NEW YORK – Mutations across many genes and pathways may provide clues to immune checkpoint immunotherapy response, new research suggests, pointing to the potential for improving treatment response prediction models with tumor exome or genome sequencing data.
"These results suggest that the use of broader diagnostics such as whole-exome or even whole-genome sequencing may significantly improve our ability to predict who will respond to immunotherapy — essentially, showing that more data does help to better predict treatment response," Weill Cornell Medicine researcher Marcin Imieliński, co-senior author of a study published in Nature Communications on Friday, said in a statement.
Researchers from Weill Cornell Medicine, the New York Genome Center, and New York University started with clinical and exome sequencing data for 319 cancer patients who had received an immune checkpoint inhibitor.
The cases included patients with melanoma, non-small cell lung cancer, bladder cancer, or head and neck cancer who took part in half a dozen past studies, they explained, and had undergone exome sequencing prior to treatment and "Response Evaluation Criteria in Solid Tumor" (RECIST) classification after treatment.
Using sequence data spanning almost 19,700 protein-coding genes and a modified version of a statistical modeling analytical method called fishHook — designed for weeding out irrelevant background mutations — the team attempted to track down and expand the repertoire of potential biomarkers for predicting immune checkpoint response across the diverse set of cancer cases considered.
They noted that response to immune checkpoint blockade immunotherapy was achieved in between 6 percent and 56 percent of the patients, including more than a dozen complete responders and 80 partial responders. Another 47 patients had stable disease, and 178 patients experienced disease progression. Treatment response was particularly common in patients with NSCLC or melanoma, as well as in older patients.
From a set of more than 129,300 tumor mutations with effects that were predicted to be high or moderate, the investigators narrowed in on six recurrently mutated genes. Four of those genes — BRAF, KRAS, TP53, and BCLAF1 — showed significant ties to checkpoint immunotherapy responses, as did alterations affecting MAP kinase, immune modulating, or p53-related pathways.
In KRAS and BRAF, for example, recurrent mutations corresponded with better-than-usual immunotherapy responses. On the other hand, relatively muted checkpoint immune blockade responses were detected in the cancer patients who had recurrent TP53 or BCLAF1 mutations in their tumors.
After identifying these treatment response-related, recurrently mutated genes and pathways, the team went on to develop a so-called "Cancer Immunotherapy Response Classifier," or CIRCLE, that appeared to boost the sensitivity and specificity of immune checkpoint blockade response models that included tumor mutational burden, tumor type, patient age, and other potential response predictors.
"Scientists have developed various biomarkers that help anticipate immunotherapy treatment response, but there's still an unmet need for a robust, clinically practical predictive model," co-senior author Neville Sanjana, a researcher affiliated with NYU and the New York Genome Center, said in a statement.
"We envision that this two-step approach and use of whole-exome sequencing will pave the way for better prognostic tools for cancer immunotherapy," Sanjana added.
The predictive potential for CIRCLE classification was further validated with exome sequence data on 165 additional immunotherapy-treated cancer patients, the researchers reported, noting that they expect to continue testing and enhancing the model using data from ever larger cohorts of cancer patients with available tumor sequence and immunotherapy response data.
"We envision that CIRCLE and, more broadly, the analysis of recurrently mutated cancer genes will pave the way for better prognostic tools for cancer immunotherapy," the authors suggested.