clinical challenge  défi clinique

clinical challenge  défi clinique

Just the Berries

Who needs an endometrial biopsy?
Sheri-Lee Samson, MD Donna Gilmour, MD, FRCSC

E ndometrial biopsy is a way to sample the endome-

trium that can be done as an office procedure. It
is an important diagnostic tool in evaluating abnor-
Risk factors for endometrial cancer
Most endometrial cancers are related to chronic
unopposed endogenous or exogenous estrogen.
mal uterine bleeding. It is used to exclude diseases, Prolonged exposure to unopposed estrogen occurs
such as endometrial cancer and its precursors, espe- in obese patients and those with polycystic ovary syn-
cially atypical endometrial hyperplasia. Endometrial drome or chronic anovulation. Unopposed exogenous
biopsy is also used in evaluating infertile patients to estrogen is a serious risk factor, so postmenopausal
diagnose luteal phase defects. This article, however, women with a uterus should not be taking estrogen
focuses on the role of endometrial biopsy in deter- without appropriate progesterone. Other women at
mining causes of abnormal uterine bleeding. increased risk of endometrial cancer include those
with a relatively long period of exposure to estro-
Endometrial biopsy gen, such as women with early menarche and late
versus dilation and curettage menopause, and nulliparous women. The pattern of
The criterion standard for sampling the endometrium abnormal bleeding is also important because women
used to be dilation and curettage (D&C) under gen- with irregular menstrual cycles are at higher risk of
eral anesthetic. It is now recognized, however, that endometrial cancer than women with regular cycles.7
D&C is really just another blind technique that often Age is also an important risk factor for endometrial
samples less than half the endometrium.1 cancer; peak age for endometrial cancer is 60 years.8
Endometrial biopsy has replaced D&C as the first- Most physicians agree that women 40 years and older
line diagnostic test for evaluating abnormal bleeding. are at increased risk. Other risk factors include diabetes
The two tests have been shown to have similar accu- and hypertension. Women with diabetes have a 2.8-fold
racy; several studies comparing endometrial biopsy increased risk of endometrial cancer.9 In two studies,9,10
and D&C samples have hypertension was not an
found excellent agree- “Just the Berries” for Family Physicians originated at St independent risk factor
ment, from 83% to 90% Martha’s Regional Hospital in 1991 as a newsletter for when adjusted for body
of samples2-5 (level III members of the Department of Family Medicine. Its pur- weight. Women using
evidence). Endometrial pose was to provide useful, practical, and current informa- tamoxifen as adjuvant
tion to busy family physicians. It is now distributed by the
biopsy has several advan- Medical Society of Nova Scotia to all family physicians in therapy for breast cancer
tages over D&C. It is safer Nova Scotia. Topics discussed are suggested by family are also at increased risk of
because there is no need physicians and, in many cases, articles are researched and endometrial cancer11; risk
for general anesthetic written by family physicians. is highest in women who
or cer vical dilation, and Just the Berries has been available on the Internet for sev- have used tamoxifen for
eral years. You can find it at www.theberries.ns.ca. Visit the
there is markedly less site and browse the Archives and the Berries of the Week. more than 2 years.
risk of hemorrhage, infec- We are always looking for articles on topics of interest to
tion, and perforation.6 family physicians. If you are interested in contributing an Indications for
Endometrial biopsy is article, contact us through the site. Articles should be short endometrial biopsy
also more convenient and (350 to 1200 words), must be referenced, and must include Postmenopausal bleed-
levels of evidence and the resources searched for the data.
saves time for both physi- All articles will be peer reviewed before publication. ing. Any woman with
cians and patients. postmenopausal bleeding

Dr Samson is a third-year resident in Obstetrics and Gynecology at Dalhousie University in Halifax, NS. Dr Gilmour is a
urogynecologist and an Assistant Professor in the Department of Obstetrics and Gynecology at Dalhousie University.

