Correlation of clinicopathological and prognostic characteristics between endometriosis-associated and primary ovarian cancer

Background The main aim of this study was to establish the clinicopathological and prognostic correlations between endometriosis-associated and non-endometriosis-associated primary ovarian cancer, with a view to provid- ing a reference guide for revision of diagnostic criteria for malignant transformation of endometriosis.

Clinicopathological and follow-up data of 174 patients with clear cell and endometrial ovarian cancer were retrospectively extracted. Cases were divided into endometriosis-associated and non-endometriosis-associated primary ovarian cancer for comparative analysis of clinicopathological characteristics and prognosis.

Average age and post-menopausal rate in the endometriosis-associated ovarian cancer group were lower relative to the primary ovarian cancer group (P < 0.05). Body mass index, age at menopause, operation history, dysmenorrhea, complications, tumor size, tumor side, ascites, CA125, HE4, CA19.9, stage, differentiation, expression of ER, PR, P53, P16, Ki67, MMR, HNF-1β and Napsin A were not significantly different between the groups (P > 0.05). Furthermore, rates of resistance to platinum chemotherapy, relapse, progression-free survival and overall survival were comparable between the two groups (P > 0.05).

Endometriosis-associated and primary ovarian cancers of the same pathological type are speculated to be homologous in terms of origin from malignant transformation of endometriosis. It may therefore be necessary to revise the diagnostic criteria for ovarian endometriosis malignancy.

Endometriosis-associated ovarian cancer, Ovarian clear cell carcinoma, Ovarian endometrioid carcinoma, Diagnostic criteria Open Access © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecom- mons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. BMC Cancer †Huimin Wang and Cong Chen are co-first authors. *Correspondence: Danbo Wang [email protected] Yanmei Zhu [email protected] Full list of author information is available at the end of the article Page 2 of 11Wang et al. BMC Cancer (2023) 23:1210

Endometriosis is a common disease in women of child - bearing age, with an incidence of 15–20% [1]. Stud - ies to date have shown that endometriosis has the same molecular biological and genetic background as ovarian cancer and presents a high risk factor for ovarian cancer development [2]. According to the current internation - ally recognized Sampson and Scott diagnostic criteria: (1) there must be a clear example of endometriosis in association with or close proximity to the cancer, (2) no other primary tumor site must exist and the histology of the tumor should be consistent with an endometrial ori - gin, (3) endometriosis associated with cancers must show morphologic progression from benign to malignant in a contiguous fashion [3, 4], with a malignant transforma - tion rate of 0.5–1.0% [5]. However, due to the significant heterogeneity of endometriosis and atypical hyperplasia, high rates of missed pathological diagnosis are inevita - ble and the actual incidence of malignant transformation of endometriosis may be higher [6, 7], which presents a challenge in establishing accurate diagnostic criteria. The major pathological types of endometriosis-associ - ated ovarian cancer are endometrioid and clear cell can - cer, accounting for 75–90% cases [8]. However, according to the diagnostic criteria of Sampson and Scott, only 50–70% ovarian clear cell and endometrioid carcinoma cases are diagnosed as endometriosis-associated ovarian cancer (EAOC). Based on the theory of external origin of ovarian cancer, clear cell and endometrioid carcinoma types are closely related to endometriosis, potentially resulting from endometriosis-associated malignancy [9]. Our research is focused on the malignant transforma - tion of endometriosis. Previous studies have reported comparable abnormal expression of EAOC-related genes, such as mismatch repair gene human mutL homolog 1 (hLMH1) and runt-related transcription factor3 (RUNX3), between EAOC and non-endometriosis-asso - ciated primary ovarian cancer (non-EAOC) groups [10, 11], giving rise to the speculation that all clear cell and endometrioid carcinomas of the ovary originate from malignant transformation of endometriosis. Since the pathological changes of endometriosis are heterogene - ous, particularly after malignant transformation, tumor tissues grow vigorously and destroy the original tissues, and the histological basis of endometriosis thus remains unclear. Pathological sampling has a number of limita - tions. For instance, cancer and endometriosis lesions are not obtained at the same time and concurrent endome - triosis is often overlooked, resulting in a low diagnostic rate of malignant transformation of endometriosis using the standard criteria of Sampson and Scott. Further rele - vant clinical and basic research is thus warranted to vali - date this theory.

