Endometrial cancer

Overview

• Epidemiology (US 2025): ~67,000 cases, ~13,000 deaths. Most common gynecologic cancer and 4th most common cancer overall in US women; incidence rising (obesity epidemic). Mostly postmenopausal. Incidence historically White > Black but now roughly equal (higher in Black women when corrected for hysterectomy prevalence); mortality nearly 2x higher in Black women.
• Diagnosis: Postmenopausal bleeding needs evaluation (do not screen asymptomatic women): TVUS (endometrial stripe ≤4 mm has >99% NPV, permits observation on an initial episode) and/or endometrial biopsy; persistent or recurrent bleeding needs histology regardless. Postmenopausal bleeding is the cardinal symptom; workup with transvaginal US (endometrial thickness >4 mm) and/or endometrial biopsy.
• MMR testing at diagnosis by IHC on all specimens (NCCN). If MMR-deficient, do reflex MLH1 promoter methylation for MLH1/PMS2 loss: methylated = likely sporadic; unmethylated MLH1 loss or MSH2/MSH6/PMS2 loss, or young age/suggestive history, refer to genetics for Lynch. Even low-risk patients after surgery should be referred to rule out Lynch.

Histologic types

• Type I (endometrioid) ~80 to 85%: estrogen-dependent (HR+), obesity, tamoxifen, associated with endometrial hyperplasia. Usually low grade (grade 1 to 2), good prognosis (low or intermediate risk). Endometrioid histology only.
  • Mutations: PTEN loss (~60%), PIK3CA (~40%), KRAS (~20%), FGFR2 (~15%), MMR (~45%), beta-catenin accumulation (≤50%).
• Type II (non-endometrioid) ~10 to 15%: clear cell, papillary serous, carcinosarcoma. Estrogen-independent, often older/thinner patients. All considered high grade, poor prognosis (high risk). Serous behaves like high-grade ovarian.
  • Mutations: p53 (~90%), PIK3CA (~50%), PTEN loss (~25%), HER2 overexpression/amplification (~25 to 30% of serous).
• Risk factors: unopposed estrogen (obesity, anovulation, tamoxifen, estrogen-only HRT), early menarche, late menopause, nulliparity, Lynch syndrome (MLH1/MSH2/MSH6/PMS2), Cowden syndrome (PTEN).

Molecular classification (four classes)

• POLE-ultramutated: excellent prognosis; consider de-escalation or observation even with higher-grade histology.
• MMRd / MSI-H (~25 to 30%): intermediate prognosis; immunotherapy-responsive. Overlaps with Lynch.
• NSMP (no specific molecular profile, copy-number low, endometrioid): intermediate prognosis; often ER/PR-driven.
• p53-abnormal (copy-number high, serous-like, TP53-mutated): worst prognosis; treat aggressively regardless of stage.
• Prognosis: POLE best, p53abn worst; MMRd and NSMP are intermediate and overlapping (PORTEC-3 5-yr RFS: POLE 98%, NSMP 74%, MMRd 72%, p53abn 48%).

Staging (FIGO)

• IA (FIGO 2009): confined to uterus, <50% myometrial invasion. FIGO 2023 restages by histotype: IC = aggressive histotype without myoinvasion; POLE-mutated and p53-abnormal findings modify stages I to II.
• IB: confined to uterus, ≥50% myometrial invasion.
• II: cervical stromal invasion.
• IIIA: invasion of serosa or adnexa.
• IIIB: vaginal or parametrial involvement.
• IIIC1: pelvic LN involvement. IIIC2: para-aortic LN involvement.
• IVA: invasion of bladder or bowel mucosa.
• IVB (FIGO 2009): distant mets or inguinal LN. FIGO 2023: IVB = abdominal (extrapelvic) peritoneal metastasis; IVC = distant metastasis including inguinal/distant nodes; pelvic peritoneal spread = IIIB2.
• FIGO 2023 update: now incorporates molecular subtype, histology, and lymphovascular space invasion (LVSI).

Workup

• Imaging: pelvic MRI, CT chest/abdomen/pelvis; PET-CT for advanced.
• Sentinel LN biopsy (SLNB) preferred over routine pelvic LND for staging.

