- Email: [email protected]
The Impact of EGFR Mutation Status on Outcomes in Patients With Resected Stage I Non-Small Cell Lung Cancers
The Impact of EGFR Mutation Status on Outcomes in Patients With Resected Stage I Non-Small Cell Lung Cancers Benjamin Izar, MD, PhD, Lecia Sequist, MD, MPH, Mihan Lee, BS, Alona Muzikansky, PhD, Rebecca Heist, MD, MPH, John Iafrate, MD, PhD, Dora Dias-Santagata, PhD, Douglas Mathisen, MD, and Michael Lanuti, MD GENERAL THORACIC
Department of Medicine, Division of Hematology/Oncology, Division of Thoracic Surgery, Biostatistics Center, and Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
Background. Mutations of the epidermal growth factor hormone receptor (EGFR) gene have been associated with improved treatment response and prognosis in advanced non-small lung cancer (NSCLC). However, their prognostic role in early-stage NSCLC is not well defined. In this study we sought to identify the pure prognostic role of EGFR mutation in patients with completely resected stage I NSCLC who received no adjuvant therapy. Methods. Mutation status was tested in treatmentnaïve patients who had complete resection of stage I (T1–2aN0) NSCLC (from 2004 to 2011) using direct sequencing or multiplex polymerase chain reaction–based assay. Recurrence rates, disease-free survival, and overall survival were compared between EGFRmutant and wild-type patients using Kaplan-Meier methods and Cox regression models. Results. Three hundred seven patients were included in this study; 62 harbored tumors with EGFR mutations and 245 had wild-type EGFR. Tumors in patients with EGFR mutations were associated with a significantly
lower recurrence rate (9.7% versus 21.6%; p [ 0.03), greater median disease-free survival (8.8 versus 7.0 years; p [ 0.0085), and improved overall 5-year survival (98% versus 73%; p [ 0.003) compared with wild-type tumors. Lobectomy was the most frequently performed procedure, accounting for 209 of 307 operations. Among these patients, EGFR mutation was associated with superior overall survival (hazard ratio, 0.45; 95% confidence interval, 0.13 to 0.83; p [ 0.017), with an estimated 5-year survival of 98% versus 70%. The presence of EGFR mutation (p [ 0.026) and tumor size less than 2 cm (p [ 0.04) were identified as independent prognostic markers for disease-free survival, whereas age, sex, and smoking status were not. Conclusions. Completely resected stage I EGFR mutation-positive NSCLC patients have a significant survival advantage compared with EGFR wild-type patients. Mutation of the EGFR gene is a positive prognostic marker in completely resected stage I NSCLC.
C
genotype-specific adjuvant therapy with the hope of making a more dramatic impact on patient outcomes. Advanced-stage NSCLC harboring mutations in the epidermal growth factor receptor (EGFR) gene shows significant and durable responses to treatment with the EGFR tyrosine-kinase inhibitors (TKIs) erlotinib or gefitinib [6–8]. In fact, EGFR mutations may also impart an improved overall prognosis for advanced NSCLC patients compared with EGFR wild-type tumors [9]. The predictive and prognostic role of EGFR mutation status in stage I through III NSCLC remains less well defined, with conflicting results in studies reported to date [10–17]. In many of these studies, the administration of neoadjuvant or adjuvant chemotherapy or TKIs clouds the interpretation of the impact of EGFR status on prognosis [11–16]. We undertook this study to examine the
omplete surgical resection remains the best curative treatment for patients with early-stage non-small cell lung cancer (NSCLC). However, 30% to 75% of these patients experience recurrence and die within 5 years despite successful surgery, with the risk of relapse increasing exponentially with increased stage [1, 2]. Patients with resected stage II and III NSCLC have modestly improved survival if they also receive cisplatin-based adjuvant chemotherapy [3]. However, recent advances in the understanding of NSCLC biology have highlighted the importance and potential therapeutic success when matching systemic targeted therapies to cancer genotype [4, 5]; hence, emerging clinical trials are examining
(Ann Thorac Surg 2013;96:962–8) Ó 2013 by The Society of Thoracic Surgeons
Accepted for publication May 24, 2013. Presented at the Poster Session of the Forty-ninth Annual Meeting of The Society of Thoracic Surgeons, Los Angeles, CA, Jan 26–30, 2013. Address correspondence to Dr Lanuti, 55 Fruit St, Blake 1570, Boston, MA 02114; e-mail: [email protected].
Ó 2013 by The Society of Thoracic Surgeons Published by Elsevier Inc
Dr Iafrate discloses a financial relationship with BioReference Laboratories.
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.05.091
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
963
pure prognostic value of EGFR mutation status in patients with surgically resected stage I NSCLC.
validated in clinical specimens and has been reported to approach 100% [20]. All genotype results are available in the electronic medical record.
Patients and Methods
Covariates and Data Collection
This is a retrospective study examining outcomes of surgically resected patients by EGFR genotype status. Under an institutional review board–approved protocol, we reviewed 1,865 consecutive patients who underwent surgical resection for NSCLC at Massachusetts General Hospital between August 2004 and June 2011. We included patients in our study who had pathologic stage I NSCLC (AJCC version 7), underwent surgical resection with curative intent, and were tested for EGFR mutation status contemporary with routine pathologic review of the resected specimen. Epidermal growth factor hormone receptor gene testing was initially limited to resected adenocarcinoma and patients with less than 30 pack-years of tobacco use, and subsequently broadened to include all patients with resected lung cancer since early 2010. We excluded all patients with stage II, III, and IV NSCLC, those who were not referred for EGFR mutation testing or had nondiagnostic genotyping results, those who received adjuvant or neoadjuvant systemic therapy of any sort, those with squamous cell lung carcinoma, and those who had no follow-up recorded in our electronic medical record. Given the absence of EGFR mutations in squamous cell lung cancers [18], we initially excluded these cancers from the analysis. Patients with at least one surveillance imaging study at our institution or who had imaging from other hospitals uploaded and reviewed by our radiologists were included in the study. Charts were reviewed for demographics, medical history, tumor pathology and genotype, and follow-up information.
For all patients included in the study, baseline patient characteristics were collected, including age, sex, selfidentified race/ethnicity (white, black, Asian, Hispanic, or non-Hispanic), ECOG/Zubrod Performance Status, smoking history, pulmonary function testing (forced expiratory volume in 1 second and diffusing capacity of the lung for carbon monoxide), and medical comorbidities, including hypertension, coronary artery disease, congestive heart failure, diabetes mellitus, and chronic obstructive pulmonary disease. Smoking status was characterized as: (1) never smokers, fewer than 100 cigarettes in their lifetime, (2) former smokers, quit more than 1 year before diagnosis, and (3) current smokers, ongoing smoking habit or quit less than 1 year before diagnosis. Tumor stage was categorized according to the American Joint Committee on Cancer Staging Manual version 7; T and N staging was based primarily on the results from mediastinoscopy, lymph node dissection, and surgical resection, with additional clinical information abstracted from fluorodeoxyglucose positron emission tomography, computed tomography, and magnetic resonance imaging. Date and type of surgery (wedge resection, segmentectomy, lobectomy, or pneumonectomy) were recorded. The EGFR status and type of EGFR mutation, if present, as well as genotyping modality (sequencing versus SNaPshot), were collected. Patients were assessed for recurrence, progression-free survival, and overall survival (OS). Recurrence was defined as radiographic evidence of cancer relapse on surveillance imaging (computed tomography, computed tomographypositron emission tomography, or magnetic resonance imaging) and/or pathologic tumor evidence on a biopsy. Recurrence rate was the fraction of recurrent patients among the total. Disease-free survival (DFS) was defined as the time from surgery until recurrence or death. Overall survival was defined as time from surgery until death. Patients were censored at their last known alive date. Date of death was obtained from the electronic medical record and confirmed on the Social Security Death Index website. For those patients who had a recurrence, subsequent treatments including surgical resection, radiation therapy, chemotherapy, and use of EGFR TKIs were recorded.