Research objects In total, 174 cases of ovarian clear cell carcinoma and ovarian endometrioid carcinoma were diagnosed and treated in China Medical University Cancer Hospital and Affiliated Shengjing Hospital from January 2008 to November 2018. Participants were divided into EAOC (74 patients) and non-EAOC (100 patients) groups according to the standards of Sampson and Scott, as follows: (1) coexistence of cancer and endometriosis in the same ovary, (2) a similar histological pattern, (3) exclusion of secondary tumors metastatic to the ovary, and (4) histopathological evidence demonstrating the transition from benign endometriosis to malignancy. Patients that met the above criteria were classified as EAOC and the remaining patients as non-EAOC. According to the different pathological types, endome - triosis-associated ovarian clear cell carcinoma (EOCC) and endometriosis-associated ovarian endometrioid carcinoma (EOEC) groups were further defined. Non- endometrium-related primary ovarian cancer was further grouped into clear cell carcinoma (OCC) and endometrioid carcinoma (OEC) groups. Our patient population included 39 EOCC, 35 EOECC, 35 OCC and 65 OECC cases. Samples were fully encoded to protect patient confidentiality. Inclusion criteria were as follows: (1) definitive patho - logical diagnosis, limited to clear cell carcinoma and endometrioid carcinoma, (2) complete case data, and (3) completion of the initial treatment plan (surgery and chemotherapy). Exclusion criteria were as follows: (1) the presence of other histological types of ovarian malignan - cies and borderline tumors of non-clear cell carcinoma and endometrioid carcinoma, (2) other primary malig - nant tumors, and (3) metastatic ovarian cancer. Sam - ples of the selected cases were re-assessed by the same gynecological pathologist to further confirm diagnosis. Data collection Clinicopathological data of all patients were collected. Clinical parameters included age, body mass index (BMI), dysmenorrhea or not, menopausal status, age of menopause, endometriosis history, operation or not (for instance, cesarean section, hysterectomy, endometriosis), complications (for instance, diabetes, heart disease, high blood pressure), related serum tumor markers (CA125, CA19.9, HE4), type of surgery (complete, suboptimal, and optimal), and endometriosis lesions during surgery. Pathological parameters included tumor size, tumor side, ascites, FIGO stage, histological classification, and immunohistochemical results (ER, PR, P53, P16, Ki67, MMR, HNF-1, Napsin A). Page 3 of 11 Wang et al. BMC Cancer (2023) 23:1210 Treatments All the included patients received standard surgi - cal treatment. Patients in the early stage (FIGO I-II) received comprehensive staging surgery while those in the advanced stage (FIGO III-IV) received tumor cell reduction surgery. Postoperative chemotherapy regi - mens containing platinum were adopted according to international guidelines as follows: paclitaxel and car - boplatin (TC regimen), docetaxel and carboplatin (DC regimen), 6–8 cycles of chemotherapy, with an interval of 21 days. Follow‑up All cases were followed up until the end of recurrence, death or the end of follow-up (up to 31/3/2019). The follow-up time was 6-132 months, with median of 67 months. Cases requiring reoperation to obtain histo - pathological and/or imaging evidence of new develop - ments and/or continued abnormal elevation of tumor markers were considered tumor recurrence. Statistical analysis Data were analyzed using SPSS 25.0 software. Enu - meration data are expressed in terms of rates and measurement data as mean ± standard deviation (± sd). Student’s T test was used for comparison between groups. The enumeration data were compared using the Chi-square test and Fisher’s exact test. Survival analysis was performed using Kaplan-Meier curves and differ - ences were assessed with the log-rank test. Differences were considered significant at P < 0.05.

