Background Patients with recurrent clear cell ovarian cancer have poor prognosis and limited effective systemic treatment options.
Objectives To characterize patterns of recurrence and compare overall survival and post-recurrence survival parameters in patients with recurrent ovarian clear cell carcinoma.
Methods Clinical data on patients with ovarian clear cell carcinoma between June 1995 and August 2014 were collected. Patients with clear cell ovarian cancer recurrence were included in this study. Patients with different histologic sub-type, persistent or progressive disease on completion of the initial treatment were excluded. Descriptive statistics, univariate and multivariable analyses, and Kaplan-Meier survival probability estimates were completed. The log-rank test was used to quantify survival differences on univariable analysis. To search for significant covariates related to the overall survival and post-recurrence survival, a univariable Cox proportional hazard model was performed.
Results A total of 209 patients met inclusion criteria. Of these, 61 (29%) patients who were free of disease at completion of the initial treatment had recurrence. Patterns of recurrence were as follows: 38 (62%) patients had multiple-site recurrence, 12 (20%) had single-site recurrence, and 11 (18%) had nodal recurrence only. The median overall survival was 44.7 months (95% CI 33.4 to 64.2) and was significantly associated with pattern of recurrence (p=0.005). The median post-recurrence survival was 18.4 months (95% CI 12.5 to 26.7): 54.4 months (95% CI 11 to 125.5) in single-site recurrence, 13.7 months (95% CI 6.8 to 16.5) in multiple-site recurrence, and 30.1 (95% CI 7.2 to 89) months in nodal recurrence (p=0.0002). In the multivariable analysis, pattern of recurrence was a predictor of post-recurrence survival.
Six patients (9.8%) had a prolonged disease-free interval after recurrence (disease-free for more than 30 months after completion of treatment for recurrence). Prolonged recurrences were noted in 4 (33%) of 12 patients with single-site recurrence, 1 (9%) of 11 patients with nodal recurrence, and in 1 (2.7%) of 38 patients with multiple-site recurrence. Three of the six patients with a prolonged disease-free interval after recurrence were treated surgically at the time of recurrence.
Conclusion Ovarian clear cell carcinoma predominantly recurs in multiple sites and it is associated with a high mortality rate and short post-recurrence survival. When recurrences are limited to a single site, or only to lymph nodes, the median post-recurrence survival is longer. Disease-free interval after recurrence is longer in patients with single-site recurrence who are treated surgically at the time of recurrence.
- clear cell
Statistics from Altmetric.com
Disease-free interval after recurrence is longer in patients with single-site recurrences who are treated surgically at the time of recurrence.
Ovarian clear cell carcinoma most commonly recurs at multiple sites.
Single-site recurrence and nodal recurrence only have the most favorable post-recurrence and overall survival.
Ovarian clear cell carcinoma is a distinct histologic sub-type of epithelial ovarian cancer, representing 4–12% of cases in North America.1–4 The 5-year overall survival rate for stage I patients is 85%.1 5–8 However, patients with advanced stage disease have worse overall survival than patients with high-grade serous ovarian-cancer.2 7 9–13 This difference is attributed to the fact that patients with ovarian clear cell carcinoma are often resistant to chemotherapy, with response rates of 22–56% to platinum-based regimens.14 15
Adjuvant treatment for patients with early stage ovarian clear cell carcinoma has not been standardized. The role of adjuvant radiation in a sub-set of patients with early-stage disease is still controversial.16 17 The response rate to chemotherapy is lower in the recurrent setting and reported as <10% in patients with platinum-sensitive ovarian clear cell carcinoma, and 1% in patients with platinum-resistant disease.18 19 Generally, patients with recurrent clear cell ovarian cancer are known to have poor prognosis. However, this is a heterogeneous group, and some sub-groups might have more favorable outcomes. Characterizing patterns of recurrence in patients with ovarian clear cell carcinoma enables us to better understand disease behavior and survival outcome in different sub-groups. This information may change the way in which patients are treated and challenge the current paradigm of poor outcome in patients with recurrent ovarian clear cell carcinoma.
In our study, we aimed to describe the patterns of recurrence in patients with ovarian clear cell carcinoma. We examined survival outcomes based on patterns of recurrence and evaluated the effect of different post-recurrence treatment modalities on survival.
An institutional review board approved this retrospective analysis of the Cancer Registry database at two tertiary cancer centers in Toronto, Canada: Princess Margaret Cancer Center and Odette Cancer Center. A total of 209 patients with ovarian clear cell carcinoma were treated between June 1995 and August 2014 and comprised the cohort for this study. Comprehensive review of the electronic medical records was performed. Only patients with pure ovarian clear cell carcinoma with a documented recurrence were included. Recurrence was diagnosed based on either imaging only or imaging and biopsy. All cases were reviewed by an expert gynecologic pathologist.