Treatment: localized (early-stage)

• Surgery is done for final staging: hysterectomy, BSO, pelvic washings, examination of the entire abdominal cavity, with SLNB (preferred). Consider pelvic and para-aortic LND in deep invasion, LVSI+, or grade 3 tumors (papillary serous, clear cell).
• Non-endometrioid (serous, clear cell, carcinosarcoma) is high-risk histology, but adjuvant therapy depends on stage, myoinvasion, molecular class, and staging completeness: fully staged stage IA without myoinvasion may get vaginal brachytherapy or observation; chemo (with or without RT) for myoinvasive or stage IB and higher disease.
• Adjuvant therapy by risk group:
  • Low-risk (stage IA, endometrioid grade 1 to 2): surgery alone then observation. 5-year OS ~95% with surgery alone.
  • Intermediate-risk (5-year OS ~85% with surgery alone): stage I grade 3 endometrioid <50% invasion, observation with or without vaginal brachytherapy.
  • High-intermediate risk: stage IB grade 1 to 2 with >50% invasion, stage IB grade 3, stage II. Consider RT. If pelvic nodes positive (stage IIIC), treat as high risk (systemic therapy with or without RT). If nodes negative, choose observation, vaginal brachytherapy, or pelvic EBRT by uterine risk factors (grade, depth, LVSI, cervical stromal invasion) and molecular class, not nodal status alone; EBRT for stage IB grade 3, stage II, or substantial LVSI even if nodes are negative.
  • High-risk (LVSI, non-endometrioid histology, cervical stromal involvement, grade 3, age >60): stage I/II endometrioid, give RT (EBRT and/or vaginal brachytherapy) with or without chemo; myoinvasive non-endometrioid (stage IA with myometrial invasion and higher), any stage III (any grade), and any stage IV, give chemo with or without RT. Role of chemoRT is not fully clear; chemoRT with cisplatin then carbo/paclitaxel for 4 cycles is used in more advanced stage IIIB disease.
• Check HER2 in serous histology: add trastuzumab to carbo/paclitaxel for HER2-positive advanced (stage III/IV) or recurrent uterine serous carcinoma (Fader); not a universal addition for every HER2-positive localized tumor.
• Intravaginal RT (IVRT) indicated for endometrioid grade 3, or any grade of other histologies (carcinosarcoma, clear cell, dedifferentiated, serous).
• Stage III/IV resected: adjuvant chemo (carbo/paclitaxel for 6 cycles), with or without RT (EBRT and/or vaginal brachytherapy) individualized to local-regional recurrence risk.
• Fertility preservation (very selected stage IA grade 1): megestrol or levonorgestrel IUD with close surveillance, then definitive surgery after childbearing.

Advanced, inoperable, recurrent, or metastatic

• Chemotherapy backbone: carboplatin + paclitaxel for chemo-naive disease (GOG 209 established carbo/paclitaxel as non-inferior to doxorubicin/cisplatin/paclitaxel with less toxicity).
• Carboplatin + paclitaxel + dostarlimab or pembrolizumab is now the 1L standard for advanced/recurrent disease (except carcinosarcoma for the pembrolizumab NRG-GY018 population):
  • RUBY (dostarlimab, Mirza NEJM 2023; PMID: 36972026): dostarlimab + carbo/paclitaxel for 6 cycles then dostarlimab maintenance for 3 years. dMMR cohort mPFS HR 0.28 (~70% reduction); all-comers mOS 44.6 vs 28.2 mo. FDA Jul 31, 2023 for dMMR; expanded Aug 1, 2024 to all-comers regardless of MMR status.
  • NRG-GY018 / KEYNOTE-868 (pembrolizumab, Eskander NEJM 2023; PMID: 36972022): pembro + carbo/paclitaxel for 6 cycles then pembro maintenance for 14 cycles. dMMR mPFS HR 0.30; pMMR mPFS HR 0.54. FDA Jun 17, 2024 for all-comers (1L advanced/recurrent).
  • NRG-GY018 OS update (Eskander ASCO 2025): NRG-GY018 OS (immature, not significant): dMMR HR 0.55 (95% CI 0.25 to 1.19), pMMR HR 0.79 (95% CI 0.53 to 1.17), both CIs crossing 1; the 1L benefit is driven by PFS. Solidifies the 1L IO + chemo backbone for both MMR groups.
  • AtTEnd (atezolizumab, Colombo Lancet Oncol 2024): atezolizumab + carbo/paclitaxel then atezo maintenance. dMMR mPFS HR 0.36; all-comers PFS HR 0.74. Confirms class effect; not FDA-approved.
  • DUO-E (Westin JCO 2024; NCT04269200): durvalumab + carbo/paclitaxel then durvalumab with or without olaparib maintenance vs chemo then placebo. dMMR PFS HR 0.42 (durva arm); pMMR HR 0.77 (durva) and HR 0.57 (durva + olaparib maintenance). FDA Jun 14, 2024 for durvalumab + carbo/pac then durva maintenance in dMMR. Durva + olaparib not separately labeled but signals PARP + IO utility, especially in pMMR.
• Carbo/paclitaxel + trastuzumab for stage III/IV or recurrent HER2+ uterine serous carcinoma (Fader trial); NCCN extends to HER2+ carcinosarcoma (not studied in Fader).
• Hormonal therapy for advanced/inoperable low-grade ER+/PR+ disease: megestrol acetate, tamoxifen, aromatase inhibitors, fulvestrant. Everolimus + letrozole ORR ~32%.

