MFI-Alb as a novel predictor combining frailty and albumin for prognosis in advanced elderly ovarian cancer

Advanced elderly ovarian cancer patients often experience adverse postoperative outcomes and prognosis. Currently, although multiple frailty assessment tools and nutritional serum biomarkers have been employed for risk assessment and outcome prediction, there remains no consensus on the optimal "gold standard" predictive tool. A retrospective study was conducted to collect general information, clinical pathological data, and follow-up information of 117 advanced elderly ovarian cancer patients aged ≥65, who underwent ovarian tumor cytoreductive surgery at Wuhan University Zhongnan Hospital between 2012 and 2021. The 5-MFI was used to assess frailty. Preoperative albumin was used to assess nutritional status.The MFI-Alb model was developed as a composite predictor integrating 5-MFI and albumin.Cox regression and logistic regression were used to investigate predicting factors.The predictive performance of the models was evaluated using AUC and ROC curves. Logistic regression analysis showed that 5-MFI and MFI-Alb were independent predictors for hospital stay and OS. The AUC for the MFI-Alb predicting hospital stay was 0.742 (95% CI 0.646-0.837), for 5-MFI was 0.697 (95% CI 0.603-0.791), and for preoperative albumin was 0.583 (95% CI 0.471-0.694), with the MFI-Alb model outperforming the other two. In terms of OS, the AUC for the MFI-Alb model predicting OS was 0.746 (95% CI 0.659-0.834), for 5-MFI was 0.734 (95% CI 0.650-0.818), and for preoperative albumin was 0.520 (95% CI 0.415-0.626), again showing superior predictive performance for the MFI-Alb model. The MFI-Alb model, which integrates frailty and nutritional status, exhibits superior predictive performance for both hospital stay length and OS compared to using 5-MFI or albumin alone.