PD-L1 strong expressions affect the clinical outcomes of osimertinib in treatment naïve advanced EGFR-mutant non-small cell lung cancer patients

The impact of strong Programmed Death-ligand 1 (PD-L1) expression on the clinical outcomes of osimertinib in treatment naïve advanced Epidermal Growth Factor Receptor (EGFR)-mutant Non-small Cell Lung Cancer (NSCLC) patients remains uncertain. We enrolled advanced NSCLC patients who harbored sensitizing EGFR mutation and were treated first-line with osimertinib between 2017 and 2021. The PD-L1 expression level was also tested. A total of 85 patients were included. The objective response rate to osimertinib was 78.9%, with the disease control rate being 90.8%. Median Progression-free Survival (PFS) was 22.1 months, while median Overall Survival (OS) was not reached (NR). Patients with the exon 19 deletion experienced better PFS than those with the exon 21 L858R mutation (NR vs 12.4 months, aHR 0.24 (95% CI, 0.10 to 0.57); p = 0.001). Seventy-one of these 85 patients had reported on their PD-L1 expression. Patients with a PD-L1 < 50% experienced longer PFS than patients with a PD-L1 ≧50% (26.5 vs 9.7 months, aHR 0.19 (95% CI, 0.06 to 0.67); p = 0.009). Additionally, patients with a PD-L1 < 50% experienced better OS than those with a PD-L1 ≧50% (NR vs 25.4 months, aHR 0.09 (95% CI, 0.01 to 0.70); p = 0.021). Strong expressions of PD-L1 in treatment naïve advanced EGFR-mutant NSCLC patients were associated with poor prognoses in those undergoing treatment with osimertinib as first-line therapy.


Material and methods
Study design and patients. This study was a retrospective, single-center, observational study performed at Taichung Veterans General Hospital (TCVGH) in Taiwan. The study was conducted ethically in accordance with the World Medical Association Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of TCVGH, Taiwan (IRB No. CF12019). All patients provided written informed consent for genetic testing, as well as use of their clinical data.
For the study, lung cancer patients were enrolled between December 2017 and December 2021. The inclusion criteria for patients were: (a) a diagnosis of histologically and cytologically confirmed NSCLC, (b) recurrence of, or diagnosed with, stage IV lung cancer according to the 8th edition of the American Joint Committee for Cancer (AJCC) staging system, (c) activating EGFR mutation with exon 19 deletion or exon 21 L858R point mutation, and (d) use of osimertinib as their first-line treatment. Patients were excluded if they were diagnosed with EGFR mutations other than exon 19 deletion and L858R mutation, were receiving other systemic treatments prior to their use of osimertinib, or were diagnosed with another active malignancy. Computed tomography of the chest was performed every three months in order for patients to qualify for National Health Insurance reimbursement. Treatment response to osimertinib was evaluated by the Response Evaluation Criteria in Solid Tumors (Version 1.1) 25 .
We collected for analysis each patient's demographic and clinical data, including age, gender, smoking status, Eastern Cooperative Oncology Group Performance Status (ECOG PS), clinical stage, condition of brain metastasis at baseline, EGFR mutation status at baseline, PD-L1 expression status, as well as their PFS and OS results in regards to osimertinib treatment. PFS was defined as the time from the first dose of osimertinib to progression of the disease or death, while OS was defined as the time measured from the first dose of osimertinib to death.
PD-L1 expression and the EGFR mutation test. PD

Statistical analyses.
To assess the inter-group differences (PD-L1 ≧50% group and PD-L1 < 50% group) in patient characteristics and demographic data, we used the Fisher's exact test for age, gender, smoking status, ECOG PS, clinical stage, condition of brain metastasis at baseline and EGFR mutation status at baseline. Survival curves for both PFS and OS were estimated using the Kaplan-Meier method. The Cox proportional hazard model was used to evaluate differences in survival times of PFS and OS. All statistical tests were performed using SPSS 23.0 software (SPSS Inc., Chicago, IL, USA). Two-tailed tests and P values < 0.05 were considered significant.

Results
Baseline clinical characteristics of patients with osimertinib as first-line EGFR-TKI treatment. In  The clinical efficacy of osimertinib. Nine patients could not be evaluated for their treatment response to osimertinib. As for the others, one (1.3%) had a complete response, while 7 (9.2%) suffered from primary resistance. Fifty-nine (77.6%) patients experienced partial response, with 9 (11.9%) having a stable disease under osimertinib. The overall ORR was 78.9%, and the Disease Control Rate (DCR) was 90.8% (Table 1). The estimated median PFS was 22.1 months (95% Confidence Interval (CI), 16.6 to 27.6) (Fig. 1A), while the estimated median OS was not reached (Fig. 1B).

