Clinicopathological characterization and prognostic risk modeling in patients with synchronous ovarian and endometrial cancer: a population-based study.

Ovarian cancer (OC) and endometrial cancer (EC) are two common malignant tumors of the female reproductive system, with high incidence rates in China and around the world. According to the relevant literature, the occurrence of two or more primary malignant tumors of the female reproductive system at the same time is relatively rare, accounting for about 1%–2% of gynecological malignant tumors [ 1 ] . The most common gynecological malignant tumors are synchronous endometrial and ovarian cancer (SEOC), accounting for about 50%–70% [ 2 ] . The concept was first introduced by Ulbright and Roth [ 3 ] in 1985, whose pathological features are dominated by low-grade tumors and low pathological stages. About 2.6%–10% [ 4 ] of patients with OC may have primary EC. The diagnostic criteria of SEOC distinctly diverge from those of single primary ovarian tumors, with notable variations also observed in the available clinical treatment options. Compared to single-primary tumors, SEOC presents unique diagnostic and therapeutic challenges, emphasizing the need for focused research on its clinicopathological features and prognostic implications, as this distinction significantly impacts staging and treatment planning [ 4 ] . Existing studies have established key diagnostic criteria for SEOC, primarily based on tumor independence (e.g. absence of myometrial invasion or ovarian surface involvement). However, several critical knowledge gaps remain unresolved. First, the morphological overlap between SEOC and metastatic tumors frequently leads to misclassification, resulting in inaccurate staging and suboptimal treatment strategies. Second, while SEOC generally exhibits superior overall survival (OS) compared to advanced OC, the underlying mechanisms for the observed prognostic heterogeneity remain unclear, particularly regarding the impact of molecular subtypes (e.g. mismatch repair deficiency, POLE mutations) on clinical outcomes. Finally, there is currently no consensus on optimal surgical approaches (e.g. extent of lymphadenectomy) or adjuvant therapy regimens, especially for intermediate-risk cases, highlighting significant clinical decision-making challenges. Using the SEER database, this study aims to [ 1 ] : systematically analyze the demographic, histological, and staging patterns of SEOC compared to single-primary tumors to clarify clinicopathological correlates [ 2 ] ; evaluate key prognostic determinants including age, histotype, and treatment modality that influence overall and cancer-specific survival; and [ 3 ] bridge existing knowledge gaps by providing evidence-based insights for refining diagnostic algorithms and developing risk-stratified treatment protocols. Through comprehensive integration of population-level epidemiological data with detailed clinicopathological characteristics, our findings will significantly enhance diagnostic accuracy, guide more precise therapeutic decision-making, and ultimately improve prognostic stratification for SEOC patients. This study was conducted without the use of artificial intelligence tools in data collection, analysis, or manuscript preparation, in compliance with the TITAN Guidelines 2025 [ 5 ] .

In this study, we leveraged the comprehensive data provided by the United States Surveillance, Epidemiology, and End Results (SEER) Program, accessible via https://seer.cancer.gov/ . Utilizing the SEER*Stat 8.4.3 client selection database, specifically the Incidence-SEER Research Data from 17 Registries, November 2022 Submission, covering the period from 2000 to 2020, we extracted pertinent clinical information pertaining to patients with double primary tumors. This cross-sectional study utilized data from the SEER database (2010–2015). A meticulous screening process identified 447 patients with complete clinical, survival, and prognosis data, employing primary site International Classification of Diseases for Oncology codes C56 and C54. These patients were then subjected to a rigorous comparative study and analysis, with a benchmark cohort of 20 769 cases of single primary OC. The inclusion criteria were stringent, requiring OC as the initial primary tumor and EC (identified by histological codes 8380/3, 8382/3, 8383/3) as the subsequent primary tumor, with a diagnosis year