Epidermal Growth Factor Hormone Receptor Gene Mutation Analysis Epidermal growth factor hormone receptor gene mutation testing in a Clinical Laboratory Improvement Amendments (CLIA) -certified institutional laboratory has been clinically available on provider request at our institution since August 2004. Initially this was accomplished through direct sequencing of EGFR exons 18 through 21 [19]. In 2009, clinical testing switched to the allele-specific, polymerase chain reaction–based SNaPshot platform (Applied Biosystems, Foster City, CA), which detects common exon 19 deletions and missense mutations in exon 21 (L858 and L861), exon 20 (T790), and exon 18 (G719) [20]. A sizing assay is used to confirm exon 19 deletions and to detect exon 20 insertions and deletions. The SNaPshot testing used for the detection of EGFR mutations is highly sensitive and specific. The sensitivity of the test was defined as the lowest percentage of mutation in a tested sample yielding a mutant allele peak that was more than three times the background in wild-type samples. The average sensitivity for mutations tested with SNaPshot is approximately 4.64%, which exceeds the sensitivity of conventional sequencing. For EGFR mutations it ranges between 1.4% and 8.5%. The specificity of the test was
Statistical Analysis Differences between covariates and recurrence rate in patients with EGFR-mutant and wild-type tumors were analyzed using Fisher’s exact test, Freeman-Halton extension of Fisher’s exact test, and Student’s t test. The DFS and OS were analyzed using the Kaplan-Meier method, and a log-rank test was used to compare the EGFR-mutant and wild-type group. Cox multivariate analysis was performed with five variables (age, sex, EGFR mutation status, smoking status, and tumor size) to test for independent markers of progression-free survival. Analyses were performed using Prism software (v5.0;
GENERAL THORACIC
Ann Thorac Surg 2013;96:962–8
964
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
Ann Thorac Surg 2013;96:962–8
GENERAL THORACIC
GraphPad Software, San Diego, CA) and SAS software (v9.2; SAS Institute Inc, Cary, NC).
Table 1. Demographics of Patients Including Comorbidities, Tumor Characteristics, and Surgical Intervention
Results
Variable
Patients
Age (y), mean SD 66.6 10.65 Sex, n (%) Male 11 (17.7) Female 51 (82.3) Tobacco history, n (%) Never smoked 35 (56.5) Past smoker 24 (38.7) Current smoker 3 (4.8) Pack-years, mean SD 9.6 17.6 Race White 56 (90.3) Black 0 Asian 6 (9.6) Hispanic 0 Zubrod score 0.37 0.52 Tumor size in cm, n (%) <2 38 (61.3) 2–3 12 (19.3) >3 12 (19.3) Surgical procedure, n (%) Lobectomy 48 (77.4) Wedge resection 12 (19.4) Segmentectomy 2 (3.2) Pneumonectomy 0 EGFR testing, n (%) Direct sequencing 15 (24.2) SNaPshot 47 (75.8) Comorbid conditions Hypertension 26 (41.9) Coronary artery disease 2 (3.2) CHF 1 (1.6) Diabetes mellitus 3 (4.8) COPD 4 (6.4) PFTs and functional status, mean SD 99.8 20.1 FEV1 (% predicted) 89.2 18.8 DLCO (% predicted)
We examined 1,865 consecutive patients who underwent surgical resection for NSCLC between 2004 and 2011 and identified 1,351 with stage I disease. Of these, 307 patients met the inclusion criteria for this study, and follow-up information was collected through October 31, 2012. Sixty-two (20.2%) patients in the study cohort were positive for an EGFR mutation (EGFR-mutant group), and 245 (79.8%) had no mutation in EGFR (wild-type group). Demographic features of the study population are summarized in Table 1. Both groups had a median age of 66 years. As expected, there were more women in the EGFR-mutant group (82% versus 62%; p ¼ 0.004) and more never-smokers (57% versus 10%; p ¼ 0.0001). Although the majority of patients in the study were white, the EGFR-mutant group had an increased proportion of Asians (10% versus 1%; p ¼ 0.01). Tumor size was comparable between genotype groups (p ¼ 0.14). The majority of patients underwent lobectomy as their surgical procedure, and there was no significant difference in the distribution of type of surgery by genotype (p ¼ 0.34). The frequency of some comorbidities, including hypertension, congestive heart failure, and diabetes mellitus, were comparable among genotype groups, whereas the number of patients with chronic obstructive pulmonary disease and coronary artery disease was higher in the wild-type group, likely attributable to the heavier incidence of smoking. There was no difference in the functional status of patients in both groups as measured by the Zubrod score (p ¼ 0.46).
Genotype Results Genotyping was performed using the SNaPshot assay in 238 of 307 patients and by direct sequencing in 69 cases. Sixty-two (20%) patients had an EGFR mutation, of which 47 were detected by SNaPshot analysis and 15 by direct sequencing (Table 2). Among the EGFR-mutant group, 60 had a single mutation and 2 patients had double mutations in EGFR. The most frequently observed EGFR mutations were the exon 21 point mutation L858R, found in 30 (48%) patients, and deletions in the LREA region of exon 19 (del 19), found in 24 (39%) patients. One patient with L858R and 1 patient with del 19 also had the exon 20 mutation T790M, which has been implicated in primary and secondary resistance to EGFR TKIs. The remaining 8 patients had other mutations, including 4 (6%) with exon 20 insertion mutations, 2 (3%) with L861Q, and 2 (3%) with G719C.
Recurrence Rate, Progression-Free Survival, and Overall Survival Median follow-up time was 30 months. There were 85 progression-free survival events observed, including 59 cases of radiographic disease recurrence and 26 deaths.
EGFR Mutant (n ¼ 62)
Wild-Type (n ¼ 245) 66.6 10.1
p Value 0.785 0.004
94 (38.4) 151 (61.6) <0.0001 25 (10.2) 166 (67.8) 54 (22) 43.8 29.4 240 (98) 1 (0.4) 3 (1.2) 1 (0.4) 0.43 0.56
0.0001 0.01 1.0 0.0028 0.46 0.14
116 (47.3) 60 (24.5) 69 (28.2) 0.34 161 68 15 1
(65.7) (27.8) (6.1) (0.4) 0.735
54 (22) 191 (78) 114 44 2 32 56
(47) (18) (0.8) (13) (22.8)
0.192 0.0023 0.493 0.076 0.023
82 22.9 70 22.8
0.0001 0.0001
CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; DLCO ¼ diffusing capacity of the lung for carbon monoxideof % predicted; EGFR ¼ epidermal growth factor receptor gene; FEV1 ¼ forced expiratory volume in 1 second of % predicted; PFTs ¼ pulmonary function tests; SD ¼ standard deviation.
The recurrence rate was 6 of 62 (9.7%) in the EGFRmutant group and 53 of 245 (21.6%) in the wild-type group (p ¼ 0.0315). In 44 of 59 patients (75%) with radiographic disease recurrence, biopsies were obtained and confirmed malignancy. The median DFS in the wild-type group was 7.0 years (95% confidence interval [CI], 4.93 to not reached) compared with 8.83 years (95% CI, 6.2 to not reached) in the EGFR-mutant group (logrank p ¼ 0.0085; Fig 1). This resulted in a hazard ratio (HR) for DFS of 0.45 (95% CI, 0.25 to 0.81). The percentage of patients who remained disease-free at
Ann Thorac Surg 2013;96:962–8
965
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
Table 2. Frequency of Different EGFR Mutations in Resected Stage I Tumors
Table 3. Multivariate Analysis for Predictors of Disease-Free Survival
EGFR Mutation
n
Variable
30 2 2 4 24 2
Age EGFR mutation Sex Smoking status Tumor size
L858 L861G lns G719C Ex20 ins Ex19*del T790Mb a
One patient in this group had a second T790M mutation. patients had a second mutation.
b
Both
EGFR ¼ epidermal growth factor receptor gene.
12 months and 5 years were 97% and 81% in the EGFR-mutant group compared with 94% and 60% in the wild-type group, respectively. In a Cox regression model for DFS using age, sex, smoking status, tumor size (<2 cm, 2 to 3 cm or > 3 cm), and EGFR status as independent variables, only EGFR mutation and tumor size were significant (Table 3). Patients with EGFR mutation were 3.1 times less likely to have disease progression compared with patients with wild-type disease (HR, 0.33; 95% CI, 0.12 to 0.87; p ¼ 0.026), and larger tumor size was associated with an increased risk of disease progression (HR, 1.37; 95% CI, 1.01 to 1.84; p ¼ 0.04). Patients with small malignancies (<2 cm) were 1.9 times less likely to have disease progression (HR, 0.53; 95% CI, 0.29 to 0.97; p ¼ 0.04), whereas age, sex, and smoking status were not independent factors associated with disease progression. Of the 53 patients with disease recurrence in the wildtype group, details about subsequent therapy were available for 52. In the wild-type group, 15 (29%) patients underwent additional surgery, 26 (50%) patients had radiation therapy, and 18 (35%) patients had systemic therapy after disease recurrence, including 17 patients who had intravenous chemotherapy and 3 patients treated with an EGFR TKI (Table 4). In the EGFR-mutant group, 4 patients had additional surgery, 1 patient had radiation therapy, and 3 patients underwent systemic chemotherapy, of which 1 patient received erlotinib.