General data analysis Analysis of epidemiological data revealed a lower aver - age age and proportion of menopause of patients in the EAOC than non-EAOC group (P 0.05), as shown in Table 1 . Further intra-group and inter-group analyses of the same pathological types showed a significantly lower average age of onset and proportion of menopause in the EOCC than OCC group (P 0.05; Table 2). Analysis of clinical features No significant differences were evident between the two groups in terms of tumor size, tumor side, ascites, type of surgery, CA125, HE4 and CA19.9 levels (all P > 0.05). The collective data are presented in Table 3 . Further intra-group and inter-group analyses dis - closed significantly higher incidence of bilateral tumors in the EOEC than EOCC group (40.0% vs. 17.9%, P < 0.05). Moreover, the HE4 level in the EOEC group was higher than that in the EOCC group to a significant extent (355.4 vs. 89.4 pmol/L, P < 0.05). No significant differences were found in the remaining parameters, including tumor size, tumor side, ascites and type of surgery (Table 4). Pathological characteristics Pathological characteristics were comparable between the two groups in terms of FIGO stage, differentiation degree, ER, PR, P53, P16, Ki67, MMR, HNF-1, and Nap - sin A-positive expression (all P > 0.05; Table 5; Fig. 1). The rates of ER and PR positivity in the EOEC group were significantly higher than those in the EOCC group (71.4% vs. 17.9%, 60.0% vs. 7.7%; P < 0.05). Conversely, the rates of HNF-1 and Napsin A positivity in the EOCC group were markedly higher relative to the EOEC group (28.2% vs. 20.0%, 30.8% vs. 7.7%; P < 0.05). HNF-1- and Napsin A-positive rates in the OCC group were significantly higher than those in the OEC group (74.2% vs. 16.9%, 80.0% vs. 4.6%; P < 0.05), as shown in Table 6. Table 1 Comparative analysis of the epidemiological data of EAOC and non-EAOC patients (n, x ± s) Characteristics EAOC Non‑EAOC P 74 100 Age(year) 49.4 ± 7.7 54.0 ± 10.0 0.001 BMI 23.6 ± 3.3 23.5 ± 3.7 0.731 History of dysmenorrhea 0.064 Yes 29 26 No 55 74 Menopausal status 0.001 Yes 32 68 No 42 32 Age of menopause 49.1 ± 4.3 49.3 ± 3.9 0.815 Endometriosis history 0.980 Yes 11 15 No 63 85 History of surgery 0.115 Yes 19 16 Caesarean section 11 6 Hysterectomy 4 7 Endometriosis 4 3 No 65 84 Complications 0.683 Yes 31 45 No 43 55 Page 4 of 11Wang et al. BMC Cancer (2023) 23:1210 Prognostic characteristics Overall, 74 patients in the EAOC group were subjected to initial analyses, 64 of whom were followed up (follow- up rate of 86.5%). Rates of platinum therapy resistance, recurrence and mortality were determined as 6.25%, 39.1%, and 28.1%, respectively. Among the 100 patients in the non-EAOC group, 85 were followed up (85.0%). The platinum resistance rate was determined as 9.41%, recur - rence rate as 54.1%, and mortality rate as 40.0%, which were not significantly different between the two groups (all P > 0.05; Table 7). Kaplan-Meier analysis and log-rank test showed that the average overall survival (OS) of the EAOC group was 91.6 months (95% CI: 76.9–106.5 months) while that of the non-EAOC group was 77.8 months (95% CI: 66.1– 90.0 months), with no significant differences (P = 0.068, > 0.05). The median progression-free survival (PFS) of the EAOC group was 78.4 months (95% CI: 62.2–94.5 months), which was not significantly different from the non-EAOC group (64.0 months; 95% CI: 50.7–77.1 months) (P = 0.216, > 0.05), as shown in Fig. 2. Intra-group and inter-group analyses of the same pathological types via Kaplan-Meier and log-rank tests Table 2 Intra-group and inter-group analyses of EAOC and non-EAOC patients (n, x ± s) a :EOEC vs. EOCC, b:EOEC vs. OEC, c:EOCC vs. OCC Characteristics EOEC EOCC OEC OCC P 35 39 65 35 Age(year) 50.6 ± 8.4 48.2 ± 6.7 54.3 ± 10.7 53.3 ± 8.6 0.177a/0.060b/<0.01c BMI 23.9 ± 2.9 23.3 ± 3.6 23.7 ± 3.8 22.8 ± 3.1 0.430a/0.782b/0.570c History of dysmenorrhea 0.076a/0.667b/0.075c Yes 10 19 16 10 No 25 20 49 25 Postmenopausal status 0.178a/0.199b/<0.01c Yes 18 14 42 26 No 17 25 23 9 Age of menopause 48.8 ± 4.7 49.2 ± 3.8 49.9 ± 4.1 48.2 ± 3.2 0.802a/0.408b/0.363c Endometriosis history 0.894a/0.883b/0.894c Yes 5 6 10 5 No 30 33 55 30 History of surgery 0.599a/0.106b/0.810c Yes 8 11 7 9 Caesarean section 3 8 2 4 0.149a/0.340b/0.290c Hysterectomy 3 1 3 4 0.339a/0.420b/0.183c Endometriosis 2 2 2 1 0.911a/0.610b/0.475c No 27 28 58 26 Complications 0.528a/0.916b/0.892c Yes 16 15 29 14 No 19 24 36 21 Table 3 Comparative analysis of clinical features of EAOC and non-EAOC patients (n, x ± s) R0: no residual lesions; R1: residual lesions 1 cm Characteristics EAOC Non‑EAOC P 74 100 Ascites 0.666 Yes 39 56 No 35 44 Tumor size 0.769 ≥ 10 cm 45 63 <10 cm 29 37 Tumor side 0.355 Unilateral 53 65 Bilateral 21 35 Type of surgery 0.804 R0 66 87 R1 9 10 R2 2 3 Biomaker CA125(U/ml) 448.9 ± 980.0 739.8 ± 1141.4 0.080 HE4(pmol/L) 203.4 ± 240.0 264.5 ± 217.5 0.620 CA19.9(U/ml) 194.0 ± 460.4 274.2 ± 546.7 0.465 Page 5 of 11 Wang et al. BMC Cancer (2023) 23:1210 showed that OS of the EOCC group was 90.8 months (95% CI, 69.9–111.8 months), EOEC group was 96.8 months (95% CI, 76.8–116.8 months), OCC group was 77.4 months (95% CI, 51.49–103.30 months), and OEC group was 81.10 months (95% CI, 68.0–94.2 months). We observed no significant differences in OS in EOCC vs. EOEC, EOCC vs. OCC, and EOEC vs. OEC groups (P = 0.290, 0.262, 0.070, all P > 0.05). PFS of patients in the EOCC group was 80.9 months (95% CI, 61.2–100.6 months), EOEC group was 82.3 months (95% CI, 59.4– 105.2 months), OCC group was 85.9 months (95% CI, 68.4–108.3 months), and OEC group was 60.0 months (95% CI, 45.50–74.6 months). We observed no significant differences in PFS in EOCC vs. EOEC, EOCC vs. OCC, and EOEC vs. OEC groups (P = 0.222, 0.675, 0.071, all P > 0.05; Fig. 3).

Several recent studies have highlighted a significantly increased risk of ovarian cancer in endometriosis patients [12]. As a precancerous lesion, endometriosis is closely related to ovarian clear cell and endometrioid car- cinomas. The theory of external origin of ovarian cancer hypothesizes that both clear cell and endometrioid car - cinoma of the ovary originate from malignant transfor - mation of endometriosis, which poses a challenge to the current diagnostic criteria for endometriosis-associated malignant transformation. However, further clinical and basic research evidence is needed to substantiate this theory. In the current study, the two ovarian cancer types (clear cell and endometrioid carcinomas) most closely associated with endometriosis were examined as a whole. According to the diagnostic criteria of Sampson and Table 4 Intra-group and inter-group clinical features analyses of EAOC and non-EAOC patients (n, x ± s) a :EOEC vs. EOCC, b:EOEC vs. OEC, c:EOCC vs. OCC. R0: no residual lesions; R1: residual lesions 1 cm Characteristics EOEC EOCC OEC OCC P 35 39 65 35 Ascites 0.469a/0.295b/0.209c Yes 20 19 44 12 No 15 20 21 23 Tumor size 0.892a/0.881b/0.363c ≥ 10 cm 21 24 38 25 <10 cm 14 15 27 10 Tumor side 0.036a/0.303b/0.316c Unilateral 21 32 33 32 Bilateral 14 7 32 3 Type of surgery 0.894a/0.883b/0.590c R0 30 33 55 32 R1 3 6 8 2 R2 2 0 2 1 Biomaker CA125(U/ml) 618.0 ± 1309.8 297.0 ± 509.6 870.3 ± 895.0 493.7 ± 1479.0 0.161a/0.255b/0.437c HE4(pmol/L) 355.4 ± 319.5 89.4 ± 69.1 498.9 ± 547.0 132.7 ± 259.0 0.039a/0.575b/0.651c CA19.9(U/ml) 248.1 ± 525.4 160.6 ± 423.0 276.9 ± 586.0 220.5 ± 547.0 0.556a/0.877b/0.663c Table 5 Comparative analysis of pathological characteristics of EAOC and non-EAOC patients(n, %) Characteristics EAOC Non‑EAOC P 74 100 FIGO stage 0.417 I/II 51 63 III/IV 23 37 Differentiation 0.932 Low 18 21 Middle 34 52 High 22 27 Biomarker ER(+) 32 52 0.253 PR(+) 24 45 0.094 P53(+) 59 68 0.085 P16(+) 41 54 0.854 MMR(-) 19 37 0.114 Ki67(%) 30.0 ± 18.0 30.8 ± 18.94 0.767 Napsin A(+) 18 37 0.075 HNF-1β(+) 15 31 0.113 Page 6 of 11Wang et al. BMC Cancer (2023) 23:1210 Fig. 1 IHC expression of related molecules in ovarian tissues (SP 200X). a ER-positive; b ER-negative; c PR-positive; d PR-negative; e P53-positive; f P53-negative; g P16-positive; h P16-negative; i. Ki67-positive; j MLH1-positive; k MLH1-negative; l. MSH2-positive; m MSH2-negative; n MSH6-positive; o MSH6-negative; p PMS2-positive; q PMS2-negative; r HNF-1β-positive; s HNF-1β-negative; t Napsin A-positive; u Napsin A-negative Page 7 of 11 Wang et al. BMC Cancer (2023) 23:1210 Scott, samples were divided into EAOC and non-EAOC groups for comparison of clinicopathologic features and prognosis. Our results showed no significant differences between the groups, supporting the theory that both non-endometriosis-associated primary ovarian endome - trial carcinoma and ovarian clear cell carcinoma poten - tially have the same origin as EAOC from endometriosis. We retrospectively analyzed the clinical features of 174 patients from EAOC and non-EAOC groups. The average age and proportion of menopausal patients in Table 6 Intra-group and inter-group pathological characteristics of EAOC and non-EAOC patients(n, %) a :EOEC vs. EOCC, b:EOEC vs. OEC, c:EOCC vs. OCC Characteristics EOEC EOCC OEC OCC P 35 39 65 35 FIGO stage I/II 21 30 37 26 0.116a/0.767b/0.792c III/IV 14 9 28 9 Differentiation Low 11 7 18 3 0.177a/0.695b/0.239c Middle 17 17 40 12 High 7 15 7 20 Biomarker ER(+) 25 7 49 3 < 0.01a/0.667b/0.311c PR(+) 21 3 40 5 < 0.01a/0.880b/0.635c P53(+) 29 30 43 25 0.526a/0.076b/0.589c P16(+) 16 25 32 22 0.112a/0.710b/0.912c MMR(-) 8 11 22 15 0.072a/0.548b/0.291c Ki67(%) 29.1 ± 20 30.7 ± 14 29.6 ± 19 33.0 ± 16.5 0.718a/0.907b/0.564c Napsin A(+) 7 11 11 26 < 0.01a/0.459b/0.230c HNF-1β(+) 3 12 3 28 < 0.01a/0.317b/0.273c Table 7 Prognosis comparison between EAOC and non-EAOC patients(n) Characteristics EAOC non‑EAOC P 74 100 Platinum resistance 4 8 0.483 Relapse 25 39 0.405 Death 19 34 0.193 Fig. 2 Differences in OS and PFS between EAOC and non-EAOC patients. a OS; b PFS Page 8 of 11Wang et al. BMC Cancer (2023) 23:1210 the EAOC group was lower relative to the non-EAOC group (P < 0.05). These