ABSTRACT
The association between the spatial heterogeneity of programmed cell death ligand 1 (PD‐L1) expression and the efficacy of third‐generation epidermal growth factor receptor tyrosine kinase inhibitors (EGFR‐TKIs) in EGFR‐mutant non‐small cell lung cancer (NSCLC) remains elusive. This retrospective study analyzed data from 4171 NSCLC patients with EGFR‐sensitive mutations treated at Shanghai Chest Hospital from August 2019 to September 2023. Among them, 182 patients receiving third‐generation EGFR‐TKIs monotherapy as a first‐line treatment were enrolled. Patients were categorized by biopsy sites into primary lung lesions (n = 112) and metastatic lymph nodes (n = 70). PD‐L1 expression was stratified based on tumor cell proportion score (TPS): < 1%, 1%–49%, and ≥ 50%. The median progression‐free survival (PFS) for the entire cohort was 18.33 months. In the PD‐L1 TPS group, PFS was 18.87 months for TPS < 1%, 17.6 months for TPS 1%–49%, and 13.6 months for TPS ≥ 50%, with significant differences across groups (p = 0.026). Moreover, multivariate analysis identified smoking history [HR = 1.653, 95% CI (1.132–2.414), p = 0.009] and TPS ≥ 50% [HR = 2.069, 95% CI (1.183–3.618), p = 0.011] as independent risk factors. In primary lesions, the median PFS was 21.93 months for TPS < 1%, 18.57 months for TPS 1%–49%, and 10.17 months for TPS ≥ 50%, with significant differences (p < 0.001). However, PD‐L1 expression in metastatic lymph nodes was not associated with PFS (p = 0.973). In advanced EGFR‐mutant NSCLC, high PD‐L1 expression may suggest reduced efficacy of third‐generation EGFR‐TKIs. The spatial heterogeneity of PD‐L1 expression could influence its predictive accuracy for third‐generation EGFR‐TKI efficacy.
Keywords: EGFR, heterogeneity, NSCLC, PD‐L1, third‐generation EGFR‐TKIs
In advanced EGFR‐mutant NSCLC, high PD‐L1 expression may indicate reduced efficacy of third‐generation EGFR‐TKIs. PD‐L1 expression in primary lesions may serve as a prognostic marker, whereas expression in metastatic lymph nodes does not. This spatial heterogeneity could affect PD‐L1's predictive accuracy for third‐generation EGFR‐TKIs efficacy.
Abbreviations
- ARMS
amplification refractory mutation system
- CI
confidence interval
- CR
complete response
- CT
computed tomography
- DCR
disease control rate
- ECT
emission computed tomography
- EGFR
epidermal growth factor receptor
- EMT
epithelial‐mesenchymal transition
- IASLC
International Association for the Study of Lung Cancer
- ICIs
immune checkpoint inhibitors
- LUAD
lung adenocarcinoma
- MHC‐I
MHC class I molecules
- MRI
magnetic resonance imaging
- NGS
next‐generation sequencing
- NR
not reached
- NSCLC
non‐small cell lung cancer
- ORR
objective response rate
- OS
overall survival
- PD
progressive disease
- PD‐1
programmed death 1
- PD‐L1
programmed cell death‐ligand 1
- PFS
progression‐free survival
- PR
partial response
- PS
performance status
- RECIST v1.1
response evaluation criteria in solid tumors version 1.1
- SD
stable disease
- TILs
tumor‐infiltrating lymphocytes
- TIME
tumor immune microenvironment
- TKIs
tyrosine kinase inhibitors
- TPS
tumor cell proportion score
- Tregs
regulatory T cells
1. Introduction
As is well documented, lung cancer is one of the leading causes of cancer‐related mortality worldwide, with non‐small cell lung cancer (NSCLC) accounting for approximately 80%–85% of cases. Among NSCLC patients, lung adenocarcinoma (LUAD) is the most prevalent pathological subtype and exhibits a high frequency of oncogenic driver mutations. Approximately 11% of Caucasian and 50% of Asian LUAD patients harbor epidermal growth factor receptor (EGFR) mutations [1, 2], primarily involving exon 19 deletions (19del) and exon 21 L858R point mutation [3]. Compared to chemotherapy, first‐ through third‐generation tyrosine kinase inhibitors (TKIs) targeting EGFR mutations have significantly prolonged progression‐free survival (PFS) and are currently established as the standard first‐line treatment for advanced EGFR‐sensitive NSCLC [4], with third‐generation EGFR‐TKI monotherapy being the preferred regimen [5]. While resistance to targeted therapies is inevitable, substantial variations in PFS were observed even among patients receiving identical third‐generation TKIs [6].
Programmed cell death‐ligand 1 (PD‐L1), as the primary ligand of programmed death 1 (PD‐1), is a critical immune checkpoint molecule expressed on the surface of tumor cells. PD‐L1 expression levels in tumor cells are currently recommended as a biomarker for predicting the response of driver gene‐negative LUAD patients to immune checkpoint inhibitors (ICIs) [7, 8]. Compared to patients without driver gene mutations, PD‐L1 expression levels in EGFR‐mutant NSCLC patients are typically lower [9]. However, the potential of PD‐L1 expression levels in EGFR‐mutant patients to predict EGFR‐TKI efficacy remains inconclusive, with inconsistencies across studies likely attributable to PD‐L1 expression heterogeneity [10, 11, 12, 13, 14].
PD‐L1 expression in NSCLC frequently exhibits spatial heterogeneity, with differing expression levels across various tumor sites [15]. Numerous studies have identified discrepancies in PD‐L1 expression levels between primary lung lesions and metastatic locations, as well as among various metastatic sites [15, 16, 17, 18]. For instance, Moutafi et al. reported that PD‐L1 expression levels were generally higher in metastatic lymph nodes, pleural effusions, and adrenal glands, whereas they were relatively lower in bone, brain, and liver metastases [17]. Prior research has predominantly focused on the heterogeneity of PD‐L1 expression in driver gene‐negative NSCLC patients and its potential implications for immunotherapy [15]. However, research examining the relationship between the heterogeneity of PD‐L1 expression and the efficacy of third‐generation EGFR‐TKI treatment in advanced EGFR‐mutant NSCLC patients is scarce.
