Biological consistency and molecular heterogeneity in synchronous endometrial and ovarian carcinomas: a clinicopathological study within the FIGO 2023 framework.

Approved by the Ethics Committee of West China Second University Hospital (No. 2023 − 140), this retrospective study analyzed 48 SEOC cases with complete clinicopathological, and follow-up data treated between January 2015 and December 2022. The inclusion criteria included: (1) age ≥ 18 years; (2) histopathological confirmation of synchronous primary endometrial and ovarian carcinomas according to Scully’s diagnostic criteria [ 4 ]; (3) availability of complete clinicopathological data and reliable follow-up. Exclusion criteria comprised: (1) incomplete medical records or unreliable follow-up; (2) non-epithelial malignancies (borderline ovarian tumors, sex cord-stromal tumors, mesenchymal tumors, mixed epithelial-stromal tumors, Brenner tumors, uterine sarcomas, or carcinosarcomas); (3) primary/secondary surgeries performed outside our institution. Comprehensive clinicopathological parameters were extracted from medical records, including: patient demographics (age, body mass index (BMI), reproductive history, menopausal status, family history, comorbidities); preoperative assessments (imaging: ultrasound/ computed tomography (CT) / magnetic resonance imaging (MRI); tumor markers: CA125/CA19-9); surgical details (approach, intraoperative findings, residual tumor status); adjuvant therapy (chemotherapy regimens/cycles; radiotherapy protocols and courses); oncologic outcomes (recurrence timing/location, survival data); and pathological characteristics (histologic type, grade, FIGO stage, tumor dimensions, lymph node metastasis, LVSI, depth of myometrial invasion (MI), cervical stromal involvement, presence of atypical endometrial hyperplasia (AEH) or ovarian endometriosis, and immunohistochemistry (IHC): p53, estrogen receptor, progesterone receptor, mismatch repair (MMR). All histopathological slides underwent centralized review by at least two pathologists. The study population was identified through a two-step rigorous diagnostic and staging process. First, to distinguish synchronous primary cancers from metastatic disease, we applied the comprehensive clinicopathologic criteria proposed by Scully et al. [ 4 ] (detailed in Table 1 ). While ‘histologic dissimilarity’ is a hallmark of independent primaries, for cases with concordant histology (specifically low-grade endometrioid carcinomas in both sites), we relied on a multi-parametric approach. This included evaluating the absence of deep MI, absence of LVSI, and presence of associated precursor lesions (e.g., AEH or ovarian endometriosis) to confirm their independent origins. Subsequently, these cases were stratified according to the 2023 FIGO system to isolate the Stage IA3 subgroup for prognostic evaluation. Table 1 Histopathological diagnosis and differentiation criteria for synchronous endometrial carcinoma and ovarian cancer Independent Primary Tumors Endometrial Primary Ovarian Secondary Ovarian Primary Endometrial Secondary 1.Histologic dissimilarity of the tumors 1.Histologic similarity of the tumors 1. Histologic similarity of the tumors 2. No or only superficial myometrial invasion of endometrial tumor 2. Large endometrial tumor - small ovarian tumor(s) 2. Large ovarian tumor - small endometrial tumor 3.No vascular space invasion of endometrial tumor 3. Atypical endometrial hyperplasia additionally present 3. Ovarian endometriosis present 4.Atypical endometrial hyperplasia additionally present 4. Deep myometrial invasion a. Direct extension into adnexa b. Vascular space invasion in myometrium 4. Location in ovarian parenchyma 5.Absence of other evidence of the spread of endometrial tumor 5. Spread elsewhere in typical pattern of endometrial carcinoma 5. Direct extension from ovary predominantly into outer wall of uterus 6.Ovarian tumor unilateral (80 to 90% of cases) 6. Ovarian tumors bilateral and/or multinodular 6. Spread elsewhere in the typical pattern of ovarian carcinoma 7.Ovarian tumor located in parenchyma 7. Hilar location, vascular space invasion, surface implants, or combination in ovary 7. Ovarian tumor unilateral (80 to 90% of cases) and forming single mass 8.No vascular space invasion, surface implants, or predominant hilar location in ovary 8. Ovarian endometriosis absent 8. No atypical hyperplasia in endometrium 9.Absence of other evidence of spread of ovarian tumor 9. Aneuploidy with similar DNA indices or diploidy of both tumors* 9. Aneuploidy with similar DNA indices or diploidy of both tumors* 10.Ovarian endometriosis present 10. Similar molecular genetic or karyotypic abnormalities in both tumors 10. Similar molecular genetic or karyotypic abnormalities in both tumors 11. Different ploidy or DNA indices, if aneuploid, of the tumors* 12.Dissimilar molecular genetic or karyotypic abnormalities in the tumors *The possibility of tumor heterogeneity must be considered in the evaluation of the ploidy findings Histopathological diagnosis and differentiation criteria for synchronous endometrial carcinoma and