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.

BACKGROUND: Osimertinib monotherapy is currently the standard of care as a first-line treatment for patients harboring epidermal growth factor receptor (EGFR) mutations; however, some EGFR-mutated non-small cell lung cancer (NSCLC) patients exhibit primary resistance and an insufficient response to EGFR-tyrosine kinase inhibitors (EGFR-TKIs). Elevated programmed death-ligand 1 (PD-L1) expression in tumors was reported as a negative predictive factor for outcomes of first- or second-generation EGFR-TKIs.
METHODS: We prospectively assessed advanced NSCLC patients with EGFR mutations who were treated with osimertinib at 14 institutions in Japan between September 2019 and December 2020. Relationships between outcomes of osimertinib monotherapy and patients' characteristics were reviewed.
RESULTS: Seventy-one patients who underwent the tumor PD-L1 test were enrolled. Multivariate analysis identified tumor PD-L1 expression as an independent predictor for progression-free survival (PFS) with osimertinib treatment (P=0.029). The objective-response and disease-control rates for osimertinib treatment were significantly lower in patients demonstrating elevated PD-L1 levels relative to those with low or negative PD-L1 level (P=0.043 and P=0.007, respectively). Furthermore, among patients treated with osimertinib, those with high PD-L1 levels exhibited shorter PFS relative to those with low plus negative PD-L1 level (median PFS: 5.0 vs. 17.4 months; P<0.001).
CONCLUSIONS: Elevated tumor PD-L1 expression is associated with poor outcomes of osimertinib monotherapy in previously untreated advanced NSCLC patients with EGFR mutation. Further clinical trials are warranted to accumulate evidence demonstrating the effectiveness of combination therapy with osimertinib for EGFR-mutated advanced NSCLC patients with elevated tumor PD-L1 expression. TRIAL REGISTRATION: UMIN000043942. 2021 Translational Lung Cancer Research. All rights reserved.

Introduction

Lung cancer is the number one cause of cancer-related death worldwide (1), and non-small cell lung cancer (NSCLC) is the most common subtype, accounting for ~85% of all lung cancer cases (2). Improved clinical outcomes in NSCLC patients harboring epidermal growth factor receptor (EGFR) mutations, including major subtypes, such as exon 19 deletion or point mutation in exon 21 resulting in L858R substitution, have contributed to the development of EGFR-targeted therapy. NSCLC patients with activating EGFR mutations exhibited better responses to first- and second-generation EGFR-tyrosine kinase inhibitors (EGFR-TKIs) than to systemic platinum-based chemotherapy (3,4). Treatment with the third-generation EGFR-TKI osimertinib showed better outcomes than those with first-generation EGFR-TKIs, such as gefitinib or erlotinib, in first-line treatment of advanced EGFR-mutated NSCLC patients (5). Therefore, osimertinib has been approved as a therapy for untreated EGFR-mutated advanced NSCLC patients in countries, including the United States and Japan. Although osimertinib monotherapy represents a promising treatment modality, ~20% of EGFR-mutated NSCLC patients exhibit primary resistance to osimertinib (5). To date, other therapeutic strategies have been approved in several countries, including the USA and Japan, for first-line treatment of EGFR-mutated NSCLC patients, including initial combination therapy with an anti-angiogenesis agent and chemotherapy to overcome the above issues and other EGFR-TKIs (6,7). Therefore, it is of important clinical relevance to determine an optimal initial therapeutic strategy for patients with EGFR-mutated advanced NSCLC.

Although EGFR-mutated advanced NSCLC cells respond well to osimertinib initially, a small percentage of cells can survive and expand, leading to acquired drug resistance and tumor heterogeneity, ultimately promoting tumor recurrence. As for intrinsic resistance to EGFR-TKIs, EGFR-T790M mutation, EGFR-exon20 insertions, and BIM deletion polymorphism have been reported as contributory factors (8-10). Based on previous reports, acquired-resistance mechanisms can be broadly classified into resistance caused by the treatment target EGFR [EGFR-T790M secondary resistance gene mutation (11)], resistance via non-EGFR bypass signal [Met gene amplification (12), HGF overexpression (13), HER2 gene amplification (14), GAS6-AXL signal activation (15)], and other resistance [transformation to small cell lung cancer (16) and epithelial-to-mesenchymal transition (17)].

