Clinical overview
Gynaecological sarcomas are malignant tumours of mesenchymal (rather than epithelial) origin. They are uncommon — uterine sarcomas account for only about 3–8% of all uterine malignancies — but they are aggressive, prognostically poor, and clinically treacherous because their commonest presentation, a rapidly enlarging uterus with abnormal bleeding in a peri- or postmenopausal woman, is indistinguishable on history alone from the overwhelmingly more common benign leiomyoma. The clinical importance of this objective for the FCOG(SA) candidate is therefore disproportionate to the rarity of the disease: the registrar who removes "a fibroid" by morcellation and only later receives a histology report of leiomyosarcoma has converted a potentially curable Stage I tumour into disseminated peritoneal disease. The diagnosis is almost always made by the pathologist, not the clinician, so understanding the pathological features — gross appearance, histogenesis, microscopic criteria, immunohistochemistry and increasingly molecular signature — is the core of safe practice.
In the South African context several factors sharpen the problem. Access to expert gynae-oncological histopathology is concentrated in NHLS academic laboratories, so a peripheral specimen may take time to reach a sub-specialist soft-tissue pathologist. Late presentation with bulky disease is common. And the high background prevalence of HIV, while far more relevant to HPV-driven cervical disease, contributes to a population in whom non-HPV malignancies are also seen at younger ages. The pragmatic message is that any "fibroid" behaving atypically — rapid growth, growth after menopause, bleeding out of proportion, degenerate or haemorrhagic appearance on imaging — must be treated as a potential sarcoma until the histology says otherwise. This chapter focuses on the pathology; staging and treatment principles are touched on so the picture is complete, but the diagnostic morphology is the examinable heart of the topic.
Core knowledge
Classification and histogenesis
Figure D11.1 — The uterine sarcomas compared: leiomyosarcoma, low- and high-grade endometrial stromal sarcoma, adenosarcoma and carcinosarcoma — cell of origin and behaviour.
Gynaecological sarcomas are classified by the tissue they recapitulate and by their site, following the WHO Classification of Tumours of the Female Genital Tract, 5th edition (2020). The uterus is by far the commonest site. The clinically and pathologically important entities the registrar must be able to discuss are:
- Leiomyosarcoma (LMS) — malignant smooth-muscle tumour; the commonest uterine sarcoma (roughly 60–70% of cases).
- Low-grade and high-grade endometrial stromal sarcoma (LGESS / HGESS) — arising from endometrial stroma.
- Undifferentiated uterine sarcoma (UUS) — a high-grade tumour lacking specific differentiation, a diagnosis of exclusion.
- Adenosarcoma — a mixed tumour with benign epithelium and malignant (sarcomatous) stroma.
- Carcinosarcoma (malignant mixed Müllerian tumour, MMMT) — now reclassified by WHO 2020 as a metaplastic/dedifferentiated carcinoma, NOT a true sarcoma. It is included here because exams still test the distinction; biologically it behaves and is treated as a high-grade, often serous-like, epithelial carcinoma.
Beyond the uterus, vulvar and vaginal sarcomas are rare; embryonal rhabdomyosarcoma (sarcoma botryoides) is the classic vaginal sarcoma of infancy and young childhood, presenting as a grape-like haemorrhagic polypoid mass. Ovarian sarcomas are very rare and usually arise within a carcinosarcoma or teratoma.
Leiomyosarcoma — gross and microscopic features
Figure D11.2 — Leiomyosarcoma vs benign leiomyoma: the Stanford criteria (atypia, ≥10 mitoses/10 HPF, coagulative tumour-cell necrosis), STUMP, and the morcellation hazard.
Gross. A leiomyosarcoma is typically a single, large (often >10 cm), softer, fleshy mass with a variegated cut surface showing necrosis and haemorrhage, frequently with an ill-defined, infiltrative margin rather than the crisp shelling-out plane of a benign fibroid. The classic benign leiomyoma, by contrast, is firm, whorled, white and well-circumscribed. These gross clues are exactly why specimens that look unusual at theatre should be sent for urgent histology and why power morcellation of an undiagnosed mass is hazardous.
Microscopic. The diagnosis of uterine LMS rests on the Stanford triad of three features, of which at least two are required:
- Significant cytological atypia (diffuse, moderate-to-severe nuclear pleomorphism).
- High mitotic count — conventionally ≥10 mitoses per 10 high-power fields.
