In early-line oncology clinical trials, evaluation of treatment effect on overall survival (OS) is often confounded by the issue of treatment drop-in, where patient enter into next line of treatment after they fail the current study treatment due to either disease progression or toxicity. In this presentation, the impact of treatment drop-in to the inference of study treatment effect in both superiority and non-inferiority trials will be discussed. Different statistical methods on how to address the treatment drop-in issue as well as real-life examples will be provided.