Recently, immune-checkpoint inhibitor (ICI) therapy has made rapid advances in several cancers, including lung cancer, according to improved clinical outcomes, such as prolonged survival and a more durable treatment response (18-22). The identification of promissing biomarkers for detecting respondents to ICI treatment is currently underway, with programmed death ligand-1 (PD-L1) expression in tumors clinically identified as a positive predictive biomarker for advanced NSCLC patients treated with ICIs, especially for NSCLC patients with wild-type driver oncogenes (21). Elevated PD-L1 expression in tumors suppresses T cell activation and growth via apoptosis of effector T cells, which interferes with tumor immune responses (23,24), thereby identifying PD-L1 as a negative regulator of immune response. Preclinical studies have shown that activation of EGFR signaling pathways is involved in the induction of PD-L1 expression in NSCLC cells (25). Meanwhile, tumor PD-L1 expression was identified as a negative predictor of outcome in EGFR-mutated advanced NSCLC patients treated with first- or second-generation EGFR-TKIs (26-30). However, the effect of tumor PD-L1 level on the efficacy of osimertinib monotherapy in EGFR-mutated advanced NSCLC patients remains unknown. In this prospective study, we identified biomarkers of osimertinib efficacy as first-line treatment for EGFR-mutated advanced NSCLC patients. We present the following article in accordance with the STROBE reporting checklist (available at https://dx.doi.org/10.21037/tlcr-21-461).

Methods

Patients

We prospectively assessed 71 advanced or postoperative recurrent NSCLC patients harboring an EGFR-activating mutation, who were treated with osimertinib at 14 institutions in Japan between September 2019 and December 2020. Osimertinib administration and assessment of its efficacy and toxicity were performed by each investigator. Image evaluation was stipulated by every 8 to 12 weeks, including complete response, partial response, stable disease, and progressive disease, using either conventional computed tomography or magnetic resonance imaging, according to criteria outlined in Response Evaluation Criteria in Solid Tumors (v.1.1). Progression-free survival (PFS) was defined as the time from initiation of osimertinib treatment to the date of objective disease progression or death, regardless of whether the patient withdrew from osimertinib treatment or received another anticancer therapy prior to progression. Among the 70 EGFR-mutant NSCLC patients showing disease progression within 90 days or during >90-day follow-up, seven were identified as exhibiting primary resistance to osimertinib treatment and categorized as “disease progression within 90 days”. The inclusion criteria in this study are as follows; (I) patients without any systemic treatment, (II) symptomatic brain metastases are allowed, (III) any Eastern Cooperative Oncology Groups performance status (ECOG PS) is allowed, (IV) EGFR mutations, including L858R point mutation in exon 21 and exon 19 deletions, in addition to the other types of mutation, such as G719X in exon18, S768I in exon 20, L861Q in exon 21, were included. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). This study was approved by the institutional review board in Kyoto Prefectural University of Medicine (ERB-C-1242) and each respective hospital and registered at the University Medical Hospital Information Network (UMIN) Clinical Trials Registry (UMIN000043942). In addition, we had performed opt-out informed consent at each hospital from the trial initiation. Written informed consent was obtained from all participants.

EGFR-mutation analysis

EGFR mutations were detected using polymerase chain reaction (PCR) of tumor samples by sequencing exons 18 through 21, with the sequencing performed in commercial clinical laboratories (SRL, Inc., Tokyo, Japan; and BML, Inc., Tokyo, Japan). Deletions in exon 19 or a leucine to arginine substitution (L858R) in exon 21 are referred to as common mutations, and the other mutations are referred to as uncommon mutations.

