The effect of neoadjuvant therapy on PD-L1 expression and CD8+lymphocyte density in non-small cell lung cancer

Abstract

PD-L1 expression is the routine clinical biomarker for the selection of patients to receive immunotherapy in non-small cell lung cancer (NSCLC). However, the application and best timing of immunotherapy in the resectable setting is still under investigation. We aimed to study the effect of chemotherapy on PD-L1 expression and tumor infiltrating lymphocytes (TILs), which is to date still poorly understood. Our retrospective, single-centre neoadjuvant cohort comprised 96 consecutive patients with NSCLC resected 2000–2016 after neoadjuvant therapy, including paired diagnostic chemo-naïve specimens in 53 cases. A biologically matched surgical cohort of 114 primary resected cases was included. PD-L1 expression, CD8 + TILs density and tertiary lymphoid structures were assessed on whole slides and correlated with clinico-pathological characteristics and survival. Seven/53 and 12/53 cases had lower respectively higher PD-L1 expressions after neoadjuvant therapy. Most cases (n = 34) showed no changes in PD-L1 expression, the majority of these harboring PD-L1 < 1% in both samples (21/34 [61.8%]). Although CD8 + TILs density was significantly higher after chemotherapy (p = 0.031) in resections compared to diagnostic biopsies, this might be due to sampling and statistical bias. No difference in PD-L1 expression or CD8 + TILs density was detected when comparing the neoadjuvant and surgical cohort. In univariable analyses, higher CD8 + TILs density, higher numbers of tertiary lymphoid structures but not PD-L1 expression were significantly associated with longer survival. Increased PD-L1 expression after neoadjuvant chemotherapy was not significantly associated with shorter 5-year survival, but the number of cases was very low. In multivariable analysis, only pT category and age remained independent prognostic factors. In summary, PD-L1 expression was mostly unchanged after neoadjuvant chemotherapy compared to diagnostic biopsies. The sample size of cases with changed PD-L1 expression was too small to draw conclusions on any prognostic value.

Longitudinal analysis of PD-L1 expression in patients with relapsed NSCLC

“…5 One possible explanation is the fact that PD-L1 expression in NSCLC is not static but seems to undergo dynamic changes. [5][6][7][8][9] Previous studies found that PD-L1 expression may be influenced by several factors including Interferon-γ and other cytokines release, [10][11][12] activation of oncogenic drivers 6 13 and specific drugs. 14 Some studies even showed that oncological treatment may alter PD-L1 expression.…”

Rethinking the role of biomarkers for operable non-small cell lung carcinoma: an effective collaboration with artificial intelligence algorithms

“…In this issue of Modern Pathology, Zens et al 8 provide further insights into the role of predictive immune biomarkers in operable NSCLC. Specifically, they investigated the effect of neoadjuvant chemotherapy on PD-L1 expression and CD8+ TILs density in a retrospective cohort of NSCLC patients.…”

“…Considering their results (manual scoring of clone SP263 on paired samples) and their review of the literature, the authors conclude that it seems unlikely that neoadjuvant chemotherapy induces PD-L1 expression. In view of the potential importance of this conclusion, it must be emphasized that a broadly held consensus has yet to emerge (summarized in Table 1 by Zens et al 8 ).…”

“…Is there increased CD8+ TILs density after chemotherapy?. Zens et al 8 evaluated CD8+ TILs per mm 2 using a semi-automated approach with an open-source AI software. Regions of interest were manually annotated and only peritumoral stromal TILs were analyzed.…”

“…Despite the lack of harmonization, AI solutions are paradoxically re-focusing digital quantification of TILs as a predictive biomarker for ICIs in patients with advanced NSCLC, with a special interest on the spatial distribution of lymphocytes and their relationship with tumor cells [14][15][16][17][18] . Therefore, the data presented by Zens et al 8 should help implement reproducible annotation protocols for AI in operable NSCLC. This is not a trivial task, as the number of questions pathologists are asking is increasing: how many fields of view do I need to score?, shall I consider only peritumoral stromal TILs?, why not including intra-epithelial TILs?, how far from the tumor margin is too far?, why are acinar patterns more time-consuming?…”