Development of Web-Based Nomograms to Predict Treatment Response and Prognosis of Epithelial Ovarian Cancer

• Published in: Cancer research and treatment Vol. 51; no. 3; pp. 1144 - 1155
• Main Authors: Kim, Se Ik, Song, Minsun, Hwangbo, Suhyun, Lee, Sungyoung, Cho, Untack, Kim, Ju-Hyun, Lee, Maria, Kim, Hee Seung, Chung, Hyun Hoon, Suh, Dae-Shik, Park, Taesung, Song, Yong-Sang
• Format: Journal Article
• Language: English
• Published: Korea (South) Korean Cancer Association 01.07.2019; 대한암학회
• Subjects: Adult; Aged; Carcinoma, Ovarian Epithelial - pathology; Carcinoma, Ovarian Epithelial - therapy; Cytoreduction Surgical Procedures; Disease-Free Survival; Female; Humans; Internet; Middle Aged; Nomograms; Original; Ovarian Neoplasms - pathology; Ovarian Neoplasms - therapy; Platinum - therapeutic use; Prognosis; Retrospective Studies; ROC Curve; Survival Rate; Tertiary Care Centers; Treatment Outcome; 의약학; Survival outcomes; Treatment response; Prognosis; Ovarian epithelial cancer; Ovarian neoplasms; Nomograms
• ISSN: 1598-2998; 2005-9256; 2005-9256
• DOI: 10.4143/crt.2018.508

Discovery of models predicting the exact prognosis of epithelial ovarian cancer (EOC) is necessary as the first step of implementation of individualized treatment. This study aimed to develop nomograms predicting treatment response and prognosis in EOC. We comprehensively reviewed medical records of 866 patients diagnosed with and treated for EOC at two tertiary institutional hospitals between 2007 and 2016. Patients' clinico-pathologic characteristics, details of primary treatment, intra-operative surgical findings, and survival outcomes were collected. To construct predictive nomograms for platinum sensitivity, 3-year progression-free survival (PFS), and 5-year overall survival (OS), we performed stepwise variable selection by measuring the area under the receiver operating characteristic curve (AUC) with leave-one-out cross-validation. For model validation, 10-fold cross-validation was applied. The median length of observation was 42.4 months (interquartile range, 25.7 to 69.9 months), during which 441 patients (50.9%) experienced disease recurrence. The median value of PFS was 32.6 months and 3-year PFS rate was 47.8% while 5-year OS rate was 68.4%. The AUCs of the newly developed nomograms predicting platinum sensitivity, 3-year PFS, and 5-year OS were 0.758, 0.841, and 0.805, respectively. We also developed predictive nomograms confined to the patients who underwent primary debulking surgery. The AUCs for platinum sensitivity, 3-year PFS, and 5-year OS were 0.713, 0.839, and 0.803, respectively. We successfully developed nomograms predicting treatment response and prognosis of patients with EOC. These nomograms are expected to be useful in clinical practice and designing clinical trials.