Analysis of PD-L1 expression

PD-L1 expression in tumors was assessed using pretreatment tumor samples by performing PD-L1 immunohistochemistry (IHC) using a 22C3 pharmDx assay at a commercial clinical laboratory (SRL, Inc., Tokyo, Japan). Tumor PD-L1 expression was given as a percentage in at least 100 viable tumor cells used for complete or partial membrane staining. Pathologists at the commercial vendor interpreted tumor PD-L1 expression according to assay results. Patients were categorized into the following three groups based on PD-L1 tumor-proportion score (TPS): high (≥50%), low (1–49%), and negative (<1%).

Statistical analysis

To analyze PFS, times to events were estimated using the Kaplan-Meier method and compared using the log-rank test. Hazard ratios (HRs) for PFS were determined using a univariate Cox proportional hazards model. Cox proportional hazards models evaluating several patient factors were used. To construct the multivariate model, we selected the most relevant factors related to PFS, identified in the results of univariate analysis. All statistical analyses were performed using GraphPad Prism software (v.8.0; GraphPad Software, San Diego, CA, USA). P<0.05 was considered significant.

Results

Patient characteristics

The median age of the 71 EGFR-mutant advanced NSCLC patients enrolled in this study was 71.0 years (range, 35.0–87.0 years). Forty-five patients (63.4%) were female. Most patients (81.7%) indicated an ECOG PS of 0 or 1, and 40 patients (56.3%) were non-smokers (). The most prevalent history of disease included incidence of adenocarcinoma (94.4%), and 9 patients (12.7%) experienced relapse after surgery. According to EGFR-mutation status, 32 patients (45.1%) harbored exon 19 deletion, 36 patients (50.7%) harbored a point mutation in exon 21 resulting in L858R substitution, and 3 patients (4.2%) had a point mutation in exon 18 at G719C (uncommon).

Table 1

Patients’ characteristics

Characteristics	n=71
Median age, years (range)	71.0 (35.0–87.0)
Age categorization, n (%)
<75	45 (63.4)
≥75	26 (36.6)
Sex, n (%)
Male	26 (36.6)
Female	45 (63.4)
ECOG PS, n (%)
0	28 (39.4)
1	30 (42.3)
2, 3	13 (18.3)
Disease stage, n (%)
III	2 (2.8)
IV	60 (84.5)
Postoperative relapse	9 (12.7)
Brain metastasis, n (%)
Positive	21 (29.6)
Negative	50 (70.4)
Histology, n (%)
Adenocarcinoma	67 (94.4)
Others	4 (5.6)
EGFR mutation, n (%)
19del	32 (45.1)
L858R	36 (50.7)
G719C	3 (4.2)
Smoking status, n (%)
Current or former	31 (43.7)
Never	40 (56.3)
PD-L1 TPS, n (%)
≥50%	15 (21.1)
1–49%	26 (36.6)
<1%	30 (42.3)

ECOG PS, Eastern Cooperative Oncology Groups Performance Status; EGFR, epidermal growth factor receptor; 19del, exon 19 deletion; L858R, exon 21 L858R mutation; G719C, exon18 G719C mutation; PD-L1, programmed death-ligand 1; TPS, tumor proportion score.

Predictive factor for initial osimertinib treatment in EGFR-mutant advanced NSCLC patients

We then examined the predictive factors of osimertinib treatment in EGFR-mutant advanced NSCLC patients. The median follow-up time for this study was 15.5 months (range, 1.2–25.1 months). Fifty-four patients were followed up for more than 1 year, and 3 patients for more than 2 years (Figure S1). Median overall survival time (OS) was not evaluable (NE) (95% CI: 22.4–NE) (Figure S2A), and 17 patients (23.9%) were successively treated. Univariate analysis identified ECOG PS, EGFR-mutation status, and tumor PD-L1 expression as predictors of PFS for osimertinib monotherapy (P=0.010, P<0.001, and P=0.003, respectively) (), and multivariate analysis demonstrated that EGFR-mutation status and PD-L1 expression were independent predictive factors for PFS in osimertinib treatment [HR: 2.05, 95% confidence interval (CI): 1.06–3.97, P=0.034; and HR: 2.40, 95% CI: 1.09–5.25, P=0.029, respectively] (). These findings demonstrated that tumor PD-L1 expression was related to the efficacy of osimertinib treatment in EGFR-mutated NSCLC patients.

