In this study, weused a nationwide registry to evaluate the outcomes of early-stage EOC patients who underwent FSS and RCS. Patients in the FSS group were younger and mostly had stage I disease. The RCS group included more cases of stage II and high grade (grade 3) disease, and more frequent adjuvant chemotherapy. The most common histologic type was mucinous carcinoma in the FSS group, compared to endometrioid and clear cell carcinomas in the RCS group. Stage was a risk factor for poor outcome for mucinous and clear cell histologies, but not for serous or endometrioid histology. Patients with grade 3 endometrioid ovarian cancer had a poorer prognosis compared to patients with grade 1/2 tumors. Among patients with early-stage clear cell carcinoma, CSS was non-inferior and even better after FSS compared to RCS, yet adjuvant chemotherapy was necessary.We also foundthat22 of the 401 women who underwent FSS developed a second malignancy later in life.
Several previous retrospective studies have compared the oncologic outcomes of early-stage EOC patients who undergo FSS or RCS, and have reported comparable results10–17. The majority of these prior studies have only enrolled patients with stage I disease, and have found that FSS is adequate treatment for stage I EOC, without compromising survival outcomes [10, 12–15]. Ditto et al.11 and Bogani et al.16analyzed the outcomes of FSS in patients with stage I disease, and small numbers of patients with stage II and III disease. Theyreportedthat FSS did not influence the progression-free survival (PFS) compared to complete staging surgery, among women with high-risk ovarian cancer, with FIGO stage IA/IB grade 3 or stage IC/II diseases11. Bogani et al. also reported that the type of surgery did not affect the disease-free survival (DFS) or overall survival (OS) of patients with grade 3 tumors or stage IC/II diseases after over 10 years of follow-up16. Similar to these past studies, in our present series,we found that FSS can be a safe procedure for patients with early-stage EOC who had a desire for fertility preservation (FSS vs. RCS, HR of CSS 1.09, p = 0.73, Table 2).
We found that disease stage was an independent risk factor affecting outcome in patients with early-stage EOC. Our previous results showedworse 5-year overall survival (OS) instage II disease compared to stage IA/IB disease4. Ditto et al. also reported poorer OS in stage IC/II diseases than stage IA/IB diseases11.In the present study, we found that stage IC and II disease were poor prognostic factors for CSS, withHR values of 2.40 and 3.60 relative to stage IA/IB diseases. We further demonstrated that stage was an independent risk factor for poor CSS in cases of mucinous and clear cell histologies, but not cases of the serous or endometrioid type. Kajiyama et al. also reported that stage IC disease was associated with poorer OS than stage IA/IB disease in cases of the mucinous carcinoma and clear cell carcinoma histologies19,20. This indicates that the capsule status during operation significantly influences the outcomes, especially in mucinous and clear cell carcinomas19,20. Therefore, we recommend that surgeons should remove ovarian tumors as carefully as possible to avoid intraoperative tumor rupture, especially for patients undergoing FSS.
Tumor grade was another risk factor for poor outcome in early-stage EOC patients. In our study, grade 3 tumors were associated with worse CSS than grade 1/2 tumors, especially in mucinous and endometrioid types (Table 3). Chen et al. also found that grade 3 tumors were a poor prognostic factor compared to grade 1/2 in cases of ovarian endometrioid carcinoma21. Among cases of stage I endometroid carcinoma, Chao at el. also found that grade 3 tumors were an independent poor prognostic factor in terms of PFS22. However, it remains controversial whether grade correlates with unfavorable survival outcomes in mucinous carcinomas23. Moreover, the grading system for ovarian mucinous carcinomasis globally inconsistent23.Busca et al. compared two widely used grading systems: the International Federation of Gynecology and Obstetrics (FIGO) system24, which was the same grading system used for endometrioid carcinoma, and the Silverberg grading system25. They found that only the Silverberg grading system appeared to correlate with outcome in cases of mucinous carcinoma23.
