| Objective This study was performed to clarify the role of secondary cytoreduction in patients with recurrent ovarian cancer and to establish scoring systems intending to evaluate the survival risk and the resectability for the patients undergoing secondary cytoreduction.Methods The patients with recurrent ovarian cancer with a complete clinical remission of more than 6 months who underwent SCR between 1986 and 2007 were entered in this retrospective study. The clinical data were reviewed including the general information, information of initial diagnosis and primary treatment, characteristics at the diagnosis of disease relapse, information of SCR and the salvage chemotherapy, and the survival data. Disease-specific survival was estimated by the Kaplan-Meier method, and the difference was determined by the Log-Rank method. Cox proportion hazards regression was used for multivariate analysis, which was the basis for the prognostic scoring system. Univariate and multivariate logistic regression were used to estimate the influence of clinical and pathologic factors on the resectability of SCR. Multivariate logistic regression was the basis for the surgery feasibility scoring system. We validated the models and tested the scoring systems on the original database.The English publications about SCR were collected from the online database. The articles meeting study inclusion criteria were reviewed and the data were extracted including the general information of articles, clinical and pathological characteristics of patients underwent SCR, the information of SCR and survival. The variance analysis (One-way ANOVA) was used to identify the difference of mean weighted survival between optimal and suboptimal cytoreductive groups. Univariate and multivariate linear regression analysis were used to evaluate the correlationship between the post-recurrence survival and the clinical pathological variables.Results 290 patients met the inclusion criteria of current study, with a median age of 51 years at recurrence. During the SCR, 64.8% patients rendered residual disease less than 1cm, and 24.1% patients rendered macroscopic disease free. The median overall survival for the whole cohort was 21.65 months. The MOS of patients with no macroscopic disease(R0), residual disease 0.1-1cm (R1), and residual disease >1cm(R2) were 44.12 months, 23.01 months and 12.0 months, respectively. And the differences for stratified comparisons were all significant (R1 vs. R0: P < 0.0001, R2 vs. R0: P< 0.0001, R2 vs. R1: P < 0.0001).Cox proportion hazards regression model revealed that the post-recurrent survival was significantly influenced by the residual disease (R2 vs. R0: HR = 2.84, P = 0.001), the specialization of gynecologic oncologist (the technology developed less than 5 years vs. more than 10 years: HR = 4.48, P < 0.001), the histological differentiation at diagnosis (Grade 3 vs. Grade 1: HR = 2.47, P = 0.025), the progression-free interval (PFI<12 months vs. PFI>24 months: HR = 2.26, P = 0.001), performance status (ECOG 2 vs. 0: HR = 5.58, P < 0.001), and the extent of recurrent disease (multiple vs. solitary: HR=1.70, P = 0.029). A prognostic scoring system was generated with the six variables and validated on the original database. According to the aggregate score, the patients were divided into low risk group (0-4 points, with MOS of 45.9 months), mid risk group (5-7 points, MOS 21.5 months) and high risk group (≥8 points,MOS 9.8 months), and the differences for stratified comparisons were all significant (low risk group vs. mid risk group: X2=29.58, P <0.001; mid risk group vs. high risk group, X2= 88.80,P< 0.001).Logistics regression revealed that the surgical outcome was significantly influenced by specialization of gynecologic oncologist, FIGO stage, performance status, serum CA125, ascites, extent of recurrent disease, and the size of recurrent disease. A scoring system for predicting the feasibility of SCR was generated with these variables and was validated on the original database. The optimal cytoreduction was achieved in 95.9% of patients with 1-4 points, 79.8% of patients with 5-7 points (OR = 5.94, P = 0.019), 51.8% of patients with 8-9 points (OR = 21.88, P < 0.001) and 29.6% of patients with > 10 points (OR = 55.95, P < 0.001), respectively.45 articles and one piece of unpublished data were included in the Meta-analysis and 2421 patients were identified. The median proportion of patients underwent optimal and complete cytoreduction were 71.5% and 50.8%, respectively. The post-recurrent survival of each cohort ranged from 10 to 60 months, with a median value of 27.75 months. The mean weighted median OS of patients with optimal and suboptimal debulking were 38.3 and 12.6 months, which were significantly different (P < 0.001). The post-recurrence survival was significantly associated with the proportion of patients underwent optimal debulking (β=29.084, P=0.002) and the proportion of patients with complete debulking (β=31.439, P<0.001). There is a trend towards better prognosis with a smaller cutoff point for optimal cytoreduction (β=-6.669, P=0.007). Multivariate linear regression shows that the year of publication (β=1.29, P=0.009), the proportion of patients with advanced stage (β=-32.84, P=0.028), the median progression free interval (β=0.67, P=0.016) and the proportion of patients with complete cytoreduction (β=26.53, P=0.008) were significantly associated with the overall survival.Conclusion Optimal secondary cytoreduction could significantly improve the survival in selected REOC patients. The survival benefit was maximally optimized in the patients rendered macroscopic disease free. The prognostic scoring system and the resectability scoring system that we had established could accurately evaluate the survival risk and the opportunity of optimal cytoreduction.
|
PDF Full Text Request