Women with recurrent postmenopausal bleeding should be re-investigated but are not more likely to have endometrial cancer

Abstract

Objectives

We aimed to determine the prevalence of endometrial cancer in patients presenting with recurrent postmenopausal bleeding (PMB) after initial negative investigations.

Study design

This was a retrospective study carried out in a dedicated PMB clinic. All patients presenting with recurrent PMB between 1 January 2003 and 31 May 2009 were studied. All patients had transvaginal ultrasound, speculum examination and endometrial sampling attempted.

Main outcome measure

The prevalence of endometrial cancer in patients presenting with recurrent PMB after initial negative investigations.

Results

A total of 1536 women with PMB were seen over a period of 56 months. Sixty-six required definitive treatment after their first visit. One hundred and twenty-six re-presented with recurrent bleeding. The prevalence of endometrial cancer in women presenting for the first time with PMB was 3.04% whereas that in women presenting with recurrent PMB after initial negative investigations was 4%. One of these women re-presented within six months of initial visit and another within one year. All the other cancers were diagnosed more than one year after initial negative investigations (range 19.6–57.1 months). One woman had an endometrial cancer even though the endometrial thickness was less than 3 mm.

Conclusions

Women with recurrent PMB after initial negative investigations are no more likely to have endometrial cancer than those presenting for the first time but re-investigation is indicated if six months has elapsed.

Keywords

Recurrent postmenopausal bleeding transvaginal ultrasound endometrial biopsy endometrial cancer postmenopausal bleeding

Introduction

Postmenopausal bleeding (PMB) is one of the most common reasons for referral to gynaecological services. All women presenting with PMB require to be investigated to exclude lower genital tract malignancy.^1–3

The Scottish Intercollegiate Guideline Network (SIGN) recommend transvaginal ultrasound scan to estimate endometrial thickness (ET), but only if the ET is greater than 3 mm should endometrial biopsy, either with endometrial samplers or with formal hysteroscopy and curettage (H&C), be performed.⁴

All investigations for PMB carry a false-negative rate for malignancy^5–8 and hence it is important to re-investigate women with recurrent PMB, but evidence to dictate the time interval is lacking.⁴ Furthermore, there is a belief that women presenting with recurrent PMB are more likely to develop endometrial cancer than those with single episode of PMB, but there is no evidence in the literature to support this.^9,10

The aim of this study was to determine the prevalence of endometrial cancer in patients presenting with recurrent PMB after initial negative investigations.

Methods

This was a retrospective study carried out in the South Glasgow PMB Clinic. It is a dedicated PMB clinic where all the patients' data (demographic details, medical history, ultrasound finding, biopsy result and final diagnosis) are collected on a password-protected Microsoft Access database. All patients presenting with recurrent PMB between 1 January 2003 and 31 May 2009 were studied.

All patients presenting with PMB to the clinic underwent transvaginal ultrasound scan (TVS), followed by clinical examination including speculum examination and attempted endometrial biopsy using a Pipelle^TM, regardless of ET. Those with a scan suggestive of an endometrial polyp underwent formal H&C under general anaesthesia. Similarly patients with ET of more than 3 mm in whom we were unable to perform Pipelle underwent H&C. Patients on tamoxifen underwent H&C directly as per the SIGN guideline.

Patients presenting to the clinic with recurrent PMB after initial negative investigations underwent the same investigations. These patients were identified from the pre-existing database and were the basis of the study.

Mean values and standard deviations were determined using SPSS (Version 8.0) statistical program and data were analysed by the authors using Student's t-test and χ ²-test as appropriate.

Results

During the study period 1536 individual women were seen at the clinic. Of the women in whom the initial investigations were negative, 126 (8.2%) re-presented with PMB during the study period. Hence, 1410 women only had one visit to the clinic, regardless of the outcome of any pathology.

It was possible to perform a TVS in all but three women who were virgo intacta. Hence, 99.8% had a TVS. It was impossible to perform a Pipelle biopsy in 406 women (26.4%): 274 because of cervical stenosis, 20 because of an inability to see the cervix because of prolapse/obesity and 112 because of an inability to see the cervix because of discomfort.

Of the 1410 women who only had one visit, pathology warranting immediate definitive treatment was found in 66 (4.7%: 43 with endometrial cancer, 9 with atypical endometrial hyperplasia, 7 with ovarian cancer, 6 with cervical cancer and 1 with fallopian tube cancer). Their mean age at diagnosis was 66.1±9.1 years (range, 48.4–82.2 years). The prevalence of endometrial cancer in this group was 3.04%.

