Clinical Impact of EGFR Mutation Subtypes on Treatment Outcomes in Advanced Non-Small Cell Lung Cancer: An Austrian Real-World Study

1. Introduction

Lung cancer is with 1.8 million deaths the leading cause of cancer related deaths [1]. As the data indicate, lung carcinoma is one of the top aggressive cancers with poor outcomes. Almost half of the non-small cell lung cancer (NSCLC) patients have stage IV at initial diagnosis [2]. The 5-year survival rate persists at a very low level [3] and roughly 75% of patients diagnosed with advanced lung cancer have a bad prognosis [4]. Chemotherapy only achieves median survival rates up to 14 months [5]. The approval of epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitors (TKIs) led to a decrease of mortality and transformed the landscape of NSCLC treatment and prognosis [6,7]. According to phase III clinical trials, EGFR-TKIs are superior to chemotherapy regarding to prolonging progression-free survival (PFS) [8,9] and overall survival (OS) [10].

EGFR mutations are the second most prevalent oncogenic drivers in NSCLC. The most common activating mutations (exon 19 deletion and L858R point mutation) comprise the vast majority of EGFR mutations (80-85%), whereas uncommon mutations (Exon 18, 20, 19 and 21) are less frequent (15-20%) [11,12]. EGFR exon 20 insertions are the third most common genotype and occur in 1-10% of NSCLC patients [13]. Since most of the patients with EGFR positive NSCLC carry either an exon 19 deletion or exon 21 L858R point mutation, data of uncommon EGFR genotypes are lacking[14].

Common mutations are defined as predictors for good clinical response to EGFR-TKIs [15]. As reported by previous studies, NSCLC patients with exon 19 deletion showed better OS and PFS than patients with L858R mutation [16,17]. Based on a meta-analysis, the EGFR mutation prevalence is higher in the Asian population (49.1%) than in Western countries (12.8%) [7]. Therefore, data on survival outcome for patients in Europe harboring advanced NSCLC and treated with EGFR-TKIs in first-line palliative treatment have been lacking.

It is expected that the results of this prospective real-world study will bring more evidence for the use of EGFR-TKIs in first-line treatment in European advanced NSCLC patients with EGFR mutations. Additionally, the comparison of the EGFR genotypes regarding survival outcome will deliver important prognostic information.

2. Materials and Methods

This observational cohort study evaluated real-world overall survival (rwOS) and real-world progression-free survival (rwPFS) survival outcome of EGFR-positive advanced (stage IIIB or higher) NSCLC patients who were treated with TKIs in first-line regarding to EGFR subtype. The study followed a prospective design, all patients signed an Informed Consent Form (ICF). Patient data were collected at two high-volume clinics in Austria (Department of Respiratory and Critical Care Medicine, Klinik Floridsdorf, Vienna; Department of Internal Medicine with Pulmonology, Klinik Ottakring, Vienna) between November 2020 and the last follow-up in February 2025. The EGFR mutation status was ascertained by next generation sequencing from tumor tissue (ThermoFisher OncomineTM Focus assay, Ion AmpliSeqTM lung cancer panel). The clinical characteristics and treatment data were predefined and retrospectively extracted from Longitudinal Analysis of LUng CAncer registry data (LALUCA, EK 20-061-VK, city of Vienna) [18], which was implemented by the Karl Landsteiner Institute for Lung Research and Pulmonary Oncology in Vienna, Austria.

The survival analysis was estimated with Kaplan-Meier curves for rwOS and rwPFS. For both, rwOS and rwPFS, EGFR exon 19 deletion was compared with EGFR exon 21 L858R mutation and each of them with uncommon mutations in EGFR exon 18-21.

Univariate and multivariable Cox proportional hazards regression was calculated to identify independent prognostic factors for OS and PFS, the results were additionally demonstrated in a forest plot for visualization.

2.2. Statistical Analysis

Statistical analysis was performed in the open-source software R-4.4.1 and IBM SPSS Statistics version 30.0. The demographic and clinical characteristics of the study population (n = 53), as well as the results of the corresponding group comparisons, were conducted with Fisher’s exact test.

