CAP Laboratory Improvement Programs. Worldwide Frequency of Commonly Detected EGFR Mutations

Transcription

1 CAP Laboratory Improvement Programs Worldwide Frequency of Commonly Detected EGFR Mutations Rondell P. Graham, MBBS; Amanda L. Treece, MD; Neal I. Lindeman, MD; Patricia Vasalos, BS; Mu Shan, BS; Lawrence J. Jennings, MD, PhD; David L. Rimm, MD, PhD Context. Recurrent epidermal growth factor receptor (EGFR) mutations are seen in a subset of pulmonary adenocarcinomas. These mutations are targeted by EGFR inhibitors and are a biomarker for response to EGFR inhibitor therapies. Initial data have indicated an increased frequency of activating EGFR mutations in nonsmoking Asian females. However, there are very few studies of global scope that address the question of mutation distribution across the population of lung cancer. Objective. To determine the frequency of EGFR mutations in exons 18 through 21 detected in clinical laboratories participating in the College of American Pathologists proficiency testing program for EGFR in calendar year Design. We reviewed the surveys from 170 clinical laboratories from 20 countries that participated in the College of American Pathologists EGFR proficiency testing program. The proficiency testing includes questions regarding the total numbers of tests performed at each common mutation site, including both activating and resistance mutations, and their frequency. Countries were grouped into regional groups in order to assess frequency of mutation by type, and to indirectly assess ethnic differences in mutation frequencies. Results. Among the treatment-sensitive activating mutations, the most common are exon 19 mutations (n ¼ of cases; 7.9% of total cases tested) and the exon 21 L858R mutation (n ¼ of cases; 7.6% of total cases tested) and the least common are exon 20 mutations (n ¼ 466 of cases; 0.3% of total cases tested). The T790M mutation in exon 20 is the more common resistance mutation (n ¼ 1010 of cases; 0.7% of all cases tested). The highest activating mutation frequency is seen in southern Asia (n ¼ 4260 of 9337 cases; 46%) and the lowest activating mutation frequencies are in South and North America (n ¼ 113 of 1439 cases and 7926 of cases; 8% and 9%, respectively). Conclusions. Our data confirm that activating EGFR mutations are more common in southern Asia and that the distribution of activating EGFR mutations varies significantly across the regions. Similarly, the frequency and distribution of resistance mutations also show significant variation when comparing southern Asia with other regions. (Arch Pathol Lab Med. 2018;142: ; doi: / arpa cp) Recurrent epidermal growth factor receptor (EGFR) mutations characterize a subset of pulmonary adenocarcinomas and represent an effective therapeutic target for some tyrosine kinase inhibitors. 1,2 In patients whose tumors harbor somatic activating mutations in EGFR, treatment Accepted for publication April 26, Published as an Early Online Release November 6, From the Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota (Mr Graham); the Department of Pathology and Laboratory Medicine, Children s Hospital Colorado, Denver (Dr Treece); the Department of Pathology, Brigham and Women s Hospital, Boston, Massachusetts (Dr Lindeman); the Surveys Department, College of American Pathologists, Northfield, Illinois (Mss Vasalos and Shan); the Department of Pathology and Laboratory Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois (Dr Jennings); and the Department of Pathology, Yale University School of Medicine, New Haven, Connecticut (Dr Rimm). Mr Graham and Dr Treece contributed equally to this work. The authors have no relevant financial interest in the products or companies described in this article. Reprints: Rondell P. Graham, MBBS, Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First St SW, Rochester, MN ( graham.rondell@mayo.edu). with the inhibitors gefitinib and erlotinib provided significantly better response rates compared with those tumors without mutations. 1,2 The response to anti-egfr therapy was also superior to that seen with standard platinum-based chemotherapy. 