PD-L1: A Biomarker of Response to Frontline Osimertinib in EGFR-Mutant NSCLC?

The therapeutic paradigm for stage IV EGFR-mutant non–small cell lung cancer (NSCLC) has evolved rapidly over the past decade, driven largely by targeted therapies such as the third-generation EGFR tyrosine kinase inhibitor (TKI) osimertinib (Tagrisso). Yet even within this molecularly defined subset, clinical outcomes remain heterogeneous. Increasingly, investigators are examining whether biomarkers beyond EGFR mutation status may help explain and potentially predict this variability.

One such biomarker is PD-L1 expression. Although PD-L1 has become central to treatment decision-making in EGFR wild-type NSCLC, its role in EGFR-mutant disease has been less clear. Findings from a recent analysis published in Cancer Treatment and Research Communications suggest that PD-L1 expression may have prognostic significance in this targeted therapy–driven setting, raising new questions about tumor biology and optimal treatment strategies.¹

Investigating PD-L1 in EGFR-Mutant Disease

The study, titled “Comprehensive Clinical Characteristics and Outcomes of Stage IV EGFR-Mutant NSCLC Based on PD-L1 Expression,” evaluated clinical characteristics and outcomes in patients with metastatic EGFR-mutant NSCLC treated with frontline osimertinib. Investigators sought to determine whether PD-L1 expression correlates with treatment outcomes in this population.

“EGFR-mutated NSCLC is a very heterogeneous disease,” Jonathan Lee, MD, MSc, lead study author and chief hematology/oncology fellow at NewYork-Presbyterian/Weill Cornell Medicine in New York, New York, said in an interview with OncLive^®. “It comprises multiple different types of EGFR mutations, and the biology of each of these diseases can be quite distinct.”

The research question arose from observations in routine clinical practice. PD-L1 testing is performed worldwide in NSCLC, even in patients with actionable driver mutations. Yet clinicians have historically lacked clear guidance on interpreting PD-L1 expression in EGFR-mutant disease.

“One of the questions that came from clinic was [based on the fact that] we see patients with EGFR-mutated disease who have very different PD-L1 expression levels. What does that mean about the tumor?” Lee said. “Does this marker, which is widely tested across the world, tell us something about tumor biology, and can it serve as a predictive or prognostic biomarker when patients receive frontline osimertinib?”

Preclinical data also helped motivate the investigation. Findings from laboratory studies suggest that PD-L1 and EGFR signaling pathways may interact at the molecular level, prompting researchers to explore whether PD-L1 expression influences responses to EGFR-targeted therapy.

Signals of Worse Outcomes With PD-L1 Positivity

In the study, patients were categorized according to PD-L1 expression status: PD-L1–negative (n = 52) or PD-L1–positive (n = 11). Investigators then evaluated clinical outcomes, including overall response rate (ORR), progression-free survival (PFS), and overall survival (OS), among patients receiving first-line osimertinib.

The results revealed a consistent trend: Patients whose tumors expressed PD-L1 experienced numerically worse outcomes across multiple end points. The data indicated that the ORR was 92.3% in the negative cohort vs 77.6% in the positive cohort (OR, 0.29; 95% CI, 0.08-0.92; P = .046). The median PFS was 22.7 months (95% CI, 18.8-32.0) and 15.4 months (95% CI, 12.0-27.2), respectively (HR, 1.49; 95% CI, 0.87-2.56; P = .10). The median OS was 38.8 months (95% CI, 31.9-not reached) in the PD-L1–negative population vs 31.4 months (95% CI, 28.5-43.2) in the –positive population (HR, 1.53; 95% CI, 0.89-2.62; P = .12).

“We found that patients with higher PD-L1–expressing tumors had a lower response rate, shorter PFS, and shorter OS [than] patients whose tumors were PD-L1 negative,” Lee said.

Although the differences did not reach statistical significance, they approached the threshold, suggesting the analysis may have been underpowered.