falling within the range of 2010–2015. Exclusion criteria included cases with a diagnostic interval exceeding 6 months, incomplete clinical or follow-up data, or ambiguous histological diagnoses. These criteria were established to minimize confounding variables and ensure data consistency for reliable prognostic modeling. The inclusion variables encompassed demographic characteristics, oncological features, and therapeutic interventions. Demographic characteristics comprised age, race, marital status, and income. Oncological features encompassed histological type, degree of tumor differentiation, SEER stage, AJCC TNM stage, tumor size, regional lymph node status, number of lymph node dissections, and CA125 levels. Therapeutic interventions consisted of radiotherapy and chemotherapy. Furthermore, outcome variables, namely follow-up time and survival status, were also encompassed within the study. HIGHLIGHTS Clinical relevance: Our findings reveal that synchronous endometrial and ovarian cancer (SEOC) patients exhibit distinct clinicopathological characteristics (e.g. younger age, early-stage endometrioid histology) and superior survival outcomes compared to single-primary ovarian cancer, challenging the conventional view that SEOC primarily represents metastatic disease. Predictive model: We developed a nomogram with high discriminatory power (area under the receiver operating characteristic curve 0.872–0.880) for predicting 1-, 2-, and 3-year mortality, offering clinicians a tool to stratify high-risk patients and tailor therapeutic strategies. Public health implications: By identifying age, tumor stage, and histological type as independent prognostic factors, this study underscores the need for early detection and personalized management of SEOC, aligning with the journal’s focus on cancer prevention and control. Clinical relevance: Our findings reveal that synchronous endometrial and ovarian cancer (SEOC) patients exhibit distinct clinicopathological characteristics (e.g. younger age, early-stage endometrioid histology) and superior survival outcomes compared to single-primary ovarian cancer, challenging the conventional view that SEOC primarily represents metastatic disease. Predictive model: We developed a nomogram with high discriminatory power (area under the receiver operating characteristic curve 0.872–0.880) for predicting 1-, 2-, and 3-year mortality, offering clinicians a tool to stratify high-risk patients and tailor therapeutic strategies. Public health implications: By identifying age, tumor stage, and histological type as independent prognostic factors, this study underscores the need for early detection and personalized management of SEOC, aligning with the journal’s focus on cancer prevention and control. Descriptive statistics characterized the cohort. Survival analyses employed Kaplan–Meier methods and Cox proportional hazards models to identify prognostic factors. Data were processed and analyzed utilizing the statistical software R (version 4.3.1). The optimal cut-off values for the continuous variables (age and tumor size) were identified and categorized employing the X-Tile software. A comparative analysis of the disparities in clinical features, pathological characteristics, and prognosis was undertaken between the DPG and SPG, leveraging propensity score matching (PSM) to balance the baseline variables between the two groups in a 1:1 ratio. The variables matched included age, race, marital status, histological type, tumor differentiation grade, SEER staging, AJCC staging, TNM (the size and local extent of the primary tumor (T category), the involvement of regional lymph nodes (N category), and the presence of distant metastases (M category)) staging, tumor size, regional lymph node status, lymph node dissection count, CA125 level, and radiotherapy status. A secondary analysis of the baseline characteristics was conducted for the matched cohorts to facilitate comparison. Survival analysis was conducted using both the Kaplan–Meier methodology and the Cox proportional hazards regression model to identify the independent prognostic factors in patients with dual-primary malignancies. Ovarian epithelial malignancies (corresponding to histological codes 8441/3, 8460/3, 8461/3, 8470/3, 8371/3, 8480/3, 8380/3, 8382/3, 8383/3, and 8310/3) within the DPG were screened, and their baseline demographics and clinical characteristics were analyzed. A P -value of less than 0.05 was deemed statistically significant.