VOL 48: MAY • MAI 2002  Canadian Family Physician • Le Médecin de famille canadien 885
 clinical challenge  défi clinique

who is not receiving hormone replacement therapy younger than 40 with no risk factors have minimal
(HRT) requires endometrial sampling. About 7% of risk of endometrial cancer.7 For these women, it
such bleeding is caused by malignancy,12 so post- might be appropriate to initiate hormone treatment
menopausal bleeding should be considered to be after an endocrine workup and sample the endome-
from endometrial cancer until proven otherwise. trium only if there is no response to treatment. On
Most women taking cyclic combined HRT experi- the other hand, women with one or more risk factors
ence regular withdrawal bleeding. They should be are at increased risk of endometrial cancer and prob-
counseled about this beforehand and reassured that ably should be sampled. The most important risk fac-
it is normal. Patients with irregular bleeding, not tor in premenopausal women is irregular menstrual
associated with progesterone withdrawal, should be cycles, which are associated with a 14% risk of abnor-
considered to have abnormal patterns of bleeding mal results of endometrial biopsy (including both
and should undergo endometrial sampling.6,8 benign and malignant lesions).7 Hence, endometrial
In women taking continuous combined HRT, the biopsy should be considered for almost all women
endometrium often becomes atrophic, and these with irregular cycles.
women become amenorrheic; 75% of these women
develop amenorrhea after 6 months.13 If bleeding Tamoxifen. Because of the estrogenlike effects of
continues beyond 6 months, it should be consid- tamoxifen on the endometrial lining, patients taking
ered abnormal, and endometrial sampling should tamoxifen who experience abnormal vaginal bleeding
be considered.6 The Society of Obstetricians and should have an endometrial biopsy.8
Gynaecologists of Canada recommends endometrial
biopsy for patients who continue to experience bleed- Abnormal Pap smear cytology. In postmenopausal
ing after the first 6 months of treatment with continu- women, presence of any endometrial cells on Pap
ous combined HRT.8 smear is an indication for endometrial sampling.6,9 In all
women, presence of atypical endometrial cells on Pap
Perimenopausal bleeding. During the perimeno- smear warrants endometrial biopsy. Patients with malig-
pausal years, waxing and waning ovarian function nant endometrial cells on Pap smear are at serious risk
affects menstrual cycles. Women often have cycles of endometrial cancer, often high-grade malignancy.15
that vary in length and fluctuating amounts and dura-
tion of flow. Anticipated changes in menstrual cycles Follow up. Women who have been treated for endo-
sometimes make it difficult to determine whether metrial hyperplasia with hormone therapy require
patients are experiencing abnormal uterine bleeding. a follow-up endometrial biopsy in 3 to 6 months to
Perimenopausal women with abnormal bleeding are ensure the hyperplasia has regressed.
at increased risk of endometrial cancer secondary to their
age and anovulatory cycles. All women with abnormal Screening with endometrial biopsy. Currently,
uterine bleeding in the perimenopausal period require there is no reason to screen women with no abnor-
endometrial sampling.9,14 The most worrisome patterns mal bleeding. Asymptomatic women are at low risk
are persistently increased menstrual flow, decreased of endometrial cancer,16 and endometrial biopsy has
menstrual interval, and intermenstrual bleeding.14 poor accuracy for identifying well differentiated, low-
volume, minimally invasive tumours, which makes it
Premenopausal bleeding. Indications for endome- a poor screening test.17
trial biopsy in premenopausal women with abnormal
bleeding are not straightforward. Adolescents gener- Accuracy
ally do not require sampling because their abnormal The accuracy of endometrial biopsy to detect endo-
bleeding is often due to anovulation secondary to an metrial disease, especially endometrial cancer, is
immature hypothalamic-pituitary-ovarian axis or, less extremely good. Studies comparing endometrial
commonly, to an underlying inherited coagulopathy. biopsy samples with specimens taken during hys-
After adolescence, endometrial cancer should be terectomy indicate that endometrial biopsy has a
considered in the differential diagnosis of abnormal sensitivity ranging from 83% to 96% for detection of
bleeding because up to 10% of women with endome- endometrial cancer.18-21
trial cancer are diagnosed before age 45.10
Risk factors for endometrial cancer should be Indications for D&C
considered when determining whether endometrial Although endometrial biopsy has replaced D&C as
biopsy is needed for premenopausal women. Women the first-line test for abnormal bleeding, there are still