ovarian cancer 2. No or only superficial myometrial invasion of endometrial tumor 2. Large endometrial tumor - small ovarian tumor(s) 3. Atypical endometrial hyperplasia additionally present 4.Atypical endometrial hyperplasia additionally present 4. Deep myometrial invasion a. Direct extension into adnexa b. Vascular space invasion in myometrium 5. Spread elsewhere in typical pattern of endometrial carcinoma 5. Direct extension from ovary predominantly into outer wall of uterus 7. Hilar location, vascular space invasion, surface implants, or combination in ovary *The possibility of tumor heterogeneity must be considered in the evaluation of the ploidy findings All 48 patients were surgically staged. Historically, endometrial carcinoma was staged according to the 2009 FIGO criteria, and ovarian carcinoma according to the 2014 FIGO system. For the purpose of this study, all cases were retrospectively re-evaluated and re-staged according to the updated FIGO 2023 staging system for endometrial cancer. Given that molecular classification (e.g., POLE mutation status, p53) was not routinely performed for all patients during the study period (2015–2022), staging was assigned based on the “Morphology-based Algorithm” as permitted by the FIGO 2023 guidelines for settings where molecular testing is unavailable or incomplete. Following the 2023 FIGO staging system and to address biological heterogeneity, the 48 confirmed SEOC cases were stratified into three distinct groups: Group A (FIGO IA3, n = 12): Strictly defined as low-grade concordant endometrioid carcinomas limited to the uterus and a unilateral ovary, with no or only superficial myometrial invasion (< 50%), absence of substantial LVSI, and no additional metastases [ 7 ]. Group B (Advanced Concordant, n = 26): Cases with identical histology at both sites that failed the IA3 criteria due to features such as deep MI, presence of LVSI, or bilateral ovarian involvement. Group C (Independent Discordant, n = 10): Cases with different histotypes in the ovary and endometrium. Due to the retrospective nature of the study and the availability of paraffin-embedded tissue blocks, a subset of 17 patients with paired samples from both endometrial and ovarian lesions was selected for molecular analysis. This sub-cohort was used to evaluate the concordance of MMR proteins and p53 expression between the two tumor sites. Molecular Discordance Definition: Inter-tumor molecular status was defined as ‘discordant’ if there was a discrepancy in any of the markers (MMR proteins or p53) between the endometrial and ovarian components. For p53, any difference between wild-type and mutation-type (null, overexpression) was recorded as discordant. All 48 patients underwent radical surgery, comprising total hysterectomy, bilateral salpingo-oophorectomy, with or without omentectomy, with or without pelvic lymphadenectomy, with or without para-aortic lymph node sampling/dissection, with or without appendectomy. Given the rarity of this malignancy and limited literature, adjuvant therapy plans were formulated through multidisciplinary tumor board discussions based on intraoperative findings and final pathology, incorporating patient-centered shared decision-making. Clinical data regarding adjuvant therapies, including chemotherapy regimens and radiotherapy, were extracted from medical records. Patient follow-up was conducted through outpatient clinic reviews and telephone interviews. The survival period was calculated in months, with the surgery date as the starting point and December 31, 2024, as the cutoff date. Progression-free survival (PFS) was defined as time from surgery to disease recurrence, metastasis, or death. Cases without events were censored at the last follow-up. Overall survival (OS) was defined as time from surgery to all-cause death. Surviving patients were censored at the last follow-up. Follow-up was conducted once every 3 months for the first 2 years postoperatively, once every 6 months during years 3–5, and annually thereafter. The follow-up assessments include Pelvic examination, serum tumor markers (CA125, CA19-9), and imaging (CT, vaginal color Doppler ultrasound). Statistical analyses were performed using IBM SPSS Statistics 27 (IBM Corp., Armonk, NY, USA). Survival curves (Fig.  1 ) were generated using GraphPad Prism software (version 10.0, GraphPad Software, Boston, MA, USA). The heatmap for molecular profiling (Fig.  2 ) was visualized using Anaconda Navigator (Anaconda3, version 2.x). Categorical variables were compared using Fisher’s exact test. Continuous variables, such as age and BMI, were analyzed using the Mann-Whitney U test. OS and PFS were estimated using the Kaplan-Meier method, with differences between groups evaluated by the log-rank test. To specifically compare the prognostic impact of the dual tumor components, clinicopathological parameters of the endometrium and ovary were evaluated independently as potential predictors of survival. Univariate analyses were conducted to identify significant risk factors within each site. A p -value < 0.05 was considered statistically significant.