HR (95% CI) 1.009 0.326 1.15 1.14 1.37
p Value
(0.982–1.037) (0.121–0.874) (0.67–1.97) (0.71–1.82) (1.01–1.84)
0.51 0.026 0.62 0.59 0.04
CI ¼ confidence interval; EGFR ¼ epidermal growth factor receptor gene; HR ¼ hazard ratio.
There were 36 deaths observed, 35 in the wild-type group and 1 in the EGFR-mutant group (p ¼ 0.023). The OS was significantly higher in the mutant group compared with the wild-type group (HR, 0.30; 95% CI, 0.14 to 0.67; p ¼ 0.003). The estimated 5-year survival rate was 98% in the mutant group and 73% in the wild-type group (Fig 2). Lobectomy was the most common surgical treatment implemented in this study, with 77% of EGFR-mutant group patients and 66% of wild-type group patients undergoing this procedure. Among patients treated with lobectomy, the OS was superior in the EGFR-mutant group (HR, 0.45; 95% CI, 0.13 to 0.83; p ¼ 0.017) compared with the EGFR wild-type group (Fig 3). The estimated 5-year survival rate was higher in the EGFR-mutant group compared with the wild-type group (98% versus 70%). There was no difference in OS by genotype among those undergoing sublobar resection (wedge or segmentectomy; data not shown). Furthermore, there were no significant differences in OS or DFS among all patients who had lobectomy compared with all patients who had sublobar resection (not segregated by EGFR status; data not shown).
Comment In this study, we investigated the prognostic implication of EGFR mutations in a cohort of 307 patients with completely resected stage I NSCLC and known EGFR mutation status. To focus solely on postoperative prognostic differences by genotype, the cohort was restricted to patients who had no systemic chemotherapy and no radiation therapy. We found that EGFR-mutant lung Table 4. Treatment Modality in Patients After Disease Recurrencea Therapy After Recurrence Surgical resection Radiation therapy Chemotherapy Intravenous EGFR TKI Supportive
Fig 1. Overall survival in resected stage I non-small cell lung cancer. (wt/WT ¼ wild-type.)
EGFR Mutant (n ¼ 6)
Wild-Type (n ¼ 52)
4 1 3 2 1 2
15 26 18 17 3 8
a Note that the sum of different treatments does not equal the number of recurrences as some patients received multiple treatment modalities.
EGFR TKI ¼ epidermal growth factor receptor tyrosine-kinase inhibitor.
GENERAL THORACIC
a
966
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
GENERAL THORACIC
Fig 2. Disease-free survival in resected stage I non-small cell lung cancer. (wt/WT ¼ wild-type.)
cancer patients had lower recurrence rates and significantly prolonged DFS and OS compared with those with EGFR wild-type cancers. Positivity for EGFR mutation and small tumors (<2 cm) were identified as independent predictors of DFS when adjusting for potential confounders, including age, sex, and smoking status. To our knowledge, this is the first study to demonstrate a DFS and OS benefit for resected stage I lung cancer patients who harbor EGFR mutations compared with those with wild-type disease. The first suggestion of a possible prognostic role for EGFR mutations emerged from an analysis of patients with stage IIIB and IV NSCLC who participated in the phase III TRIBUTE trial [21], in which patients were randomly assigned to receive firstline carboplatin and paclitaxel with either erlotinib or placebo. Here, EGFR-mutant disease was associated with longer OS and improved response rate regardless of treatment arm [9]. More recent studies have confirmed that advanced NSCLC patients with EGFR mutations seem to fare better on chemotherapy than patients with wild-type EGFR [5], but trials performed in patients with advanced disease cannot inform us of whether the mutation status is an inherent prognostic factor or a predictor of improved outcome on chemotherapy because all patients in these trials receive treatment of some sort for their metastatic cancer. Hence, there is great interest in looking at outcomes in an earlier stage
Fig 3. Overall survival in stage I non-small cell lung cancer after lobectomy. (wt/WT ¼ wild-type.)
Ann Thorac Surg 2013;96:962–8
of the disease to understand the true prognostic significance of EGFR mutations. Previous studies examining the prognostic value of EGFR mutation status in resected stage I through III NSCLC are limited, and their results are somewhat conflicting [10–17]. Several have suggested that EGFR mutations impart an improved prognosis after surgery. A Japanese study was fairly small, with only 53 patients, but had an impressive median follow-up time of more than 60 months and found that mutation of the EGFR gene was associated with a strong trend toward improved survival in stage I through III disease (71% versus 56%) and a significantly superior 5-year survival rate among the 36 patients with stage I disease (92% versus 57%; p ¼ 0.037) [10]. Another study of 294 patients with stage I through III disease found an improved 3-year survival rate (p ¼ 0.02) and OS (p ¼ 0.031) for the patients with EGFR-mutated tumors (n ¼ 40) compared with those with KRAS-mutant or wild-type disease; however, 40% of patients included in this cohort received cytotoxic chemotherapy in addition to surgery [14]. A second group evaluating the prognostic role of EGFR and KRAS mutations in 164 Taiwanese patients with stage I through III disease showed a significantly higher 3-year survival rate (p ¼ 0.02) in EGFR-mutated NSCLC compared with wild-type and KRAS-mutant tumors [16]. There was a trend toward longer OS in the EGFR group (55 months) compared with tumors that were KRAS and EGFR wildtype (35 months); however, this difference was not statistically significant (p ¼ 0.2). The analysis in this study is clouded by use of chemotherapy, radiation, and EGFR TKIs in 6%, 11%, and 9% of patients, respectively. Finally, a recently published large study (n ¼ 1,118) from Memorial Sloan Kettering investigated outcomes among patients of all stages, although more than 97% had stage I through III disease. In that study, EGFR-mutated disease was associated with a significantly improved OS compared with wild-type tumors (p < 0.001) and remained significant after adjusting for stage [12]. However, a significant proportion of patients in this cohort again received chemotherapy (21%), highlighting the chief difficulty in interpretation of this body of literature. Because EGFR-mutant tumors appear more sensitive to chemotherapy than wild-type tumors, it is challenging to understand prognosis when patients have been exposed to this potentially confounding treatment. A second group of studies have found that apparent prognostic advantages for EGFR-mutant cancers may be lost after adjusting for potential confounding factors [14, 15, 17]. In a study evaluating the prognostic importance of EGFR, KRAS, and TP53 mutations in 397 Japanese patients with stage I through IV NSCLC, EGFR-mutant disease (n ¼ 196) was associated with significantly improved OS (p ¼ 0.006), particularly in patients with the two major EGFR mutations, L858R and exon 19 deletions [17]. Of note, 56 patients were treated with gefitinib in this cohort and were not included in the survival analysis owing to concern over confounding factors. After adjusting for other potential confounders, including sex, stage (I versus II through IV), tumor differentiation, and smoking history, the survival benefit
Ann Thorac Surg 2013;96:962–8
967
provides further evidence to support routine EGFR mutation testing on resected stage I adenocarcinomas. Financial support for this study was provided by the Division of Thoracic Surgery at the Massachusetts General Hospital. We would like to acknowledge our data manager and research coordinator, Sheila Cann and Diane Davies, respectively, for their diligence and dedication toward compiling and maintaining the Thoracic Surgery Database.
