Introduction/Background Ovarian cancer has high mortality rates since the majority of patients are diagnosed with advanced disease and develop a recurrence. Identification of specific recurrence patterns that impact prognosis could be useful to distinguish between patients with a relatively good versus a poor prognosis.

Methodology 179 patients with epithelial ovarian cancer diagnosed between January 2008 and December 2013 received cytoreductive surgery in the University Medical Center Utrecht. 124 of these patients developed recurrent disease of whom 89 patients met the inclusion criteria and formed the study population. Clinical and pathological data was obtained and >40 anatomical sites were scored on the presence of tumour on Computed Tomography scan (CT-scan). A Cox Proportional Hazard model was used to study the influence of the recurrence pattern on survival.

Results The median survival of patients with mainly intraperitoneal (n=62), hematogenous (n=13) or lymphatic (n=14) disease at recurrence was 25.8, 27.6 and 52.9 months, respectively. Univariate Cox regression identified the following prognostic factors for overall survival (OS): main type of disease at recurrence (p=0.03), residual tumour at initial surgery (p=0.001), ascites at primary disease(p<0.001), BRCA mutation status (p<0.001) and tumour histology (p=0.04). In the multivariate analysis including these covariates and age, main type of disease retained a prognostic role for OS. Patients with complete debulking had more often mainly lymphatic disease at recurrence (26 vs 7%), as compared to patients with incomplete debulking who had more often mainly peritoneal disease (76 vs 63%, p=0.045).

Conclusion Patients with mainly lymphatic disease at recurrence had a better OS than patients with mainly peritoneal or hematogenous disease. Classification of the main type of disease at recurrence can help to distinguish between patients with a relatively good versus a poor prognosis.

Disclosure Nothing to disclose.