How to decide whether to offer and use "nonstandard" therapies in patients with advanced sarcomas and gastrointestinal stromal tumors: global variations in clinical practice, assessment, and access to therapies in diseases with limited incidence and data.

As cancer is more generally recognized as a collection of various rare diseases rather than a homogeneous illness, sarcomas have become a model for the manner in which data can and cannot be used to drive clinical decision making. In this article, we explore the limitations of data generated in rare diseases such as sarcomas to provide an evidence base for clinical practice. How should patients be treated if there is no "standard" that offers "proof" of clinical benefit? By asking this question, we also raise the issue of what constitutes "clinical benefit"-and how to measure that-for patients with sarcomas and other rare diseases. As physicians become more accountable for decisions-and yet are always accountable to the patients and families who rely on them to provide the best and most appropriate care-oncologists must be cognizant of the limitations of data in rare diseases and be ready to justify actions that are in the best medical and social interests of patients.