A TMB-high cutoff of 10 mutations per megabase doesn't predict immunotherapy responses universally well across cancer types, according to a paper published in JAMA Oncology.
The transcription-based subtypes could differentiate which SCLC patients respond to treatments, such as immune checkpoint inhibitors and PARP inhibitors.
FLEX will include full transcriptome data paired with 800 data points from 10,000 patients, offering researchers the chance to study diverse patient subsets.
The tool, which has been validated in two independent patient cohorts, uses artificial intelligence to predict patients likely to benefit from upfront FOLFOX.