A total of 48 patients with SEOC were identified and stratified into three cohorts based on the FIGO 2023 framework and clinicopathological findings. Group A (Stage IA3) included 12 patients (25.0%), Group B (Advanced Concordant) included 26 patients (54.2%), and Group C (Independent Discordant) included 10 patients (20.8%). As shown in Table  2 , there were no significant differences among the three groups regarding age ( p  = 0.859), BMI ( p  = 1.000), parity ( p  = 0.548), or menopausal status ( p  = 0.318). Preoperative CA125 levels were elevated (≥ 35 U/mL) in the majority of patients across all groups (83.3%, 84.6%, and 70%, respectively; p  = 0.544). Table 2 Clinicopathological characteristics of SEOC patients stratified by subgroups ( N  = 48) Characteristics Group A ( n =12) Group B ( n =26) Group C ( n =10) p Age (years) Median (Range) 46 (31–52) 46.5(27–54) 50(28–59) 0.859 BMI (kg/m 2 )  Median (Range) 21.3 (19.3–29.9) 23.4 (16.4–35.2) 22.2 (18.4–34.8) 1.000 Parity 0.548  Nulliparous 3 (25%) 8(30.8%) 1 (10%)  Multiparous 9 (75%) 18(69.2%) 9 (90%) Menopausal State 0.318  Premenopausal 12 (100%) 22(84.6%) 8 (80%)  Postmenopausal 0 (0%) 4(15.4%) 2 (20%) Initial Symptom 0.293  Vaginal bleeding 5(41.7%) 17(65.4%) 4(40%)  Others 7(58.3%) 9(34.6%) 6(60%) Preoperative CA125 (U/mL) 0.544  <35 2 (16.7%) 4(15.4%) 3(30%)  ≥ 35 10 (83.3%) 22(84.6%) 7(70%) Preoperative CA19-9 (U/mL) 0.388  <34.1 2 (16.7%) 5(19.2%) 4(40%)  ≥ 34.1 10 (83.3%) 21(80.8%) 6(60%) Adjuvant Therapy 1.000  Yes 11(91.7%) 24(92.3%) 9(90%)  None 1 (8.3%) 2(7.7%) 1(10%) Endometrial Cancer Features  Histotype, n (%) 0.208   Endometrioid 12(100%) 26(100%) 9(90%)   Non-Endometrioid 0(0%) 0(0%) 1(10%) Grade, n (%) 0.040*  Grade 1–2 12(100%) 26(100%) 8(80%)  Grade 3 0(0%) 0(0%) 2(20%) Stage, n (%) 0.004*  Ⅰ 12(100%) 18(69.2%) 4(40%)  Non-Ⅰ 0(0%) 8(30.8%) 6(60%) MI, n (%) 0.157  < 50% 12(100%) 25(96.2%) 8(80%)  ≥ 50% 0(0%) 1(3.8%) 2(20%) LVSI, n (%) 0.023*  Negative 12(100%) 23(88.5%) 6(60%)  Positive 0(0%) 3(11.5%) 4(40%) Cervical Stromal Invasion, n (%) 0.125  Negative 12(100%) 23(88.5%) 7(70%)  Positive 0(0%) 3(11.5%) 3(30%) Ovarian Cancer Features  Histotype, n (%) <0.01*   Endometrioid 12(100%) 26(100%) 1(10%)   Non-Endometrioid 0(0%) 0(0%) 9(90%) Grade, n (%) 0.006*  Grade 1–2 12(100%) 23(88.5%) 5(50%)  Grade 3 0(0%) 3(11.5%) 5(50%) Stage, n (%) 0.003*  Ⅰ 12(100%) 14(53.8%) 4(40%)  Non-Ⅰ 0(0%) 12(46.2%) 6(60%) Laterality 0.089  unilateral 12(100%) 19(73.1%) 9(90%)  bilateral 0(0%) 7(26.9%) 1(10%) Lymph Node Metastasis, n (%) 0.002*†  Negative 12(100%) 24(92.3%) 5(50%)  Positive 0(0%) 0(0%) 4(40%)  Unknown 0(0%) 2(7.7%) 1(10%) Cytology of Ascites, n (%) 0.048*†  Negative 6(50%) 17(65.4%) 4(40%)  Positive 0(0%) 5(19.2%) 4(40%)  Unknown 6(50%) 4(15.4%) 2(20%) Associated Pathology, n (%)  AEH 3 (25%) 16(61.5%) 2(20%) 0.028*  Ovarian endometriosis 8(66.7%) 7(26.9%) 3(30%) 0.065 Data are presented as median (range) or number (percentage). BMI: Body Mass Index; Categorical variables were compared using Fisher’s exact test; Continuous variables were compared using Mann-Whitney U test. * p  < 0.05 indicates statistical significance. † p -values for Lymph Node Metastasis and Cytology of Ascites were calculated comparing Positive vs. Negative cases, excluding patients with unknown status Clinicopathological characteristics of SEOC patients stratified by subgroups ( N  = 48) Data are presented as median (range) or number (percentage). BMI: Body Mass Index; Categorical variables were compared using Fisher’s exact test; Continuous variables were compared using Mann-Whitney U test. * p  < 0.05 indicates statistical significance. † p -values for Lymph Node Metastasis and Cytology of Ascites were calculated comparing Positive vs. Negative cases, excluding patients with unknown status Detailed baseline clinicopathological features are summarized in (Table  2 ). By design, variables utilized to define the clinical-biological strata—such as myometrial invasion depth, LVSI status, tumor grade, and histological type—demonstrated distinct distributions across Groups A, B, and C. Because these variations are intrinsic to the sub-grouping criteria themselves, they represent expected architectural baselines rather than independent novel findings. For instance, the absence of deep myometrial invasion and substantial LVSI in Group A, or the predominant non-endometrioid histotypes in Group C (90%), are reflective of the morphology-based algorithm within the FIGO 2023 framework. To evaluate the true validity and clinical utility of this framework, our analysis strictly relies on outcome metrics completely independent of the initial grouping definitions: namely, long-term survival rates and paired molecular profiling patterns. Analysis of associated precursor lesions revealed that AEH was significantly more frequent in Group B (61.5%) compared to Group A (25.0%) and Group C (20.0%) ( p  = 0.028). Conversely, concurrent ovarian endometriosis was most prevalent in the Stage IA3 cohort (Group A, 66.7%), although the overall difference across all three groups narrowly missed the threshold for statistical significance ( p  = 0.065). Regarding