Table 2

Cox proportional hazard models for PFS in patients with non-small cell lung cancer harbording EGFR mutation who received osimertinib monotherapy, univariate analysis

Characteristics	Patient’s No.	Median PFS (95% CI), months	P value
Age categorization			0.895
<75 years	45	15.4 (8.9–NE)
≥75 years	26	15.6 (11.1–NE)
Sex			0.790
Male	26	15.6 (13.1–NE)
Female	45	14.7 (10.3–NE)
ECOG PS			0.010
0	28	NE (14.8–NE)
1	30	12.5 (9.9–17.4)
2, 3	13	6.5 (2.4–20.1)
Disease stage			0.812
III	2	11.9 (11.9–NE)
IV	60	15.4 (11.1–20.1)
Postoperative relapse	9	NE (2.4–NE)
Brain metastasis			0.136
Positive	21	12.9 (5.0–NE)
Negative	50	19.9 (12.5–NE)
Histology			0.188
Adenocarcinoma	67	15.6 (12.5–NE)
Others	4	5.5 (1.6–NE)
EGFR mutation			<0.001
19del	32	20.1 (12.9–NE)
L858R	36	13.8 (9.9–NE)
G719C	3	1.1 (1.0–NE)
Smoking status			0.165
Current or former	31	12.9 (7.5–17.4)
Never	40	19.9 (11.9–NE)
PD-L1 TPS			0.003
≥50%	15	5.0 (1.6–13.8)
1–49%	26	15.1 (11.1–NE)
<1%	30	19.9 (15.4–NE)

PFS, progression-free survival; EGFR, epidermal growth factor receptor; CI, confidential interval; NE, not evaluable; ECOG PS, Eastern Cooperative Oncology Groups Performance Status; 19del, exon 19 deletion; L858R, exon 21 L858R mutation; G719C; exon18 G719C mutation, PD-L1, programmed death-ligand 1; TPS, tumor proportion score.

Table 3

Cox proportional hazard models for PFS in patients with non-small cell lung cancer harbording EGFR mutation who received osimertinib monotherapy, multivariate analysis

Items	PFS, hazard ratio (95% CI)	P value
ECOG PS ≥2	1.71 (0.78–3.73)	0.180
EGFR mutation status (19del vs. L858R vs. uncommon mutation)	2.05 (1.06–3.97)	0.034
PD-L1 TPS ≥50%a	2.40 (1.09–5.25)	0.029

a, PD-L1 TPS ≥50% versus all others except for unknown. PFS, progression-free survival; EGFR, epidermal growth factor receptor; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Groups Performance Status; 19del, exon 19 deletion; L858R, exon21 L858R mutation; PD-L1, programed death-ligand 1; TPS, tumor proportional score.

The significance of tumor PD-L1 expression on clinicopathological features and osimertinib efficacy

Of the 71 patients, 15, 26, and 30 patients were classified into PD-L1 TPS high (≥50%), low (1–49%), and negative (<1%) groups, respectively. We assessed correlations of clinicopathological features by comparing the PD-L1 groups. There was no significant difference between the three groups ().