References 1. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg 1995;60:615–23. 2. Goldstraw P, Crowley J, Chansky K, et al; International Association for the Study of Lung Cancer International Staging Committee; Participating Institutions. The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM Classification of malignant tumours. J Thorac Oncol 2007;2:706–14. 3. Song WA, Zhou NK, Wang W, et al. Survival benefit of neoadjuvant chemotherapy in non-small cell lung cancer: an updated meta-analysis of 13 randomized control trials. J Thorac Oncol 2010;5:510–6. 4. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010;363:1693–703. 5. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009;361:947–57. 6. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129–39. 7. Paez JG, J€ anne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497–500. 8. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain. PLoS Med 2005;2:e73. 9. Eberhard DA, Johnson BE, Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic indicators in patients with non-small-cell lung cancer treated with chemotherapy alone and in combination with erlotinib. J Clin Oncol 2005;23:5900–9. 10. Sonobe M, Nakagawa M, Takenaka K, et al. Influence of epidermal growth factor receptor (EGFR) gene mutations on the expression of EGFR, phosphoryl-Akt, and phosphorylMAPK, and on the prognosis of patients with non-small cell lung cancer. J Surg Oncol 2007;95:63–9. 11. Janjigian YY, Park BJ, Zakowski MF, et al. Impact on diseasefree survival of adjuvant erlotinib or gefitinib in patients with resected lung adenocarcinomas that harbor EGFR mutations. J Thorac Oncol 2011;6:569–75. 12. D’Angelo SP, Janjigian YY, Ahye N, et al. Distinct clinical course of EGFR-mutant resected lung cancers: results of testing of 1118 surgical specimens and effects of adjuvant gefitinib and erlotinib. J Thorac Oncol 2012;7:1815–22. 13. Lim KH, Huang MJ, Liu HC, Kuo HT, Tzen CY, Hsieh RK. Lack of prognostic value of EGFR mutations in primary resected non-small cell lung cancer. Med Oncol 2007;24:388–93. 14. Marks JL, Broderick S, Zhou Q, et al. Prognostic and therapeutic implications of EGFR and KRAS mutations in resected lung adenocarcinoma. J Thorac Oncol 2008;3:111–6. 15. Kim YT, Seong YW, Jung YJ, et al. The presence of mutations in epidermal growth factor receptor gene is not a prognostic factor for long-term outcome after surgical
GENERAL THORACIC
for EGFR-mutant disease lost statistical significance, although tumor stage remained an independent marker for OS. Median follow-up was relatively short in this study (32 months), and the use of chemotherapy and radiation was not reported, although it is conceivable that a significant portion of patients received systemic therapy given that 149 patients had stage II through IV disease, of which the majority had stage III and IV disease (n ¼ 105). Similar results were reported in a cohort of 863 Korean patients with stage I through IV NSCLC [15]. A total of 354 patients had an EGFR mutation, and the majority had early-stage disease. The EGFR-mutated tumors were associated with a significantly improved OS compared with the wild-type tumors (p ¼ 0.001). However, in a multivariate analysis adjusting for potential confounders, EGFR mutation status did not remain an independent prognostic marker. This study was also limited by a relatively short median follow-up period of 24 months. Furthermore, there was a proportional imbalance of patients with stage II through IV cancers between the genotype groups (36% versus 48%), suggesting a likely significant difference in the receipt of systemic treatment that could have blunted the interpretation of prognosis. These collective results stress the need for carefully studying a patient population best suited to minimize the primary confounding variable of prior research and systemic treatment exposure. Our study looked at a uniform stage of disease (stage I only) and excluded any patient who received chemotherapy, radiation therapy, or EGFR TKIs to focus solely on the biology of the cancer genotype and its impact on recurrence and survival. Although type of surgery can affect recurrence rates [1], three quarters of patients in our cohort received a lobectomy, and separate analysis of this population was also significant for improved survival among the EGFR-mutant group. Our study must be interpreted in the context of its limitations, such as its retrospective design and its singlecenter patient cohort. We analyzed outcomes based on the EGFR mutation status alone and did not investigate the influence of other potentially relevant genotype biomarkers such as KRAS. Furthermore, given the aim and design of this study, the patient population was smaller compared with some of the other studies discussed above, which included multiple stages from I to IV. In summary, our study examined a large cohort of stage I NSCLC patients in which no patient received adjuvant or neoadjuvant therapy of any sort. We identified EGFR mutation-positive status as an independent prognostic marker associated with decreased recurrence and improved progression-free survival and OS. Although past studies have shown little benefit for cisplatin-based adjuvant chemotherapy in stage I NSCLC, our data suggest it may be reasonable to design future prospective trials of adjuvant chemotherapy to include only patients with EGFR wild-type tumors because the natural history prognosis of stage I EGFR-mutant tumors is excellent. In addition, clinical trials examining the benefit of adjuvant EGFR TKIs will require that comparisons be made only within genotype-similar groups. This analysis also
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
968
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
resection of non-small-cell lung cancer. J Thorac Oncol 2013;8:171–8. 16. Liu HP, Isaac Wu HD, Chang JWC, et al. Prognostic implications of epidermal growth factor receptor and KRAS gene mutations and epidermal growth factor receptor gene copy numbers in patients with surgically resectable non-small cell lung cancer in Taiwan. J Thorac Oncol 2010;5:1175–84. 17. Kosaka T, Yatabe Y, Onozato R, Kuwano H, Mitsudomi T. Prognostic implication of EGFR, KRAS, and TP53 gene mutations in a large cohort of Japanese patients with surgically treated lung adenocarcinoma. J Thorac Oncol 2009;4:22–9. 18. Marchetti A, Martella C, Felicioni L, et al. EGFR mutations in non–small-cell lung cancer: analysis of a large series of cases and development of a rapid and sensitive method for
Ann Thorac Surg 2013;96:962–8
diagnostic screening with potential implications on pharmacologic treatment. J Clin Oncol 2005;23:857–65. 19. Sequist LV, Joshi VA, J€ anne PA, et al. Epidermal growth factor receptor mutation testing in the care of lung cancer patients. Clin Cancer Res 2006;12(14 Pt 2):4403s–8s. 20. Dias-Santagata D, Akhavanfard S, David SS, et al. Rapid targeted mutational analysis of human tumours: a clinical platform to guide personalized cancer medicine. EMBO Mol Med 2010;2:146–58. 21. Herbst RS, Prager D, Hermann R, et al. TRIBUTE Investigator Group. TRIBUTE: a phase III trial of erlotinib hydrochloride (OSI-774) combined with carboplatin and paclitaxel chemotherapy in advanced non-small-cell lung cancer. J Clin Oncol 2005;23:5892–9.
GENERAL THORACIC
INVITED COMMENTARY As lung cancer remains the number-one worldwide cause of cancer-related deaths, and non-small cell lung cancer (NSCLC) accounts for 85% of primary lung cancer cases, significant recent efforts have been put toward developing new therapeutic options to improve outcomes from this disease. Specifically, investigators have taken great interest in attempts to recognize specific driver mutations that foster neoplastic transformation and contribute to tumor progression. The dominant theme of such investigations has been aimed toward identifying specific molecular targets to develop new agents with the capacity to kill cancer cells with minimal toxicity to normal cells. Epidermal growth factor receptor (EGFR) mutations have been identified in 13% to 22% of patients with NSCLC, and the development of EGFR tyrosine kinase inhibitors as successful therapeutic agents for patients harboring this mutation has led many institutions to implement routine testing for this genetic aberrancy. A key purpose for such genetic testing is to determine susceptibility to available treatments; however, additional information may potentially be gained from such investigations, and, as exemplified by Izar and colleagues [1], may provide us with highly informative prognostic data. Such data allow us to better identify patients’ risk of recurrence—ultimately guiding us in establishing optimal treatment strategies, determining the most appropriate follow-up surveillance plans, and communicating effectively with patients regarding their futures. While previous authors have shown EGFR mutations to be associated with improved treatment response in advanced NSCLC, the effect of EGFR mutation status on stage I disease has remained largely unclear. In this insightful report, the authors aimed to identify the prognostic role of EGFR mutations among patients with completely resected, early-stage NSCLC, which has significant relevance to us as thoracic surgeons. Direct sequencing and multiplex polymerase chain reaction-based assays were used to determine mutation status for 307 treatment-naive patients with T1-2a N0 disease who underwent operative resection with curative intent. The authors subsequently measured recurrence rates, progression-free survival (PFS), and overall survival in patients with EGFR mutations (20% of the study population) compared with the wild-type individuals. Ó 2013 by The Society of Thoracic Surgeons Published by Elsevier Inc
Over a median follow-up of 30 months, the individuals with EGFR mutations were less likely to have tumor recurrence (9.7% vs 21.6%) and experienced longer progression-free survival (8.8 vs 7.0 years) than wild-type patients. Cox regression analysis was used to examine several tumor/patient variables, finding only tumor size and EGFR status were associated with risk of disease progression. EGFR mutations also conveyed improved overall survival at 5 years of 98% vs 73%. Findings were consistent regardless of the extent of the resection performed. By studying a large cohort of patients who received neither adjuvant nor neoadjuvant therapy, the authors have examined the true biologic effect of EGFR status on outcome, contributing greatly to our understanding of the prognostic implication of EGFR mutations in NSCLC. However, this provocative study leaves several questions unanswered and has set a foundation for future investigations. The molecular mechanisms leading to more indolent growth of EGFR-mutated tumors have not been defined, and further bench work is required. Further, with evidence that patients with early-stage NSCLC and EGFR mutations are less likely to experience recurrence, we are left with questions about how we ought to alter our treatment of these individuals. Should their surveillance plans be tailored accordingly? Should we expand our adjuvant chemotherapeutic strategies for EGFR wild-type patients? Like all excellent studies, this report has raised more questions than it answers, providing impetus for a number of future trials. Mara B. Antonoff, MD Department of Surgery Division of Cardiothoracic Surgery Washington University St. Louis, MO 63110 e-mail: [email protected]
Reference 1. Izar B, Sequist L, Lee M, et al. The impact of EGFR mutation status on outcomes in patients with resected Stage I non-small cell lung cancers. Ann Thorac Surg 2013;96:962–8.
