(W-014) Model-Based Meta Analysis for the IPSOS Trial in Platinum Ineligible NSCLC Patients

Russell Wada, PhD – CEO, QuanTx; Nancy Zhang, PhD – Vice President, Senior Consultant, QuanTx; Vilma Graupner, PhD – Principal Clinical Scientist, Roche; Stefanie Morris, PhD – Principal Global Scientific Director, Roche; Youyou Hu, PhD – Principal Data Scientist, Roche; Wei Zhang, PhD – Senior Principal Data Scientist, Genentech; Nastya Kassir, PhD – Senior Director And Distinguished Scientist, Genentech; Benjamin Wu, PhD – Director And Distinguished Scientist, Genentech; Phyllis Chan, PhD – Distinguished Scientist Modeling & Simulation, Genentech

Introduction/

Objectives: IPSOS (NCT03191786) is a Phase III trial that compared atezolizumab (atezo) monotherapy (mono) vs. single-agent chemotherapy (chemo) (gemcitabine or vinorelbine) in patients with treatment-naïve locally advanced or metastatic NSCLC deemed unsuitable for platinum-doublet chemo. The study demonstrated significant OS improvement in the atezo arm compared to single-agent chemo (gemcitabine or vinorelbine); stratified HR of 0.78 (95% CI: 0.63, 0.97; p-value of 0.028) [1].

The IPSOS control arm only allowed gemcitabine or vinorelbine, while other single agent chemo could have been used. Thus, this model-based meta-analysis (MBMA) extracted OS from published literature in patients similar to the IPSOS trial, adjusting for heterogeneity across trials, to estimate OS HR between the IPSOS active and control arms vs. historical control arms in which other single agent chemo were used. The aim was to demonstrate the non-inferiority of the IPSOS control arm vs. historical control, and assess atezo treatment benefits.

Methods: A literature search was based on the following criteria of the clinical trial participants: chemo-naïve; advanced or metastatic NSCLC; platinum-ineligible; elderly (≥ 70 years of age); ECOG ≥ 2; taking paclitaxel, docetaxel, gemcitabine, pemetrexed, or vinorelbine mono. Summary-level OS data were extracted by digitizing Kaplan-Meier curves and only trials with ≥ 50 participants were included in the MBMA database. A list of potential covariates were pre-specified, and the database was augmented by imputing missing covariate values. A nonparametric approach, modeling the conditional probability of OS data [2] was implemented.

Results: The assembled database consisted of summary-level data from 26 trials with 41 arms and 3637 participants. ECOG was the only significant covariate identified. After adjusting for ECOG, the model-predicted HR for the IPSOS control arm relative to historical trials was 0.543 (95% CI: 0.435,0.677). In addition, the HR for the IPSOS atezo mono arm relative to historical trials was 0.418 (95%CI: 0.335,0.522).

Conclusions: Overall, the results of the MBMA support the clinical benefit seen in the IPSOS trial.

Citations:

1. 
   Lee SM, Schulz C, Prabhash K, et al. First-line atezolizumab monotherapy versus single-agent chemotherapy in patients with non-small-cell lung cancer ineligible for treatment with a platinum-containing regimen (IPSOS): a phase 3, global, multicentre, open-label, randomised controlled study [published correction appears in Lancet. 2023 Aug 5;402(10400):450]. Lancet. 2023;402(10400):451-463.
   
   


2. 
   Turner DC, Wada R, Zhou H, et al. Model-based meta-analysis of non-small cell lung cancer with standard of care PD-1 inhibitors and chemotherapy for development decision making. CPT Pharmacometrics Syst Pharmacol. 2023;12(11):1751-1763.