Ovarian carcinosarcoma is a distinct form of ovarian cancer with poorer survival compared to tubo-ovarian high-grade serous carcinoma

Abstract

Background: Ovarian carcinosarcoma (OCS) is an uncommon, biphasic and highly aggressive ovarian cancer type, which has received relatively little research attention.

Methods: We curated the largest pathologically confirmed OCS cohort to date, performing detailed histopathological characterisation, analysis of features associated with survival and comparison against high-grade serous ovarian carcinoma (HGSOC).

Results: Eighty-two OCS patients were identified; overall survival was poor (median 12.7 months). In all, 79% demonstrated epithelial components of high-grade serous (HGS) type, while 21% were endometrioid. Heterologous elements were common (chondrosarcoma in 32%, rhabdomyosarcoma in 21%, liposarcoma in 2%); chondrosarcoma was more frequent in OCS with endometrioid carcinomatous components. Earlier stage, complete resection and platinum-containing adjuvant chemotherapy were associated with prolonged survival; however, risk of relapse and mortality was high across all patient groups. Histological subclassification did not identify subgroups with distinct survival. Compared to HGSOC, OCS patients were older (P < 0.0001), more likely to be FIGO stage I (P = 0.025), demonstrated lower chemotherapy response rate (P = 0.001) and had significantly poorer survival (P < 0.0001).

Conclusion: OCS represents a distinct, highly lethal form of ovarian cancer for which new treatment strategies are urgently needed. Histological subclassification does not identify patient subgroups with distinct survival. Aggressive adjuvant chemotherapy should be considered for all cases, including those with early-stage disease.

Fig. 1. Case flow diagram for ovarian carcinosarcoma (OCS) cohort.

IHC immunohistochemistry, H&E haematoxylin and eosin.

Fig. 2. Histopathological features of ovarian carcinosarcoma (OCS).

a, b WT1-positive epithelial component of high-grade serous type. c, d WT1-negative epithelial component of endometrioid type. e, f OCS demonstrating S100-positive malignant cartilage. g, h OCS with myogenin-positive rhabdomyoblasts. i OCS demonstrating liposarcoma (confirmed cytokeratin-negative) adjacent to endometrioid epithelial component. j OCS demonstrating squamous differentiation. Scale bar represents 100 µm.

Fig. 3. Clinical and histopathological landscape of ovarian carcinosarcoma.

STIC serous tubal intraepithelial carcinoma, FIGO International Federation of Gynecology and Obstetrics, NE non-evaluable, FT fallopian tube, RD residual disease, NA not available. For epithelial p53 staining, 8 NE for cytoplasmic p53 staining, and 6 NE for other reasons (significant artefacts due to sample age or fixation, null tumour staining without corresponding stromal positivity or other uninterpretable staining patterns).

Fig. 4. Ovarian carcinosarcoma patient survival.

a Overall survival (OS) by histological subgrouping. Labelled hazard ratio (HR) represents comparison of high-grade serous (HGS) epithelial with homologous sarcoma (HGS-homologous) versus endometrioid (endo) epithelial with heterologous sarcoma (endo-heterologous); P = 0.380. b OS by residual disease (RD) status following debulking surgery; P = 0.001. c OS by stage at diagnosis. Labelled HR represents comparison of stage I/II versus stage III; P = 0.015. d OS by age at diagnosis; P = 0.538. Patients grouped according to median age at diagnosis across the cohort (69 years). e OS by first-line treatment regime. Labelled HR represents comparison of platinum–taxane versus surgery only; P < 0.001. Additionally, HR for other platinum regimens (26 single-agent platinum, 2 other platinum combinations) versus surgery only is 0.34, 95% CI 0.19–0.61, P < 0.001. f Forest plot of multivariable overall survival analysis, stratified by diagnosis period. Chemo chemotherapy, macro macroscopic, NVRD no visible residual disease.

Fig. 5. Comparison of ovarian carcinosarcoma (OCS) and high-grade serous ovarian carcinoma (HGSOC).

a Age at diagnosis. b FIGO stage at diagnosis; labelled P value represents comparison of the frequency of stage I cases. c Response to first-line chemotherapy; labelled P value represents comparison of overall response rate (complete response [CR] plus partial response [PR] versus stable disease [SD] plus progression disease [PD]). d Overall survival; labelled HR represents multivariable analysis of tumour type (HGSOC versus OCS; P < 0.0001), age at diagnosis and stage at diagnosis, stratified by residual disease status.