Abstract
Background/Aim: Real-world evidence regarding the prevalence of epidermal growth factor receptor (EGFR) mutation-positive status (M+) and the clinicopathological characteristics associated with the presence of EGFR mutations in advanced non-small cell lung cancer (NSCLC) is scarce, especially among Caucasian populations. The present study aimed to bridge this gap, as well as to record treatment patterns and outcomes in routine-care settings. Patients and Methods: REASON (NCT01153399) was a prospective study of patients with stage IIIB/IV NSCLC and known EGFR mutation status. Clinicopathological, treatment characteristics and clinical outcomes were recorded and correlated with EGFR mutation testing results. Results: Of 575 enrolled patients, EGFR mutations were detected in 15.7% of them. Male gender (p=0.008) and smoking (p<0.001), but not adenocarcinoma, were associated with EGFR M+ status. In the EGFR M+ subpopulation (n=88), absence of bone and/or brain metastasis and presence of exon 19 EGFR M+ status at diagnosis were independently associated with longer progression-free survival (PFS) (p=0.011 and p=0.040, respectively). Conclusion: In our population, males and smokers had decreased odds of harboring an EGFR mutation, while adenocarcinoma histology was not a significant predictor of EGFR M+ status. EGFR M+ patients with bone and/or brain metastases at diagnosis or mutations other than exon 19 deletions were at increased risk for earlier disease progression.
In 2012, 353,000 deaths in Europe were attributed to lung cancer, the most common cause of cancer-related deaths. In Greece, lung cancer was estimated to be the leading cause of cancer-related deaths among men (age standardized rate (ASR): 67.7 per 100,000), and the second leading cause of cancer-related deaths among women (ASR: 11.8 per 100,000), while its incidence was ranked highest in men (ASR: 74.7 per 100,000), and as the third highestin women (ASR: 13.2 per 100,000) after breast and colorectal cancer (1).
Non-small cell lung cancer (NSCLC) comprises about 85% of all lung cancer diagnoses (2). Most NSCLC patients are initially diagnosed at an unresectable locally advanced (stage IIIB) or metastatic (stage IV) stage (3). Prognosis of stage IIIB/IV NSCLC is poor, with a median overall survival (OS) of about 10 months (4), and a 5-year relative survival rate of metastatic disease of merely 4.5% based on 2007-2013 data from the Surveillance, Epidemiology, and End Results Program (5).
Routine treatment strategies for stage IIIB/IV NSCLC include chemotherapy, radiotherapy and targeted therapy and are guided by tumor histological subtype, molecular profiling and genetics, as well as the patient's age, performance status (PS) and preferences (6-8). The addition of targeted agents to the treatment armamentarium of NSCLC was a major breakthrough, offering clinically meaningful benefits for patients harboring epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase 1 (ALK1) mutations, as well as ROS1 rearrangements. EGFR tyrosine kinase inhibitors (EGFR-TKIs) and ALK1-TKIs have provided a paradigm shift in the management of advanced NSCLC, representing the pioneers of personalized treatment options and solidifying the importance of molecular testing as part of the diagnostic algorithm (9, 10).
The frequency of EGFR mutation-positive (M+) tumors is much higher among Asian-Pacific (30-50%) (3,11-14) than Caucasian (10-17%) populations (15-20). Additionally, a positive EGFR mutation status has been associated with a non-smoking history, female sex and adenocarcinoma histology (3, 10). High NSCLC incidence and mortality in Greece along with the anticipated interethnic variations in terms of genetic profile and treatment outcomes of those patients fueled the need for conduct of real-life epidemiological studiesat a country level. Thus,essential information could be recorded in order to support evidence-informed decision-making for the routine-care management of advanced NSCLC disease.
Under this perspective, the present study sought to collect epidemiological data regarding EGFR mutation status from patients with stages IIIB/IV NSCLC in Greece and determine its association with smoking status, gender, and tumor histology. In addition, the study aimed to capture information on clinical outcome data (progression-free survival (PFS), OS and disease control rate (DCR), first-line treatment patterns employed in stage IIIB/IV NSCLC patients, regardless of EGFR mutation status and gain preliminary insight on the healthcare resource utilization of EGFR M+ patients treated in routine care settings.
