Prognostic analysis of patients with non-small cell lung cancer harboring exon 19 or 21 mutation in the epidermal growth factor gene and brain metastases

Lung cancer is the cancer with the world’s highest morbidity and mortality [1]. Non-small cell lung cancer (NSCLC) accounts for 80–85% of the cases of lung cancer. NSCLC mainly affects adults > 65 years old, tobacco smokers, and men [2, 3]. In China, in 2014, approximately 2,114,000 men and 1,690,000 women have been diagnosed with lung cancer, representing 10,422 new cases each day; in addition, there were 2,296,000 deaths attributable to lung cancer in 2014 [4].

Mutation in the epidermal growth factor receptor (EGFR) gene is now a key target in the treatment of NSCLC. Indeed, afatinib, erlotinib, gefitinib, icotinib, and osimertinib have been shown to improve the prognosis and survival of patients harboring EGFR sensitizing mutations [2, 5].

Brain metastases (BMs) are the main form of distant metastases in lung cancer and is one of the main causes of treatment failure [2, 3]. Approximately 25% of patients with NSCLC suffer from BM, and its occurrence influences survival [2, 3]. As early as 1997, the Radiation Therapy Oncology Group (RTOG) put forward a recursive partitioning analysis (RPA) for the classification of BMs [6], which was the first prognostic scoring system for assessing the prognosis of patients with BM, but this system does not take into account the presence of EGFR mutations. The therapeutic modalities to control BMs include whole-brain radiotherapy (WBRT), stereotactic radiosurgery (SRS), surgery, and chemotherapy, and the best approach has to be tailored to each patient based on the number of lesions, their size, their exact location, and the extent of invasion [2, 3]. Furthermore, the optimal treatment is unknown for EGFR-mutated patients with NSCLC and BMs [7], and this important research question remains to be examined.

Many patients with lung cancer develop BMs, which impacts the quality of life and shortens survival. Despite therapy, the prognosis of NSCLC patients with BMs is poor, and the 1-year survival rate is < 20% [12]. Previous studies found a significant association between EGFR mutations and the risk of BM [13, 14] and pointed out the distinct clinical features of EGFR-mutated tumors in terms of BM [15‐18]. Therefore, it is speculated that BMs in these patients exhibit their own characteristics in occurrence, treatment, and prognosis. In 1997, the RTOG put forward the recursive partitioning analysis classification for the prognosis of BMs, but this system does not take into account the epidermal growth factor receptor (EGFR) mutations present. Accordingly, this study aimed to summarize the factors affecting the prognosis of these patients with EGFR-mutated lung adenocarcinoma after BM. Furthermore, this study explored the optimal treatment for these patients.

Our results indicated that the number of BMs and treatment after BM were associated with overall survival after BMs. Previous studies concluded that the performance status [6, 19‐21], age [6, 19‐21], extracranial metastases [6, 19‐21], and primary tumor control [19, 20] affected survival. Other studies [12, 22, 23] indicated that the number of BMs influenced survival. The choice of treatment should be based on the current guidelines and tailored to the clinical reality of each patient. Better physical strength generally means better tolerance. Nevertheless, our results were different from previous studies, probably because previous studies did not target patients with NSCLC harboring EGFR mutation and BMs. Few studies discussed the treatment factors influencing the prognosis of patients with BM and EGFR mutation. Gong et al [24]. indicated that the number of chemotherapy cycles and combined targeted therapy was key prognostic factors influencing survival. Our results indicate that the treatments after BM were associated with mOS. Due to the relatively small number of patients in each group, we were unable to exhaustively assess the factors that were correlated with the prognosis of patients with BM.

The first-generation EGFR-TKIs available in China during the study period included gefitinib, erlotinib, and icotinib. The CTONG0901 study compared the PFS and OS of gefitinib and erlotinib and found that the two were equivalent [25]. The ICOGEN study was a randomized, controlled phase III clinical trial comparing gefitinib to icotinib in previously treated patients with locally advanced or metastatic non-small cell lung cancer. The results showed that there was no significant difference in PFS and OS between gefitinib and icotinib [26]. The WJOG5108L clinical trial also showed that gefitinib and erlotinib were equivalent in PFS [27]. In clinical practice, which TKI a patient chooses is related to the patient’s choice and the doctor’s prescription habits. In the present study, gefitinib was used in 20 patients, erlotinib was used in 25 patients, and icotinib was used in 8 patients. Due to the price advantage of icotinib, some patients chose to use it. During treatment, no advantage in the efficacy of a certain drug was found, and all three drugs were not found to have grade III-IV adverse reactions.