- Tumour-cell (coagulative) necrosis — geographic necrosis with an abrupt transition from viable tumour to dead cells, distinct from the gradual hyalinised infarct-type necrosis seen in degenerating fibroids.
The intersection of these features is the crux of the differential with benign and borderline smooth-muscle tumours. A tumour with mitotic activity but no atypia or tumour-cell necrosis is a mitotically active leiomyoma. A tumour that fulfils some but not all criteria, or whose features are difficult to interpret (e.g. epithelioid or myxoid variants), is labelled smooth muscle tumour of uncertain malignant potential (STUMP) — a genuinely uncertain category that requires careful follow-up, not reassurance. The myxoid and epithelioid variants are important pitfalls because they can be malignant at much lower mitotic counts; the usual numerical thresholds do not apply, and infiltrative margins or any tumour-cell necrosis should raise alarm.
Immunohistochemistry. LMS expresses smooth-muscle markers — desmin, h-caldesmon, smooth-muscle actin (SMA) — and often shows aberrant p53 and high Ki-67 proliferation index. ER/PR may be positive but are usually weaker than in benign smooth muscle. There is no single diagnostic immunostain; the diagnosis remains morphological, with IHC supporting smooth-muscle lineage and excluding mimics.
Endometrial stromal tumours
These arise from the stromal cells of the proliferative-phase endometrium and span a biological spectrum.
- Endometrial stromal nodule — benign, well-circumscribed, no myometrial or vascular invasion.
- Low-grade endometrial stromal sarcoma (LGESS) — cells resembling proliferative endometrial stroma, with the characteristic "tongue-like" permeative infiltration of the myometrium and prominent lymphovascular invasion (worm-like plugs in vessels). Mitoses are usually low. It is typically ER/PR positive and CD10 positive, and the great majority carry the recurrent JAZF1–SUZ12 (JAZF1–JJAZ1) gene fusion. Prognosis is relatively good but late recurrence is characteristic, so long follow-up is needed; hormone-driven growth means oestrogen should be avoided.
- High-grade endometrial stromal sarcoma (HGESS) — more atypical, higher mitotic rate, often carrying the YWHAE–NUTM2 (YWHAE–FAM22) fusion or, in a more recently described group, BCOR alterations. It is more aggressive than LGESS but still a distinct entity from UUS.
- Undifferentiated uterine sarcoma (UUS) — high-grade, pleomorphic, no specific line of differentiation and no defining molecular fusion; a diagnosis of exclusion with a poor prognosis.
The molecular fusions are increasingly central to classification under WHO 2020 because morphology and IHC overlap; identifying a fusion can confirm an otherwise ambiguous diagnosis and carries prognostic weight.
Adenosarcoma
Adenosarcoma is a biphasic (mixed) tumour with a benign epithelial component and a malignant stromal (sarcomatous) component. Grossly it is often a polypoid mass filling the endometrial cavity, sometimes prolapsing through the cervix. The hallmark microscopic pattern is leaf-like (phyllodes-like) clefts lined by benign epithelium, with stromal condensation ("cuffing") around the glands and increased mitotic activity in that periglandular stroma. Prognosis is generally better than for the homologous true sarcomas unless there is sarcomatous overgrowth (a sarcomatous component occupying ≥25% of the tumour, typically high-grade), which markedly worsens outcome and behaves like a pure high-grade sarcoma.
Carcinosarcoma — why it is no longer counted as a sarcoma
Carcinosarcoma (MMMT) contains both malignant epithelial (carcinomatous) and malignant mesenchymal (sarcomatous) elements. The sarcomatous component may be homologous (tissues native to the uterus, e.g. stromal/leiomyosarcomatous) or heterologous (tissues foreign to the uterus — rhabdomyosarcoma, chondrosarcoma, osteosarcoma). Monoclonality studies and WHO 2020 have established that the tumour is fundamentally a carcinoma that has undergone sarcomatous (epithelial–mesenchymal) metaplasia — the two components share the same clonal origin. It is therefore staged and treated as a high-grade endometrial carcinoma (commonly a p53-abnormal, serous-like aggressive phenotype), which is why the molecular framework of endometrial carcinoma is relevant to it. The examinable point is to be able to state clearly that carcinosarcoma is epithelial in origin, not a true sarcoma, while recognising its sarcomatous morphology.