Materials and Methods
Study design, population and setting. REASON (NCT01153399) was a multicenter, prospective observational study, carried out by 22 hospital-based physicians specializing in oncology (n=19) or lung diseases (n=3) from representative geographic regions of Greece. Consecutive enrollment of patients attending the study sites that met the study specific eligibility criteria was employed as means to control for and minimize patient selection bias. Routine assessments were performed under real-life conditions without intervening in patient selection, diagnostic procedures employed, or therapeutic decision-making.
At enrollment, for all patients with histologically confirmed stage IIIB/IV NSCLC and EGFR mutation-negative (M−) or EGFR non-evaluable (Mx) status, as well as those with EGFR mutation-postive (M+) status who wished to participate in any interventional trial, data pertaining to clinicopathological characteristics and first line treatments were collected; for these patients, participation was completed at the enrollment visit. For EGFR M+ patients for whom participation in an interventional study was not foreseen, study participation ended one year after the last patient was included into the study, unless the patient i) wished to end his/her participation in the study earlier, ii) experienced disease progression or died, or iii) was lost to follow-up. Study follow-up information for these EGFR M+ patients included response evaluation based on the treating physician's routine assessments and without mandating the use of standardized tumor response evaluation criteria, survival status and healthcare resource utilization. All study data were recorded on paper case report forms.
The study was performed in accordance with the International Society for Pharmacoepidemiology guidelines for Good Pharmacoepidemiology Practice, the ICH-GCP guidelines (where applicable) the ethical principles of the Declaration of Helsinki and all standing regulations. As per the national regulations, the original study protocol including the final version of the patient's Informed Consent Form (ICF), were reviewed and approved by the competent institutional review boards of the participating Hospital Sites and by the Greek National Organization for Medicines (EOF), before the enrollment of any patient into the study and the performance of any study-related procedure. There was one protocol amendment (extending the recruitment period and increasing the number of participating sites in order to meet the study target) which was approved by the IRBs of the participating hospitals as per the standing national regulations.
Study population. The eligible study population comprised of newly diagnosed and untreated males and females aged ≥18 years with histologically confirmed stage IIIB/IV NSCLC and known EGFR mutation status (i.e. EGFR M+, EGFR M− or Mx), who at enrollment were treated in the first-line setting and whose tumor was not amenable to curative surgery or radio-chemotherapy. Patients with mixed histology of small and non-small cell lung cancer were excluded from the study.
Study objectives and endpoints. The study primarily aimed to collect epidemiological data on the frequency of EGFR M+ NSCLC in a population of predominantly Caucasian ethnicity, and to elucidate the association of smoking status, gender and tumor histology with EGFR mutation status. Secondary study objectives were to capture the real-life management patterns in the overall population and EGFR M+, M− and Mx subpopulations, and to assess clinical outcomes (PFS, OS, DCR) and healthcare resource utilization in terms of hospitalizations and outpatient visits in routine care settings of Greece among EGFR M+ patients.
Statistical methods. All enrolled patients with histologically confirmed stage IIIB/IV NSCLC and with available EGFR mutation status information have been included in the analysis of the primary aim of the study (Full Analysis Set – FAS). Patients fulfilling all eligibility criteria have been included into the dataset for the evaluation of secondary endpoints (Per Protocol Analysis Set – PP).
The association of smoking status, gender and histological type with the EGFR mutation status was examined through simple logistic regression analysis as well as by a multiple logistic regression model. In order to estimate the median PFS and OS times, the Kaplan–Meier method was applied. Association of age at the time of diagnosis (>65 years vs. ≤65 years), smoking status at enrollment (never smoker versus smokers), histological subtype at initial diagnosis (non-adenocarcinoma versus adenocarcinoma), presence of bone and/or brain metastases, presence of exon 19 mutation with PFS was assessed through a multivariable Cox proportional hazard model estimating the hazard ratios (HR) and the relevant 95% confidence intervals (CIs). DCR has been defined as the percentage of patients who had achieved at least a complete response, partial response or stable disease. Clopper-Pearson 95% exact CIswere calculated. The hospitalization and outpatient visit rates expressed in person-years have been calculated by dividing the total number of hospitalizations and outpatient visits respectively, by the ‘patient-year at risk’ time, i.e. the days elapsed from enrollment to study completion period divided by 365.25 to obtain the actual period in years. No imputation of missing data has been performed with the exception of partial dates. All statistical tests were two-sided and were performed at a 0.05 significance level. Statistical analysis has been conducted using SAS® v9.3 (SAS Institute, Cary, NC, USA).
