The T-cell repertoire in NSCLC: Therapeutic implications

SAN FRANCISCO – An analysis of the T-cell repertoire in nearly 400 patients with stage I-III non–small cell lung cancer (NSCLC) suggests that patients with a more tumor-focused repertoire have better outcomes.

The findings, which suggest that patients with fewer T cells and with lower clonality in tumor-adjacent normal lung tissue fare better, could have implications for the use of tumor-infiltrating lymphocyte (TIL) therapy and checkpoint blockade – and possibly other therapies – in these patients, according to Alexandre Reuben, PhD, of the University of Texas MD Anderson Cancer Center, Houston.

Studying the T-cell repertoire in the lung can be rather “messy,” because many T cells may be responding to the outside environment, but comparing findings in the normal lung with those in tumor tissue helps to clarify things, Dr. Reuben said at the ASCO-SITC Clinical Immuno-Oncology Symposium.

Why study the T-cell repertoire?

The successes seen with immune checkpoint blockade in recent years are largely a result of the ability of these therapies to enhance the antitumor T-cell response. Interestingly, checkpoint blockade works better in tumor types like lung cancer that have a high mutational load, Dr. Reuben said, explaining that this is largely attributable to the ability of the mutations to increase tumor immunogenicity through generation of tumor-specific antigens, which can then be targeted by the T-cell response.

This relationship between the mutational and neoantigen burden and patient outcomes has been described, but the role of the T-cell repertoire and how it relates to patient outcomes is less clear.

Hypothesizing that T-cell repertoire would be associated with survival in patients with NSCLC, he and his colleagues collected peripheral blood, normal lung, and tumor tissue, and performed T cell–receptor sequencing, among other analyses, in 398 patients.

“T cells recognize antigens through their T-cell receptor, as a result of which they undergo clonal expansion. Therefore, by sequencing the variable region of the T-cell receptor, one can gain insight into the T cells that are responding within the sample, as well as the overall T-cell repertoire,” he explained.

This provides information on T-cell density and richness, and thus on the diversity of the T-cell repertoire, he said, adding that it is possible to go beyond that and plot T cells based on their frequency in a sample to study clonality.

Since these T cells tend to expand clonally as a result of activation, uneven distribution would be associated with a reactive T-cell repertoire, as described by a high T-cell clonality.

An assessment to determine how the T-cell repertoire relates to clinical characteristics and response revealed a few interesting correlations. For example, adenocarcinomas tended to be more densely infiltrated than squamous cell carcinomas, smaller tumors were also more densely infiltrated by T cells than were their larger counterparts, and in smokers the T-cell repertoire appeared much more reactive than in nonsmokers.

“But ultimately, we didn’t see the [direct correlations between the T-cell repertoire and outcomes] we were hoping to see,” Dr. Reuben said, noting that this could be because of environmental influences.

“Obviously the lung is exposed to the outside environment, which could be masking some of the antitumor T-cell responses we were hoping to study,” he explained. “So we used a more holistic approach, integrating the peripheral blood and normal lung with tumor repertoire going forward.”

T cells in normal lung vs. tumor

Measuring the proportion of the T-cell repertoire that is shared in peripheral blood vs. normal lung and vs. tumor tissue showed that there is very little in common between them.

“However, when you compare the normal lung to the tumor, there’s much more homology in the T-cell repertoire,” he said, noting that, given the T-cell expansion resulting from antigenic stimulation, focusing on the most dominant cells in a sample highlights those most likely to be responding to antigens. “When we did that ... we saw even more of an enrichment in the homology between the normal lung and tumor T-cell repertoire, suggesting certain parallels in the ongoing immune responses across both these compartments.”

Further, T-cell density and diversity were actually higher in the tumor than in the normal lung in about two-thirds of patients, he said.

“However, surprisingly ... clonality appears to be much higher in the normal lung than in the tumor,” he added, noting that this was the case in about 75% of patients.

These findings raise three key questions:

Why is clonality higher in the normal lung?

T cells are not confined to a specific part of the host and are free to circulate, Dr. Reuben said.

“However, the closer you get to a site of inflammation, the higher the enrichment for T cells that are relevant to that specific site of inflammation, so you can use these statistical methods to enrich for T cells that are more relevant and try to subtract out T cells that are simply circulating through the organ,” he noted.

He and his colleagues used these methods and compared both normal lung and tumor to the peripheral blood, focusing only on clones that were statistically enriched in these two compartments “to really eliminate a lot of the background that may have been caused by the low-frequency T cells in these samples.”

When you look at the lung enriched T-cell repertoire between normal lung vs. tumor, the homology increases quite significantly, suggesting that by subtracting these T cells that are circulating through the host and not likely relevant to the antigenic response, you’re increasing the homology and further highlighting some of the aforementioned parallels in the ongoing immune responses between both sites, he said.

“Now if you look at clonality, there’s really no clear trend ... in the total T-cell repertoire or the enriched repertoire focusing on the normal lung, but if you look at the tumor, there’s a trend toward increased clonality in all patients – to the extent where you no longer see a difference in clonality between the normal lung and tumor, suggesting that this enrichment is allowing us to focus increasingly on T cells relevant to the antitumor response,” he added.

T-cell clonality is highly reliant on the ability of T cells to expand as a result of antigenic stimulation, and immune profiling showed that programmed cell death–1 (PD-1) and programmed death–ligand 1 (PD-L1) were higher within the tumor, suggesting that there is some dysfunction on both sides of this interaction, which could also explain the lower clonality originally seen within the tumor, he said.