CT imaging characteristics of ovarian small cell carcinoma: correlation with clinicopathology and prognosis.

This retrospective study included patients who underwent surgery at the First Affiliated Hospital of the University of Science and Technology of China(USTC) between April 2018 and January 2025, with a pathological diagnosis of small cell carcinoma of the ovary (SCCO). Subtypes were classified according to the 2020 World Health Organization (WHO) criteria into pulmonary (SCCOPT) and hypercalcemic (SCCOHT) types. Inclusion criteria: (1) definitive pathological diagnosis of SCCO, with SCCOPT defined by positivity for neuroendocrine markers (Synaptophysin[Syn], Chromogranin A [CgA], CD56, INSM1) and SCCOHT defined by loss of SMARCA4 (BRG1) expression; (2) pre-treatment contrast-enhanced abdominopelvic CT with adequate imaging quality; and (3) complete clinical, pathological, and follow-up data. Exclusion criteria: (1) receipt of radiotherapy, chemotherapy, or other antitumor therapy prior to surgery; (2) CT images of inadequate quality; or (3) concomitant malignancies interfering with ovarian imaging. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the institutional ethics committee. The requirement for informed consent was waived owing to its retrospective design. A comprehensive literature search was conducted in PubMed and the China National Knowledge Infrastructure (CNKI) for studies published up to September 1, 2025. The search strategy combined disease-specific keywords (“ovarian small cell carcinoma,” “SCCOHT,” “SCCOPT”) with imaging-related terms (“CT,” “MRI,” “ultrasound,” “radiologic findings,” “imaging analysis”). After removing duplicates, two reviewers independently performed a two-stage screening process, first evaluating titles and abstracts and then assessing the full texts of potentially eligible records. Any disagreements were resolved through consensus discussion, and the reasons for study exclusion were documented. After screening, a total of 24 studies comprising 34 cases were included [ 3 , 10 – 32 ], together with 8 cases from our institution, yielding 42 patients with SCCO. A flowchart detailing the study selection process is summarized in Fig. 1 . Fig. 1 Flow diagram of the study selection process Flow diagram of the study selection process Clinical characteristics included age at diagnosis, presenting symptoms, preoperative International Federation of Gynecology and Obstetrics (FIGO, 2014) stage, and serum tumor markers such as CA125 and neuron-specific enolase (NSE). Treatment information encompassed the type of surgery, chemotherapy regimen, and the number of administered cycles. Pathological data comprised gross findings, microscopic morphology, histological subtype, differentiation status, lymphovascular invasion, perineural invasion, and immunohistochemical results (SMARCA4/BRG1, Syn, CgA, CD56, INSM1, Ki-67, and others). Follow-up information was collected through outpatient visits, inpatient records, and telephone interviews. Overall survival (OS) was defined as the time from diagnosis to death or last follow-up, and disease-free survival (DFS) as the interval from completion of initial treatment with complete remission to the first documented recurrence or metastasis. For the included studies, relevant data were extracted from the text. When necessary, we contacted the corresponding authors to obtain missing information. For patients included from our institution, contrast-enhanced abdominal CT examinations were performed using multidetector CT scanners (Discovery CT750 and Optima CT680, GE Healthcare). Acquisition parameters were standardized as follows: tube voltage, 100–120 kVp; tube current, 200–350 mAs; slice thickness, 5 mm; and reconstruction thickness, 1.25 mm. A non-ionic iodinated contrast agent (iohexol, or iodixanol) was administered intravenously at a dose of 1.5 mL/kg and an injection rate of 3.0 mL/s. Detailed acquisition parameters were not uniformly available for the cases retrieved from the literature review. Two senior radiologists, with 10 and 15 years of experience in gynecologic oncology imaging, independently reviewed all CT scans. Both readers were blinded to pathological findings and clinical outcomes. In cases of disagreement, consensus was reached with the involvement of a third expert radiologist. To reduce heterogeneity arising from differences in CT acquisition parameters and contrast phases across institutional and literature-derived cases, tumor enhancement was assessed using a relative visual grading approach rather than absolute quantitative metrics (e.g., Hounsfield units), which were not consistently available or comparable across scanners and published figures. Specifically, the enhancement of the solid tumor component was graded by comparison with an internal reference tissue on the same enhanced phase as: mild enhancement (lower than reference), moderate enhancement (approximately similar to reference), and marked enhancement (higher than reference). The normal myometrium was used as the preferred reference tissue; when the uterus was absent or myometrial attenuation was potentially unreliable (e.g., affected by leiomyoma or adenomyosis), the iliopsoas muscle at the same level was used as an alternative reference to ensure consistency. Image features were evaluated according to the following criteria: Tumor characteristics: laterality (left, right, bilateral), maximum diameter (cm), shape (round, oval, lobulated, irregular), and margin (well-defined, ill-defined). Internal architecture: Composition: predominantly solid, mixed solid–cystic, or predominantly cystic. Solid component: relative enhancement degree (mild/moderate/marked) and enhancement pattern (homogeneous, heterogeneous, peripheral, septal, progressive, or rapid wash-in/wash-out when assessable). Cystic degeneration/necrosis: presence and extent ( 50%). Calcification: presence and morphology (punctate, patchy, curvilinear). Peritumoral and distant findings: Ascites: presence and volume (mild, moderate, massive). Lymph node metastasis: size, morphology, and enhancement features of pelvic and para-aortic nodes. Peritoneal dissemination: peritoneal thickening, nodularity, or omental caking. Distant metastasis: suspected metastatic lesions in the liver, lung, bone, or other organs. Tumor characteristics: laterality (left, right, bilateral), maximum diameter (cm), shape (round, oval, lobulated, irregular), and margin (well-defined, ill-defined). Internal architecture: Composition: predominantly solid, mixed solid–cystic, or predominantly cystic. Solid component: relative enhancement degree (mild/moderate/marked) and enhancement pattern (homogeneous, heterogeneous, peripheral, septal, progressive, or rapid wash-in/wash-out when assessable). Cystic degeneration/necrosis: presence and extent ( 50%). Calcification: presence and morphology (punctate, patchy, curvilinear). Composition: predominantly solid, mixed solid–cystic, or predominantly cystic. Solid component: relative enhancement degree (mild/moderate/marked) and enhancement pattern (homogeneous, heterogeneous, peripheral, septal, progressive, or rapid wash-in/wash-out when assessable). Cystic degeneration/necrosis: presence and extent ( 50%). Calcification: presence and morphology (punctate, patchy, curvilinear). Peritumoral and distant findings: Ascites: presence and volume (mild, moderate, massive). Lymph node metastasis: size, morphology, and enhancement features of pelvic and para-aortic nodes. Peritoneal dissemination: peritoneal thickening, nodularity, or omental caking. Distant metastasis: suspected metastatic lesions in the liver, lung, bone, or other organs. Ascites: presence and volume (mild, moderate, massive). Lymph node metastasis: size, morphology, and enhancement features of pelvic and para-aortic nodes. Peritoneal dissemination: peritoneal thickening, nodularity, or omental caking. Distant metastasis: suspected metastatic lesions in the liver, lung, bone, or other organs. Statistical analyses were conducted using R (version 4.3.0). Associations were assessed with Spearman’s correlation [ 33 ]. To ensure the reliability of imaging assessment, interobserver agreement was quantified using Cohen’s kappa (κ) or weighted quadratic κ, with values interpreted per Landis–Koch criteria. The Kaplan-Meier method and log-rank test were used for survival analysis. Variables showing a significant or borderline association in the univariate analysis ( p < 0.1) were selected as candidates for the multivariate Cox model to identify independent predictors of survival. Model performance evaluated by Harrell’s concordance index (C-index). A two-sided p-value < 0.05 was considered statistically significant. All results are presented as hazard ratios (HRs) with 95% confidence intervals (CIs).

A total of 42 patients were analyzed, including 8 from our center and 34 from the literature, comprising 22 SCCOPT and 20 SCCOHT cases (Fig.  1 ). Marked clinicopathologic differences were identified between the two subtypes. Compared with SCCOPT, SCCOHT occurred at a younger age (median 25.5 vs. 54.0 years, p  < 0.0001), with larger tumors (15.0 vs. 11.1 cm, p  = 0.039), and more frequent palpable pelvic masses (35.0% vs. 9.1%, p  = 0.041). Hypercalcemia was distinctive for SCCOHT (55.0% vs. 0%, p  = 0.0002). Tumor laterality also differed significantly ( p  = 0.0375). Despite these contrasts, both subtypes commonly presented at advanced FIGO stages with similar symptoms—mainly abdominal pain and distension—and showed comparable rates of elevated serum CA-125 (Table  1 ). Overall, 38 patients underwent surgery, including 9 after neoadjuvant chemotherapy with interval debulking. Prognosis remained poor across subtypes, with no significant difference in median OS (14.4 months for SCCOPT vs. 11.0 months for SCCOHT, p  = 0.431). Comprehensive clinical characteristics, treatments, and outcomes from both our cohort and the literature are