Duration of systemic therapy in patients with metastatic EGFR-mutated non-small cell lung cancer and brain metastases

Introduction

Brain metastases occur in approximately one-quarter of patients with epidermal growth factor receptor (EGFR)-mutated metastatic non-small cell lung cancer (mNSCLC) at diagnosis (1). Osimertinib, a third-generation EGFR tyrosine kinase inhibitor (TKI), was established as the standard first-line systemic therapy by the FLAURA trial, demonstrating superior survival compared with first-generation TKIs such as erlotinib and gefitinib (1,2). In patients with baseline brain metastases, osimertinib showed substantially greater intracranial activity, with central nervous system (CNS) objective response rates exceeding 90% among those with measurable CNS disease, compared with approximately 68% for erlotinib or gefitinib, as well as longer CNS progression-free survival (3). These findings support the routine use of first-line osimertinib to optimize intracranial disease control and improve survival outcomes in this high-risk population.

However, outcomes observed in clinical trials may not fully reflect real-world effectiveness, given selective enrolment, treatment at academic centres, and protocol-driven care (4). This efficacy-effectiveness gap underscores the importance of population-based studies to evaluate treatment performance in routine practice. To address this, we conducted a population-level cohort study using a large, comprehensive Canadian administrative database to compare systemic therapy outcomes among patients with de novo brain metastases treated with first-line osimertinib vs. gefitinib. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0107/rc).

Methods

Data sources

Details of data sources and methods have been published previously (5). We conducted a population-based retrospective cohort study using administrative databases linked through ICES in Ontario, Canada. Patients diagnosed with lung cancer between January 2007 and August 2024 were identified from the Ontario Cancer Registry (OCR), which captures 95% of Ontario cancer diagnoses, using ICD-O-3 topography codes C34.0–C34.4, C34.8, and C34.9. Only non-small cell lung cancer (NSCLC) histologies were included. Exclusions applied to patients younger than 18 years or older than 99 years, those with fewer than six months of follow-up in the absence of death (suggesting out-of-province migration), and those with another malignancy diagnosed within five years before NSCLC diagnosis or during follow-up. The study adhered to the Declaration of Helsinki and its subsequent amendments, was approved by the institutional ethics review board at Sunnybrook Health Sciences Centre (No. REB 6148) and complied with ICES privacy requirements. Individual consent was waived due to the retrospective design.

Study subjects

Patients receiving the EGFR TKIs gefitinib or osimertinib were included. Drug exposure was identified through the Ontario Drug Benefit (ODB) and Cancer Activity Level Reporting (ALR) databases and used as a surrogate for EGFR mutation positivity, as molecular pathology data are not available in administrative datasets. Radiotherapy data, including treatment dates, target site, dose, fractionation, and technique, were obtained from Cancer ALR and Ontario Health Insurance Plan (OHIP) billing records.

Brain-directed radiotherapy was used as a surrogate for brain metastases. To reduce misclassification of primary intracranial tumors, only contemporary dose-fractionation regimens commonly used for brain metastases (15–35 Gy in 10 fractions) were included. Patients who received their first course of brain radiotherapy within four weeks of initiating TKI therapy were classified as having de novo brain metastases. Subsequent courses meeting the same criteria were assumed to represent treatment for progressive intracranial disease.

Demographic and clinical variables included age and year of diagnosis (OCR); sex, rurality index, and vital status from the Registered Persons Database; neighborhood income quintile derived from the Canadian Census; and comorbidities assessed using the Elixhauser and Charlson comorbidity indices, constructed from the National Ambulatory Care Reporting System, CIHI Discharge Abstract Database, CIHI Same Day Surgery database, and OHIP billing data.

Outcomes

The primary outcome was duration of first-line systemic therapy. Secondary outcomes included receipt of subsequent brain-directed radiotherapy, CNS radiation progression-free survival (CNS-rPFS), and overall survival (OS).

Statistical analysis

Baseline characteristics were summarized using measures of central tendency. Duration of systemic therapy was calculated from the first to last recorded date of EGFR TKI use, with censoring at the last recorded date for patients remaining on treatment; results were reported as median with interquartile range (IQR). OS was estimated using the Kaplan-Meier method from TKI initiation to death from any cause or censoring, with comparisons using the log-rank test. CNS-rPFS was calculated from the first day of brain radiotherapy to subsequent brain radiotherapy, death, or censoring.