Geographic distribution of study sites and enrolled population.
Sample size. Under the assumption that, in the present study, the proportion of EGFR mutation positive (M+) NSCLC subjects would be 12%, the assessment of 450 subjects was required in order to detect this rate at the significance level of 0.05, with 80% power and an approximate±0.03 points (95%CI=9-15%) precision, using a two-tailed test (Relative Standard Error: 12.77%). Therefore, accounting for a 25% non-evaluable rate, approximately 600 patients were finally proposed to be included in the study.
Results
Patient characteristics. Between 13 October 2010 and 19 December 2013 a total of 589 Caucasian patients were enrolled in the study by 22 study sites located throughout Greece (Figure 1). The overall study duration was approximately 4 years, with the last patient last visit occurring on 18 December 2014. Patient disposition in the FAS population (N=575) and the PP population (N=564) is illustrated in Figure 2.
The median age of the overall population (FAS) at enrollment was 65.6 years (range=35.9-87.0 years); 73.4% (422/575) were males, and 82.6% (475/575) were either current or former smokers (Table I). Patients with adenocarcinoma comprised 81.4% (468/575) of the overall population and those with ECOG PS 0 or 1, 88.3% (508/575). Among patients with known data, the primary tumor was mainly localized in the right upper lobe (41.8%; 233/557), left upper lobe (27.5%; 153/557) and the right inferior lobe (24.2%; 135/557). At initial diagnosis, the vast majority of the overall population (93.6%; 538/575) presented with late stage (IIIB/IV) disease at initial diagnosis (37 patients presented with early stage disease at initial diagnosis). The most common sites of metastases were the bones (31.1%; 179/575) and the brain (17.7%; 102/575). Sociodemographic, anthropometric and clinical characteristics of the subpopulations per EGFR mutation status are reported in Table I.
Patient disposition per EGFR mutation status and study analysis sets. FAS: Full analysis set; PP population: per protocol population (eligible patients).
Sociodemographic, anthropometric and clinical characteristics of the overall population and subpopulations per EGFR mutation status.
The EGFR mutation status was positive in 15.7% (90/575) (95%CI=12.7-18.6), negative in 80.3% (462/575) (95%CI=77.1-83.6), and not evaluable (EGFR Mx) in the remaining 4.0% (23/575) (95%CI=2.4-5.6) of the patients. Simple logistic regression analysis did not demonstrate a statistically significant association between adenocarcinoma tumor histology and EGFR mutation status; on the other hand, smokers (current or former smokers) were less likely than never smokers (odds ratio (OR)=0.10; 95%CI=0.06-0.17; p<0.001), and males were less likely than females (OR=0.20; 95%CI=0.12-0.32; p<0.001) to harbor EGFR mutations. Multiple logistic regression model including gender, smoking status and adenocarcinoma histology, confirmed the above findings with males and smokers shown to be less likely to be EGFR M+ than EGFR M− (ORadjusted=0.68; 95%CI=0.51-0.90; p=0.008; and ORadjusted=0.40; 95%CI=0.30-0.53; p<0.001, respectively), and adenocarcinoma histology was not shown to be a predictive factor of EGFR mutation positivity (ORadjusted=0.95; 95% CI=0.67-1.33; p=0.751).
Frequency of EGFR mutations in exons 18-21: EGFR M+ subpopulation.
