Outcomes of endometrial assessment in women with unscheduled bleeding on hormone replacement therapy

Abstract

Objective

This study correlates the transvaginal ultrasound findings with histopathology results in women who present with unscheduled bleeding on hormone replacement therapy.

Study design

Retrospective analysis of 469 consecutive cases with unscheduled bleeding on hormone replacement therapy (203 patients on sequential hormone replacement therapy (seq-HRT) and 266 patients on continuous combined hormone replacement therapy (con-HRT)).

Main outcome measures

Outcomes of endometrial assessment in women with unscheduled bleeding on hormone replacement therapy.

Results

Normal appearance of the endometrium on pelvic ultrasound was seen in 62% patients on seq-HRT and 43% of women on con-HRT. These women required no further assessment and were discharged. Histological assessment showed normal endometrial tissue in 22% of women on seq-HRT and 22% of con-HRT group. Benign endometrial polyps were noted in 8% of women on seq-HRT versus 18% of women on con-HRT. Hyperplasia without atypia was noted in 0.5% of woman on seq-HRT versus 0.4% of women on con-HRT while atypical hyperplasia/endometrial cancer was noted in 2% of women on seq-HRT versus 1% of women on con-HRT.

Conclusion

Women who present with unscheduled bleeding on hormone replacement therapy both on sequential and continuous combined regimens can be reassured that the risk of pathology is low.

Keywords

Bleeding histopathology hormone replacement therapy

Introduction

Hormone replacement therapy (HRT) is indicated for managing menopausal symptoms in peri- and post-menopausal women. Unscheduled bleeding is a common side-effect and a major factor in women ceasing HRT.¹ Sequential HRT (seq-HRT) gives up to 90% of women a cyclical bleed whereas women on continuous combined HRT (con-HRT) preparations are expected to be amenorrhoeic due to endometrial atrophy.^2,3 Unscheduled bleeding is defined as abnormal bleeding on seq- or con-HRT six months post initiation of therapy or after a period of amenorrhoea.⁴ An initial phase of irregular bleeding is common in women in the first six months of treatment² while persistent bleeding after this period of treatment should always be investigated.³ A change in bleeding pattern including heavy and/or painful bleeding or change in the timing of the withdrawal bleed should also be investigated.⁵ The National Institute for Health and Care Excellence directs healthcare professionals to advise women to report unscheduled vaginal bleeding at the three-month review appointment or promptly if it occurs after the first three months.⁶ The Scottish Intercollegiate Guidelines Network⁷ suggests that women with unscheduled bleeding need to be referred for investigations six months post initiation of therapy or after established amenorrhoea.

Benign pathology such as endometrial polyps and atrophic vaginitis are common differential diagnosis. The incidence of endometrial hyperplasia is low and has been reported between 1% and 5% in previous studies.⁸ In women on seq-HRT, for more than five years, there has been an association with a small increase in endometrial hyperplasia. In contrast, those on con-HRT have been reported to have a significantly lower risk of endometrial cancer.⁹

Studies have shown that up to 50% of women cease their HRT within 6–12 months due to unscheduled bleeding,¹ with complaints of unscheduled bleeding reported in 38% of women on seq-HRT and 41% of women con-HRT.¹⁰

Initial investigations of the woman with unscheduled bleeding on HRT includes: a pelvic examination; vulval examination for signs of vulval conditions such as urogenital atrophy, lichen sclerosus/planus or vulval cancer; speculum examination to rule out cervical polyps or cervical cancer and to obtain a cervical smear test if indicated as well as a genital infection if the history suggests this. Further investigation when warranted includes a transvaginal ultrasound scan (TVUS) to assess the endometrial thickness (EMT) and morphology, the uterus for pathology and the adnexa for ovarian/tubal tumours.

In women who present with post-menopausal bleeding, a normal EMT has been agreed at a cut-off of <4 mm on TVUS.^1–13 In our unit, an outpatient endometrial biopsy is obtained at an EMT ≥5 mm in women with unscheduled bleeding on HRT in line with the The British Gynaecological Cancer Society Uterine Cancer guideline.¹⁴ A careful assessment of the morphology of the endometrium and the timing of a withdrawal bleed is made in those patients on seq-HRT.¹⁰

This study aims to investigate the prevalence of endometrial pathology in women who present with unscheduled bleeding on HRT in our population.

