The T-cell repertoire in NSCLC: Therapeutic implications

Why is the T-cell repertoire so similar across normal lung and tumor (and what are these T cells really recognizing)?

“Well, we performed whole-exome sequencing and it’s really no surprise that mutational load is substantial in the tumor, but what was surprising was the amount of mutations we detected in the normal lung,” Dr. Reuben said.

The number was lower than in the tumor, though still considerable, and included a large proportion that were shared mutations between the normal lung and the tumor, he noted.

A closer look at the shared mutations showed that they correlated positively with the proportion of shared dominant T cells between the normal lung and the tumor, suggesting that some of the shared T cells may be targeting shared mutations between the normal lung and the tumor. The correlation was weak, but statistically significant, so while it doesn’t account for all of the overlap, it likely accounts for some of the homology, he said.

In a paper published last year, Mark M Davis, PhD, of Stanford (Calif.) University and his colleagues went beyond standard analysis of the T-cell repertoire and identified residues specific to certain antigens in order to classify T cells based on their likely reactivity. Dr. Reuben and his colleagues collaborated with that group to determine whether T cells were predominantly viral or nonviral.

“If you focus on the normal lung and tumor, you don’t see much of a trend. In some patients there are more viral motifs, and in others are more nonviral motifs, but what was striking was the enrichment for viral motifs that we saw when we focused on the T cells that were shared between the normal lung and tumor,” Dr Reuben said.

In fact, 88% of patients had more viral motifs within their shared T cells vs. only 33% in the normal lung and 30% in tumor.

“So T cells that are shared may be recognizing a combination of shared mutations and/or viruses,” he explained.

How does the T-cell repertoire relate to outcomes?

A focus on the normal lung showed that patients with fewer T cells and lower clonality had better outcomes.

“What does this mean? It suggests that potentially, in these patients, the immune response in the lung is less distracted by outside pathogens and agents unrelated to the tumor, potentially providing the opportunity for a more focused antitumor T-cell response,” Dr. Reuben said, concluding that “T-cell density is higher, but clonality is lower in tumor vs. normal lung, there’s a substantial overlap in the T-cell repertoire between the normal lung and the tumor (including many T cells which may be reactive to shared mutations and/or viruses), and it seems like a more tumor-focused T-cell repertoire in the lung may be associated with improved outcomes.”

In an interview, Dr. Reuben said the findings have certain therapeutic implications, because most current therapies target the T-cell response whether by design or consequence.

“Considering the large proportion of T cells found in lung tumors which are unrelated to tumor responses, expansion of the wrong T cells – whether these target viruses or shared mutations between the normal lung and tumor – could potentially offer no benefit to the patient, because it would likely not contribute to eradicating their tumor,” he explained. “Furthermore, targeting T cells (through checkpoint blockade or TIL therapy) that are reactive to shared mutations could increase the potential for toxicity within these patients. Therefore, a better understanding of the T-cell repertoire in the lung is necessary to increase the specificity and success rates of current immunotherapies.”

Invited discussant, Antoni Ribas, MD, of the University of California, Los Angeles, suggested that the finding of a substantial number of shared T-cells is likely a baseline phenomenon, and that on-therapy biopsies in patients who respond to treatment would better separate and expand the T cells that responded from those that did not.

In fact, Dr. Reuben and his colleagues have expanded their research in this manner.

“We are now studying this phenomenon longitudinally in patients receiving checkpoint blockade to see how these factors evolve over the course of therapy,” he said.

Dr. Reuben reported having no disclosures. Dr. Ribas owns stock in Advaxis, Arcus Ventures, Compugen, CytomX Therapeutics, Five Prime Therapeutics, FLX Bio, and Kite Pharma, and has served as a consultant or advisor for Amgen, Genentech/Roche, Merck, Novartis, and Pierre Fabre.

sworcester@frontlinemedcom.com

SOURCE: Reuben A et al. Clinical Immuno-Oncology Symposium Abstract 140.