Comments
Abstract
Rationale: Patients with chronic obstructive pulmonary disease (COPD) have a higher prevalence of lung cancer. The chronic inflammation associated with COPD probably promotes the earliest stages of carcinogenesis. However, once tumors have progressed to malignancy, the impact of COPD on the tumor immune microenvironment remains poorly defined, and its effects on immune-checkpoint blockers’ efficacy are still unknown.
Objectives: To study the impact of COPD on the immune contexture of non–small cell lung cancer.
Methods: We performed in-depth immune profiling of lung tumors by immunohistochemistry and we determined its impact on patient survival (n = 435). Tumor-infiltrating T lymphocyte (TIL) exhaustion by flow cytometry (n = 50) was also investigated. The effectiveness of an anti–PD-1 (programmed cell death-1) treatment (nivolumab) was evaluated in 39 patients with advanced-stage non–small cell lung cancer. All data were analyzed according to patient COPD status.
Measurements and Main Results: Remarkably, COPD severity is positively correlated with the coexpression of PD-1/TIM-3 (T-cell immunoglobulin and mucin domain–containing molecule-3) by CD8 T cells. In agreement, we observed a loss of CD8 T cell–associated favorable clinical outcome in COPD+ patients. Interestingly, a negative prognostic value of PD-L1 (programmed cell death ligand 1) expression by tumor cells was observed only in highly CD8 T cell–infiltrated tumors of COPD+ patients. Finally, data obtained on 39 patients with advanced-stage non–small cell lung cancer treated by an anti–PD-1 antibody showed longer progression-free survival in COPD+ patients, and also that the association between the severity of smoking and the response to nivolumab was preferentially observed in COPD+ patients.
Conclusions: COPD is associated with an increased sensitivity of CD8 tumor-infiltrating T lymphocytes to immune escape mechanisms developed by tumors, thus suggesting a higher sensitivity to PD-1 blockade in patients with COPD.
The immune system is strongly involved in the establishment of chronic inflammation in chronic obstructive pulmonary disease (COPD) and in the control of tumor burden in lung cancer. However, despite the strong epidemiologic link between these two diseases, the impact of COPD-associated chronic inflammation on the immune contexture of lung cancer remains poorly defined.
Here, we report that COPD disrupts the immune microenvironment of non–small cell lung cancer (NSCLC), and we identify CD8 tumor-infiltrating lymphocytes (CD8 TILs) as the most affected population. Indeed, we observed higher exhaustion of CD8 TILs, identified by PD-1 (programmed cell death-1)/TIM-3 (T-cell immunoglobulin and mucin domain–containing molecule-3) coexpression, in patients with NSCLC with coexisting moderate to severe COPD. In agreement, the prognostic value of intratumor CD8+ T cells that has been found favorable in most cancer types and particularly in NSCLC has no impact on the survival of patients with coexisting COPD. Together, our data point out patients with COPD as a potential NSCLC patient population to treat with immune-checkpoint blockers. In accordance with this hypothesis, data obtained in a cohort of 39 nivolumab-treated patients might suggest a higher efficacy of anti–PD-1 treatment in patients with NSCLC with coexisting COPD.
Despite abundant evidence that the immune system plays a central role in controlling tumor burden (1–3), it may also have a dark side linked to the maintenance of deleterious inflammation. For instance, patients with inflammatory bowel disease (4) or chronic pancreatitis (5) showed increased risk of developing colorectal and pancreatic cancer, respectively. Similarly, chronic obstructive pulmonary disease (COPD) is considered to be an important risk factor for lung cancer (6, 7). This inflammatory condition is linked to a more pronounced destructive inflammation of the lung, compared with non-COPD smokers, characterized by a strong release of TNF-α (tumor necrosis factor-α) and CXCL-8 by epithelial cells and alveolar macrophages leading to the recruitment of inflammatory monocytes and neutrophils (8). The presence of B cells in lymphoid follicles has been reported in the airways and parenchyma of patients with COPD (9), illustrating the involvement of adaptive immunity in COPD pathophysiology. This chronic inflammation may promote the earliest stages of carcinogenesis (8) through an increased expression of genes involved in cell proliferation and survival, including NF-κB (nuclear factor-κB) and STAT3, which are activated by cytokines, such as IL-6 and TNF-α.
Once tumors have progressed to malignancy, COPD was shown to worsen the survival of patients with early stage non–small cell lung cancer (NSCLC) (10) and emphysema was shown to be associated with increased lung cancer mortality (11). Mechanisms governing this prognostic impact, including the role of the immune system, are currently undefined. Although the tumor immune contexture in NSCLC has been extensively characterized, the COPD status of patients has not been taken into account. Nevertheless, a high density of CD8 tumor-infiltrating T lymphocytes (CD8 TILs), together with a concomitant high density of DC-Lamp+ cells that signals the presence of tertiary lymphoid structures within tumor tissues, identified patients with the best prognostic outcome (12). However, overexpression of inhibitory receptors by tumor-infiltrating T cells, also called immune checkpoints, can keep the immune system under control (13). In NSCLC, their cumulative expression, including PD-1 (programmed cell death-1), TIM-3 (T-cell immunoglobulin and mucin domain–containing molecule-3), CTLA-4 (cytotoxic T lymphocyte–associated antigen-4), and LAG-3 (lymphocyte activation gene-3), has been described as being a hallmark of dysfunctional T cells and tumor progression (14).
Drugs targeting immune checkpoints, in particular the PD-1/PD-L1 (programmed cell death-1/programmed cell death ligand 1) pathway, can unleash antitumor immunity and mediate durable cancer regression (15–17). Nevertheless, these new treatments are not efficient in all patients and identifying factors that predict clinical response to these therapies remains a challenge. In melanoma, an association between high PD-L1 expression and clinical response to pembrolizumab had been reported (18). However, patients with PD-L1–negative tumors may also achieve durable responses. In NSCLC, several efforts have also been made recently to more accurately identify patients that would respond to checkpoint therapy. The focus here has largely been on the identification of predictive markers for response to anti–PD-1, such as tumor mutational burden (TMB) (19), PD-L1 expression by tumor cells (20, 21), and gene signature reflecting adaptive immunity (22).
In this context, the present study investigated the potential impact of COPD on the immune microenvironment of NSCLC and, thus, on patients’ outcome. Our work reveals that COPD severity is positively correlated with the level of CD8 TIL exhaustion. In agreement, we observed a complete loss of CD8 T cell–associated favorable clinical outcome in COPD+ patients. Finally, data obtained on 39 patients with advanced-stage NSCLC treated by an anti–PD-1 might suggest a higher sensitivity to this treatment in patients with COPD.
