Article Text

Download PDFPDF

903 Identification of non-small cell lung cancer (NSCLC) patient subgroups with poor response to immunotherapy using subgroup-level meta-analyses
  1. Kinisha Gala1,
  2. Darko Medin2,
  3. Ankit Kalucha2,
  4. Judith Pérez Granado2,
  5. Bruno Larvol2 and
  6. Mark Gramling2
  1. 1Silver Spring, MD, USA
  2. 2LARVOL, San Francisco, CA, USA
  • Journal for ImmunoTherapy of Cancer (JITC) preprint. The copyright holder for this preprint are the authors/funders, who have granted JITC permission to display the preprint. All rights reserved. No reuse allowed without permission.


Background Given the heterogeneity of clinical outcomes for patients with NSCLC treated with immunotherapy, there remain subpopulations for which immunotherapy has poor efficacy. Identification of these subgroups allows for better stratification within the NSCLC treatment paradigm and may also allow for the development of novel subject-specific therapeutics. We aimed to identify subgroups with poor responses to immunotherapies by performing subgroup-level meta-analyses.

Methods A systematic search of LARVOL CLIN was performed to identify randomized phase 3 trials comparing at least one immunotherapeutic agent (pembrolizumab, cemiplimab, nivolumab, atezolizumab, durvalumab, tislelizumab, toripalimab or avelumab) to non-immunotherapy regimens for NSCLC. Trials that did not report overall survival (OS) hazard ratios (HR) for patient subgroups were excluded. The data was organized into 24 binarized subgroup pairs based on subject demographics, clinical factors and mutation status and evaluated for each of the eight drugs. The DerSimonian and Laird method was used to perform meta-analyses for each subgroup pair across all eight immunotherapuetic agents. The relative performance of each subgroup was assessed based on the difference in HR across all trials.

Results 175 trials of NSCLC involving immunotherapeutic agents were identified of which 34 met criteria. The subgroup pair of male versus female had the largest number of trials at 26 trials across the eight agents. EGFR mutation (HR difference: 0.26, p-value<0.05), PD-L1<50% (HR difference: 0.14, p-value<0.001), and TMB<16 (HR difference: 0.24, p-value<0.001) were all identified as subgroups with poor responses to immunotherapies across all eight agents. The presence of NOTCH 1–4 mutations was identified as a subgroup with favorable response to immunotherapy across all eight agents (HR difference: -0.53, p-value<0.01) (figure 1).

Conclusions Consistent with previous data, our subgroup-level meta-analyses identified low PD-L1 expression, low tumor mutation burden and the presence of EGFR mutations as subgroups that do not have favorable responses to immunotherapy. Additionally, we found that NOTCH 1–4 mutations predict a favorable response to immunotherapy. These data present the most comprehensive study of NSCLC subgroups response to immunotherapy and the methodology presented here can be expanded upon to identify previously unknown subgroups that may benefit from alternative treatment paradigms.

Abstract 903 Figure 1

Subgroup-level meta-analysis by drug for binarized NSCLC subgroups. Effect size for subgroups was considered as the different between the OS HRs and was presented as a color on the heatmap. Red represents non-favorable outcomes, green represents favorable outcomes, and yellow represents neutral favorablility. *p<0.5, **p<0.01, ***p<0.001.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.