One hundred and twenty-six women (8.2%), in whom the initial investigations were negative, re-presented with PMB during the study period. Of these, 112 had only one further episode of bleeding and 14 had more than one episode. Their mean age at first visit was 59.2±8.9 years (range, 41.4–85.5 years).

Overall, 282 women (20%) were on hormone replacement therapy (HRT) at their first visit. In the group with recurrent PMB, 27 (21%) were on HRT at first presentation.

Table 1 shows that there was no difference in age, body mass index or mean ET in women who only had one visit compared with those who had recurrent PMB after initial negative investigations.

Table 1

Comparison of women who only had one visit and the initial investigations were negative with those who had recurrent PMB

	Single visit, negative investigations	Women with recurrent PMB	Difference in means (95% confidence interval for difference in means) [P]
N	1410	126
Age (years)	59.9±10.5	59.2±8.9	0.7 (⁻1.0–⁺2.4) [P = 0.41]
BMI (kg/m²)	28.1±6.6	28.2±7.8	⁻0.1 (⁻1.5–⁺1.30 [P = 0.89]
Endometrial thickness (mm)	4.0±5.0	4.8±4.9	⁻0.8 (⁻1.7–⁺0.1) [P = 0.08]

PMB, postmenopausal bleeding; BMI, body mass index

Data presented as mean±standard deviation

The mean ET for the women with recurrent PMB at their subsequent visit was 4.3±4.2 mm (95% confidence interval for the difference in means: ⁻1.6–⁺0.6 comparing first and subsequent visits, P = NS). The mean time interval for women representing with recurrent PMB was 21.4±14.5 months (range: 2.0–61.6 months). Eight women (6%) were referred within six months of their initial visit, 34 (27%) within one year and the remainder referred after 12 months from initial visit.

In the recurrent PMB group, six women (5%) required subsequent hysterectomy (5 with endometrial cancer and one with atypical hyperplasia). Their mean age at the first visit (negative test) was 69.5±12.5 years and at diagnosis (positive test) 71.5±11.2 years (range, 59.2–85.6 years). The prevalence of endometrial cancer in this group of women was 4% (chi-squared = 0.32, P = 0.43). Hence women presenting with recurrent PMB were no more likely to have significant pathology compared with women presenting for the first time and of those who did have significant pathology at the subsequent visit, their age was no different from the 66 women who had significant pathology diagnosed at their first visit.

The mean time interval from first visit (negative investigations) to diagnosis of significant pathology at the subsequent visit was 24.3±19.5 months. Only one woman with a subsequent endometrial carcinoma presented within six months of her first visit. At her first visit the ET was 9.6 mm, but it was impossible to pass a Pipelle and she had a hysteroscopy which showed an atrophic cavity. She was referred again by her general practitioner 133 days later (4.3 months). This time the ET was 14.8 mm and repeat hysteroscopy confirmed a cancer. At hysterectomy she had a bicornuate uterus with the cancer confined to the inner half of one uterine horn.

One woman had a cancer diagnosed eight months after an initial negative hysteroscopy (ET 11.3 mm but was unable to tolerate a speculum examination). The final histology (ET 12.0 mm) showed an otherwise normal uterus with the cancer confined to the inner third of the myometrium but, other than human error, we cannot explain the misdiagnosis at first procedure.

All the other cancers were diagnosed more than one year after initial negative investigations (range 19.6–57.1 months). One woman had ET of 9 mm at her first visit. Pipelle biopsy could not be carried out due to cervical stenosis and she underwent formal H&C at which a benign polyp was removed. At the second visit the ET was 2 mm but it was possible to pass a Pipelle and the subsequent histology showed malignant cells.

Discussion

In our study the prevalence of endometrial cancer in women presenting for the first time with PMB was 3.04% and overall, 4.7% required definitive treatment for malignant or premalignant disease. In comparison, the prevalence of endometrial cancer in women from this original cohort, presenting with recurrent PMB after initial negative investigations was 4%. There was no statistically significant difference in the percentage of women with endometrial cancer in the two groups. Two women with recurrent PMB investigated within one year of initial negative investigations (both had hysteroscopy) had a cancer but, based on the ultrasound of their ET, it is probable that both of them had cancers at the initial visit but these were missed.

The SIGN guidelines recommend that endometrial sampling should be performed only if the ET is more than 3 mm, but they acknowledge that all methods of endometrial assessment will miss cancers and no ET threshold on ultrasound can completely exclude early endometrial carcinoma.⁴ Furthermore, all endometrial biopsy techniques carry false-negative rates.^11–13 Hence it is important to re-investigate women presenting with recurrent PMB.

A Pipelle^TM sampler is inexpensive and adds little to the time of an assessment and we have shown previously that even if the ET is less than 3 mm, malignancy may be present.¹⁴ It is assumed that women presenting with recurrent PMB carry an increased risk of developing endometrial cancer than those presenting for the first time, but there are few studies in the literature addressing this issue.

