High PD-L1 Expression Correlates with an Immunosuppressive Tumour Immune Microenvironment and Worse Prognosis in ALK-Rearranged Non-Small Cell Lung Cancer

High tumour programmed cell death-ligand 1 (PD-L1) expression is associated with poor progression-free survival (PFS) after tyrosine kinase inhibitor (TKI) therapy in ALK-rearranged non-small cell lung cancer (NSCLC). However, the characteristics of the tumour microenvironment (TME) and their prognostic values in ALK-rearranged NSCLC are unknown. Here, we collected tumour tissues from pretreated ALK-rearranged NSCLC patients, immunohistochemical staining was used to assess PD-L1 expression, and tumour-infiltrating immune cells were determined via multiplex immunofluorescence staining (mIF). Our data showed that the median values of PFS for the high PD-L1 group and low PD-L1 group who received ALK-TKI treatment were 4.4 and 16.4 months, respectively (p = 0.008). The median overall survival (OS) of the two groups was 24.0 months and not reached, respectively (p = 0.021). Via univariate and multivariate analyses, a high PD-L1 expression and a worse ECOG PS were determined to be independent prognostic factors of OS (HR = 3.35, 95% CI: 1.23–9.11, p = 0.018; HR = 6.42, 95% CI: 1.45–28.44, p = 0.014, respectively). In addition, the high PD-L1 group had increased Tregs and exhausted CD8+ T cells in both the tumour and stroma (all p < 0.05). High PD-L1 expression was an adverse predictive and prognostic biomarker for ALK-rearranged NSCLC. The characteristics of the TME in patients with high PD-L1 expression were shown to have an immunosuppressive status.


Introduction
Lung cancer is currently the main leading cause of cancer-related deaths worldwide [1].Non-small cell lung cancer (NSCLC) comprises approximately 85% of lung cancers overall [2].Approximately 3-7% of patients with NSCLC harbour rearrangements in the anaplastic lymphoma kinase (ALK) gene, a potent oncogenic driver [3], and ALK tyrosine kinase inhibitors (TKIs) have markedly improved the survival of these patients.However, not all patients with ALK-rearranged NSCLC can benefit from ALK-TKIs, and it is important to identify these patients to improve their survival outcomes [4].
Programmed cell death-ligand 1 (PD-L1) in tumours was identified as a positive predictive biomarker for advanced NSCLC patients treated with immune checkpoint inhibitors (ICIs) [5].PD-L1 expressed on tumour cells interacts with programmed cell death protein 1 (PD-1), which is expressed on cytotoxic T cells to suppress T cell activation and growth, and negatively regulates the antitumour immune response.Several studies have shown that high tumour PD-L1 expression is linked to worse progression-free survival (PFS) in response to ALK-TKIs [6][7][8][9][10].However, the association between PD-L1 expression and prognosis in ALK-rearranged NSCLC has not been well studied.A prior study showed that the status of PD-L1 expression alone did not influence the OS of ALK-positive NSCLC patients [11].However, another recent study found that high baseline PD-L1 expression was associated with shorter overall survival (OS) in ALK-rearranged lung adenocarcinoma [10].Hence, the relationship between PD-L1 expression and prognosis in ALK-rearranged NSCLC remains unclear.
Previous have studies stratified the tumour microenvironment (TME) based on tumourinfiltrating lymphocytes (TILs) and PD-L1 expression, in which the PD-L1+TILs+ TME is most likely to benefit from ICI therapy [12,13].Mazzaschi et al. observed that resected NSCLC patients with high levels of CD8+ lymphocytes lacking PD-1 had a longer OS [14].Yang et al. reported that tumour PD-L1 expression and increased CD4+ T cell infiltration were independent predictors of better OS in SCLC [15].Based on this evidence, it appears that the TME is associated with prognosis.However, the association between the TME and prognosis in ALK-rearranged NSCLC remains poorly understood.Yang et al. performed multiplex immunofluorescence staining (mIF) for three cases, and found that cases with high PD-L1 expression (PD-L1 TPS 80%) showed less infiltration of CD20+ B cells, CD8+ T cells, and CD4+ T cells [8].Most ALK-rearranged NSCLC with high tumour PD-L1 expression (TPS ≥ 50%) seems to lack high levels of TILs [12,16].However, the difference in the TME between patients with high and low PD-L1 expression is unclear.Hence, there is a lack of studies evaluating the TME of tumours with high PD-L1 expression in ALK-rearranged NSCLC and the impact of the TME on the prognosis of these patients.
Therefore, we evaluated the association between baseline PD-L1 expression and prognosis in patients with ALK-rearranged advanced NSCLC, performed a comprehensive analysis of the TME of patients with a high PD-L1 expression, and analysed the impact of the TME on prognosis.