A retrospective consecutive cohort of 435 NSCLC untreated patients seen between June 2001 and December 2005 at the Department of Thoracic Surgery of Hôtel-Dieu Hospital (Paris, France) was used to study by immunohistochemistry the immune composition of the tumor microenvironment. A second cohort of fresh tumor samples, distant nontumoral lung samples, and peripheral blood was obtained from 50 patients with untreated NSCLC who underwent surgery between March 2014 and December 2015. These samples were used to perform flow cytometry experiments. A third cohort of 39 patients with advanced-stage NSCLC receiving an anti–PD-1 antibody (nivolumab) was used to assess the effectiveness of this treatment according to COPD status. Additional details are provided in the online supplement.
The Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria were used to assess the presence of COPD and to evaluate the severity of airflow obstruction (23). Additional details are provided in the online supplement.
Serial sections of paraffin-embedded NSCLC tumors were stained as previously described (24). Additional details are provided in the online supplement.
Calopix software (Tribvn) was used to count CD66b+ and CD68+ cells in the whole tumor section; CD8 T cells were counted separately in the tumor nests and in tumor stroma. DC-Lamp+ cells were counted manually in the whole tumor section. The proportion of PD-L1+ cells among tumor cells was determined manually by at least two independent observers (J.B., H.O., or D.D.). The positivity threshold was fixed at greater than or equal to 1%. Additional details are provided in the online supplement (see Table E1 in the online supplement).
Multiple stainings on isolated mononuclear cells from tumor, nontumor distal lung tissue (NT), and blood were performed using various antibodies (see Table E2 and Figure E1), as previously described (24). Additional details are provided in the online supplement.
Genomic DNA from tumors was isolated from formalin-fixed paraffin-embedded blocks using Maxwell 16 FFPE Tissue LEV DNA Purification Kit (Promega), according to the manufacturer’s instructions. DNA whole-exome sequence data were sequenced on the Novaseq 6000 platform (25). Additional details are provided in the online supplement.
Categorical data were compared using chi-square test or Fisher exact test, as appropriate, whereas they were compared according to COPD stages using exact Cochran-Armitage trend test. For log-rank tests, the prognostic value of continuous variables was assessed by a quartile stratification. For Cox proportional hazards models, immune cell densities were log-transformed. Multivariate analysis for overall survival (OS) was adjusted for age, sex, vascular emboli, and smoking history, and stratified on the tumor stage.
In flow cytometry experiments, according to data distribution, a parametric test (ANOVA, Student’s t test) or a nonparametric test (Kruskal-Wallis, Mann-Whitney), with appropriate post hoc comparisons, was used to compare quantitative variables across the different groups. Correlations between quantitative parameters were assessed by using the Spearman test. Additional details are provided in the online supplement.
We first investigated the impact of COPD on the composition of the tumor immune microenvironment in a retrospective cohort of 435 patients with NSCLC. Among them, 45% had COPD, and in the COPD+ group, 29% had a COPD GOLD stage I (COPD+ I), 60% a COPD GOLD stage II (COPD+ II), and 11% a COPD GOLD stage III (COPD+ III) (Table 1; see Table E3). The mean age and the percentages of male and of smokers were higher for COPD+ than COPD− patients (Table 1). Coexisting COPD was associated with significant worse survival only for NSCLC stage I patients (see Figure E2). Because of the small number of COPD+ III patients, the COPD+ II and COPD+ III groups were merged for most of the subsequent analyses. COPD+ and COPD− patients did not differ in density of neutrophils (CD66b+ cells), macrophages (CD68+ cells), mature DCs (DC-Lamp+ cells), and CD8 T cells in tumor nests (CD8Tu) and in stroma (CD8S), regardless of GOLD stage (see Figure E3).
| Characteristics | COPD− | COPD+ | P Value |
|---|---|---|---|
| Subjects, n (%) | 238 (55) | 197 (45) | |
| Sex, M/F, n (%) | 165/73 (69/31) | 174/23 (88/12) | <0.0001* |
| Age, yr, mean ± SD | 61 ± 12 | 64 ± 9 | 0.01 |
| Smokers, n (%) | 182 (78) | 173 (91) | 0.0003* |
| Smoking history, pack-years, mean ± SD | 36 ± 24 | 49 ± 24 | <0.0001 |
| Histologic subtype, n (%) | |||
| ADC | 173 (73) | 136 (69) | 0.70* |
| SSC | 53 (22) | 50 (25) | |
| Others | 12 (5) | 11 (6) | |
| pTNM stages, n (%) | |||
| I | 102 (43) | 94 (48) | 0.05* |
| II | 58 (24) | 59 (30) | |
| III-IV | 78 (33) | 44 (22) | |
| Vascular emboli, n (%) | |||
| Yes | 125 (53) | 118 (60) | 0.21* |
| No | 98 (41) | 65 (33) | |
| ND | 15 (6) | 14 (7) | |
| Pleural invasion, n (%) | |||
| Yes | 124 (52) | 110 (56) | 0.70† |
| No | 110 (46) | 83 (42) | |
| ND | 4 (2) | 4 (2) | |
| Lobectomy, n (%) | 210 (88) | 171 (87) | 0.65* |
| Pneumonectomy, n (%) | 28 (12) | 26 (13) | |
| % FEV1/FVC, mean ± SD | 79 ± 5 | 60 ± 8 | <0.0001 |
| FEV1 (% of a predicted value), mean ± SD | 90 ± 17 | 71 ± 17 | <0.0001 |
To indirectly investigate whether the functionality of immune cells in the tumor microenvironment could be modified by coexisting COPD, we first determined their impact on patient survival according to COPD status and GOLD stage (Figure 1; see Tables E4 and E5). In the whole retrospective cohort and in the COPD− group, univariate Cox regression analysis showed that CD8Tu, CD8S, and DC-Lamp+ cell densities were all associated with favorable prognostic value, whereas neutrophil and macrophage density had no impact on patient survival (Figures 1A and 1B). Strikingly, CD8Tu cell density did not affect patient survival in COPD+ patients (Figure 1C). Furthermore, CD8Tu, CD8S, and DC-Lamp+ cell densities were not significantly associated with improved survival in COPD+ II–III patients (Figure 1D).
Figure 1. Prognostic value of immune cell densities in non–small cell lung cancer according to chronic obstructive pulmonary disease (COPD) status (retrospective cohort). (A–D) Forest plots of univariate Cox regression analysis showing the impact of neutrophil (n = 435), macrophage (n = 435), CD8Tu (n = 427), CD8S (n = 435), and DC-Lamp+ cell (n = 435) density on overall survival in the whole cohort (all patients) (A) and COPD− (B), COPD+ (C), and COPD+ II–III (D) patients. (E–G) The quartile method was used to stratify patients into four groups by density of CD8Tu cells, from the lowest (first quartile, black curves) to the highest density of CD8Tu cells (fourth quartile, red curves). Kaplan-Meier curves of overall survival according to CD8Tu cell density in COPD− (E), COPD+ (F), and COPD+ II–III (G) patients. The horizontal dashed lines represent the median survivals. (H–J) Forest plots of multivariate Cox regression analysis showing the impact of CD8Tu, CD8S, and DC-Lamp+ cell density on overall survival adjusted for age, sex, vascular emboli, smoking history, and stratified on the stage of the tumor, in COPD− (H), COPD+ (I), and COPD+ II–III (J) patients. P < 0.05 were considered statistically significant and appear in bold. CI = confidence interval; HR = hazard ratio.