Thus, this study aimed to investigate the relationship between PD‐L1 expression in primary lung tumors and metastatic lymph nodes and treatment outcomes in patients with advanced EGFR‐mutant NSCLC undergoing first‐line therapy with third‐generation EGFR TKIs.
2. Materials and Methods
2.1. Patients
In the present study, the medical records of 4171 patients diagnosed with EGFR mutations of NSCLC at Shanghai Chest Hospital between August 2019 and September 2023 were reviewed. The inclusion criteria were as follows: (I) patients with advanced lung adenocarcinoma diagnosed as stage IIIB‐IV in accordance with the 8th edition International Association for the Study of Lung Cancer (IASLC) TNM classification system; (II) Patients harboring EGFR‐sensitive mutations; (III) Tissue samples obtained from primary lung lesions or metastatic lymph nodes at the time of initial diagnosis; (IV) PD‐L1 testing conducted on tumor tissues at the time of initial diagnosis; and (V) first‐line treatment with third‐generation EGFR‐TKI monotherapy. The exclusion criteria were as follows: patients who did not undergo PD‐L1 testing, those who received first‐ or second‐generation EGFR‐TKI monotherapy, individuals undergoing targeted combination therapy, and those with incomplete data (see Figure 1).
FIGURE 1.
Flow chart of the study. NSCLC, non‐small cell lung cancer; EGFR, epidermal growth factor receptor; PD‐L1, programmed cell death‐ligand 1; TKIs, tyrosine kinase inhibitors.
This retrospective study was approved by the Ethics Committee of Shanghai Chest Hospital (approved ID: IS24111) and was conducted in accordance with the Declaration of Helsinki (2008 revision). The requirement for informed consent from patients was waived.
2.2. Clinical Assessments
Prior to the initiation of targeted therapy, all included patients were staged according to the 8th edition of the IASLC TNM classification system. Patients received third‐generation EGFR‐TKI monotherapy as first‐line treatment. Throughout the treatment course, chest computed tomography (CT) and abdominal ultrasound were performed every 2 to 3 months, with additional cranial magnetic resonance imaging (MRI) and bone emission computed tomography (ECT) conducted as necessary. These assessments were continued until disease progression, treatment discontinuation, or the final follow‐up. Treatment responses were evaluated using the Response Evaluation Criteria in Solid Tumors (RECIST, version 1.1) and were categorized into complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). Assessment endpoints included progression‐free survival (PFS), objective response rate (ORR), and disease control rate (DCR). Given that the majority of patients were treated with targeted therapy within the last 3 years, overall survival (OS) was excluded as an observational endpoint due to insufficient maturation of data. The final follow‐up for this study was completed on August 31, 2024.
2.3. Gene Testing and PD‐L1 Expression Tumor Proportion Score (TPS)
All patients underwent tissue biopsy before initial treatment. Genetic testing was performed using either the amplification refractory mutation system (ARMS) or next‐generation sequencing (NGS), and data were recorded for EGFR mutation types as well as other genetic alterations. The expression levels of PD‐L1 in tumor cells were assessed by two experienced pathologists using the Tumor Cell Proportion Score (TPS) (DACO PD‐L1 IHC 22C3 pharmDx.) and categorized as TPS < 1% (negative expression), TPS 1%–49% (low expression), and TPS ≥ 50% (high expression).
2.4. Statistical Analysis
Statistical analyses were performed using SPSS version 24.0 (IBM, Armonk, NY) software. Age was treated as a continuous variable and presented as the median, whereas the remaining variables were categorical and expressed as counts and percentages. PFS was calculated from the initiation of third‐generation EGFR TKIs until disease progression, regimen change, death, or the last follow‐up, whichever occurred first. ORR was defined as the sum of the proportions of patients achieving CR and PR, whereas the DCR was defined as the sum of the proportions of patients achieving CR, PR, and SD. Baseline characteristics between groups were compared using t‐tests for continuous variables and either Pearson's chi‐square test or Fisher's exact test for categorical variables. The Kaplan–Meier method was employed to generate survival curves for PFS. Univariate and multivariate Cox regression analyses were conducted to identify factors associated with PFS in patients receiving third‐generation EGFR TKIs. A two‐sided p‐value of less than 0.05 was considered statistically significant.
3. Results
3.1. Characteristics for All Patients
After a median follow‐up of 25.1 months (95% Confidence Interval, CI: 24.33–36.43), a total of 182 patients with advanced NSCLC harboring EGFR mutations who received first‐line treatment with third‐generation EGFR‐TKI monotherapy were enrolled in this study. All recruited patients were diagnosed with adenocarcinoma. Their median age was 61 years, with females accounting for 53.8% (98/182) and nonsmokers comprising 72.5% (132/182) of the cohort. Among all enrolled patients, 97.8% (178/182) were stage IV, 1.6% (3/182) were stage IIIB, and 0.5% (1/182) were stage IIIC. At baseline, 63 patients (34.6%) had brain metastases, 130 patients (71.4%) had bone metastases, and 23 patients (12.6%) had liver metastases. All patients underwent genetic testing following the initial histopathological diagnosis. Among them, 94 patients (51.6%) harbored the EGFR 19del mutation, whereas the remaining 88 patients (48.4%) harbored the EGFR 21L858R mutation. Furthermore, PD‐L1 immunohistochemical analysis of tumor tissues at diagnosis revealed that 101 patients (55.5%) were classified in the TPS < 1% group, 60 patients (33%) in the TPS 1%–49% group, and 21 patients (11.5%) in the TPS ≥ 50% group. Baseline characteristics of all enrolled patients are presented in Table 1.
TABLE 1.
