens were obtained without a proctoscope making contamination from the cervix a distinct possibility. This may be of greater importance since the introduction of highly sensitive assays. ⁹

In 1981 Quinn and colleagues ¹⁰ stated that ‘in adults the main route of sexual transmission and acquisition of rectal CT is likely to be unprotected anal intercourse, although in women there is the possibility of spread of cervical chlamydial infection via the perianal region to the rectum’.

The high concordance between cervical/rectal chlamydia specimens in our study would support this and it seems likely that by taking cervical swabs only, most cases of rectal CT would be coincidentally treated. However, our numbers are small and given the high prevalence of rectal CT we cannot be complacent until further large studies are carried out.

We also do not know whether rectal CT with non-LGV serovars is less likely to respond to standard therapy than cervical chlamydia since good data on the efficacy of the recommended drugs (a week of doxycycline or single dose azithromycin) is lacking. ¹¹ With rectal gonococcal infection there is evidence that is more persistent at this site, and this may also be the case with rectal CT infection. ^12,13 Clearly, further studies are needed to assess the prevalence of rectal CT in women and to evaluate optimal treatment and follow-up regimens.

Finally, given limited resources, if rectal chlamydia swabs in women are limited for fiscal reasons, the groups at greatest risk seem to be high-risk women aged less than 20 years and women with proven gonococcal infection in whom we found a prevalence of 22.6% and 30%, respectively.