[More] [Minimize]Then, patients were stratified by quartiles of CD8Tu cell density. In COPD− patients, CD8Tu cell density was associated with longer OS as soon as the second quartile was reached (Figure 1E), whereas in COPD+ and COPD+ II–III patients the survival curves for all quartiles merged together (Figures 1F and 1G). DC-Lamp+ cell and CD8S cell densities were not associated with significant prognostic value in COPD+ II–III patients (see Figure E4). Multivariate Cox regression analysis adjusted for age, sex, vascular emboli, and smoking history and stratified on tumor stages highlighted the absence of CD8 T-cell prognostic value for the stroma and tumor nests in COPD+ patients (Figures 1H–1J). Together, these data might suggest that the protective impact of a high adaptive immune cell infiltration in NSCLC is altered in COPD+ patients and identify CD8Tu cells as the most affected population.
Based on the previously mentioned results, we investigated whether effector functions of TILs were altered in COPD+ patients using a prospective cohort of 50 patients with NSCLC (Figure 2; see Figure E5 and Tables E6 and E7). Regardless of COPD status, within the tumor tissue (tumor), the proportion of CTLA-4+, LAG-3+, PD-1+, and TIM-3+ cells among CD4 (see Figure E5A) and CD8 T cells (Figures 2A and 2B) was consistently higher than in the other anatomic sites (blood and NT). A marked increase of CD4+ FoxP3+ regulatory T cells frequency among total CD4 T cells in tumor was also observed (see Figures E6A and E6B). Regarding cytokine secretion, the frequency of CD4 (see Figure E5B) and CD8 T cells (Figures 2C and 2D) positive for granzyme B, TNF-α, IFN-γ, and IL-17 was lower in tumor than in NT. As shown by the correlation matrix exposed on Figure 2E, among CD8 TILs, PD-1 and TIM-3 expression was strongly positively correlated, as was the frequency of IFN-γ+ and TNF-α+ cells. Remarkably, CD8 TILs coexpressing PD-1 and TIM-3 were restricted to tumor (Figure 2F), and for this cell subset only, there was a significant inverse correlation with both IFN-γ and TNF-α secretion (Figure 2G). Overall, similar results were observed regarding CD4 TILs (see Figures E5C and E5D). However, fewer CD4 TILs coexpressed PD-1/TIM-3 (see Figure E5D), and the relationship between cytokine secretion and PD-1/TIM-3 coexpression was weaker (see Figures E5C and E5E).
Figure 2. Characterization of CD8 tumor-infiltrating T lymphocytes (CD8 TILs) in non–small cell lung cancer (prospective cohort). (A) Histograms showing, for one representative patient, the frequency of CTLA-4+, LAG-3+, PD-1+, and TIM-3+ cells among CD8 T cells in tumor, nontumor distal lung tissue (NT), and blood. (B) Frequency of CTLA-4+, LAG-3+, PD-1+, and TIM-3+ cells among CD8 T cells in tumor, NT, and blood. (C) Histograms showing, for one representative patient, the frequency of granzyme B+, TNF-α+, IFN-γ+, and IL-17+ cells among CD8 T cells in tumor, NT, and blood. (D) Frequency of granzyme B+, TNF-α+, IFN-γ+, and IL-17+ cells among CD8 T cells in tumor, NT, and blood. (E) Spearman-correlation matrix of parameters studied by flow cytometry. Each colored square illustrates the correlation between two parameters. Red illustrates a strong positive correlation, and green illustrates a strong negative correlation. (F) Frequencies of PD-1+ TIM-3+, PD-1+ TIM-3−, PD-1− TIM-3+, and PD-1− TIM-3− cells among CD8 T cells in tumor (black circle; n = 50), NT (gray circle; n = 47), and blood (white circle; n = 41). (G) Correlation between the frequencies of PD-1+ TIM-3+, PD-1− TIM-3+, PD-1+ TIM-3−, and PD-1− TIM-3− cells among CD8 TILs and frequencies of granzyme B+, IL-17+, TNF-α+, and IFN-γ+ cells among CD8 TILs. In B, D, and F, data are expressed as mean and an ANOVA or a Kruskal-Wallis test followed by an appropriate correction was applied based on Shapiro normality test. In E and G, a Spearman test was applied. *P < 0.05, **P < 0.01, and ***P < 0.001. CTLA-4 = cytotoxic T lymphocyte–associated antigen-4; LAG-3 = lymphocyte activation gene-3; PD-1 = programmed cell death-1; TIM-3 = T-cell immunoglobulin and mucin domain–containing molecule-3; TNF-α = tumor necrosis factor-α; Tu = tumor.
[More] [Minimize]Based on these results, we investigated the link between COPD and TIL exhaustion. In COPD, airflow obstruction severity is inversely correlated with the FEV1 % predicted (see methods section). Remarkably, FEV1 % predicted was inversely correlated with the proportion of CD8 TILs expressing PD-1 and coexpressing PD-1/TIM-3 in COPD+ patients only (Figure 3A). In agreement, in COPD+ patients, FEV1 % predicted was positively correlated with the proportion of CD8 TILs secreting TNF-α and IFN-γ (Figure 3A). Regarding CD4 TILs, only IFN-γ was positively correlated with FEV1 % predicted (see Figure E7A). Interestingly, frequency of TIM-3+, PD-1+, and TIM-3+/PD-1+ cells among CD4 and CD8 TILs was higher in COPD+ II–III patients than in COPD− patients (Figure 3B; see Figure E7B). The proportion of regulatory T cells among CD4 TILs was not different according to patients’ COPD status (see Figure E7C). Overall, these results demonstrate that COPD severity is strongly correlated with TIL exhaustion, and that this association is more pronounced for CD8 TILs.