management, all 48 patients underwent primary surgery comprising hysterectomy with bilateral salpingo-oophorectomy, plus variable combinations of pelvic and para-aortic lymphadenectomy (87.5%), omentectomy (91.7%), and appendectomy (72.9%). Hysterectomy types included infra-fascial (85.4%), modified radical (12.5%), and subtotal (2.1%, due to renal transplant adhesions). Peritoneal cytology was performed in 75.0% of cases (18.8% positive). Adjuvant therapy was determined by multidisciplinary review: 79.2% (38/48) received platinum-based chemotherapy, 10.4% (5/48) received chemoradiation, and 10.4% (5/48) received no adjuvant therapy. Notably, within the Stage IA3 subgroup (Group A), 91.7% (11/12) of patients received adjuvant chemotherapy, reflecting the historical treatment intensity for this population prior to the implementation of the FIGO 2023 staging system. With a median follow-up of 59.5 months (cutoff: December 31, 2024), 47 patients (97.9%) completed follow-up. Five deaths and six disease progressions/recurrences occurred. The 5-year OS and PFS rates for the entire cohort were 89.1% and 87.5%, respectively. Kaplan-Meier estimates of overall survival in patients are presented in (Fig.  1 ). Fig. 1 Kaplan-Meier survival analysis of overall survival (OS) in SEOC patients. A Stratification by FIGO 2023-based subgroups: Group C (70.0%) showed a lower 5-year OS compared to Group A (90.9%) and Group B (95.5%) ( p = 0.051). B Stratification by ovarian histological grade: High-grade (G3) ovarian tumors were associated with significantly worse survival than low-grade (G1-2) tumors (62.5% vs. 94.5%, p = 0.003). C Stratification by lymph node metastasis: Node-positive patients exhibited significantly reduced survival compared to node-negative patients (50.0% vs. 97.0%, p < 0.01). Note: p -values were calculated using the log-rank test Kaplan-Meier survival analysis of overall survival (OS) in SEOC patients. A Stratification by FIGO 2023-based subgroups: Group C (70.0%) showed a lower 5-year OS compared to Group A (90.9%) and Group B (95.5%) ( p = 0.051). B Stratification by ovarian histological grade: High-grade (G3) ovarian tumors were associated with significantly worse survival than low-grade (G1-2) tumors (62.5% vs. 94.5%, p = 0.003). C Stratification by lymph node metastasis: Node-positive patients exhibited significantly reduced survival compared to node-negative patients (50.0% vs. 97.0%, p < 0.01). Note: p -values were calculated using the log-rank test As shown in Fig.  1 , survival outcomes exhibited divergent trends among the three clinical cohorts. Group B (Advanced Concordant) demonstrated a favorable prognosis, with a 5-year OS of 95.5% and a 5-year PFS of 92.3%. Group A (Stage IA3) followed closely, with both 5-year OS and PFS reaching 90.9%. In contrast, Group C (Independent Discordant) exhibited the poorest outcomes, with both 5-year OS and PFS dropping to 70.0%. Although these survival variations between the three stratification groups demonstrated statistical significance for PFS in univariate analysis (Table  3 ), the inter-group difference for OS achieved marginal statistical significance (log-rank p  = 0.051 for OS; p  = 0.158 for PFS across the total model, while specific group comparisons showed localized separation). Table 3 Univariate analysis affecting the prognosis of patients Factors N 5-year OS(%) p 5-year PFS(%) p Age 0.291 0.647  ≤45y 21 95.2 90.5  ＞45y 27 84.7 85.2 BMI 0.494 0.969  ＜25 32 87.5 87.5  ≥25 16 92.9 87.1 Parity 0.815 0.526  Nulliparous 12 91.7 83.3  Multiparous 36 88.3 88.9 Menopausal Status 0.076 0.137  Postmenopausal 6 66.7 66.7  Premenopausal 42 92.8 90.4 Presenting Symptoms 0.764 0.513  Vaginal bleeding 26 87.4 84.6  Other 22 90.5 90.5 CA 125 0.225 0.220  ＜35 9 100 100  ≥35 39 86.4 84.4 CA 19-9 0.908 0.713  ＜34.1 11 90.9 90.9  ≥34.1 37 88.6 86.2 Adjuvant Therapy 0.422 0.531  Yes 43 87.8 88.2  No 5 100 80 Endometrial carcinoma Histological Type 0.733 0.712  Endometrioid 47 88.9 87.1  Non- Endometrioid 1 100 100 Histological Grade 0.625 0.597  Grade 1-2 46 88.6 86.8  Grade 3 2 100 100 FIGO Stage (2023) 0.658 0.892  Ⅰ 34 91.1 88.1  Non-Ⅰ 14 85.1 85.7 MI 0.544 0.512  ＜1/2 45 88.3 86.5  ≥1/2 3 100 100 Cervical stromal invasion 0.597 0.738  Positive 6 83.3 83.3  Negative 42 89.8 87.9 LVSI 0.084 0.175  Positive 7 68.6 71.4  Negative 41 92.6 90.1 Ovarian carcinoma  Histological Type 0.008* 0.031*   Endometrioid 39 94.3 92.2   Non- Endometrioid 9 66.7 66.7 Histological Grade 0.003* 0.015*  Grade 1-2 40 94.5 92.4  Grade 3 8 62.5 62.5 FIGO Stage (2014) 0.032* 0.132  Ⅰ 30 96.6 93.2  Non-Ⅰ 18 75.0 77.8 Lymph node metastasis ＜0.01*† 0.007*†  Positive 4 50.0 50  Negative 38 97.0 94.7  Unknown 6 66.7 66.7 Cytology of ascites 0.482† 0.676†  Positive 9 74.1 77.8  Negative 27 92.4 88.7  Unknown 12 91.7 91.7 Lymph node dissection 0.029* 0.082  Yes 42 92.4 90.4  No 6 66.7 66.7 Concordant Stage 0.222 0.560  Dual Stage I (I/I) 22 95.2 90.7  Advanced / Discordant Stage 26 83.5 84.6 SEOC Subgroup 0.051 0.158  Group A 12 90.9 90.9  Group B 26 95.5 92.3  Group C 10 70 70 Group A: FIGO IA3; Group B: Advanced Concordant; Group C: Independent Discordant. OS, overall