Methods

This is a retrospective analysis of a consecutive case series of 469 patients presenting with unscheduled bleeding on HRT (203 patients on seq-HRT and 266 patients on con-HRT) at a specialist menopause clinic in a London teaching hospital. Cases were identified from our clinical database (Viewpoint, GE Healthcare). The demographics of the patients and type of HRT use were recorded. Presenting complaint was recorded as single, multiple or ‘other’ episodes of unscheduled bleeding on HRT when the number of episodes could not be determined from the captured data. TVUS assessment of the endometrium was performed by experienced gynaecologists. The EMT was measured as the maximum measurement in the sagittal plane with inclusion of both endometrial layers. The following were assessed: whether the endometrium was clearly visualised or not, the EMT if clearly visualised, the presence or absence of a suspected endometrial polyp or inadequate visualisation of the endometrium for which a hysteroscopy would be requested. The clinical protocol was to investigate the following groups of women by endometrial biopsy±hysteroscopy: (1) women with an EMT ≥5 mm or repeat the ultrasound scan assessment after a withdrawal bleed in those on seq-HRT, (2) women with evidence of an endometrial polyp and (3) women with an inadequate visualisation of the endometrium. The histopathological diagnoses were categorised as normal endometrium, endometrial polyp, hyperplasia without atypia, atypical hyperplasia/cancer and insufficient sample.

Results

The mean age for the entire study cohort was 57 years old ± 7 years, SD (range 45–85 years). Patients were similar in their demographic characteristics (Table 1).

Table 1.

Demographics of patients in both seq- and con-HRT groups.

Patient characteristics	Seq-HRT (n = 203)	Con-HRT (n = 266)
Age (years)^a	54 (45–76)	56 (45–85)
Menopause years^a	5 (2–10)	8 (3–13)
Parity^b
P0	64 (32%)	85 (32%)
P1	47 (23%)	34 (13%)
P2	38 (19%)	73 (27%)
P ≥ 3	54 (26%)	74 (28%)
• Unscheduled bleeding
• Single episode	11 (5%)	64 (24%)
• Multiple episodes	37 (18%)	67 (25%)
• ‘Other’ episodes^c	155 (77%)	135 (51%)

^aAge/menopause years given as median (range).

^bParity.

^cNumber of unscheduled bleeding episodes not specified.

Normal appearance of the endometrium on TVUS was seen in 125/203 (62%) patients on seq-HRT, median thickness of 3.9 mm ± 1.7 mm, SD (range 1–10.6 mm). In the con-HRT group, 115/268 (42%) patients had a normal endometrium, median thickness of 3.1 mm ± 0.9 mm, SD (range 0.8–4.9 mm) (p < 0.01, 95% confidence interval (CI) 1.5–3.1). These patients were discharged as per protocol.

On TVUS, a thick EMT was seen in 53/203 (26%) of patients on seq-HRT, with a median EMT of 8.3 mm ± 4.9 mm, SD (range 5 mm–29.5 mm). In comparison, 110/266 (41%) of patients in con-HRT group had a thick endometrium measuring 7.1 mm ± 2.7 mm, SD (range 5.1 mm–24.1 mm) (p < 0.01, 95% CI 0.3–0.7). Endometrial polyps were suspected on scan in 25/203 (12%) patients on seq-HRT as compared to 41/266 (15%) patients on con-HRT (p = 0.3, 95% CI 0.3–0.6).

An outpatient endometrial biopsy was obtained in 30/203 (15%) patients on seq-HRT versus 56/266 (21%) in the con-HRT. In those who were unable to tolerate or with failed outpatient endometrial biopsy, a hysteroscopy was carried out in 48/203 (24%) versus 95/266 (36%) in each group, respectively.

In those patients with abnormal TVUS results (EMT ≥5 mm or polyps), the histology results revealed 45/78 (58%) versus 58/151 (38%) with normal endometrial tissue in the seq-HRT and con-HRT, respectively. Benign endometrial polyps were diagnosed in 17/78 (22%) patients on seq-HRT versus 47/151 (31%) patients on con-HRT. There were 11/78 (14%) of patients in the seq-HRT and 42/151 (28%) in the con-HRT group with insufficient samples. There were 1/78 (1%) and 1/151 (1%) patients who had hyperplasia without atypia in each group, respectively. A total of 4/78 (5%) patients were diagnosed with atypical hyperplasia/endometrial cancer in the seq-HRT group, with 1/78 (1.3%) of those having atypical hyperplasia. In the con-HRT group there were 3/151 (2%) patients diagnosed with endometrial cancer. The histopathology results are summarised in Table 2.

