Skip to main content
main-content
Top

24-08-2018 | Oncology | News | Article

Editor's pick

Radiomic CD8-cell signature could predict immunotherapy response

medwireNews: Researchers have developed a radiomic CD8-cell signature that could help predict the immune phenotype of solid tumours and thereby a patient’s clinical response.

This use of radiomics, which involves the translation of medical images into quantitative data, offers a means of developing noninvasive image-driven biomarkers for immunotherapy, the team comments in The Lancet Oncology.

Eric Deutsch (Gustave Roussy Cancer Campus, Villejuif, France) and co-researchers used computed tomography scans of biopsy samples and RNA sequencing data from 135 patients with advanced solid tumours who were enrolled in the MOSCATO trial to develop the eight-feature radiomics signature based on CD8 cell expression. The eight features included five radiomic features, tumour volume from two locations of interest (adenopathy and head and neck), and the peak kilovoltage variable.

This radiomic signature was able to predict high levels of CD8-cell infiltration in a group of 119 patients from The Cancer Genome Atlas database with 67% accuracy. And when tested in 100 randomly selected patients from their own registry, the radiomic score was found to be significantly higher among patients who had previously been described as having immune-inflamed (with dense CD8 infiltrate) tumour phenotypes versus those with immune-desert (low CD8 infiltrate) phenotypes, and discriminated between the two groups with 76% accuracy.

With regard to clinical response, among an independent cohort of 137 patients treated with anti-programmed cell death protein 1 or anti-programmed cell death ligand 1 immunotherapy in phase I trials, the proportion achieving an objective response at 3 months, relative to stable disease and disease progression, was significantly greater among those with high radiomic signature scores (above the median) at baseline. This was also the case at 6 months for objective response and controlled disease versus disease progression.

Patients with high baseline radiomic signature scores also had improved overall survival, at a median 24.3 months compared with 11.5 months among patients with radiomic signature scores below the median, giving a significant hazard ratio for death of 0.52, after taking into account lines of treatment and the Royal Marsden Hospital prognostic score.

“[O]ur study suggests that radiomics could be an efficient, non-invasive, cost-effective, and reliable way to evaluate patients’ responses to precision medicine”, say Deutsch and colleagues.

This was echoed in a related comment by Issam El Naqa and Randall Ten Hoken, both from the University of Michigan in Ann Arbor, USA, who say the study “has provided a stimulating analysis of the potential role of radiomics for personalising immunotherapy.”

They continue: “[R]adiomics has the potential to be a digital biopsy technique that is both spatially guided and non-invasive and that can quantify T cell infiltration of tumours, support personalised design of immunotherapy interventions, and longitudinally monitor and assess immune checkpoint blockade response.”

Both the commentators and researchers stress, however, that further evaluation of the findings in prospective clinical trials is needed.

By Lucy Piper

medwireNews is an independent medical news service provided by Springer Healthcare. © 2018 Springer Healthcare part of the Springer Nature group