3 Prior to the identification of EGFR mutations and the use of these novel EGFR inhibitors, the clinical response of patients to chemotherapy was dismal and survival was poor. 4 EGFR inhibitors have significantly improved the clinical outcome of these patients, and as such, EGFR mutation analysis is now commonly performed for cases of metastatic pulmonary adenocarcinoma. 5 Initial data indicated an increased frequency of activating EGFR mutations in nonsmoking females of Asian descent. 1,6,7 However, studies of global scope since this original description are lacking. The College of American Pathologists (CAP) provides proficiency test material and collects results of proficiency assessment and accompanying survey data from molecular laboratories in the United States and other countries enrolled in the CAP proficiency test for EGFR mutations in lung adenocarcinoma. The accompanying survey includes a number of questions about the workload of each participating laboratory. This study was designed to use that Arch Pathol Lab Med Vol 142, February 2018 Detection Rates of EGFR Mutations in Lung Cancer Graham et al 163

2 Table 1. Number of Cases Tested for EGFR Mutations in the Respective Countries a Country No. of Cases Regional Grouping Belgium 380 Europe Brazil 1439 South America Canada 6306 North America China 2702 Northern Asia France 80 Europe Germany 22 Europe Hong Kong 770 Southern Asia India 2077 Southern Asia Israel 84 Africa Japan Northern Asia Jordan 104 Africa Netherlands 19 Europe Philippines 569 Southern Asia Portugal 529 Europe Saudi Arabia 41 Africa Singapore 818 Southern Asia South Africa 94 Africa South Korea 2113 Northern Asia Taiwan 5103 Southern Asia United States North America a The laboratory in Switzerland was excluded because no lung cancer specimens were tested during the study period. survey to determine the frequency of EGFR mutations in exons 18 through 21 detected in clinical laboratories participating in the CAP proficiency testing program for EGFR in a 12-month period. MATERIALS AND METHODS We surveyed 170 participating clinical laboratories from 20 countries regarding the number of clinical samples tested for EGFR mutations in a 12-month period (March 2013 March 2014); the number of cases with activating mutations in exons 18, 19, 20, and 21 (specifically L858R and L861Q); and the number of cases with the resistance mutations S768I and T790M in exon 20. The frequencies of activating mutations and resistance mutations for each laboratory were determined. Countries were grouped into regional groups in order to assess indirectly for ethnic differences in mutation frequencies. The significance of differences in activating and resistance mutation distribution among regions was examined using v 2 tests. Bonferroni correction was applied such that a significant P value was P.001. Also, the activating mutations responsible for these differences were analyzed by v 2 tests with Bonferroni correction. Bonferroni correction is a statistical method used to reduce the likelihood of false-positive statistical associations due to multiple pairwise comparisons. For this study, comparison of multiple continents/regions necessitated the application of this statistical method. Regions with fewer than 100 cases evaluated were excluded from statistical comparisons. RESULTS During the study period, tests were performed in the surveyed clinical laboratories. The greatest volumes of clinical tests were performed in the United States (n ¼ of ; 59%) and Japan (n ¼ of ; 24%). Table 1 shows the test volume in each country. The countries are combined largely by regions for the remainder of the analyses, split into 6 regions including northern Asia, southern Asia, Europe, Africa (including the Middle East), South America, and North America. Taiwan had the highest rate of detection of EGFR-activating mutations at 55% (2802 of 5103 total cases tested there), followed by China 37% (1009 of 2702 total cases), then Japan with 29% (9644 of total cases) and India with the same rate of 29% (605 of 2077 total cases). Activating Mutations There were cases that were positive for any of the activating