Characteristics of all patients.
| Characteristics | Overall |
|---|---|
| Age (mean) | 61 |
| Gender, n (%) | |
| Male | 84 (46.2%) |
| Female | 98 (53.8%) |
| Smoking history, n (%) | |
| Smoker | 50 (27.5%) |
| Nonsmoker | 132 (72.5%) |
| Pathologic diagnostic process, n (%) | |
| TBLB/TBB | 52 (28.5%) |
| TBNA | 26 (14.3%) |
| Biopsy of supraclavicular lymph node | 54 (29.7%) |
| Percutaneous lung biopsy | 50 (27.5%) |
| Biopsy sites, n (%) | |
| Lung | 112 (61.5%) |
| Lymph nodes | 70 (38.5%) |
| TNM stage, n (%) | |
| IIIB | 3 (1.6%) |
| IIIC | 1 (0.5%) |
| IV | 178 (97.8%) |
| Brain metastasis, n (%) | |
| Brain metastasis | 63 (34.6%) |
| No brain metastasis | 119 (65.4%) |
| Bone metastasis, n (%) | |
| Bone metastasis | 130 (71.4%) |
| No bone metastasis | 52 (28.6%) |
| Liver metastasis, n (%) | |
| Liver metastasis | 23 (12.6%) |
| No liver metastasis | 159 (87.4%) |
| EGFR mutation, n (%) | |
| 19del | 94 (51.6%) |
| 21L858R | 88 (48.4%) |
| TPS, n (%) | |
| TPS < 1% | 101 (55.5%) |
| TPS 1%–49% | 60 (33%) |
| TPS ≥ 50% | 21 (11.5%) |
| Best response, n (%) | |
| PR | 114 (62.6%) |
| SD | 65 (35.7%) |
| PD | 3 (1.6%) |
| Third‐generation EGFR‐TKIs, n (%) | |
| Osimertinib | 155 (85.2%) |
| Almonertinib | 11 (6.0%) |
| Furmonertinib | 16 (8.8%) |
3.2. Characteristics of the Primary Lung Lesion Group and the Metastatic Lymph Node Group
Based on the sites of pathological tissue sampling, patients were categorized into two groups, namely the primary lung lesion group (n = 112) and the metastatic lymph node group (n = 70). Demographic characteristics, TNM stage, distant metastases, EGFR mutation subtypes, performance status (PS) scores, and TPS were comparable between the two groups (see Table 2).
TABLE 2.
Characteristics of patients in different groups.
| Characteristics | Primary lung group (n = 112) | Lymph nodes group (n = 70) | p |
|---|---|---|---|
| Age (mean) | 62 | 61 | 0.372 |
| Gender, n (%) | 0.312 | ||
| Male | 55 (49.1%) | 29 (41.4%) | |
| Female | 57 (50.9%) | 41 (58.6%) | |
| Smoking history, n (%) | 0.674 | ||
| Smoker | 32 (28.6%) | 18 (25.7%) | |
| Nonsmoker | 80 (71.4%) | 52 (74.3%) | |
| TNM stage, n (%) | 0.265 | ||
| IIIB | 1 (0.9%) | 2 (2.9%) | |
| IIIC | 0 (0%) | 1 (1.4%) | |
| IV | 111 (99.1%) | 67 (95.7%) | |
| Brain metastasis, n (%) | 0.375 | ||
| Brain metastasis | 36 (32.1%) | 27 (38.6%) | |
| No brain metastasis | 76 (67.9%) | 43 (61.4%) | |
| Bone metastasis, n (%) | 0.312 | ||
| Bone metastasis | 77 (68.8%) | 53 (75.7%) | |
| No bone metastasis | 35 (31.2%) | 17 (24.3%) | |
| Liver metastasis, n (%) | 0.148 | ||
| Liver metastasis | 11 (9.8%) | 12 (17.1%) | |
| No liver metastasis | 101 (90.2%) | 58 (82.9%) | |
| EGFR mutation, n (%) | 0.511 | ||
| 19del | 60 (53.6%) | 34 (48.6%) | |
| 21L858R | 52 (46.4%) | 36 (51.4%) | |
| TPS, n (%) | 0.565 | ||
| < 1% | 65 (58%) | 36 (51.4%) | |
| 1%–49% | 36 (32.1%) | 24 (34.3%) | |
| ≥ 50% | 11 (9.8%) | 10 (14.3%) | |
| PS score, n (%) | 0.182 | ||
| PS = 0 | 12 (10.7%) | 6 (8.6%) | |
| PS = 1 | 100 (89.3%) | 62 (88.6%) | |
| PS = 2 | 0 (0%) | 2 (2.9%) |
In the primary lung lesion group, patients were further stratified based on PD‐L1 expression into three subgroups: TPS < 1%, 1%–49%, and ≥ 50%. A comparative analysis of demographic characteristics, TNM stage, EGFR mutation subtypes, and distant metastases was performed among the three subgroups. The analysis revealed a significant difference in only the distribution of EGFR mutation subtypes, whereas the remaining variables were similar across all subgroups (see Table 3).
TABLE 3.
Characteristics of patients in the primary lung group.
| Characteristics | TPS < 1% (n = 65) | TPS 1%–49% (n = 36) | TPS ≥ 50% (n = 11) | p |
|---|---|---|---|---|
| Age (mean) | 62 | 62 | 60 | 0.797 |
| Gender, n (%) | 0.773 | |||
| Male | 33 (50.8%) | 16 (44.4%) | 6 (54.5%) | |
| Female | 32 (49.2%) | 20 (55.6%) | 5 (45.5%) | |
| Smoking history, n (%) | 0.307 | |||
| Smoker | 22 (33.8%) | 7 (19.4%) | 3 (27.3%) | |
| Nonsmoker | 43 (66.2%) | 29 (80.6%) | 8 (72.7%) | |
| EGFR mutation, n (%) | 0.040 | |||
| 19del | 29 (44.6%) | 22 (61.1%) | 9 (81.8%) | |
| 21L858R | 36 (55.4%) | 14 (38.9%) | 2 (18.2%) | |
| TNM stage, n (%) | 0.098 | |||
| III | 0 (0%) | 0 (0%) | 1 (9.1%) | |
| IV | 65 (100%) | 36 (100%) | 10 (90.9%) | |
| Brain metastasis, n (%) | 0.919 | |||
| Brain metastasis | 20 (30.8%) | 12 (33.3%) | 4 (36.4%) | |
| No brain metastasis | 45 (69.2%) | 24 (66.7%) | 7 (63.6%) | |
| Bone metastasis, n (%) | 0.272 | |||
| Bone metastasis | 43 (66.2%) | 28 (77.8%) | 6 (54.5%) | |
| No bone metastasis | 22 (33.8%) | 8 (22.2%) | 5 (45.5%) | |
| Liver metastasis, n (%) | 0.433 | |||
| Liver metastasis | 7 (10.8%) | 2 (5.6%) | 2 (18.2%) | |
| No liver metastasis | 58 (89.2%) | 34 (94.4%) | 9 (81.8%) | |
| mPFS | 21.93 | 18.57 | 10.17 | < 0.001 |
| ORR | 63.1% | 61.1% | 81.8% | 0.433 |
| DCR | 98.5% | 97.2% | 90.9% | 0.357 |
The same stratification method was applied to the metastatic lymph node group, wherein patients were similarly assigned to the TPS < 1%, 1%–49%, and ≥ 50% subgroups. Comparative analysis demonstrated that all comparative variables were similar across these subgroups (see Table 4).
