Bayesian Analysis of Factors Affecting Long-Term and Short-Term Survival of Breast Cancer Patients Using the Smooth Semi-Nonparametric Mixture Cure Model



Fs Kamel Fatemeh SEDIGHI ; A Rasekhi ALIAKBAR ; Sh Haghighat Shahpar H
Source: Iranian Journal of Breast Diseases Published:2024

Abstract

Introduction: Today, according to the advancements in cancer treatment, a fraction of patients never experience an adverse event, such as death, even when the duration of the disease is prolonged. Cure models are used in the analysis of these types of diseases. In this study, we examined the survival of patients, the cure probability, and the affecting factors among breast cancer patients. Methods: We analyzed the data of 1,247 breast cancer patients who referred to Motamed Jihad University Research Institute in Tehran between 1995 and 2013 and followed them up until 2018. Data analysis was done using R version 4.3.0 software to check the survival time of uncured patients and the cure rate and to identify the effective factors with the Bayesian estimation method by fitting the semi-nonparametric smooth mixture cure model. Results: The results of this study showed that out of 1,247 patients with breast cancer, 82.8% of the patients were censored, and 17.2% of the patients died. The cure rate was 58%, according to the Kaplan-Meier curve. Examining the factors affecting the death of patients showed that the patient's high weight, more advanced stages of the disease, involvement of lymph nodes, and breast-conserving surgery were effective on the time to death and short-term survival. Conclusion: Based on the results of this, there are several crucial prognostic factors associated with breast cancer that play a significant role in identifying high-risk patients and choosing the type of treatment in the short term. © 2024 Elsevier B.V., All rights reserved.