Five features predict response to checkpoint inhibitor (CPI) chemotherapy across a wide range of cancers, research indicates.
The findings, in Nature Genetics, could drive the personalization of cancer treatment by identifying people most likely to benefit from immunotherapy.
They could also widen the use of immunotherapy, as several groups of patients who would not currently be considered as candidates were flagged as potentially benefiting from treatment.
The researchers believe that the five features constitute a reference framework for current and future biomarkers of CPI response and survival.
“So far, many studies have focused on identifying and reporting individual biomarkers, but our results suggest that many of these biomarkers might be different versions of the same underlying factors,” said researcher Abel González-Pérez, PhD, a bioinformatician from The Barcelona Institute of Science and Technology in Spain.
CPIs have had a tremendous impact on cancer therapy but patient response varies considerably, and treatment can result in significant immune-related adverse events.
It has become increasingly clear that CPI response and subsequent survival is mediated by latent factors that include tumor characteristics, its microenvironment, and the host.
Biomarkers identified across different studies may in fact represent different measurements of these underlying aspects of a tumor, making it difficult to determine how many truly independent markers there are.
For example, the expression of genes and gene sets previously identified as biomarkers may collectively represent the degree of infiltration of cytotoxic cells in the tumor.
To investigate further, the researchers studied thousands of molecular and clinical features in 479 patients with metastatic tumors who were part of the Center for Personalized Cancer Treatment study.
Participants received anti-PD1/PDL1 or a combination of anti-PD1/PDL1 and anti-CTLA4 therapy.
The analysis of genomics, transcriptomics and clinical data revealed that all significantly associated features collapsed into one of five independent latent factors that are relevant across all the tumor types.
These were: the tumor mutation burden (TMB); effective T cell infiltration; whether the patients received any prior treatment; the activity of transforming growth factor-beta (TGF-β) in the tumor microenvironment; and the proliferative potential of the tumor.
Their association with CPI response and survival was validated across six independent cohorts, comprising 1491 individuals, which spanned six major cancer types.
The researchers examined how the five latent factors combined using multivariate machine-learning models to predict the response, overall survival or progression-free survival of the original group of patients.
When applied to patients in the HMF cohort who did not receive CPIs, this revealed an important proportion of the patients with skin (35%), bladder (42%) and lung (16%) tumors who were highly likely to respond to the treatment.
Patients suffering from other metastatic malignancies, some of whom would not usually be considered as candidates for CPI, were also identified as potentially good responders.
For example, 4% of breast cancer patients, 3% of those with colorectal cancer, 19% of those with kidney tumors, and 15% of those with liver cancer exhibited high probability of response to CPI.
“This study represents an important step in understanding how different tumour characteristics influence treatment response,” said researcher Nuria Lopez-Bigas, PhD, also from The Barcelona Institute of Science and Technology.
“In the future, we hope that these five factors will be integrated into clinical practice to guide therapeutic decisions.”