Pathological mimics and the morcellation hazard
The pathology of these tumours matters most when it is used to separate benign from malignant smooth-muscle disease. Benign mimics that produce alarming gross or clinical pictures include cellular leiomyoma, mitotically active leiomyoma, atypical (symplastic) leiomyoma, bizarre leiomyoma, and the deceptive benign metastasising leiomyoma and intravenous leiomyomatosis — benign smooth-muscle proliferations with disconcerting "malignant-sounding" behaviour but bland cytology. The single most important practical pathology lesson is that uterine sarcoma cannot be reliably excluded pre-operatively, so undiagnosed presumed-fibroid masses should not be electromechanically morcellated, because dispersing fragments of an occult sarcoma upstages it and worsens survival.
Assessment
Figure D11.3 — Uterine sarcoma clues, spread and prognosis: rapid growth/postmenopausal red flags, haematogenous spread, and the prognosis ladder.
There is no reliable pre-operative test that confirms a uterine sarcoma; the diagnosis is essentially always histological and frequently retrospective on a hysterectomy/myomectomy specimen. Nonetheless the registrar's job is to raise suspicion and avoid harm.
- History and examination. The red-flag history is a rapidly enlarging uterine mass, particularly new growth after the menopause, abnormal uterine bleeding, and pelvic pain or pressure. (Rapid growth in a pre-menopausal woman has poor specificity — most are benign — but rapid growth after menopause, when oestrogen withdrawal should cause fibroids to shrink, is far more worrying.) Examine for a fixed, irregular, bulky uterus and any prolapsing cervical/vaginal mass (adenosarcoma; sarcoma botryoides in a child).
- Imaging. Pelvic ultrasound is first-line and may show a large, heterogeneous, vascular mass with areas suggesting necrosis. MRI gives the best soft-tissue characterisation and assesses local extent and myometrial relationship; features such as ill-defined margins, necrosis and restricted diffusion are suggestive but not diagnostic. CT chest/abdomen/pelvis is used to assess metastatic spread (lung is a favoured site for LMS). Imaging never replaces histology.
- Endometrial sampling. Pipelle/curettage or hysteroscopic biopsy may detect endometrial-cavity tumours (ESS, adenosarcoma) but is insensitive for intramural LMS, which lies within the myometrium and is often missed. A negative sample does not exclude sarcoma.
- The pathology request. When suspicion exists, communicate it on the request form so the NHLS laboratory prioritises and the pathologist applies the Stanford criteria, ancillary IHC (desmin, h-caldesmon, CD10, ER/PR, p53, Ki-67) and, where available, molecular/fusion testing for stromal tumours. Tumour grade and the FIGO uterine-sarcoma staging assignment come from the resection specimen.
Management
Management is included here for completeness; the pathological diagnosis drives every step.
- Surgery is the primary treatment. Total hysterectomy with bilateral salpingo-oophorectomy and removal of the tumour intact (no morcellation) is the standard for LMS and ESS confined to the uterus. The principle that flows directly from the pathology is en-bloc, intact removal: never morcellate an undiagnosed mass. Routine systematic lymphadenectomy is not standard for LMS (nodal spread is uncommon and haematogenous spread dominates) but is considered in ESS and where nodes are abnormal. In a young woman with apparent early LGESS, fertility/ovarian conservation is contentious because the tumour is hormone-driven.
- Hormonal therapy is central in LGESS, which is ER/PR positive: oophorectomy and anti-oestrogen / progestin or aromatase-inhibitor therapy for advanced or recurrent disease, and oestrogen (including HRT and tamoxifen) is avoided.
- Adjuvant radiotherapy and chemotherapy have a more limited and selective role in uterine sarcomas than in epithelial cancers; chemotherapy (e.g. doxorubicin-based, or gemcitabine–docetaxel regimens) is used for high-grade, advanced or recurrent LMS and UUS. Decisions should be made in a gynae-oncology multidisciplinary meeting.
- Carcinosarcoma, being epithelial, is managed along high-grade endometrial carcinoma lines — surgical staging plus platinum/taxane chemotherapy — which links it to the FIGO 2023 endometrial staging (see endometrial-carcinoma) that now advises molecular classification (POLEmut, MMRd, NSMP, p53abn) for endometrial carcinomas; carcinosarcomas are typically p53abn with poor prognosis. The ESGO–ESTRO–ESP endometrial carcinoma guideline (2025 update), which integrates molecular class into risk groups, governs this pathway. Where mismatch-repair deficiency (dMMR/MSI-H) is found, checkpoint-inhibitor strategies established in endometrial carcinoma trials (dostarlimab + carboplatin–paclitaxel, ENGOT-EN6/RUBY; pembrolizumab + chemotherapy, NRG-GY018; and pembrolizumab + lenvatinib, KEYNOTE-775 for the proficient-MMR setting) become relevant, underscoring why accurate pathological and molecular characterisation matters even for the sarcoma-mimicking tumours.