A total of 21 390 patients from the SEER database who met the eligibility criteria were included in this study. The baseline characteristics are presented in Table 1 , while the constituent components are illustrated in Figure 1 . The proportion of patients younger than 55 years old in the dual-primary group (DPG) (66.2%) was significantly higher than that in the single-primary group (SPG) (38.2%). In the DPG, the majority of histological types were ovarian ECs, accounting for 60.9%. Furthermore, among tumor grades, hyperdifferentiated/moderately differentiated tumors were the most prevalent, comprising 54.4% of cases. Regional lymph nodes were predominantly negative, comprising 62.4% of cases. Additionally, CA125 levels were predominantly positive, accounting for 69.8% of cases. The majority of patients in the DPG were classified as early-stage, with focal disease (208 [46.5%]) and AJCC stage I (228 [51.0%]). Furthermore, the majority of patients were categorized as T1 (241 [53.9%]), N0 (396 [88.6%]), and M0 (414 [92.6%]) in terms of tumor staging. In the SPG, in contrast to the DPG, plasmacytoid histological type was the most prevalent (51%), and the majority of tumors were classified as poorly differentiated/undifferentiated (52.3%). Furthermore, the regional lymph nodes were predominantly positive (65.7%). The tumors were predominantly in intermediate to advanced stages, with a high incidence of distant metastases (61.4%) and classified as AJCC stage III/IV (64.6%). Furthermore, the majority of tumors were classified as T3 stage (54.3%). The median survival time for patients in the DPG was significantly longer, at 80 months, compared to 56 months in the SPG. Figure 1. The proportion of dual-primary group and single-primary group. (A) The composition ratio of the dual-primary group (left) and the single-primary group (right) before PSM. (B) The composition ratio of the dual-primary group (left) and the single-primary group (right) after PSM. Table 1 Clinicopathological characteristics of patients in the dual-primary group vs. single-primary group Before PSM After PSM Characteristic Double Single P Double Single P 447 20 769 425 425 Age (%) <0.001 0.617  55 151 (33.8) 12 831(61.8) 147 (34.6) 160 (37.7) Race (%) 0.005 0.485  White 367 (82.1) 16 878 (81.3) 347 (81.6) 362 (85.2)  Black 18 (4.0) 1663 (8.0) 18 (4.2) 16 (3.8)  Other 53 (11.9) 1883 (9.1) 51 (12.0) 42 (9.9)  Chinese 9 (2.0) 345 (1.7) 9 (2.1) 5 (1.2) Marital_status (%) <0.001 0.481  Married 209 (46.8) 10 441 (50.3) 202 (47.5) 188 (44.2)  Unmarried 139 (31.1) 4506 (21.7) 128 (30.1) 128 (30.1)  Other 99 (22.1) 5822 (28.0) 95 (22.4) 109 (25.7) Diagnose-treatment a (%) <0.001 0.720  <1 370 (82.8) 14 723 (70.9) 352 (82.8) 347 (81.6)  ≥1 77 (17.2) 6046 (29.1) 73 (17.2) 78 (18.4) Hist b (%) <0.001 0.985  Serous 49 (11.0) 10 590 (51.0) 49 (11.5) 44 (10.4)  Mucinous 17 (3.8) 1187 (5.7) 17 (4.0) 18 (4.2)  Endometrioid 272 (60.9) 1788 (8.6) 250 (58.8) 254 (59.7)  Clear cell 15 (3.4) 1327 (6.4) 15 (3.5) 14 (3.3)  Other 94 (21.0) 5877 (28.3) 94 (22.1) 95 (22.4) Grade (%) <0.001 0.792  Well/moderately 243 (54.4) 3835 (18.5) 226 (53.2) 232 (54.6)  Poorly/undifferentiated 130 (29.1) 10 861 (52.3) 129 (30.4) 120 (28.2)  Unk 74 (16.6) 6073 (29.2) 70 (16.5) 73 (17.2) Laterality (%) <0.001 0.096  Paired 124 (27.7) 9488 (45.7) 121 (28.5) 94 (22.1)  Left 159 (35.6) 5513 (26.5) 146 (34.4) 164 (38.6)  Right 164 (36.7) 5768 (27.8) 158 (37.2) 167 (39.3) Tumor_size (%) <0.001 0.628  250 78 (17.4) 5775 (27.8) 77 (18.1) 67 (15.8) Nodes_examined c (%) <0.001 0.404  Yes 317 (70.9) 10 839 (52.2) 297 (69.9) 309 (72.7)  No 130 (29.1) 9930 (47.8) 128 (30.1) 116 (27.3) Nodes_positive d (%) <0.001 0.154  Negative 279 (62.4) 7115 (34.3) 259 (60.9) 280 (65.9)  Positive 168 (37.6) 13 654 (65.7) 166 (39.1) 145 (34.1) LN_Sur e (%) 3 260 (58.2) 8128 (39.1) 245 (57.6) 263 (61.9)  Unk 12 (2.7) 592 (2.9) 11 (2.6) 12 (2.8) CA_125 f (%) 0.003 0.918  Negative 22 (4.9) 1924 (9.3) 22 (5.2) 24 (5.7)  Positive 312 (69.8) 14 385 (69.3) 298 (70.1) 300 (70.6)  Unk 113 (25.3) 4460 (21.5) 105 (24.7) 101 (23.8) SEER_stage (%) <0.001 0.181  Distant 119 (26.6) 12 756 (61.4) 117 (27.5) 105 (24.7)  Localized 120 (26.8) 3451 (16.6) 108 (25.4) 132 (31.1)  Regional 208 (46.5) 4562 (22.0) 200 (47.1) 188 (44.2) AJCC_stage (%) <0.001 0.600  I 228 (51.0) 5484 (26.4) 213 (50.1) 231 (54.4)  II 84 (18.8) 1851 (8.9) 79 (18.6) 77 (18.1)  III 102 (22.8) 8192 (39.4) 100 (23.5) 86 (20.2)  IV 33 (7.4) 5242 (25.2) 33 (7.8) 31 (7.3) T (%) <0.001 0.747  T1 241 (53.9) 5915 (28.5) 226 (53.2) 242 (56.9)  T2 98 (21.9) 2697 (13.0) 93 (21.9) 86 (20.2)  T3 107 (23.9) 11 283 (54.3) 105 (24.7) 96 (22.6)  Tx 1 (0.2) 772 (3.7) 1 (0.2) 1 (0.2)  T0 0 (0.0) 102 (0.5) 0 (0.0) 0 (0.0) N (%) <0.001 0.376  N0 396 (88.6) 14 827 (71.4) 375 (88.2) 386 (90.8)  N1 45 (10.1) 4579 (22.0) 44 (10.4) 36 (8.5)  Nx 6 (1.3) 1363 (6.6) 6 (1.4) 3 (0.7) M (%) <0.001 0.897  M0 414 (92.6) 15 527 (74.8) 392 (92.2) 394 (92.7)  M1 33 (7.4) 5242 (25.2) 33 (7.8) 31 (7.3) DX_bone g (%) 0.049 0.616  Yes 0 (0.0) 155 (0.7) 0 (0.0) 0 (0.0)  No 444 (99.3) 20 272 (97.6) 422 (99.3) 424 (99.8)  Unk 3 (0.7) 342 (1.6) 3 (0.7) 1 (0.2) DX_liver g (%) 0.419 0.404  Yes 1 (0.2) 39 (0.2) 1 (0.2) 1 (0.2)  No 442 (98.9) 20 377 (98.1) 420 (98.8) 423 (99.6)  Unk 4 (0.9) 353 (1.7) 4 (1.0) 1 (0.2) DX_brain g (%) 0.419 0.404  Yes 1 (0.2) 39 (0.2) 1 (0.2) 1 (0.2)  No 442 (98.9) 20 377 (98.1) 420 (98.8) 423 (99.6)  Unk 4 (0.9) 353 (1.7) 4 (1.0) 1 (0.2) DX_lung g (%) 0.001 0.605  Yes 8 (1.8) 1052 (5.1) 8 (1.9) 8 (1.9)  No 436 (97.5) 19 347 (93.2) 414 (97.4) 416 (97.9)  Unk 3 (0.7) 370 (1.8) 3 (0.7) 1 (0.2) Radia_recode h (%) <0.001 0.417  Yes 26 (5.8) 264 (1.3) 15 (3.5) 10 (2.4)  No 421 (94.2) 20 505 (98.7) 410 (96.5) 415 (97.6) Chemo_recode h (%) <0.001 0.240  Yes 313 (70.0) 16 181 (77.9) 297 (69.9) 280 (65.9)  No 134 (30.0) 4588 (22.1) 128 (30.1) 145 (34.1) Income (%) 0.006 0.960  69 999 235 (52.6) 9507 (45.8) 223 (52.5) 224 (52.7) Survival_months (median [IQR]) 80.0 [60.0,102.0] 56.0 [21.0,86.0] <0.001 80.0 [59.0, 102.0] 78.0 [50.0, 98.0] 0.414 a Diagnose-treatment: Time of treatment from diagnosis. b Hist: Type of histopathology. c Nodes_examined: For a regional lymph node examination. d Nodes_positive: Whether the regional lymph nodes were positive. e LN_Sur: Number of lymph node dissection during surgery. f CA_125: Serological levels of the tumor marker CA125. g DX_bone/liver/brain/lung: Whether a distant metastasis occurs. h Radia/Chemo_recode: Whether radiotherapy/chemotherapy. The proportion of dual-primary group and single-primary group. (A) The composition ratio of the dual-primary group (left) and the single-primary group (right) before PSM. (B) The composition ratio of the dual-primary group (left) and the single-primary group (right) after PSM. Clinicopathological characteristics of patients in the dual-primary group vs. single-primary group Diagnose-treatment: Time of treatment from diagnosis. Hist: Type of histopathology. Nodes_examined: For a regional lymph node examination. Nodes_positive: Whether the regional lymph nodes were positive. LN_Sur: Number of lymph node dissection during surgery. CA_125: Serological levels of the tumor marker CA125. DX_bone/liver/brain/lung: Whether a distant metastasis occurs. Radia/Chemo_recode: Whether radiotherapy/chemotherapy. All clinicopathological variables were comprehensively encompassed within the univariate analysis. As depicted in Table 2 , age, marital status, time interval from diagnosis to treatment, histological types, grade, tumor laterality, tumor size, status of regional lymph nodes, tumor stage, TNM staging, and liver metastasis emerged as significant factors correlated with the prognosis of patients within the DPG. Subsequently, variables displaying a trend of statistical significance ( P < 0.05) in the univariate analysis were incorporated into a multivariate Cox proportional hazards regression model. Age, marital status, histological types, tumor size, SEER stage of the tumor, and AJCC stage were identified as independent prognostic factors. Table 2 Univariable and multivariate Cox regression analysis based on all variables of SEOC patients Uni-cox Mul-cox Characteristic HR CI P HR CI P Age 2.12 1.48–3.04 0 1.48 1–2.2 0.0509 Race 1.21 0.93–1.56 0.1564 Marital_status 1.41 1.07–1.84 0.013 1.41 1.06–1.86 0.0174 Diagnose-treatment a 1.91 1.15–3.16 0.0125 1.42 0.81–2.46 0.2172 Hist b 0.42 0.33–0.53 0 0.78 0.59–1.03 0.0785 Grade 1.53 1.19–1.97 0.0011 1.3 0.94–1.79 0.1122 Laterality 0.75 0.57–0.99 0.0451 1.13 0.84–1.52 0.4323 Tumor_size 1.38 1.06–1.79 0.0172 1.44 1.08–1.92 0.0128 Nodes_examined c 2.62 1.69–4.07 0 1 0.25–4.03 0.9962 Nodes_positive d 3.98 2.52–6.28 0 1.17 0.43–3.13 0.7602 LN_Sur e 0.62 0.5–0.78 0 0.87 0.51–1.48 0.5998 CA_125 f 1.04 0.68–1.59 0.8442 SEER_stage 0.45 