Histopathologic patterns of female genital tuberculosis with clinical correlation: a 10-year (2013–2022) retrospective cross-sectional study – BMC Women’s Health

This study analyzed 122 cases of Female Genital Tuberculosis (FGTB) at AAU in Ethiopia. The diagnosis of FGTB presents challenges and requires clinical suspicion and diagnostic tools. The objective of the study is to assess the prevalence, trends, and clinical/histopathologic patterns of FGTB in gynecology specimens received by the pathology department.

The study found that TB was present in 0.94% of female genital tract specimens analyzed over the past ten years. The annual percentage has been declining, with the highest occurrence in 2013. Similarly, a previous study conducted in the same department in 1999 reported a prevalence of 1.38% [9]. These decreasing trends align with the national decline in TB incidence since 1995 [12]. In contrast, in New Delhi, India, TB endometritis was identified in 2.3% of nonpregnant endometrial samples [13]. Likewise, a study in the TASH gynecology department in 2018 reported varying prevalence rates of endometrial TB using different diagnostic methods: 4.6% using PCR, 2.6% using culture, and only 1.3% using histological examination [14].

Furthermore, a study by Radhika et al. compared different diagnostic methods for female genital tuberculosis, revealing that combining PCR with BACTEC increased sensitivity to 52% [15]. Similarly, Sethi et al. emphasized the inadequacy of relying on a single test for diagnosing genital tuberculosis and stressed the importance of clinical evaluation in conjunction with a combination of conventional and newer diagnostic methods [16].

Considering the relatively low proportion of TB cases observed in this study, it is reasonable to suggest that the absence of alternative diagnostic methods, such as culture and more sensitive molecular techniques, may have resulted in an underestimation of TB prevalence. Therefore, future research should incorporate these advanced modalities to improve the accuracy of diagnosing female genital tuberculosis.

In this study, most of the patients were between the ages of 20 and 39, accounting for 75% of the cases. A total of 9.3% were over the age of 50, with no cases documented under the age of 15. This finding is consistent with the study conducted here in Ethiopia by Abebe et al., which shows that the majority of cases (83.18%) occurred between the ages of 20 and 40 [9]. Similarly, in a study conducted in Pakistan by Qureshi et al., the majority of cases (75%) were between the ages of 20 and 45 [17]. Furthermore, in a study performed in Uganda by Othieno et al., the most common age group was 21–30 years, accounting for 48.7% of the cases [18]. The young age observed in these populations’ patients could be explained by early marriages and pregnancies, as well as the overall young age of the population. In addition, the atrophic endometrium in elderly women is not a favorable environment for the development of mycobacterium TB, which explains the rarity of tuberculosis in postmenopausal patients [19]. Our finding is different from the study in Sweden by Falk et al., which showed that GTB was more common after menopause [10].

In this study, the most common presentations were menstrual disturbance (46.3%), abdominopelvic pain (34.3%), and infertility (32.4%). In the study by Abebe et al., abnormal uterine bleeding (45.5%) and infertility (36%) were the most common presentations [9]. Similarly, in the study by Othieno et al., the most common presentations were abnormal uterine bleeding (35.9%), infertility (28.2%), and abdominal distention (28%) [18]. However, in most studies, the most common complaint in women with GTB is infertility, followed by abdominopelvic pain, and abnormal uterine bleeding. In a study in India by Mondal et al., 65–70% of patients presented with infertility, 50–55% with pelvic/abdominal pain, and 20–25% with menstrual disturbances [20]. Similarly, Qureshi et al. observed that infertility was the most common presenting symptom (42.5%), followed by abdominal pain (42%) [21]. Because the patients’ marital status was not recorded in this study, it was difficult to pinpoint a cause for the documented low infertility rate compared to most studies. Nevertheless, it is generally agreed that menstrual disorder occurs in approximately 40–50% of women who suffer from GTB, which is consistent with our study.

In the present study, three patients with GTB were misdiagnosed as having malignant disease, resulting in aggressive surgical management. Two patients underwent TAH with SO, and one patient underwent TAH with SO and colectomy. On histopathology, they were found to have ovarian TB with peritoneal involvement and, in one case, a concomitant benign ovarian cyst. Similar to our study, retrospective studies conducted in Malaysia and Turkey found that patients with tubo-ovarian TB associated with peritoneal involvement (ascites) and a frequently associated high level of serum CA-125 are frequently misdiagnosed with ovarian carcinoma and are subjected to unnecessary and aggressive surgery [22, 23]. Furthermore, case reports by Lobo et al. from Singapore and Yang et al. from China documented the synchronous occurrence of benign cystic neoplasm and ovarian TB and the diagnostic difficulty it poses [24, 25]. In the literature thus far, few cases of ovarian TB and coexisting benign cystic neoplasm of the ovary have been reported, and all cases were regarded as mere coincidences [24,25,26]. In this study, which was conducted in a country with a high TB prevalence, 31.3% of ovarian TB patients had concurrent benign cystic neoplasms. Hence, it is advisable to carry out more research to examine this relationship because it may be more than just a coincidence.