Patients
We reviewed the medical records of 4261 consecutive patients diagnosed with stage IIIB-IVB NSCLC who were tested for ALK rearrangement and PD-L1 expression using immunohistochemical staining at the West China Hospital between December 2014 and August 2020.

Immunohistochemical Staining of ALK
Tumour ALK status was determined via immunohistochemistry (IHC) staining using the ALK (D5F3) assay with a Ventana BenchMark XT automated staining instrument (Ventana, AZ, USA), according to the manufacturer's instructions.

Immunohistochemical Staining of ALK
Tumour ALK status was determined via immunohistochemistry (IHC) staining using the ALK (D5F3) assay with a Ventana BenchMark XT automated staining instrument (Ventana, AZ, USA), according to the manufacturer's instructions.

Tissue Imaging and Analysis
The stained slides were scanned using the Mantra System (PerkinElmer, Waltham,

Tissue Imaging and Analysis
The stained slides were scanned using the Mantra System (PerkinElmer, Waltham, MA, USA).Multispectral images were separated using a spectral library established from a single stained tissue image for each reagent using inForm software (PerkinElmer, Waltham, MA, USA).For batch analysis, inForm software can automatically detect and segment specific tissue compartments into the tumour parenchyma and stroma based on the epithelial cell marker pancytokeratin and improve phenotyping using DAPI-based cell segmentation.The intratumour region was defined as malignant cell nests, and the stroma was characterised by the fibrous tissue present between malignant cells.For each staining marker, a positive threshold within the nucleus or cytoplasm was set, and the entire image set was analysed.The generated data included the count, density and percentage of positive cells for each fluorescent marker.The single-cell-level data exported from inForm were further processed using the R statistical programming language (Version 3.1).Using Olympus, 2-3 regions of interest (ROIs) per slide were selected when each tumour tissue was scanned using low magnification (4×).The invasive margin (IM) was defined as the area within 250 µm on each side of the border separating the host tissue from malignant tissue, and then two distinct regions were obtained: IM tumour (IM-T) and IM stroma (IM-S).

Evaluation of Immune Cell Spatial Distribution
To evaluate the spatial relationships between immune cells and tumour cells, we conducted spatial analysis, including calculating the average number of immune cells distributed within a 30 µm radius from the nuclear centre of any given tumour cell and using "nearest neighbour distance" (NND) analysis; this was performed by calculating the distance (µm) from each tumour cell to its nearest neighbour immune cell.We selected a radius of 30 µm for study, as it has been previously suggested that this is a reasonable physiologic distance to identify immune cells that may be able to engage in direct and cell-to-cell interactions with tumour cells [19].

Treatment Evaluation
Clinical response to ALK-TKIs was evaluated based on a computed tomography (CT) scan follow-up every 3 months during treatment according to the Response Evaluation Criteria in Solid tumours, version 1.1 (RECIST 1.1) [20].PFS was defined as the period from the start of ALK-TKI treatment to disease progression or death.OS was defined as the interval between diagnosis and death or the last follow-up.The objective response rate (ORR) was defined as the percentage of patients achieving complete response (CR) or partial response (PR).The disease control rate (DCR) was defined as the percentage of patients achieving CR, PR or stable disease (SD).

