One Episode of Bleeding Five Years After Menopause
A 57-year-old woman comes in for one episode of vaginal bleeding last week; it was light, red-brown, and lasted about a day. Her last menstrual period was at age 51, with no bleeding since, and she is not on hormone therapy. BMI 38, hypertension, type 2 diabetes for 8 years (A1c 7.8%). She has never been pregnant. Her mother had 'stomach cancer' at age 70. She had one Pap test 12 years ago, which was normal, and has not been screened since. Exam: BP 146/88, no abdominal masses, vulva and vagina normal, cervix appears normal with no visible lesion and no bleeding at the time of exam, uterus small and mobile, adnexa not palpable.
1. What is the priority diagnosis, and how does the history shift your probability estimate?
Read the worked answer
Priority diagnosis: endometrial cancer until excluded. Any bleeding after 12 months of amenorrhea in a postmenopausal woman meets the definition of postmenopausal bleeding, regardless of how light or brief it was. Endometrial cancer is the cause in roughly 1 in 10 of these patients overall, and it is the diagnosis whose delay causes the most harm.
Her history pushes the probability upward:
- BMI 38 - peripheral conversion of androgens to estrogen, unopposed by progesterone
- Type 2 diabetes and hypertension - both part of the classic risk cluster
- Nulliparity - unopposed estrogen exposure across the reproductive years
- Age 57 - risk rises with age
Conditions that are more common but must be diagnosed, not assumed: atrophic vaginitis, endometrial or cervical polyps, submucosal fibroids, and endometrial hyperplasia. The brief, light nature of the bleeding is not reassuring - early endometrial cancer often presents exactly this way.
Do not attribute this to atrophy on the basis of the exam. A normal-appearing cervix and small uterus do not exclude an endometrial source, and a visual exam cannot see the cavity.
Reasoning: the exam tests whether a single short bleeding episode is dismissed as trivial. It should not be.
2. What is the initial workup, and how do you interpret a transvaginal ultrasound showing a 3 mm endometrial stripe?
Read the worked answer
First-line: office endometrial biopsy (pipelle or equivalent). It is minimally invasive, has high sensitivity for endometrial cancer, and provides tissue for histology - which the ultrasound cannot. Order it now, along with a CBC and a review of her cancer screening status; she is overdue for cervical screening and is still within the age range for it.
If a transvaginal ultrasound is obtained and the stripe is 3 mm: a thin stripe has a high negative predictive value for endometrial cancer in a postmenopausal patient with bleeding, and a thickness of 4 mm or less is generally considered reassuring. But it does not exclude the diagnosis in this patient, and it does not cancel the need for tissue in her case:
- She has multiple strong risk factors, which raise the pre-test probability
- A thin stripe does not exclude polyps, submucosal pathology, or cervical sources, and ultrasound cannot distinguish a small focal lesion from normal endometrium
- If bleeding recurs, tissue sampling is mandatory regardless of stripe thickness - a reassuring ultrasound cannot be used repeatedly to defer biopsy in a patient who keeps bleeding
Reasoning: the exam likes to place a normal-looking test in front of you. The correct move is to weigh it against the pre-test probability, not to treat it as a rule-out.
3. The office biopsy is performed but the pathology report says 'insufficient for diagnosis.' What now, and what would you avoid?
Read the worked answer
Insufficient tissue is a non-result, not a negative result. Treat it as an unfinished workup and escalate.
Next step: hysteroscopy with directed biopsy, with dilation and curettage if needed. This allows direct visualization of the cavity, sampling of focal lesions, and definitive tissue. Refer to gynecology; do not simply repeat an office biopsy that already failed, and do not file the case as complete.
What to avoid:
- Reassuring her that the ultrasound was thin, so nothing further is needed
- Empirical hormone therapy or a 'trial of estrogen' to see whether bleeding stops
- Watchful waiting for recurrent bleeding before acting
- Ordering a CT or MRI as a substitute for tissue - imaging does not replace histology here
Meanwhile: address the modifiable risk factors you can act on at this visit - blood pressure 146/88, A1c 7.8%, and weight - and bring her cervical screening up to date. These matter both for her overall risk and because they are the parts of the plan she can control while awaiting the procedure.
Close the loop explicitly: tell her what an insufficient sample means, give her a date by which she should have heard the result of the next test, and document the referral.
Reasoning: the exam tests follow-through on an indeterminate result, which is where real-world diagnostic delays occur.