After initial surgery, treatment options included adjuvant chemotherapy, radiation, a combination of both, or observation. At the time of initial diagnosis, chemotherapy was predominantly platinum-based doublet, with carboplatin (area under curve=5–6) and paclitaxel (135–175 mg/m2) every 3 weeks, for six cycles. In this manuscript, adjuvant chemotherapy refers to chemotherapy in the initial setting (before the recurrence). The radiation treatment at initial diagnosis was in keeping with standard protocols.17 Patients were followed up every 3–4 months for the first 2 years and every 6 months thereafter. At 5 years, patients were either discharged back to primary care or continued follow-up with the oncology team on a yearly interval.
For all patients with epithelial ovarian cancer, we use a local follow-up practice, which includes documentation of new symptoms and a physical examination. Serum CA125 was not routinely drawn, and routine imaging was not mandatory. Patients underwent CT imaging of chest, abdomen, and pelvis, if they exhibited symptoms concerning for recurrence, elevated serum CA125, or an abnormal physical examination as documented by the physician or reported by the patients. Histopathologic confirmation of recurrence was optional before initiation of treatment for cases of multiple-sites recurrence. Patients with one site of recurrence underwent a CT- or ultrasound-guided biopsy, unless they were candidates for secondary debulking surgery. Treatment options at time of recurrence included surgery, radiation, systemic treatment (chemotherapy with or without targeted therapy in the setting of clinical trials), or a combination of these modalities.
In total, there were 61 recurrences and an additional 17 patients who had persistent or progressive disease on completion of the initial treatment. The main objective of our study was to analyze the pattern of recurrence and the effect of treatment for the recurrence, hence, only patients who had no evidence of disease at the end of their initial treatment were included in the cohort for the analysis. We divided our study population into three groups: single-site recurrence, multiple-site recurrence, and nodal recurrence. Single-site recurrence included all patients with single site, non-nodal recurrence in the abdomen and pelvis. Multiple-sites recurrence included all patients with more than one disease site, including lymph nodes. Patients with lymph node recurrences only, were included in the nodal recurrence group. This included lymph nodes in any region, local and/or distant, single or multiple nodes. Lymph node metastases were considered present when patients had a CT scan showing a lymph node >1 cm in short-axis diameter that progressively increased on follow-up CT, or was biopsied and was consistent with disease. Post-recurrence survival was defined as the time from date of recurrence to death or to the last follow-up. We chose to use post-recurrence survival rather than progression-free survival, since we wanted to evaluate prognosis after recurrence. Overall survival was defined as time elapsed from the date of initial diagnosis to the date of last follow-up, or death (in months).
To compare demographics and clinical factors, descriptive analysis using Wilcoxon rank-sum or Fisher exact test was applied for continuous or categorical variables.
Survival estimates were plotted using the Kaplan-Meier method. The log-rank test was used to quantify survival differences on univariable analysis. To search for significant covariates related to the overall survival and post-recurrence survival, a univariable Cox proportional hazard model was performed. HR and 95% confidence intervals (CIs) of the HR were also calculated for each covariate. Cox proportional hazard models were used for multivariable regression analyses. A p value <0.05 was considered statistically significant.
All analyses were conducted using Statistical Analysis Software (SAS version 9.4 for Windows).
A total of 209 patients with ovarian clear cell carcinoma were identified in the dataset. The majority, 165 (79%) patients had early stage (stage I–II) at diagnosis, and 44 (21%) patients had advanced stage (stage III–IV) disease. Among the patients with early-stage disease, 43 (26%) cases recurred. Twenty-seven patients with advanced disease had a complete response to primary treatment. By the time of the analysis, disease had recurred in 18 of them (recurrence rate of 67%). Initial surgical outcomes in these 18 patients whose disease later recurred was: no visible residual disease in 14 patients, <1 cm residual disease in 2 patients, and sub-optimal (>1 cm residual disease) in 2 patients.
Of the 43 of 165 patients with early-stage disease who recurred, the initial adjuvant treatment (before recurrence) following the surgery was chemotherapy, radiation, combination, or observation. For the patients with advanced disease, the treatments included chemotherapy, chemotherapy and radiation, or observation (Table 1). Median follow-up for the 61 patients who recurred was 41.2 months (range 3.6–181.9). At the time of analysis, 46 patients had died and 15 patients were still alive. Of the 15 patients who survived, one was lost to follow-up and two were discharged to a hospice. The post-recurrence mortality was 80.7%. Median post-recurrence survival was 18.4 months (95% CI 12.52 to 26.74) and 5-year post-recurrence survival was 17% (95% CI 9% to 33%). The median overall survival was 44.7 months (95% CI 33.4 to 64.2), with 38.4% 5-year overall survival.