0.34–0.6 0 1.49 1–2.22 0.0506 AJCC_stage 2.67 2.16–3.31 0 2.52 1.36–4.68 0.0034 T 2.96 2.28–3.85 0 1.36 0.76–2.45 0.3032 N 2.47 1.68–3.65 0 1.01 0.5–2.04 0.9808 M 6.27 3.69–10.67 0 0.51 0.21–1.28 0.1519 DX_bone g 1.93 0.27–13.92 0.5122 DX_liver g 0.1 0.02–0.39 0.001 0.38 0.1–1.43 0.1529 DX_brain g 1.93 0.27–13.92 0.5122 DX_lung g 0.39 0.11–1.4 0.1492 Radia_recode h 0.88 0.36–2.18 0.7837 Chemo_recode h 0.81 0.5–1.33 0.4126 Income 0.72 0.52–1.01 0.0544 a Diagnose-treatment: Time of treatment from diagnosis. b Hist: Type of histopathology. c Nodes_examined: For a regional lymph node examination. d Nodes_positive: Whether the regional lymph nodes were positive. e LN_Sur: Number of lymph node dissection during surgery. f CA_125: Serological levels of the tumor marker CA125. g DX_bone/liver/brain/lung: Whether a distant metastasis occurs. h Radia/Chemo_recode: Whether radiotherapy/chemotherapy. Univariable and multivariate Cox regression analysis based on all variables of SEOC patients Diagnose-treatment: Time of treatment from diagnosis. Hist: Type of histopathology. Nodes_examined: For a regional lymph node examination. Nodes_positive: Whether the regional lymph nodes were positive. LN_Sur: Number of lymph node dissection during surgery. CA_125: Serological levels of the tumor marker CA125. DX_bone/liver/brain/lung: Whether a distant metastasis occurs. Radia/Chemo_recode: Whether radiotherapy/chemotherapy. With the relevant factors influencing prognosis serving as independent variables and OS as the dependent variable, Kaplan–Meier (K–M) curves were generated for the comparative analysis of the two groups. The log-rank test was then employed to statistically assess the differences between the groups (Figs. 2 – 4 ). Furthermore, a forest plot illustrating the comparison of survival rates between the two groups was constructed, utilizing the 5-year OS rate as the basis (Fig. 5 ). Within the DPG, the 5-year OS rate was approximately 49% for patients with serous OC and 90% for those with endometrioid OC. The 5-year OS rate was significantly higher among highly/moderately differentiated patients compared to poorly differentiated/undifferentiated patients (90% vs. 54%). It was also higher in patients with negative CA125 levels than in those with positive levels (91% vs. 78%) and higher among patients with negative lymph node involvement than in those with positive involvement (88% vs. 64%). The 5-year OS rates were approximately 94% and 85% for tumors staged as regional and focal, respectively, while the 5-year OS rate for patients with distant metastases was approximately 52%. Figure 2. K–M curves and cumulative death risk curves for overall survival in dual-primary group and single-primary group. Life tables for patients at risk are given below each plot. (A) The K–M curves of dual-primary group before PSM. (B) The cumulative death risk curves of dual-primary group before PSM. (C) The K–M curves of dual-primary group after PSM. (D) The cumulative death risk curves of dual-primary group after PSM. Figure 3. K–M curves for overall survival in different subgroups of two groups before PSM. (A) Age. (B) Marital status. (C) SEER stage. (D) Time of treatment from diagnosis. (E) Status of the regional lymph nodes. (F) Tumor size. (G) Grade. (H) AJCC stage. (I) Histological types. Figure 4. K–M curves for overall survival in different subgroups of two groups after PSM. (A) SEER stage. (B) Status of the regional lymph nodes. (C) AJCC stage. Figure 5. ForestMap of subgroup analysis based on 5-year overall survival in two groups. (A) Before PSM. (B) After PSM. K–M curves and cumulative death risk curves for overall survival in dual-primary group and single-primary group. Life tables for patients at risk are given below each plot. (A) The K–M curves of dual-primary group before PSM. (B) The cumulative death risk curves of dual-primary group before PSM. (C) The K–M curves of dual-primary group after PSM. (D) The cumulative death risk curves of dual-primary group after PSM. K–M curves for overall survival in different subgroups of two groups before PSM. (A) Age. (B) Marital status. (C) SEER stage. (D) Time of treatment from diagnosis. (E) Status of the regional lymph nodes. (F) Tumor size. (G) Grade. (H) AJCC stage. (I) Histological types. K–M curves for overall survival in different subgroups of two groups after PSM. (A) SEER stage. (B) Status of the regional lymph nodes. (C) AJCC