New research seeks to explain difference in ICB therapy response

Tempus developed an integrative model of tumor-extrinsic and tumor-intrinsic features associated with longer time to progression in NSCLC patients.
Authors Michelle Stein, PhD
Director, Computational Biology, Tempus


Research summary

In metastatic non-small cell lung cancer (mNSCLC), response rates to PD-(L)1 immune checkpoint blockade (ICB) therapies vary widely, and the mechanisms of response and resistance are not well understood. Previous studies have proposed disruption of HLA class I antigen presentation as an important mechanism of immune escape and ICB resistance, as this may interfere with direct killing of tumor cells by cytotoxic CD8+ T cells.1-3

With rapid advances in DNA and RNA sequencing, the use of linked clinical, genomics, and transcriptomics datasets, like those found in the Tempus Multimodal Database, can be used to facilitate analysis of transcriptional and genomic signatures associated with ICB response, supplementing evidence generated in clinical trials or other studies.

My coauthors and I sought to identify HLA-I-independent features that were associated with ICB response in mNSCLC. We turned to Tempus’s biological modeling laboratory to generate single-cell multiomic profiling datasets from NSCLC tumor specimens – scRNAseq, T cell receptor sequencing, and surface protein profiling – to characterize the tumor and T cell compartments in NSCLC tumors from 10 patients. We discovered a novel population of tumor-infiltrating, clonally expanded CD4+ helper T cells aberrantly expressing genetic programming similar to classical cytotoxic CD8+ killer T cells. In addition, tumor cells from these lung cancers also showed expression of the HLA class II, allowing them to be susceptible to these cytotoxic CD4+ T cells. These findings add to the emerging evidence (Oh et al., Cohen et al., Awad et al.) that cytotoxic CD4+ T cells are a noteworthy component of the tumor immune microenvironment and may play a role in anti-tumor immune responses following treatment with ICB.

We leveraged the single-cell multiomic data to develop a specific transcriptomic signature that captured both CD4+ and CD8+ cytotoxic T cells populations in tumors and applied it to a real-world cohort of mNSCLC patients selected from the Tempus Clinico-genomic  Database (n=123). By combining this gene signature for cytotoxicity with tumor mutational burden (TMB), we developed an integrative model of tumor-extrinsic (cytotoxic gene signature) and tumor-intrinsic TMB features that is associated with longer time to progression in a real-world cohort of mNSCLC patients treated with ICB regimens, including those with disrupted class I HLA. These results demonstrate that integrating tumor-extrinsic and tumor-intrinsic features may be an informative biomarker for identifying mNSCLC patients who are more likely to respond to ICB and can remain effective even in populations where tumor HLA-LOH is common.

We are now working with our biopharma partners to apply this novel signature into their programs to improve patient selection. For example, biopharma is:

  • Collaborating with Tempus to design and rapidly deploy novel single-cell experiments and compare real-world multimodal data aimed at translating ICB insights into clinical biomarkers.
  • Applying our model of ICB response to enhance existing biomarkers of ICB therapies in both the metastatic and neoadjuvant setting in NSCLC and other cancers.
  • Further exploring biomarkers of response or resistance to ICB or other therapies of interest in the Tempus Multimodal Database.
  • Testing and further developing existing preclinical transcriptional signatures into actionable biomarkers.

Next steps

Contact Tempus to discuss this principled approach with one of our computational biologists. We can dive even deeper with you as it relates to results for your immunotherapy portfolio or design bespoke projects to advance your research and clinical programs.


  1. Sade-Feldman, M. et al. Resistance to checkpoint blockade therapy through inactivation of antigen presentation. Nat. Commun. 8, 1–11 (2017).
  2. Gettinger, S. et al. Impaired HLA class I antigen processing and presentation as a mechanism of acquired resistance to immune checkpoint inhibitors in lung cancer. Cancer Discov. 7, 1420–1435 (2017).
  3. Zaretsky, J. M. et al. Mutations associated with acquired resistance to PD-1 blockade in melanoma. N. Engl. J. Med. 375, 819–829 (2016).


Lau, D., Khare, S., Stein, M.M. et al. Integration of tumor extrinsic and intrinsic features associates with immunotherapy response in non-small cell lung cancer. Nat Commun 13, 4053 (2022). https://doi.org/10.1038/s41467-022-31769-4

Resource Center

View more