A real-world study found just one-third of patients remained on adjuvant osimertinib at 3 years, with early discontinuation linked to worse disease-free survival.
Among patients with resected, EGFR-mutated non–small cell lung cancer (NSCLC) treated with adjuvant osimertinib (Tagrisso) in US community and academic practice, only one-third remained on therapy at 36 months despite a recommended 3-year treatment course, according to findings from a retrospective cohort study presented in a poster session at the
The study, which used the US Flatiron Health Research Database to evaluate real-world duration of therapy (rwDOT) and real-world disease-free survival (rwDFS) in 822 patients who initiated adjuvant osimertinib between January 1, 2021 and November 30, 2024, found a median rwDOT of 33.0 months (95% CI, 29.6-35.2). Kaplan-Meier estimates showed 75% of patients remained on treatment at 12 months, 62% at 24 months, and 33% at 36 months.
Patients who discontinued osimertinib before the 12-month or 24-month landmarks had a significantly higher adjusted risk of recurrence or death compared with those who continued therapy (12-month landmark HR, 2.34; 95% CI, 1.28-4.29; 24-month landmark HR, 2.74; 95% CI, 1.22-6.13). The association was not statistically significant at the 34-month landmark, which investigators attributed to limited follow-up and fewer events beyond that time point.
Investigators, led by Eric K. Singhi, MD, of The University of Texas MD Anderson Cancer Center, conducted the retrospective cohort study using de-identified electronic health record data from the nationwide Flatiron Health Research Database. Eligible patients were 18 years or older with stage I to IIIA EGFR-mutant NSCLC who underwent surgical resection and initiated adjuvant osimertinib, with documented EGFR mutation status and a minimum of 3 months of potential follow-up.
The index date was defined as the date of adjuvant osimertinib initiation, with follow-up continuing until death, loss to follow-up, or the data cutoff of February 28, 2025. rwDOT was measured from index date to treatment discontinuation, and rwDFS from index date to recurrence or death; both were estimated using the Kaplan-Meier method. Associations between discontinuation and rwDFS were assessed with adjusted Cox proportional hazards models at 12-, 24-, and 34-month landmarks to mitigate immortal time bias, adjusting for age, NSCLC stage, prior adjuvant or neoadjuvant therapy, prior radiation, and smoking history (the 12- and 24-month models were additionally adjusted for socioeconomic status). An exploratory multistate survival analysis using Nelson-Aalen cumulative hazard curves further characterized hazard accumulation across treatment, discontinuation, recurrence, and death states.
At baseline, the study population had a median age of 71.0 years (IQR, 64.0-76.0); 73.5% were female, and 47.2% had a smoking history. Disease stage at diagnosis was stage I in 39% of patients, stage II in 34%, and stage IIIA in 27%. Most patients (77.6%) were treated in community oncology practices.
Kaplan-Meier estimated rwDFS probability was 95% at 12 months, 89% at 24 months, and 78% at 36 months, rates the investigators noted were comparable to disease-free survival outcomes reported in the phase 3 ADAURA trial (NCT02511106) at 24 and 36 months (90% and 85%, respectively).² ADAURA established the 3-year treatment duration as standard of care after demonstrating disease-free and overall survival benefit over placebo in resected EGFR-mutant NSCLC.²,³
Among the 336 patients who discontinued osimertinib during follow-up, 39 (12%) had a documented recurrence within 30 days before or after discontinuation. Of the remaining 297 patients without a recurrence in that window, 154 (51.9%) had a documented real-world adverse effect (rwAE) and 143 (48.1%) had neither a recurrence nor a rwAE recorded. Among patients who discontinued before 24 months specifically, a higher proportion had an rwAE documented at the time of discontinuation than a recurrence, a pattern the investigators said underscores the importance of adverse effect (AE) management in supporting treatment persistence. The most frequently reported rwAEs within 30 days of discontinuation across the overall population were fatigue (22.9%), loss of appetite (15.8%), diarrhea (15.2%), rash (14.5%), and nausea/vomiting (11.8%). The exploratory multistate survival analysis was consistent with the landmark findings, showing a faster rate of recurrence following treatment discontinuation than during active treatment.
The investigators concluded that in US real-world practice, a substantial proportion of patients discontinued adjuvant osimertinib before completing the recommended 3-year course, and that early discontinuation was associated with more than a two-fold higher risk of recurrence or death. They noted that proactive management of AEs, together with effective clinician-patient communication about the benefits and risks of continuing therapy, is likely important for supporting treatment persistence and improving outcomes in this population.
The authors acknowledged several limitations, including the absence of explicit, investigator-confirmed reasons for discontinuation; recurrence and rwAEs near discontinuation were used as a proxy rather than a confirmed cause as well as potential residual confounding from unmeasured factors, incomplete capture of variables including EGFR mutation subtype, race, and ECOG performance status, and limited generalizability beyond the US community oncology setting.