Table 2.

Outcomes of assessment.

Outcomes	Seq-HRT ^a (n=203)	Con-HRT ^a (n=266)	P-value	OR, 95% CI^b
Scan
Normal scan	125 (62%)	115 (43%)	<0.001	2.1, 1.5–3.1
Abnormal scan	78 (38%)	151 (57%)	<0.001	0.5, 0.33–0.69
	Seq-HRT ^a (n=78)	Con-HRT ^a (n=151)	P-value	OR, 95% CI^b
Histopathology
Normal	45	58	0.92	1.1, 0.65–1.58
	(58%)^c	(38%)^c	<0.01	2.2, 1.3–4
	(22%)^d	(22%)^d	1	1, 0.5–1.9
Benign endometrial polyp	17	47	<0.001	0.4, 0.24–0.75
	22%)^c	(31%)^c	0.1	0.3–1.2
	(8%)^d	(18%)^d	0.04	0.4, 0.2–1
Insufficient	11	42	<0.001	0.1, 0.07–0.43
	(14%)^c	(28%)^c	0.02	0.4, 0.2–0.86
	(5%)^d	(16%)^d	0.02	0.3, 0.1-0.8
Hyperplasia (without atypia)	1	1	0.9	1.2, 0.07–18.8
	(1%)^c	(0.7%)^c	1	1, 0.06–16
	(0.5%)^d	(0.4%)^d	1	1, 0.1–16
Endometrial cancer	4	3	0.5	1.8, 0.29–10.6
	(5%)^c	(2%)^c	0.3	2.6, 0.5–13
	(2%)^d	(1%)^d	0.5	2, 0.2–22

^aGiven as n (%).

^bOdds ratio and 95 percent confidence intervals of proportion tabulated.

^cPercentage of women who required endometrial biopsy.

^dPercentage of total women with bleeding on Seq-/Con-HRT.

Discussion

This retrospective study investigated the findings of endometrial assessment in women who presented with unscheduled bleeding on HRT. The results showed that endometrial pathology was low in women who presented with unscheduled bleeding on HRT both with seq- as well as con-HRT regimens.

The majority of women who presented with unscheduled bleeding on HRT had a normal endometrium on TVUS, with a significantly higher proportion of women on seq-HRT having a normal appearance of the endometrium on ultrasound scan. Equally there was a significantly higher number of patients in the con-HRT group with a thick endometrium on pelvic ultrasound scan. The findings of polyps were comparable in both groups.

Overall, approximately half of the patients in the study required an endometrial outpatient biopsy or hysteroscopy for endometrial assessment. Of those, half had a normal endometrium on histopathology. Patients on con-HRT had a significantly higher proportion of endometrial polyps and of insufficient samples on histology. There were six cases of cancer diagnosed in this study, three diagnosed in women having seq-HRT and three in women taking con-HRT.

The findings in our study can help clinicians counsel and alleviate women’s anxieties around bleeding on HRT. It is reassuring that the histopathology diagnosis in most patients who presented with unscheduled bleeding yielded a low risk of endometrial pathology, including cancer. This is useful information to help triage women appropriately and reduce the risk of unnecessary investigations.

The retrospective data presented above has not allowed us to differentiate as to whether a patient was scanned before or after a withdrawal bleed in those on seq-HRT. Despite these limitations, and the confounding factors of a retrospective study, the data provide an objective analysis into the histopathology diagnoses following episode(s) of unscheduled bleeding.