Figure 3. CD8 tumor-infiltrating T lymphocytes (CD8 TILs) exhaustion according to the chronic obstructive pulmonary disease (COPD) status of the patients (prospective cohort). (A–E) CD8 TIL characterization in COPD− (n = 30), COPD+ I (n = 8), and COPD+ II–III (n = 12) patients. (A) Graphical representation of Spearman correlations between the FEV1 % predicted and the frequencies of PD-1+, TIM-3+, PD-1+ TIM-3+, granzyme B+, IFN-γ+, and TNF-α+ cells among CD8 TILs, in COPD− patients and COPD+ patients. Red and blue colors indicate positive and negative correlations, respectively; the lighter the color, the less significant the corresponding correlation. The filled fraction of the circle in each of the pie charts corresponds to the absolute value of the associated Spearman correlation coefficient. (B) Frequencies of PD-1+, TIM-3+, and PD-1+ TIM-3+ cells among CD8 TILs. (C) Graphical representation of Spearman correlations in COPD− and COPD+ patients, between CD8Tu cell densities and flow cytometry data, including the frequencies of PD-1+, of TIM-3+, of PD-1+ TIM-3+, granzyme B+, IFN-γ+, and TNF-α+ cells among CD8 TILs. Graphical representation is the same as in A. (D and E) The median CD8Tu cell density was used to stratify patients into CD8TuLow (D) or CD8TuHigh (E) groups. (D and E) Histograms and radar plots showing the frequencies of PD-1+, TIM-3+, PD-1+ TIM-3+, IFN- γ +, and TNF-α+ cells among CD8 TILs according to COPD status in CD8TuLow group and CD8TuHigh group. In A and C, a Spearman test was performed. In B, data are expressed as mean ± SEM and a parametric test (ANOVA with post hoc Bonferroni correction) or a nonparametric test (Kruskal-Wallis test followed by a post hoc Dunn test) was applied based on Shapiro normality test. In D and E, data are expressed as mean, and a parametric test (Student’s test) or a nonparametric test (Mann-Whitney test) was applied based on Shapiro normality test. *P < 0.05, **P < 0.01, and ***P < 0.001. PD-1 = programmed cell death-1; TIM-3 = T-cell immunoglobulin and mucin domain–containing molecule-3; TNF-α = tumor necrosis factor-α
[More] [Minimize]An association between CD8 TIL exhaustion (PD-1+ cell frequency) and the immune composition of the tumor microenvironment (density of CD8+ T cells) was recently reported in colorectal cancer (26). We investigated this interrelation in our prospective cohort and then studied the impact of COPD. First, CD8 TIL exhaustion (based on PD-1 and TIM-3 expression) and cytokine secretion were only linked to CD8Tu cell and CD8S cell densities (see Figure E8A). Regarding CD4 TILs, none of the immune cell densities studied was associated with their exhaustion, and only their cytokine secretion was slightly inversely correlated with CD8S cell density (see Figure E7D). Because of the strong association between CD8 TIL exhaustion and their density in the tumor nests, we then focused our analysis on CD8 TILs. Importantly, CD8Tu cell density and CD8 TIL exhaustion were more strongly associated in COPD+ patients than in COPD− patients (Figure 3C; see Figures E8B and E8C).
To confirm these results, median CD8Tu cell density was used to separate patients into two groups according to a low (Figure 3D) or a high (Figure 3E) CD8Tu cell density. In the CD8TuLow group, the level of CD8 TIL exhaustion did not differ according to COPD status (Figure 3D). In the CD8TuHigh group, the frequencies of CD8 TILs expressing TIM-3 and coexpressing PD-1/TIM-3 were significantly higher in COPD+ patients than in COPD− patients (Figure 3E). Overall, CD8 TIL exhaustion was restricted to highly CD8 T-cell infiltrated tumors and this phenomenon was exacerbated in COPD+ patients.
The strong impact of immunosuppression on tumor burden is based on TIL exhaustion, but also on concomitant mechanisms that malignant cells develop to avoid immune surveillance. The most-studied mechanism is probably PD-L1 expression by malignant cells (see Figures E9A–E9D). No difference of PD-L1 expression by tumor cells was observed according to patients’ COPD status and GOLD stages (retrospective cohort) (see Figure E9F). Consistent with previous studies, we found that high CD8 T-cell density is associated with high PD-L1 expression by tumor cells (see Figure E9G) (27, 28). The frequency of tumor cells expressing PD-L1 was also higher, but to a lesser extent, in neutrophilHigh, macrophageHigh, and DC-LampHigh groups. Additionally, among these highly infiltrated groups, PD-L1 expression was similar between COPD− and COPD+ patients (see Figure E9G).
We then investigated whether coexisting COPD modified the prognostic value of PD-L1 expression by tumor cells. Whatever the group of patients considered, PD-L1 expression was not associated with significant prognostic value (Figures 4A and 4B). Because PD-L1 expression was strongly linked to CD8 T-cell density (see Figure E9G), we then deciphered the prognostic value of PD-L1 according to a high/low CD8Tu cell density, and also to patient COPD status. For CD8TuLow groups, PD-L1 expression was not associated with significant prognostic value in COPD− or in COPD+ patients (Figures 4C and 4D). Interestingly, for CD8TuHigh groups, PD-L1 expression did not affect survival for COPD− patients (Figure 4E) but was associated with a reduced OS for COPD+ patients (Figure 4F). Moreover, in COPD− patients, the prognostic value of CD8Tu and of CD8S cell density was similar whether tumor cells expressed PD-L1 or not (Figure 4G). Remarkably, for COPD+ patients, CD8Tu and CD8S cell densities were associated with extended OS for those with PD-L1− tumors, whereas these prognostic values were not observed in the PD-L1+ groups (Figure 4H). Finally, these results were confirmed in subgroups of patients defined by a cutoff of PD-L1+ tumor cell frequency greater than or equal to 5% (see Figures E10A and E10B) and greater than or equal to 10% (see Figure E10C and E10D).
Figure 4. Immune cell prognostic value according to chronic obstructive pulmonary disease (COPD) status and PD-L1 (programmed cell death ligand 1) expression by tumor cells (retrospective cohort). Kaplan-Meier curves of overall survival (OS) in patients with non–small cell lung cancer according to PD-L1 expression by tumor cells, in COPD− patients (n = 234) (A) and in COPD+ patients (n = 192) (B). (C and D) Kaplan-Meier curves of OS in the CD8TuLow group according to PD-L1 expression by tumor cells in COPD− (n = 111) (C) and COPD+ patients (n = 101) (D). (E and F) Kaplan-Meier curves of OS in the CD8TuHigh group according to PD-L1 expression by tumor cells in COPD− (n = 123) (E) and COPD+ patients (n = 91) (F). (G and H) Forest plots of univariate Cox regression analyses showing the impact of CD8Tu cell, CD8S cell, and DC-Lamp+ cell density on the OS according to PD-L1 expression by tumor cells, in COPD− patients (G) and in COPD+ patients (H). P < 0.05 were considered statistically significant and appear in bold. CI = confidence interval; HR = hazard ratio.
[More] [Minimize]We investigated the impact of COPD on the effectiveness of an anti–PD-1 antibody from a cohort of 39 patients with advanced-stage NSCLC receiving nivolumab (see Table E8). The percentage of smokers and the number of pack-years were higher for COPD+ than COPD− patients, whereas no significant differences were observed between the two groups of patients regarding NSCLC subtypes, the duration of follow-up, the age, and the sex (see Table E8). At the completion of the study, a significantly longer progression-free survival (PFS) (Figures 5A and 5B) and a higher percentage of patients still alive (Figure 5C; see Tables E8 and E9) were observed in the COPD+ group. However, we did not observe any impact of COPD severity, assessed using the FEV1 % predicted, on nivolumab efficacy (see Table E9).