Most patients with initial stage I disease, (22/35, 62.9%) recurred in multiple sites. With a total of 5 of 35 (14.3%) patients recurring at a single site, and 8 of 35 (22.9%) patients recurring only in the lymph nodes. A total of 4 of 8 (50%) patients with initial stage II disease recurred at a single site and 4 of 8 (50%) patients at multiple sites. Most patients (11/15) with stage III disease recurred in multiple sites, one at a single site and three patients recurred in lymph nodes alone. Two of three (66%) patients with stage IV had recurred at a single site; however, they were considered stage IV owing to pleural effusion without proven cytology. One stage IV patient (33%) recurred in multiple sites.
The majority of patients had multiple-sites recurrence (38/61, 62%), including multiple extra-pelvic recurrences (18), pelvic and extra-pelvic recurrences (15), intrathoracic recurrences (4), brain and leptomeningeal spread (1). Twelve (20%) patients had a single-site recurrence, all limited to the pelvis, including one vaginal cuff and one ascending colon recurrence. Eleven patients (18%) had nodal recurrence only: five patients in a single lymph node basin (four in para-aortic and one in groin lymph node) and six in multiple lymph-node sites: pelvic, para-aortic, supra-diaphragmatic, groin). The median post-recurrence survival was 54.4 versus 30.1 versus 13.7 months in single-site, nodal-, and multiple-sites recurrence, respectively. The median time to first recurrence was 24.5 versus 30.5 versus 16.9 months in single-site, nodal-, and multiple-sites recurrence, respectively. The 5-year post-recurrence survival was 37.4%, 23.8% and 8.5% in single-site recurrence, nodal recurrence, and multiple-sites recurrence, respectively (p=0.0002) (Figure 1). The median overall survival was 94.4 versus 35.5 months versus not reached (all % >50%) in single-site, multiple-sites, and nodal recurrence, respectively. The 5-year overall survival was 74%, 22.4% and 61% in single-site recurrence, multiple-sites recurrence, and nodal recurrence; respectively (p=0.003) (Figure 2). There was a significant association between overall survival and pattern of recurrence (p=0.0056). Patients with multiple-sites recurrence had a higher risk of death than those with nodal recurrence (HR=3.89, p=0.01) or with single-site recurrence (HR=2.68, p=0.015). No significant difference in overall survival was found between single-site recurrence and nodal recurrence (HR=1.45, p=0.55). In the multiple-sites recurrence, 2 patients had surgical resection for the recurrence, 3 patients chose to have observation only, and 33 patients had systemic treatment. In the single-site recurrence, 9 patients had surgery, with or without systemic treatment or radiation, 2 patients had systemic treatment only, and 1 had radiation only. In the nodal recurrence, 1 patient had surgery, 5 patients had systemic treatment, 2 patients had radiation, and 3 patients chose observation.
In univariable analysis for post-recurrence survival, including age, sites of recurrence, initial stage (early vs advanced), initial surgical outcome, adjuvant chemotherapy, and post-recurrence treatment (surgical vs non-surgical), there was a significant difference in post-recurrence survival among sites of recurrence. Patients in multiple-sites recurrence had worse post-recurrence survival than those with nodal recurrence (HR=3.56, p=0.014) and single-site recurrence (HR=4.12, p=0.001). No significant difference was found between single-site recurrence and nodal recurrence in post-recurrence survival. Patients with initial advanced stage (HR=1.91, p=0.04), without post-recurrence surgical treatment (HR=2.36, p=0.05), or with adjuvant chemotherapy (HR=2.87, p=0.005), were more likely to have shorter post-recurrence survival, based on the univariable analysis. However, those factors were not significant in multivariable analysis. Pattern of recurrence was the only factor that was significantly associated with post-recurrence survival (HR=4.7; 95% CI 1.5 to 14.7; p=0.007; and HR=2.8; 95% CI 0.98 to 7.9; p=0.055, for multiple versus single recurrence and multiple versus nodal recurrence, respectively) (Table 2). At the time of the analysis, 13 patients were alive. Table 3 shows the descriptive factors for these patients.
Six (9.8%) patients had a prolonged disease-free interval after recurrence (no evidence of disease >30 months after recurrence). One of the patients was disease free for over 10 years and died of a stroke. In descriptive and univariable analysis, patients with single-site recurrence have higher probabilities of a prolonged disease-free interval after recurrence than those with multiple-sites recurrence and nodal recurrence (odds ratio=7; 95% CI 1.2 to 47.4; p=0.0351). Multivariable analysis differed, given the limited number of survivors (n=6). A summary of the entire cohort is presented in online supplementary figure S1 (supplementary appendix). We searched for an association between sites of recurrence and predictive factors. Initial stage, (early vs advanced), initial surgical outcome, lymph node dissection, adjuvant chemotherapy, and adjuvant radiation were not found to be significantly associated with specific site of recurrence pattern in descriptive and univariate analysis (Online supplementary table S1) (supplementary appendix). Cytology was performed in 19 patients: 4 patients had negative and 15 patients had positive cytology. Two of the four patients with negative cytology recurred in multiple sites; however, both were not staged and one of them had surface involvement. Eleven of fifteen (73%) patients with positive cytology recurred in multiple sites, 3 of 15 (20%) in lymph nodes, and 1 of 15 (6.7%) in a single site.