Gull et al. followed up 339 women with PMB for 10 years. They concluded that women with recurrent PMB were at higher risk of developing endometrial cancer. However, those women with ET of ≤4 mm at initial scan did not have increased risk of endometrial cancer in spite of recurrent PMB.⁹

Van Doorn et al. followed up 249 women with PMB who had ET of ≤4 mm at initial scan. Twenty-five out of 249 patients had recurrent PMB and two of these (8%) had endometrial cancer. From this small sample, they concluded that women with recurrent PMB bore a considerable risk of carcinoma.¹⁰

Fung et al. ¹⁵ performed 40 hysterectomies for persistent PMB, but there were no cases of endometrial cancer diagnosed in the final pathology. Gull et al. followed up 163 women referred with PMB who had ET ≤ 4 mm for 12 months. Endometrial biopsy was performed because of recurrent bleeding in 6% and no cancer was diagnosed during the follow-up period.¹⁶

It has previously been presumed that women with recurrent PMB were more likely to have endometrial cancer than those presenting for the first time, but this study shows that the incidence is no different and women can be reassured accordingly. Nevertheless, two cancers were missed at initial hysteroscopy. Neither had advanced disease at the time of definitive diagnosis, but it is impossible to determine the subsequent extent of spread if their symptoms had been ignored so we believe that, even if initial investigations are negative, women should be re-investigated for recurrent PMB if six months has elapsed.

Competing interests

None declared.

References

Danero

, Ricci

, La Rosa

, Critical review of dilatation and curettage in the diagnosis of malignant pathology of the endometrium. Eur J Gynaecol Oncol 1986;7:162–5

O'Connell

, Fries

, Zeringue

, Brehm

. Triage of abnormal postmenopausal bleeding: a comparison of endometrial biopsy and transvaginal sonohysterography versus fractional curettage with hysteroscopy. Am J Obstet Gynecol 1998;178:956–61

Smith-Bindman

, Kerlikowske

, Feldstein

, Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280:1510–7

Scottish Intercollegiate Guideline Network. Investigation of post-menopausal bleeding. Guideline 2002;61

Haller

, Matecjcic

, Rukavina

, Transvaginal sonography and hysteroscopy in women with postmenopausal bleeding. Int J Gynaecol Obstet 1996;54:155–9

Guner

, Tiras

, Karabacak

, Endometrial assessment by vaginal ultrasonography might reduce endometrial sampling in patients with postmenopausal bleeding: a prospective study. Aust NZ J Obstet Gynaecol 1996;36:175–8

Tsuda

, Kawabata

, Umesaki

, Endometrial assessment by transabdominal ultrasonography in postmenopausal women. Eur J Obstet Gynecol Reprod Biol 1993;52:201–4

Sladkevicius

, Valentin

, Marsal

. Endometrial thickness and Doppler velocimetry of the uterine arteries as discriminators of endometrial status in women with postmenopausal bleeding: a comparative study. Am J Obstet Gynecol 1994;171:722–8

Gull

, Karlsson

, Milsom

, Granburg

. Can ultrasound replace dilation and curettage? A longitudinal evaluation of postmenopausal bleeding and transvaginal sonographic measurement of the endometrium as the predictors of endometrial cancer. Am J Obstet Gynecol 2003;188:401–8

10.

Van Doorn

, Timmermans

, Opmeer

What is the recurrence rate of postmenopausal bleeding in women who have thin endometrium during a first episode of postmenopausal bleeding?

Acta Obstet Gynecol Scand 2008;87:89–93

11.

Stock

, Kanbour

. Prehysterectomy curettage. Obstet Gynecol 1975;45:537–41

12.

MacKenzie

, Bibby

. Critical assessment of dilatation and curettage in 1029 women. Lancet 1978;2:566–8

13.

Stoval

, Solomon

, Ling

. Endometrial sampling prior to hysterectomy. Obstet Gynecol 1989;73 (3 Pt 1):405–9

14.

Sulaiman

, Chong

K.-W.

, Gaudoin

. One-stop postmenopausal bleeding clinics reduce patient waiting times and operating costs. Scot Med J 2004;49:152–4

15.

Fung

, Fung

, Burnett

, Faught

. Does persistent postmenopausal bleeding justify hysterectomy? Eur J Gynaecol Oncol 1997;18:26–8

16.

Gull

, Carlsson

, Karlsson

Transvaginal ultrasonography of the endometrium in women with postmenopausal bleeding: is it always necessary to perform an endometrial biopsy?

Am J Obstet Gynecol 2000;182:509–15