Propensity score matching was performed using logistic regression incorporating age, sex, rurality index, Elixhauser and Charlson comorbidity indices, and neighborhood income quintile. One-to-one nearest-neighbor matching with a caliper of 0.2 was applied, and balance assessed using standardized mean differences. Sensitivity analyses were performed by (I) redefining the index date as the start of first brain metastasis radiotherapy and (II) restricting analyses to patients treated in 2019 and after.

Results

Between January 2007 and August 2024, 3,541 patients in Ontario were treated with gefitinib or osimertinib. Among these, 481 patients (259 gefitinib; 222 osimertinib) received brain radiotherapy within four weeks of TKI initiation and were included in the primary analysis (Figure 1). Baseline characteristics are summarized in Table 1. Patients treated with osimertinib were more frequently diagnosed in 2019 or later, reflecting temporal adoption of the drug; otherwise, baseline characteristics were well balanced between cohorts. The median interval between systemic therapy initiation and brain radiotherapy was 14 days (IQR, 2–23 days). After propensity score matching, 206 patients remained in each cohort with good covariate balance across all variables (standardized mean differences <0.10; Table S1, Figure S1).

Figure 1 CONSORT diagram.

OS and CNS-rPFS outcomes are presented in Figures 2,3.

Figure 2 OS in unmatched cohort (A) and matched cohort (B). GEFI, gefitinib; OS, overall survival; OSIM, osimertinib.

Figure 3 CNS-rPFS in unmatched cohort (A) and matched cohort (B). CNS-rPFS, central nervous system radiation progression-free survival; GEFI, gefitinib; OSIM, osimertinib.

In the unmatched cohort, the median follow-up time was 15.9 months [95% confidence interval (CI): 14.7–17.6], and was 13.6 months (95% CI: 11.9–15.3) and 24.7 months (95% CI: 17.0–26.8) for gefitinib and osimertinib groups, respectively. First-line systemic therapy duration was significantly longer among patients treated with osimertinib compared with gefitinib [median 12.6 months (IQR, 4.7–24.1 months) vs. 6.4 months (IQR, 1.0–12.1 months); P<0.001]. OS was also superior with osimertinib, with a median OS of 24.7 months (95% CI: 17.0–26.8) vs. 13.6 months (95% CI: 11.9–15.3) for gefitinib [hazard ratio (HR) 0.59, 95% CI: 0.48–0.73; P<0.001; Figure 2A]. A greater proportion of patients receiving osimertinib underwent two or more courses of brain radiotherapy (20.3%) compared with those treated with gefitinib (7.7%; P<0.001; Table S2). Median CNS-rPFS was significantly longer in the osimertinib cohort at 21.7 months (95% CI: 16.3–26.2) vs. 13.5 months (95% CI: 12.2–15.6) in the gefitinib cohort (HR 0.64, 95% CI: 0.52–0.79; P<0.001; Figure 3A).

Findings were consistent in the propensity score-matched analysis. Patients treated with osimertinib had a longer median duration of first-line therapy [12.9 months (IQR, 4.8–24.6 months) vs. 6.8 months (IQR, 1.0–13.5 months); P<0.001]. Median OS remained significantly improved with osimertinib at 25.6 months (95% CI: 17.4–27.5) compared with 14.6 months (95% CI: 12.2–16.3) for gefitinib (HR 0.59, 95% CI: 0.47–0.74; P<0.001; Figure 2B). A higher proportion of osimertinib-treated patients received two or more courses of brain radiotherapy (19.9% vs. 6.8%; P<0.001), and CNS-rPFS was longer [24.9 months (95% CI: 16.7–27.3) vs. 14.5 months (95% CI: 12.4–16.3); HR 0.64, 95% CI: 0.51–0.80; P<0.001; Figure 3B].

In sensitivity analyses redefining the index date to the start of brain radiotherapy, osimertinib remained associated with longer first-line treatment duration (median 380 vs. 204 days; P<0.001), with consistent results after matching. Median OS and CNS-rPFS were also longer in the osimertinib as compared to the gefitinib cohort (Figure S2). When restricting the analysis to patients treated in 2019 or later (n=248; osimertinib, n=218, and gefitinib, n=34), the median treatment duration for osimertinib was 393 vs. 77 days for gefitinib (P<0.01). On matched analysis, a significant difference in median OS and CNS-rPFS was observed with osimertinib vs. gefitinib (Figure S3).

