Fig. 4.

Deeper understanding of treatment effect and exposure–response patterns through tumor burden dynamics modeling: longitudinal modeling allows for robust, quantitative characterization of data-rich pre- and on-study tumor size information on the importance of understanding pre-baseline tumor size trajectory. This hypothetical example illustrates the concept of variability in the tumor growth trajectory at study start and consequent possibility to declare RECIST progressive disease in a patient likely benefiting more from treatment (e.g. black) than another who would be classified with a more favorable stable disease designation (e.g. green). More precise estimates of treatment effect and dose/exposure–response relationships can be brought through a quantitative understanding of a patient’s entire tumor size trajectory, including available pre-baseline scans