detailed in Table  2 and Table S1 . Among the eight cases of SCCO managed at our center, five were SCCOPT, one was SCCOPT associated with a Brenner tumor, and two were SCCOHT. Notably, both SCCOHT cases were initially misdiagnosed as sex cord–stromal tumors. Morphologically, SCCOPT closely resembled small cell carcinoma of the lung. The tumor cells were small with scant cytoplasm, arranged in compact sheets; nuclei were hyperchromatic with the characteristic “salt-and-pepper” chromatin pattern, accompanied by frequent mitotic figures and extensive necrosis (Fig.  2 A and C). Immunohistochemically, SCCOPT consistently expressed epithelial markers and exhibited a typical neuroendocrine phenotype. In this series of 22 SCCOPT cases, Syn was positive in 86.4% (19/22), CgA in 77.3% (17/22), and CD56 in 68.2% (15/22). SMARCA4 (BRG1) expression was retained in all evaluated cases, and the Ki-67 index remained uniformly high (70–90%). PAX8 positivity was also observed in some cases, supporting an ovarian origin (Fig.  2 D and F). By contrast, the 20 SCCOHT cases were composed of uniform small round cells, occasionally forming follicle-like structures (Fig.  3 A and C). Their immunopurified was characterized by complete loss of SMARCA4 expression and diffuse positivity for EMA, WT1, and INI1. The Ki-67 index ranged from 40% to 90%, further indicating high proliferative activity (Fig.  3 D and F). Detailed immunohistochemical findings are summarized in Table  3 and Table S2 . Table 1 Clinicopathologic characteristics of patients with SCCOPT and SCCOHT Characteristic SCCOPT ( n  = 22) SCCOHT ( n  = 20) p value * Age , years < 0.0001 Median 54.0 [46.5–64.5] 25.5 [18.25–33.25] Mean 55.6 ± 11.8 26.4 ± 11.8 Range 32–80 10–60 Tumor laterality 0.038 Right 10 (45.4) 11 (55.0) Left 6 (27.3) 9 (45) Bilateral 6 (27.3) 0.435 FIGO stage 0.099 I 2 (9.1) 0 (0) II 1 (4.5) 6 (30) III 8 (36.4) 6 (30) IV 11 (50.0) 27 (20.8) Symptoms Abdominal pain 15 (68.2) 13 (65.0) 0.827 Abdominal distension 9 (40.9) 9 (45.0) 0.789 Pelvic mass 2 (9.1) 7 (35.0) 0.041 Other 3 (13.6) 5 (25.0) 0.349 Hypercalcemia 0.0002 Yes 0 (0) 11 (55.0) No 17 (77.3) 5 (25.0) Unknown 5 (22.7) 4 (20.0) Epithelial marker CA-125 0.088 Elevated 15 (68.2) 11 (55.0) Normal 0 (0) 4 (20.0) Unknown 7 (31.8) 5 (25.0) Tumor size (cm) 0.039 Median 11.1 [8.3–15.3] 15.0 [11.5–15.8] Mean 11.6 ± 5.1 14.9 ± 4.7 Range 3.5–22.0 8.8–25.0 Duration of follow up (months) 30.0[19.0–48.0] 43.0[13.0-inf] 0.729 Median OS (months) 14.4[8.0-inf] 11.0[8.0–26.0] 0.431 Note: Data are median (interquartile range) or n (%). SCCOPT, small cell carcinoma of the ovary–pulmonary type; SCCOHT, small cell carcinoma of the ovary–hypercalcemic type; CA-125, carbohydrate antigen 125; FIGO, International Federation of Gynecology and Obstetrics; OS, overall survival. *Mann–Whitney U or Student’s t test for continuous variables and Chi-squared test or Fisher exact test for categorical variables Clinicopathologic characteristics of patients with SCCOPT and SCCOHT Note: Data are median (interquartile range) or n (%). SCCOPT, small cell carcinoma of the ovary–pulmonary type; SCCOHT, small cell carcinoma of the ovary–hypercalcemic type; CA-125, carbohydrate antigen 125; FIGO, International Federation of Gynecology and Obstetrics; OS, overall survival. *Mann–Whitney U or Student’s t test for continuous variables and Chi-squared test or Fisher exact test for categorical variables Table 2 Clinical characteristics, treatment modalities, and outcomes of eight patients with SCCO at the First Affiliated Hospital of USTC Patient Age(year) Patho Dx Presentation CA125 (U/ml) Ca (mmol/L) FIGO Operation Chemotherapy DFS(m) OS(m) State 1 63 SCCOPT Abdominal pain 391 2.42 III TAH, BSO, OMT, LYM TC * 8 79.20 79.56 NED 2 53 SCCOPT Abdominal distension 192 2.24 III TAH, RSO, OMT, AP, LYM TP * 10 12.46 14.36 DOD 3 45 SCCOPT Mucus-containing stool 340 2.09 IV TAH, BSO, LAR, LYM NA 0.99 0.99 DOD 4 67 SCCOPT Pelvic mass 317 2.23 IV TAH, BSO, OMT, LYM, AP, S2H, PIR EP * 6 6.53 12.69 DOD 5 39 SCCOPT Abdominal distension 476 2.2 III TAH, BSO, OMT, LYM, AP TC * 6 43.39 43.39 NED 6 68 SCCOPT abdominal distension and pain 578 2.0 IV Laparoscopic exploration TC * 6 NA 3.43 DOD 7 34 SCCOHT abdominal distension 262 3.49 I RSO, OMT NA 2.23 7.76 DOD 8 28 SCCOHT Abdominal pain, Pelvic mass 71 2.73 III RSO, OMT, LYM IEP * 6 6.53 7.62 DOD Note: SCCO: Small cell carcinoma of the ovary, including SCCOPT (pulmonary type) and SCCOHT (hypercalcemic type); TAH: Total abdominal hysterectomy; BSO: Bilateral salpingo-oophorectomy; RSO: Right salpingo-oophorectomy; OMT: Omentectomy; LYM: Lymphadenectomy; AP: Appendectomy; LAR: Low anterior resection; S2H: Segment II hepatectomy; PIR: Partial ileal resection; TC: Carboplatin and paclitaxel; TP: Cisplatin and paclitaxel; EP: Etoposide and cisplatin; IEP: Ifosfamide, etoposide, and cisplatin; DOD: Dead of disease; NED: No evidence of disease; DFS: Disease-free survival; OS: Overall survival. Ca: serum calcium level (mmol/L); Patho Dx: pathological diagnosis Clinical characteristics, treatment modalities, and outcomes of eight patients with SCCO at the First Affiliated Hospital of USTC Note: SCCO: Small cell carcinoma of the ovary, including SCCOPT (pulmonary type) and SCCOHT (hypercalcemic type); TAH: Total abdominal hysterectomy; BSO: Bilateral salpingo-oophorectomy; RSO: Right salpingo-oophorectomy; OMT: Omentectomy; LYM: Lymphadenectomy; AP: Appendectomy; LAR: Low anterior resection; S2H: Segment II hepatectomy; PIR: Partial ileal resection; TC: Carboplatin and paclitaxel; TP: Cisplatin and paclitaxel; EP: Etoposide and cisplatin; IEP: Ifosfamide, etoposide, and cisplatin; DOD: Dead of disease; NED: No evidence of disease; DFS: Disease-free survival; OS: Overall survival. Ca: serum calcium level (mmol/L); Patho Dx: pathological diagnosis Fig. 2 Histopathological and immunohistochemical profile of SCCOPT. ( A – C ) H&E staining (×40, ×100, ×200) showing nested small tumor cells with scant cytoplasm, “salt-and-pepper” chromatin, and focal necrosis. ( D ) Strong nuclear BRG1 (SMARCA4) expression (×100) confirming protein retention. ( E ) Diffuse cytoplasmic Synaptophysin expression (×100) indicating neuroendocrine differentiation. ( F ) Focal nuclear PAX8 staining (×100) supporting an ovarian origin Histopathological and immunohistochemical profile of SCCOPT. ( A – C ) H&E staining (×40, ×100, ×200) showing nested small tumor cells with scant cytoplasm, “salt-and-pepper” chromatin, and focal necrosis. ( D ) Strong nuclear BRG1 (SMARCA4) expression (×100) confirming protein retention. ( E ) Diffuse cytoplasmic Synaptophysin expression (×100) indicating neuroendocrine differentiation. ( F ) Focal nuclear PAX8 staining (×100) supporting an ovarian origin Fig. 3 Histopathological and immunohistochemical profile of SCCOHT. ( A – C ) H&E staining (×40, ×100, ×200) showing diffuse sheets of small tumor cells and characteristic follicle-like structures containing eosinophilic material. ( D ) Complete loss of nuclear BRG1 (SMARCA4) expression (×100) in tumor cells, with stroma serving as an internal positive control. ( E ) Negative Synaptophysin staining (×100) ruling out neuroendocrine differentiation. ( F ) Diffuse nuclear WT1 positivity (×100) supporting an ovarian origin Histopathological and immunohistochemical profile of SCCOHT. ( A – C ) H&E staining (×40, ×100, ×200) showing diffuse sheets of small tumor cells and characteristic follicle-like structures containing eosinophilic material. ( D ) Complete loss of nuclear BRG1 (SMARCA4) expression (×100) in tumor cells, with stroma serving as an internal positive control. ( E ) Negative Synaptophysin staining (×100) ruling out neuroendocrine differentiation. ( F ) Diffuse nuclear WT1 positivity (×100) supporting an ovarian origin Table 3 Immunohistochemical characteristics of eight patients at the First Affiliated Hospital of USTC Patient Pathology Immunohistochemical characteristics 1 SCCOPT(ROV) CD56 (+), Syn (–), CgA (–), EMA (+), CK7 (–), PAX-8 (+), Inhibin-α (–), Melan-A (–), ER (+), PR (–), p16 (+), p53 (+, ~ 80%), Vimentin (+), TTF-1 (–), Ki-67 (+, ~ 70%) 2 SCCOPT(ROV) Syn (+), CgA (+), Vimentin (-),PAX-8 (+), CK7 (-),CK20 (-),CDX-2 (-), EMA (+),CD56 (+), BRG1 (+), Inhibin (–), TTF-1 (–), Ki-67 (+, ~ 60%), 3 SCCOPT(LOV) CK (pan) (+), Syn (+), CD56 (+), CK20 (+), SATB2 (+), PAX-8 (+), p40 (+), LCA (–), CK7 (–), CDX-2 (–), p16 (–), CgA (–), TTF-1 (–), Ki-67 (+, ~ 80%) 4 SCCOPT(ROV) CK (+), CgA (+), CD56 (+), Syn (+), TTF-1 (+), p40 (–), ER (–), PR (–), p16 (+), p53 (–), WT1 (–), Ki-67 (+, ~ 90%) 5 SCCOPT with Brenner tumor (ROV) CgA (+), Syn (+), Napsin A (–), ER (–), PR (+), HNF1-b (+), p16 (+), p53 (+), WT1 (+), HNF1-β (–), PAX-8 (+), BRG1 (+), TTF-1 (–), Ki-67 (+, ~ 70%) 6 SCCOPT(Bilateral) CK (pan) (+), Syn (+), CgA (+), CD56 (+), TTF-1 (-), PAX-8 (+), p53 (+), Desmin (–), CD99 (–), WT1 (–), ER (–), Ki-67 (+, ~ 90%) 7 SCCOHT(ROV) EMA (+), BRG1 (-), INI1 (+), CK (–), PAX-8 (–), CA199 (–), SF-1 (–), Inhibin-α (+), WT1 (+), CK7 (–), Ki-67 (+, ~ 60%) 8 SCCOHT(ROV) CK (pan) (+), EMA (+), WT1 (+), LCA (–), BRG1 (–), Syn (-), CgA (–), SALL4 (+), OCT4 (–), SF-1 (–), Calretinin (+), INI1 (+), PAX-8 (+), PLAP (–), CD117 (+), p53 (+), ER (–), Ki-67 (+, ~ 70%) Note: SCCO = small cell carcinoma of the ovary; SCCOPT = pulmonary type; SCCOHT = hypercalcemic type. Expression results are based on immunohistochemical staining; (+) = positive; (–) = negative; Immunohistochemical characteristics of eight patients at the First Affiliated Hospital of USTC Note: SCCO = small cell carcinoma of the ovary; SCCOPT = pulmonary type; SCCOHT = hypercalcemic type. Expression results are based on immunohistochemical staining; (+) = positive; (–) = negative; SCCOPT showed markedly aggressive imaging features with frequent dissemination. Bilateral ovarian involvement occurred in 6/22 cases (27.3%), irregular morphology in 14/22 (63.6%), and ill-defined margins in 13/22 (59.1%), consistent with