The influence of TKIs as well as EGFR mutation subtypes upon OS and PFS in EGFR-positive NSCLC patients was estimated with Kaplan-Meier analyses. The potential differences between the survival curves were analyzed by using the log-rank test. OS was calculated from the start of first-line EGFR-TKI treatment to death from any cause. PFS was defined as the time between initiation of first-line EGFR-TKI treatment and disease progression or death. Participants who were alive were censored at the last date known alive. Cox proportional hazard regression model was assessed for the identification of possible independent prognostic factors of OS and PFS. For p-values, the established 5% α-level was used in order to draw conclusions about statistical significance.

3. Results

In total, 145 patients with an EGFR mutation (11%)were identified out of 1,267 patients (844 adenocarcinoma, 302 squamous cell carcinoma, 117 NOS or large cell carcinoma, 4 adenosquamous carcinoma) with NSCLC . The diagnosed EGFR variants were categorized into EGFR common und EGFR uncommon mutations, shown in Table 1. Of the 145 patients with an EGFR mutation, 87 (60%) had a common EGFR mutation and 58 (40%) harbored an uncommon EGFR mutation.

Among 145 patients with pathologically confirmed advanced NSCLC with known EGFR mutation status, 53 cases were included for evaluation, based on the following eligibility criteria: harboring advanced NSCLC, somatic EGFR mutation and having received TKIs as first-line-treatment (palliative therapy), illustrated in Figure 1. The final sample size was subdivided into the common mutations (n = 36) and uncommon mutations (n = 17) groups.

Demographic and clinical characteristics are detailed in Table 2., tumor characteristics, including the stage at initial diagnosis, location of metastasis, sites of metastasis, histologic subtype and PD-L1 status are presented in Table 3. The patients received second-generation and third-generation TKIs in first-line, which is summarized in Table 4, including response.

The median age at initial diagnosis was 73 years (range, 52-85 years), patients ≥ 65 years (66%) were predominant. The vast majority of the patients were Caucasian (91%) and female (60%). Approximately half of the patients (57%) had a body mass index (BMI) ranging from 18.5 to 24.9. Most of the patients were never smoker (43%) or former smoker (38%); 19% were heavy smokers (≥ 30 pack years). Based on the performance status (ECOG), the majority of the patients had ECOGs of 0 (60%), 30% had ECOGs of 1, 6% had ECOGs ≥ 2 and 4% were not evaluated. Most of the patients with exon 19 deletion were female (68%), whereas the gender dissemination of patients with an exon 21 L858R mutation or uncommon mutations was almost equally distributed.

All variables (age, sex, ethnicity, smoking status, pack years, BMI and ECOG performance status) did not significantly differ between the EGFR exon 19 deletion, EGFR exon 21 L858R mutation and uncommon mutations. (p > 0.05) groups.

As shown in Table 3, most of the patients (93%) had stage IVA (42%) or stage IVB (51%), only 8% had stage IIIB. Most patients harbored an adenocarcinoma (91%), followed by squamous cell carcinoma (6%) and large cell carcinoma (4%). A negative PDL-1 expression (< 1%) was found in 25 patients (47%), 21 patients (40%) had a PDL-1 expression of 1-49%. In total, 6% had a high (50-89%) or a very high PDL-1 expression (≥90%). The primary sites of metastasis were pleura (42%), brain (26%), and bones (34%). Rare locations of metastasis were distant lymph nodes (9%), adrenal gland (6%) and the liver (11%). The cumulative percentages for metastatic locations will not sum to 100%, since patients can exhibit multiple metastatic sites simultaneously.

All variables (stage at initial diagnosis, histologic type, PDL-1 status, location and site of metastasis) did not significantly differ between the EGFR exon 19 deletion, EGFR exon 21 L858R mutation and uncommon mutations. (p > 0.05) groups.