Figure 5. Nivolumab efficacy in advanced-stage patients with non–small cell lung cancer according to a coexisting chronic obstructive pulmonary disease (COPD) and to smoke exposure. (A and B) Kaplan-Meier curves of progression-free survival (PFS) and overall survival (OS) in COPD− (n = 20) and COPD+ (n = 19) patients. PFS was defined as the time from the start date of nivolumab treatment to the date of the first documented event of tumor progression. (C) Characteristics of the response to nivolumab treatment according to patients’ COPD status and to smoking exposure. (D–G) Kaplan-Meier curves of PFS (D and E) and of OS survival (F and G) in COPD− (n = 20) (D and F) and COPD+ (n = 19) (E and G) patients according to smoking exposure. Patients were stratified into two groups, smokerHigh (>30 pack-years) and smokerLow (≤30 pack-years). Tick marks indicate censoring events. P < 0.05 were considered statistically significant and appear in bold. Mu = mutated-tumor; NA = not available; TMB = tumor mutation burden expressed as number of nonsynonymous mutations per megabase and determined using whole-exome sequencing experiments; WT = wild-type tumor.
[More] [Minimize]In addition, it was previously shown that the efficacy of pembrolizumab, another anti–PD-1 antibody, was greater in patients with a smoking-associated mutational signature or with a higher nonsynonymous mutation burden in tumor (19). Consequently, we investigated whether the increased PFS seen with nivolumab in COPD+ patients was linked to the higher smoke exposure observed in this group (see Table E8). In the whole cohort, a smoke exposure greater than 30 pack-years was associated with a better PFS and OS (see Table E9), whereas in patients without COPD, smoke exposure greater than 30 pack-years was not associated significantly with a better PFS or OS (Figures 5D and 5F; see Table E9). Remarkably, in the COPD+ group, a smoke exposure greater than 30 pack-years was associated with a better PFS and also with a dramatic improvement of OS (Figures 5E and 5G; see Table E9). Regarding PFS, these results were confirmed when smoke exposure was assessed using number of pack-years as a continuous variable (see Table E9).
A strong relationship between tobacco smoke exposure and the number of somatic mutations was previously reported in NSCLC (29). Consequently, we tried to investigate whether the number of nonsynonymous mutations per megabase (TMB) differ according to patients’ COPD status, by performing whole-exome sequencing experiments. Among the 22 patients for whom enough DNA was available (Figure 5C), we did not observe any difference regarding the TMB in COPD+ patients compared with patients without COPD (see Figure E11A). Remarkably, the number of pack-years was significantly correlated with the TMB only in patients without COPD (see Figures E11C–E11E). Additionally, it has been shown that TP53 and/or KRAS-mutated tumors, two mutations strongly associated with tobacco smoke exposure (30, 31), had a better response to PD-1 blockade (32). Among the 31 patients who had a molecular interrogation of their tumor before starting anti–PD-1 treatment (Figure 5C), using next-generation sequencing, we did not detect a differential distribution of TP53 and/or KRAS-mutated tumors according to patients’ COPD status (see Figure E11B).
Our preliminary data might suggest a differential link between smoking history and response to nivolumab, but also between TMB and smoking history, in COPD+ patients versus patients without COPD. Consequently, we investigated whether COPD and tobacco had a synergistic impact on CD8 TIL exhaustion and on immune cell prognostic value. First, in the prospective cohort, the number of pack-years was positively correlated with the proportion of CD8 TILs coexpressing PD-1/TIM-3 in COPD+ patients only (see Figure E12A). Moreover, a higher CD8 TILs exhaustion was observed in COPD+ group compared with COPD− group, in patients with a number of pack-years greater than 60 (see Figure E12B). Second, in our retrospective cohort of 435 patients with NSCLC, we investigated whether immune cell prognostic value (CD8Tu, CD8S, and DC-Lamp+ cells) was impacted by a strong smoke exposure (>30 pack-years). In heavy smokers, immune cell prognostic value was stronger in patients without COPD (see Figure E12C), than the one observed in Figure 1B. Conversely, in COPD+ patients, the CD8S cell prognostic value was not significant in heavy smokers, and was completely absent for CD8Tu cells (hazard ratio, 1.01; P = 0.948) (see Figure E12D). Altogether, patients with NSCLC with COPD, a group characterized by a complete loss of CD8 T cell–associated favorable clinical outcome in heavy smokers probably caused by their marked exhaustion, also showed a longer PFS after nivolumab treatment.
Our main objective was to evaluate the potential impact of COPD on the immune contexture of NSCLC. First, immune cell densities did not differ according to the COPD status of the patients. Immune cell recruitment into the malignant lesion is probably driven by tumor cells according to their immunogenicity linked to their mutational burden, thereby attenuating the impact of coexisting COPD. Cooccurring genetic alterations in KRAS-mutant lung adenocarcinoma, were associated with different tumor immune patterns, and could be a first argument supporting this hypothesis (33). However, our study showed higher TIL exhaustion in the COPD+ II–III group, an impaired protective impact of immune cells in patients with COPD, and identified CD8 TILs as the most affected population.
The characteristics of the NSCLC immune contexture linked to CD8 TIL exhaustion are not completely defined, and the role of COPD in this phenomenon is not completely elucidated. We identified that CD8 TIL exhaustion was restricted to highly CD8 T-cell infiltrated tumors, and these findings were exacerbated in COPD+ patients. Interestingly, PD-1 expression, which is upregulated on T cells after TCR (T-cell receptor) ligation (34), is also upregulated in activated T cells by IL-6 through STAT3-dependent mechanisms (35). Accordingly, high frequency of exhausted TILs observed in COPD+ II–III patients could be driven in part by the increased amounts of IL-6 previously reported in the sputum of COPD+ patients (36). It is also conceivable that when a tumor forms near an emphysematous/inflammatory lesion, the surrounding inflammation, including numerous cytokines (IL-6, TNF-α, and IL-1β) and chemokines (CXCL8 and CXCL1) (37) modifies autocrine and paracrine interactions between malignant cells and infiltrating leukocytes. Interestingly, ex vivo infections with influenza virus of lung resections showed an impaired antiviral function of CD8+ T cells in COPD+ patients compared with patients without COPD, through an upregulation of PD-1 expression. Moreover, if these PD-1 expressing CD8 T cells, coming from the NT, are overrepresented in the tumor immune microenvironment of COPD+ patients, they might participate in the deviation of the antitumor immune response observed in patients with COPD (38). Nevertheless, orthogonal approaches, including gene expression analysis related to the immune response in cancer, are required to precisely identify the characteristics of the inflammation disrupting the tumor immune contexture of patients with NSCLC and COPD.