infiltrative growth. Tumors were predominantly heterogeneous solid masses, sometimes with cystic components (Fig.  4 A). On contrast-enhanced CT, both primary tumors and metastases demonstrated moderate-to-marked enhancement in all cases (22/22, 100%), producing a characteristic “concurrent enhancement” pattern. Dissemination was common, including peritoneal/omental/mesenteric seeding in 17/22 (77.3%), lymph node metastases in 16/22 (72.7%), and bowel invasion in 7/22 (31.8%) (Fig.  4 B and D). In contrast, SCCOHT exhibited a more circumscribed, “benign-like” appearance despite aggressive behavior. All cases presented as large unilateral cystic–solid pelvic masses (20/20, 100%; median diameter, 15 cm). Lesions were usually round or lobulated (14/20, 70%) with well-defined margins (18/20, 90%), suggesting expansile growth. On CECT, peripheral solid components and thick septa showed mild-to-moderate heterogeneous enhancement in 19/20 cases (95%), while central cystic/necrotic regions remained non-enhancing; >25% necrosis was present in 10/20 (50%) (Fig.  4 E and F). A detailed summary of CT findings is provided in Table S3 . To support the objectivity of qualitative imaging assessment, interobserver agreement between two independent readers was evaluated and is summarized in Table S4 . Overall reliability was high across key CT features (κ range 0.81–1.00, almost perfect agreement), including enhancement degree and necrosis extent assessed with quadratic weighted κ. Fig. 4 Contrast-enhanced CT findings of SCCOPT and SCCOHT. ( A – D ) SCCOPT: bilateral irregular adnexal masses (9.8 × 9.6 × 10.4 cm) with ill-defined margins and heterogeneous enhancement ( A ), enlarged retroperitoneal lymph nodes ( B ), a heterogeneously enhancing hepatic segment II metastasis ( C ), and coronal reconstruction confirming bilateral adnexal involvement ( D ). ( E – F ) SCCOHT: a large pelvic mass (11.7 × 7.5 × 8.5 cm) with heterogeneous enhancement ( E ), and coronal reconstruction showing a bulky but localized lesion without nodal or distant metastasis ( F ). Arrows indicate tumor lesions Contrast-enhanced CT findings of SCCOPT and SCCOHT. ( A – D ) SCCOPT: bilateral irregular adnexal masses (9.8 × 9.6 × 10.4 cm) with ill-defined margins and heterogeneous enhancement ( A ), enlarged retroperitoneal lymph nodes ( B ), a heterogeneously enhancing hepatic segment II metastasis ( C ), and coronal reconstruction confirming bilateral adnexal involvement ( D ). ( E – F ) SCCOHT: a large pelvic mass (11.7 × 7.5 × 8.5 cm) with heterogeneous enhancement ( E ), and coronal reconstruction showing a bulky but localized lesion without nodal or distant metastasis ( F ). Arrows indicate tumor lesions To assess the diagnostic and prognostic value of CT imaging in SCCO, imaging features of 42 patients were correlated with histological subtype, immunophenotype, and clinical outcomes. As shown in Fig.  5 A, Spearman correlation analysis revealed several significant associations. BRG1 loss, indicative of SMARCA4-deficient SCCOHT, was negatively correlated with enhancement ( r = − 0.64, p  < 0.001), peritoneal metastases ( r = − 0.40, p  = 0.012), and bowel invasion ( r = − 0.41, p  = 0.011), suggesting less aggressive behavior and reduced angiogenic activity. In contrast, high Ki-67 expression correlated positively with aggressive CT features, including irregular morphology, marked enhancement, peritoneal seeding, distant metastasis, and bowel invasion ( r  = 0.33–0.47, all p  < 0.05), linking proliferative activity to invasive potential. Notably, in the SCCOPT subgroup ( n  = 22), CgA expression showed a negative correlation with margin clarity ( r = − 0.45, p  = 0.035), suggesting that reduced or absent CgA expression, indicative of poorer neuroendocrine differentiation, was associated with ill-defined, infiltrative tumor margins on CT, consistent with greater malignant potential. Among the 37 patients with follow-up data (88%), the median follow-up duration was 30 months. Cox regression analysis identified tumor morphology and bowel invasion as independent predictors of OS. Irregular morphology was associated with a higher risk of mortality (HR = 3.75, 95% CI: 1.10–12.77, p  = 0.03, C-index = 0.66, 95% CI: 0.56–0.75), whereas bowel invasion conferred an even greater risk (HR = 3.13, 95% CI: 1.13–8.66, p  = 0.01; C-index = 0.62, 95% CI: 0.52–0.71) (Table  4 ). Kaplan–Meier survival analysis further confirmed these associations, demonstrating significantly shorter OS in patients with irregular compared to regular morphology (median 10 vs. 26 months, p  = 0.010) and in those with bowel invasion compared to no invasion (median 8 vs. 24 months, p  = 0.0078) (Fig.  5 B and C). Among 36 patients who achieved R0 resection, 24 (67%) developed recurrence or metastasis, with a median DFS of 6.5 months, reflecting the poor prognosis associated with SCCO. Table 4 Univariate and multivariate Cox proportional hazards analysis of prognostic factors for OS ( n  = 