TABLE 4.
Characteristics of patients in the lymph nodes group.
| Characteristics | TPS < 1% (n = 36) | TPS 1%–49% (n = 24) | TPS ≥ 50% (n = 10) | p |
|---|---|---|---|---|
| Age (mean) | 61 | 62 | 59 | 0.715 |
| Gender, n (%) | 0.995 | |||
| Male | 15 (41.7%) | 10 (41.7%) | 4 (40%) | |
| Female | 21 (58.3%) | 14 (58.3%) | 6 (60%) | |
| Smoking history, n (%) | 0.945 | |||
| Smoker | 9 (25%) | 6 (25%) | 3 (30%) | |
| Nonsmoker | 27 (75%) | 18 (75%) | 7 (70%) | |
| EGFR mutation, n (%) | 0.076 | |||
| 19del | 17 (47.2%) | 9 (37.5%) | 8 (80%) | |
| 21L858R | 19 (52.8%) | 15 (62.5%) | 2 (20%) | |
| TNM stage, n (%) | 0.660 | |||
| III | 1 (2.8%) | 2 (8.3%) | 0 (0%) | |
| IV | 35 (97.2%) | 22 (91.7%) | 10 (100%) | |
| Brain metastasis, n (%) | 0.315 | |||
| Brain metastasis | 13 (36.1%) | 8 (33.3%) | 6 (60%) | |
| No brain metastasis | 23 (63.9%) | 16 (66.7%) | 4 (40%) | |
| Bone metastasis, n (%) | 0.323 | |||
| Bone metastasis | 28 (77.8%) | 16 (66.7%) | 9 (90%) | |
| No bone metastasis | 8 (22.2%) | 8 (33.3%) | 1 (10%) | |
| Liver metastasis, n (%) | 0.271 | |||
| Liver metastasis | 7 (19.4%) | 2 (8.3%) | 3 (30%) | |
| No liver metastasis | 29 (80.6%) | 22 (91.7%) | 7 (70%) | |
| mPFS | 17.0 | 16.05 | 15.95 | 0.973 |
| ORR | 61.1% | 62.5% | 50% | 0.780 |
| DCR | 100% | 100% | 100% | 1.000 |
3.3. High PD‐L1 Expression Predicted Poor Efficacy of Third‐Generation EGFR‐TKIs
By the final follow‐up, 70.9% (129/182) of patients had experienced disease progression following first‐line therapy. The ORR for the TPS < 1%, 1%–49%, and ≥ 50% groups was 62.4%, 61.7%, and 66.7%, respectively, while the DCR was 99.0%, 98.3%, and 95.2%, respectively. Interestingly, no statistically significant differences were noted in ORR or DCR among the three groups (see Figure 2).
FIGURE 2.
Response rate across different groups. (A) ORR for all patients, patients in the primary lung lesion group, and those in the lymph node group. No statistically significant difference was observed across the three groups. (B) DCR for all patients, patients in the primary lung lesion group, and those in the lymph node group. No statistically significant difference was observed among the three groups. DCR, disease control rate; LN, lymph node; ORR, objective response rate; TPS, tumor cell proportion score.
The median PFS for all patients was 18.33 months (95% CI: 15.77–20.43). Specifically, the median PFS for the TPS < 1%, 1%–49%, and ≥ 50% groups was 18.87 months (95% CI: 17.00–22.87), 17.6 months (95% CI: 12.83–23.27), and 13.6 months (8.83—Not Reached, NR), respectively, with statistically significant differences among the three groups (p = 0.026). Additionally, post hoc analysis unveiled a significant difference in median PFS between the TPS < 1% and TPS ≥ 50% groups (p = 0.03), whereas no statistically significant differences were observed between the TPS < 1% and TPS 1%–49% groups (p = 0.503) or between the TPS 1%–49% and TPS ≥ 50% groups (p = 0.392), as illustrated in Figure 3A.
FIGURE 3.
Kaplan–Meier estimates of PFS. (A) PFS for all patients. A statistically significant difference in median PFS was observed among the three groups. (B) PFS in the primary lung lesion group. A statistically significant difference in median PFS was noted among the three groups. (C) PFS in the lymph node group. No statistically significant differences in median PFS were observed among the three groups. OS, overall survival; PFS, progression‐free survival; TPS, tumor cell proportion score.
On the one hand, univariate analysis of PFS indicated that age, gender, EGFR mutation subtype, brain metastasis, and bone metastasis were not significantly associated with PFS. On the other hand, a history of smoking [HR = 1.522, 95% CI (1.049–2.209), p = 0.027], liver metastasis [HR = 1.628, 95% CI (1.018–2.602), p = 0.042], and TPS ≥ 50% [HR = 2.034, 95% CI (1.187–3.484), p = 0.01] were associated with shorter median PFS. Meanwhile, multivariate analysis of PFS identified a history of smoking [HR = 1.653, 95% CI (1.132–2.414), p = 0.009] and TPS ≥ 50% [HR = 2.069, 95% CI (1.183–3.618), p = 0.011] as independent prognostic factors for patients with advanced EGFR‐mutant NSCLC receiving third‐generation EGFR TKIs (see Figure 4).
FIGURE 4.
Univariate Cox regression analysis of PFS. EGFR, epidermal growth factor receptor; PFS, progression‐free survival; TPS, tumor cell proportion score.
