Abstract
In a longitudinal analysis of patients with metastatic melanoma who received sequential therapies targeting CTLA-4, then PD-1 upon disease progression, researchers have discovered that high copy-number loss in recurrent regions of the tumor genome is associated with resistance to immune checkpoint blockade.
In a longitudinal analysis of patients with metastatic melanoma who received sequential therapies targeting CTLA-4, then PD-1 upon disease progression, researchers have discovered that high copy-number loss in the tumor genome is associated with resistance to these immune checkpoint blockers.
“We wanted to get a handle on what the molecular correlates might be of responders versus nonresponders to immunotherapy,” says co–senior author Andrew Futreal, PhD, chair of genomic medicine at The University of Texas MD Anderson Cancer Center in Houston. “There's been much slicing and dicing of data on known mutations in melanoma, but nothing as far as correlation goes.”
Futreal and his team first sequenced a genetically variable region of the T-cell receptor (TCR) in serial tumor samples from their 56-patient cohort—obtained before, during, and after treatment—to assess clonality. They observed enrichment of particular TCRs, an indicator of robust T-cell activity, in patients who responded well to PD-1 blockade after anti–CTLA-4 therapy.
“It fits other studies showing that a more clonal T-cell repertoire—potentially primed by CTLA-4 blockade in this instance—is predictive of response to anti–PD-1 therapy,” Futreal says. “However, we also wanted to figure out what the tumor genome itself might bring to the table, so we began looking at changes in copy number.”
The team reported that high copy-number loss, but not gain, correlated with resistance to both CTLA-4 and PD-1 blockade. These were recurrent losses in specific regions, chiefly chromosomes 6q, 10q, and 11q, which house well-known tumor suppressor genes such as FOXO3, PTEN, and CBL.
Using sequencing results from an independent 110-patient cohort given only anti–CTLA-4 therapy, the researchers validated their findings. “We reran that dataset through our system; the same correlation panned out for single-agent treatment—high copy-number loss indicates who's unlikely to respond,” Futreal says. “At the genomic level, then, the tumor is an active participant in its own fate.”
Copy-number loss plus neoantigen load may yield “a combinatorial biomarker that helps guide clinical use of immunotherapy,” says Margaret Callahan, MD, a medical oncologist at Memorial Sloan Kettering Cancer Center (MSKCC) in New York, NY: When Futreal's team used these two features to retrospectively analyze patients from the independent dataset, they found that clinical benefit from CTLA-4 blockade was greatest in those with low copy-number loss and a high neoantigen load.
This study “examines more deeply some of the immunogenomic factors thought to influence the likelihood of response to checkpoint blockade,” says Timothy Chan, MD, PhD, director of translational oncology at MSKCC. “The results will be an important resource for fine-tuning predictive models of which patients derive the most benefit from these therapies.”
Futreal notes that “rather than extrapolate the rules of the road from melanoma, we need to see if similar correlations emerge in other cancers.” He's also interested in probing a role for the germline, in addition to T cells and tumor—specifically, how the “interplay of three genomes” influences an individual's ability to mount an immune response. –Alissa Poh