Statistical Analysis
The chi-squared test was used to analyse ORR and DCR according to PD-L1 expression.Kaplan-Meier curves and the log-rank test were used to analyse PFS and OS according to PD-L1 expression and to provide median survival times based on immune cell exposures.The prognostic factors of PFS and OS were analysed via univariate and multivariate Cox regression analyses.Variables with p < 0.05 in the univariate analysis were included in the multivariate Cox regression model.The hazard ratio (HR) and 95% confidence interval (CI) were calculated for all variables in the regression model.The Mann-Whitney U test was used to examine the composition, density, location and spatial distribution of immune cells according to PD-L1 expression.The Spearman correlation test was used to assess the correlations between immune cells.The value of the coefficient ranged from −1 to 1, and an absolute value of the correlation coefficient of ">0.60" was defined as a strong correlation in our study.The optimal cut-off values of the immune cell percentage were obtained using the surv_cutpoint function.All tests were two sided, and statistical significance was set at p < 0.05.All statistical analyses were performed using SPSS Statistics (version 25.0), GraphPad Prism (version 9.0) and R version 4.1.2.

Features of the Tumour Immune Microenvironment in Patients with High PD-L1 Expression
Of the 52 patients, 18 patients (10 PD-L1hi and 8 PD-L1lo) had sufficient tissue available for mIF analysis (Figure 3A,B).The percentages of Tregs and exhausted CD8+ T cells in the PD-L1hi group were higher than those in the PD-L1lo group (2.55 × 10 −3 vs. 6.60 × 10 −5 , p = 0.005; 1.31 × 10 −3 vs. 1.65 × 10 −4 , p = 0.005, respectively), while the percentages of CD3+ T cells, B cells, NK cells and macrophages were not significantly different between the two groups (all p > 0.05) (Figure 3C-J).The segmentation of tumours from the tumour nest and stroma showed that the percentages of Treg and exhausted CD8+ T cells were higher in the PD-L1hi group than in the PD-L1lo group in both the tumour and stroma (all p < 0.05) (Figure 3K-L).The stromal Treg/CD8 ratio was higher in patients with high PD-L1 expression than in patients with low PD-L1 expression (12.0% vs. 0.01%, p = 0.015) (Figure 3M).
No significant difference was observed in the percentages of Treg or exhausted CD8+ T cells between the PD-L1hi group and PD-L1lo group in the invasive margin (all p > 0.05, Figure 4A,B).We performed an NND analysis measuring the distance from each tumour cell to the closest Treg and exhausted CD8+ T cells (Figure 4C) and calculated the numbers of Treg and exhausted CD8+ T cells within a 30 µm radius of tumour cells (Figure 4D).Fourteen patients (8 PD-L1hi and 6 PD-L1lo) had computable Treg and exhausted CD8+ T cells in NND analysis, and Tregs and exhausted CD8+ T cells in the PD-L1hi group were located closer to tumour cells than those in the PD-L1lo group, but the difference was not statistically significant (p = 0.181 and 0.081, respectively) (Figure 4E,F).However, the numbers of Tregs and exhausted CD8+ T cells around tumour cells within a 30 µm range were significantly higher in the PD-L1hi group than in the PD-L1lo group (4.55 × 10 −2 vs. 2.33 × 10 −4 , p = 0.003; 1.04 × 10 −2 vs. 0, p = 0.001, respectively) (Figure 4G,H).In the PD-L1hi group, we found a strong correlation between intratumoural exhausted CD8+ T cells and stromal B cells (r = −0.762,p = 0.037) (Figure 4I).In addition, a strong correlation between intratumoural exhausted CD8+ T cells and intratumoural B cells was also observed (r = −0.833,p = 0.015) (Figure 4J).
Biomolecules 2023, 13, x FOR PEER REVIEW 9 of 17 (all p < 0.05) (Figure 3K-L).The stromal Treg/CD8 ratio was higher in patients with high PD-L1 expression than in patients with low PD-L1 expression (12.0% vs. 0.01%, p = 0.015) (Figure 3M).No significant difference was observed in the percentages of Treg or exhausted CD8+ T cells between the PD-L1hi group and PD-L1lo group in the invasive margin (all p > 0.05, Figure 4A,B).We performed an NND analysis measuring the distance from each tumour cell to the closest Treg and exhausted CD8+ T cells (Figure 4C) and calculated the numbers of Treg and exhausted CD8+ T cells within a 30 µm radius of tumour cells (Figure 4D).Fourteen patients (8 PD-L1hi and 6 PD-L1lo) had computable Treg and exhausted CD8+ T cells in NND analysis, and Tregs and exhausted CD8+ T cells in the PD-L1hi group were located closer to tumour cells than those in the PD-L1lo group, but the difference was not statistically significant (p = 0.181 and 0.081, respectively) (Figure 4E,F).However, the numbers of Tregs and exhausted CD8+ T cells around tumour cells within a 30 µm range were significantly higher in the PD-L1hi group than in the PD-L1lo group (4.55 × 10 −2 vs. 2.33 × 10 −4 , p = 0.003; 1.04 × 10 −2 vs. 0, p = 0.001, respectively) (Figure 4G,H).In the PD-L1hi group, we found a strong correlation between intratumoural exhausted CD8+ T cells and stromal B cells (r = −0.762,p = 0.037) (Figure 4I).In addition, a strong correlation between intratumoural exhausted CD8+ T cells and intratumoural B cells was also observed (r = −0.833,p = 0.015) (Figure 4J).