37) Variable Univariable Multivariable HR(95%CI) p value HR(95%CI) p value FIGO Stage (III–IV vs. I–II) 3.37(2.91–3.91) 0.27 Age (years) 0.99(0.97–1.01) 0.54 Size(cm) 1.02(0.95–1.09) 0.65 Operation(Yes vs. No) 0.46(0.13–1.58) 0.22 Tumor Morphology (Irregular vs. Regular) 3.33(1.28–8.67) 0.01 3.75(1.10-12.77) 0.03 Margin (Ill-defined vs. well-defined) 1.02(0.45–2.32) 0.31 Internal architecture(Cystic solid vs. Solid) 1.02(0.45–2.32) 0.96 Enhancement degree (Mild vs. Moderate/Marked) 1.31(0.55–3.11) 0.54 Necrosis(≥ 25% vs. <25%) 0.75(0.33–1.72) 0.50 Lymph Node Metastasis (Yes vs. No) 2.47(1.01–6.07) 0.05 2.56(0.89–7.38) 0.08 Omental Metastasis (Yes vs. No) 2.38(0.87–6.49) 0.09 0.41(0.08–2.11) 0.29 Distant Metastasis (Yes vs. No) 2.17(0.91–5.13) 0.08 1.48(0.47–4.71) 0.50 Bowel Invasion (Yes vs. No) 3.22(1.30–8.01) 0.01 3.13(1.13–8.66) 0.03 Note: Variables with univariate p  < 0.10 were included in the multivariable Cox model; p  < 0.05 was considered significant. OS, overall survival; HR, hazard ratio; CI, confidence interval; FIGO, International Federation of Gynecology and Obstetrics Univariate and multivariate Cox proportional hazards analysis of prognostic factors for OS ( n  = 37) Note: Variables with univariate p  < 0.10 were included in the multivariable Cox model; p  < 0.05 was considered significant. OS, overall survival; HR, hazard ratio; CI, confidence interval; FIGO, International Federation of Gynecology and Obstetrics Fig. 5 Correlation of CT imaging features with immunophenotype and prognosis in SCCO. ( A ) Correlation network diagram visualizing the associations between qualitative CT features and quantitative radiomic features. To improve clarity and interpretability, only statistically significant correlations ( p   0.3 are displayed. Red and blue lines indicate positive and negative correlations, respectively. The width of the connecting lines is proportional to the correlation strength, and statistical significance is denoted by asterisks (* p  < 0.05, ** p  < 0.01, *** p  < 0.001). ( B ) Kaplan–Meier overall survival (OS) curves stratified by tumor morphology. ( C ) Kaplan–Meier OS curves stratified by bowel invasion. SCCO, small cell carcinoma of the ovary Correlation of CT imaging features with immunophenotype and prognosis in SCCO. ( A ) Correlation network diagram visualizing the associations between qualitative CT features and quantitative radiomic features. To improve clarity and interpretability, only statistically significant correlations ( p   0.3 are displayed. Red and blue lines indicate positive and negative correlations, respectively. The width of the connecting lines is proportional to the correlation strength, and statistical significance is denoted by asterisks (* p  < 0.05, ** p  < 0.01, *** p  < 0.001). ( B ) Kaplan–Meier overall survival (OS) curves stratified by tumor morphology. ( C ) Kaplan–Meier OS curves stratified by bowel invasion. SCCO, small cell carcinoma of the ovary

Small cell carcinoma of the ovary (SCCO) is an exceedingly rare and highly aggressive malignancy, accounting for less than 0.01% of all ovarian cancers [ 1 ]. Histologically, SCCO is classified into two distinct subtypes: the hypercalcemic type (SCCOHT) and the pulmonary type (SCCOPT). SCCOHT predominantly affects adolescents and young women and is frequently associated with paraneoplastic hypercalcemia, whereas SCCOPT occurs more commonly in perimenopausal and postmenopausal women [ 2 – 4 ]. Despite these differences, both subtypes exhibit highly aggressive biological behavior, extremely poor prognoses, and only rare reports of long-term survival. The nonspecific clinical presentation and extreme rarity of SCCO often result in limited clinician awareness, frequent misdiagnosis, and delays in initiating appropriate treatment, thereby compromising patient outcomes [ 5 , 6 ]. Although the histopathological features of SCCO have been partially elucidated, its imaging characteristics—particularly those observed on CT—remain poorly understood [ 7 , 8 ]. The current literature consists largely of isolated case reports, in which the imaging findings are typically described as nonspecific and often resemble those of other ovarian malignancies, including epithelial carcinomas, sex cord–stromal tumors, and germ cell tumors [ 9 ]. This absence of well-defined radiologic descriptors poses significant challenges for early recognition. Nevertheless, given the highly malignant nature and rapid clinical progression of SCCO, it is plausible that certain characteristic CT patterns may exist. When assessed in conjunction with clinical indicators—such as patient age, hypercalcemia, bilateral ovarian involvement, and neuroendocrine differentiation—radiologists may be able to improve diagnostic accuracy. A systematic analysis of CT features is therefore warranted to better define the imaging spectrum of SCCO and to explore their potential clinical implications. The present study aims to systematically analyze the CT