Discussion
Our research in real-world practice has demonstrated that osimertinib provided satisfactory efficacy in advanced NSCLC patients harboring EGFR mutation. PD-L1 expression levels and ECOG PS influenced clinical outcomes of osimertinib when used as first-line treatment. Advanced EGFR mutated NSCLC patients with PD-L1≧50% had poorer PFS and OS than those with PD-L1 < 50%.
Furthermore, our previous study showed that PD-L1 expression level was associated with the frequency of primary resistance to EGFR-TKIs in treatment naïve advanced lung adenocarcinoma patients with EGFR mutation 23 . Patients with a PD-L1≧1% experience a higher incidence of primary resistance to EGFR-TKIs than those with a PD-L1 < 1% (Odds Ratio (OR), 5.95; p < 0.001). This phenomenon persisted while the cutoff value was changed to either 25% (OR, 11.96; p = 0.001) or 50% (OR, 16.47; p = 0.008). In the study performed by Su www.nature.com/scientificreports/  20 . Kang et al., also found that early progression rates, defined as progressive disease occurring within 6 months after first-line EGFR-TKI treatment had begun, were 12.7%, 2.7%, and 37.5% in patients with negative, weak, and strong PD-L1 expression, respectively (p = 0.004) 21 . Based on the above findings, not only PFS but also the chance of de novo resistance to first-line EGFR-TKIs, were correlated with PD-L1 expressions in EGFR-mutant NSCLC patients.
Although previous research has discussed the relationship between PD-L1 expression levels and the clinical efficacy of EGFR-TKIs, there have been few papers focusing on the third-generation EGFR-TKI osimertinib. In the FLAURA study, post-hoc analysis showed that in advanced treatment naïve EGFR-mutant NSCLC patients with PD-L1≧1%, the median PFS was 18.4 months with osimertinib, and 6.9 months with gefitinib and erlotinib (HR 0.30 (95% CI, 0.15 to 0.60)) 24 . Among patients with PD-L1 negative, the median PFS of osimertinib was 18.9 months, and 10.9 months in patients receiving gefitinib and erlotinib (HR 0.37 (95% CI, 0.17 to 0.74)). The ORR to osimertinib in patients with PD-L1≧1% was 79%, while in patients with PD-L1 < 1% it was 85%. Brown et al., concluded that the clinical outcomes of first-line osimertinib treatment for advanced EGFR-mutant NSCLC was not affected by PD-L1 expression status. However, these authors did not analyze the impact of strong PD-L1 expression (≧50%) on the clinical efficacy of osimertinib. In our present study, we found that advanced EGFRmutant NSCLC patients with PD-L1 < 50% experienced a longer PFS than those with PD-L1≧50% (26.5 versus 9.7 months, respectively, aHR 0.19 (95% CI, 0.06 to 0.67); p = 0.009). Patients with PD-L1 < 50% had better OS than patients with PD-L1≧50% (NR versus 25.4 months, respectively, aHR 0.09 (95% CI, 0.01 to 0.70); p = 0.021). Strong PD-L1 expressions resulted in statistically significant poorer prognoses in patients being treated with osimertinib as their first-line EGFR-TKI treatment.
Previous studies have demonstrated several mechanisms which could result in primary resistance to EGFR-TKI in EGFR-mutant NSCLC patients. These mechanisms include, Phosphatase and Tensin homolog gene (PTEN) loss, Erb-b2 receptor tyrosine kinase 2 gene (ERBB2) amplification, de novo T790M mutation, MET amplification, v-myc avian myelocytomatosis viral oncogene homolog gene (MYC) amplification, BIM deletion polymorphism, and overexpression of CRIPTO1 [30][31][32] . Previous studies have also reported that EGFR activation upregulated PD-L1 expression through the p-ERK1/2/p-c-Jun pathway, while MET activation also promoted the expression of PD-L1 33,34 . In Kang's study, strong PD-L1 expressions in EGFR positive NSCLC specimens were related to both the JAK-STAT pathway and MUC16 mutation frequency 21 . The authors found that the activation of the JAK-STAT pathway may play the role of a de novo resistance mechanism to EGFR-TKIs 21 . The above findings may explain the mechanisms surrounding drug resistance to EGFR-TKIs in EGFR-mutant NSCLC patients www.nature.com/scientificreports/ with strong PD-L1 expressions. Additional studies are still needed in order to obtain a complete understanding of the mechanisms involved. In clinical practice, we should notice that NSCLC patients with EGFR-mutant and high PD-L1 expression may experience relatively poor outcomes than patients with PD-L1 < 50%. In the future, we should figure out the possible treatment strategies for these patients. Combination treatment with anti-angiogenic agents or chemotherapy may overcome the downside. However, we need clinical trials to confirm our idea. There were certain limitations to our study. First, it was a single center, retrospective study with inevitable bias, as compared with prospective studies. Second, only Taiwanese subjects were studied. Therefore, our findings may not be generalized to include other ethnic populations. Third, osimertinib treatment patients began to be reimbursed for their treatment costs in April 2020. For this economic reason, many patients did not receive osimertinib treatment prior to this time, accounting for the relatively small number of patients enrolled for our analysis. Fourth, the follow-up time period was not long enough, with the maturation rate of PFS reaching only 41.2% and the OS measuring 12.9%. Fifth, not all patients given osimertinib as first-line EGFR-TKI treatment could provide data regarding PD-L1 expression. Finally, PD-L1 expression was determined using two different kits (Ventana SP263 and Dako 22C3). However, the Blueprint project had previously reported that three PD-L1 IHC assays (Ventana SP263, Dako 22C3 and  were closely aligned in their results which involved tumor cell staining 35 . Thus, we believe that the different kits we used did not influence our results.
In conclusion, our study has demonstrated that patients receiving osimertinib as first-line treatment had satisfactory ORR and PFS. Of particular note, patients with strong PD-L1 expression displayed poor outcomes after osimertinib treatment in terms of both PFS and OS. We still need multi-center retrospective studies or clinical trials to confirm our finding.

Data availability
The datasets generated during and/or analysed during the current study are not publicly available due to patients' privacy but are available from the corresponding author on reasonable request. www.nature.com/scientificreports/