Diagnostic Hysteroscopy In Abnormal Uterine Bleeding & It'S Histopathologic Correlation: Our Experience

Introduction

Abnormal uterine bleeding is the commonest complaint which is noticed in a gynecology out patient setting. Goldstein et al¹ has defined abnormal uterine bleeding as “patients having either metrorraghia defined as vaginal bleeding separated from expected menses or menorraghia defined as patient's subjective complaints of either increased duration or increased volume of flow or both. Patients usually present themselves to the gynecologist when there is variation in the normal cyclical pattern. The variation could be due to physiological, hormonal change or may be due to benign or malignant condition. Age specific association of endometrial lesions are known to occur².Hence, this needs a proper evaluation. Hysteroscopy guided biopsy is the recommended diagnostic test in the present day. The traditional blind dilatation and curettage can miss focal intrauterine lesions like polyp, sub mucous fibroid and cancer. The other diagnostic tests involve transvaginal sonography (TVS) and saline infusion sonography (SIS)³. Hysteroscopy illuminates the darkness of the uterine cavity. So hysteroscopy is considered as a gold standard in the evaluation of abnormal uterine bleeding in perimenopausal and postmenopausal age group.

Aims and Objectives

1. To study the accuracy of hysteroscopy in evaluation of abnormal uterine bleeding in perimenopausal and postmenopausal women.

2. To correlate hysteroscopic findings with histopathologic results.

Materials and methods

It is a retrospective study of 175 cases who have attended the department of OBG at K.S.Hegde Medical Academy, Mangalore in the last 6 years with complaints of abnormal uterine bleeding. All patient's history and clinical examination findings were noted. All patients had an ultrasonography done. Those patients who underwent diagnostic hysteroscopy followed by dilatation and curettage were selected for this study. There hysteroscopy findings and histopathological reports were analyzed.

Patients who underwent hysteroscopy for infertility were excluded. Patients who had an obvious cause of bleeding from cervix or vagina were also excluded. Patients who were on anticoagulant therapy, hormone replacement treatment and who were known case of bleeding dyscrasias were excluded.

Records of hysteroscopy findings were classified as normal, hyperplasia, atrophy, polyp, fibroid, cancer. Histopathological diagnosis was taken as gold standard to determine accuracy of hysteroscopy findings for diagnosing endometrial abnormalities. Their sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were assessed. Data collected was statistically compiled using SPSS 20 version.

Results

In our study, 29 cases (16.6%) were postmenopausal and remaining 146 cases (83.4%) were perimenopausal. Women in age group of 41-50 years formed a major group of 95 cases (54.3%). Majority of the patients were parous, that is 159 (90.9%). Among the study group, 27 cases (15.4%) had co-morbid medical disorders.11.4% were hypertensive, 1.7% were diabetic, whereas 2.3% patients had both diabetes and hypertension and 0.6% were hypothyroid. Only 2 patients out 175 had a family history of breast cancer. Among 12 patients who had endometrial cancer on histopathology, 7 patients had associated medical co-morbidities (Table 1).

The commonest symptom for hysteroscopy was menorraghia about 87(49.7%) followed by postmenopausal bleeding 29(16.6%). Other indications for hysteroscopy were irregular cycles, polymenorraghia, continuous bleeding per vaginum and dysmenorrhoea (Table 2). All hysteroscopy were done in operation theatre under general anesthesia. Normal saline was used as the distending medium for diagnostic hysteroscopy. For all patients, endometrial tissue was sent for histopathological examination.

The hysteroscopy findings were tabulated as normal, hyperplasia, atrophy, polyp, fibroid and cancer (Table 3). 53 cases of hyperplasia were diagnosed on histopathology, but 108 cases were suspected as hyperplasia on hysteroscopy. Out 53 cases of hyperplasia on histopathology, 35 cases were simple hyperplasia without atypia, 10 cases were simple hyperplasia with atypia, 5 cases of complex hyperplasia without atypia and 3 cases of complex hyperplasia with atypia. Out of85 cases of normal endometrium, only 25 cases were diagnosed as normal on hysteroscopy. Out of 16 cases which were diagnosed as atrophy on hysteroscopy, 5 were confirmed as atrophic on histopathological examination (HPE) &6 were insufficient endometrium obtained again probably due to atrophic endometrium (Table 4).

13 cases of polyp were seen on hysteroscopy, whereas on HPE, 6 cases were confirmed to be polyp. In case of detection of submucous fibroid, 10 cases were diagnosed on hysteroscopy. Since none underwent resection of the fibroid at the same sitting it could not be confirmed by HPE. Of 12 cases of endometrial cancer diagnosed on HPE, 3 cases were suspected as cancerous growth on hysteroscopy, 8 appeared hyperplasia &1 was diagnosed as submucous fibroid on hysteroscopy.

