2691 - Quantitative spatial profiling of lymphocyte-activation gene 3 (LAG-3)/major histocompatibility complex class II (MHC II) interaction in gastric and urothelial tumors

Background

LAG-3, expressed on exhausted T cells, negatively regulates effector T-cell activation and may promote regulatory T-cell activity. MHC II, one of the ligands for LAG-3, is expressed by antigen-presenting cells and is heterogeneously upregulated on tumor cells in a variety of cancers. The LAG-3/MHC II interaction may activate LAG-3 and inhibit antitumor immunity. We performed digital spatial analysis to define the geographic association of LAG-3+ tumor-infiltrating lymphocytes (TILs) with individual MHC II+ and MHC II− tumor cells.

Methods

Bladder and gastric tumor samples (n = 20 each) of varying MHC II+ tumor expression (0–100%) were serially sectioned and stained by IHC for LAG-3, MHC II (human leukocyte antigen-DP, -DQ, and -DR), and Pan-cytokeratin. Slides were scanned via an Aperio AT2 scanner using a 20× objective and whole slide images were digitally aligned and analyzed via HALO software. LAG-3 engagement scores for the density of LAG-3+ TILs (LAG-3-D) and the proportion in close proximity (within 30 µm) to tumor cells (LAG-3-P) were computed for each sample using R software.

Results

MHC II was expressed by at least 1% of tumor cells in 55% of bladder and 70% of gastric samples. LAG-3-D and LAG-3-P within an individual tumor were significantly greater when associated with MHC II+ tumor cells (median [interquartile range] LAG-3-D = 6.53 [1.76, 24.9] cells/mm²; LAG-3-P = 46.7 [30.1, 70.4] % engaged) compared to MHC II− tumor cells (LAG-3-D = 0.616 [0.213, 2.38] cells/mm² [P < 0.001]; LAG-3-P = 17.5 [6.09, 30.1] % engaged [P < 0.001]). Furthermore, we found significantly lower LAG-3/MHC II− tumor engagement by LAG-3-P in gastric compared to urothelial tumors (P < 0.001).

Conclusions

Digital spatial analysis of tumor cells and TILs in the tumor microenvironment is feasible without multiplex assays and can capture cell–cell relationships in tumors with heterogeneous MHC II staining. These data suggest preferential localization of LAG-3-expressing TILs to MHC II+ tumor cells within a proximity that may allow engagement and activation of LAG-3 and help define the importance of spatial analysis in predictive biomarker development for immunotherapy.

Clinical trial identification

Editorial acknowledgement

Editorial assistance was provided by Kathryn Woods, PhD, of Complete HealthVizion, funded by Bristol-Myers Squibb.

Legal entity responsible for the study

Bristol-Myers Squibb.

Funding

Bristol-Myers Squibb.

Disclosure

C.V. Hedvat: Shareholder / Stockholder / Stock options: Bristol-Myers Squibb; Full / Part-time employment: Bristol-Myers Squibb. G. Lee: Shareholder / Stockholder / Stock options: Bristol-Myers Squibb; Full / Part-time employment: Bristol-Myers Squibb. V. Baxi: Full / Part-time employment: Bristol-Myers Squibb. K. Dziuba: Full / Part-time employment: Bristol-Myers Squibb. D. Locke: Shareholder / Stockholder / Stock options: Bristol-Myers Squibb; Full / Part-time employment: Bristol-Myers Squibb. B. Li: Shareholder / Stockholder / Stock options: Bristol-Myers Squibb; Full / Part-time employment: Bristol-Myers Squibb. R. Edwards: Shareholder / Stockholder / Stock options: Bristol-Myers Squibb; Full / Part-time employment: Bristol-Myers Squibb.