PFS and OS of ALK-Rearranged NSCLC Based on TME
Of the 18 patients, 16 patients received crizotinib, and 2 patients received alectinib.To evaluate the clinical response of the patients to ALK-TKIs based on TME status, we performed survival analysis in patients treated with crizotinib.Kaplan-Meier curves revealed that patients with a high Treg level had a shorter PFS and OS than those with a low Treg level (9.8 vs. 35.5 months, p = 0.008; 24.0 months vs. NR, p = 0.007, respectively) (Figure 5A,C).In addition, patients with higher exhausted CD8+ T cell levels had a worse PFS than those with low exhausted CD8+ T cell levels (10.2 vs. 51.1 months, p = 0.005) (Figure 5B,D).

PFS and OS of ALK-Rearranged NSCLC Based on TME
Of the 18 patients, 16 patients received crizotinib, and 2 patients received alectinib.To evaluate the clinical response of the patients to ALK-TKIs based on TME status, we performed survival analysis in patients treated with crizotinib.Kaplan-Meier curves revealed that patients with a high Treg level had a shorter PFS and OS than those with a low Treg level (9.8 vs. 35.5 months, p = 0.008; 24.0 months vs. NR, p = 0.007, respectively) (Figure 5A,C).In addition, patients with higher exhausted CD8+ T cell levels had a worse PFS than those with low exhausted CD8+ T cell levels (10.2 vs. 51.1 months, p = 0.005) (Figure 5B,D).High Treg levels were present in 71.4% (5/7) of patients within the PD-L1hi group and in 22.2% (2/9) of patients within the PD-L1lo group.In addition, 42.9% (3/7) of cases belonging to the PD-L1hi group showed a high level of exhausted CD8+ T cells compared with 33.3% (3/9) of the PD-L1lo group.Our data revealed that patients with low exhausted CD8+ T cell levels showed trends towards an association with a longer PFS than patients with high exhausted CD8+ T cell levels in the PD-L1lo group (26.4 vs. 51.1 months, p = 0.077) (Figure 6C).Although not reaching a statistically significant impact on OS (Supplementary Figure S1), we found that patients with a high PD-L1 expression, low Treg levels and high exhausted CD8+ T cell levels exhibited the longest OS, at 51.4 months (Supplementary Table S1).High Treg levels were present in 71.4% (5/7) of patients within the PD-L1hi group and in 22.2% (2/9) of patients within the PD-L1lo group.In addition, 42.9% (3/7) of cases belonging to the PD-L1hi group showed a high level of exhausted CD8+ T cells compared with 33.3% (3/9) of the PD-L1lo group.Our data revealed that patients with low exhausted CD8+ T cell levels showed trends towards an association with a longer PFS than patients with high exhausted CD8+ T cell levels in the PD-L1lo group (26.4 vs. 51.1 months, p = 0.077) (Figure 6C).Although not reaching a statistically significant impact on OS (Supplementary Figure S1), we found that patients with a high PD-L1 expression, low Treg levels and high exhausted CD8+ T cell levels exhibited the longest OS, at 51.4 months (Supplementary Table S1).