Table 4

Clinicopathological features comparing tumor PD-L1 expression

Characteristics	Tumor PD-L1 expression			P value
	≥50% (n=15)	1–49% (n=26)	<1% (n=30)
Median age, years (range)	69.0 (48.0–83.0)	74.0 (35.0–87.0)	70.0 (38.0–86.0)	0.249
Age categorization, n (%)				0.139
<75 years	8 (53.3)	14 (53.8)	23 (76.7)
≥75 years	7 (46.7)	12 (46.2)	7 (23.3)
Sex, n (%)				0.596
Male	7 (46.7)	8 (30.8)	11 (36.7)
Female	8 (53.3)	18 (69.2)	19 (63.3)
ECOG PS, n (%)				0.383
0	4 (26.7)	10 (38.5)	14 (46.7)
1	6 (40.0)	13 (50.0)	11 (36.7)
2, 3	5 (33.3)	3 (11.5)	5 (16.7)
Disease stage, n (%)
III	0 (0.0)	1 (3.8)	1 (3.3)	0.67
IV	14 (93.3)	20 (76.9)	26 (86.7)
Postoperative relapse	1 (6.7)	5 (19.2)	3 (10.0)
Brain metastasis, n (%)				0.09
Positive	7 (46.7)	4 (15.4)	10 (33.3)
Negative	8 (53.3)	22 (84.6)	20 (66.7)
Histology, n (%)				0.193
Adenocarcinoma	13 (86.7)	26 (100.0)	28 (93.3)
Others	2 (13.3)	0 (0.0)	2 (6.7)
Smoking status, n (%)				0.174
Current or former	9 (60.0)	8 (30.8)	14 (46.7)
Never	6 (40.0)	18 (69.2)	16 (53.3)
Response, n (%)				0.038
CR	1 (6.7)	0 (0.0)	2 (6.7)
PR	7 (46.7)	22 (84.6)	19 (63.3)
SD	3 (20.0)	3 (11.5)	6 (20.0)
PD	4 (26.7)	0 (0.0)	1 (3.3)
NE	0 (0.0)	1 (3.8)	2 (6.7)
ORR (95% CI)	53.3% (26.6–78.7%)	88.0% (68.8–97.5%)	75.0% (55.1–89.3%)	0.051
DCR (95% CI)	73.3% (44.9–92.2%)	100.0% (88.7–100.0%)	96.4% (81.7–99.9%)	0.007

PD-L1, programmed death-ligand 1; ECOG PS, Eastern Cooperative Oncology Groups Performance Status; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; NE, not evaluable; ORR, objective response rate; CI, confidence interval; DCR, disease control rate.

We then examined the effect of tumor PD-L1 expression on osimertinib efficacy. In all EGFR-mutated NSCLC patients, the objective-response rate (ORR) and disease-control rate (DCR) for osimertinib treatment were 72.1% and 92.6%, respectively (Table S1). The ORR values for osimertinib treatment tended to be low in high-PD-L1 patients compared to those in PD-L1-low and -negative patients (high, low, and negative: 53.3%, 88.0%, and 75.0%, respectively; P=0.051). Additionally, the DCR values for osimertinib treatment were significantly lower in high-PD-L1 patients than those in PD-L1-low and negative patients (high, low, and negative: 73.3%, 100.0%, and 96.4%, respectively; P=0.007) ( and ). Moreover, the ORR values for osimertinib treatment were significantly lower in high-PD-L1 patients relative to both those in PD-L1-low and -negative patients (53.3% vs. 81.1%; P=0.043), and the DCR values for osimertinib treatment were significantly lower in high-PD-L1 patients relative to those in both PD-L1-low and -negative patients (73.3% vs. 98.1%; P=0.007) ().

Figure 1

Osimertinib efficacy according to tumor PD-L1 expression. (A) ORR and DCR for osimertinib in PD-L1-high (≥50%), -low (1–49%), and -negative (<1%) patients. (B) ORR and DCR for osimertinib in PD-L1-high and all others group. (C) The frequency of primary resistance to osimertinib treatment in PD-L1-high (≥50%), -low (1–49%), and -negative (<1%) patients. (D) Kaplan-Meier curve for PFS of EGFR-mutated NSCLC patients according to tumor PD-L1 expression (high, low, and negative). Median PFS following osimertinib treatment was 5.0 months (PD-L1-high; 95% CI: 1.6–13.8), 15.1 months (PD-L1-low; 95% CI: 11.1–NE), and 19.9 months (PD-L1-negative; 95% CI: 15.3–NE) according to tumor PD-L1 expression (high vs. low and high vs. negative; P=0.006 and P=0.003, respectively). (E) Kaplan-Meier curve for PFS of EGFR-mutated NSCLC patients classified according to tumor PD-L1 expression (high and low + negative). Median PFS following osimertinib treatment was 5.0 months (PD-L1-high; 95% CI: 1.6–13.8) and 17.4 months (PD-L1-low and -negative; 95% CI: 13.1–NE) according to tumor PD-L1 expression (P<0.001). PD-L1, programmed death-ligand 1; ORR, objective response rate; DCR, disease control rate; PFS, progression-free survival; EGFR, epidermal growth factor receptor; NSCLC, non-small cell lung cancer; CI, confidence interval; NE, not evaluable.