Endometrial cancer is present in 3–10% of women who present with post-menopausal bleeding, with more than 90% experiencing bleeding as a first sign.¹⁴ In women on HRT, this risk is lower as reported in previous studies. In a study by Burbos et al.,³ 4847 post-menopausal women with bleeding were investigated with an endometrial biopsy, including 750 (15%) women on HRT. The study found that women using HRT preparations were significantly less likely to be diagnosed with endometrial cancer (p < 0.001). This study did not differentiate between patients on seq- or con-HRT. The Million Women Study⁵ reported a reduced risk of endometrial cancer in those on con-HRT, and no difference in those on seq-HRT as compared to women on no HRT. Furthermore, in the Osteoporosis Prevention and Arterial effects of tiboLone¹⁵ study the effects of tibolone, con-HRT and placebo on the endometrium was assessed. There was no difference found in the risk of developing endometrial proliferation, hyperplasia or cancer between the three groups during the three-year study period. Our study found that overall 1.3% of patients on HRT with unscheduled bleeding were diagnosed with endometrial cancer. This is lower than the endometrial cancer background risk in patients not taking HRT.

The overall prevalence of endometrial hyperplasia in women on HRT is also low. Most endometrial samples from women on seq-HRT show weakly proliferative changes.⁸ The prevalence of endometrial hyperplasia associated with seq-HRT has been reported as anything from 5.4% to 16%.^8,16 The risk of endometrial hyperplasia has been reported to be lower in those women on con-HRT (0.8%)¹⁷ and may help to normalise the endometrium of women who have developed hyperplasia on sequential therapy.⁸ In our study population, we found a low prevalence of endometrial hyperplasia without atypia (0.5%).

The EMT threshold for which to obtain an endometrial biopsy in women with unscheduled bleeding on HRT is not well defined. The probability of malignancy with an EMT <5 mm is less than 1% in women with post-menopausal bleeding.¹⁸ The Postmenopausal Estrogen/Progestin Interventions (PEPI)¹⁹ study found that women on HRT who were asymptomatic with an EMT of <5 mm had a high negative predictive value (99%) for pathology and found no cases of atypical hyperplasia or cancer in this group. In a large multi-center study, Granberg et al.²⁰ reported the histopathology findings of women who presented with unscheduled bleeding with con-HRT (n = 202) as compared to those on estriol only or no HRT. Across all three groups, endometrial pathology was more likely to be found with an EMT >8 mm and endometrial cancer was diagnosed more frequently in women not receiving HRT (p < 0.001).^20–22

More than two-thirds of endometrial biopsy samples in women with an EMT of <5 mm on TVUS are reported as insufficient.¹⁸ In our study, there was a significantly higher proportion of women on con-HRT with an insufficient sample (p < 0.001) which is to be expected as con-HRT induces atrophy of the endometrium.

On the contrary, the endometrium in patients on seq-HRT showed a similar pattern to pre-menopausal women during the menstrual cycle. These women have a maximum EMT on days 13–23 with a greater variation than women on con-HRT and controls.^4,16 It is therefore preferable to scan women on seq-HRT after a withdrawal bleed as the endometrium is then comparable to the endometrium of women on con-HRT.¹ A large meta-analysis by Smith-Bindman et al.²² concluded that an EMT ≤5 mm can exclude endometrial pathology in women with vaginal bleeding on HRT with a 91% sensitivity and 77% specificity (95% CI 89–93). As there have been no large prospective studies to determine a cut-off EMT in women with unscheduled bleeding on HRT, an EMT of <5 mm can be used to reassure and safely discharge women on both seq- and con-HRT.

Following endometrial biopsy with pipelle and/or hysteroscopy, up to 70% of patients in one study discontinued treatment.¹ An episode of unscheduled bleeding by itself whilst taking HRT does not signify the development of endometrial abnormalities;²³ however, persistent bleeding should always be investigated. In addition to carrying out diagnostic tests in patients with unscheduled bleeding on HRT, consideration should also be given to reviewing the HRT regimen and modifying it where required to improve compliance and lower the chances of women discontinuing treatment.

Conclusion

Women who present with unscheduled bleeding with seq- or con-HRT regimens can be reassured that the risk of pathology is low. This information can be useful in counselling women and guiding clinicians.

Footnotes

Acknowledgements

We would like to thank Dr F Baird and Miss S Trainor for their assistance in data analysis.

Contributorship

SAI and HH researched literature and conceived the study. SAI wrote the first draft of the manuscript. HH reviewed and edited the manuscript and approved the final version of the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

The study was approved by King’s College Hospital Research and Development Department and was exempt from Ethics Committee approval.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Guarantor

SAI.

References

Jurkovic

Valentin

L and

Vyas

Gynaecological ultrasound in clinical practice: ultrasound imaging in the management of gynaecological conditions. London: RCOG, 2009.