Patients included in this study received second-generation and third-generation TKIs in the first-line. Details regarding the treatment pattern are summarized in Table 4. Most of the study population received Osimertinib (third-generation TKI) (58%) or Afatinib (second-generation TKI) (32%). Of the 53 patients, four (8%) out of six patients with an exon 20 insertion were treated with Mobocertinib. One patient harbored additionally an ALK translocation and was treated with Brigatinib, a second-generation ALK TKI.

None of the patients had a complete response in the first-line, 36% had a partial response, followed by 11% with a stable disease and 4% with progressive disease.

3.1. Survival Analysis

3.1.1. Overall Survival Analysis

The median rwOS for the whole cohort (n = 53) was 17.7 months (95% CI, 10.4-24.9 months). Patients harboring an EGFR exon 19 deletion had a significantly longer overall survival compared with patients with EGFR uncommon mutations (32.5 vs. 11.6 months, p = 0.002; Figure 2b). There was no significant difference in the median overall survival between patients with EGFR exon 21 L858R mutation and uncommon mutations (17.4 vs. 11.6 months, p = 0.094; Figure 2c), and between the EGFR exon 19 deletion and EGFR exon 21 L858R mutation groups (32.5 vs. 17.4 months, p = 0.74; Figure. 2a).

A summary of the median OS times are shown in Table 5, including the respective 95% CI. EGFR exon 19 deletion showed the longest median overall survival with 32.5 months (CI: 30.8-34.3 months), followed by EGFR exon 21 L858R mutation with 17.4 months (CI: 17.0 – 18.1 months). The shortest overall survival was of patients with uncommon mutations with 11.6 months (CI: 9.2-14.0 months).

The univariate and multivariable Cox proportional hazards analysis of overall survival, including the variables age, sex, histologic type, BMI, ECOG performance status, sites of metastasis, PDL-1 expression and EGFR mutation subtype is shown in Table 6. The univariate analysis revealed that overall survival was longer in patients with common mutation (HR: 3.7; 95% CI: 1.6-8.7; p = 0.002), ECOG status of 0 (HR: 2.3; 95% CI: 1.1-4.5; p = 0.018) and patients with adenocarcinoma (HR: 4.3; 95% CI: 1.4-14; p = 0.013). In accordance with the univariate analysis, the multivariable analysis showed that NSCLC patients with a common mutation had a longer overall survival compared to the uncommon mutation subgroup (HR: 3.7; 95% CI: 1.23-11.2; p = 0.02). No significant difference regarding OS could be identified for age, sex, histologic type, ECOG status, sites of metastasis, PDL-1 expression and BMI.

In conclusion, an EGFR common mutation subtype (HR = 3.71) was an independent predictor of better overall survival. The results are visualized in Figure 3.

3.1.2. Progression-Free Survival Analysis

The median rwPFS in the entire cohort (n = 53) was 14.2 months (95% CI: 7.4-20.9 months). There was no significant difference in the median progression-free survival between all groups: patients with EGFR exon 19 deletion and EGFR exon 21 L858R mutation (23.1 vs. 17.0 months, p = 0.989; Figure 4a), patients with EGFR exon 19 deletion and EGFR uncommon mutations (23.1 vs. 10.0 months, p = 0.078; Figure 4b), patients with EGFR exon 21 L858R mutation and uncommon mutations (17.0 vs. 10.0 months, p = 0.307; Figure 4c).

Median PFS times are shown in Table 7, including the respective 95% CI. EGFR exon 19 deletion showed the longest median progression-free survival with 23.1 months (CI: 6.8-39.4 months), followed by EGFR exon 21 L858R mutation with 17.0 months (CI: 5.2 – 28.7 months). The shortest OS (10 months) had also patients with uncommon mutations (CI: 3.4-16.4 months).

The univariate and multivariable Cox proportional hazards analysis of progression-free survival, including the variables age, sex, histologic type, BMI, ECOG performance status, sites of metastasis, PDL-1 expression and EGFR mutation subtype is demonstrated in Table 8. Both, univariate and multivariable analysis revealed no significant associations.

The results are visualized in Figure 5.