In agreement with other studies (17, 27, 39), we observed that PD-L1 expression by NSCLC tumor cells highlights the presence of an active tumor immune microenvironment in lung cancer, independently of COPD. This is probably because cancer cells may upregulate PD-L1 expression in response to IFN-γ secretion by TILs (39). Studies of the prognostic value of PD-L1 expression in patients with NSCLC have yielded inconsistent data (40–44). These conflicting results could be because the amount of CD8 TILs and patients’ COPD status had not been taken into account. In our study, the prognostic value of PD-L1 expression was restricted to COPD+ patients belonging to the CD8TuHigh group. It probably reflects the effectiveness of mechanisms that cancer cells develop to avoid immune surveillance in a subpopulation of patients characterized by a strong CD8 TIL exhaustion. In CD8TuLow, the lack of PD-L1 prognostic value was probably linked to the lowest impact of the PD-1/PD-L1 pathway on a weakly active antitumor immune response. In this situation, PD-L1 expression is probably driven more through oncogenic pathways, including inactivation of STK11/LKB1 (45) or loss of function of the tumor suppressor PTEN (46), and not associated to a strong PD-1 expression by CD8 TILs. Moreover, immune cell prognostic value was completely absent for tumors from COPD+ patients expressing PD-L1, probably because the level of TIL exhaustion was higher in this group. Interestingly, in melanoma, preexisting CD8 TILs in tumor microenvironment were required for tumor regression after treatment with pembrolizumab (27), and CD8 TILs coexpressing PD-1/CTLA-4 had been proposed as a biomarker to predict response to anti–PD-1 (47). Altogether, for patients with NSCLC with moderate to severe COPD, our results support an increased predictive potential of PD-L1 expression by tumor cells for the response to checkpoint inhibitors targeting the PD-l/PD-L1 pathway (20).
In accordance with this assumption, our preliminary data obtained in a cohort of 39 nivolumab-treated patients with NSCLC showed a longer PFS in patients with coexisting COPD (19). Nevertheless, we did not identify any impact of COPD severity on the response to anti–PD-1. It is conceivable that we had not enough patients to observe this kind of effect. Moreover, other confounding factors not explored in our work could explain the lack of association between COPD severity and the response to nivolumab, including among others, CD8 TIL density and PD-L1 expression by tumor cells. Remarkably, at the time of publication, the study from Mark and coworkers (48), also observed that the presence of COPD was associated with longer PFS interval in patients treated with anti–PD-1 or anti–PD-L1. However, a smoking-associated mutational signature had previously been suggested to signal a better response to immunotherapy. Because tobacco and COPD may be confounding factors in the response to nivolumab, we tried to investigate the interplay between these two factors, and if possible their respective impact. Our study suggests that the impact of tobacco on the response to nivolumab would be mainly observed in COPD+ patients. In agreement with this assumption, when anti–PD-1/PD-L1 impact was evaluated only in former smokers, Mark and coworkers (48) still observed a longer PFS and OS in COPD+ patients. We also showed a stronger impact on patients’ survival of the immune microenvironment in heavy smokers without COPD. In this situation, tumors are probably more immunogenic and the presence of a strong specific adaptive antitumor immune response even more important for patients’ survival. Interestingly, in COPD+ patients, the number of pack-years was positively correlated with the level of CD8 TIL exhaustion, and the impact of CD8Tu cell density on patients’ survival was completely absent in heavy smokers. Altogether, the higher nivolumab efficacy observed in COPD+ patients probably reflects the effectiveness of PD-1 blockers to unleash antitumor CD8 T-cell response in a subpopulation of patients characterized by strong CD8 TIL exhaustion.
Finally, we tried to investigate whether the longer PFS observed in nivolumab COPD+ treated patients was linked to a higher TMB induced by a stronger tobacco smoke exposure. Unfortunately, we were able to perform such work only on 22 anti–PD-1–treated patients. However, our preliminary data are not in favor of a higher TMB in COPD+ patients. Moreover, the frequency of KRAS and TP53 mutations, two mutations strongly linked to tobacco smoke exposure (30, 31) that have been suggested as being associated with longer PFS in anti–PD-1–treated patients (32), was not enriched in COPD+ patients. In agreement with these results, our previous work did not detect a higher frequency of TP53 or KRAS mutations in COPD+ patients, again characterized by a stronger smoke exposure, in a cohort of 282 lung adenocarcinomas (31). Beyond the scope of the present work, studies using larger cohorts of patients are mandatory to precisely determine whether the longer PFS after PD-1 blockade observed in COPD+ patients is mostly linked or not to their higher tobacco smoke exposure. Such studies will also allow to determine whether tobacco smoke exposure differentially impacts CD8 TIL exhaustion, TMB, and response to PD-1 blockers in COPD+ patients versus patients without COPD.
However, there were some limitations to our study. First, our cohort of anti–PD-1–treated patients is restricted to 39 patients, mainly because we needed a follow-up of at least 1 year, combined to fully characterized respiratory functions. Second, the use of three cohorts inevitably increased the number of comparisons and the number of false discovery rate, but we tried to reduce this risk using the most appropriate statistical methodology and by applying appropriate adjustments for multiple comparisons. The impact of histologic subtypes on our findings was not fully analyzed, even if it does not seem to impact our main results (data not shown). In fact, we did not address this point to avoid an increase of multiple comparisons. Another limitation, inherent to this kind of work, is related to tissue heterogeneity, slide thickness, and surface area covered. However, one strength of our study is that in immunochemistry experiments, we worked on whole tumor sections and not on tumor microarrays, with senior pathologists, using automates for staining, scanning, and counting. All these precautions were intended to increase the accuracy and reproducibility of our data.
By deciphering the immune network of NSCLC, we pinpointed that the use of immunologic biomarkers to evaluate patient prognosis (49) and to predict the response to therapy should definitively take into account the coexistence of COPD. This chronic inflammatory disease of the lung is not fully understood and its impact on tumor immune microenvironment functions and on response to immune checkpoint inhibitors deserves additional experiments. We recommend that clinical trials should investigate whether CD8 TIL density and smoke exposure, together with coexisting COPD, may identified the best responders to therapies targeting immune checkpoints. Overall, our study emphasizes the need to consider the impact of coexisting chronic inflammation on the tumor immune microenvironment in other cancer types. In the era of precision medicine, such studies should extend the clinical success of immunotherapies in cancer.