Mattson

LÅ

Ipsen

Granqvist

et al. ; Study Group. Ultra-low-dose estradiol and norethisterone acetate: bleeding patterns and other outcomes over 52 weeks of therapy. Climacteric 2015; 18: 419–425.

Burbos

Musonda

Duncan

et al . Postmenopausal vaginal bleeding in women using hormone replacement therapy. Menopause Int 2012; 18: 5–9.

Romer

Hormone replacement therapy and bleeding disorders.

Gynecol Endocrinol 2006; 22: 140–24.

Whitehead

M and

Farmer

The Million Women Study: a critique.

Endocrine 2004; 24: 187–293

National Institute for Health and Care Excellence (NICE). Menopause: diagnosis and management. UK: National Institute for Health and Care Excellence, 2015.

Intercollegiate Guidelines Network. Investigation of post-menopausal bleeding, a national clinical guideline. Edinburgh: Scottish Intercollegiate Guidelines Network, 2002.

Feeley

Wells

Hormone replacement therapy and the endometrium.

J Clin Pathol 2001; 54: 435–440.

Hamoda

Panay

Arya

et al . The British Menopause Society & Women’s Health Concern 2016 recommendations on hormone replacement therapy in menopausal women. Post Reprod Health 22.

10.

Ettinger

Klein

Unexpected vaginal bleeding and associated gynecological care in postmenopausal women using hormone replacement therapy: comparison of cyclic versus continuous combined schedules.

Fertil Steril 1998; 69: 865–869.

11.

Otify

Fuller

Ross

et al . Endometrial pathology in the postmenopausal woman – an evidence based approach to management. Obstet Gynaecol 2015; 17: 29–38

12.

Sladkevicius

Valentin

Ultrasound assessment of endometrial morphology and vascularity to predict endometrial malignancy in women with postmenopausal bleeding and sonographic endometrial thickness ≥4.5 mm.

Ultrasound Obstet Gynaecol 2007; 30: 332–340.

13.

The British Gynaecological Cancer Society (BGCS). Uterine cancer guidelines: recommendations for practice. London: The British Gynaecological Cancer Society.

14.

Doubilet

PM.

Society of Radiologists in Ultrasound consensus conference statement on postmenopausal bleeding.

J Ultrasound Med 2001; 20: 1037–1042.

15.

Langer

Landgren

Rymer

et al. ; OPAL Investigators. Effects of tibolone and continuous combined conjugated equine estrogen/medroxyprogesterone acetate on the endometrium and vaginal bleeding: results of the OPAL study. Am J Obstet Gynecol 2006; 195: 1320–1327

16.

Levine

Gosink

Johnson

LA.

Change in endometrial thickness in postmenopausal women undergoing hormone replacement therapy.

Radiology 1995; 197: 3.

17.

Brynhildsen

Hamar

Low dose transdermal estradiol/norethisterone acetate treatment over 2 years does not cause endometrial proliferation in postmenopausal women.

Menopause 2002; 2: 137–144.

18.

Van den Bosch

Van Schoubroeck

Domali

et al . A thin and regular endometrium on ultrasound is very unlikely in patients with endometrial malignancy. Ultrasound Obstet Gynecol 2007; 29: 674–679.

19.

Langer

Pierce

O’Hanlan

et al . Transvaginal ultrasonography compared with endometrial biopsy for the detection of endometrial disease. Postmenopausal Estrogen/Progestin Interventions Trial. N Engl J Med 1997; 337: 1792–1798.

20.

Granberg

Yliistalo

Wikland

et al . Endometrial sonographic and histologic findings in women with and without hormonal replacement therapy suffering from postmenopausal bleeding. Maturitas 1997; ▪: 35–40.

21.

Mossa

Imperato

Marziani

et al . Hormonal replacement therapy and evaluation of intrauterine pathology in postmenopausal women: a ten‐year study. Eur J Gynaecol Oncol 2003; 24: 507–512.

22.

Smith‐Bindman

Kerlikowske

Feldstein

et al . Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998; 280: 1510–1517.

23.

Omodei

Ferrazzi

Becorpi

et al . Endometrial evaluation with transvaginal ultrasound during hormone therapy: a prospective multicenter study. Fertl Steril 2004; 81: 1632–27.