Objective Measurement and Clinical Significance of TILs in Non–small Cell Lung Cancer

Kurt A. Schalper; Jason Brown; Daniel Carvajal-Hausdorf; Joseph McLaughlin; Vamsidhar Velcheti; Konstantinos N. Syrigos; Roy S. Herbst; David L. Rimm


J Natl Cancer Inst. 2015;107(3) 

In This Article

Abstract and Introduction


Background Tumor-infiltrating lymphocytes (TILs) are usually measured using subjective methods. Studies suggest that TIL subtypes have independent roles in cancer and that they could support the use of novel immunostimulatory therapies. We simultaneously measured TIL subtypes in non–small cell lung cancer (NSCLC) samples using objective methods and determined their relationship with clinico-pathologic characteristics and survival.

Methods Using multiplexed quantitative fluorescence (QIF), we measured the levels of CD3, CD8, and CD20 in 552 NSCLC from two independent collections represented in tissue microarrays (YTMA79, n = 202 and YTMA140, n = 350). The level of TILs was obtained in different tumor compartments using cytokeratin stain to define tumor cells and 4',6-Diamidino-2-Phenylindole. Association of TILs with clinical parameters was determined using univariate and multivariable analyses. All statistical tests were two-sided.

Results In both NSCLC collections there was a low correlation between the three TIL markers (linear regression coefficients (R2) = 0.19–0.22, P < .001 for YTMA79 and R2 = 0.23–0.32, P < .001 for YTMA140). No consistent association between the level of TIL subtypes and age, sex, smoking history, tumor size, stage, and histology type was found. In univariate analysis, an elevated CD3 or CD8 signal was statistically significantly associated with longer survival in both collections. However, only CD8 was independent from age, tumor size, histology, and stage in multivariable analysis. High CD20 was associated with longer survival in the YTMA79 cohort.

Conclusions Increased levels of CD3 and CD8 + TILs are associated with better outcome in NSCLC, but only CD8 is independent from other prognostic variables. Objective measurement of TIL subpopulations could be useful to predict response or evaluate the local immune effect of anticancer immune checkpoint inhibitors.


Increased tumor-infiltrating lymphocytes (TILs) have been consistently associated with better outcome in diverse human neoplasms, including melanoma, colorectal, and triple- negative carcinomas.[1] Similarly, increased total TILs has been associated with longer survival in non–small cell lung cancer (NSCLC).[2–4] However, studies using chromogenic immunohistochemistry (IHC) to characterize the prognostic effect of specific TIL subtypes in NSCLC have reported conflicting results.[5–8] Diverse studies indicate that in tumors, immune cell subpopulations are strategically distributed within different tissue compartments.[1] Some immune cell subtypes are located predominantly within the tumor core and others in the surrounding stroma or at the invasive front. Moreover, structurally organized and vascularized B cell–rich aggregates of lymphoid cells mimicking secondary lymphoid follicles, termed tertiary lymphoid tissues/structures, have been recognized and positively associated with outcome in human tumors.[1,9–10] This suggests that different immune cell populations have different biological roles in tumor control and that the spatial orientation of immune cell subtypes with respect to cancer cells might be of biological relevance.

Novel immunostimulatory therapies using monoclonal antibodies targeting specific immune checkpoint targets such as CTLA-4, PD-1, and PD-L1 (alone or in combination with chemotherapy) have shown durable responses in a subset of patients with heavily pretreated, advanced NSCLC.[11] There has been considerable interest in these agents because in general they are well tolerated and have demonstrated clinical activity in adenocarcinomas and also squamous tumors currently lacking effective targeted therapies. Preliminary observations indicate that clinical responses to immune checkpoint blockers are associated with elevated tumor levels of immune inhibitory signals, such as IDO-1 and PD-L1, and with increased TILs or specific TIL subtypes.[11–14] Moreover and as expected, side effects associated with such immunotherapies are also commonly associated with increased inflammatory infiltrates in the affected tissue/organ. Therefore, methods for accurate characterization and measurement of TIL subpopulations in the context of cancer tissues and targeted immunotherapies are needed and might help in prognosis assessment and prediction of treatment response.

The traditional way to determine TILs in formalin-fixed paraffin-embedded (FFPE) tissue samples is by pathologist scoring of hematoxylin and eosin (H&E)–stained preparations. However, the information provided by this method is only semiquantitative and subject to considerable interobserver variation. Measurement of TILs using chromogenic IHC provides excellent cellular detail and potentially quantitative output, but is limited to simultaneous assessment of only one to two cellular subtypes without the architectural context. Herein, we simultaneously characterize the TIL subpopulations in FFPE samples from two independent NSCLC cohorts using multiplexed quantitative immunofluorescence (QIF). In addition, we determine the clinico-pathological associations and prognostic value of different TIL subtypes in NSCLC.