EGFR mutation status screening and detection. The tissue sample for EGFR mutation testing had originated from the primary tumor for 85.2% (490/575) of the patients, and from a metastatic lesion in the remaining 14.8% (85/575). In the overall population, direct sequencing had been employed as the EGFR detection method in 69.0% (397/575), followed by high-resolution melt analysis (HRMA) in 13.0% (75/575), targeted methods (such as an amplification refractory mutation system (ARMS); cobas® and TheraScreen®) in 12.2% (75/575), polymerase chain reaction (PCR) in 10.6% (61/575) and pyrosequencing in 3.1% (18/575),while the method was unknown in 9.9% (57/575). Notably, for 61.7% (355/575) direct sequencing had been used without any targeted method, while in 7.3% (42/575) both direct sequencing and a targeted method had been employed. In the EGFR M+ subpopulation, direct sequencing had been used in 64.4% (58/90), HRMA in 14.4% (13/90); targeted methods in 13.3% (12/90); PCR in 11.1% (10/90); pyrosequencing in 5.6% (5/90); the method was unknown in 14.4% (13/90). The most prevalent EGFR mutation site was exon 19 (59.6%; 53/89), followed by exon 21 (29.2%; 26/89) (Table II).
First-line treatment patterns. In the eligible patient population (PP; N=564), first line treatment had been initiated at a median of 0.8 (interquartile range (IQR)=0.4-1.4) months following histological confirmation of disease diagnosis for the EGFR M+, a median of 1.0 (IQR=0.5-1.8) month for the EGFR M-subpopulation, and 1.7 (IQR=0.9-3.4) months for the EGFR Mx. The most common first-line treatment pattern in the patient subpopulations per EGFR mutation status were: EGFR-TKI monotherapy (67.0%; 59/88) for the EGFR M+ population; and multi-agent chemotherapy in both the EGFR M− (61.7%; 280/454) and EGFR Mx subpopulations (86.4%; 19/22) (Figure 3A). The three most commonly prescribed agents were gefitinib (47.7%; 42/88), carboplatin (28.4%; 25/88), and erlotinib (19.3%; 17/88) in the EGFR M+ subpopulation; carboplatin (64.5%; 293/454), pemetrexed (39.0%; 177/454), and bevacizumab (27.8%; 126/454) in the EGFR M− subpopulation. The first line treatment patterns in the overall population per ECOG performance status are displayed in Figure 3B. Of the EGFR M+ patients, 68.6% (24/35) of those with PS0, 64.6% (31/48) with PS1, 75.0% (3/4) with PS2, and the single patient with PS3 received EGFR-TKI containing therapy.
Clinical response to therapy in the EGFR M+ subpopulation. Over a median 8.8 months (range=0.5-42.2 months) of exposure to first line treatment, the Kaplan-Meier estimated that the median PFS time in the eligible EGFR M+ population (n=88) was 9.67 (95%CI=7.90-11.77) months (Figure 4). A Cox multivariable proportional hazards model (n=86) was used to examine the association of factors of interest withthe PFS. Presence versus absence of bone and/or brain metastasis at initial diagnosis was shown to confer a higher risk of disease progression (HR=1.93; 95%CI=1.16-3.22; p=0.011), while presence versus absence of exon 19 EGFR mutation a lower risk of disease progression (HR=0.56; 95%CI=0.32-0.97; p=0.040). On the other hand, age (>65 versus ≤65 years) at initial NSCLC diagnosis (HR=0.71; 95%CI=0.42-1.20; p=0.200); smoking status (never smoker versus smoker) (HR=1.05; 95%CI=0.62-1.76; p=0.866); and non-adenocarcinoma histology at initial diagnosis (versus adenocarcinoma histological type) (HR=0.63; 95%CI=0.30-1.31; p=0.215) were not shown to be associated with PFS. During the study observation period (median of 6.9 months; range=0.03-43.7 months), a total of 12 deaths (13.6%) were reported; the Kaplan-Meier median OS time was not estimable due to data immaturity. The DCR among eligible EGFR M+ patients with available clinical response data (n=80) was 67.5% (95%CI=57.2-77.8).
First-line treatment patterns at enrollment by treatment categories (A) in the overall eligible patient population and subpopulations by EGFR mutation status and (B) by ECOG performance status at enrollment.
Healthcare resource utilization in the EGFR M+ subpopulation. Over a cumulative post-enrollment study observation period of 73.4 years, a total of 102 hospitalizations (median: 2.0; IQR=1.0-3.5) were reported for 32/82 (39.0%) of the eligible EGFR M+ patients with available data, yielding a hospitalization incidence rate of 1.39 per person-year. Similarly, a total of 244 outpatient visits (median: 4.0; IQR=1.0-5.0) were reported for 50/82 (61.0%) of the EGFR M+ patients, yielding an outpatient visit incidence rate of 3.32 per person-year.