The authors thank Patricia Bonjour, Béatrice Marmey (Department of Pathology, Hôpital Cochin, Paris), Sarah Leseurre (Department of Thoracic Surgery, Hôpital Cochin, Paris), Nathalie Jupiter, and Samantha Knockaert (UMRS 1138, Cordeliers Research Center, Team 13, Paris) for technical assistance.
| 1. | Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B, Lagorce-Pagès C, et al. Type, density, and location of immune cells within human colorectal tumors predict clinical outcome. Science 2006;313:1960–1964. |
| 2. | Azimi F, Scolyer RA, Rumcheva P, Moncrieff M, Murali R, McCarthy SW, et al. Tumor-infiltrating lymphocyte grade is an independent predictor of sentinel lymph node status and survival in patients with cutaneous melanoma. J Clin Oncol 2012;30:2678–2683. |
| 3. | Sharma P, Shen Y, Wen S, Yamada S, Jungbluth AA, Gnjatic S, et al. CD8 tumor-infiltrating lymphocytes are predictive of survival in muscle-invasive urothelial carcinoma. Proc Natl Acad Sci USA 2007;104:3967–3972. |
| 4. | Kim ER, Chang DK. Colorectal cancer in inflammatory bowel disease: the risk, pathogenesis, prevention and diagnosis. World J Gastroenterol 2014;20:9872–9881. |
| 5. | Pinho AV, Chantrill L, Rooman I. Chronic pancreatitis: a path to pancreatic cancer. Cancer Lett 2014;345:203–209. |
| 6. | de Torres JP, Marín JM, Casanova C, Cote C, Carrizo S, Cordoba-Lanus E, et al. Lung cancer in patients with chronic obstructive pulmonary disease: incidence and predicting factors. Am J Respir Crit Care Med 2011;184:913–919. |
| 7. | Young RP, Hopkins RJ, Christmas T, Black PN, Metcalf P, Gamble GD. COPD prevalence is increased in lung cancer, independent of age, sex and smoking history. Eur Respir J 2009;34:380–386. |
| 8. | Houghton AM. Mechanistic links between COPD and lung cancer. Nat Rev Cancer 2013;13:233–245. |
| 9. | Cosio MG, Saetta M, Agusti A. Immunologic aspects of chronic obstructive pulmonary disease. N Engl J Med 2009;360:2445–2454. |
| 10. | Zhai R, Yu X, Shafer A, Wain JC, Christiani DC. The impact of coexisting COPD on survival of patients with early-stage non-small cell lung cancer undergoing surgical resection. Chest 2014;145:346–353. |
| 11. | Turner MC, Chen Y, Krewski D, Calle EE, Thun MJ. Chronic obstructive pulmonary disease is associated with lung cancer mortality in a prospective study of never smokers. Am J Respir Crit Care Med 2007;176:285–290. |
| 12. | Goc J, Germain C, Vo-Bourgais TKD, Lupo A, Klein C, Knockaert S, et al. Dendritic cells in tumor-associated tertiary lymphoid structures signal a Th1 cytotoxic immune contexture and license the positive prognostic value of infiltrating CD8+ T cells. Cancer Res 2014;74:705–715. |
| 13. | Fourcade J, Sun Z, Benallaoua M, Guillaume P, Luescher IF, Sander C, et al. Upregulation of Tim-3 and PD-1 expression is associated with tumor antigen-specific CD8+ T cell dysfunction in melanoma patients. J Exp Med 2010;207:2175–2186. |
| 14. | Thommen DS, Schreiner J, Müller P, Herzig P, Roller A, Belousov A, et al. Progression of lung cancer is associated with increased dysfunction of T cells defined by coexpression of multiple inhibitory receptors. Cancer Immunol Res 2015;3:1344–1355. |
| 15. | Brahmer JR, Tykodi SS, Chow LQM, Hwu W-J, Topalian SL, Hwu P, et al. Safety and activity of anti-PD-L1 antibody in patients with advanced cancer. N Engl J Med 2012;366:2455–2465. |
| 16. | Eggermont AMM, Chiarion-Sileni V, Grob J-J, Dummer R, Wolchok JD, Schmidt H, et al. Adjuvant ipilimumab versus placebo after complete resection of high-risk stage III melanoma (EORTC 18071): a randomised, double-blind, phase 3 trial. Lancet Oncol 2015;16:522–530. |
| 17. | Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, et al. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N Engl J Med 2012;366:2443–2454. |
| 18. | Daud AI, Wolchok JD, Robert C, Hwu W-J, Weber JS, Ribas A, et al. Programmed death-ligand 1 expression and response to the anti-programmed death 1 antibody pembrolizumab in melanoma. J Clin Oncol 2016;34:4102–4109. |
| 19. | Rizvi NA, Hellmann MD, Snyder A, Kvistborg P, Makarov V, Havel JJ, et al. Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science 2015;348:124–128. |
| 20. | Borghaei H, Paz-Ares L, Horn L, Spigel DR, Steins M, Ready NE, et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N Engl J Med 2015;373:1627–1639. |
| 21. | Reck M, Rodríguez-Abreu D, Robinson AG, Hui R, Csőszi T, Fülöp A, et al.; KEYNOTE-024 Investigators. Pembrolizumab versus chemotherapy for PD-L1-positive non-small-cell lung cancer. N Engl J Med 2016;375:1823–1833. |
| 22. | Prat A, Navarro A, Paré L, Reguart N, Galván P, Pascual T, et al. Immune-related gene expression profiling after PD-1 blockade in non-small cell lung carcinoma, head and neck squamous cell carcinoma, and melanoma. Cancer Res 2017;77:3540–3550. |
| 23. | Rabe KF, Hurd S, Anzueto A, Barnes PJ, Buist SA, Calverley P, et al.; Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 2007;176:532–555. |
| 24. | de Chaisemartin L, Goc J, Damotte D, Validire P, Magdeleinat P, Alifano M, et al. Characterization of chemokines and adhesion molecules associated with T cell presence in tertiary lymphoid structures in human lung cancer. Cancer Res 2011;71:6391–6399. |
| 25. | Guo G, Sun X, Chen C, Wu S, Huang P, Li Z, et al. Whole-genome and whole-exome sequencing of bladder cancer identifies frequent alterations in genes involved in sister chromatid cohesion and segregation. Nat Genet 2013;45:1459–1463. |
| 26. | Mlecnik B, Bindea G, Angell HK, Maby P, Angelova M, Tougeron D, et al. Integrative analyses of colorectal cancer show Immunoscore is a stronger predictor of patient survival than microsatellite instability. Immunity 2016;44:698–711. |
| 27. | Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJM, Robert L, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 2014;515:568–571. |
| 28. | Thompson ED, Zahurak M, Murphy A, Cornish T, Cuka N, Abdelfatah E, et al. Patterns of PD-L1 expression and CD8 T cell infiltration in gastric adenocarcinomas and associated immune stroma. Gut 2017;66:794–801. |