Discussion
Our study demonstrated that high PD-L1 expression was independently associated with poor prognosis in ALK-rearranged patients receiving ALK-TKI treatment.Patients with high PD-L1 expression had a large number of Tregs and exhausted CD8+ T cells in the TME, which was associated with prognosis in ALK-rearranged patients.
We found that high PD-L1 expression was associated with a worse OS in ALK-rearranged NSCLC patients.Recently, accumulating evidence has suggested that PD-L1 expression could impact the efficacy of ALK-TKIs in ALK-rearranged NSCLC [8,9].However, the impact of PD-L1 expression on the prognosis of ALK-TKIs remains controversial.A prior study showed that the status of PD-L1 expression alone did not influence the OS of ALK-positive NSCLC patients [11].Zhou et al. observed that patients with a high baseline PD-L1 expression status (TPS ≥ 50%) had a significantly shorter median OS after crizotinib treatment [10].This difference could be due to the number of cases analysed in the studies.In our study, we found that high PD-L1 expression was identified as an independent prognostic biomarker, wherein patients with a high PD-L1 expression had a worse OS.We speculate that an immunosuppressed TME may contribute to the poorer prognosis of patients with high PD-L1 expression.The abnormal activation of the PD-1/PD-Ll signal-

Discussion
Our study demonstrated that high PD-L1 expression was independently associated with poor prognosis in ALK-rearranged patients receiving ALK-TKI treatment.Patients with high PD-L1 expression had a large number of Tregs and exhausted CD8+ T cells in the TME, which was associated with prognosis in ALK-rearranged patients.
We found that high PD-L1 expression was associated with a worse OS in ALKrearranged NSCLC patients.Recently, accumulating evidence has suggested that PD-L1 expression could impact the efficacy of ALK-TKIs in ALK-rearranged NSCLC [8,9].However, the impact of PD-L1 expression on the prognosis of ALK-TKIs remains controversial.A prior study showed that the status of PD-L1 expression alone did not influence the OS of ALK-positive NSCLC patients [11].Zhou et al. observed that patients with a high baseline PD-L1 expression status (TPS ≥ 50%) had a significantly shorter median OS after crizotinib treatment [10].This difference could be due to the number of cases analysed in the studies.In our study, we found that high PD-L1 expression was identified as an independent prognostic biomarker, wherein patients with a high PD-L1 expression had a worse OS.We speculate that an immunosuppressed TME may contribute to the poorer prognosis of patients with high PD-L1 expression.The abnormal activation of the PD-1/PD-Ll signalling pathway can inhibit the proliferation and differentiation of T cells and induce their apoptosis in a variety of different ways, thereby mediating the immune escape of tumours and disease progression.A previous study showed that PD-L1 upregulation is correlated with the dysfunction of tumour-infiltrating CD8+ T lymphocytes in NSCLC patients [21].Such evidence suggests that immunological factors could play an important role in patients' response to ALK-TKIs.It is essential to explore the differences in the TME of ALK-rearranged NSCLC with high and low PD-L1 expression.
We found that the number of Tregs in patients with high PD-L1 expression was higher than that in patients with low PD-L1 expression.The association between PD-L1 expression and Treg infiltration has been shown in several cancers [22][23][24].However, it has not been reported in lung cancer.Similar to other studies, we demonstrated that tumour-associated PD-L1 expression correlated with increased Treg infiltration in ALK-rearranged NSCLC.These results were supported by the findings that PD-L1 could increase Foxp3 expression and that PD-L1 could convert naïve T cells to the iTreg phenotype [25,26].Previous studies reported that the TME of ALK-positive NSCLC was immunosuppressive with a high population of immunosuppressive cells, such as Tregs, and a high expression of immunosuppressive markers [27,28].Budczies et al. compared the immune gene expression profile and specific immune cell population levels of ALK-positive and ALK-negative lung adenocarcinoma, and found that the proportion of Tregs increased significantly in ALKpositive tumours [28].Consistent with prior studies, our data indicated that there is a large number of Tregs in the TME of ALK-rearranged NSCLC patients with high PD-L1 expression.A previous study showed that Tregs are immunosuppressive cells that play an important role in shaping the antitumour immune response [29].We found that the number of Tregs around tumour cells within a 30 µm range was higher in the PD-L1hi group than in the PD-L1lo group.These results indicated that the immunosuppressive function of Tregs in the TME was probably based not only on their number, but also on their spatial distribution.
