No abstract

, 382 patients met inclusion criteria. Of those, 134 (35%) had complete gross resection with no RD. On multivariate analysis, 5 criteria were found to be associated with RD: widespread small bowel serosal involvement (OR 11.2, 95% CI 10.4-12.1); presacral extraperitoneal disease (OR 4.2, 95% CI 1.2-14); lesions in the gallbladder fossa/liver intersegmental fissure (OR 2.2, 95% CI 2-2.3); disease in the small bowel mesentery, around its root, or the root of the superior mesenteric artery (OR 2, 95% CI 1.3-3); and parenchymal splenic/perisplenic tumor (OR 1.3, 95% CI 1.3-1.3). Because of the high association between widespread small bowel involvement and RD (OR 11.2), a 2-step predictive model was developed. In the first step, the cohort was assessed for the presence of widespread small bowel involvement; among the 102 patients with that finding, the rate of having any RD was 96%. In the second step, among the remaining 280 patients, a 'predictive score' was assigned to each of the 4 other criteria, which was based on their multivariate OR. A total predictive score was then calculated for each patient using their individual findings, and the rate of having any RD for patients who had a total score of 0-1, 2-3, 4-5, and 6 or more was 42%, 63%, 75%, and 89%, respectively (Table 1). A receiver operating characteristic curve generated for the model showed an AUC of 0.77. Conclusions: In 2 high-volume ovarian cancer centers, we identified 5 surgical criteria associated with RD at PDS. We developed a multivariate model in which the rate of having any RD was directly proportional to a predictive score. With further confirmation, this model could form the basis of a laparoscopic assessment to help determine resectability and triaging to laparotomy and attempted PDS versus neoadjuvant chemotherapy.

Methods: Participants with high grade ovarian, fallopian tube, primary peritoneal cancer who relapsed within 12 months of previous platinum-based therapy were randomised, with stratification for germline BRCA1/2 mutation status, prior PARP inhibitor (PARPi) or anti-angiogenic therapy, to receive O (300mg daily) or either wP (80mg/m 2 d1,8,15, q28) or O+C (300mg/20mg daily respectively). Treatment continued until disease progression or unacceptable toxicity. Primary endpoint was Progression Free Survival (PFS). Secondary endpoints were safety and tolerability of the combination of O+C, overall survival, objective response rate, and quality of life. A sample size of 138 (46 per arm) was sufficient to observe HR 0.64 in favour of O+C compared to O alone and 1.44 for wP compared to O (20% one-sided type I error, 80% power, 15% dropout, one-sided p-value <0.2).Results: OCTOVA recruited 139 participants, median 65 years (IQR:38-84), from 15 UK centres over 34 months. 31 (22%) had had prior PARPi therapy, 47 (34%) prior antiangiogenic therapy. 41 (29%) had known germline BRCA1/2 mutations. 125 (90%) had relapsed <6 months after prior platinum. Median 2 prior lines of chemotherapy (range 1-7). There was no difference in PFS between wP and O (HR¼0.97, 60%CI: 0.79, 1.19; p¼0.55). PFS was increased for O+C compared to O (HR¼0.70; 60%CI: 0.57, 0.86; p¼0.08). There were 239, 176, and 137 treatment related adverse events (all grades) in the O+C, wP and O arms respectively. Conclusions:The OCTOVA trial demonstrated that the combination of O+C showed greater efficacy than O alone, but we did not find evidence that wP was inferior to O in women with multiply relapsed OC.

No abstract