Recently, the incidences of ovarian carcinoma have been increasing (1). The prognoses of ovarian carcinoma have not been improved in spite of development for anti-cancer treatment, particularly in advanced stages (2). Several predictive factors for prognoses in ovarian carcinomas have been identified: FIGO stage, residual tumor diameter, and histological subtypes, and so on (3,4). Among all histological subtypes, clear cell carcinoma (CCC) has been recognized as a subtype showing worse prognoses (3). Additionally, CCC is a distinct subtype with lower response and short response duration even in responder against chemotherapy (3,5,6).

Endometriosis is well-known as precursor of CCC (7,8), and there was a report suggesting that genetic back ground in CCC derived from endometriosis was different from those without endometriosis (7). Some reports have shown negative correlation of endometriosis with prognoses in CCC (9–11), and other investigations have shown positive association with better prognoses (12–15). So the clinical impact of the complication with endometriosis in upon prognosis of CCC has not been determined.

Recently, it has been pointed out that post-progression survival (PPS), defined as duration from the date of recurrence to the date of death, was relatively longer in ovarian cancers (16). Herein, we investigated the relationship with complication with endometriosis, and clinicopathological factors in CCC, including evaluation of PPS.

In this study, a total of 105 cases were enrolled. The median follow-up time was 60 months (range, 2–276 months). Of all cases, 45 cases (42.8%) were classified into EM-group, and 60 cases (57.1%) were in non-EM group (Table I). The clinicopathological factors of EM and non-EM patients were shown in Table I. EM-group patients were diagnosed at younger ages (P=0.03), and at earlier stages (P<0.01), compared with non-EM cases. Other factors such as residual tumor and adjuvant therapy regimens were not statistically different between two groups. There were no significant differences in PFS (P=0.36, Fig. 1A) and OS (P=0.79, Fig. 1B) between two groups by univariate analysis. Using multivariate analyses for PFS, FIGO stage was identified as the only independent prognostic factor. The complication with endometriosis was selected as the only independent prognostic factor for OS in multivariate analyses (Table II).

During the study period, 48 cases developed recurrence: 18 (40%) in 45 EM-group cases and 30 (50%) in 60 non-EM cases. The clinicopathological features of recurrent cases were shown in Table III. No cases received radiation therapy and most of the cases received chemotherapy. Some cases received secondary debulking surgery for recurrent tumors. There were no significant differences of clinicopathological factors including age, FIGO stage, debulking status at primary debulking surgery, and platinum-sensitivity between two groups. PPS of recurrent EM-group cases was significantly worse than that of recurrent non-EM cases (P<0.01, Fig. 2A). Median PPS was 6 months in recurrent EM-group, 18 months in recurrent non-EM cases, respectively. PPS of the patients that received therapeutic modality for recurrent tumors were shown in Figure 2B. PPS of EM-group cases was also significantly worse than that of non-EM cases (p<0.01, Fig. 2B).

Multivariate analysis for PPS revealed that complication with endometriosis was an independent worse prognostic factor (Hazard ratio, 4.57; 95% Confidence Interval, 1.93–10.97; P<0.01), in addition to platinum-sensitivity (Table IV).

In the present study, the cases with EM were diagnosed at younger age and at earlier stages. Nevertheless, the complication with endometriosis was not a predictive factor of prognosis including PFS and OS. Until now, several studies had evaluated clinical significance of the correlation with endometriosis in prognoses of CCC, however, the conclusions have not been determined (7,11,14,16). According to recent meta-analysis conducted by Kim et al, the complication with endometriosis was not related with prognosis in all subtypes of ovarian cancers (7).

PPS of clear cell carcinoma was considered to be shorter than that of serous carcinoma (17). This reason was caused by not only lower response rate, regardless of platinum-sensitivity status, but also short response duration in cases with recurrent CCC (3,5,9). In general, PPS of ovarian carcinomas including all histological subtypes was >20 months (18). However, PPS of CCC was much shorter, especially in the cases with endometriosis. A report suggested that PPS was more highly associated than PFS with OS in the first-line chemotherapy for advanced epithelial ovarian cancer (18). Short PPS in CCC could have led to shorter OS in CCC cases, especially in EM-group patients. Although genetic difference in the tissue specimens was not evaluated in the present study, a report suggested that some patients carried driver mutations in ARID1A, PIK3CA, KRAS genes in benign deep infiltrating endometriosis (19). It is speculated that tumor samples of EM-group had potentially aggressive phenotype compared with non-EM group tumors.

Although there was no difference in PFS between EM-group and non-EM group, PPS was significantly shorter in EM-group patients. Primary therapy using anti-cancer agents might alter genetic profile in CCC tumors, and cause refractory phenotype after recurrence. A report suggested CCC patients with Met gene amplification showed chemoresistant phenotype and worse prognosis (19). Thus, the alteration of genetic profiles affected by primary therapy might influence PPS in the patients with CCC. Additionally, phosphatidylinositol-3-kinase (PI3K)/Akt signaling pathway, and the receptor tyrosine kinase (RTK)/Ras signaling pathway were identified as prognostic biomarkers for CCC tumors using whole-genome sequencing (21). Novel chemotherapeutic agents inhibiting these pathways might improve survival of the CCC patients.

The limitation of this study included a retrospective study and a small number of the patients enrolled in a single-institutional analysis. Further prospective investigation is needed to confirm the impact of complication with endometriosis upon PPS of CCC patients.

In conclusion, the complication with endometriosis was the independent poor prognostic factor for PPS in CCC. Longer PPS could have led to longer OS in EM-group CCC. Further prospective investigation is necessary to confirm the significance of complication with endometriosis on PPS in CCC and develop the new therapy for recurrent CCC with endometriosis should also be considered.