In this study, the endometrium was the most frequently affected organ, accounting for 55.6% of the cases, followed by the fallopian tube (23.1%), the ovary (14.8%), the cervix (14.8%), and the vulva (1.9%). Similar results to ours were also observed in studies conducted by Abebe et al., Turkmen et al., & Mondal et al. [9, 20, 27]. In contrast, the study by Nogales et al. found that the fallopian tube was the most frequently affected [17]. When FGT occurs, the fallopian tube is involved in nearly all patients, and involvement of the endometrium is usually secondary to tubal disease [19]. Thus, the fact that tubal involvement was less common in our study than endometrial involvement may be attributed to sampling bias due to the easy accessibility of the endometrium for biopsy, and the majority of the specimens received in our department were endometrial biopsies. In the endometrium, 46.7% of cases showed well-formed to multicentric granulomas with varying degrees of caseous necrosis, whereas Mondal et al. found endometrial caseation in only 2.6% of cases [20]. The histopathology of FGTB typically resembles tuberculosis in other tissues, but advanced-stage caseation, fibrosis, and calcification are uncommon in the endometrium during the reproductive period due to regular cyclical shedding [28]. Progesterone has a “flaring” effect on the endometrial tuberculosis located at the basal layer. The highest levels of progesterone occur in the middle of the second half of the menstrual cycle. Therefore, the optimal time to perform an endometrial biopsy to assess would be in the middle of the luteal phase.

The higher percentage of caseation observed in our study may be due to the presence of amenorrheic or postmenopausal women among those with caseation. However, in line with most studies, the majority of patients in our study (53.3%) had early-stage immature granulomas. Therefore, it is advisable to conduct biopsies just before menstruation or during the late secretory phase.

In our study, AFB was detected in 42.6% of the FGT tissues with a histopathological diagnosis of TB. In a study conducted by Eshete in Ethiopia on 60 lymph node tissues with a histopathological diagnosis of TB, AFB was detected in 61.7% [29]. Similarly, in a study performed by Krishnaswanni in India on 128 lymph node tissues and by Rasool in Pakistan on 50 lymph node tissues, both with a histopathological diagnosis of TB, AFB was detected in 79% and 56% of the cases, respectively [30, 31]. In another study conducted in England by Greenwood and Fox on 70 cases with a histopathological diagnosis of TB from various anatomical sites (including 21 lymph nodes, 9 kidneys, 6 gastrointestinal tracts, 5 epididymis, 4 livers, 4 omenta, 3 synoviums, 1 brain, 1 vagina, 1 heart, 1 spleen, and 1 skin), AFB was detected in 60% [32]. Similarly, a study conducted in America by Koch et al. on 136 tissues (100 from the lung, 28 from lymph nodes, 5 from the spleen, and 1 each from the heart, liver, and bone marrow) from different anatomic sites found that 34.5% were positive for AFB [33].

The relatively lower rate of AFB detection in our study compared to other studies may reflect the paucibacillary nature of FGTB. Consistent with the paucibacillary nature of the disease in our study, other studies have found a low overall detection rate of FGTB by AFB staining. According to Malhotra et al., the AFB detection rate was 2.7%, 1.6% according to Bhanu et al., and 8.3%according to Thangappah [34,35,36]. The higher detection rate of AFB in our study compared to these studies is most likely because the AFB stain was performed on tissues with a histopathological diagnosis of TB, unlike the other studies that were conducted on clinically suspected tissues. Furthermore, in the other studies, only the ZN stain was used, whereas in this study, both ZN and auramine stain were used, which may have contributed to the observed higher rate.

In this study, auramine detected AFB in 22 tissues, and ZN detected AFB in 21 tissues (17 cases stained positive with both ZN and auramine, 4 cases with ZN only, and 5 cases with auramine stain only). This finding supports the higher sensitivity of the auramine stain, as documented in the literature. A study conducted in India by Krishnaswanni compared ZN and fluorescent techniques for AFB demonstration on 128 lymph node tissues with a TB histopathology diagnosis found 71.1% positive for AFB after ZN staining and 79.7% positive for AFB on fluorescent techniques [30]. Similarly, a study conducted in England by Greenwood and Fox on 70 tissues from various anatomical sites with a TB histopathology diagnosis found 47.1% positive for AFB with the ZN staining technique and 60.0% positive for AFB with the fluorescent technique [32]. A study conducted in Pakistan by Rasool on 50 lymph node tissues with a TB histopathology diagnosis revealed that 54% were positive for AFB after ZN staining and 56% were positive for AFB after fluorescence staining [31]. Bacteria are easier to visualize with the auramine stain, so tissues that stain positive with ZN also stain positive with auramine. The four tissues in our study that stained positive for ZN but not for auramine could be attributed to technical issues, which might have contributed to the slight difference in auramine and ZN positivity compared to the other studies.

In this study, the ovary had the highest rate of AFB detection, followed by the fallopian tube, endometrium, and cervix. AFB was detected in 35.3% of tissues with stage 1–2 granulomas and 51.9% of tissues with stage 3–4 granulomas. Therefore, the relationship between AFB positivity and granuloma stage was not statistically significant, with a p-value of 0.1. Although we observed a trend toward higher AFB detection in stage 3–4 granulomas, further studies with larger sample sizes are needed to confirm this finding. Among the 7 TB/HIV coinfected women, 1 had stage-1 granulomas, 2 had stage-4 granulomas, and the remaining 4 had stage-2 granulomas. AFB was not detected in the four tissues that were examined. However, it is important to note that the histologic features of TB granulomas can vary depending on the patient’s immune status. Therefore, conducting a larger study with a stratified study population based on immunosuppression levels would provide a better characterization of the histologic features of FGTB with HIV coinfection.

The strength of this study, compared to other similar studies, is the inclusion of patients over ten years resulting in a relatively large sample size, which increases the study’s importance. However, a limitation of this study is that definitive evidence of mycobacterium TB infection using more sensitive tests such as PCR/culture was not performed due to unavailability issues.