survival; PFS, progression-free survival; BMI, body mass index; FIGO, International Federation of Gynecology and Obstetrics; SEOC, synchronous endometrial and ovarian carcinoma. The 5-year OS and PFS rates were estimated using the Kaplan-Meier method, and p -values were calculated using the log-rank test. * p <0.05 indicates statistical significance. † p -values for Lymph Node Metastasis and Cytology of Ascites were calculated comparing Positive vs. Negative cases, excluding patients with unknown status Univariate analysis affecting the prognosis of patients Group A: FIGO IA3; Group B: Advanced Concordant; Group C: Independent Discordant. OS, overall survival; PFS, progression-free survival; BMI, body mass index; FIGO, International Federation of Gynecology and Obstetrics; SEOC, synchronous endometrial and ovarian carcinoma. The 5-year OS and PFS rates were estimated using the Kaplan-Meier method, and p -values were calculated using the log-rank test. * p <0.05 indicates statistical significance. † p -values for Lymph Node Metastasis and Cytology of Ascites were calculated comparing Positive vs. Negative cases, excluding patients with unknown status To characterize the unexpected treatment failures, we scrutinized the granular profiles of the single recurring instances within the low- and intermediate-risk cohorts. In Group A (Stage IA3), the single progressing case (case 27) occurred in a 47-year-old premenopausal female presenting with elevated preoperative tumor markers (CA125: 158.6 U/mL; CA199: 120.1 U/mL). Histopathological evaluation confirmed concordant well-to-moderately differentiated (Grade 1–2) dual endometrioid adenocarcinomas. The endometrial component exhibited superficial myometrial invasion (< 50%) but demonstrated notable mucosal involvement of the lower uterine segment. Intraoperatively, severe acute pelvic suppurative infection and dense inflammatory adhesions necessitated an incomplete surgical staging paradigm; pelvic lymphadenectomy was selectively omitted to mitigate the risk of infectious dissemination and major vascular injury. However, sampling of 12 palpable para-aortic lymph nodes was successfully performed and pathologically confirmed to be tumor-free. Molecularly, paired immunohistochemistry revealed a concordant mismatch repair-deficient (dMMR) phenotype characterized by the concurrent loss of MLH1 and PMS2 expression (with intact MSH2 and MSH6), alongside diffuse estrogen receptor (ER) and progesterone receptor (PR) positivity. The patient subsequently developed early disease recurrence at 16 months and succumbed to the disease at 18 months post-surgery. In Group B (Intermediate-Risk), the single fatal case (case 36) involved a 52-year-old postmenopausal female (BMI 28.6) with preoperative CA125 of 316.7 U/mL and CA199 of 1990 U/mL. Pathology indicated independent dual primary Grade 1–2 endometrioid adenocarcinomas, accompanied by squamous differentiation in the endometrium. The endometrial tumor infiltrated nearly 1/2 of the myometrium, extended to the cervical-uterine junction, and demonstrated lymph vascular space invasion (LVSI). The left ovarian tumor displayed macroscopic involvement of the contralateral ovary and uterine serosa. Peritoneal washing cytology was positive for malignant cells, and surgical cytoreduction was suboptimal. All 44 harvested regional lymph nodes were negative for metastasis. Molecular analysis demonstrated a dMMR profile in both tumor components with a loss of MLH1 and PMS2, wild-type p53 expression, and ER/PR positivity. Disease recurrence was observed at 25 months, and the patient died at 30 months post-surgery. To identify the determinants of survival, we independently evaluated the clinicopathological features of the endometrial and ovarian components (Table  3 ). None of the endometrial parameters were significantly associated with OS or PFS. Specifically, the histological type ( p  = 0.733), histological grade ( p  = 0.625), and MI (≥ 1/2 vs. < 1/2, p  = 0.544) showed no statistical correlation with prognosis. Notably, the FIGO 2023 Stage (Stage I vs. Non-Stage I) also failed to predict OS ( p  = 0.658) or PFS ( p  = 0.892). In contrast, ovarian clinicopathological features showed a strong association with survival within the overall cohort. Ovarian histological grade was strongly associated with OS (Grade 1–2: 94.5% vs. Grade 3: 62.5%, p  = 0.003) and PFS ( p  = 0.015). Furthermore, the presence of lymph node metastasis was a critical negative prognostic factor, with a 5-year OS of 50.0% compared to 97.0% in patients without metastasis ( p  < 0.01). Ovarian staging (Stage I vs. Non-Stage I) was also significantly correlated with OS ( p  = 0.032). Other clinical variables, including age ( p  = 0.291), BMI ( p  = 0.494), parity ( p  = 0.815), and preoperative CA 125 levels ( p  = 0.225), did not significantly impact survival outcomes. Although 91.7% of the cohort received adjuvant therapy, its administration was not a statistically significant independent factor for OS ( p  = 0.422) or PFS ( p  = 0.531) in this specific cohort. We attempted to perform a multivariate Cox regression analysis incorporating