The correlation between hysteroscopy findings and histopathological examination of endometrial curetting were studied (Table 5). Of those 108 patients with hyperplasia on hysteroscopy, 45 patients (41.6%) had hyperplasia, 52 patients (48.1%) had normal endometrium and 8 patients (7.4%) had cancer. Of 25 patients with normal hysteroscopies, 18 (72%) had normal endometrium and 4 patients (16%) had hyperplasia on hispathological examination. On hysteroscopy 16 patients were suspected to have atrophic endometrium, 5 were confirmed , 6 turned out to be insufficient endometrium which is probably because of atrophy.13 patients were diagnosed to have polyp on hysteroscopy, however 6 (46.1%) were confirmed to have polyp on curettage and 2 (15.3%) each were normal and hyperplastic. Submucous fibroid was diagnosed in 10 patients on hysteroscopy but none were confirmed on curettage. Of the 12 patients (6.9%) with histological diagnosis of cancer, the hysteroscopic study showed cancer in 3 patients (25%), hyperplasia in 8 patients (66.6%) and submucous fibroid as in 1 patient (8.3%). However, in 3 patients where hysteroscopic impression was cancer, all 3 were confirmed as cancer on histology too.

Hysteroscopy was highly specific in diagnosing cancer however sensitivity was low. The sensitivity of hysteroscopy in diagnosing cancer was 16.7%, specificity was 100%, positive predictive value (PPV) being 100% and negative predictive value (NPV) was 94.2%. Hysteroscopy was found to be highly specific in diagnosing normal endometrium (92.2%), atrophy (96.9%), polyp (95.9%) and cancer (100%). It was found to have low specificity in diagnosing hyperplasia (48.4%). The sensitivity of hysteroscopy in detecting normal endometrium and cancer were 21.2% and 16.7% respectively which is extremely low. The sensitivity in diagnosing polyp was found to be 100% and hyperplasia was 84.9%. Overall, hysteroscopy was highly sensitive and specific in detecting polyp. Following table shows the sensitivity, specificity, PPV and NPV of all the parameters except submucous fibroid as none were confirmed by histological examination (Table 6).

Discussion

In gynecology, the most frequently encountered problem is abnormal uterine bleeding. Abnormal uterine bleeding contributes to 30-40% of all gynecological problems.

In our study, 16.6% had postmenopausal bleeding & the rest had other menstrual abnormalities as an indication for hysteroscopy. Among which, menorraghia (49.7%) being the commonest complaint which was found similar to study done in Bahrain⁴ where menorraghia was seen in 62% patients & postmenopausal bleeding in 14%. The age group ranged from 30 to 75years in our study. We did not include hysteroscopy done for infertility purpose. So in our study, the perimenopausal age group of 41 to 50 years formed the major bulk (54.3%). Most of the study group patients were multiparous (90.9%).

Most studies have shown hysteroscopy was more sensitive in diagnosing uterine abnormalities like polyps & submucous fibroid^{4, 5}. Even in our study, sensitivity in diagnosing polyp on hysteroscopy was 100% but submucous fibroid could not be assessed as therapeutic hysteroscopy was not attempted in the same sitting. On the other hand, hysteroscopy had a low sensitivity in diagnosing cancer (16.7%) and normal endometrium (21.2%). This is comparable with other study findings. Like in Sameera et al's study⁴, sensitivity of hysteroscopy in diagnosing cancer was 40% and Ben Yehoda et al's study⁶ was 20%. Torrjeon R et al reported hysteroscopy was 100% sensitive, with specificity 99.4% and global diagnostic precision 99.5% in diagnosing adenocarcinoma in premenopausal patients⁷. On the contrary, sensitivity of diagnosing endometrial hyperplasia was high (84.9%) in our study compared to 30% and 52% in Sameera and Ben Yehoda's reports. Hysteroscopy performed alone has reported high false positive rate for detecting endometrial hyperplasia. So hysteroscopy targeted biopsy or dilatation and curettage has been advisable⁸. Hysteroscopy showed high PPV (100%) and NPV (94.2%) in regard to cancer detection in our study. Even in postmenopausal bleeding, hysteroscopy has been reported to have an overall sensitivity of 97% and specificity of 98.6% in detecting endometrial pathologies⁹.

A similar study¹⁰ done in medical institute in Maharashtra concluded that hysteroscopy affords a more accurate diagnosis than dilatation and curettage for intrauterine pedunculated pathologies. But for hyperplasia and carcinoma endometrium, histopathology is 100% diagnostic. Diagnosis of endometrial atrophy is best made by hysteroscopy.

The limitations of this study were the hysteroscopy procedure was carried out by many gynecologists with different level of experience (junior resident to professor) and not by a consistent hysteroscopist, secondly it was a retrospective study and thirdly the curettage was performed after the hysteroscopy. Well, it is true you need to have a hysterectomy specimen for an accurate diagnosis but that is not feasible in all the cases. In our study, out of 175 cases, only 20 underwent hysterectomy at a later date. At last, the number of cases obtained in this study to analyze was limited for an accurate interpretation.

Conclusion

Hysteroscopy was more sensitive in detecting endometrial polyp, submucous fibroid and endometrial hyperplasia but it was found less sensitive than curettage in detecting cancer and normal endometrium. On the other hand, hysteroscopy was highly specific in conditions like endometrial cancer, polyp, atrophic and normal endometrium. Hysteroscopy guided biopsy and histopathology compliment each other in evaluation of a patient with abnormal uterine bleeding for accurate diagnosis and further treatment.