imaging features of 42 patients with pathologically confirmed SCCO, including 22 cases of the pulmonary subtype and 20 cases of the hypercalcemic subtype. In contrast to prior reports that were largely descriptive, this work aims to bridge radiologic phenotypes with immunophenotypic features. Beyond delineating the subtype-specific imaging spectrum, we quantitatively evaluate the associations between CT findings and key immunohistochemical markers—particularly SMARCA4/BRG1, Ki-67, and neuroendocrine markers—to assess whether distinct immunopurified are reflected in recognizable imaging patterns. We further examine the prognostic value of CT features for overall survival (OS) and disease-free survival (DFS). By integrating radiologic, clinical, and pathological data, this study seeks to provide novel and clinically actionable insights to support preoperative differential diagnosis and risk stratification for this rare ovarian malignancy.

SCCO is an exceptionally rare gynecologic malignancy with an extremely poor prognosis, making early diagnosis and accurate assessment critical for clinical management. In contrast to prior literature that is largely limited to descriptive case reports or small series, this study provides incremental contributions in three aspects: (i) a structured CT spectrum with direct subtype-level comparison between SCCOPT and SCCOHT using uniform imaging-feature definitions; (ii) quantitative associations between CT phenotypes and immunohistochemical markers (e.g., SMARCA4/BRG1, Ki-67, and neuroendocrine markers), thereby linking imaging appearances to tumor immunophenotypes; and (iii) identification of CT-visible prognostic indicators that may support preoperative risk stratification. Collectively, our findings offer preliminary, clinically actionable evidence to aid diagnosis, differential diagnosis, and prognostic evaluation in this rare ovarian carcinoma. Both SCCOPT and SCCOHT presented as large, complex ovarian masses with internal necrosis, consistent with rapid tumor growth [ 5 ]. However, significant subtype-specific radiological differences were observed. SCCOPT more frequently displayed imaging features indicative of high-grade malignancy, including irregular or lobulated contours, predominantly solid composition, and moderate-to-marked progressive enhancement. Metastatic spread was also more common in SCCOPT, with high rates of lymph node involvement, peritoneal seeding, distant metastasis, bowel invasion, and ascites, in line with its histopathological resemblance to pulmonary small cell carcinoma and its known aggressive biological behavior [ 3 , 25 ]. In contrast, SCCOHT demonstrated deceptively indolent imaging characteristics, such as well-circumscribed margins and mild enhancement, which sometimes led to misdiagnosis as a sex cord-stromal tumor. However, the apparent imaging “mildness” contrasted sharply with the dismal outcomes, as the median OS of SCCOHT patients in this study was only 11 months, consistent with previous reports describing this paradoxical discrepancy between indolent radiologic appearance and aggressive clinical behavior [ 34 ]. This paradox may be explained by its molecular profile; negative correlations between BRG1 (SMARCA4) loss and tumor enhancement suggest impaired angiogenesis. Meanwhile, its clinical aggressiveness is likely driven by other intrinsic factors, such as high Ki-67 indices and genomic instability from chromatin remodeling defects [ 35 – 37 ]. This dissociation between imaging features and clinical outcomes highlights the unique biology of SCCOHT and underscores the importance of integrated molecular–pathological assessment. Previous studies on SCCO imaging are extremely limited, largely restricted to isolated case reports that describe nonspecific solid ovarian masses with heterogeneous enhancement, resembling epithelial ovarian carcinoma or sex cord–stromal tumors [ 38 , 39 ]. Our findings are consistent with those of Chen et al. [ 7 ], who reported necrotic solid masses in SCCOPT, but extend prior work by quantitatively characterizing enhancement patterns and metastatic features. The 77.3% rate of peritoneal dissemination in our SCCOPT cohort exceeded the 46% previously reported, likely reflecting the higher proportion of advanced-stage disease (86.4% FIGO stage III–IV) in our series. The mild enhancement observed in SCCOHT is consistent with previous reports linking SMARCA4 deficiency to reduced vascularity [ 40 , 41 ]. Mechanistically, BRG1 (SMARCA4), as the ATPase subunit of the SWI/SNF chromatin-remodeling complex, has been shown to help maintain chromatin accessibility at endothelial enhancers that regulate endothelial gene expression and angiogenic capacity [ 42 ]. Accordingly, BRG1 loss may perturb these transcriptional programs, providing a biologically plausible explanation for the relatively “hypovascular” phenotype—manifesting as weaker perfusion and reduced contrast enhancement—observed in our cohort. Although