factors significant in the univariate analysis. However, likely due to the limited sample size and the low number of events (5 deaths) inherent to the favorable prognosis, no variable retained statistical significance as an independent prognostic factor ( p  > 0.05). Therefore, the identified univariate predictors, particularly ovarian pathology and lymph node status, should be interpreted as significant risk stratifiers rather than independent predictors. Molecular profiling of the 17 paired SEOC cases (Fig.  2 ) revealed that 64.7% (11/17) of the patients exhibited complete inter-tumor concordance for both MMR proteins and p53 expression. Specifically, in Group A (Stage IA3, n  = 4), molecular concordance was observed in the evaluable pairs, with p53 status unavailable for Case 27. Within Group B ( n  = 9), while molecular patterns were predominantly homogeneous, three cases exhibited divergence: Case 30 showed discordant p53 expression despite concordant dMMR status, while Cases 38 and 41 displayed discordant MMR profiles (EC-pMMR/OC-dMMR) with matched p53 patterns. Group C ( n  = 4) demonstrated a higher frequency of heterogeneity, including MMR discordance in Case 29 (EC-pMMR/OC-dMMR) and p53 discordance in Case 44. These preliminary observations highlight a spectrum of molecular stability across the subgroups, with more frequent divergence identified in the Independent Discordant group. Fig. 2 Correlation between molecular profiles and FIGO 2023 subgroups. Integrated heatmap illustrating mismatch repair (MMR) and p53 expression across 17 paired SEOC cases. Subgroups: Patients are categorized into Group A (Stage IA3, yellow), Group B (Advanced Concordant, green), and Group C (Independent Discordant, purple) based on FIGO 2023 staging. Molecular Status: Rows display mismatch repair (MLH1, MSH2, MSH6, PMS2) and p53 status for endometrial cancer (EC) and ovarian cancer (OC) components. Blue: pMMR/p53-Normal (wild-type). Red: dMMR/p53-Abnormal (null, mutation, overexpression). Grey: Data not available (Case 27) Correlation between molecular profiles and FIGO 2023 subgroups. Integrated heatmap illustrating mismatch repair (MMR) and p53 expression across 17 paired SEOC cases. Subgroups: Patients are categorized into Group A (Stage IA3, yellow), Group B (Advanced Concordant, green), and Group C (Independent Discordant, purple) based on FIGO 2023 staging. Molecular Status: Rows display mismatch repair (MLH1, MSH2, MSH6, PMS2) and p53 status for endometrial cancer (EC) and ovarian cancer (OC) components. Blue: pMMR/p53-Normal (wild-type). Red: dMMR/p53-Abnormal (null, mutation, overexpression). Grey: Data not available (Case 27)

Synchronous primary carcinomas of the female reproductive system are rare, with synchronous endometrial and ovarian carcinoma (SEOC) being the most common type [ 1 – 3 ]. For decades, the primary clinical challenge has been distinguishing between two independent primary tumors and a single primary with early metastasis. This distinction is critical, as it dictates adjuvant treatment strategies and prognostic expectations. Traditionally, clinicians relied on clinicopathological criteria, such as those proposed by Scully et al. [ 4 ], which, while useful, often involve subjective interpretation and lack biological precision. Although SEOC generally has a favorable prognosis (5-year overall survival of 83–85.9%) [ 5 , 6 ], its optimal management has historically been a subject of intense debate. Under older staging systems (International Federation of Gynecology and Obstetrics (FIGO) 2009), these dual tumors were frequently misclassified as advanced-stage single-organ disease. Consequently, low-risk, early-stage synchronous tumors were often subject to high-intensity adjuvant interventions, creating a compelling clinical need for a consensus-driven approach to identify candidates potential for treatment de-escalation. In 2023, the FIGO introduced a significant update to the staging of endometrial cancer, specifically incorporating Stage IA3 to classify synchronous low-grade endometrioid carcinomas involving both the endometrium and ovary. This stage is strictly defined by the disease being limited to the uterus and a unilateral ovary, necessitating the absence of substantial lymph vascular space invasion (LVSI) and any additional metastases [ 7 ]. Theoretically, Stage IA3 identifies a subgroup with slow-growing tumors where adjuvant therapy can be safely omitted in accordance with FIGO guidelines. However, the introduction of Stage IA3 raises two critical questions that remain under-investigated: First, does this anatomical classification truly reflect a homogenous molecular profile? Second, how do endometrial-specific risk factors (such as myometrial invasion) compare with ovarian-specific parameters in determining the overall survival (OS) of these patients? This retrospective study aims to evaluate the clinical and biological rationale of the updated FIGO 2023 staging framework within a cohort of 48 SEOC patients, stratified into three distinct clinical-biological groups (Stage IA3, Advanced Concordant, and Independent Discordant). We specifically focus on differentiating the prognostic weight of endometrial versus ovarian clinicopathological features. Additionally, in a subset of 17 cases with available paired molecular data, we evaluated MMR and p53 expression to explore the biological homogeneity of Stage IA3 and the evolutionary divergence in non-IA3 cohorts. By integrating clinical outcomes with molecular evidence, we seek to provide preliminary data to support refined risk stratification and personalized management of SEOC.