The frequency of primary resistance to osimertinib treatment was significantly higher in high-PD-L1 patients compared to that in PD-L1-low and -negative patients (33.33%, 3.85%, and 3.45%, respectively; P=0.006) ().

Median PFS with osimertinib treatment was 15.4 months [95% CI: 11.9–not evaluable (NE)] in all EGFR-mutated NSCLC patients (Figure S2B). Notably, osimertinib treatment of NSCLC patients with high PD-L1 expression (5.0 months; 95% CI: 1.6–13.8) resulted in shorter PFS relative to that of PD-L1-low and -negative patients (low: 15.1 months, 95% CI: 11.1–NE; and negative: 19.9 months, 95% CI: 15.3–NE, respectively) (high vs. low and high vs. negative; P=0.006 and P=0.003, respectively) (). Additionally, osimertinib treatment of NSCLC patients with high PD-L1 expression resulted in significantly shortened PFS as compared with that of both PD-L1-low and -negative patients (<50%; 17.4 months, 95% CI: 13.1–NE; P<0.001) (). There was no significant relationship in OS between PD-L1-high patients and PD-L1-low plus negative patients (P=0.858) (Figure S3).

Median PFS with osimertinib treatment according to EGFR mutational status was 15.4 months (95% CI: 11.9–NE) in exon 19 deletion and 13.8 months (95% CI: 9.9–NE) in exon 21 L858R mutation (Figure S4A,S4B). With respect to median PFS according to EGFR-mutation status, we found no significant correlation between PD-L1-high and PD-L1-low or -negative patients harboring exon 19 deletion in EGFR (P=0.522), whereas median PFS was significantly shorter in PD-L1-high patients relative to that in PD-L1-low and -negative patients (<50%) harboring the point mutation in exon 21 (6.5 vs. 15.6 months; P=0.024) (Figure S4C,S4D).

Discussion

The results of this prospective study revealed the clinical impact of elevated tumor PD-L1 expression as a negative predictive factor in determining the clinical outcomes of osimertinib treatment of EGFR-mutant NSCLC patients. To the best of our knowledge, this is the first study reporting that tumor PD-L1 expression is a clinically relevant predictive factor for osimertinib sensitivity.

Preclinical studies show that EGFR-mutant NSCLC cell lines with high PD-L1 expression exhibit induced epithelial-mesenchymal transition and less susceptibility to EGFR-TKIs via activation of transforming growth factor-β/SMAD canonical signaling (31). Moreover, previous clinical studies indicated that EGFR-mutant NSCLC patients exhibiting ≥50% tumor PD-L1 expression have a shorter PFS following treatment with the first-generation EGFR-TKI gefitinib, relative to patients showing tumor PD-L1 expression of <50%, which agreed with the findings of the present study (26-30). Consistent with these findings, in the present study, we found that high tumor PD-L1 expression (≥50%) was associated with poor outcomes of EGFR-TKI monotherapy in EGFR-mutant NSCLC patients. In contrast, subset analysis of data from the FLAURA clinical trial indicated that the median PFS for EGFR-mutant NSCLC patients with osimertinib was hardly affected between tumor PD-L1 expressors (≥1%) and negatives (<1%) (32). These results suggest that tumor PD-L1 expression of ≥50% might be a potent negative prognostic factor for EGFR-TKI treatment.