3.4. PD‐L1 Expression in Primary Lung Lesions Correlated With the Efficacy of Third‐Generation EGFR TKIs
In the primary lung lesion group, the ORR for the TPS < 1%, 1%–49%, and ≥ 50% groups was 63.1%, 61.1%, and 81.8%, respectively, whereas the DCR was 98.5%, 97.2%, and 90.9%, respectively. As anticipated, no statistically significant differences were observed in ORR or DCR among these groups (see Table 3; Figure 2). Following treatment with TKIs, the median PFS for the three groups was 21.93 months (95% CI: 17.17–28.43) for the TPS < 1% group, 18.57 months (95% CI: 12.73–24.67) for the TPS 1%–49% group, and 10.17 months (95% CI: 4.83—NR) for the TPS ≥ 50% group. Importantly, significant differences were identified in the median PFS across the three groups (p < 0.001). Further post hoc analysis uncovered significant differences in median PFS between the TPS < 1% and TPS ≥ 50% groups (p < 0.001) and between the TPS 1%–49% and TPS ≥ 50% groups (p = 0.047), whereas no significant difference was observed between the TPS < 1% and TPS 1%–49% groups (see Figure 3A). Univariate Cox regression analysis indicated that median PFS [HR = 3.856, 95% CI (1.803–8.246), p < 0.001] was significantly shorter in patients with TPS ≥ 50% in primary lung lesions compared to the TPS < 1% group, whereas TPS 1%–49% did not significantly impact PFS (see Table 5; Figure 4).
TABLE 5.
Association of PFS with PD‐L1 expression.
| Characteristics | Hazard ratio (95% CI) for PFS | p |
|---|---|---|
| Primary lung group | ||
| TPS < 1% | Reference | |
| TPS 1%–49% | 1.548 (0.943–2.541) | 0.084 |
| TPS ≥ 50% | 3.989 (1.884–8.447) | < 0.001 |
| Lymph nodes group | ||
| TPS < 1% | Reference | |
| TPS 1%–49% | 1.057 (0.564–1.982) | 0.862 |
| TPS ≥ 50% | 1.088 (0.489–2.420) | 0.836 |
3.5. PD‐L1 Expression in Metastatic Lymph Nodes Was Not Associated With the Efficacy of Third‐Generation EGFR‐TKIs
In the metastatic lymph node group, the ORR for the TPS < 1%, 1%–49%, and ≥ 50% groups was 61.1%, 62.5%, and 50%, respectively, with all three groups exhibiting a DCR of 100%. Of note, no statistically significant differences were observed in ORR or DCR among these three groups (see Table 4; Figure 2). After treatment with TKIs, the median PFS was 17 months (95% CI: 12.5–21.93) for the TPS < 1% group, 16.05 months (95% CI: 11.83‐NR) for the TPS 1%–49% group, and 15.95 months (95% CI: 10.27—NR) for the TPS ≥ 50% group, with no statistically significant differences observed among these groups (p = 0.973; see Figure 3C). Furthermore, no significant correlation was identified between PD‐L1 expression in lymph nodes and PFS in patients receiving third‐generation EGFR TKIs (see Table 5; Figure 4).
3.6. Baseline Co‐Mutations
Among the 182 enrolled patients, 126 underwent NGS testing before targeted therapy. Among them, 83.3% (105/126) harbored co‐mutations, with the highest prevalence of TP53 mutations at 63.5% (80/126), followed by BRCA mutations (6.3%), ATM mutations (6.3%), and RB1 mutations (5.6%). In the TP53 mutation group, 58.75% (n = 47) of patients had a TPS < 1%, 30% (n = 24) had a TPS of 1%–49%, and 11.25% (n = 9) had a TPS ≥ 50%. Statistical analysis revealed no significant correlation between the TP53 mutation and PD‐L1 expression (p = 0.997). Furthermore, no significant difference in median PFS was observed between the TP53 mutation group and the wild type group (median PFS: 20.3 months vs. 16.2 months, p = 0.238).
3.7. Changes of PD‐L1 Expression After Third‐Generation EGFR‐TKI Therapy
Among the 129 patients who experienced disease progression after first‐line therapy with third‐generation EGFR TKIs, 49 underwent re‐biopsy and repeat PD‐L1 testing. Biopsy sites included lung tissue in 29 cases, metastatic lymph nodes in 13 cases, pleural and ascitic fluid in 5 cases, and metastatic liver tissue in 2 cases. Among these 49 patients, 24 had PD‐L1 expression data collected from the same organ both before and after treatment. The results showed no statistically significant difference in PD‐L1 expression in these patients before and after receiving third‐generation targeted therapy (p = 0.325). However, changes in PD‐L1 expression were noted in 13 patients posttreatment, with 7 exhibiting upregulation and 6 showing downregulation (see Figure 5).
FIGURE 5.
Paired analysis of PD‐L1 expression in same‐site biopsies before and after third‐generation EGFR‐TKIs treatment. No statistically significant difference was observed in PD‐L1 expression before and after treatment. EGFR‐TKIs, epidermal growth factor receptor‐tyrosine kinase inhibitors; PD‐L1, programmed cell death‐ligand 1.
4. Discussion
The present study aimed to investigate the association between PD‐L1 expression levels at various sampling sites and the efficacy of first‐line monotherapy with third‐generation EGFR‐TKIs in patients with advanced NSCLC harboring EGFR mutations. Our findings demonstrated that high PD‐L1 expression in primary lung lesions was significantly associated with shorter PFS in patients treated with third‐generation EGFR‐TKIs (p < 0.001). In contrast, no correlation was observed between PD‐L1 expression in metastatic lymph nodes and the efficacy of third‐generation EGFR‐TKIs.
PD‐L1 is frequently employed as a biomarker for predicting the efficacy of immunotherapy in driver gene‐negative NSCLC patients, with its expression generally positively correlated with treatment outcomes [7, 8]. Nonetheless, the relationship between PD‐L1 expression and the efficacy of third‐generation EGFR‐TKIs in advanced EGFR‐mutant NSCLC remains controversial. Several studies have concluded that high PD‐L1 expression was correlated with poorer efficacy of third‐generation EGFR‐TKIs [11, 19, 20, 21, 22], potentially attributable to EGFR‐mutant NSCLC cells with high PD‐L1 expression inducing epithelial–mesenchymal transition (EMT) via the activation of the TGF‐β/SMAD canonical signaling pathway, eventually conferring resistance to EGFR‐TKIs [23]. Additionally, activation of the PI3K pathway [24] and persistent activation of ERK signaling via the PD‐L1/BAG‐1 axis [25] may also play a decisive role in conferring PD‐L1‐mediated resistance to TKIs in EGFR‐mutant tumor cells. Conversely, other studies have reported that PD‐L1 expression does not significantly influence the therapeutic effect of third‐generation EGFR‐TKIs [13, 26, 27]. Our results were consistent with the former perspective, implying that patients with advanced EGFR‐mutant NSCLC receiving first‐line monotherapy with third‐generation EGFR‐TKIs had a significantly longer median PFS in the TPS < 1% group compared to the TPS ≥ 50% group. Notably, TPS ≥ 50% and a history of smoking emerged as independent prognostic factors for treatment outcomes with third‐generation EGFR‐TKIs. Further analysis signaled that high PD‐L1 expression in primary lung lesions is associated with diminished efficacy of third‐generation TKIs, while PD‐L1 expression in metastatic lymph nodes exhibited no correlation. This suggests that the spatial heterogeneity of PD‐L1 may account for the discrepancies observed in studies examining the relationship between PD‐L1 expression and the efficacy of EGFR‐TKIs.