Kaplan–Meier progression-free survival plot in the eligible EGFR M+ population.
Discussion
The Greek REASON represents the widest epidemiological dataset of clinicopathological characteristics, treatment patterns and outcomes in Caucasian patients with advanced NSCLC treated in the routine care of Greece. In our study, 15.7% of the patients had an EGFR M+ status matching the 15.8% rate reported in another Greek study of approximately 1,500 NSCLC patients (21). On the other hand, the respective frequency was 10.3% in the German REASON (18), 11.6% in the Spanish REASON (19), while other studies conducted in Caucasian populations have reported frequencies ranging from 13.6% to 16.6% (15-17, 20). Variance in reported EGFR mutation prevalence rates may undoubtedly lie in inter-ethnic variations as outlined in the study's rationale, but also in inter-assay variations, underscoring the importance of precise referencing the EGFR mutational testing methods utilized (21, 22).
In uniformity to the German (18) and Spanish (19) REASON studies, the study population of the Greek REASON was mainly comprised of males and smokers (current or former), diagnosed with stage IV adenocarcinoma and with an ECOG PS of 0 or 1. Males represented 62% of the enrolled population in the German REASON and roughly 73% in the Spanish and Greek REASON; smokers comprised about 82% of the population in all three studies, while adenocarcinoma histology was reported in 58%, 69% and 81% of the Spanish, German and Greek REASON studies, respectively. In our study, an EGFR positive mutation status was identified in 9.2% of the male population, but in 33.3% of the females; in 8.6% of smokers but in 49.0% of never smokers; and in 16.4% of patients with adenocarcinoma versus 12.2% of those with non-adenocarcinoma. These trends are not dissimilar from those reported elsewhere (17-21). In particular, the reported mutation frequencies in the Spanish REASON,German REASON and the recent Greek observational study ranged from 6.4-11.7% in males and from 16.7-25.4% among females; from 6.4-11.5% in smokers and 25.6-38.1% in never smokers; and from 13.1-16.6% in adenocarcinomas to 3.8-11.5% in non-adenocarcinomas (18, 19, 21).
According to a multiple logistic regression model controlling for gender, smoking status, and histological type, males and smokers had decreased odds of harboring an EGFR mutation than females and never smokers, respectively, in alignment with the relative frequencies reported above. However, on the other hand, according to the model, adenocarcinoma histologic type was not found to be associated with an EGFR M+ status, despite the higher EGFR mutation frequency noted among patients with adenocarcinomas. The Spanish REASON study reported the same finding; however, the German REASON, in addition to females and never smokers, also demonstrated that adenocarcinoma histology increases the odds of harboring an EGFR mutation (18, 19).
More than 89% of the EGFR M- patients received first line multi-agent chemotherapy or combination chemotherapy with anti-VEGF antibody, in alignment with the contemporary ESMO guidelines recommending chemotherapy with platinum doublets, platinum-based chemotherapy with any third-generation cytotoxics, or platinum-based chemotherapy with bevacizumab as the first-line treatment options for this NSCLC population. On the other hand, approximately 33% of the patients with an EGFR M+ status were managed with first line treatment patterns which did not contain EGFR-TKIs, the guideline-recommended first line treatment option for this patient population (6, 7). Recording of the factors guiding the treatment decision-making was beyond the scope of the study, thus not allowing the reasoning for this divergence between the guideline recommendations and clinical practice to be deciphered. Nevertheless, it becomes apparent that there are still opportunities to enhance adoption of evidence-based personalized strategies in the routine care of Greece aiming at further improving the clinical outcomes in this difficult-to-treat population.
