Postmenopausal bleeding on tamoxifen with a thick endometrium and a non-diagnostic office biopsy

The thickness is not interpretable in the usual way. Tamoxifen has a partial oestrogenic effect on the endometrium, producing subepithelial stromal hypertrophy and cystic glandular spaces. On ultrasound this looks like a thickened, echogenic, heterogeneous or 'Swiss cheese' endometrium. Up to half of asymptomatic women on tamoxifen have an endometrial thickness over 5 mm, and the conventional 4–5 mm threshold used in other postmenopausal women has no useful diagnostic value here. A 14 mm measurement is therefore not by itself alarming, and equally it cannot reassure.

Bleeding is the finding that matters. Any postmenopausal bleeding in a woman on tamoxifen requires tissue diagnosis, because tamoxifen is associated with endometrial polyps (common), endometrial hyperplasia and carcinoma (relative risk of about 2–4, absolute risk low, usually low-grade and early-stage), and rarely uterine sarcoma. Bleeding cannot be attributed to the drug.

The Pipelle was inadequate and must not be accepted as negative. A blind sampler can miss a polyp, a focal lesion or a sarcoma, and the report says no tissue was obtained - that is a failed test, not a normal one.

Next step: hysteroscopy with directed biopsy, and polypectomy if a polyp is found. This allows direct inspection of the cavity, targeted sampling of any abnormal area and removal of the likely cause of bleeding in one procedure. If hysteroscopy is unavailable or she cannot tolerate it, saline infusion sonography to outline the cavity followed by a formal dilatation and curettage would be second best. Also examine the vulva, vagina and cervix - atrophic vaginitis, a cervical polyp or a vaginal lesion can all cause bleeding and are easily missed if attention is fixed on the endometrium.

Apixaban. Hysteroscopy with polypectomy is a procedure with moderate bleeding risk, and the anaesthetist and gynaecologist should agree a plan with her cardiologist or anticoagulation clinic rather than deciding informally.

• With an eGFR of 72, apixaban is cleared reasonably quickly. For a polyp resection or any procedure likely to require general anaesthesia, omit the last dose - typically 24 hours before, or 48 hours if the creatinine clearance were lower, if a larger resection were planned, or if neuraxial anaesthesia were used.
• No bridging with low molecular weight heparin is needed for a direct oral anticoagulant in a minor procedure; bridging increases bleeding without reducing stroke risk in this setting.
• For a purely diagnostic, small-calibre office hysteroscopy with no resection, many units continue the drug, but the decision should be documented.
• Restart apixaban 24 hours after the procedure provided haemostasis is secure, and check the haemoglobin and for ongoing bleeding before discharge.

Tamoxifen. She should continue it - stopping for a minor vaginal procedure is unnecessary and would interrupt a treatment with proven survival benefit. It does increase venous thromboembolism risk, particularly when combined with pelvic surgery and immobility, so if a larger operation such as hysterectomy is later required, discuss timing with her oncologist (some would interrupt for 1–2 weeks) and ensure mechanical and pharmacological thromboprophylaxis.

Explain to her that neither the tamoxifen nor the apixaban is the likely cause of the bleeding, that the apixaban makes the bleeding look heavier than it otherwise would, and that the plan is to look inside the uterus directly and take a proper sample. Obtain valid consent covering the possibility of finding and treating a polyp, the small chance of perforation, infection and bleeding, and the fact that a further procedure may be needed if the histology is malignant.

This is a favourable finding. Grade 1 endometrioid cancer arising in a polyp, without myometrial invasion, is typically stage IA and carries an excellent prognosis; tamoxifen-associated cancers are usually detected early because they present with bleeding.

Staging and surgery

• Review the histology and the hysteroscopy findings together; MRI of the pelvis helps assess myometrial and cervical involvement, and cross-sectional imaging of the chest and abdomen is reasonable if there is any suspicion of more advanced disease.
• Definitive treatment is total hysterectomy with bilateral salpingo-oophorectomy, with sentinel lymph node assessment or lymphadenectomy depending on local protocol and intraoperative findings. A minimally invasive approach is preferred if she is fit, but her BMI and previous pelvic surgery may make this technically difficult - the surgeon should be prepared to convert.
• She does not need adjuvant radiotherapy or chemotherapy for a node-negative stage IA grade 1 tumour.
• Note that her fibroid and uterine size do not change the plan; the subserosal fibroid is incidental.

Tamoxifen

• The standard advice is to continue tamoxifen. The survival benefit for her breast cancer outweighs the small excess risk of endometrial cancer, most of which is detected early and treated curatively. Stopping early would trade a certain oncological benefit for an uncertain gynaecological one.
• If she has completed five years soon, that is the natural stopping point; discuss with her oncologist. Switching to an aromatase inhibitor is an option only after hysterectomy and oophorectomy in a postmenopausal woman, and that decision belongs to oncology.
• She should be told that removing the uterus and ovaries removes the endometrial risk but not the risk of a vaginal or, rarely, peritoneal recurrence, so any further bleeding must be reported.

Follow-up - postoperative review with histology, explanation that no further treatment is likely, and clear instructions to report any vaginal bleeding, pelvic pain or discharge. Document the conversation with her oncologist so that she is not given conflicting advice about her tamoxifen.