In the present study, patients with targeted therapy combined with radiotherapy after BM had the best survival advantage. The proportion of patients with CR or PR following BMs was significantly different across treatment groups. The proportion of CR + PR was 63.0% for radiotherapy, 26.7% for chemotherapy, 50.0% for targeted therapy, and 89.7% for targeted therapy combined with radiotherapy. After BM, the median survival of the four treatment groups was 20, 9, 12, and 25 months, respectively (P < 0.05), and their median iPFS were 12, 7, 10, and 21 months, respectively. The prognosis of chemotherapy was the worst, similar to a previous report [28]. This is thought to be due to several factors, including the blood-brain barrier (BBB) and the inherent chemotherapy resistance of BM. Thus, WBRT has been used as a standard treatment in NSCLC patients with BM, resulting in an OS ranging between 3 and 6 months since the 1970s [29, 30].

TKIs are small molecules and have a good lipid-water partition coefficient. They are easily absorbed and cross the BBB. Brain metastases can damage the BBB to some extent [31]. More recently, TKI therapy for BM patients with EGFR mutations achieved effective rates of 70–80% [32] and 87.8% [33]. Furthermore, the iPFS was 14.5 months, and the OS was 21.9 months. Nearly half of the patients delayed radiation therapy for more than 1.5 years after the diagnosis of BMs by using TKI [33]. Accordingly, some experts pointed out that TKI was becoming a favorable treatment, especially for patients with EGFR mutation of BMs of lung cancer. In the present study, the radiotherapy group did show some advantages over the targeted treatment group, probably because most patients had no more than three BMs whose maximum diameter < 2 cm, and they accepted stereotactic radiotherapy. Omuro et al. [34] and Park et al. [32] also drew similar conclusions, pointing out that TKI therapy for NSCLC brain metastases leads to a high intracranial recurrence rate and short PFS. The retrospective analysis by Magnuson et al. [35] showed that radiotherapy, compared with TKI treatment, contributed to a longer survival (34.1 vs. 19.4 months). PET/CT images with ¹¹C-erlotinib as the tracer combined with the MRI images show a significant concentration of ¹¹C-erlotinib in the brain metastases, but no ¹¹C-erlotinib could be found in the normal brain tissues [36]. In mouse models, compared with other EGFR-TKIs, osimertinib reaches a higher concentration in the brain and is easier to accumulate in the brain [37]. In eight healthy adult volunteers (52 ± 8 years old), PET-CT was used to observe the distribution of ¹¹C-osimertinib in the brain after a single intravenous injection of 1.2 μg (1.1–1.4 μg) over 90 min. It was found that ¹¹C-osimertinib could distribute rapidly in the brain, with an average Tmax of 13 min and a brain/plasma AUC_{0–90 min} ratio of 8.3 ± 0.3 [38]. In the AURA3 study, the ORR of the central nervous system was 70% in the 80-mg osimertinib group and 31% in the platinum-containing dual drug chemotherapy group [39]. The ASTRIS open-label, real-world, international single-arm treatment study aimed to explore the efficacy and safety of osimertinib in T790M-positive advanced NSCLC adult patients with EGFR-TKI treatment history. The results showed that for advanced NSCLC patients with the T790M mutation and asymptomatic stable CNS metastasis treated with osimertinib, the overall ORR of T790M positive was 55% and the median PFS was 9.7 months [40]. Therefore, osimertinib could have particular benefits for patients with NSCLC and brain metastases. The role of radiotherapy in the treatment of BMs still requires additional studies, and its timing in relation to different TKIs requires additional study. Nevertheless, some studies suggested that upfront TKI and radiotherapy achieved better survival than TKI alone in patients with BMs from NSCLC [41‐43].

TKIs have a radio-sensitizing effect [44, 45], and radiotherapy can disrupt the BBB to improve TKI levels in the intracranial space, and this mechanism provides a theoretical basis for the idea of targeted treatment combined with radiotherapy [46]. Zeng et al. [47], Cai et al. [48], and Welsh et al. [49] supported the hypothesis that TKI combined with WBRT is more effective for the control of intracranial lesions and prolonging the survival than either therapy alone. The benefits seemed exceptionally high for patients with EGFR mutation rates. Taken together, the present provides a comprehensive comparison of the various treatments. Larger prospective randomized clinical trials are needed to validate our findings and confirm these suppositions.