Median PFS in the EGFR M+ population was estimated to be 9.67 (95%CI=7.90-11.77) months with approximately 67% of this population comprised of patients receiving first line EGFR-TKI containing therapy. A similar PFS (10.8 (95%CI=4.8-15.3) months)was reported in the Galician Lung Cancer Group observational study, in which 88% of the patients had received EGFR-TKI containing therapy (19), but also in the open-label phase IV IFUM study (median PFS 9.7 (95%CI=8.5-11.0)) months)of 118 Caucasian EGFR mutation positive stage III/IV patients (15). A clear benefit of EGFR-TKIs versus chemotherapy in the first line setting of EGFR M+ patients has been demonstrated in many randomized controlled trials (RCTs), with PFS ranging from 8 to 13.1 months with EGFR-TKIs versus 4.6-6.7 months with chemotherapy (23). The advantage conferred by EGFR-TKIs over chemotherapy, including not only on PFS, but also on OS and DCR, have been demonstrated in several meta-analyses of RCT data (24-27), leading to their establishment as the optimal first line treatment option for patient with EGFR mutation positive advanced NSCLC (6-8).
Regarding the identification of factors of poor prognosis in advanced EGFR M+ NSCLC, in our study, patients with bone and/or brain metastases were shown to have an approximately double risk of disease progression, while those with an exon 19 EGFR mutation were identified to have a lower risk of disease progression. Worse outcomes among EGFR M+ patients with brain metastases as well as in those with mutations other than exon 19 deletions have been previously reported (28, 29).
The inherent strength of the study's design, aiming to capture data under real-life clinical practice, and thus not mandating the employment of specific EGFR mutation screening methods or specific response criteria (e.g. RECIST), has generated limitations in the respective outcomes stemming from inter-assay and/or inter-observer variations. Furthermore, as certain study outcomes involve a relative limited number of available observations, caution should be exercised when interpretingthe statistical significance of these study outcomes. Lastly, as it pertains to the EGFR M+ subpopulation, the lack of post-withdrawal survival data collection and the short on-study observation period (median 6.9 months) have likely contributed to the inability to generate overall survival estimates.
Nonetheless, this study yielded real-world data on a patient population whose epidemiological data, long-term outcomes and treatment practice paradigms employed in Greece are understudied. It is anticipated that study outcomes may help optimize diagnostic algorithms and augment personalized management with targeted treatment options in the routine care of advanced NSCLC.
Acknowledgements
The Authors gratefully acknowledge for their support during the recruitment phase the following PIs and their teams (authors excluded): Dr Razi, Hygeia Hospital, Athens; Dr Giassas, Iaso General Hospital, Athens; Dr Athanasiadis, Mitera Hospital, Athens; Dr Touroutoglou, Diavalkaniko Hospital, Thessaloniki; Dr Ginopoulos, Ag. Andreas Hospital, Patras; Dr Linardou, Metropolitan Hospital, Piraeus; Dr Vaslamatzis,General Hospital of Athens Evaggelismos, Athens; Dr Gaga,Sotiria General Hospital of Chest Diseases, Athens; Dr Georgatou, Sotiria General Hospital of Chest Diseases, Athens; DrRapti, Sotiria General Hospital of Chest Diseases, Athens; Dr Ardavanis, Ag. Savvas Hospital, Athens; Dr Athanasiadis, General Hospital of Larissa, Larissa; Dr Koumakis, Ag. Savvas Hospital, Athens; Dr Kassarakis, General Hospital of Kavala, Kavala; Prof Pavlidis, University General Hospital of Ioannina, Ioannina; Dr Anthopoulos, General Hospital of Ptolemaida, Ptolemaida; and Dr Boukovinas,Bioclinic, Thessaloniki. This study was sponsored by AstraZeneca, Greece.
Footnotes
Conflicts of Interest
Prof. K.S declared no conflict of interst, Prof. V.G. has received speaker honorarium from MSD, Astra Zeneca and Novartis, Prof. K.Z. declared no conflict of interest, Dr P.M.has received a speaker honorarium from BMS, Novartis, Roche, Dr A.C. has received Honoraria from Brisol Mayer Squib, Pfizer, Roche Genetech and ASTRA Zeneca and Consultation fees from Boehringer Ingelheim, Dr C.C. have participated in advisory board and / or have received speaker honorarium from: Novartis, Roche, BMS, MSD, Amgen, Astra Zeneca, Pfizer, Merck, Genesis.
- Received April 8, 2018.
- Revision received May 4, 2018.
- Accepted May 7, 2018.
- Copyright© 2018, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.