stage. ForestMap of subgroup analysis based on 5-year overall survival in two groups. (A) Before PSM. (B) After PSM. In contrast, the 5-year OS rate after distant metastasis was merely 31% in the SPG, whereas the 5-year OS rates among patients with positive CA125 and detected lymph nodes (44% and 38%, respectively) were inferior to those observed in the double-primary group (78% and 64%, respectively). The overall 5-year OS was approximately 79% in the DPG versus 51% in the SPG, exhibiting a statistically significant difference in survival prognosis between the two cohorts ( P < 0.001) (Fig. 5 ). Following PSM-matching, patients in the DPG with distant metastases and detected lymph node positivity demonstrated superior 5-year OS compared to those in the SPG ([52% vs. 39%, P = 0.057] and [64% vs. 54%, P = 0.038]). Similarly, patients with AJCC stage III and T3 tumors in the DPG had significantly higher 5-year OS than their counterparts in the SPG ([60% vs. 43%, P = 0.013] and [53% vs. 37%, P = 0.022]). The K–M curves for OS did not reveal any statistically significant differences between the DPG and SPG (Fig. 2 ). Although SEOC patients exhibited better survival outcomes compared to single-primary OC patients, the SEER database has inherent limitations, including the absence of detailed treatment regimens and potential biases due to its retrospective nature [ 6 ] . These constraints necessitate cautious interpretation of the findings and underscore the need for further validation in diverse populations. In addition, the baseline characteristics of epithelial malignant tumors of the ovary in the DPG are shown in Supplemental Digital Content Table 1 , available at: http://links.lww.com/MS9/A979 . The median survival time for epithelial malignant tumors of the ovary was 82 months, with endometrioid ovarian carcinoma being the most prevalent subtype (272 cases [77.05%]), exhibiting a median survival time of 87 months. Furthermore, the K–M curves and cumulative mortality risks associated with various histological types of ovarian epithelial malignant tumors are depicted in Supplemental Digital Content Figure 1, available at: http://links.lww.com/MS9/A978 . The aforementioned factors influencing the prognosis of biprotic tumors were incorporated into an R language-based model, with variables exhibiting covariance being excluded, to develop a column-line graph model capable of personalizing the prediction of 1-, 2-, and 3-year specific mortality risks for biprotic patients. The length of each line segment in the column chart represents the extent of the impact that each factor exerts on the patient’s survival and prognosis. Each subtype of prognostic risk factor is assigned a corresponding score on the upper scale, and the cumulative total of these scores, representing the summation of all prognostic risk factors, is depicted as a vertical line along the total score scale positioned at the base of the model. This vertical line serves as an indicator for deriving the individualized risk of mortality for each patient. As depicted in Supplemental Digital Content Figure 2, available at: http://links.lww.com/MS9/A978 , the primary factors influencing the patient’s prognosis encompass AJCC stage, grading, and the status of lymph node dissection undergone by the patient. When considering Table 1 , it is evident that the percentage of patients with AJCC stage I is significantly higher in the DPG compared to the SPG (51% vs. 26.4%). Furthermore, 70.9% of patients in the double primary group underwent lymph node dissection, potentially indicative of a more favorable prognosis. When contrasted with single primary OC, the double primary group exhibited a lower proportion of poorly differentiated and undifferentiated tumors; however, the scores assigned to these tumor types in the table were higher, signifying an unfavorable prognosis. The C-index test, time-dependent ROC curve, and time-dependent AUC curve were employed to assess the discriminatory power. The AUC values corresponding to the 1-, 2-, and 3-year time points on the ROC curve were 0.880, 0.872, and 0.872, respectively. Calibration curves were utilized to validate the model’s goodness-of-fit, demonstrating a close alignment with the 45° diagonal line, thereby indicating a satisfactory degree of discrimination and calibration performance (Supplemental Digital Content Figure 2, available at: http://links.lww.com/MS9/A978 ).

