Postmenopausal Bleeding With a Non-Diagnostic Office Biopsy
A 64-year-old nulliparous woman (BMI 36, type 2 diabetes, hypertension) reports two episodes of light vaginal spotting over the past 6 weeks, 12 years after menopause. She has no pain, and examination is unremarkable. Community transvaginal ultrasound reported endometrial thickness 5 mm with a homogeneous echo pattern; serum CA-125 is 18 U/mL. An office Pipelle sample returned scant tissue, insufficient for diagnosis. She asks whether she can simply be rescanned in 3 months, since the ultrasound was normal and the bleeding has stopped. She takes no hormones and has no personal history of breast cancer. Her mother was diagnosed with colon cancer at 72. She is a poor historian about her menses and cannot recall whether the spotting was heavy enough to stain undergarments.
1. What is the most appropriate next diagnostic step, and why does the reassuring transvaginal ultrasound not justify observation?
Read the worked answer
Observation is not acceptable. Ultrasound is a triage test, not a diagnostic one. In a woman with postmenopausal bleeding, a thin endometrium lowers but does not eliminate cancer risk, and 5 mm sits at or above the threshold at which a benign explanation can be assumed. The decisive issue is the non-diagnostic sample. Blind office sampling with a Pipelle samples a small fraction of the cavity and misses a meaningful proportion of focal lesions, particularly polyps, cornual disease, and small cancers. In a symptomatic woman, a scant-tissue report is an indication for obtaining tissue, not for reassurance.
The next step is hysteroscopy with directed biopsy and curettage of the cavity, under anesthesia or sedation if needed, so that focal lesions are visualized and sampled. Repeating imaging in 3 months risks interval progression and delays diagnosis.
Her metabolic profile (BMI 36, diabetes, hypertension) raises the pre-test probability of endometrial cancer and of endometrioid intraepithelial neoplasia, and should lower the threshold for definitive sampling. Note also that bleeding described as minimal still qualifies as postmenopausal bleeding; there is no 'too little to matter' threshold.
2. Hysteroscopy shows a thickened polypoid endometrium; biopsy shows endometrioid intraepithelial neoplasia (EIN) without invasion. She does not want fertility preservation. Outline management, including the role of nodal staging and the risk of finding cancer.
Read the worked answer
EIN carries a concurrent endometrial carcinoma risk of roughly 25-45% at hysterectomy, so completion surgery is the first step.
- Procedure: total hysterectomy with bilateral salpingo-oophorectomy. Removing the ovaries eliminates the estrogen drive and addresses the small risk of synchronous adnexal disease.
- Nodes: lymphadenectomy is not indicated for a preoperative diagnosis of EIN alone. If the final specimen shows stage IA grade 1-2 disease, nodal risk is very low and dissection adds lymphedema without survival benefit. Because a substantial minority are upgraded, the uterus must be carefully sectioned; if deep invasion, grade 3, or lymphovascular invasion is found, a sentinel lymph node approach can be considered. Some surgeons place a sentinel node protocol on standby so staging can be completed in the same anesthetic if frozen section confirms invasion.
- Counselling: discuss the upgrade rate explicitly, along with weight loss, glycemic control, and thromboprophylaxis.
- If she were a poor surgical candidate: a levonorgestrel IUD with repeat sampling achieves regression in the majority, but requires surveillance and carries a persistent progression risk.
3. She asks what could have prevented this and what you advise for her 41-year-old daughter, who has BMI 33 and irregular periods.
Read the worked answer
Separate non-modifiable from modifiable risk and address the daughter as a distinct patient.
This patient: age, nulliparity, and unopposed estrogen exposure are fixed. Obesity is the dominant modifiable driver; each 5 kg/m2 increase in BMI corresponds to roughly a 60% relative increase in endometrial cancer risk. Diabetes and hypertension contribute via hyperinsulinemia and insulin resistance, and adipose aromatase converts androgens to estrone.
Family history: a mother with colon cancer at 72 does not by itself satisfy Amsterdam criteria, but if this patient's tumor is mismatch-repair deficient, germline testing becomes indicated and would change the daughter's surveillance. Ask specifically about Lynch-spectrum cancers and ages.
The daughter: the priorities are weight reduction, treatment of insulin resistance, and endometrial protection if she has chronic anovulation. Irregular cycles from polycystic ovary syndrome mean unopposed estrogen; cyclic progestin or a levonorgestrel IUD protects the endometrium and is more appropriate than unopposed estrogen. There is no proven screening program for average-risk women; no routine ultrasound or blind biopsy. The key safety message is that any abnormal bleeding warrants prompt tissue sampling rather than a scan.