Epidermal Growth Factor Receptor Mutation in Resectable Lung Cancer: Association With Survival Outcomes.

Patients and Methods
Lung adenocarcinoma patients who underwent lobectomy between January 2013 and December 2023 at the University of Pittsburgh Medical Center were retrospectively reviewed using a prospectively collected surgical database. Only those with available mutational analysis were included (Supplemental Figure 1). Patients who did not have EGFR testing performed and had evidence of recurrence prior to surgical treatment were excluded. Only lobectomies were used, as this approach represented both the curative-intent standard during the study period and allowed for longer-term follow-up, enabling robust survival analysis. This extended follow-up was facilitated by consistent data capture within a maintained institutional database. Clinical, demographic, and pathological data were included. The study was approved by the institutional review board (STUDY23100047; December 15, 2023), and informed consent was waived due to its retrospective design.
Baseline characteristics of patients with EGFRmu and EGFRwt tumors were compared using the χ2 test for categorical variables and independent sample 2-sided t tests for continuous variables. Survival analysis evaluated 5-year and 10-year survival outcomes and 3-year disease-free survival. Kaplan-Meier estimation generated survival curves, with differences assessed using the log-rank test. Multivariable Cox regression estimated hazard ratios (HRs) and 95% CIs, incorporating EGFR status while adjusting for age, tumor stage (stage IA as reference category), sex, and smoking status, known predictors of mortality in NSCLC. Statistical analysis was performed using IBM SPSS statistics, version 27.0 (IBM Corp), with P values < .05 considered statistically significant.

Results

Cohort Characteristics
From 2013 to 2023, a total of 2733 lobectomies were performed, with 1612 adenocarcinomas. Post exclusion, 578 patients were included: 112 EGFRmu and 466 EGFRwt. The cohort was predominantly female (56%), had smoking history (81%), and exhibited heterogeneous pathologic stage distribution (Table 1). EGFRmu were more likely to be female (79% vs 50%, P < .001), never-smokers (49% vs 12%, P < .001), and have lower pathologic stage (52% vs 32% in stage I, P < .001; Table 2) than EGFRwt. Additionally, EGFRmu had smaller tumor size (2.6 vs 3.0 cm, P = .01) and were more likely to present with synchronous tumors (11% vs 1%, P < .001).

Survival Analysis
Mean follow-up time for EGFRmu was 53 months, and 42 months for EGFRwt. Kaplan-Meier analysis showed significant differences in 5-year (75% vs 50%, P < .001) and 10-year (55% vs 25%, P < .001; Figure 1) overall survival (OS) between EGFRmu and EGFRwt. Stage-specific survival analysis revealed 5-year (95% vs 50%, P = .006) and 10-year (75% vs 0%, P < .001) survival advantage in only stage IB EGFRmu compared with EGFRwt (Supplemental Figure 2). Multivariable Cox regression analysis identified EGFRwt (HR, 1.74, 95% CI, 1.11-2.74; P = .02), male sex (HR, 1.36, 95% CI, 1.03-1.78; P = .03), age (HR, 1.32, 95% CI, 1.13-1.53; P < .001), stage IB (HR, 2.09, 95% CI, 1.18-3.72; P = .01), stage II (HR, 1.93, 95% CI, 1.14-3.28; P = .02), and stage-III disease (HR, 2.97, 95% CI, 1.79-4.91; P < .001) as independent predictors of worse 10-year OS (Figure 2). The 3-year disease-free survival analysis showed that EGFRmu were less likely to experience disease recurrence within 3 years (75% vs 58%, P = .003; Supplemental Figure 3). This difference was statistically significant in stage II (P = .046), while a nonsignificant trend favoring EGFRmu was observed in stage IA (P = .12; Supplemental Figure 4). Notably, in multivariable analysis, EGFRwt status was not a predictor of death within 5 years when there was no metastatic recurrence (HR, 1.05, 95% CI, 0.44-2.51; P = .92), while it remained a significant predictor of death within 5 years after metastatic recurrence occurred (HR, 1.77, 95% CI, 1.01-3.11; P = .048; Figure 2).

Comment
EGFRmu lung cancer presents a unique clinical profile, occurring disproportionately in women and never-smokers. Its unclear etiology makes it a critical area for research, as we currently lack modifiable risk factors and effective screening strategies. Data so far have been inconclusive. Two recent studies have looked at outcomes for patients with EGFRmu who undergo surgery and have found no differences in long-term survival.4,5 On the other hand, a large study conducted in Japan using a cancer registry database found improved survival in EGFRmu patients after adjusting for baseline differences in multivariable regression analysis,6 while another study in the US found improved long-term survival in EGFRmu.7 These differences are not surprising given the unique demographics that EGFR mutations are associated with, notably female sex, non-smokers, and earlier tumor stage, all of which have historically been linked to improved survival in lung cancer.8
Our findings of improved 10-year OS in EGFRmu mirrors the findings of the Japanese registry with similar baseline differences seen between EGFRmu and EGFRwt. When stratified by stage, there were stark survival differences in stage IB and a small trend in stage IA patients that could be explained by the fact that EGFRwt patients include a large subset of NSCLC cases characterized by KRAS mutations. A study by Izar and associates9 found KRAS mutations as a prognostic marker for worse disease-free survival and OS in stage I NSCLC compared with KRAS wild-type and EGFRmu, which could explain the differences seen in this study. Interestingly, our multivariable regression analysis revealed that EGFRmu was associated with improved survival only in patients who experienced disease recurrence, with no survival difference observed among those without recurrence. The observed survival benefit may be multifactorial, with one likely contributor being the use of EGFR-TKIs in EGFRmu patients. However, due to the lack of detailed treatment data, this hypothesis remains beyond the scope of the current study. The improved survival is consistent with prior research, which demonstrated significantly improved postrecurrence survival in EGFRmu patients treated with EGFR-TKIs (49 months), compared with EGFRwt or unknown cases (range, 17-20 months) which was due to TKI treatment.10
Limitations of our study include its retrospective, single-center design, moderate sample size, use of only lobectomies, lack of information regarding TKI use, and unequal follow-up times between groups. Additionally, the predominance of White and female patients may limit generalizability. Nonetheless, strengths include a wide range of tumor stages and extended follow-up, enabling meaningful comparisons.
EGFR-mutated lung cancer may be a distinct disease from tobacco smoke–induced lung cancer, and the long-term prognosis may be different. Screening and treatment might need to be different for this subset of patients, and additional studies could reveal differences in outcomes.