OBJECTIVES: Breakthrough innovation in oncology is commonly considered to be a substantial increase in overall survival (OS) above specific thresholds.  However, this growing emphasis on significant OS gain, undervalues new products and may not capture aspects of treatment that are important to patients, including quality of life, delayed progression or improvements in side-effect profiles. We argue that progressive innovation in clinical and non-clinical domains needs to be considered and valued in regulatory and health technology assessment (HTA) decisions. METHODS: Definitions of innovation used in regulatory, HTA, and industry were reviewed in the academic and gray literature and current/proposed OS thresholds applied to therapies in colorectal cancer (CRC) and non-small cell lung cancer (NSCLC).  RESULTS: Regulators and HTA agencies do not provide clear consistent definitions of innovation; however, the magnitude of OS benefit is a consistent key aspect from both policy (e.g., England’s Cancer Drugs Fund) and clinical perspectives (e.g., ASCO). Emphasis is on “clinically meaningful” change expressed as minimum thresholds: OS gain >2.5months; HR >0.8; PFS gain >3 months, HR >0.5.. Only one of six CRC drugs approved since 2000 met these thresholds (ASCO; Ellis et al.