The 2023 FIGO staging revision introduced Stage IA3 as a landmark effort to refine risk stratification in synchronous endometrial and ovarian carcinomas (SEOC); however, its biological underpinnings and clinical applicability have remained largely unvalidated in independent cohorts. In this context, we provide a multidimensional, albeit preliminary, assessment—integrating clinicopathological parameters, long-term survival, and paired molecular profiling (MMR/p53)—to explore the validity of this new framework within a single-center retrospective setting. Our observations suggest that Stage IA3 may define a biologically homogeneous, indolent subgroup, whereas progression beyond these anatomical confines appears to be characterized by increasing genomic divergence and aggressive clinical behavior. Notably, by independently evaluating dual tumor sites, we observed that ovarian-specific parameters, particularly histological grade and lymph node status, may exert a more pronounced influence on overall survival than traditional endometrial risk factors in this specific cohort. The introduction of Stage IA3 was theoretically designed to identify low-risk SEOC patients eligible for conservative management [ 7 ]. Our molecular findings offer preliminary biological context for this anatomical definition. In the evaluable Stage IA3 cases (Group A), paired immunohistochemistry showed complete concordance for MMR proteins and p53 expression, which may indicate a shared clonal origin. This trend aligns with high-throughput genomic studies employing next-generation sequencing (NGS), which have generally shown that histologically concordant, low-grade endometrioid SEOCs frequently harbor identical somatic mutations—suggesting that the ovarian lesion often represents an early, indolent metastasis rather than a truly independent primary [ 8 – 10 ]. Although our IHC-based data cannot confirm clonality with the resolution of NGS, this observed biological uniformity provides a hypothesis-generating rationale for discussing treatment de-escalation — a strategy that could potentially reduce the historical overtreatment observed in our cohort, where 91.7% of Stage IA3 patients received platinum-based chemotherapy despite their favorable prognosis. In contrast to the general uniformity observed in Stage IA3, advanced concordant and independent discordant cohorts exhibit pronounced molecular complexity and aggressive phenotypic behavior. Within our independent discordant cohort, the identification of site-specific molecular discordances—such as divergent MMR profiles or localized p53 abnormalities restricted to a single anatomical site—underlines a complex parallel evolutionary trajectory. These findings align with Moukarzel et al. [ 10 ], who noted that clonal progression in SEOC can follow intricate pathways. Furthermore, this spatial and morphological plasticity is further supported by Iacobelli et al. [ 11 ], who demonstrated that conventional histopathological classifications can be confounded by localized clonal divergence—such as the emergence of mixed or discordant non-endometrioid phenotypes—even within an evolutionary lineage originating from a single primary clone. Such molecular discordance raises two primary possibilities: first, that these tumors may possess independent biological origins; or second, that they have undergone ‘clonal evolution,’ where metastatic lesions or synchronous sites acquire novel molecular features. This biological heterogeneity is broadly reflected in our survival trends. While Group A and Group B demonstrated excellent 5-year OS rates (90.9% and 95.5%, respectively), Group C (Independent Discordant) showed a markedly poorer prognosis (70.0%, p = 0.051), reinforcing the clinical relevance of histological discordance as a potential high-risk feature [ 12 – 15 ]. Extending beyond molecular complexity, our univariate analysis revealed a notable phenotypic hierarchy: ovarian histological grade (G3 vs. G1-2, 62.5% vs. 94.5%, p = 0.003) and lymph node metastasis (50.0% vs. 97.0%, p < 0.01) were strongly associated with survival, whereas endometrial parameters—including myometrial invasion depth and LVSI—did not reach statistical significance in this limited cohort. This pattern is broadly consistent with the retrospective findings of Caldarella et al. [ 16 ] and raises the hypothesis that the ovarian microenvironment may act as a driver for clonal acceleration. Once the tumor breaches the anatomical confines of Stage IA3, the biological behavior of the ovarian component may become a pronounced risk factor propelling systemic dissemination and therapy resistance, potentially owing to the higher