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.06.059
Department of Medicine, Division of Hematology/Oncology, Division of Thoracic Surgery, Biostatistics Center, and Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
Background. Mutations of the epidermal growth factor hormone receptor (EGFR) gene have been associated with improved treatment response and prognosis in advanced non-small lung cancer (NSCLC). However, their prognostic role in early-stage NSCLC is not well defined. In this study we sought to identify the pure prognostic role of EGFR mutation in patients with completely resected stage I NSCLC who received no adjuvant therapy. Methods. Mutation status was tested in treatmentnaïve patients who had complete resection of stage I (T1–2aN0) NSCLC (from 2004 to 2011) using direct sequencing or multiplex polymerase chain reaction–based assay. Recurrence rates, disease-free survival, and overall survival were compared between EGFRmutant and wild-type patients using Kaplan-Meier methods and Cox regression models. Results. Three hundred seven patients were included in this study; 62 harbored tumors with EGFR mutations and 245 had wild-type EGFR. Tumors in patients with EGFR mutations were associated with a significantly
lower recurrence rate (9.7% versus 21.6%; p [ 0.03), greater median disease-free survival (8.8 versus 7.0 years; p [ 0.0085), and improved overall 5-year survival (98% versus 73%; p [ 0.003) compared with wild-type tumors. Lobectomy was the most frequently performed procedure, accounting for 209 of 307 operations. Among these patients, EGFR mutation was associated with superior overall survival (hazard ratio, 0.45; 95% confidence interval, 0.13 to 0.83; p [ 0.017), with an estimated 5-year survival of 98% versus 70%. The presence of EGFR mutation (p [ 0.026) and tumor size less than 2 cm (p [ 0.04) were identified as independent prognostic markers for disease-free survival, whereas age, sex, and smoking status were not. Conclusions. Completely resected stage I EGFR mutation-positive NSCLC patients have a significant survival advantage compared with EGFR wild-type patients. Mutation of the EGFR gene is a positive prognostic marker in completely resected stage I NSCLC.
C
genotype-specific adjuvant therapy with the hope of making a more dramatic impact on patient outcomes. Advanced-stage NSCLC harboring mutations in the epidermal growth factor receptor (EGFR) gene shows significant and durable responses to treatment with the EGFR tyrosine-kinase inhibitors (TKIs) erlotinib or gefitinib [6–8]. In fact, EGFR mutations may also impart an improved overall prognosis for advanced NSCLC patients compared with EGFR wild-type tumors [9]. The predictive and prognostic role of EGFR mutation status in stage I through III NSCLC remains less well defined, with conflicting results in studies reported to date [10–17]. In many of these studies, the administration of neoadjuvant or adjuvant chemotherapy or TKIs clouds the interpretation of the impact of EGFR status on prognosis [11–16]. We undertook this study to examine the
omplete surgical resection remains the best curative treatment for patients with early-stage non-small cell lung cancer (NSCLC). However, 30% to 75% of these patients experience recurrence and die within 5 years despite successful surgery, with the risk of relapse increasing exponentially with increased stage [1, 2]. Patients with resected stage II and III NSCLC have modestly improved survival if they also receive cisplatin-based adjuvant chemotherapy [3]. However, recent advances in the understanding of NSCLC biology have highlighted the importance and potential therapeutic success when matching systemic targeted therapies to cancer genotype [4, 5]; hence, emerging clinical trials are examining
(Ann Thorac Surg 2013;96:962–8) Ó 2013 by The Society of Thoracic Surgeons
Accepted for publication May 24, 2013. Presented at the Poster Session of the Forty-ninth Annual Meeting of The Society of Thoracic Surgeons, Los Angeles, CA, Jan 26–30, 2013. Address correspondence to Dr Lanuti, 55 Fruit St, Blake 1570, Boston, MA 02114; e-mail: [email protected].
Ó 2013 by The Society of Thoracic Surgeons Published by Elsevier Inc
Dr Iafrate discloses a financial relationship with BioReference Laboratories.
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.05.091
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
963
pure prognostic value of EGFR mutation status in patients with surgically resected stage I NSCLC.
validated in clinical specimens and has been reported to approach 100% [20]. All genotype results are available in the electronic medical record.
Patients and Methods
Covariates and Data Collection
This is a retrospective study examining outcomes of surgically resected patients by EGFR genotype status. Under an institutional review board–approved protocol, we reviewed 1,865 consecutive patients who underwent surgical resection for NSCLC at Massachusetts General Hospital between August 2004 and June 2011. We included patients in our study who had pathologic stage I NSCLC (AJCC version 7), underwent surgical resection with curative intent, and were tested for EGFR mutation status contemporary with routine pathologic review of the resected specimen. Epidermal growth factor hormone receptor gene testing was initially limited to resected adenocarcinoma and patients with less than 30 pack-years of tobacco use, and subsequently broadened to include all patients with resected lung cancer since early 2010. We excluded all patients with stage II, III, and IV NSCLC, those who were not referred for EGFR mutation testing or had nondiagnostic genotyping results, those who received adjuvant or neoadjuvant systemic therapy of any sort, those with squamous cell lung carcinoma, and those who had no follow-up recorded in our electronic medical record. Given the absence of EGFR mutations in squamous cell lung cancers [18], we initially excluded these cancers from the analysis. Patients with at least one surveillance imaging study at our institution or who had imaging from other hospitals uploaded and reviewed by our radiologists were included in the study. Charts were reviewed for demographics, medical history, tumor pathology and genotype, and follow-up information.
For all patients included in the study, baseline patient characteristics were collected, including age, sex, selfidentified race/ethnicity (white, black, Asian, Hispanic, or non-Hispanic), ECOG/Zubrod Performance Status, smoking history, pulmonary function testing (forced expiratory volume in 1 second and diffusing capacity of the lung for carbon monoxide), and medical comorbidities, including hypertension, coronary artery disease, congestive heart failure, diabetes mellitus, and chronic obstructive pulmonary disease. Smoking status was characterized as: (1) never smokers, fewer than 100 cigarettes in their lifetime, (2) former smokers, quit more than 1 year before diagnosis, and (3) current smokers, ongoing smoking habit or quit less than 1 year before diagnosis. Tumor stage was categorized according to the American Joint Committee on Cancer Staging Manual version 7; T and N staging was based primarily on the results from mediastinoscopy, lymph node dissection, and surgical resection, with additional clinical information abstracted from fluorodeoxyglucose positron emission tomography, computed tomography, and magnetic resonance imaging. Date and type of surgery (wedge resection, segmentectomy, lobectomy, or pneumonectomy) were recorded. The EGFR status and type of EGFR mutation, if present, as well as genotyping modality (sequencing versus SNaPshot), were collected. Patients were assessed for recurrence, progression-free survival, and overall survival (OS). Recurrence was defined as radiographic evidence of cancer relapse on surveillance imaging (computed tomography, computed tomographypositron emission tomography, or magnetic resonance imaging) and/or pathologic tumor evidence on a biopsy. Recurrence rate was the fraction of recurrent patients among the total. Disease-free survival (DFS) was defined as the time from surgery until recurrence or death. Overall survival was defined as time from surgery until death. Patients were censored at their last known alive date. Date of death was obtained from the electronic medical record and confirmed on the Social Security Death Index website. For those patients who had a recurrence, subsequent treatments including surgical resection, radiation therapy, chemotherapy, and use of EGFR TKIs were recorded.
Epidermal Growth Factor Hormone Receptor Gene Mutation Analysis Epidermal growth factor hormone receptor gene mutation testing in a Clinical Laboratory Improvement Amendments (CLIA) -certified institutional laboratory has been clinically available on provider request at our institution since August 2004. Initially this was accomplished through direct sequencing of EGFR exons 18 through 21 [19]. In 2009, clinical testing switched to the allele-specific, polymerase chain reaction–based SNaPshot platform (Applied Biosystems, Foster City, CA), which detects common exon 19 deletions and missense mutations in exon 21 (L858 and L861), exon 20 (T790), and exon 18 (G719) [20]. A sizing assay is used to confirm exon 19 deletions and to detect exon 20 insertions and deletions. The SNaPshot testing used for the detection of EGFR mutations is highly sensitive and specific. The sensitivity of the test was defined as the lowest percentage of mutation in a tested sample yielding a mutant allele peak that was more than three times the background in wild-type samples. The average sensitivity for mutations tested with SNaPshot is approximately 4.64%, which exceeds the sensitivity of conventional sequencing. For EGFR mutations it ranges between 1.4% and 8.5%. The specificity of the test was
Statistical Analysis Differences between covariates and recurrence rate in patients with EGFR-mutant and wild-type tumors were analyzed using Fisher’s exact test, Freeman-Halton extension of Fisher’s exact test, and Student’s t test. The DFS and OS were analyzed using the Kaplan-Meier method, and a log-rank test was used to compare the EGFR-mutant and wild-type group. Cox multivariate analysis was performed with five variables (age, sex, EGFR mutation status, smoking status, and tumor size) to test for independent markers of progression-free survival. Analyses were performed using Prism software (v5.0;
GENERAL THORACIC
Ann Thorac Surg 2013;96:962–8
964
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
Ann Thorac Surg 2013;96:962–8
GENERAL THORACIC
GraphPad Software, San Diego, CA) and SAS software (v9.2; SAS Institute Inc, Cary, NC).