In our study, ALK-rearranged NSCLC patients with high PD-L1 expression had increased exhausted CD8+ T cell levels.The percentage of exhausted CD8+ T cells around tumour cells within the 30 µm range in PD-L1hi patients was also higher.Our findings were in line with a previous study showing that the higher tumour cell expression of PD-L1 was correlated with a higher density of PD-1+ tumour-associated T cells [30].These findings were supported by the work of Pardoll et al., who showed that activated PD-L1 expression in tumour cells could promote CD8+ T cell exhaustion in the TME [31].A previous study reported that a significant number of PD-1-positive CD8+ T cells infiltrated the ALK-positive tumour bed [32].The combined analysis of PD-L1 and CD8+ TILs showed a significantly higher proportion of PD-L1-/TIL-tumours and a lower proportion of PD-L1+/TIL+ tumours in ALK-positive patients than in ALK-negative patients [33].The upregulation of PD-1 expression was associated with tumour antigen-specific CD8+ T cell dysfunction in NSCLC, and PD-1+ CD8+ T cells could result in immunosuppression in cancer [18,21].These results suggested that the TME in ALK-rearranged NSCLC patients with high PD-L1 expression may be inhibited.
In our study, compared with the PD-L1lo group, the PD-L1hi group showed that Tregs and exhausted CD8+ T cells were closer to tumour cells, but this difference was not statistically significant.A multitude of studies have indicated that the proximity measurement of specific immune cells may be a potential indicator for survival [34,35].Our results suggested that the spatial relationships between Tregs or exhausted CD8+ T cells and tumour cells may not contribute to survival, which is inconsistent with the outcomes reported in previous studies [36,37].A possible explanation for this might be that our study was limited by a relatively small sample size.The precise relationships between Tregs and tumour cells and between exhausted CD8+ T cells and tumour cells in the microenvironment of ALK-rearranged NSCLC require further study.
We found that high Treg levels and high exhausted CD8+ T cell levels were associated with a worse PFS and OS.A recent study analysed advanced NSCLC biopsies obtained at different TKI treatment time points and found that regulatory T cells and CD8+ T cells expressing PD-1 were enriched at the PD time point [38].Studies in NSCLC showed that patients with CD8 lymphocytes lacking PD-1 were associated with a more favourable clinical outcome than those harbouring CD8 lymphocytes expressing PD-1 [14,39].Tregs can secrete TGF-β, IL-10, and IL-35, which downregulate antitumour immunity, suppress antigen presentation by DCs and CD4+ T cell function and promote intratumoural T cell exhaustion [40,41].Exhausted CD8+ T cells have been demonstrated to lose effector functions and be unable to efficiently produce IL-2, TNF and IFN-γ [18].Thus, these findings were in agreement with the expected negative clinical impact of locally immunosuppressive cells, such as Tregs and exhausted CD8+ T cells, promoting cancer growth and progression.However, there is a lack of evidence regarding the impact of the TME on prognosis.Our results showed that high Treg levels and exhausted CD8+ T cell levels were associated with a worse OS, suggesting that Tregs and exhausted CD8+ T cells might be prognostic biomarkers in ALK-rearranged NSCLC.This finding was consistent with a previous study showing that the increased infiltration of Tregs into core tumour regions is an independent predictor of worse OS in NSCLC [36].Furthermore, we found that patients with low exhausted CD8+ T cell levels showed an association with a longer PFS in the PD-L1hi group.Such findings indicated that based on different TMEs, a subset of ALK-rearranged NSCLC patients with high PD-L1 expression may benefit from ALK-TKIs.The specific molecular mechanisms and signalling pathways that regulate the close interaction between PD-L1 and Tregs or exhausted CD8+ T cells within the TME were not explored in our study, and we need to further decipher the determinants of Treg and exhausted CD8+ T cell infiltration in patients with a high PD-L1 expression and ALK-rearranged NSCLC.
Additionally, our study found a strong association between exhausted CD8+ T cells and B cells.A previous study showed that B cells can secrete IL-10 and that blocking IL-10 can prevent and reverse T cell exhaustion [42].Our findings imply that B cells promote CD8+ T cell exhaustion and dampen antitumour activity.However, the correlation analysis in this study was only based on the quantitative expression of biomarkers without considering the influence of location.Further analysis is needed to confirm this.
There are some limitations to our study.First, this was a retrospective study with a small sample size from a single institution.Second, the biopsy specimen tissue amount was small, and the tumour and the microenvironment were heterogeneous, and therefore the expression of biomarkers in the entire tumour tissue might be overestimated or underestimated.Third, only mIF was used to characterise the TME, and other tests, such as flow cytometry and single-cell RNA sequencing, should be conducted in future studies.

