MA11.09 Increased Frequency of Bystander T Cells in the Lungs Is Associated with Recurrence in Localized Non-Small Cell Lung Cancer
A. Reuben,J. Zhang,R. Gittelman,S. Chiou,J. Li,J. Fujimoto,C. Behrens,F. Kheradmand,C. Chow,C. Bernatchez,A.K. Eterovic,S. Benzeno,M. Vignali,X. Xifeng Wu,Y. Ye,T. Cascone,R. Herbst,S. Swisher,H. Kadara,C. Moran,P. Scheet,A. Vaporciyan,B. Sepesi,D. Gibbons,H. Robins,P. Hwu,J. Heymach,P. Sharma,J. Allison,J. Lee,M. Davis,I. Wistuba,P.A. Futreal
DOI: https://doi.org/10.1016/j.jtho.2019.08.589
IF: 20.121
2019-01-01
Journal of Thoracic Oncology
Abstract:Non-small cell lung cancer (NSCLC) exhibits a high mutational burden. As a result, patients afflicted by this tumor type experience greater responses to immune checkpoint blockade. This is largely due to the ability of T cells to destroy tumor cells on the basis of antigens recognized by their T cell receptor (TCR). However, the lungs are exposed to carcinogens and pathogens which can also trigger a T cell response distinct from cancer. Therefore, a better understanding of the T cell repertoire in the lungs is needed to improve upon the success of current immunotherapies in NSCLC. We obtained peripheral blood, tumors, and adjacent uninvolved lungs from a cohort of 236 early stage NSCLC patients. Whole exome sequencing, RNA microarray, immunohistochemistry (CD3, CD4, CD8, CD57, CD68, FoxP3, CD45RO, GzmB, PD-1, and PD-L1) and T cell repertoire sequencing were performed in NSCLC patients and lungs from organ donors and COPD patients. Antigen specificity was predicted using the Grouping of Lymphocyte Interactions by Paratope Hotspot (GLIPH) algorithm. Single cell TCR and RNA sequencing as well as sequencing of the virome are underway. Clonality was associated with CD8 T cells (r=0.31; p=0.0003), GzmB (r=0.29; p=0.001) and IFN-γ (r=0.52; p<0.0001) production as well as with tumor mutational burden (r=0.19; p=0.015), HLA-B (r=0.29; p=0.0005) and β2-m expression (r=0.20; p=0.018). Patients with classical EGFR mutations exhibited lower T cell clonality (p=0.003) even after adjustment for TMB, highlighting the impact of this driver mutation on the T cell response. Surprisingly, clonality was higher in the adjacent uninvolved lung than tumor (p<0.0001), suggesting an active antigenic response outside the tumor. Comparison of the composition of the T cell repertoire between the uninvolved lung and tumor revealed 57% of the top 100 T cells in the tumor were also found in the adjacent normal lung, highlighting certain parallels in the ongoing antigenic responses. Deeper analysis suggested that shared T cells may have been reactive against mutations shared between the normal lung and tumor (r=0.23, p=0.028) or viruses (p<0.0001). Accordingly, patients with a more reactive T cell repertoire outside the tumor (i.e. bystanders) exhibited shorter disease-free survival (p=0.036) suggesting these responses against shared mutations and/or viruses may detract from the anti-tumor T cell response. Our findings highlight the importance of understanding the specificity of the T cell repertoire in the lungs in patients with NSCLC treated with immunotherapy. As a high proportion of bystander T cells appear to reside in the lungs, their reactivation could contribute to the impaired responses and/or increased toxicity observed in certain patients with NSCLC treated with immunotherapy.