It is worthwhile emphasizing that PD‐L1 expression generally exhibits spatial heterogeneity in NSCLC [18]. Hong et al. pointed out that PD‐L1 expression in metastatic lymph nodes is generally higher than that in primary lung lesions. Nevertheless, its expression in metastatic lymph nodes was not correlated with the efficacy of immunotherapy in NSCLC patients without driver gene mutations [15]. PD‐L1 expression in lung cancer is influenced by various factors, including inflammatory stimuli and oncogenic pathways, at the transcriptional, post‐transcriptional, and post‐translational levels [28]. Differences in the tumor immune microenvironment (TIME) between primary lung lesions and metastatic lymph nodes have been documented [29], potentially leading to distinct mechanisms of PD‐L1 upregulation. Compared to primary lung lesions, metastatic lymph nodes are enriched in regulatory T cells (Tregs) [30] that can enhance PD‐L1 expression in tumor cells by secreting substantial amounts of TGF‐β; PD‐L1 can, in turn, induce Tregs and amplify their immunosuppressive capabilities, creating a positive feedback loop [31, 32]. This interaction up‐regulates PD‐L1 expression in metastatic lymph nodes. Additionally, IFN‐γ released by tumor‐infiltrating lymphocytes (TILs), which are abundant in metastatic lymph node tissue, plays a critical role in inducing PD‐L1 expression in tumor cells [33]. Therefore, PD‐L1 expression in metastatic lymph nodes may be mediated by secondary modifications, potentially limiting its use as a predictor for the efficacy of third‐generation EGFR‐TKI therapy. In contrast, PD‐L1 expression in primary lung lesions may possess greater predictive value.
In addition to spatial heterogeneity, PD‐L1 expression in lung cancer exhibits temporal heterogeneity and is affected by treatments [15, 34]. The effect of third‐generation EGFR‐TKIs on PD‐L1 expression in EGFR‐mutant tumor cells remains to be elucidated. Earlier studies evinced that the third‐generation EGFR‐TKI osimertinib downregulated PD‐L1 expression by reducing mRNA levels and triggering protein degradation [35]. Conversely, PD‐L1 may be upregulated following the development of resistance to third‐generation EGFR‐TKIs, potentially linked to the upregulation of MHC class I molecules (MHC‐I) posttreatment, which could enhance CD8+ cytotoxic T cell activity in certain patients [36]. Herein, 24 patients had tissue samples available from the same organ before and after TKI treatment, allowing for the observation of changes in PD‐L1 expression. The results showed no statistically significant difference in PD‐L1 expression before and after treatment within the same organ. However, PD‐L1 expression was upregulated in 7 patients and downregulated in 6 patients. Taken together, these findings suggest that TKI therapy may alter PD‐L1 expression, highlighting the complexity of the underlying mechanisms.
Nevertheless, this study has several limitations that merit acknowledgment. To begin, the relatively small sample size may have impacted the analytical power and reliability of our results. Secondly, this retrospective study was subject to inherent biases that could have compromised the accuracy of our findings. Finally, the short follow‐up duration resulted in an insufficient number of mortality events, limiting our ability to conduct a meaningful OS analysis.
In summary, this study demonstrated that high PD‐L1 expression is associated with poor prognosis in advanced NSCLC patients harboring EGFR mutations receiving first‐line treatment with third‐generation EGFR‐TKIs. Notably, PD‐L1 expression levels in the primary lung lesions were correlated with treatment efficacy, whereas expression levels in metastatic lymph nodes were not. The spatial heterogeneity of PD‐L1 expression may significantly impact the accuracy in predicting the efficacy of third‐generation EGFR‐TKIs.
Author Contributions
Yidan Zhang: data curation, formal analysis, methodology, validation, writing – original draft. Yingqi Xu: data curation, investigation, software, visualization, writing – original draft. Hongping Jin: data curation, formal analysis, investigation, visualization, writing – original draft. Tengfei Liu: investigation, software, visualization. Hua Zhong: formal analysis, methodology, validation. Jianlin Xu: funding acquisition, methodology, project administration, resources, writing – review and editing. Yuqing Lou: funding acquisition, project administration, resources, validation, writing – review and editing. Runbo Zhong: conceptualization, project administration, resources, supervision, writing – review and editing.
Ethics Statement
This retrospective study was approved by the Ethics Committee of Shanghai Chest Hospital (ID: IS24111) and was conducted in accordance with the Declaration of Helsinki ((revised in 2013)), waiving the requirement for informed consent from the patients.
Consent
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
We would like to thank all of the investigators for their involvement in this study.
Funding: This research was supported by the National Natural Science Foundation of China (grant number 82272665) and the Talent training plan of Shanghai Chest Hospital (to Yuqing Lou).
Yidan Zhang, Yingqi Xu and Hongping Jin contributed equally to this work.
Jianlin Xu, Yuqing Lou and Runbo Zhong are co‐corresponding authors.
Contributor Information
Yuqing Lou, Email: louyq@hotmail.com.
Runbo Zhong, Email: tonic_chung@139.com.
