NHMRC Clinical Trials Centre, Sydney; Flevohospital, Almere; Academic Medical Centre, Amsterdam, Netherlands
Dirkje Willemien Sommeijer , Qian Shi , Everardo D. Saad , Elisabeth Coart , Marc E. Buyse , Tomasz Burzykowski , Jeffrey P. Meyers , Tim Maughan , Richard A. Adams , Matthew T. Seymour , Leonard Saltz , Richard M. Goldberg , Jean-Yves Douillard , Hans-Joachim Schmoll , Cornelis J. A. Punt , Christophe Tournigand , Benoist Chibaudel , Aimery De Gramont , Daniel J. Sargent , John Raymond Zalcberg
Background: To expedite drug development and improve treatment (trt) decisions, early indicators of trt efficacy are needed. We evaluated the pt-level association between early tumor shrinkage (ETS; ≥ 20% decrease from baseline), early objective tumor response (EOTR; CR/PR by RECIST), and early nonprogression status (EnPD;CR/PR/SD by RECIST) at 6, 8/9, or 12 weeks (wks) with OS, and compared these with standard endpoints (best overall response [BOR] and confirmed response [ConfR]) in pts treated with 1st-line CT. Methods: 16 phase III trials, including 10,962 pts on 5FU-LV/capecitabine alone or with oxaliplatin/irinotecan, from the ARCAD database were analyzed. Associations between early endpoints and OS were tested by stratified Cox models with a landmark approach. Prediction performance was compared by the C-index (c), a measure of relative accuracy to distinguish those at high versus low risk of death (higher value indicates better risk discrimination). Results: ETS and EOTR were significant predictors of OS with similar hazard ratios (HRs) to BOR and ConfR (p <0.0001; see Table). Pts without early progression (non-PD) showed the largest risk reduction (HRs ~ 0.3). Adjusting for age, performance status, number of metastases, and prior trt, significance remained. C-indices were similar across early endpoints, and similar to those for BOR and ConfR. Conclusions: Early responses are significantly associated with prolonged OS in mCRC pts on 1st-line chemotherapy. Early PD strongly correlates with an increased hazard of death. The pt-level prediction accuracy is as strong as for standard endpoints with the advantage of earlier tumor evaluation. Similar performance of ConfR implies the lack of utility for response confirmation. Formal surrogacy testing at trial level is ongoing.
Weeks | EOTR |
ETS |
EnPD |
BOR |
ConfR |
---|---|---|---|---|---|
HR (95% CI) [c] | HR (95% CI) [c] | HR (95% CI) [c] | HR (95% CI) [c] | HR (95% CI) [c] | |
6 | .61 (.54-.69) [.55] | .61 (.55-.67) [.57] | .31 (.27-.36) [.56] | .59 (.56-.63) [.58] | .65 (.59-.70) [.55] |
8/9 | .60 (.62-.69) [.58] | .64 (.57-.71) [.57] | .24 (.19-.30) [.56] | ||
12 | .61 (.56-.65) [.58] | .63 (.58-.67) [.57] | .35 (.32-.39) [.57] |
Disclaimer
This material on this page is ©2024 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org
Abstract Disclosures
2024 ASCO Gastrointestinal Cancers Symposium
First Author: Kohei Shitara
2024 ASCO Annual Meeting
First Author: Wenhua Li
2024 ASCO Gastrointestinal Cancers Symposium
First Author: Hendrik-Tobias Arkenau
2024 ASCO Gastrointestinal Cancers Symposium
First Author: Thierry Andre