It is not rare for OC and EC to coexist in clinical work. But one primary tumor is often misdiagnosed as a metastasis from the other, especially when histopathological types are the same [ 7 ] . Therefore, distinguishing the primary double cancers and metastatic tumors is still a diagnostic challenge. The tumor is staged as IA if it is a double primary cancer. If it is metastatic cancer, it is staged as IIIA based on the endometrium or II based on the ovaries. If diagnosed as IA, no additional treatment is required. If diagnosed as IIIA or II, additional treatment is required. Ulbright et al first proposed the pathological diagnosis of differentiating primary double cancers and metastatic tumors, on the basis of which Scully and Young et al further improved and formulated more perfect diagnostic criteria [ 8 ] , which are now commonly used in the clinic. Some scholars in the literature categorized primary double cancers into three types according to the pathological types of the two cancer foci, in which type A refers to both ovarian and endometrial endometrioid carcinoma, and in SEOC, type A is the main type, accounting for 61.9% [ 9 ] . Wang et al [ 10 ] studied 51 primary double cancer patients, with both cancers, endometrioid carcinoma accounted for 65%, and the prognosis was much better than that of patients. In the study of Khalid et al [ 11 ] , 66.6% and 83.3% of ovarian and ECs had the pathologic type of endometrioid carcinoma. In the inclusion criteria of the present study, the histological codes of endometrial carcinoma were 8380/3, 8382/3, and 8383/3, and the histological type of synchronous ovarian carcinoma was endometrioid, accounting for 77.05%, which corroborates prior research identifying endometrioid histology and early-stage disease as key favorable prognostic factors in SEOC. However, they contrast with recent molecular profiling studies suggesting SEOC primarily represents metastatic EC. This inconsistency highlights the need for further investigation into SEOC’s unique biological characteristics. In previous studies, most patients with double primary tumors were 41–54 years old [ 12 ] , and in this study, most patients were 43–55 years old at the time of their consultation, and patients with single primary OC tend to be seen at an age >55 years (61.8%). And the median age of onset of OC was reported to be 63 years old in the previous literature [ 13 ] , compared to which, patients with double primary were characterized by a younger age of onset. As mentioned in previous reports, clinical symptoms in patients with SEOC are often not specific, and the earliest signs appearing in patients are abnormal vaginal bleeding, followed by abdominal pain [ 7 ] , as well as vaginal discharge and pelvic masses, abdominal distension, ascites, and other clinical symptoms. Although the clinical signs appear early, there is no clear specificity. The possibility of double cancers cannot be ruled out when patients have multiple clinical manifestations. Meanwhile, even if there are no symptoms such as vaginal bleeding in patients with OC, endometriosis cannot be completely ruled out because, according to previous reports, some patients with primary double cancers have lower abdominal pain as their main complaint at the time of diagnosis, and double cancers are not a common disease in clinical practice, which often leads to neglect of such diseases. Therefore, in clinical practice, no matter whether the patient has obvious symptoms or not, a comprehensive preoperative evaluation should be performed to avoid underdiagnosis. In addition to the lack of specific clinical manifestations, its tumor markers also lack specificity. Literature about SEOC reported that most patients had elevated CA125 levels before surgery, but other tumor markers, such as CA19-9 and HE4, were not reported, and some studies even suggested that the preoperative level of CA125 was an independent factor affecting the prognosis of double cancers [ 14 ] . The change of CA125 level has been an important biomarker for monitoring the disease changes of epithelial OC. In this study, CA125 level was found to lack sensitivity and specificity for the diagnosis of EC, but some studies have reported that CA125 can be used as a predictor of the spread of EC outside the uterus and prognosis [ 15 ] . In this study, a one-way analysis of CA125 levels revealed that the difference between CA125-positive and CA125-negative was not statistically significant ( P > 0.05). Therefore, primary double cancers of the endometrium and ovary are characterized by a lack of specific tumor markers, and future studies with larger sample sizes are needed to further confirm the role of tumor markers in distinguishing primary double cancers from metastatic cancers. In addition, there is a lack of standardized diagnostic and treatment criteria for SEOC, which makes over-treatment or under-treatment inevitable in clinical practice. Surgery is internationally recognized as the treatment of choice, and there is controversy about whether adjuvant therapy should be performed after surgery. It has been suggested that postoperative adjuvant therapy can be determined by the presence of risk factors for recurrence in endometrial and ovarian lesions [ 16 ] . Although the majority of patients with SEOC in this study had early postoperative pathologic staging, most of them underwent postoperative adjuvant therapy under the guidance of their clinicians. Taking into account the condition of the patients and their personal wishes, it is in line with the viewpoint put forward by Lou et al that postoperative adjuvant therapy should be performed even in patients with early SEOC [ 17 ] . Although the optimal treatment for SEOC remains to be determined, a study by Chiang et al concluded that adjuvant chemotherapeutic treatment should be particularly indicated for patients with advanced disease [ 18 ] . Currently, there are limited molecular studies on SEOC. Some scholars performed high-throughput sequencing technology on all tumors of 22 patients with SEOC and found at least one shared mutation in