References
- 1. Siegel R. L., Giaquinto A. N., and Jemal A., “Cancer Statistics, 2024,” CA: A Cancer Journal for Clinicians 74, no. 1 (2024): 12–49, 10.3322/caac.21820. [DOI] [PubMed] [Google Scholar]
- 2. Yu L., Hu Y., Xu J., et al., “Multi‐Target Angiogenesis Inhibitor Combined With PD‐1 Inhibitors May Benefit Advanced Non‐Small Cell Lung Cancer Patients in Late Line After Failure of EGFR‐TKI Therapy,” International Journal of Cancer 153, no. 3 (2023): 635–643, 10.1002/ijc.34536. [DOI] [PubMed] [Google Scholar]
- 3. Shigematsu H., Lin L., Takahashi T., et al., “Clinical and Biological Features Associated With Epidermal Growth Factor Receptor Gene Mutations in Lung Cancers,” JNCI Journal of the National Cancer Institute 97, no. 5 (2005): 339–346, 10.1093/jnci/dji055. [DOI] [PubMed] [Google Scholar]
- 4. Meyer M. L., Fitzgerald B. G., Paz‐Ares L., et al., “New Promises and Challenges in the Treatment of Advanced Non‐Small‐Cell Lung Cancer,” Lancet 404, no. 10454 (2024): 803–822, 10.1016/S0140-6736(24)01029-8. [DOI] [PubMed] [Google Scholar]
- 5. Hendriks L. E., Kerr K. M., Menis J., et al., “Oncogene‐Addicted Metastatic Non‐Small‐Cell Lung Cancer: ESMO Clinical Practice Guideline for Diagnosis, Treatment and Follow‐Up,” Annals of Oncology 34, no. 4 (2023): 339–357, 10.1016/j.annonc.2022.12.009. [DOI] [PubMed] [Google Scholar]
- 6. Shah M. P. and Neal J. W., “Targeting Acquired and Intrinsic Resistance Mechanisms in Epidermal Growth Factor Receptor Mutant Non‐Small‐Cell Lung Cancer,” Drugs 82, no. 6 (2022): 649–662, 10.1007/s40265-022-01698-z. [DOI] [PubMed] [Google Scholar]
- 7. Reck M., Rodriguez‐Abreu D., Robinson A. G., et al., “Pembrolizumab Versus Chemotherapy for PD‐L1‐Positive Non‐Small‐Cell Lung Cancer,” New England Journal of Medicine 375, no. 19 (2016): 1823–1833, 10.1056/NEJMoa1606774. [DOI] [PubMed] [Google Scholar]
- 8. Mok T. S. K., Wu Y. L., Kudaba I., et al., “Pembrolizumab Versus Chemotherapy for Previously Untreated, PD‐L1‐Expressing, Locally Advanced or Metastatic Non‐Small‐Cell Lung Cancer (KEYNOTE‐042): A Randomised, Open‐Label, Controlled, Phase 3 Trial,” Lancet 393, no. 10183 (2019): 1819–1830, 10.1016/S0140-6736(18)32409-7. [DOI] [PubMed] [Google Scholar]
- 9. Soo R. A., Lim S. M., Syn N. L., et al., “Immune Checkpoint Inhibitors in Epidermal Growth Factor Receptor Mutant Non‐Small Cell Lung Cancer: Current Controversies and Future Directions,” Lung Cancer 115 (2018): 12–20, 10.1016/j.lungcan.2017.11.009. [DOI] [PubMed] [Google Scholar]
- 10. Hsu K. H., Huang Y. H., Tseng J. S., et al., “High PD‐L1 Expression Correlates With Primary Resistance to EGFR‐TKIs in Treatment Naive Advanced EGFR‐Mutant Lung Adenocarcinoma Patients,” Lung Cancer 127 (2019): 37–43, 10.1016/j.lungcan.2018.11.021. [DOI] [PubMed] [Google Scholar]
- 11. Sakata Y., Sakata S., Oya Y., et al., “Osimertinib as First‐Line Treatment for Advanced Epidermal Growth Factor Receptor Mutation‐Positive Non‐Small‐Cell Lung Cancer in a Real‐World Setting (OSI‐FACT),” European Journal of Cancer 159 (2021): 144–153, 10.1016/j.ejca.2021.09.041. [DOI] [PubMed] [Google Scholar]
- 12. D'Incecco A., Andreozzi M., Ludovini V., et al., “PD‐1 and PD‐L1 Expression in Molecularly Selected Non‐Small‐Cell Lung Cancer Patients,” British Journal of Cancer 112, no. 1 (2015): 95–102, 10.1038/bjc.2014.555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Brown H., Vansteenkiste J., Nakagawa K., et al., “Programmed Cell Death Ligand 1 Expression in Untreated EGFR Mutated Advanced NSCLC and Response to Osimertinib Versus Comparator in FLAURA,” Journal of Thoracic Oncology 15, no. 1 (2020): 138–143, 10.1016/j.jtho.2019.09.009. [DOI] [PubMed] [Google Scholar]
- 14. Cho J. H., Zhou W., Choi Y. L., et al., “Retrospective Molecular Epidemiology Study of PD‐L1 Expression in Patients With EGFR‐Mutant Non‐Small Cell Lung Cancer,” Cancer Research and Treatment 50, no. 1 (2018): 95–102, 10.4143/crt.2016.591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hong L., Negrao M. V., Dibaj S. S., et al., “Programmed Death‐Ligand 1 Heterogeneity and Its Impact on Benefit From Immune Checkpoint Inhibitors in NSCLC,” Journal of Thoracic Oncology 15, no. 9 (2020): 1449–1459, 10.1016/j.jtho.2020.04.026. [DOI] [PubMed] [Google Scholar]
- 16. Schoenfeld A. J., Rizvi H., Bandlamudi C., et al., “Clinical and Molecular Correlates of PD‐L1 Expression in Patients With Lung Adenocarcinomas,” Annals of Oncology 31, no. 5 (2020): 599–608, 10.1016/j.annonc.2020.01.065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Moutafi M. K., Tao W., Huang R., et al., “Comparison of Programmed Death‐Ligand 1 Protein Expression Between Primary and Metastatic Lesions in Patients With Lung Cancer,” Journal for Immunotherapy of Cancer 9, no. 4 (2021): e002230, 10.1136/jitc-2020-002230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Wu J., Sun W., Yang X., et al., “Heterogeneity of Programmed Death‐Ligand 1 Expression and Infiltrating Lymphocytes in Paired Resected Primary and Metastatic Non‐Small Cell Lung Cancer,” Modern Pathology 35, no. 2 (2022): 218–227, 10.1038/s41379-021-00903-w. [DOI] [PubMed] [Google Scholar]