functional and perfusion-based validation is warranted, this observation underscores an important clinical paradox: the aggressive course of SCCOHT despite its “mild” radiologic appearance suggests that prognosis may be driven by mechanisms beyond vascular proliferation, such as epigenetic dysregulation and/or immune-evasive processes, consistent with the perspective proposed by Raupach et al. [ 43 ]. These findings carry important clinical implications. For SCCOPT, recognition of moderate-to-marked enhancement, necrosis, and metastatic features on CT should alert radiologists to its high-risk nature and prompt early multidisciplinary discussion and aggressive treatment strategies, including optimal cytoreduction followed by platinum-based chemotherapy (e.g., TC or TP regimens). Correlations with Ki-67 and CgA further support the integration of immunohistochemistry into therapeutic planning, potentially guiding the use of neuroendocrine-targeted agents. Survival analyses identified tumor morphology and bowel invasion as independent predictors of overall survival, with irregular morphology and bowel invasion significantly associated with shorter OS. These findings underscore the prognostic value of CT features in risk stratification and may inform both surgical decision-making and postoperative management. For SCCOHT, imaging interpretation requires particular caution because lesions may appear deceptively indolent. Although SCCOHT commonly presents as a large unilateral cystic–solid mass with well-defined margins, mild-to-moderate peripheral/septal enhancement, and central necrosis, this “benign-like” morphology can lead to misdiagnosis as sex cord–stromal tumors, germ cell tumors, or even borderline tumors—especially in younger patients. To improve real-world differential diagnosis, we now emphasize several pragmatic cues that should trigger the consideration of SCCOHT: (i) clinical context, particularly young age and unexplained paraneoplastic hypercalcemia; (ii) disproportionate tumor bulk (often > 10–15 cm) with thick septa/peripheral solid components despite a circumscribed contour; (iii) extensive central necrosis within an apparently encapsulated mass; and (iv) a profound discordance between “mild” enhancement and rapid clinical progression. When these features are present, we recommend early coordination with pathology for SMARCA4/BRG1 immunohistochemistry, which can complement imaging and enable an earlier definitive diagnosis. Overall, by bridging CT phenotypes with immunophenotypic markers, our results provide a preliminary framework for developing imaging-informed biomarkers that may improve preoperative differential diagnosis and risk stratification in this rare but lethal ovarian malignancy. While this study offers novel insights, several limitations must be acknowledged. First, the primary limitation is the retrospective design and profound sample heterogeneity, arising from the amalgamation of our small institutional cohort with cases from numerous publications. This multi-source approach introduced confounding biases in patient selection, imaging protocols, and data quality. Second, the sample size, while substantial for an exceedingly rare tumor like SCCO, remains limited ( n = 42), which restricts the statistical power for robust inference. Third, although imaging assessments were performed by experienced radiologists, some degree of subjectivity is unavoidable. Future studies should further improve objectivity and cross-center generalizability by incorporating quantitative approaches, such as radiomics-derived texture features and perfusion-based biomarkers, to enhance reproducibility [ 44 ]. Fourth, the absence of genomic or multi-omics data restricted deeper exploration of the molecular underpinnings of SCCO, particularly the aggressive biology of SCCOHT [ 36 ]. In addition, heterogeneity in treatment approaches, owing to the long study period and individualized clinical decision-making, may have confounded survival analyses [ 5 , 37 ]. Future research should expand case numbers through multi-institutional collaboration to strengthen statistical robustness.

This study provides the first systematic analysis of CT imaging features in SCCO, delineating the differences between pulmonary and hypercalcemic subtypes in radiologic appearance, biological behavior, and prognosis. Our findings highlight the potential role of CT signs in preoperative assessment and prognostic prediction and demonstrate meaningful correlations between imaging characteristics and pathological markers such as BRG1, CgA, and Ki-67, thereby extending prior largely descriptive reports toward imaging–pathology integration. Future research should focus on expanding sample sizes and integrating advanced imaging techniques with molecular pathology to elucidate the imaging biology of SCCO further and improve early diagnosis and prognostic evaluation of this lethal disease.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4