886 Canadian Family Physician • Le Médecin de famille canadien  VOL 48: MAY • MAI 2002
 clinical challenge  défi clinique

circumstances where D&C is indicated. If an endo- References

metrial biopsy cannot be done, if the sample is insuf- 1. Stock RJ, Kanbour L. A prehysterectomy curettage. Obstet Gynecol 1975;45:537-41.
2. Fothergill DJ, Brown VA, Hill AS. Histological sampling of the endometrium—a
ficient, or if patients’ bleeding persists after negative
comparison between formal curettage and the Pipelle sampler. Br J Obstet
biopsy results, further investigation is needed. Gynaecol 1992;99:779-80.
When a biopsy cannot be per formed or a 3. Goldchmit R, Katz Z, Blickstein I, Caspi B, Dgani R. The accuracy of endometrial
sample is insuf ficient, patients should be tri- Pipelle sampling with and without sonographic measurement of endometrial thick-
ness. Obstet Gynecol 1993;82:727-30.
aged according to risk of endometrial cancer to
4. Kavak Z, Cayhan N, Pekin S. Combination of vaginal ultrasonography and Pipelle
determine the next appropriate step. If patients sampling in the diagnosis of endometrial disease. Aust N Z J Obstet Gynaecol
are at high risk of endometrial cancer, they 1996;36(1):63-6.
require D&C. If they are at relatively low risk or 5. Stovall TG, Photopoulos GJ, Poston WM, Ling FW, Sandles LG. Pipelle endo-

at increased risk of being adversely af fected by metrial sampling in patients with known endometrial carcinoma. Obstet Gynecol
1991;77:954-6.
anesthetics, a transvaginal ultrasound examina-
6. Chambers JT, Chambers SK. Endometrial sampling: Who? Where? Why? With
tion could be used to guide further management. what? Clin Obstet Gynecol 1992;35(1):28-39.
If the ultrasound examination demonstrates an 7. Farrell SA, Samson S, Ash S, Flowerdew G, Andreou P. Risk categories for abnormal
endometrial thickness of ≤ 4 mm, then the risk endometrial biopsy in dysfunctional uterine bleeding. J SOGC 2000;22(4):265-9.
8. Brand A, Duduc-Lissoir J, Ehlen TG, Plante M. Diagnosis of endometrial cancer in
of having serious endometrial disease is low, and
women with abnormal vaginal bleeding. SOGC clinical practice guidelines.
fur ther investigation is probably not required, J SOGC 2000;22(1):102-4.
other than close follow up.8,22 If the thickness is 9. Apgar BS, Newkirk GR. Endometrial biopsy. Prim Care 1997;24(2):303-26.
> 4 mm, patients probably require D&C.8,23 10. Brenton LA, Berman ML, Mortel R, Twiggs LB, Barrett RJ, Wilbanks GD, et al.

If the results of an endometrial biopsy are nega- Reproductive, menstrual, and medical risk factors for endometrial cancer: results
from a case-control study. Am J Obstet Gynecol 1992;167:1317-25.
tive and patients continue to have abnormal bleeding, 11. Fornander T, Rutquist LE, Cedermark B, Glas U, Mattson A, Silfversward C, et
studies show that there is a 10% to 11% risk that a al. Adjuvant tamoxifen in early breast cancer: occurrence of new primary cancers.
lesion has been overlooked.8,24 Such patients need to Lancet 1989;1(8630):117-20.