peritoneal metastatic potential of ovarian deposits. However, given the small number of events and the exploratory nature of our subgroup analyses, these findings should be interpreted as hypothesis-generating rather than confirmatory. While our data do not contradict the overall safety of the Stage IA3 criteria, the exceptional failures observed in advanced concordant and independent discordant cohorts highlight critical molecular vulnerabilities that pure anatomical staging may overlook. Notably, the only two fatal events within Groups A and B occurred in patients with a mismatch repair-deficient (dMMR) phenotype, highly suggestive of Lynch syndrome [ 17 ]. Despite favorable anatomical staging, these dMMR tumors exhibited aggressive kinetics [ 10 , 18 ]. In the Stage IA3 case, severe pelvic suppurative inflammation necessitated the omission of pelvic lymphadenectomy—potentially masking occult nodal metastasis—a vulnerability exacerbated by concurrent lower uterine segment (LUS) involvement, an area rich in lymphatic networks that may facilitate early occult dissemination [ 19 ]. In the other case, widespread macroscopic serosal dissemination rapidly supervened. These cautionary observations align with the emerging consensus that dMMR-driven genomic instability may confer a unique tumor microenvironment prone to rapid, cross-boundary progression [ 20 ]. Therefore, while anatomical staging provides a useful framework for initial risk stratification, we suggest that routine integration of MMR and p53 immunohistochemistry could serve as an additional biological safety check [ 21 ]. For Stage IA3 patients with dMMR tumors, treatment de-escalation should be approached with particular caution, and comprehensive surgical staging — including adequate lymphadenectomy—remains essential to rule out occult extrauterine disease. This study offers a multidimensional, albeit preliminary, evaluation of the FIGO 2023 Stage IA3 criteria by integrating clinicopathological data with paired molecular profiling from both tumor sites. However, several significant limitations warrant emphasis. The retrospective, single-center design and relatively small cohort ( n  = 48), coupled with the limited number of survival events (5 deaths), precluded informative multivariable analysis; consequently, our findings should be interpreted as preliminary risk stratifiers rather than independent predictors. Furthermore, we acknowledge an inherent methodological circularity: several clinicopathological differences among the three groups are expected by design, as they comprise the defining criteria for our cohort stratification. To mitigate this, we anchored our evaluation primarily on survival trends and paired molecular markers, which were independent of the grouping definitions. The small molecular sub-cohort ( n  = 17) and the reliance on IHC-based analysis—which lacks the genomic resolution of NGS—further limit the robustness of our molecular conclusions. Future multi-center studies with larger sample sizes and comprehensive NGS panels are urgently needed to validate these preliminary observations. Our findings support the clinical utility of the FIGO 2023 Stage IA3 criteria in identifying a low-risk subset of SEOC patients who may benefit from treatment de-escalation. Integrating p53 and MMR immunohistochemistry into routine evaluation may offer an additional biological safety check, potentially helping clinicians identify potentially aggressive variants that anatomical staging alone might overlook. Future studies should prioritize larger, multi-center cohorts to validate the prognostic impact of molecular heterogeneity in SEOC. Additionally, incorporating NGS will be essential to definitively map clonal evolution and refine the molecular boundaries of the current staging system, particularly for cases exhibiting discordant phenotypic patterns.

In conclusion, our preliminary findings offer initial clinical and biological evidence that the FIGO 2023 Stage IA3 criteria may help delineate a low-risk subset of SEOC patients with favorable outcomes and paired molecular concordance. This observation provides a hypothesis-generating rationale for future multi-center trials evaluating treatment de-escalation, although such strategies remain investigational pending validation in larger cohorts. However, as disease extends beyond Stage IA3, SEOC exhibits increasing genomic divergence and aggressive phenotypic behavior. Within the limitations of this single-center, small-sample study, ovarian-specific parameters — particularly histological grade and lymph node status — appeared to exert a more pronounced influence on prognosis than traditional endometrial risk factors. We suggest that routine integration of MMR and p53 immunohistochemistry may offer an additional biological safeguard, especially for patients with dMMR tumors or histological discordance, to better distinguish low-risk clonal SEOC from potentially aggressive independent malignancies. Nevertheless, these recommendations are exploratory and require prospective validation in adequately powered, molecularly annotated multi-center cohorts before clinical adoption.