| 29. | Govindan R, Ding L, Griffith M, Subramanian J, Dees ND, Kanchi KL, et al. Genomic landscape of non-small cell lung cancer in smokers and never-smokers. Cell 2012;150:1121–1134. |
| 30. | Le Calvez F, Mukeria A, Hunt JD, Kelm O, Hung RJ, Tanière P, et al. TP53 and KRAS mutation load and types in lung cancers in relation to tobacco smoke: distinct patterns in never, former, and current smokers. Cancer Res 2005;65:5076–5083. |
| 31. | Mansuet-Lupo A, Alifano M, Pécuchet N, Biton J, Becht E, Goc J, et al. Intratumoral immune cell densities are associated with lung adenocarcinoma gene alterations. Am J Respir Crit Care Med 2016;194:1403–1412. |
| 32. | Dong Z-Y, Zhong W-Z, Zhang X-C, Su J, Xie Z, Liu S-Y, et al. Potential predictive value of TP53andKRAS mutation status for response to PD-1 blockade immunotherapy in lung adenocarcinoma. Clin Cancer Res 2017;23:3012–3024. |
| 33. | Skoulidis F, Byers LA, Diao L, Papadimitrakopoulou VA, Tong P, Izzo J, et al. Co-occurring genomic alterations define major subsets of KRAS-mutant lung adenocarcinoma with distinct biology, immune profiles, and therapeutic vulnerabilities. Cancer Discov 2015;5:860–877. |
| 34. | Agata Y, Kawasaki A, Nishimura H, Ishida Y, Tsubata T, Yagita H, et al. Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes. Int Immunol 1996;8:765–772. |
| 35. | Austin JW, Lu P, Majumder P, Ahmed R, Boss JM. STAT3, STAT4, NFATc1, and CTCF regulate PD-1 through multiple novel regulatory regions in murine T cells. J Immunol 2014;192:4876–4886. |
| 36. | Bhowmik A, Seemungal TA, Sapsford RJ, Wedzicha JA. Relation of sputum inflammatory markers to symptoms and lung function changes in COPD exacerbations. Thorax 2000;55:114–120. |
| 37. | Brusselle GG, Joos GF, Bracke KR. New insights into the immunology of chronic obstructive pulmonary disease. Lancet 2011;378:1015–1026. |
| 38. | McKendry RT, Spalluto CM, Burke H, Nicholas B, Cellura D, Al-Shamkhani A, et al. Dysregulation of antiviral function of CD8(+) T cells in the chronic obstructive pulmonary disease lung. Role of the PD-1-PD-L1 axis. Am J Respir Crit Care Med 2016;193:642–651. |
| 39. | Taube JM, Anders RA, Young GD, Xu H, Sharma R, McMiller TL, et al. Colocalization of inflammatory response with B7-h1 expression in human melanocytic lesions supports an adaptive resistance mechanism of immune escape. Sci Transl Med 2012;4:127ra37. |
| 40. | Schmidt LH, Kümmel A, Görlich D, Mohr M, Bröckling S, Mikesch JH, et al. PD-1 and PD-L1 expression in NSCLC indicate a favorable prognosis in defined subgroups. PLoS One 2015;10:e0136023. |
| 41. | Chen YB, Mu C-Y, Huang J-A. Clinical significance of programmed death-1 ligand-1 expression in patients with non-small cell lung cancer: a 5-year-follow-up study. Tumori 2012;98:751–755. |
| 42. | Cooper WA, Tran T, Vilain RE, Madore J, Selinger CI, Kohonen-Corish M, et al. PD-L1 expression is a favorable prognostic factor in early stage non-small cell carcinoma. Lung Cancer 2015;89:181–188. |
| 43. | Velcheti V, Schalper KA, Carvajal DE, Anagnostou VK, Syrigos KN, Sznol M, et al. Programmed death ligand-1 expression in non-small cell lung cancer. Lab Invest 2014;94:107–116. |
| 44. | Mu C-Y, Huang J-A, Chen Y, Chen C, Zhang X-G. High expression of PD-L1 in lung cancer may contribute to poor prognosis and tumor cells immune escape through suppressing tumor infiltrating dendritic cells maturation. Med Oncol 2011;28:682–688. |
| 45. | Xu C, Fillmore CM, Koyama S, Wu H, Zhao Y, Chen Z, et al. Loss of Lkb1 and Pten leads to lung squamous cell carcinoma with elevated PD-L1 expression. Cancer Cell 2014;25:590–604. |
| 46. | Parsa AT, Waldron JS, Panner A, Crane CA, Parney IF, Barry JJ, et al. Loss of tumor suppressor PTEN function increases B7-H1 expression and immunoresistance in glioma. Nat Med 2007;13:84–88. |
| 47. | Daud AI, Loo K, Pauli ML, Sanchez-Rodriguez R, Sandoval PM, Taravati K, et al. Tumor immune profiling predicts response to anti-PD-1 therapy in human melanoma. J Clin Invest 2016;126:3447–3452. |
| 48. | Mark NM, Kargl J, Busch SE, Yang GHY, Metz HE, Zhang H, et al. Chronic obstructive pulmonary disease alters immune cell composition and immune checkpoint inhibitor efficacy in non–small cell lung cancer. Am J Respir Crit Care Med 2018;197:325–336. |
| 49. | Galon J, Mlecnik B, Bindea G, Angell HK, Berger A, Lagorce C, et al. Towards the introduction of the “Immunoscore” in the classification of malignant tumours. J Pathol 2014;232:199–209. |
*These authors contributed equally to this work.
‡Current address: Joan and Sanford I. Weill Department of Medicine, Division of Gastroenterology and Hepatology, Department of Microbiology and Immunology, and The Jill Robert's Institute for Research in Inflammatory Bowel Disease, Weill Cornell Medicine, Cornell University, New York, New York.
Supported by the Institut National de la Santé et de la Recherche Médicale, Paris Descartes-Paris 5 University, Pierre et Marie Curie-Paris 6 University, Cancer Research for Personalized Medicine, LabEx Immuno-oncology, the Institut National du Cancer (2011-PLBIO-06-INSERM 6-1), and MedImmune.
Author Contributions: D.D. and R. Herbst designed and supervised the study. J.B., H.O., J.G., P.D.-M., and C.G. acquired immunohistochemical data. J.B. acquired flow cytometry data. H.S., R. Halpin, and T.C. acquired and analyzed whole-exome sequence experiments. J.B., H.O., C.B.-M., C.G., R. Herbst, and D.D. analyzed the data. J.B. and A.D. performed statistical analysis. H.O., A.M.-L., M.A., J.A., F.G., P.B.-R., L.F., and D.D. were responsible for clinical data. A.M.-L. and D.D. were responsible for pathologic data. J.B., R. Herbst, and D.D. interpreted data. J.B., D.D., and R. Herbst wrote the manuscript. M.A., H.O., N.R., P.-R.B., C.G., I.C., and M.-C.D.-N. revised the manuscript.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org.
Originally Published in Press as DOI: 10.1164/rccm.201706-1110OC on March 8, 2018
Author disclosures are available with the text of this article at www.atsjournals.org.