differences may be attributed to a potential decrease in postmenopausal hormone levels, gradual atrophic degradation of ectopic endometrium [13], limitations of pathological materials, and errors due to insufficient pathological evidence of endometriosis. Further experiments on larger sample sizes are required to confirm these findings. No significant differences were evident in BMI, dysmenorrhea history, endome - triosis history, operation history, complications, ascites, tumor size, type of surgery and relevant serum tumor markers (CA125, HE4, CA19.9) between the two groups Fig. 3 Differences in OS and PFS of EOCC vs. EOEC, EOCC vs. OCC and EOEC vs. OEC groups. a OS-EOCC vs. EOEC; b OS- EOCC vs. OCC; c OS- EOEC vs. OEC; d PFS- EOCC vs. EOEC; e PFS- EOCC vs. OCC; f PFS- EOEC vs. OEC Page 9 of 11 Wang et al. BMC Cancer (2023) 23:1210 (all P > 0.05), consistent with earlier results. Our experi - ments support similar characteristics of endometrio - sis- and non-endometriosis-associated primary ovarian cancers. Furthermore, upon stratified analysis according to pathological type into EOCC and EOEC, OCC and OEC groups, relevant epidemiological and clinical char - acteristics, such as BMI, dysmenorrhea history, endome - triosis history, history of surgery, complications, ascites, tumor size, type of surgery and serum tumor marker (CA125, HE4, CA19.9) levels, were not significantly dif - ferent (P > 0.05). Our findings present further evidence that primary ovarian endometrial carcinoma and clear cell carcinoma are associated with corresponding patho - logical types in endometriosis-associated ovarian cancer with similar clinical features. These results provide pre - liminary confirmation that the two groups have compa - rable epidemiological and clinical characteristics and the age difference is potentially attributed to strict diagnostic criteria. Histopathological evaluation is currently the gold standard of ovarian cancer diagnosis and classification. Immunohistochemical analysis revealed high expression of ER and PR in endometrioid carcinoma of ovary and low expression in clear cell carcinoma [14]. Compared with other epithelial ovarian cancer types, increased HNF-1β, high expression of Napsin A and positivity for P53 were characteristic features of ovarian clear cell car - cinoma [15]. MMR expression in both ovarian endome - trial and clear cell carcinoma types is abnormally high relative to that in other subtypes of ovarian cancer [16]. The above molecules may therefore serve as useful bio - markers to distinguish epithelial ovarian cancer subtypes. Ki-67 is currently used as a positive nuclear proliferation marker and its expression reflects the biological behavior of tumor cells [17]. For analysis of differences, similari - ties and potential mechanisms of EAOC and non-EAOC, we compared the pathologies of the two groups. Our data showed no histological differences in terms of pathologic differentiation degree, ER, PR, P53, and P16, Ki67, MMR, HNF-1β and Napsin A-positive expression (P > 0.05) between EAOC and non-EAOC. The majority of previ - ous reports have investigated the malignant transforma - tion of endometriosis, with limited studies focusing on the characteristics of different pathological types of endo- metrial carcinoma and clear cell carcinoma. However, a number of differences in clinicopathological characteris - tics exist between endometrioid and clear cell carcinoma types. Further grouping analysis revealed no significant differences in expression of relevant non-specific indi - cators, such as P53, P16, Ki67 and MMR, among the groups. The positive expression rates of ER and PR were markedly different between endometrioid carcinoma and clear cell carcinoma groups, but not from the primary ovarian carcinoma of the same pathologic type. Differen - tiation of endometriosis into two tissue types is reported to occur in a dual mode regulated by sex hormones [18, 19]: (1) estrogen and progesterone receptor-positive endometriosis lesions undergo malignant transformation to form hormone-dependent endometrioid carcinoma after long-term stimulation without antagonistic estro - gen and (2) atrophic ectopic endometrial lesions negative for