Discussion

This is the largest population-based cohort of patients with EGFR-mutated NSCLC and de novo brain metastases. We observed that first-line treatment with osimertinib was associated with a significantly longer duration of first-line systemic therapy and OS as compared with gefitinib. Similarly, CNS-rPFS was longer for osimertinib with patients receiving osimertinib more likely to undergo subsequent courses of brain radiotherapy as compared to patients receiving first-line gefitinib. These findings are consistent with and externally validate, the results of the FLAURA trial. Among patients with de novo brain metastases in FLAURA, median CNS-rPFS was longer among osimertinib treated patients (not reached vs. 13.9 months with gefitinib, P=0.01) (3).

Several observational studies also corroborate our observations. Kim et al. reported on a cohort of 352 patients from 9 academic centres in the United States treated with osimertinib (n=181) or first/second-generation TKI (n=171), of which 44.2% (n=80) and 41.5% (n=71) had baseline CNS metastases, respectively (6). Compared to earlier-generation TKIs, patients treated with osimertinib with baseline CNS metastases had a significantly longer progression-free survival (PFS) [median 13.1 vs. 8.2 months, HR 0.63 (95% CI: 0.40–0.99)], duration of treatment [median 15.5 vs. 8.0 months, HR 0.42 (95% CI: 0.28–0.64)], and time to next treatment [not reached vs. 11.2 months, HR 0.44 (95% CI: 0.27–0.70)] (6). Similarly, in a study of 606 matched patients with advanced EGFR-positive NSCLC in China, 178 had baseline brain metastases. Of these patients, those treated with osimertinib had a significantly longer median PFS (21.0 vs. 8.7 months, P<0.001) and OS (40.5 vs. 25.8 months; P<0.001) compared to the cohort treated with first-generation TKIs (7). The intracranial activity of osimertinib has also been described among patients with advanced CNS disease. For instance, in a cohort study of 71 metastatic NSCLC with leptomeningeal disease, Tamura et al. described an intracranial objective response rate of 62.5% among patients treated with osimertinib, as compared to 25.7% among patients treated with first-generation TKIs (P=0.007) (8). Similar differences are observed for leptomeningeal PFS, with a median of 23.4 vs. 12.1 months, respectively (P=0.02).

In the current study, patients receiving osimertinib underwent more courses of subsequent brain radiotherapy than those treated with gefitinib. This finding is likely multifactorial and may reflect longer overall disease control with osimertinib, more frequent magnetic resonance imaging (MRI) surveillance in later years, wider adoption of stereotactic radiosurgery (SRS) and a lower threshold for focal treatment of oligoprogressive lesions. Together, these factors likely reflect an evolving treatment approach where effective systemic therapy extends the window for targeted local interventions. Future studies should clarify the optimal sequencing and timing of SRS with osimertinib to further improve intracranial control and neurologic outcomes.

This study has several strengths, including the use of a large, comprehensive population-based dataset within a universal healthcare system, allowing evaluation of real-world treatment patterns in an unselected cohort. Propensity score matching improved balance across measured demographic, comorbidity, and socioeconomic variables.

However, limitations inherent to administrative data must be acknowledged as key clinical factors, including performance status, intracranial tumor burden, extracranial disease extent, and EGFR mutation subtype are unavailable in current administrative databases. Although temporal confounding related to the later adoption of osimertinib could not be fully balanced with propensity score matching, the observed clinical benefit in the osimertinib cohort remained in sensitivity analysis restricted to patients treated from 2019 onwards. This suggests that improvements in supportive care, diagnostic pathways, and radiation techniques do not fully explain the findings. Brain radiotherapy was used as a surrogate for CNS involvement and may therefore have excluded patients with small, asymptomatic metastases managed initially with TKI alone. Despite this limitation, outcomes remained improved in the osimertinib cohort, suggesting that osimertinib remains effective even among patients with larger or more symptomatic brain metastases beyond those typically represented in the FLAURA trial. Lastly, more recent first-line trials have reported improved outcomes relative to osimertinib alone. In the FLAURA2 trial, osimertinib plus platinum-pemetrexed prolonged PFS among patients with baseline CNS metastases (9). In the MARIPOSA trial, amivantamab and lazertinib improved OS vs. osimertinib, with updated analyses also suggesting more durable intracranial disease control (10). However, these intensified regimens were associated with substantially higher rates of grade 3+ toxicity than osimertinib monotherapy which may limit tolerability in frailer patients.