• Single-agent IO for dMMR / MSI-H: pembrolizumab (KEYNOTE-158) or dostarlimab (GARNET).
• Lenvatinib + pembrolizumab (KEYNOTE-775, Makker NEJM 2022; PMID: 35045221) for pMMR / MSS recurrent disease: pMMR mOS 17.4 vs 12.0 mo, mPFS 6.6 vs 3.8 mo vs chemo (all-comers 18.3 vs 11.4 and 7.2 vs 3.8 mo) (doxorubicin or paclitaxel). FDA full approval Jul 21, 2021 for advanced EC after prior systemic therapy not curable by surgery/RT. Established option for previously treated pMMR disease (KEYNOTE-775 enrolled checkpoint-inhibitor-naive, post-platinum patients); benefit after first-line PD-1 therapy is extrapolated, not level-1, so not an automatic 2L standard. Toxicity: hypertension, fatigue, GI, hand-foot syndrome.
• HER2-amplified serous: trastuzumab + carbo/paclitaxel; T-DXd (trastuzumab deruxtecan; FDA tumor-agnostic accelerated approval Apr 5, 2024 for HER2 IHC 3+ solid tumors after prior systemic therapy, based on DESTINY-PanTumor02).
• SIENDO / ENGOT-EN5 (Vergote Lancet Oncol 2024): selinexor maintenance after platinum response. Negative overall PFS but TP53-wildtype subgroup mPFS 27.4 vs 5.2 mo (HR 0.41); XPORT-EC-042 phase 3 confirmatory ongoing. Molecularly guided maintenance signal.
• Other active agents: liposomal doxorubicin, topotecan, temsirolimus. Single-agent trastuzumab did not show significant response rates.

Key adjuvant/RT trials

• GOG 249: high-intermediate-risk, stage II (any histology), and stage I to II serous or clear cell. Whole-pelvis RT vs vaginal brachytherapy + chemo: same recurrence-free survival.
• GOG 258: stage III and IVA disease. Chemo vs chemoRT (cisplatin) then carbo/paclitaxel for 4 cycles: same recurrence-free survival, lower local relapse with chemoRT; grade ≥3 AEs similar (58% chemoRT vs 63% chemo), with more GI toxicity on chemoRT and more hematologic toxicity on chemo alone.
• PORTEC-3: stage I endometrioid grade 3 with deep (≥50%) myometrial invasion and/or LVSI, endometrioid stage II or III, or stage I to III serous/clear cell. Pelvic RT (plus vaginal brachytherapy if cervix involved) vs chemoRT (cisplatin) then carbo/paclitaxel for 4 cycles: chemoRT improved 5-year OS (81.4% vs 76.1%, HR 0.70) and failure-free survival (76.5% vs 69.1%, HR 0.70), with greatest benefit in stage III and serous cancers (de Boer Lancet Oncol 2019).

Lynch syndrome: endometrial considerations

• Endometrial cancer is often the sentinel cancer in Lynch syndrome (precedes colorectal cancer).
• Universal MMR/MSI screening on all endometrial cancer specimens.
• Discuss risk-reducing hysterectomy after childbearing, timing individualized by gene (highest risk MLH1/MSH2). Decide oophorectomy separately: insufficient evidence in MSH6, and PMS2 carriers may reasonably decline (ovarian risk not clearly increased).

High-yield endometrial pearls

• Postmenopausal bleeding: biopsy.
• Universal MMR/MSI testing at diagnosis; refer to genetics if deficient.
• Type I (endometrioid) = obesity, estrogen, indolent. Type II (serous, p53) = aggressive.
• Molecular four classes: POLE (best) > MMRd/MSI and NSMP (intermediate, overlapping) > p53abn (worst).
• RUBY / GY-018: IO + chemo 1L (dMMR strongest; pMMR modest benefit).
• Lenvatinib + pembrolizumab (KEYNOTE-775) for pMMR recurrent.
• Check HER2: add trastuzumab (serous); T-DXd for HER2 IHC 3+ after prior systemic therapy.
• Tamoxifen raises endometrial cancer risk; Lynch often presents as endometrial cancer first.