Tumor Growth Rate Informs Treatment Efficacy in Metastatic Pancreatic Adenocarcinoma: Application of a Growth and Regression Model to Pivotal Trial and Real-World Data

• Published in: The oncologist (Dayton, Ohio) Ohio), 2023-02, Vol.28 (2), p.139-148
• Main Authors: Yeh, Celine, Zhou, Mengxi, Sigel, Keith, Jameson, Gayle, White, Ruth, Safyan, Rachael, Saenger, Yvonne, Hecht, Elizabeth, Chabot, John, Schreibman, Stephen, Juzyna, Béata, Ychou, Marc, Conroy, Thierry, Fojo, Tito, Manji, Gulam A, Von Hoff, Daniel, Bates, Susan E
• Format: Article
• Language: English
• Subjects: Adenocarcinoma - drug therapy; Antimitotic agents; Antineoplastic agents; Care and treatment; Development and progression; Diagnosis; Dosage and administration; Gastrointestinal Cancer; Humans; Metastasis; Pancreatic cancer; Pancreatic Neoplasms; Pancreatic Neoplasms - pathology; Treatment Outcome
• ISSN: 1083-7159; 1549-490X
• DOI: 10.1093/oncolo/oyac217

Methods for screening agents earlier in development and strategies for conducting smaller randomized controlled trials (RCTs) are needed. We retrospectively applied a tumor growth model to estimate the rates of growth of pancreatic cancer using radiographic tumor measurements or serum CA 19-9 values from 3033 patients with stages III-IV PDAC who were enrolled in 8 clinical trials or were included in 2 large real-world data sets. g correlated inversely with OS and was consistently lower in the experimental arms than in the control arms of RCTs. At the individual patient level, g was significantly faster for lesions metastatic to the liver relative to those localized to the pancreas. Regardless of regimen, g increased toward the end of therapy, often by over 3-fold. Growth rates of PDAC can be determined using radiographic tumor measurement and CA 19-9 values. g is inversely associated with OS and can differentiate therapies within the same trial and across trials. g can also be used to characterize changes in the behavior of an individual's PDAC, such as differences in the growth rate of lesions based on metastatic site, and the emergence of chemoresistance. We provide examples of how g can be used to benchmark phase II and III clinical data to a virtual reference arm to inform go/no go decisions and consider novel trial designs to optimize and accelerate drug development.