Table 1. Demographics of Patients Including Comorbidities, Tumor Characteristics, and Surgical Intervention
Results
Variable
Patients
Age (y), mean SD 66.6 10.65 Sex, n (%) Male 11 (17.7) Female 51 (82.3) Tobacco history, n (%) Never smoked 35 (56.5) Past smoker 24 (38.7) Current smoker 3 (4.8) Pack-years, mean SD 9.6 17.6 Race White 56 (90.3) Black 0 Asian 6 (9.6) Hispanic 0 Zubrod score 0.37 0.52 Tumor size in cm, n (%) <2 38 (61.3) 2–3 12 (19.3) >3 12 (19.3) Surgical procedure, n (%) Lobectomy 48 (77.4) Wedge resection 12 (19.4) Segmentectomy 2 (3.2) Pneumonectomy 0 EGFR testing, n (%) Direct sequencing 15 (24.2) SNaPshot 47 (75.8) Comorbid conditions Hypertension 26 (41.9) Coronary artery disease 2 (3.2) CHF 1 (1.6) Diabetes mellitus 3 (4.8) COPD 4 (6.4) PFTs and functional status, mean SD 99.8 20.1 FEV1 (% predicted) 89.2 18.8 DLCO (% predicted)
We examined 1,865 consecutive patients who underwent surgical resection for NSCLC between 2004 and 2011 and identified 1,351 with stage I disease. Of these, 307 patients met the inclusion criteria for this study, and follow-up information was collected through October 31, 2012. Sixty-two (20.2%) patients in the study cohort were positive for an EGFR mutation (EGFR-mutant group), and 245 (79.8%) had no mutation in EGFR (wild-type group). Demographic features of the study population are summarized in Table 1. Both groups had a median age of 66 years. As expected, there were more women in the EGFR-mutant group (82% versus 62%; p ¼ 0.004) and more never-smokers (57% versus 10%; p ¼ 0.0001). Although the majority of patients in the study were white, the EGFR-mutant group had an increased proportion of Asians (10% versus 1%; p ¼ 0.01). Tumor size was comparable between genotype groups (p ¼ 0.14). The majority of patients underwent lobectomy as their surgical procedure, and there was no significant difference in the distribution of type of surgery by genotype (p ¼ 0.34). The frequency of some comorbidities, including hypertension, congestive heart failure, and diabetes mellitus, were comparable among genotype groups, whereas the number of patients with chronic obstructive pulmonary disease and coronary artery disease was higher in the wild-type group, likely attributable to the heavier incidence of smoking. There was no difference in the functional status of patients in both groups as measured by the Zubrod score (p ¼ 0.46).
Genotype Results Genotyping was performed using the SNaPshot assay in 238 of 307 patients and by direct sequencing in 69 cases. Sixty-two (20%) patients had an EGFR mutation, of which 47 were detected by SNaPshot analysis and 15 by direct sequencing (Table 2). Among the EGFR-mutant group, 60 had a single mutation and 2 patients had double mutations in EGFR. The most frequently observed EGFR mutations were the exon 21 point mutation L858R, found in 30 (48%) patients, and deletions in the LREA region of exon 19 (del 19), found in 24 (39%) patients. One patient with L858R and 1 patient with del 19 also had the exon 20 mutation T790M, which has been implicated in primary and secondary resistance to EGFR TKIs. The remaining 8 patients had other mutations, including 4 (6%) with exon 20 insertion mutations, 2 (3%) with L861Q, and 2 (3%) with G719C.
Recurrence Rate, Progression-Free Survival, and Overall Survival Median follow-up time was 30 months. There were 85 progression-free survival events observed, including 59 cases of radiographic disease recurrence and 26 deaths.
EGFR Mutant (n ¼ 62)
Wild-Type (n ¼ 245) 66.6 10.1
p Value 0.785 0.004
94 (38.4) 151 (61.6) <0.0001 25 (10.2) 166 (67.8) 54 (22) 43.8 29.4 240 (98) 1 (0.4) 3 (1.2) 1 (0.4) 0.43 0.56
0.0001 0.01 1.0 0.0028 0.46 0.14
116 (47.3) 60 (24.5) 69 (28.2) 0.34 161 68 15 1
(65.7) (27.8) (6.1) (0.4) 0.735
54 (22) 191 (78) 114 44 2 32 56
(47) (18) (0.8) (13) (22.8)
0.192 0.0023 0.493 0.076 0.023
82 22.9 70 22.8
0.0001 0.0001
CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; DLCO ¼ diffusing capacity of the lung for carbon monoxideof % predicted; EGFR ¼ epidermal growth factor receptor gene; FEV1 ¼ forced expiratory volume in 1 second of % predicted; PFTs ¼ pulmonary function tests; SD ¼ standard deviation.
The recurrence rate was 6 of 62 (9.7%) in the EGFRmutant group and 53 of 245 (21.6%) in the wild-type group (p ¼ 0.0315). In 44 of 59 patients (75%) with radiographic disease recurrence, biopsies were obtained and confirmed malignancy. The median DFS in the wild-type group was 7.0 years (95% confidence interval [CI], 4.93 to not reached) compared with 8.83 years (95% CI, 6.2 to not reached) in the EGFR-mutant group (logrank p ¼ 0.0085; Fig 1). This resulted in a hazard ratio (HR) for DFS of 0.45 (95% CI, 0.25 to 0.81). The percentage of patients who remained disease-free at
Ann Thorac Surg 2013;96:962–8
965
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
Table 2. Frequency of Different EGFR Mutations in Resected Stage I Tumors
Table 3. Multivariate Analysis for Predictors of Disease-Free Survival
EGFR Mutation
n
Variable
30 2 2 4 24 2
Age EGFR mutation Sex Smoking status Tumor size
L858 L861G lns G719C Ex20 ins Ex19*del T790Mb a
One patient in this group had a second T790M mutation. patients had a second mutation.
b
Both
EGFR ¼ epidermal growth factor receptor gene.
12 months and 5 years were 97% and 81% in the EGFR-mutant group compared with 94% and 60% in the wild-type group, respectively. In a Cox regression model for DFS using age, sex, smoking status, tumor size (<2 cm, 2 to 3 cm or > 3 cm), and EGFR status as independent variables, only EGFR mutation and tumor size were significant (Table 3). Patients with EGFR mutation were 3.1 times less likely to have disease progression compared with patients with wild-type disease (HR, 0.33; 95% CI, 0.12 to 0.87; p ¼ 0.026), and larger tumor size was associated with an increased risk of disease progression (HR, 1.37; 95% CI, 1.01 to 1.84; p ¼ 0.04). Patients with small malignancies (<2 cm) were 1.9 times less likely to have disease progression (HR, 0.53; 95% CI, 0.29 to 0.97; p ¼ 0.04), whereas age, sex, and smoking status were not independent factors associated with disease progression. Of the 53 patients with disease recurrence in the wildtype group, details about subsequent therapy were available for 52. In the wild-type group, 15 (29%) patients underwent additional surgery, 26 (50%) patients had radiation therapy, and 18 (35%) patients had systemic therapy after disease recurrence, including 17 patients who had intravenous chemotherapy and 3 patients treated with an EGFR TKI (Table 4). In the EGFR-mutant group, 4 patients had additional surgery, 1 patient had radiation therapy, and 3 patients underwent systemic chemotherapy, of which 1 patient received erlotinib.
HR (95% CI) 1.009 0.326 1.15 1.14 1.37
p Value
(0.982–1.037) (0.121–0.874) (0.67–1.97) (0.71–1.82) (1.01–1.84)
0.51 0.026 0.62 0.59 0.04
CI ¼ confidence interval; EGFR ¼ epidermal growth factor receptor gene; HR ¼ hazard ratio.