estrogen and progesterone receptors are stimulated by oxidative stress for a long period of time, resulting in malignant transformation and formation of non-hor - mone-dependent clear cell carcinoma. Our results were consistent with earlier literature, validating that clear cell carcinoma is a non-hormone-dependent tumor. The positive expression rates for HNF-1β and Napsin A were significantly higher in clear cell carcinoma than endome - trioid carcinoma. Accordingly, we concluded that differ - ent histological types undergo distinct mechanisms for development of endometriosis-associated ovarian can - cer, which will be the focus of follow-up investigations. The above results further confirm similar expression of pathology-related molecular indicators in EAOC and non-EAOC with the same pathological subtype. Moreo - ver, the differences in molecular expression indicated by the results of stratified analysis based on pathological subtype may be related to the pathogenesis of the differ - ent types. The majority of previous studies focused on comparing clinicopathological and prognostic differences between intrauterine endometriosis-associated ovarian cancers of a single pathologic type. However, sample sizes in earlier reports were usually small and the results obtained were inconsistent. Some studies suggest that ovarian endome - trial carcinoma and clear cell carcinoma are early phases with better prognosis than the two pathological types of primary ovarian cancer [20, 21], while other researchers report no obvious differences [22, 23]. Here, we exam - ined 174 cases of ovarian endometrial carcinoma and clear cell carcinoma, with a median follow-up period of 67 months. Our data showed no significant differences between platinum resistance, recurrence and mortal - ity rates between the EAOC and non-EAOC groups (P > 0.05). PFS and OS were comparable between the two groups (P > 0.05), supporting similar prognosis. In 2020, Hermens et al. [24] analyzed 32,419 patients with ovarian cancer and found synchronous endometriosis of ovarian cancer staging of early, higher progression-free survival and overall survival is longer, perhaps because of more frequent hospital visits of patients with endome - triosis due to simultaneous treatment with the appropri - ate drugs and long-term state of endometriosis-induced inflammation, which activates immune function, in turn, facilitating early detection and better prognosis, giving Page 10 of 11Wang et al. BMC Cancer (2023) 23:1210 rise to the theory that ovarian cancer with endometriosis may have different pathophysiological features relative to other ovarian cancer types. Earlier studies have estab - lished a low early incidence of epithelial ovarian cancer. In this research, the early incidence rates of EAOC and non-EAOC were 68.9% and 63.0%, respectively, which were significantly higher than the average early inci - dence of ovarian cancer (30.0%). Overall survival rates of the EAOC and non-EAOC groups were 71.9% and 60.0%, which were markedly higher than the average five-year survival rate of ovarian epithelial carcinoma (44- 50%) [25]. Considerable evidence suggests that ovarian clear cell carcinoma and endometrioid carci - noma are sources of malignant transformation of endo - metriosis and can therefore be detected earlier. However, accelerated growth of tumor tissue destroys the tissue of origin owing to malignant transformation. Pathological sampling cannot remove cancer and ectopic foci lesions at the same time. Consequently, a proportion of clear cell and endometrioid carcinomas of the ovary are not diag - nosed as endometriosis-associated ovarian cancer. Our