Conclusions
In conclusion, our study found an association between high PD-L1 expression and poor prognosis in ALK-rearranged NSCLC patients.Patients with high PD-L1 expression demonstrated an immunosuppressive TME, which was characterised by a higher Treg level and exhausted CD8+ T cell infiltration.The characteristics of the TME may help to identify patients who would greater benefit from ALK-TKIs.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biom13060991/s1, Figure S1: Kaplan-Meier analysis of OS based on TME in patients treated with crizotinib; Table S1: Clinical outcomes of 16 patients with ALKrearranged NSCLC treated with crizotinib.
Author Contributions: X.T. and Y.L. had equal contribution and are coauthors as the first author; Conceptualization, Y.L. and P.T.; methodology, H.Z. and Q.W.; software, Q.H.; formal analysis, Q.H.; investigation, H.Z. and Q.W.; writing-original draft preparation, X.T. and Y.L.; writing-review and editing, P.T.; visualization, X.T. and Y.L.; supervision, P.T.; funding acquisition, P.T.All authors have read and agreed to the published version of the manuscript.

Figure 1 .
Figure 1.Flow chart of the study.

Figure 2 .
Figure 2. Survival analysis of 52 patients with ALK−rearranged NSCLC.(A) Maximum percentage reduction from the baseline sum of lesion diameters in 52 patients receiving ALK−TKI treatment.(B) Clinical response in patients receiving ALK−TKI treatments.(C,D) PFS (C) and OS (D) in patients with ALK−rearranged NSCLC stratified by PD-L1 TPS 50%.

Figure 2 .
Figure 2. Survival analysis of 52 patients with ALK−rearranged NSCLC.(A) Maximum percentage reduction from the baseline sum of lesion diameters in 52 patients receiving ALK−TKI treatment.(B) Clinical response in patients receiving ALK−TKI treatments.(C,D) PFS (C) and OS (D) in patients with ALK−rearranged NSCLC stratified by PD-L1 TPS 50%.