Our study suggests that patients with ovarian clear cell carcinoma with a single-site recurrence or nodal recurrence have favorable prognosis. Poor survival for recurrent ovarian clear cell carcinoma has been described previously. Kajiyama et al12 reported a 13% 5-year post-recurrence survival. Jenison et al20 compared clinical features of 41 ovarian clear cell carcinoma of patients with serous ovarian cancer. A total of 25 of 41 patients recurred: 72% had diffuse abdominal spread, 40% lymph node involvement, and 40% distant organ involvement The frequency of lymph nodes and involvement of distant organs in the recurrent setting was found to be significantly higher in the group with ovarian clear cell carcinoma than in those with serous ovarian cancer. Our data confirm these findings. Other reports by Hoskins et al16 and Macrie et al21 provide limited data on recurrence distribution, but again, describe a mix of diffuse disease and more localized disease at recurrence.
In our study, multiple-sites recurrence was found to have worse overall survival and post-recurrence survival. Initial stage was not found to be significant for post-recurrence survival in multivariate analysis. There was no significant difference in overall survival or post-recurrence survival between single-site recurrence and nodal recurrence. Nevertheless, while post-recurrence survival was similar for single-site recurrence and nodal recurrence, the prolonged disease-free interval after recurrence (defined as no evidence of disease >30 months post-recurrence) rate was better for single-site recurrence than for nodal recurrence. A possible explanation for this finding is that patients with nodal recurrence have a more indolent course, and while the disease is not completely eradicated, it is usually slowly growing, hence, those patients have a prolonged overall survival when they are alive with disease.
Nodal-only recurrences are uncommon in patients with epithelial ovarian cancer. This pattern of recurrence, was previously described by Blanchard et al,22 who reported that among 640 patients, 27 experienced isolated lymph node recurrence (4.2%). Median survival after relapse was 26 months. A subset of patients who declined treatment for recurrence achieved overall survival of 91 months. The authors suggested that isolated lymph node recurrence might have a better prognosis in epithelial ovarian cancer. We found an 18% rate of nodal recurrence only, which is higher than the 4.2% reported in serous ovarian cancer recurrence.22 23 Similar to the group with epithelial ovarian cancer, we found favorable survival parameters in patients with ovarian clear cell carcinoma with this pattern of recurrence, in comparison with the entire cohort. This finding supports the idea that variability in tumor biology might contribute to the type of recurrence and affect survival. The small sub-group of patients with single-site recurrence who were treated surgically were more likely to have prolonged disease-free post-recurrence survival. Given the small numbers in this group, there is no statistical test to support the descriptive findings. Bias in favor of the surgical group having better performance status, localized and small volume disease at time of recurrence, and favorable disease biology cannot be excluded.
Our proposal based on the results of this study is the following: if the only site of recurrence is the pelvis, surgical resection should be considered; radiation constitutes a reasonable alternative in cases with considerable surgical morbidity. If recurrence is limited to lymph nodes, sites of the involved lymph node(s) should be documented: if more than one site of lymph nodes is involved, surgery, systemic treatment, or radiation should be considered. In the case of a persistent single lymph node involvement, surgical treatment, radiation or a combined approach should be contemplated. If the single lymph node recurrence is of small volume and the patient is asymptomatic, observation and re-imaging in a timely manner is also a reasonable approach. For patients with multiple-sites recurrence, systemic treatment should be considered. The current paradigm of patients with ovarian clear cell recurrence having a poor prognosis with limited treatment options is challenged by our findings.
We provided evidence that this is a diverse group of patients, with different clinical behavior and survival profiles. Some sub-groups achieve long-term survival and as such, should be offered additional treatment options, include aggressive local therapies. Most patients with recurrent ovarian clear cell carcinoma recur in a multifocal pattern, and have poor prognosis. However, 20% of the patients recur in the pelvis only, and have significantly better survival.
Contributors Study conception and design LH and MQB. Acquisition of data SN-M and LH. Analysis and interpretation of data LH, MQB, DV, and GKL. All authors provided critical feedback and helped shape the research, analysis, and manuscript. LH took the lead in writing the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Sunnybrook Health science center 380-2014 University Health Network 14-8346.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.