There were 36 deaths observed, 35 in the wild-type group and 1 in the EGFR-mutant group (p ¼ 0.023). The OS was significantly higher in the mutant group compared with the wild-type group (HR, 0.30; 95% CI, 0.14 to 0.67; p ¼ 0.003). The estimated 5-year survival rate was 98% in the mutant group and 73% in the wild-type group (Fig 2). Lobectomy was the most common surgical treatment implemented in this study, with 77% of EGFR-mutant group patients and 66% of wild-type group patients undergoing this procedure. Among patients treated with lobectomy, the OS was superior in the EGFR-mutant group (HR, 0.45; 95% CI, 0.13 to 0.83; p ¼ 0.017) compared with the EGFR wild-type group (Fig 3). The estimated 5-year survival rate was higher in the EGFR-mutant group compared with the wild-type group (98% versus 70%). There was no difference in OS by genotype among those undergoing sublobar resection (wedge or segmentectomy; data not shown). Furthermore, there were no significant differences in OS or DFS among all patients who had lobectomy compared with all patients who had sublobar resection (not segregated by EGFR status; data not shown).
Comment In this study, we investigated the prognostic implication of EGFR mutations in a cohort of 307 patients with completely resected stage I NSCLC and known EGFR mutation status. To focus solely on postoperative prognostic differences by genotype, the cohort was restricted to patients who had no systemic chemotherapy and no radiation therapy. We found that EGFR-mutant lung Table 4. Treatment Modality in Patients After Disease Recurrencea Therapy After Recurrence Surgical resection Radiation therapy Chemotherapy Intravenous EGFR TKI Supportive
Fig 1. Overall survival in resected stage I non-small cell lung cancer. (wt/WT ¼ wild-type.)
EGFR Mutant (n ¼ 6)
Wild-Type (n ¼ 52)
4 1 3 2 1 2
15 26 18 17 3 8
a Note that the sum of different treatments does not equal the number of recurrences as some patients received multiple treatment modalities.
EGFR TKI ¼ epidermal growth factor receptor tyrosine-kinase inhibitor.
GENERAL THORACIC
a
966
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
GENERAL THORACIC
Fig 2. Disease-free survival in resected stage I non-small cell lung cancer. (wt/WT ¼ wild-type.)
cancer patients had lower recurrence rates and significantly prolonged DFS and OS compared with those with EGFR wild-type cancers. Positivity for EGFR mutation and small tumors (<2 cm) were identified as independent predictors of DFS when adjusting for potential confounders, including age, sex, and smoking status. To our knowledge, this is the first study to demonstrate a DFS and OS benefit for resected stage I lung cancer patients who harbor EGFR mutations compared with those with wild-type disease. The first suggestion of a possible prognostic role for EGFR mutations emerged from an analysis of patients with stage IIIB and IV NSCLC who participated in the phase III TRIBUTE trial [21], in which patients were randomly assigned to receive firstline carboplatin and paclitaxel with either erlotinib or placebo. Here, EGFR-mutant disease was associated with longer OS and improved response rate regardless of treatment arm [9]. More recent studies have confirmed that advanced NSCLC patients with EGFR mutations seem to fare better on chemotherapy than patients with wild-type EGFR [5], but trials performed in patients with advanced disease cannot inform us of whether the mutation status is an inherent prognostic factor or a predictor of improved outcome on chemotherapy because all patients in these trials receive treatment of some sort for their metastatic cancer. Hence, there is great interest in looking at outcomes in an earlier stage
Fig 3. Overall survival in stage I non-small cell lung cancer after lobectomy. (wt/WT ¼ wild-type.)
Ann Thorac Surg 2013;96:962–8
of the disease to understand the true prognostic significance of EGFR mutations. Previous studies examining the prognostic value of EGFR mutation status in resected stage I through III NSCLC are limited, and their results are somewhat conflicting [10–17]. Several have suggested that EGFR mutations impart an improved prognosis after surgery. A Japanese study was fairly small, with only 53 patients, but had an impressive median follow-up time of more than 60 months and found that mutation of the EGFR gene was associated with a strong trend toward improved survival in stage I through III disease (71% versus 56%) and a significantly superior 5-year survival rate among the 36 patients with stage I disease (92% versus 57%; p ¼ 0.037) [10]. Another study of 294 patients with stage I through III disease found an improved 3-year survival rate (p ¼ 0.02) and OS (p ¼ 0.031) for the patients with EGFR-mutated tumors (n ¼ 40) compared with those with KRAS-mutant or wild-type disease; however, 40% of patients included in this cohort received cytotoxic chemotherapy in addition to surgery [14]. A second group evaluating the prognostic role of EGFR and KRAS mutations in 164 Taiwanese patients with stage I through III disease showed a significantly higher 3-year survival rate (p ¼ 0.02) in EGFR-mutated NSCLC compared with wild-type and KRAS-mutant tumors [16]. There was a trend toward longer OS in the EGFR group (55 months) compared with tumors that were KRAS and EGFR wildtype (35 months); however, this difference was not statistically significant (p ¼ 0.2). The analysis in this study is clouded by use of chemotherapy, radiation, and EGFR TKIs in 6%, 11%, and 9% of patients, respectively. Finally, a recently published large study (n ¼ 1,118) from Memorial Sloan Kettering investigated outcomes among patients of all stages, although more than 97% had stage I through III disease. In that study, EGFR-mutated disease was associated with a significantly improved OS compared with wild-type tumors (p < 0.001) and remained significant after adjusting for stage [12]. However, a significant proportion of patients in this cohort again received chemotherapy (21%), highlighting the chief difficulty in interpretation of this body of literature. Because EGFR-mutant tumors appear more sensitive to chemotherapy than wild-type tumors, it is challenging to understand prognosis when patients have been exposed to this potentially confounding treatment. A second group of studies have found that apparent prognostic advantages for EGFR-mutant cancers may be lost after adjusting for potential confounding factors [14, 15, 17]. In a study evaluating the prognostic importance of EGFR, KRAS, and TP53 mutations in 397 Japanese patients with stage I through IV NSCLC, EGFR-mutant disease (n ¼ 196) was associated with significantly improved OS (p ¼ 0.006), particularly in patients with the two major EGFR mutations, L858R and exon 19 deletions [17]. Of note, 56 patients were treated with gefitinib in this cohort and were not included in the survival analysis owing to concern over confounding factors. After adjusting for other potential confounders, including sex, stage (I versus II through IV), tumor differentiation, and smoking history, the survival benefit
Ann Thorac Surg 2013;96:962–8
967
provides further evidence to support routine EGFR mutation testing on resected stage I adenocarcinomas. Financial support for this study was provided by the Division of Thoracic Surgery at the Massachusetts General Hospital. We would like to acknowledge our data manager and research coordinator, Sheila Cann and Diane Davies, respectively, for their diligence and dedication toward compiling and maintaining the Thoracic Surgery Database.