indicate similar rates of recurrence and survival in EAOC and non-EAOC groups. In summary, this study found no significant differ - ences in epidemiological or pathological features and comparable prognosis between endometriosis- and non-endometriosis-associated ovarian cancer diag - nosed according to the Sampson and Scott criteria, leading to the speculation that primary ovarian endo - metrial and clear cell carcinomas have the same origin as EAOC. Our findings support the theory of dualism of ovarian cancer suggesting that both endometrial carcinoma and clear cell carcinoma originate from progression of endometriosis. Based on our findings, we propose that the Sampson and Scott diagnostic criteria for endometriosis malignancy are too strict and the actual incidence of malignant transformation of endometriosis may be higher. Therefore, it is worth exploring whether the stringent diagnostic criteria for malignant transformation of endometriosis require fur - ther revision. In view of the collective data, the diag - nostic criteria for endometriosis malignancy could include the following guidelines: (1) no other primary tumor site must exist, (2) the histology of the tumor should be consistent with an endometrial origin, and exclude: there must be a clear example of endometrio - sis in association with or close proximity to the cancer and endometriosis associated with cancers must show morphologic progression from benign to malignant in a contiguous manner. The retrospective nature of this study is a major limi - tation. Moreover, the results represent single-center data and the number of included cases is relatively low. Further large-scale, prospective multicenter clinical and molecular biology studies are therefore required to validate the associations of malignant transformation of endometriosis with ovarian clear cell and endome - trial carcinomas, which should aid in clarification of the underlying biological mechanisms and development of individualized treatments for patients with endometri - osis-associated ovarian cancer. Abbreviations EAOC Endometriosis-associated ovarian cancer non-EAOC Non-endometriosis-associated primary ovarian cancer EOCC Endometriosis associated ovarian clear cell carcinoma EOEC Endometriosis associated ovarian endometrioid carcinoma OCC Clear cell carcinoma OEC Endometrioid carcinoma BMI Gody Mass Index OS Overall survival PFS Progression-free survival

Not applicable. Authors’ contributions HW and CC carried out most parts of the experiment; YZ participated in the experiment; DW participated in the design of the study; HW, CC and PC performed the statistical analysis. All authors read and approved the final manuscript. Funding This work was supported by grants from The National Natural Science Founda- tion of China (Grant Number 81771556) and the Natural Fund of Liaoning Province (2020-ZLLH-36). Availability of data and materials The datasets generated and/or analyzed in the current study are not publicly available due to the data also forming part of an ongoing study, but can be made available by the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Samples were fully encoded to protect patient confidentiality. All methods were carried out in accordance with the relevant guidelines and regulations. Due to the retrospective nature of this study and preserved anonymity of patients, a waiver of informed consent was obtained from the Research Ethics committees of Liaoning Cancer Hospital & Institute. The study and associated protocols were approved by the Research Ethics committees of Liaoning Cancer Hospital & Institute (2020G0322). Consent for publication Not applicable. Competing interests The authors declare no competing interests. Author details 1 Department of Gynecology, Liaoning Cancer Hospital & Institute, 44 Xiao- heyan Road, Dadong District, Shenyang, Liaoning 110004, China. 2 Department of Pathology, Liaoning Cancer Hospital & Institute, Shenyang, Liaoning, China. Received: 23 April 2023 Accepted: 14 November 2023 Page 11 of 11 Wang et al. BMC Cancer (2023) 23:1210 • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit y our researc hReady to submit y our researc h ? Choose BMC and benefit fr om: ? Choose BMC and benefit fr om:

1. Giudice LC. Endometriosis. N Engl J Med. 2010;362:2389–98. 2. Erzen M, Rakar S, Klancnik B, Syrjänen K. Endometriosis-associated ovarian carcinoma (EAOC): an entity distinct from other ovarian carcinomas as suggested by a nested case-control study. Gynecol Oncol. 2001;83:100–8. 3. Sampson J. Endometrial carcinoma of the ovary arising in endo metrial tissue in that organ. Arch Surg. 1925;10:1–72. 4. Scott R. Malignant change in endometriosis. Obstet Gynecol. 1953;2:293–9. 5. Swiersz LM. Role of endometriosis in cancer and tumor development. Ann N Y Acad Sci. 2002;955:281–92. 6. Gadducci A, Lanfredini N, Tana R. Novel insights on the malignant trans- formation of endometriosis into ovarian carcinoma. Gynecol Endocrinol. 2014;30:612–7. 7. Vercellini P , Viganò P , Buggio L, Makieva S, Scarfone G, Cribiù FM, et al. Perimenopausal management of ovarian endometriosis and associated cancer risk: when is medical or surgical treatment indicated?Best practice and research. Clin Obstet Gynaecol. 2018;51:151–68. 8. Bas-Esteve E, Pérez-Arguedas M, Guarda-Muratori GA, Acién M, Acién P . Endometriosis and ovarian cancer:their association and relationship. Eur J Obstet Gynecol Reprod Biol X. 2019;3:100053. 9. Mallen A, Soong TR, Townsend MK, Wenham RM, Crum CP , Tworoger SS. Surgical prevention strategies in ovarian cancer. Gynecol Oncol. 2018;151:166–75. 10. Ren F, Wang D, Jiang Y, Ren F. Epigenetic inactivation of hMLH1 in the malignant transformation of ovarian endometriosis. Arch Gynecol Obstet. 2012;285:215–21. 11. Guo C, Ren F, Wang D, Li Y, Liu K, Liu S, et al. RUNX3 is inactivated by promoter hypermethylation in malignant transformation of ovarian endometriosis. Oncol Rep. 2014;32:2580–8. 12. Samartzis EP , Labidi-Galy SI, Moschetta M, Uccello M, Kalaitzopoulos DR, Perez-Fidalgo JA, et al. Endometriosis-associated ovarian carcinomas: insights into pathogenesis, diagnostics, and therapeutic targets-a narra- tive review. Ann Transl Med. 2020;8:1712. 13. Wang S, Qiu L, Lang JH, Shen K, Yang JX, et al. Clinical analysis of ovarian epithelial carcinoma with coexisting pelvic endometriosis. Am J Obstet Gynecol. 2013;208:411–5. 14. Chen S, Dai X, Gao Y, Shen F, Ding J, Chen Q. The positivity of estrogen receptor and progesterone receptor may not be associated with Metasta- sis and recurrence in epithelial ovarian cancer. Sci Rep. 2017;7:16922. 15. Ip PPC, Wang SY, Wong OGW, Chow KL, Lee HHY, Cheung ANY, et al. Napsin A, Hepatocyte nuclear factor-1-beta (HNF-1β), estrogen and pro- gesterone receptors expression in Arias-Stella reaction. Am J Surg Pathol. 2019;2019(43):325–33. 16. Bennett Ja, Morales-Oyarvide V, Campbell S, Longacre TA, Oliva E. Mismatch repair protein expression in clear cell carcinoma of the ovary: incidence and morphologic associations in 109 cases. Am J Surg Pathol. 2016;40:656–63. 17. Munjishvili V, Barabadze E, Musashvili T, Gachechiladze M, Burkadze G. Morphophenotypic characteristics of ovarian serous borderline tumors. Georgian Med News. 2019;290:20–5. 18. Suzuki F, Akahira JI, Akahira I, Miura T, Suzuki K, Ito K, Hayashi SI, et al. Loss of estrogen receptor beta isoform expression and its correlation with aberrant DNA methylation of the 5’-untranslated region in human epithelial ovarian carcinoma. Cancer Sci. 2008;99:2365–72. 19. Tanase Y, Yamada Y, Shigetomi H, Kajihara H, Oonogi A, Yoshizawa Y, et al. Modulation of estrogenic action in clear cell carcinoma of the ovary. Exp Ther Med. 2012;3:18–24. 20. Ren T, Wang S, Sun J, Qu JM, Xiang Y, Shen K, et al. Endometriosis is the independent prognostic factor for survival in Chinese patients with epithelial ovarian carcinoma. J Ovarian Res. 2017;10:67. 21. Park JY, Kim DY, Suh DS, Kim JH, Kim YM, Kim YT, et al. Significance of ovarian endometriosis on the prognosis of ovarian clear cell carcinoma. Int J Gynecol Cancer. 2018;28:11–8. 22. Li Q, Sun Y, Zhang X, Wang L, Wu W, Meng C, et al. Endometriosis-asso- ciated ovarian cancer is a single entity with distinct clinicopathological characteristics. Cancer Biol Ther. 2019;20:1029–34. 23. Zhao T, Shao Y, Liu Y, Wang X, Guan L, Lu Y. Endometriosis does not con- fer improved prognosis in ovarian clear cell carcinoma: a retrospective study at a single institute. J Ovarian Res. 2018;11:53. 24. Hermens M, van Altena AM, van der Aa M, Bulten J, van Vliet HAAM, Sie- bers AG, et al. Ovarian cancer prognosis in women with endometriosis: a retrospective nationwide cohort study of 32,419 women. Am J Obstet Gynecol. 2021;224:284.e1-284.e10. 25. Moufarrij S, Dandapani M, Arthofer E, Gomez S, Srivastava A, Lopez- Acevedo M, et al. Epigenetic therapy for ovarian cancer: promise and progress. Clin Epigenetics. 2019;11:7. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.