Advances in Mixture Cure Models: Modeling Time-to-Event Data with Long-Term Survivors

Medical breakthroughs in recent decades have led to cures for many diseases including cancer. For example, various groups have shown that advances in therapy for cancer have increased overall survival rates and some groups have identified factors related to long-term survival, defined as survival exceeding three years, which included that patients treated on newer treatment regimens were more likely to be long-term survivors. In fact, some argue that these improved outcomes indicate that some cancer patients can be considered "potentially cured." The mixture cure model (MCM) is a time-to-event model that is used when a cured fraction exists. MCMs assume the population consists of two subgroups, those cured will not experience the event of interest and those susceptible to the event of interest. Therefore, "cured" can be considered synonymous with attaining long-term relapse-free survival. Thus, there are two regression components in MCMs, which are characterized by their ability to simultaneously model the probability of long-term survival and the latency distribution for those susceptible to the event. Thus, MCMs offer a powerful means of analysis. The primary goal of this session is to inform analysts about recent developments for modeling time-to-event outcomes when long-term survivors comprise the dataset. The talks will discuss the statistical challenges associated with different types of mixture cure models and inferential questions, including a) evaluating assumptions associated with mixture cure models; b) fitting mixture cure models to clustered survival data; and c) fitting models with both long-term survivors and competing risks. The session will provide an overview of advances in mixture cure models and how they can provide users with better, more informative insights on variables that impact clinical time-to-event outcomes.