There is long-standing debate regarding the safety of using FSS to treat patients with early-stage ovarian cancer, especially clear cell carcinoma. Clear cell histology was considered a contraindication for FSS due to its relativelyworse outcome compared to other histologic types26.However, many recent retrospective studies have compared the outcomes of FSS and RCS for stage I clear cell carcinoma, and the results reveal non-inferior survival outcomes following FSS compared to RCS, although with limited patient numbers20,27−30.Nasioudis et al. used the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) database to evaluate cases of stage I clear cell carcinoma treated with uterus- or ovary-preserving staging surgery31. Their study included a total of 741 patients with ovarian clear cell carcinoma, including 96 with uterus preservation. The five-year cancer-specific survival rates did not significantly differ between patients with or without uterus preservation (90.8% vs. 87.7%, p = 0.29)31.Moreover, uterine preservation was not associated with worse survival even after controlling for the disease sub-stage31. However, in their review article, Satoh and Yoshikawa reported a higher cumulative relapse rate for patients with stage IC clear cell carcinoma (22.6%) compared to patients with stage IA clear cell carcinoma (11.1%)32.Thus, they did not recommend FSS for stage IC clear cell carcinoma32, and this indication remains controversial. Our present study included a large series obtained from the nationwide database, and our results showed that patients who underwent FSS for early-stage clear cell carcinoma had survival outcomes similar to those of patients treated without fertility preservation. Within our nationwide database, the patients of reproductive age who underwent FSS were significantly younger than the patients who underwent RCS; however, our results showed no effect of age. Notably, among clear cell carcinoma patients, the FSS group even showed a significantly better survival outcome than the RCS group (HR 0.28, 95% CI: 0.06–0.82, p = 0.040).Further subgroupanalysisrevealedthat the FSS and RCS groups had similarly good CSS rates in stage IA/IB (Fig. 3A, p = 0.19) and stage II (Fig. 3C, p = 0.06) clear cell carcinomas.In contrast, among stage IC clear cell carcinomas, the FSS group had better CSS than the RCS group (Fig. 3B, p = 0.006). A likely explanation is that the FSS patients may have been selected by surgeons during surgery, rather than randomly selected. Surgeons may choose the most suitable and low-risk patients to undergo FSS, such as patients with an intact capsule and with minimal pelvic adhesion, to avoid potential spreading of cancer during surgery.
Patients who undergo FSS might develop second malignancies in the uterine corpus or the contralateral ovary.In our study, ovarian endometrioid carcinoma was the histology most likely to develop second malignancies. Of the eight second malignancies in cases of ovarian endometrioid cancer, seven were uterine cancer. These cases illustrated the risk of uterus preservation with this histologic type. It is difficult to distinguish whether the second malignancy was a metastatic endometrioid endometrial cancer from the ovary or a synchronous cancer. Zhao et al. reported thatpatients with stage I ovarian endometrioid carcinoma had a 19.3% rate of synchronous early stage and well-to-moderate differentiated endometrial carcinoma22. Notably, in our study, uterine cancer also developed in patients with ovarian cancerof the other histologies. Three patients developed ovarian cancer in the contralateral ovary. During surgery, biopsy of the contralateral ovary was usually not recommended in the absence of gross abnormalities, based on the low risk of microscopic involvement in a contralateral ovary with normal appearance13, and concerns about infertility caused by postoperative adhesions on the remaining ovary33,34. However, patients should be informed that there is a risk of recurrence in the preserved contralateral ovary. Bentivegna et al. reported an 11.6% recurrence rate following FSS in patients with stage I–II disease. Among these recurrences, 38% were isolated in the spared ovary, and 62% occurred at an extraovarian site, which was associated with a worse survival outcome35. Based on these findings, we suggest regular postoperative surveillance of the remaining ovary and of the endometrium using sonography or computerized tomography. Additionally, we recommend theperformance of endometrial biopsy before or during FSS to identify synchronous endometrial and ovarian carcinomas, particularly in patients with ovarian endometrioid carcinoma with clinical symptoms, such as abnormal vaginal bleeding.
The present study had several strengths. The first strength was that it was a nationwide population-based study performed using the TCR database, which includes over 90% of cancer patients in Taiwan18. The database is periodically subjected to field data audits,and is thus a high-quality and reliable data source18. Additionally, this study has a large sample size, providing sufficient statistical power. Due to ethical problems, it is almost impossible to perform a prospective study comparing the outcomes of FSS and RCS12; therefore,we think that our study providesimportantnew insights on this issue. Taiwan exhibits a higher incidence of clear cell carcinoma than many other countries36.Thus, this study was able to includea large number of ovarian clear cell carcinoma patients, enabling comparison of outcomes after FSS or RCS specifically among ovarian clear cell carcinoma patients. We used cancer-specific survival (CSS) as the main outcome measure, instead of overall survival (OS), because the proportion of death from other causes is relatively high in patients with early-stage cancers37.Therefore, CSS more precisely reflects the survival outcome related to early-stage ovarian cancer.
One limitation of this study was its retrospective nature. There may have been some unavoidable difficulties in the coding and grouping of the patients, such as an inaccurate diagnosis or cancer staging, and loss of patients to follow-up, which may lead to systemic bias. Additionally, we lacked details of the pathologic findings, information regarding tumor recurrence, chemotherapy regimen and dosage, and the subsequent pregnancy records due to the nature of the TCR system.