Compartmental Analysis of T-cell Clonal Dynamics as a Function of Pathologic Response to Neoadjuvant PD-1 Blockade in Resectable Non-Small Cell Lung Cancer

Abstract

Purpose: Neoadjuvant PD-1 blockade is a promising treatment for resectable non-small cell lung cancer (NSCLC), yet immunologic mechanisms contributing to tumor regression and biomarkers of response are unknown. Using paired tumor/blood samples from a phase II clinical trial (NCT02259621), we explored whether the peripheral T-cell clonotypic dynamics can serve as a biomarker for response to neoadjuvant PD-1 blockade.

Experimental design: T-cell receptor (TCR) sequencing was performed on serial peripheral blood, tumor, and normal lung samples from resectable NSCLC patients treated with neoadjuvant PD-1 blockade. We explored the temporal dynamics of the T-cell repertoire in the peripheral and tumoral compartments in response to neoadjuvant PD-1 blockade by using the TCR as a molecular barcode.

Results: Higher intratumoral TCR clonality was associated with reduced percent residual tumor at the time of surgery, and the TCR repertoire of tumors with major pathologic response (MPR; <10% residual tumor after neoadjuvant therapy) had a higher clonality and greater sharing of tumor-infiltrating clonotypes with the peripheral blood relative to tumors without MPR. Additionally, the posttreatment tumor bed of patients with MPR was enriched with T-cell clones that had peripherally expanded between weeks 2 and 4 after anti-PD-1 initiation and the intratumoral space occupied by these clonotypes was inversely correlated with percent residual tumor.

Conclusions: Our study suggests that exchange of T-cell clones between tumor and blood represents a key correlate of pathologic response to neoadjuvant immunotherapy and shows that the periphery may be a previously underappreciated originating compartment for effective antitumor immunity.See related commentary by Henick, p. 1205.

Conflict of interest statement

Disclosure of Potential Conflicts of Interest

J.M.T. receives research funding from BMS, and is a consultant/advisory board member for Bristol-Myers Squibb (BMS), Merck, and MedImmune/AstraZeneca. J. C. is a consultant for AstraZeneca, BMS, and Genentech and received research funding from AstraZeneca, BMS, Genentech, and Merck. T. M. is a consultant for Leap Therapeutics, Immunos Therapeutics and Pfizer, and co-founder of Imvaq therapeutics; has equity in Imvaq therapeutics; reports grants from BMS, Surface Oncology, Kyn Therapeutics, Infinity Pharmaceuticals, Peregrine Pharmeceuticals, Adaptive Biotechnologies, Leap Therapeutics and Aprea; is inventor on patent applications related to work on oncolytic viral therapy, alphavirus-based vaccines, neo-antigen modeling, CD40, GITR, OX40, PD-1 and CTLA-4. J.N. receives research funding from Merck and AstraZeneca, is a consultant/advisory board member for BMS, Roche/Genentech, and AstraZeneca, and has received honoraria from AstraZeneca and BMS. V.A. receives research funding from BMS. M.D.H. has received research funding from BMS; is a paid consultant to Merck, BMS, AstraZeneca, Roche/Genentech, Janssen, Nektar, Syndax, Mirati, and Shattuck Lab; has received travel support/honoraria from AstraZeneca and BMS; and a patent has been filed by MSK related to the use of tumor mutation burden to predict response to immunotherapy (PCT/US2015/062208), which has received licensing fees from PGDx. P.M.F receives research funding from AZ, BMS, Corvus, Kyowa, and Novartis and is a consultant/advisory board member for Abbvie, AstraZeneca, BMS, Boehringer, EMD Serono, Iniviata, Janssen, Lilly, Merck, and Novartis. J.R.B receives research funding (to institution) from BMS, Merck, AstraZeneca, and is on consulting/advisory boards of BMS (uncompensated), Merck and Genentech. D.M.P. reports grant and patent royalties through institution from BMS, grant from Compugen, stock from Trieza Therapeutics and Dracen Pharmaceuticals, and founder equity from Potenza; being consultant for Aduro Biotech, Amgen, Astra Zeneca (Medimmune/Amplimmune), Bayer, DNAtrix, Dynavax Technologies Corporation, Ervaxx, FLX Bio, Rock Springs Capital, Janssen, Merck, Tizona, and Immunomic- Therapeutics; being on the scientific advisory board of Five Prime Therapeutics, Camden Nexus II, WindMil; being on the board of director for Dracen Pharmaceuticals. V.E.V. is a founder of Personal Genome Diagnostics, a member of its Scientific Advisory Board and Board of Directors, and owns Personal Genome Diagnostics stock, which are subject to certain restrictions under university policy. V.E.V. is an advisor to Takeda Pharmaceuticals. Within the last five years, V.E.V. has been an advisor to Daiichi Sankyo, Janssen Diagnostics, and Ignyta. K.N.S. has received travel support/honoraria from Neon Therapeutics and Illumina. The terms of these arrangements are managed by Johns Hopkins University in accordance with its conflict of interest policies.

©2019 American Association for Cancer Research.

Figures

Figure 1.. Clonality of the TCR repertoire and association with the anti-tumor response.