In the SA public sector, drug access is constrained by the Essential Medicines List; doxorubicin and platinum/taxane agents are available, whereas novel immunotherapy access is limited. All cases should be referred to a tertiary gynae-oncology unit, and pathology reviewed by a sub-specialist soft-tissue pathologist before definitive non-surgical treatment.
Red flags / pitfalls
- Morcellating an undiagnosed "fibroid." The single most consequential error. Power morcellation of an occult LMS disseminates tumour, upstages disease and worsens survival. If a mass is atypical, remove it intact (consider in-bag/contained extraction if morcellation is unavoidable for a presumed-benign specimen).
- Reassuring on the basis of a normal endometrial biopsy. A negative Pipelle/curettage does NOT exclude an intramural leiomyosarcoma.
- Misreading "rapid growth." In pre-menopausal women rapid fibroid growth is usually benign; the truly worrying scenario is a uterine mass growing after the menopause.
- Confusing degenerative necrosis with tumour-cell necrosis. Hyalinised infarct-type necrosis in a degenerating fibroid is benign; abrupt coagulative tumour-cell necrosis is one of the Stanford malignancy criteria. This is a pathology distinction the clinician should understand to interpret a report.
- Applying ordinary mitotic thresholds to myxoid and epithelioid smooth-muscle tumours. These variants can be malignant at low mitotic counts; standard numerical cut-offs do not apply.
- Calling carcinosarcoma a sarcoma. It is a metaplastic high-grade carcinoma (WHO 2020) and is staged/treated as endometrial carcinoma — a frequent exam trap.
- Dismissing a vaginal mass in a child. A grape-like haemorrhagic polypoid vaginal mass in an infant/toddler is sarcoma botryoides (embryonal rhabdomyosarcoma) until proven otherwise and needs urgent specialist referral.
- Under-resourced histology turnaround. In SA, ensure the NHLS request flags clinical suspicion so the specimen reaches a sub-specialist and molecular/fusion testing where indicated; do not start non-surgical treatment on imaging alone.
Evidence anchors
- WHO Classification of Tumours of the Female Genital Tract, 5th edition (2020) — the authoritative histological typing used throughout this chapter (smooth-muscle tumours including STUMP, endometrial stromal tumours with their fusion-defined groups, adenosarcoma, and the reclassification of carcinosarcoma as a metaplastic carcinoma rather than a true sarcoma).
- FIGO 2023 endometrial staging (Berek et al., IJGO 2023) and the ESGO–ESTRO–ESP endometrial carcinoma guideline, 2025 update (Lancet Oncol) — relevant to carcinosarcoma, which is staged and risk-stratified as a high-grade (typically p53abn) endometrial carcinoma with molecular classification (POLEmut, MMRd, NSMP, p53abn).
- Immunotherapy evidence relevant to dMMR/MSI-H endometrial-type tumours: dostarlimab + carboplatin–paclitaxel (ENGOT-EN6/RUBY), pembrolizumab + chemotherapy (NRG-GY018), and pembrolizumab + lenvatinib (KEYNOTE-775) — included because carcinosarcoma falls under the endometrial-carcinoma treatment framework.
- South African context: NHLS academic histopathology and molecular/fusion testing access; the SA EML (Hospital Level, Adults) governs chemotherapy availability (doxorubicin, platinum/taxane) and constrains novel immunotherapy access in the public sector. Referral to a tertiary gynae-oncology MDT is standard.
Note on numerical thresholds: the "≥10 mitoses per 10 HPF" and "Stanford triad (at least two of atypia, mitoses, tumour-cell necrosis)" diagnostic criteria for uterine leiomyosarcoma, and the JAZF1–SUZ12 / YWHAE–NUTM2 / BCOR fusion associations for endometrial stromal sarcomas, derive from the WHO 2020 classification and standard surgical-pathology criteria rather than from a guideline listed individually in VERIFIED-SOURCES.md; they are presented as established pathology and should be confirmed against the current WHO 5th edition before being asserted as exact cut-offs.