be reevaluated. Options for further evaluation include 12. Choo YC, Mak KC, Hsu C, Wong TS, Ma HK. Postmenopausal uterine bleeding
of nonorganic cause. Obstet Gynecol 1985;66:225-8.
repeating endometrial biopsy, transvaginal ultra-
13. Udeff L, Langenberg P, Adashi EY. Combined continuous hormone replacement
sound examination, or hysteroscopy and D&C. therapy: a critical review. Obstet Gynecol 1995;86:306-16.
14. Bealy PS. Diseases of the uterus. In: Scott JR, Danforth DN, editors. Danforth’s
Conclusion obstetrics and gynecology. 8th ed. Philadelphia, Pa: Lippincott, Williams & Wilkins;
1999. p. 837-55.
Endometrial biopsy is an important diagnostic tool
15. DuBeshtes B, Warshal DP, Angel C, Dvoretsky PM, Lin JY, Raubertas RF.
in investigating abnormal uterine bleeding. It has Endometrial carcinoma: the relevance of cervical cytology. Obstet Gynecol 1991;77:
replaced D&C as the first-line diagnostic test because 458-62.
it has similar accuracy, yet is safer, quicker, and more 16. Archer DF, McIntyre-Seltman K, Wilborn WW Jr, Dowling EA, Cone F, Creasy

convenient. Primary care physicians and gynecolo- EW, et al. Endometrial morphology in asymptomatic postmenopausal women. Am
J Obstet Gynecol 1991;165:317-22.
gists should be aware of the risk factors for endome- 17. Ferry J, Farnsworth A, Webster M, Wren B. The efficacy of the Pipelle endo-
trial cancer so that they know which women require metrial biopsy in detecting endometrial carcinoma. Aust N Z J Obstet Gynaecol
endometrial biopsy. All postmenopausal women with 1993;33(1):76-8.
uterine bleeding who are not taking HRT require 18. Kaunitz AM, Masciello A, Ostrowski M, Rovira EZ. Comparison of endometrial
biopsy with the endometrial Pipelle and Vabra aspirator. J Reprod Med 1988;33:
endometrial sampling. All women taking HRT with
427-31.
unexpected uterine bleeding need endometrial 19. Koss LG, Schreiber K, Oberlander SG, Missouris HF, Lesser M. Detection of
biopsy. All perimenopausal women with abnormal endometrial carcinoma and hyperplasia in asymptomatic women. Obstet Gynecol
bleeding should be sampled. Premenopausal women 1984;64:1-11.

with irregular cycles or with one or more other risk 20. Stovall TG, Ling FW, Morgan PL. A prospective, randomized comparison of the
Pipelle endometrial sampling device with the Novak curette. Am J Obstet Gynecol
factors for endometrial cancer should have endo- 1991;165:1287-9.
metrial biopsy. When endometrial biopsy cannot be 21. Rodriquez GC, Yaqub N, King ME. A comparison of the Pipelle device and the
performed or produces insufficient tissue, or patients Vabra aspirator as measured by endometrial denudation in hysterectomy speci-

have persistent bleeding after negative biopsy results, mens. Am J Obstet Gynecol 1993;168:55-9.
22. Emanual MH, Verdel MJ, Wamsteker K, Lammes FB. A prospective comparison
transvaginal ultrasound can help guide decisions
of transvaginal ultrasonography and diagnostic hysteroscopy in the evaluation of
about who to sample with D&C and who to monitor patients with abnormal uterine bleeding: clinical implications. Am J Obstet Gynecol
with close follow up. 1995;172(2 Pt 1):547-52.
23. Winkler B, Alvarez S, Richart RM, Crum CP. Pitfalls in the diagnosis of endome-
trial neoplasia. Obstet Gynecol 1984;64:185-94.
Acknowledgment
24. Feldman S, Chapter A, Welch WR, Berkowitz RS. Two-year follow up of 263
We thank Dr John Hickey and Dr Don Wescott of Antigonish, patients with post/perimenopausal vaginal bleeding and negative initial biopsy.
NS, for reviewing the draft of this article. Gynecol Oncol 1994;55:56-9.

VOL 48: MAY • MAI 2002  Canadian Family Physician • Le Médecin de famille canadien 887