A. Flow chart of the phase 2 clinical trial and biospecimen collection, along with correlative studies performed at each timepoint. B. Correlation between productive clonality in the tumor bed at the time of resection (after anti-PD-1) with the number of non-synonymous sequence alterations (Spearman’s rho, 0.70; P=0.025). Each patient is represented by a black dot (n=10). The blue line indicates the linear regression line, and the gray area indicates the upper and lower boundaries of the 95% confidence interval. C. Correlation between productive clonality in the post-treatment tumor bed and the percent residual tumor (n=10, Spearman’s rho, −0.65; P=0.041). The blue line indicates the linear regression line, and the gray area indicate the upper and lower boundaries of the 95% confidence interval. Productive clonality: Clonality as determined by using the productive amino acid (AA) sequence of the CDR3. D. Productive clonality of the TCR repertoire in the post-treatment tumor bed and normal lung for patients with major pathologic response (MPRs; blue) and without MPR (non-MPR; red). E, Occupied clonal space (the total frequency among all intratumoral T cells) of ITCs according to their percent rank (top 1% ranked ITCs vs top 1–2% ranked ITCs vs top 2–5% ranked ITCs vs >5% ranked ITCs) in the post-treatment tumor bed of each patient. F. Comparison of the clonal space occupied by ITCs segregated by frequency-ranks between MPRs (n=9) and non-MPRs (n=5).

Figure 2.. Dynamics of ITCs across tissue compartments and in longitudinal peripheral blood.

A. Proportion of ITCs shared between pre-treatment blood and normal lung, comparing non-top 1% ITCs (blue) and top 1% ITCs (red) (n=14). B. Top 1% ITCs shared between pre-treatment blood and resected normal lung, comparing MPRs (blue) and non-MPRs (red) (n=14). C. Proportion of top 1% ITCs by their shared compartment (pre-treatment blood+resected normal lung, pretreatment blood, resected normal lung, and tumor resident only). D. Clonal space of top 1% ITCs by shared compartment between MPRs and non-MPRs (n=14). The proportion of top 1% ITCs that are shared with the pre-treatment peripheral blood and resected normal lung is higher in MPRs vs. non-MPRs. However, an inverse correlation was observed for top 1% frequency-ranked ITCs that were only found in the tumor bed. E. Temporal dynamics of the ITCs in the longitudinal blood (n=14). The percent change of top 1% ITCs shared with the peripheral repertoire over time was calculated as compared to pre-treatment blood. Data show the mean +/− standard error for all patients (left panel) and by MPRs vs. non-MPRs (right panel) for top 1% ITCs (red) and non-top 1% ITCs (blue). F. The pattern and degree of peripheral remodeling at week 2 and week 4 following treatment initiation for all patients. The fold change of each clonotype was calculated, and the means of the fold change were plotted on a logarithmic scale (log10), stratified by the abundance ranks in the post-treatment tumor bed (n=14). The left and right panels show the dynamic magnitude in clones that underwent peripheral expansion and contraction, respectively. *: p

Fig. 3.. Differentially dynamic clones and their association with tumor infiltration and tumor accumulation.

A. The proportion of dynamic clones vs non-dynamic clones that infiltrated tumor bed upon PD-1 blockade in MPR and non-MPR (n=12). B. The total proportion of clones in the tumor is shown for significantly contracted clones, significantly expanded clones, and non-dynamic clones at week 2 (dark grey) and week 4 (light grey) after treatment initiation (n=12). C. Association of intratumoral clonal space of peripherally expanded clones and percent residual tumor by time window (pre-treatment~W2, Spearman’s rho=−0.35, W2~W4 Spearman’s rho=−0.62). Each patient is represented by a black dot. The blue line indicates the robust linear regression line (fitted using R function ‘rlm’ from ‘MASS’ package based on M estimator), and the gray area indicates the upper and lower boundaries of the 95% confidence interval (n=12). D. The clonal space occupied by ITCs with different peripheral dynamic patterns (expanded, contracted, non-dynamic or tumor resident-only) at 2–4 weeks after immunotherapy initiation is shown for MPRs and non-MPRs (n=12). E. The clonal space occupied by normal lung T cells with differential patterns in the periphery is shown for MPRs and non-MPRs at 2–4 weeks after treatment initiation (n=12). F. The proportion of top 1% ITCs and their peripheral dynamic pattern at 2–4 weeks after treatment initiation for MPRs and non-MPRs (n=12). G. Proposed schema of peripheral activation of the anti-tumor repertoire homing back to the tumor bed. ns, not significant. *: p

Figure 4.. PD1 sorting revealed differential patterns of intra-clonal PD1+ positivity between complete pathologic responder and non-responder.

A. TCRseq was performed on sorted PD-1 positive and negative CD4+ and CD8+ peripheral blood T cells obtained from patient MD01–005 (0% residual tumor) prior to treatment (left panel). The frequency of clonotypes detected in the PD-1+ population is shown in serial peripheral blood (center) and biopsied and resected tissues (right) is shown. All 5 differentially expanded ITCs identified in this patient were detected in the PD1+ sorted population. B. Pie charts showing the intra-clonal PD-1 positivity (the proportion of total reads of each clone that was detected in the PD-1+ vs. PD-1− sorted population) for the 5 peripherally expanded, top 1% ITCs is shown. C. TCRseq was performed on sorted PD-1 positive and negative CD4+ and CD8+ peripheral blood T cells obtained from patient MD01–024 (100% residual tumor) prior to treatment (left panel). The frequency of clonotypes detected in the PD-1+ population is shown in serial peripheral blood (center) and resected tissues (right). D. Pie charts showing the intraclonal PD-1 positivity for peripherally expanded, top 1 % ITCs. E. The read proportion of top 1% ITCs in the normal lung and tumor bed is shown for patient MD01–005 and patient MD01–024. Comparisons between normal lung and tumor bed were evaluated using the Wilcoxon signed-rank test.