Previous studies reported a correlation between EGFR-TKI insensitivity and high PD-L1 expression. Specifically, in addition to EGFR, activation of other oncogenes promoted EGFR-TKI resistance associated with high PD-L1 expression, which led to the accumulation of other genetic alternations. Additionally, evolution of the tumor microenvironment, including immune cells, induced EGFR-TKI resistance via loss of tumor-antigen presentation and increased numbers of tumor-associated macrophages as a result of high tumor PD-L1 expression (30,33-35). Moreover, changes in intra-tumoral heterogeneity influence the therapeutic response of EGFR-mutated NSCLC tumors exhibiting high PD-L1 expression to ICIs and EGFR-TKIs (36). These observations suggest that the effectiveness of each targeted therapy might be influenced by the resident EGFR mutation or PD-L1 expression of each respective tumor.

To further improve clinical outcomes for EGFR-mutant NSCLC patients, several novel therapeutic approaches are being considered. Elevated tumor PD-L1 expression is a well-known biomarker associated with the response to ICIs, whereas ICI treatment is generally less effective in EGFR-mutated NSCLC patients (37). A previous report showed that tumor PD-L1 expression increases in EGFR-mutated NSCLC patients exhibiting high tumor PD-L1 expression after attaining resistance to EGFR-TKIs (38,39), suggesting that ICI treatment might be effective in osimertinib-resistant EGFR-mutated NSCLC patients exhibiting high tumor PD-L1 expression. However, another retrospective study showed that the duration of response to previous EGFR-TKIs was a negative predictor of ICI efficacy in EGFR-mutant NSCLC patients (40). Therefore, the utility of PD-L1 expression as a surrogate marker for response to therapeutic PD-L1-blockade in EGFR-mutated NSCLC patients remains controversial. Further clinical studies are needed to confirm the response to ICI or combined ICI+osimertinib treatment of EGFR-mutated NSCLC patients, especially those with high tumor PD-L1 expression.

Regulatory T cells (Tregs) are crucial mediators of immune suppression, contribute to tumor immune evasion, and represent poor prognostic factors for various malignancies (41,42). By contrast, treatment with an anti-vascular endothelial growth factor (VEGF) antibody inhibits Treg proliferation and leads to immune activation, which inactivates Tregs (43). A recent study showed that Treg frequency in tumor microenvironments is a reliable biomarker of clinical responses to the anti-VEGF receptor (VEGFR)2 antibody ramucirumab (44). In a subset analysis of phase 3 trial, the combination of immunochemotherapy plus anti-VEGF antibody bevacizumab improved PFS, compared to immunochemotherapy in advanced NSCLC patients with EGFR mutation [NE (95% CI: 17.0–NE) vs. 21.4 months (95% CI: 13.8–NE)] (45). Additionally, several clinical trials demonstrated that the frequency of primary resistance to combination therapy using an anti-VEGF/VEGFR antibody and EGFR-TKIs was lower relative to that observed for treatment with EGFR-TKI alone in EGFR-mutated NSCLC patients, suggesting that inhibition of VEGF-related signaling might play an important role in regulating immunomodulatory and/or anti-angiogenic factors (6,45,46). Another study reported that PD-L1 expression is associated with FOXP3-expressing Treg infiltration in tumors and poor prognosis in soft tissue sarcoma (47). Therefore, combined therapy with osimertinib and an anti-VEGF/VEGFR antibody might represent a promising therapeutic option for untreated EGFR-mutated NSCLC patients exhibiting high tumor PD-L1 expression.

This study has several limitations. First, the study involved a limited cohort of 71 cases, although this is prospective study. Second, all patients in the cohort were Japanese. Third, EGFR mutation status was detected using PCR analysis, which has limitations in the detection of compound mutations. Finally, two patients with a follow-up time of less than six months were enrolled. However, the novel findings regarding patient response to osimertinib are notable and could be useful for addressing clinical issues.

Conclusions

Our prospective data demonstrated that tumor PD-L1 expression is significantly associated with osimertinib efficacy in untreated advanced NSCLC patients harboring EGFR mutation. Further clinical trials are required to accumulate clinical evidence demonstrating the effectiveness of combination therapy with osimertinib to improve clinical outcomes for EGFR-mutated advanced NSCLC patients exhibiting high tumor PD-L1 expression.

The article’s supplementary files as