Patient population
From a multi-center international database (collaborative database of Peritoneal Surface Oncology Group International (PSOGI) and BIG-RENAPE working groups [
9]), the “CRS plus HIPEC” group represented patients with PM treated with CRS and HIPEC. The “CRS alone” patients with PM treated with CRS but without HIPEC were retrieved from hospital database of the Institut de Cancérologie de Lorraine. Ethics approval was obtained from the participating institutions through their institutional review boards or through the chairpersons of their ethics committees.
The inclusion criteria were patients with primary or advanced peritoneal carcinomatosis of endometrial origin, giving consent to the procedure and without contraindications either to CRS alone or HIPEC. Patients with pre-operative extra-abdominal metastasis, unresectable disease or lack of fitness for the procedure were excluded.
A total of 44 patients in the “CRS plus HIPEC” group and 90 patients in the “CRS alone” fulfilled inclusion and exclusion criteria. In order to control the potential confounding factors, patients of “CRS plus HIPEC” and “CRS alone” groups were 1:1 matched by the global optimal algorithm [
18] based on propensity score. The exact matching was performed on three criteria: age at diagnosis (± 10 years), histological type (endometrioid vs adenocarcinoma vs other carcinosarcoma), and year of surgery (± 5 ans); the propensity score was computed by a multivariate logistic regression with group as dependant parameter and all patients and clinical characteristics as independent parameters.
The main clinical data were collected retrospectively from patients treated for peritoneal carcinomatosis. Age, histological type, tumor histology, peritoneal cancer index (PCI), surgical procedure, HIPEC techniques and drugs, completeness of cytoreduction (CC) score, data regarding primary treatment, chemotherapy, postoperative complications according to the common terminology criteria for adverse events (CTCAE) v3.0 of the National Institute of Health and complete follow-up information were collected. Staging was performed on imaging data, including computed tomography (CT), magnetic resonance imaging, positron emission CT or laparoscopic exploration for resectability evaluation. Approval of treatment were established at multidisciplinary meetings.
All surgical explorations and procedures were under the direction a senior surgeon. All patients were judged to be completely resectable during surgical exploration. The extent of carcinomatosis was assessed using the Peritoneal Cancer Index (PCI), obtaining a score between 0 and 39 [
19]. Surgery was performed in order to obtain a complete resection of all visible tumor nodules. Peritonectomy procedures were performed when the peritoneal surfaces were macroscopically affected. After completion of the surgical cytoreduction, the Completeness of Cytoreduction Score (CC-S) was evaluated by the surgeon before HIPEC perfusion and was classified as follows: CC-0 = no macroscopic residual cancer, CC-1 = residual nodules < 2.5 mm, CC-2 = residual nodule between 2.5 and 25 mm, CC-3 = residual nodule > 25 mm.
HIPEC was delivered at the end of surgery according to centers preferences and the technic previously described [
20]. The intraperitoneal chemotherapy protocol used cisplatin, doxorubicin or mitomycin. The mixture was placed in contact with the peritoneal cavity at a dose of 2 l/m
2 of body surface for 60 to 90 min at a controlled temperature between 41 and 43 °C.
The overall survival (OS) was evaluated from the date of surgery to the date of death or last follow-up, and reported at 3 and 5 years. The progression free survival (PFS) was evaluated from the date of surgery to the date of documented disease progression or recurrence assessed on cross-sectional imaging.
Statistical analysis
Quantitative parameters were described as mean and standard deviation or by median and interquartile range (IQR) and qualitative parameters as frequency and percentage. Normality of the distribution was assessed by Shapiro–Wilk test. Patients’ characteristics at surgery were compared between the two groups with paired sample Student t-test or paired sample Wilcoxon test or Mac Nemar test in order to take into account the matching and paired differences were computed.
OS and PFS were described by the Kaplan Meier method and compared by univariate Cox proportional hazards regression model using a robust sandwich‐type variance estimator for the clustering within matched groups. Results were adjusted on the remaining unbalanced characteristics between the two groups by a multivariate Cox proportional hazards regression model. Results were expressed as hazard ratio (HR) and 95% confidence interval (95% CI) with “CRS only” group as reference.
The percentage of Grade 3 and 4 complications was compared with Mac Nemar test.
Significance level was set at 5%. The analyses were performed with SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA).