Clinicopathologic predictors of early relapse in advanced epithelial ovarian cancer: development of prediction models using nationwide data

1

Introduction

Epithelial ovarian cancer (EOC) is the most lethal gynecologic malignancy in the western world [ 1 , 2 ] Worldwide, approximately 240,000 new cases and 185,000 disease-related deaths from EOC occur annually [ 2 ]. The mortality rate remains high as the vast majority of patients is still diagnosed with advanced stage (i.e. International Federation of Gynecology and Obstetrics (FIGO) stages IIB-IV) disease and a very high likelihood to develop recurrent disease [ 345 ]. In advanced EOC, standard treatment includes cytoreductive surgery combined with platinum-based chemotherapy. While most patients respond to treatment, 15–20 % have intrinsic resistance toward platinum and often succumb to the disease shortly after diagnosis [ 6 ]. Many others experience disease recurrence after initial response to treatment (∼60−80%). One-fourth of these recurrences occur within six months after completing first-line treatment [ 3 ]. Quantifying risk of early relapse in patients with advanced disease could assist in the counselling of individual EOC patients, leading into a more personalized care for them.

Prior studies on prognostic factors of early relapse (defined as a PFI ≤ 6 months) have primarily focused on biomarkers, molecular or genetic factors that contribute to the development of early progressive or recurrent disease [ 4 , 5 , 7 , 8 ]. While numerous clinicopathologic factors have been studied for progression-free and overall survival in advanced EOC, it is uncertain if these factors could also be used to accurately predict the risk towards early relapse. For instance, studies have suggested that widespread disease that is inaccessible for primary cytoreductive surgery and >1 cm residual disease after surgery lead to an increased risk of early relapse [ 9 , 10 ]. However, these studies are often hampered by their limited sample size and their extent of missing data. No studies that quantify associations between clinicopathologic factors and early relapse (defined as a PFI ≤ 6 months) to develop clinical prediction models using population-based data have been conducted.

If patients who are expected to derive little to no benefit from standard platinum-based treatment are identified early, then intervention with alternative approaches (e.g. novel targeting therapies or dose-dense chemotherapy) or even discontinuation of chemotherapy might be considered. Therefore, the aim of this study is to develop and internally validate two prediction models (a pretreatment and a postoperative model) for early relapse in advanced stage EOC patients during or after first-line treatment using nationwide data.