- 19. Hsu K. H., Tseng J. S., Yang T. Y., et al., “PD‐L1 Strong Expressions Affect the Clinical Outcomes of Osimertinib in Treatment Naive Advanced EGFR‐Mutant Non‐Small Cell Lung Cancer Patients,” Scientific Reports 12, no. 1 (2022): 9753, 10.1038/s41598-022-13102-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Yoshimura A., Yamada T., Okuma Y., et al., “Impact of Tumor Programmed Death Ligand‐1 Expression on Osimertinib Efficacy in Untreated EGFR‐Mutated Advanced Non‐Small Cell Lung Cancer: A Prospective Observational Study,” Translational Lung Cancer Research 10, no. 8 (2021): 3582–3593, 10.21037/tlcr-21-461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Lakkunarajah S., Truong P. T., Bone J. N., et al., “First‐Line Osimertinib for Patients With EGFR‐Mutated Advanced Non‐Small Cell Lung Cancer: Efficacy and Safety During the COVID‐19 Pandemic,” Translational Lung Cancer Research 12, no. 7 (2023): 1454–1465, 10.21037/tlcr-23-81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Papazyan T., Denis M. G., Sagan C., Raimbourg J., Herbreteau G., and Pons‐Tostivint E., “Impact of PD‐L1 Expression on the Overall Survival of Caucasian Patients With Advanced EGFR‐Mutant NSCLC Treated With Frontline Osimertinib,” Targeted Oncology 19, no. 4 (2024): 611–621, 10.1007/s11523-024-01072-x. [DOI] [PubMed] [Google Scholar]
- 23. Zhang Y., Zeng Y., Liu T., et al., “The Canonical TGF‐β/Smad Signalling Pathway Is Involved in PD‐L1‐Induced Primary Resistance to EGFR‐TKIs in EGFR‐Mutant Non‐Small‐Cell Lung Cancer,” Respiratory Research 20, no. 1 (2019): 164, 10.1186/s12931-019-1137-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Ding W., Yang P., Zhao X., et al., “Unraveling EGFR‐TKI Resistance in Lung Cancer With High PD‐L1 or TMB in EGFR‐Sensitive Mutations,” Respiratory Research 25, no. 1 (2024): 40, 10.1186/s12931-023-02656-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Lin P. L., Wu T. C., Wu D. W., Wang L., Chen C. Y., and Lee H., “An Increase in BAG‐1 by PD‐L1 Confers Resistance to Tyrosine Kinase Inhibitor in Non‐Small Cell Lung Cancer via Persistent Activation of ERK Signalling,” European Journal of Cancer 85 (2017): 95–105, 10.1016/j.ejca.2017.07.025. [DOI] [PubMed] [Google Scholar]
- 26. Teranishi S., Sugimoto C., Nagaoka S., et al., “Retrospective Analysis of Independent Predictors of Progression‐Free Survival in Patients With EGFR Mutation‐Positive Advanced Non‐Small Cell Lung Cancer Receiving First‐Line Osimertinib,” Thoracic Cancer 13, no. 19 (2022): 2741–2750, 10.1111/1759-7714.14608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Chamorro D. F., Cardona A. F., Rodriguez J., et al., “Genomic Landscape of Primary Resistance to Osimertinib Among Hispanic Patients With EGFR‐Mutant Non‐Small Cell Lung Cancer (NSCLC): Results of an Observational Longitudinal Cohort Study,” Targeted Oncology 18, no. 3 (2023): 425–440, 10.1007/s11523-023-00955-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Yi M., Niu M., Xu L., Luo S., and Wu K., “Regulation of PD‐L1 Expression in the Tumor Microenvironment,” Journal of Hematology & Oncology 14, no. 1 (2021): 10, 10.1186/s13045-020-01027-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Haragan A., Field J. K., Davies M. P. A., Escriu C., Gruver A., and Gosney J. R., “Heterogeneity of PD‐L1 Expression in Non‐Small Cell Lung Cancer: Implications for Specimen Sampling in Predicting Treatment Response,” Lung Cancer 134 (2019): 79–84, 10.1016/j.lungcan.2019.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Reticker‐Flynn N. E., Zhang W., Belk J. A., et al., “Lymph Node Colonization Induces Tumor‐Immune Tolerance to Promote Distant Metastasis,” Cell 185, no. 11 (2022): 1924–1942, 10.1016/j.cell.2022.04.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Evanno E., Godet J., Piccirilli N., et al., “Tri‐Methylation of H3K79 Is Decreased in TGF‐β1‐Induced Epithelial‐To‐Mesenchymal Transition in Lung Cancer,” Clinical Epigenetics 9, no. 1 (2017): 80, 10.1186/s13148-017-0380-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Cai J., Wang D., Zhang G., and Guo X., “The Role of PD‐1/PD‐L1 Axis in Treg Development and Function: Implications for Cancer Immunotherapy,” Oncotargets and Therapy 12 (2019): 8437–8445, 10.2147/OTT.S221340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Chen S., Crabill G. A., Pritchard T. S., et al., “Mechanisms Regulating PD‐L1 Expression on Tumor and Immune Cells,” Journal for Immunotherapy of Cancer 7, no. 1 (2019): 305, 10.1186/s40425-019-0770-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. John N., Schlintl V., Sassmann T., et al., “Longitudinal Analysis of PD‐L1 Expression in Patients With Relapsed NSCLC,” Journal for Immunotherapy of Cancer 12, no. 4 (2024): e008592, 10.1136/jitc-2023-008592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Jiang X. M., Xu Y. L., Huang M. Y., et al., “Osimertinib (AZD9291) Decreases Programmed Death Ligand‐1 in EGFR‐Mutated Non‐Small Cell Lung Cancer Cells,” Acta Pharmacologica Sinica 38, no. 11 (2017): 1512–1520, 10.1038/aps.2017.123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Isomoto K., Haratani K., Hayashi H., et al., “Impact of EGFR‐TKI Treatment on the Tumor Immune Microenvironment in EGFR Mutation‐Positive Non‐Small Cell Lung Cancer,” Clinical Cancer Research 26, no. 8 (2020): 2037–2046, 10.1158/1078-0432.CCR-19-2027. [DOI] [PubMed] [Google Scholar]