Introduction: Surgical resection is the primary treatment with curative intent for early-stage NSCLC; however, many patients experience recurrence even after resection. The goal for this study was to assess the real-world disease-free survival (rwDFS) and patient characteristics, including epidermal growth factor receptor mutation (EGFRm) status, associated with rwDFS in stages IB-IIIA NSCLC patients. Methods: This was a retrospective cohort study using COTA’s real-world US oncology database through 31 Dec 2021. Adults diagnosed with stages IB–IIIA NSCLC (based on AJCC criteria at the time of diagnosis) from 1 Jan 2013 to 31 Dec 2018 who had undergone complete surgical resection (index date) of the primary disease were identified. Median rwDFS (time from the index date to the earliest of death or disease recurrence) was estimated using the Kaplan-Meier method. A multivariable Cox-PH model stratified by year of diagnosis was developed to assess patient characteristics associated with rwDFS. Results: 703 patients met the study criteria (median age 67 years, female (56%), white (82%), treated in the northeast region (67%), median follow-up of 37.4 months from the index date). 47.9% of patients developed recurrent disease (N =284) [mode of recurrence: distant/metastatic (77.8%); localized (10.9%); regional (10.6%); unknown (0.7%) or death (N = 53)] with a median rwDFS of 42.9 (95% CI: 37.4–52.2) months. 17.4% of the study population was EGFRm positive [stage IB (33.6%); II (27.9%); IIIA (38.5%)], 57.5% EGFRm negative [stage IB (24.3%); II (36.8%); IIIA (38.9%)], and 25.2% EGFRm unknown. The rate of rwDFS events was slightly higher among EGFRm positive patients (50.8%), and the median rwDFS for this group was 31.5 months, compared to 47.9 and 48.5 months for the EGFRm negative group (rwDFS event rate: 46.3%) and unknown EGFRm status (rwDFS event rate: 49.7%), respectively. After adjustment for key patient covariates; disease stage and adjuvant therapy status were significantly associated with rwDFS in the study population (N =703) (Table 1). Conclusions: This analysis shows that early-stage NSCLC patients with later stage diagnosis and without adjuvant therapy had higher risk for disease recurrence or death. Further investigations are needed to explore additional patient characteristics that might be associated with the decision to use adjuvant therapy and rwDFS for these patients.
Stratified multivariable Cox-PH model to assess characteristics associated with rwDFS in patients with surgically resected NSCLC patients (state IB–IIIA)†
*p<0.05; # Patients with evidence of chemo or radiation therapy; † for the purpose of abstract, only selected variables/levels from the COX regression analysis are shown.