“The P values were around .06,” Lee noted. “We think we were probably underpowered to definitively answer the question, but there was certainly a trend.”

Taken in isolation, the findings might be interpreted cautiously. However, the investigators emphasized that their results align with a growing body of literature suggesting a similar pattern.

“Taking our study in conglomerate with other studies, [we also saw] this sort of trend…that having higher levels of PD-L1 is prognostic for worse outcomes with first-line targeted therapies,” Ashish Saxena, MD, PhD, senior study author, associate professor of clinical medicine at Weill Cornell Medical College, and assistant attending physician at NewYork-Presbyterian Hospital, said. “At this point, it’s more hypothesis generating and something that needs to be followed up on in larger studies [so that we can better] understand the biological rationale of why having more PD-L1 expression makes you less responsive to single-agent osimertinib.”

Context Within the Broader Literature

Data from previous analyses have suggested that PD-L1 expression may correlate with more aggressive tumor biology in EGFR-mutant NSCLC. For example, findings from retrospective studies have demonstrated that high PD-L1 expression may be associated with shorter PFS in patients treated with EGFR TKIs.

A large multicenter study evaluated outcomes among patients with EGFR-mutant NSCLC receiving EGFR TKIs. The study findings, published in the Journal of Thoracic Oncology, reported that high PD-L1 expression correlated with inferior clinical outcomes compared with those who had PD-L1–negative disease.² Specifically, a lower ORR was seen in patients with positive (n = 14) vs weak (n = 18) or negative (n = 52) expression, at 35.7% vs 66.7% vs 67.3%, respectively (P = .001), as was shorter PFS, with a median of 3.8 months vs 6.0 months vs 9.5 months (P < .001).

Similarly, research data in Lung Cancer showed that PD-L1 expression may reflect underlying tumor biology associated with resistance to EGFR-targeted therapy.³ The analysis evaluated patients with primary resistance to EGFR TKI therapy and those who achieved either stable disease or partial response to EGFR TKI therapy. Fifty-five (22.7%) patients had a PD-L1 tumor proportion score (TPS) of at least 50% in the primary resistance group vs 1 (1.8%) in the disease control group (P < .001). Twenty (30.3%) patients had a PD-L1 TPS of at least 25% in the primary resistance group vs 2 (3.5%) in the disease control group (P < .001). Thirty (45.5%) patients had a PD-L1 TPS of at least 1% in the primary resistance group vs 7 (12.3%) in the disease control group (P = .001).

Patients with PD-L1 expression of at least 1% had a higher likelihood of having primary resistance to EGFR TKIs than those with a PD-L1 expression level below 1% (OR, 5.95; 95% CI, 2.35-15.05; P < .001). The same pattern was seen when the cutoff value was changed to 25% (OR, 11.96; 95% CI, 2.65-53.87; P = .001) and 50% (OR, 16.47; 95% CI, 2.10-129.16; P = .008). The estimated median PFS was 7.3 months in patients with PD-L1 expression below 1%, 2.1 months in patients with expression of at least 1%, 1.8 months in patients with expression of at least 25%, and 1.6 months in patients with expression of at least 50%.

Despite these observations, PD-L1 remains a controversial biomarker in EGFR-mutant NSCLC.⁴ Consequently, current guidelines from organizations such as the National Comprehensive Cancer Network continue to recommend EGFR-targeted therapy as the preferred frontline treatment strategy for EGFR-mutant metastatic NSCLC, regardless of PD-L1 expression status.

Implications for Treatment Strategy

One key question is why PD-L1 expression would correlate with poorer outcomes in patients treated with EGFR TKIs. The underlying mechanisms remain uncertain.

“That is the million-dollar question,” Lee said. “What is the mechanistic explanation for what we’re seeing? We often think about PD-L1 expression purely in the context of immune evasion. But the question is, what else does it tell us about the intrinsic biology of the disease?”

Although the study focused specifically on frontline osimertinib monotherapy, its findings raise questions about how PD-L1 expression might inform broader treatment strategies.