References 1. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg 1995;60:615–23. 2. Goldstraw P, Crowley J, Chansky K, et al; International Association for the Study of Lung Cancer International Staging Committee; Participating Institutions. The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM Classification of malignant tumours. J Thorac Oncol 2007;2:706–14. 3. Song WA, Zhou NK, Wang W, et al. Survival benefit of neoadjuvant chemotherapy in non-small cell lung cancer: an updated meta-analysis of 13 randomized control trials. J Thorac Oncol 2010;5:510–6. 4. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010;363:1693–703. 5. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009;361:947–57. 6. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129–39. 7. Paez JG, J€ anne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497–500. 8. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain. PLoS Med 2005;2:e73. 9. Eberhard DA, Johnson BE, Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic indicators in patients with non-small-cell lung cancer treated with chemotherapy alone and in combination with erlotinib. J Clin Oncol 2005;23:5900–9. 10. Sonobe M, Nakagawa M, Takenaka K, et al. Influence of epidermal growth factor receptor (EGFR) gene mutations on the expression of EGFR, phosphoryl-Akt, and phosphorylMAPK, and on the prognosis of patients with non-small cell lung cancer. J Surg Oncol 2007;95:63–9. 11. Janjigian YY, Park BJ, Zakowski MF, et al. Impact on diseasefree survival of adjuvant erlotinib or gefitinib in patients with resected lung adenocarcinomas that harbor EGFR mutations. J Thorac Oncol 2011;6:569–75. 12. D’Angelo SP, Janjigian YY, Ahye N, et al. Distinct clinical course of EGFR-mutant resected lung cancers: results of testing of 1118 surgical specimens and effects of adjuvant gefitinib and erlotinib. J Thorac Oncol 2012;7:1815–22. 13. Lim KH, Huang MJ, Liu HC, Kuo HT, Tzen CY, Hsieh RK. Lack of prognostic value of EGFR mutations in primary resected non-small cell lung cancer. Med Oncol 2007;24:388–93. 14. Marks JL, Broderick S, Zhou Q, et al. Prognostic and therapeutic implications of EGFR and KRAS mutations in resected lung adenocarcinoma. J Thorac Oncol 2008;3:111–6. 15. Kim YT, Seong YW, Jung YJ, et al. The presence of mutations in epidermal growth factor receptor gene is not a prognostic factor for long-term outcome after surgical
GENERAL THORACIC
for EGFR-mutant disease lost statistical significance, although tumor stage remained an independent marker for OS. Median follow-up was relatively short in this study (32 months), and the use of chemotherapy and radiation was not reported, although it is conceivable that a significant portion of patients received systemic therapy given that 149 patients had stage II through IV disease, of which the majority had stage III and IV disease (n ¼ 105). Similar results were reported in a cohort of 863 Korean patients with stage I through IV NSCLC [15]. A total of 354 patients had an EGFR mutation, and the majority had early-stage disease. The EGFR-mutated tumors were associated with a significantly improved OS compared with the wild-type tumors (p ¼ 0.001). However, in a multivariate analysis adjusting for potential confounders, EGFR mutation status did not remain an independent prognostic marker. This study was also limited by a relatively short median follow-up period of 24 months. Furthermore, there was a proportional imbalance of patients with stage II through IV cancers between the genotype groups (36% versus 48%), suggesting a likely significant difference in the receipt of systemic treatment that could have blunted the interpretation of prognosis. These collective results stress the need for carefully studying a patient population best suited to minimize the primary confounding variable of prior research and systemic treatment exposure. Our study looked at a uniform stage of disease (stage I only) and excluded any patient who received chemotherapy, radiation therapy, or EGFR TKIs to focus solely on the biology of the cancer genotype and its impact on recurrence and survival. Although type of surgery can affect recurrence rates [1], three quarters of patients in our cohort received a lobectomy, and separate analysis of this population was also significant for improved survival among the EGFR-mutant group. Our study must be interpreted in the context of its limitations, such as its retrospective design and its singlecenter patient cohort. We analyzed outcomes based on the EGFR mutation status alone and did not investigate the influence of other potentially relevant genotype biomarkers such as KRAS. Furthermore, given the aim and design of this study, the patient population was smaller compared with some of the other studies discussed above, which included multiple stages from I to IV. In summary, our study examined a large cohort of stage I NSCLC patients in which no patient received adjuvant or neoadjuvant therapy of any sort. We identified EGFR mutation-positive status as an independent prognostic marker associated with decreased recurrence and improved progression-free survival and OS. Although past studies have shown little benefit for cisplatin-based adjuvant chemotherapy in stage I NSCLC, our data suggest it may be reasonable to design future prospective trials of adjuvant chemotherapy to include only patients with EGFR wild-type tumors because the natural history prognosis of stage I EGFR-mutant tumors is excellent. In addition, clinical trials examining the benefit of adjuvant EGFR TKIs will require that comparisons be made only within genotype-similar groups. This analysis also
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
968
IZAR ET AL EGFR MUTATION AND STAGE I NSCLC
resection of non-small-cell lung cancer. J Thorac Oncol 2013;8:171–8. 16. Liu HP, Isaac Wu HD, Chang JWC, et al. Prognostic implications of epidermal growth factor receptor and KRAS gene mutations and epidermal growth factor receptor gene copy numbers in patients with surgically resectable non-small cell lung cancer in Taiwan. J Thorac Oncol 2010;5:1175–84. 17. Kosaka T, Yatabe Y, Onozato R, Kuwano H, Mitsudomi T. Prognostic implication of EGFR, KRAS, and TP53 gene mutations in a large cohort of Japanese patients with surgically treated lung adenocarcinoma. J Thorac Oncol 2009;4:22–9. 18. Marchetti A, Martella C, Felicioni L, et al. EGFR mutations in non–small-cell lung cancer: analysis of a large series of cases and development of a rapid and sensitive method for
Ann Thorac Surg 2013;96:962–8
diagnostic screening with potential implications on pharmacologic treatment. J Clin Oncol 2005;23:857–65. 19. Sequist LV, Joshi VA, J€ anne PA, et al. Epidermal growth factor receptor mutation testing in the care of lung cancer patients. Clin Cancer Res 2006;12(14 Pt 2):4403s–8s. 20. Dias-Santagata D, Akhavanfard S, David SS, et al. Rapid targeted mutational analysis of human tumours: a clinical platform to guide personalized cancer medicine. EMBO Mol Med 2010;2:146–58. 21. Herbst RS, Prager D, Hermann R, et al. TRIBUTE Investigator Group. TRIBUTE: a phase III trial of erlotinib hydrochloride (OSI-774) combined with carboplatin and paclitaxel chemotherapy in advanced non-small-cell lung cancer. J Clin Oncol 2005;23:5892–9.
GENERAL THORACIC
INVITED COMMENTARY As lung cancer remains the number-one worldwide cause of cancer-related deaths, and non-small cell lung cancer (NSCLC) accounts for 85% of primary lung cancer cases, significant recent efforts have been put toward developing new therapeutic options to improve outcomes from this disease. Specifically, investigators have taken great interest in attempts to recognize specific driver mutations that foster neoplastic transformation and contribute to tumor progression. The dominant theme of such investigations has been aimed toward identifying specific molecular targets to develop new agents with the capacity to kill cancer cells with minimal toxicity to normal cells. Epidermal growth factor receptor (EGFR) mutations have been identified in 13% to 22% of patients with NSCLC, and the development of EGFR tyrosine kinase inhibitors as successful therapeutic agents for patients harboring this mutation has led many institutions to implement routine testing for this genetic aberrancy. A key purpose for such genetic testing is to determine susceptibility to available treatments; however, additional information may potentially be gained from such investigations, and, as exemplified by Izar and colleagues [1], may provide us with highly informative prognostic data. Such data allow us to better identify patients’ risk of recurrence—ultimately guiding us in establishing optimal treatment strategies, determining the most appropriate follow-up surveillance plans, and communicating effectively with patients regarding their futures. While previous authors have shown EGFR mutations to be associated with improved treatment response in advanced NSCLC, the effect of EGFR mutation status on stage I disease has remained largely unclear. In this insightful report, the authors aimed to identify the prognostic role of EGFR mutations among patients with completely resected, early-stage NSCLC, which has significant relevance to us as thoracic surgeons. Direct sequencing and multiplex polymerase chain reaction-based assays were used to determine mutation status for 307 treatment-naive patients with T1-2a N0 disease who underwent operative resection with curative intent. The authors subsequently measured recurrence rates, progression-free survival (PFS), and overall survival in patients with EGFR mutations (20% of the study population) compared with the wild-type individuals. Ó 2013 by The Society of Thoracic Surgeons Published by Elsevier Inc
Over a median follow-up of 30 months, the individuals with EGFR mutations were less likely to have tumor recurrence (9.7% vs 21.6%) and experienced longer progression-free survival (8.8 vs 7.0 years) than wild-type patients. Cox regression analysis was used to examine several tumor/patient variables, finding only tumor size and EGFR status were associated with risk of disease progression. EGFR mutations also conveyed improved overall survival at 5 years of 98% vs 73%. Findings were consistent regardless of the extent of the resection performed. By studying a large cohort of patients who received neither adjuvant nor neoadjuvant therapy, the authors have examined the true biologic effect of EGFR status on outcome, contributing greatly to our understanding of the prognostic implication of EGFR mutations in NSCLC. However, this provocative study leaves several questions unanswered and has set a foundation for future investigations. The molecular mechanisms leading to more indolent growth of EGFR-mutated tumors have not been defined, and further bench work is required. Further, with evidence that patients with early-stage NSCLC and EGFR mutations are less likely to experience recurrence, we are left with questions about how we ought to alter our treatment of these individuals. Should their surveillance plans be tailored accordingly? Should we expand our adjuvant chemotherapeutic strategies for EGFR wild-type patients? Like all excellent studies, this report has raised more questions than it answers, providing impetus for a number of future trials. Mara B. Antonoff, MD Department of Surgery Division of Cardiothoracic Surgery Washington University St. Louis, MO 63110 e-mail: [email protected]
Reference 1. Izar B, Sequist L, Lee M, et al. The impact of EGFR mutation status on outcomes in patients with resected Stage I non-small cell lung cancers. Ann Thorac Surg 2013;96:962–8.
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.06.059