PTEN, AKT1, P1K3CA, KRAS, TP53, and ARID1A in all the tumors, confirming the clonal origins of uterine and ovarian tumors in SEOC [ 19 ] . Anglesio et al [ 20 ] found clonal linkage in group 17/18 SEOC by targeted deep gene sequencing, suggesting that most of the SEOC were not independent primary tumors but metastatic tumors. Schultheis et al [ 6 ] found clonal linkage in all 17 cases of SEOC by massively parallel sequencing. Recently, some scholars not only found that 92% of SEOC had clonal association, but also found that their molecular profiles were extremely similar to those of EC in The Cancer Genome Atlas (TCGA), and therefore concluded that SEOC was actually metastatic EC [ 21 ] . However, the molecular finding that SEOC is essentially a metastatic tumor contradicts its better clinical prognosis. Given the view that most disseminated SEOCs are essentially endometrial metastatic carcinomas, as confirmed by the WES technique in recent years, we believe that assessment of prognostic risk in female patients may be more clinically relevant than determining the primary site. The strength of our study is that it no longer overemphasizes tumor primary versus metastasis, but is based on regression analysis to identify factors that influence prognosis and groups at higher risk of recurrence to guide treatment decisions. This may be a more reliable model for guiding clinical management than traditional categorization methods. This study was based on the SEER database to determine the independent risk factors of patients with double primary tumors, but this study also had the following shortcomings [ 1 ] : this study was a retrospective study based on the SEER database, and some cases with incomplete information were screened out when collecting the clinical data of the patients, which resulted in a selective bias [ 2 ] ; although there are Asian populations in the database, they account for a relatively small proportion, and most of them are European and American populations, so it is necessary to improve the applicability by increasing the sample size of Asian populations [ 3 ] ; the database lacks detailed information on specific treatment regimens, such as neoadjuvant radiotherapy, chemotherapy, and immunotherapy, and the types of drugs used in these regimens, which may limit the assessment of treatment effects [ 4 ] ; databases usually have some delay in updating data, and the results of the study may not be fully consistent with the latest medical practices and treatment protocols. Our study leveraged the extensive dataset provided by the SEER program, offering a robust platform for analyzing SEOC. While the SEER database’s large sample size strengthens the statistical power of our analysis, we acknowledge a critical limitation regarding the underrepresentation of certain ethnic groups, particularly Asians. This demographic imbalance may introduce biases that affect the generalizability of our findings to these populations. To ensure the broad representativeness of our research conclusions, we recommend further studies with larger Asian cohorts to validate our results. Collaborative efforts with international registries, including those with significant Asian populations, could provide a more comprehensive assessment of the SEOC prognosis and the applicability of our prognostic model across different ethnicities [ 22 ] . Additionally, we suggest that future research employ stratified analyses to explore potential ethnic differences in SEOC outcomes. Such analyses could reveal unique risk factors or prognostic indicators specific to underrepresented groups, thereby enhancing the precision of clinical management strategies for SEOC patients globally. In summary, while our study provides valuable insights into SEOC prognosis, the potential biases introduced by the SEER database’s demographic composition must be considered. We advocate for increased diversity in cancer registry data to ensure that research conclusions are applicable to all populations affected by SEOC. In summary, the pathologic type of double primary tumors was mostly highly differentiated/moderately differentiated early endometrioid carcinoma, with a high rate of regional lymph node negativity, and fewer distant metastases appeared, so the prognosis was good. Age, marital status, histological typing, tumor size, and tumor stage were independent factors affecting prognosis. However, it is still necessary to expand the sample size of the study to explore the clinicopathological characteristics of double primary tumors, and new technologies such as second-generation sequencing can be used to analyze and validate the molecular level, in order to increase the understanding of this disease from multiple perspectives and in multiple dimensions, so as to provide the theoretical basis for the establishment of molecular diagnostic criteria, and assist in the establishment of more convenient and effective diagnostic and treatment modes.

SEOC exhibits distinct clinicopathological characteristics, predominantly featuring endometrioid histology, early-stage presentation, and favorable prognosis. Key prognostic determinants include age, marital status, histological subtype, tumor size, and disease stage. Our findings highlight three critical implications: first, establishing precise pathological-molecular diagnostic criteria to differentiate between synchronous primary tumors and metastatic disease is of paramount clinical importance; second, age-adjusted diagnostic vigilance requires special emphasis; third, the development of risk-stratified treatment algorithms is imperative. Future research should focus on [ 1 ] : advancing molecular characterization using next-generation sequencing technologies [ 2 ] ; validating current findings in ethnically diverse populations; and [ 3 ] developing integrated diagnostic models that combine pathological and molecular criteria. These findings provide a significant foundation for optimizing SEOC management paradigms while identifying key scientific questions that necessitate multinational collaborative research to address.