Figure 4 .
Figure 4. Location, spatial distribution and correlation analysis of Treg and exhausted CD8+ T cells.(A,B) Treg (A) and exhausted CD8+ T cell (B) percentages in the tumour and stroma of the invasive margin (IM) region.(C) Representative image showing the NND analysis, with the NND calculated from tumour cells to their nearest Treg and exhausted CD8+ T cells.(D) Cartoons depicting the number of cells within a 30 µm radius (tumour cells (red), Tregs (green), and exhausted CD8+ T cells (blue)).(E,F) Boxplots of the average NND between Tregs (E), exhausted CD8+ T cells (F) and tumour cells.(G,H) Boxplots showing the number of Tregs (G) and exhausted CD8+ T cells (H) within a 30 µm radius of tumour cells.(I,J) The correlation between exhausted CD8+ T cells and B cells in PD−L1hi patients.

Figure 4 .
Figure 4. Location, spatial distribution and correlation analysis of Treg and exhausted CD8+ T cells.(A,B) Treg (A) and exhausted CD8+ T cell (B) percentages in the tumour and stroma of the invasive margin (IM) region.(C) Representative image showing the NND analysis, with the NND calculated from tumour cells to their nearest Treg and exhausted CD8+ T cells.(D) Cartoons depicting the number of cells within a 30 µm radius (tumour cells (red), Tregs (green), and exhausted CD8+ T cells (blue)).(E,F) Boxplots of the average NND between Tregs (E), exhausted CD8+ T cells (F) and tumour cells.(G,H) Boxplots showing the number of Tregs (G) and exhausted CD8+ T cells (H) within a 30 µm radius of tumour cells.(I,J) The correlation between exhausted CD8+ T cells and B cells in PD−L1hi patients.

Figure 5 .
Figure 5. Kaplan-Meier analysis of PFS and OS according to the percentage of Treg and exhausted CD8+ T cells in patients treated with crizotinib.(A,B) The difference in PFS between patients with high and low levels of tumour-infiltrating Tregs (A) and exhausted CD8+ T cells (B).(C,D) The difference in OS between patients with high and low levels of tumour-infiltrating Tregs (C) and exhausted CD8+ T cells (D).

Figure 5 .
Figure 5. Kaplan-Meier analysis of PFS and OS according to the percentage of Treg and exhausted CD8+ T cells in patients treated with crizotinib.(A,B) The difference in PFS between patients with high and low levels of tumour-infiltrating Tregs (A) and exhausted CD8+ T cells (B).(C,D) The difference in OS between patients with high and low levels of tumour-infiltrating Tregs (C) and exhausted CD8+ T cells (D).

Figure 6 .
Figure 6.Kaplan-Meier analysis of PFS based on TME in patients treated with crizotinib.(A,B) The difference in PFS between patients with high and low levels of tumour-infiltrating Tregs in the PD-L1lo group (A) and PD-L1hi group (B).(C,D) The difference in PFS between patients with high and low levels of tumour-infiltrating exhausted CD8+ T cells in the PD-L1lo group (C) and PD-L1hi group (D).

Figure 6 .
Figure 6.Kaplan-Meier analysis of PFS based on TME in patients treated with crizotinib.(A,B) The difference in PFS between patients with high and low levels of tumour-infiltrating Tregs in the PD-L1lo group (A) and PD-L1hi group (B).(C,D) The difference in PFS between patients with high and low levels of tumour-infiltrating exhausted CD8+ T cells in the PD-L1lo group (C) and PD-L1hi group (D).

Table 2 .
Baseline characteristics in the subgroups of patients with low and high PD-L1 TPS.

Table 3 .
Univariate and multivariate analyses for PFS.

Table 3 .
Univariate and multivariate analyses for PFS.

Table 4 .
Univariate and multivariate analyses for OS.