The therapeutic landscape for EGFR-mutant NSCLC has expanded significantly in recent years. Findings from phase 3 trials such as FLAURA2 (NCT04035486) and MARIPOSA-2 (NCT04988295) have demonstrated improved outcomes with combination approaches, including EGFR TKIs paired with chemotherapy or targeted antibodies.^5,6

These developments have introduced new complexity into treatment sequencing decisions. Within this context, PD-L1 expression might help identify patients who could benefit from more aggressive up-front strategies.

“Our ideal outcome would be that PD-L1 expression could [help us decide whether] to intensify therapy with a combination vs single-agent TKI therapy,” Saxena explained. “But we’re not quite there yet.”

Challenges in Defining High PD-L1 Expression and Looking Ahead

One barrier to clinical implementation is the lack of a standardized definition of high PD-L1 expression in EGFR-mutant NSCLC. Different studies have used varying thresholds ranging from 1% to 50% TPS to define PD-L1 positivity.

“Our data looked at PD-L1–expressing vs nonexpressing tumors,” Lee explained. “But other studies have used cutoffs of 25% or 50%, so the jury is still out on what counts as high PD-L1 expression.”

Establishing a biologically meaningful cutoff will likely require integration of clinical outcomes data with mechanistic insights.

To further clarify the relationship between PD-L1 expression and outcomes in EGFR-mutant NSCLC, researchers are pursuing additional collaborative analyses. Lee and Saxena are working with international partners to pool data from multiple cohorts and explore potential mechanistic explanations. These efforts are particularly important given the rapidly evolving treatment landscape for EGFR-mutant NSCLC.

“As the clinical landscape of EGFR-mutated disease continues to change, questions like this will only become more relevant,” Lee said.

For practicing oncologists, the immediate clinical implications remain modest but potentially meaningful.

“I think PD-L1 expression should inform clinical practice,” Lee said. “There [are] now enough data from multiple groups showing a similar trend that higher PD-L1 expression may be associated with poorer outcomes on single-agent osimertinib. If I saw a patient with EGFR-mutant metastatic NSCLC whose tumor had higher PD-L1 expression, I would consider a combination approach, a more intense treatment regimen than single-agent osimertinib, unless there are other strong reasons not to.”

References

1. Lee JW, Jin X, Bogdan S, et al. Comprehensive clinical characteristics and outcomes of stage IV EGFR-mutant NSCLC based on PD-L1 expression. Cancer Treat Res Commun. 2026;46:101109. doi:10.1016/j.ctarc.2026.101109
2. Su S, Dong ZY, Xie Z, et al. Strong programmed death ligand 1 expression predicts poor response and de novo resistance to EGFR tyrosine kinase inhibitors among NSCLC patients with EGFR mutation. J Thorac Oncol. 2018;13(11):1668-1675. doi:10.1016/j.jtho.2018.07.016
3. Hsu KH, Huang YH, Tseng JS, et al. High PD-L1 expression correlates with primary resistance to EGFR-TKIs in treatment naïve advanced EGFR-mutant lung adenocarcinoma patients. Lung Cancer. 2019;127:37-43. doi:10.1016/j.lungcan.2018.11.021
4. Liu J, Itchins M, Nagrial A, et al. Relationship between PD-L1 expression and outcome in EGFR-mutant lung cancer patients treated with EGFR tyrosine kinase inhibitors. Lung Cancer. 2021;155:28-33. doi:10.1016/j.lungcan.2021.03.004
5. Jänne PA, Planchard D, Kobayashi K, et al; FLAURA2 Investigators. Survival with osimertinib plus chemotherapy in EGFR-mutated advanced NSCLC. N Engl J Med. 2026;394(1):27-38. doi:10.1056/NEJMoa2510308
6. Yang JCH, Lu S, Hayashi H, et al; MARIPOSA Investigators. Overall survival with amivantamab-lazertinib in EGFR-mutated advanced NSCLC. N Engl J Med. 2025;393(17):1681-1693. doi:10.1056/NEJMoa2503001