4. Discussion

In the present study, we examined a prospective cohort of 53 patients with advanced NSCLC harboring an EGFR exon 18-21 mutation from two high-volume institutions in Vienna, Austria.

The beneficial survival outcome of EGFR-TKIs in NSCLC patients in first-line as well as the prognostic impact of clinicopathological characteristics and EGFR subtype differences have been examined previously, but predominantly in the Asian, Middle East and North Africa, Moroccan and especially in the Chinese population [19,20,21]. To the best of our knowledge, this is one of the first real-world analyses to investigate the survival outcome after first-line EGFR-TKI treatment (second and third-generation) as well as the prognostic role of EGFR genotypes and clinical and pathological characteristics in advanced NSCLC patients in Austria.

In the present cohort, the majority of the patients were female (60%), Caucasians (91%) and never or former smoker (81%). Differences between common and uncommon EGFR mutations failed to reach significance between age groups, sexes, ethnicities, smokers and non-smokers, pack-years, or for varying ECOG status and BMI. Previous studies unveiled an association between distinct EGFR genotypes and smoking habits. Albeit not nominally significant, more than half of the patients with EGFR uncommon mutations are smokers or former smokers in contrast to patients harboring common mutations, where more than half of the patients are never smokers, supporting prior evidence [15,19,22,23,24,25]. The higher number of females in our sample reflects the previously shown [19,22,25,26,27] increased likelihood in women with NSCLC of developing EGFR mutation compared to men. Also in line with previous findings, we provide tentative evidence for higher rates of EGFR exon 19 deletion within females [19,22,28]. A substantial number of the patients had adenocarcinoma (91%), stage IV (93%) and one metastasis (55%). For both, the clinicopathological characteristics, as well as tumor characteristics, no significant differences in the number of common and uncommon mutations could be found.

Within the scope of the present study, we also investigated potential prognostic factors. Multivariable cox regressions revealed that an EGFR common genotype is an independent prognostic factor for improved median overall survival (mOS). No significant association could be found concerning age, sex, histologic type, ECOG performance status, BMI, sites of metastasis and PDL-1 expression. The observations align with a previous study, where female gender is additionally mentioned as an independent prognostic factor for improved overall survival [19]. The strongest independent factor for prolonged overall survival is the EGFR common genotype, which is consistent with previous investigations [2,29]. Regarding to median progression-free survival (mPFS), independent prognostic factors could not be identified.

The survival analysis of the study demonstrated that patients with common EGFR mutations exhibited significantly longer mOS compared with those with an uncommon EGFR genotype. Moreover, within EGFR common mutations, exon 19 deletion surpassed (although not nominally significant) the mOS compared to exon 21 L858R mutation cohort. These results are concordant with previous real-world studies and clinical trials [4,19,20,29,30]. Patients with EGFR exon 19 deletion achieved better mPFS, followed by EGFR exon 21 L858R; uncommon mutations had the poorest mPFS. These findings are overall in line with foregoing studies [16,17,30].

Limitations

Some limitations have to be acknowledged. Firstly, the relatively small sample size may restrict the statistical power for various of the subgroup analyses conducted. Furthermore, the treatment options have changed over the period the data were collected. The vast majority of patients were treated with Afatinib (second-generation) or Osimertinib (third-generation), which may have had an impact on survival outcome. According to previous study data, the treatment with Osimertinib in first-line resulted in a significantly longer mPFS and mOS than with comparator TKIs [10,31,32]. Moreover, therapy options for patients harboring rare EGFR mutations are more restricted than those for common mutations, which may be one reason for poorer survival outcome. In addition, patients carrying rare EGFR mutations are less responsive to TKIs [22,24].

5. Conclusions

The findings of the present survival analysis in a real-world setting are in agreement with the results of clinical trials. Our results underline the prognostic role of distinct EGFR genotypes and the relevance of corresponding analyses in NSCLC patients. The study also highlights the challenges regarding to EGFR uncommon mutations and the resulting need for further research to investigate alternative treatment options.