In:
Journal of Clinical Oncology, American Society of Clinical Oncology (ASCO), Vol. 38, No. 15_suppl ( 2020-05-20), p. e15067-e15067
Abstract:
e15067 Background: Approvals of immune checkpoint inhibitors (ICI) were made based on positive clinical trial results analyzed by the Cox proportional hazards (PH) model. With ICI data, however, long tails and early crossover in survival curves, which violate the Cox PH assumption, can lead to misinterpretation of clinical significance of findings. Here we introduce the Cox-TEL and show the differences of study results before and after Cox-TEL adjustment using KEYNOTE 042 and 045 as examples. Methods: Cox-TEL is built on the mathematical foundation of Taylor expansion. As an easily implemented alternative of PH cure model, it not only infers associations between survival probabilities of the two study arms among patients without long-term survival (poor-responders), but also estimates differences in proportion (DP) between arms among patients in the long-tail segment of the survival curve (true-responders). Results: In KEYNOTE 042, the Cox-TEL HRs for death were statistically insignificant across all subgroups. The trend of DP, on the other hand, is positively related to that of PD-L1 TPS and inverted related to that of Cox HR when the PD-L1 ≥50% cohort is covered. In KEYNOTE 045, the Cox-TEL HRs suggested that for the poor-responders, pembrolizumab did not do better than chemotherapy in terms of overall survival (OS) and might do harm to the patients in terms of progression-free survival (PFS). For the true-responders, DPs of OS and PFS were both statistically significant (Table). Conclusions: Our data demonstrated the biases derived from insufficient data analyses and strengthened the necessity of analytic model revisits in the new oncology era of which cure for advanced cancers is no longer impossible. [Table: see text]
Type of Medium:
Online Resource
ISSN:
0732-183X
,
1527-7755
DOI:
10.1200/JCO.2020.38.15_suppl.e15067
Language:
English
Publisher:
American Society of Clinical Oncology (ASCO)
Publication Date:
2020
detail.hit.zdb_id:
2005181-5