Conclusions

Overall, our real-world findings reinforce the CNS activity of osimertinib that has been demonstrated in clinical trials with patients treated with osimertinib having longer duration on first-line systemic therapy, CNS-rPFS and OS as compared to prior-generation TKIs. Contemporary practice guidelines are increasingly prioritizing CNS-penetrant agents such as osimertinib for patients with asymptomatic brain metastases, often deferring local therapies to time of intracranial progression (11). These shifts underscore the importance of multidisciplinary input for metastatic NSCLC patients with brain metastases, where collaboration between medical and radiation oncology is essential to optimize outcomes. Future studies should evaluate the real-world effectiveness of newer systemic therapy regimens, such as those evaluated in FLAURA2 and MARIPOSA, which have demonstrated improvements in CNS outcomes.

Acknowledgments

This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH). This document used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under license from ©Canada Post Corporation and Statistics Canada. Parts of this material are based on data and information compiled and provided by the Ontario Ministry of Health. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0107/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0107/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0107/prf

Funding: This study was supported by grant funding from the Ontario Institute for Cancer Research (OICR).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0107/coif). A.V.L. serves as an unpaid editorial board member of Translational Lung Cancer Research from November 2024 to October 2026. A.S. reports consulting or advisory roles, research funding, honoraria for speaking and lectures, and travel reimbursement from Elekta; consulting or advisory roles, honoraria for speaking and lectures, and travel reimbursement from Brainlab SE; honoraria for speaking and lectures from AstraZeneca Canada Inc.; research funding and honoraria for speaking and lectures from Seagen Inc.; honoraria for speaking and lectures and travel reimbursement from Cerapedics Inc.; honoraria for speaking and lectures from Carbofix; and honoraria for speaking and lectures from Les Laboratoires Servier Canada Inc. A.V.L. reports honoraria for speaking and lectures from AstraZeneca Canada Inc. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional ethics review board at Sunnybrook Health Sciences Centre (No. REB 6148), and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Soria JC, Ohe Y, Vansteenkiste J, et al. Osimertinib in Untreated EGFR-Mutated Advanced Non-Small-Cell Lung Cancer. N Engl J Med 2018;378:113-25. [Crossref] [PubMed]
2. Ramalingam SS, Vansteenkiste J, Planchard D, et al. Overall Survival with Osimertinib in Untreated, EGFR-Mutated Advanced NSCLC. N Engl J Med 2020;382:41-50. [Crossref] [PubMed]
3. Reungwetwattana T, Nakagawa K, Cho BC, et al. CNS Response to Osimertinib Versus Standard Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors in Patients With Untreated EGFR-Mutated Advanced Non-Small-Cell Lung Cancer. J Clin Oncol 2018;36:3290-7.
4. Templeton AJ, Booth CM, Tannock IF. Informing Patients About Expected Outcomes: The Efficacy-Effectiveness Gap. J Clin Oncol 2020;38:1651-4. [Crossref] [PubMed]
5. Tjong MC, Doherty M, Tan H, et al. Province-Wide Analysis of Patient‐Reported Outcomes for Stage IV Non‐Small Cell Lung Cancer. Oncologist 2021;26:e1800-e1811.
6. Kim K, Syeed S, Au T, et al. Real-world comparative outcomes of EGFR-TKIs for first-line treatment of EGFR+ metastatic non-small-cell lung cancer. Cancer Treat Res Commun 2025;43:100898. [Crossref] [PubMed]
7. Zhang D, Liu X, Shen F, et al. Osimertinib versus comparator first-generation epidermal growth factor receptor tyrosine kinase inhibitors as first-line treatment in patients with advanced EGFR-mutated non-small cell lung cancer: a Chinese, multicenter, real-world cohort study. Transl Lung Cancer Res 2023;12:2229-44. [Crossref] [PubMed]
8. Tamura K, Yoshida T, Masuda K, et al. Comparison of clinical outcomes of osimertinib and first-generation EGFR-tyrosine kinase inhibitors (TKIs) in TKI-untreated EGFR-mutated non-small-cell lung cancer with leptomeningeal metastases. ESMO Open 2023;8:101594. [Crossref] [PubMed]
9. Planchard D, Jänne PA, Cheng Y, et al. Osimertinib with or without Chemotherapy in EGFR-Mutated Advanced NSCLC. N Engl J Med 2023;389:1935-48. [Crossref] [PubMed]
10. Gadgeel S, Cho BC, Lu S, et al. OA02.03 Amivantamab Plus Lazertinib vs Osimertinib in First-line EGFR-mutant Advanced NSCLC: Longer Follow-up of the MARIPOSA Study. J Thorac Oncol 2024;19:S10-S11.
11. Vogelbaum MA, Brown PD, Messersmith H, et al. Treatment for Brain Metastases: ASCO-SNO-ASTRO Guideline. J Clin Oncol 2022;40:492-516. [Crossref] [PubMed]