Immune checkpoint inhibitor (ICI) therapy, such as nivolumab, pembrolizumab, atezolizumab or combinations with ipilimumab, can be extremely impactful for treating cancer patients, but it is not without limitations. ICIs elicit immune-related adverse events and unfortunately, they do not work for every patient. Some patients treated with ICIs will not respond.
A recent study published in Nature Genetics aimed to identify genetic biomarkers that could distinguish ICI-responders from non-responders. Using an unbiased genetic approach, the investigators analyzed genomic and transcriptomic data among treated patients to differentiate factors representing ICI-responders versus non-responders. Their analysis revealed five latent factors: three are associated with ICI response, and two reflect tumor, microenvironment, and host factors that influence patient survival.
Three latent factors associated with ICI response:
Tumor Mutation Burden – among the genetic results obtained, the single feature with the highest correlation overall was tumor mutation burden. An increase in tumor mutation burden is associated with a higher probability of response to ICIs and an increase in survival.
Prior Treatment – prior exposure to systemic therapy and prior exposure to any therapy are negatively associated with ICI response and survival. The rationale for this is an increase in tumor aggression and/or a decline in the patient’s condition.
T cell Effective Infiltration – based on the gene expression of the cluster identified, effective T cell infiltration appears to be critical. An increase in T cell infiltration into the tumor is associated with a higher probability of a response to ICIs and longer patient survival.
Two latent factors that influence the survival of patients:
Transforming Growth Factor-Beta Activity in the Microenvironment – the highest correlation between genes in a defining gene cluster showed TGF-β activity in fibroblasts. Low values of TGF-β activity in the microenvironment are associated with longer survival of patients treated with ICIs. Therefore, TGF-β activity in this context has a negative association with response and survival.
Tumor Proliferative Potential – a group of genes from the tumor clustered with other genes sets representing cell cycle and cell proliferation and it was dubbed ‘Proliferation Potential.’ In this set and the validation cohorts that had transcriptomic data available, tumor proliferative potential was negatively associated with survival among patients.
How did researchers uncover these biomarkers?
Among the 5,288 samples in the Hartwig Medical Foundation, researchers used 479 patients with metastatic cancer who received ICI therapy and had survival information available. Among these, 191 were diagnosed with melanoma, 110 with lung cancer, 88 with bladder cancer, and the remaining 90 with other cancer types. The whole transcriptome was sequenced for 396 of these samples.
Using these samples, the following features were collected: single nucleotide variants, base insertions or deletions in the genome, copy number status of 2,415 genomic regions, somatic mutation features of the tumor, whether known driver variants occurred, and genomic instability. Other features included were the clinical characteristics of the tumor, gene expression, metastatic features, and features affecting the immune response to the cancer.
Did the authors validate independent cohorts of other tumor types?
Yes, the authors validated the five latent factors in independent cohorts of the same and other tumor types. These cohorts included INSPIRE, Lyon, MARI-ATHASAN, PARKER ICI, RAVI, and VHIO. In total, 1,491 patients with metastatic tumors were used to validate their findings.
Why does this study matter?
Having biomarkers to identify patients with a low probability of responding to ICI therapy would help rapidly move them to another option with a chance for an improved outcome. Avoiding ICIs would also spare these patients from the potential of immune-related adverse events. Importantly, avoiding a likely unsuccessful therapy can help reduce the financial burden of providers and patients.
The possibility of identifying patients with tumors not usually amenable to ICI therapy but whose genetics indicate a high probability of response to ICIs is intriguing. The authors in this study applied their analysis to patients in their original cohort who did not receive ICIs. The results identified a fraction of patients with skin (35%), bladder (42%), and lung (16%) tumors that showed a high likelihood of response to ICIs. In addition, 4% of patients with breast cancer, 3% of patients with colorectal cancer, 19% of patients with renal cancer, and 15% of patients with liver cancer also showed a high probability of response to ICIs, even though these patients were previously considered not candidates for ICIs.
Although the scientific findings are far from ready for clinical incorporation, the work provides a frame of reference for biomarkers that indicate a high likelihood of a positive response to ICI therapy. Having biomarkers is especially important as certain ICIs have more than 40 indications for a single drug in at least 20 different types of tumors. Therefore, the potential of ICIs to treat other cancer types could be increased with information to differentiate responders from non-responders.
Reference: This summary was based on the article by Usset J, Huber AR, Andrianova MA et al. Five latent factors underlie response to immunotherapy. Nat Genet. 2024;56(10):2112-2120. doi: 10.1038/s41588-024-01899-0.