mutations listed (17% of total cases; ). Activating mutations were most frequent in southern Asia and were identified in 4260 of the 9337 cases tested there (46%), followed by of cases (30%) in northern Asia. Activating mutations were least common in South America, present in only 113 of 1439 cases (7.9%; Figure). Table 2 summarizes the relative frequency of the various EGFR-activating mutations. The distributions of mutations and mutation frequencies across the various regions were significantly different (P,.001). The pairwise v 2 testing was performed by using conservative Bonferroni adjustment. Southern Asian laboratories reported a significantly higher activating mutation rate compared with all other regions. The bar graph shows the mutation detection rate across the regions evaluated in this study. 164 Arch Pathol Lab Med Vol 142, February 2018 Detection Rates of EGFR Mutations in Lung Cancer Graham et al

3 Region Table 2. Country Distribution and Frequency of Various Activating EGFR Mutations Mutations Exon 18 Exon 19 Exon 20 Exon 21 L858R Exon 21 L861Q Total Cases Tested (% Mutation Detection Rate) Northern Asia China Japan South Korea Total (%) 750 (2.0) 4841 (12.8) 27 (0.07) 5253 (13.9) 397 (1.0) (30.0) North America Canada United States Total (%) 445 (0.5) 3888 (4.5) 335 (0.4) 2989 (3.4) 269 (0.3) (9.2) South America Brazil 8 (0.6) 60 (4.2) 4 (0.3) 37 (2.6) 4 (0.3) 1439 (7.9) Europe Belgium France Germany Netherlands Portugal Total (%) 7 (0.7) 93 (9.0) 3 (0.3) 34 (3.3) 1 (0.1) 1030 (13.4) Africa and Middle East Israel Jordan Saudi Arabia South Africa Total (%) 2 (0.6) 24 (7.4) 6 (1.9) 17 (5.3) 3 (0.9) 323 (16.1) Southern Asia India Hong Kong Philippines Singapore Taiwan Total (%) 198 (2.1) 1896 (20.3) 91 (1.0) 2021 (21.6) 54 (0.6) 9337 (45.6) Total (% of all cases tested) 1410 (1.0) (7.9) 466 (0.3) (7.6) 728 (0.5) (17.4) Exon 18 Mutations. Exon 18 mutations were significantly more commonly detected in northern Asia (n ¼ 750 of ; 2.0%) and southern Asia laboratories (n ¼ 198 of 9337; 2.1%) compared with laboratories in Europe (n ¼ 7of 1030; 0.7%), North America (n ¼ 445 of ; 0.5%), and South America (n ¼ 8 of 1439; 0.6%). Exon 19 Mutations. Laboratories in southern Asia had a significantly higher exon 19 mutation detection rate (n ¼ 1896 of 9337; 20.3%) compared with all other regions. Exon 19 mutations were also significantly more commonly detected in northern Asia (n ¼ 4841 of ; 12.8%) compared with Europe (n ¼ 93 of 1030; 9.0%), North America (n ¼ 3888 of ; 4.5%), and South America (n ¼ 60 of 1439; 4.2%). Also, from these data, one can appreciate that European laboratories detected exon 19 mutations more frequently than their North American and South American counterparts. Exon 20 Mutations. Exon 20 mutations were significantly more commonly detected in Africa (including the Middle East; n ¼ 6 of 323; 1.9%) compared with Europe (n ¼ 3 of 1030; 0.3%), North America (n ¼ 335 of ; 0.4%), South America (n ¼ 4 of 1439; 0.3%), and northern Asia (n ¼ 27 of ; 0.07%). L858R and L861Q Mutations. Exon 21 L858R mutations were significantly more commonly detected in southern Asia (n ¼ 2021 of 9337; 21.6%) and northern Asia (n ¼ 5253 of ; 13.9%) compared with all other regions. Southern Asia had a significantly higher exon 21 L858R mutation rate compared with northern Asia. Exon 21 L861Q mutations were significantly more common in northern Asia (n ¼ 397 of ; 1.0%) compared with Europe (n ¼ 1of 1030; 0.1%), North America (n ¼ 269 of ; 0.3%), and southern Asia (n ¼ 54 of 9337; 0.6%). Most Common Activating Mutations. Exon 19 mutations (n ¼ of ; 45% of detected mutations) and exon 21 L858R mutations (n ¼ of ; 44% of detected mutations) were the most common activating mutations overall, both identified in approximately 8% of total cases (n ¼ and , respectively, of ). The exon 21 L858R mutation was the single most common mutation detected in both northern (n ¼ 5253 of cases; 14%) and southern Asia (n ¼ 2021 of 9337 cases; 22%), whereas exon 19 mutations were the most common in all other regions (4% 9% of cases; 3888 of in North America, 60 of 1439 in South America, 93 of 1030 in Europe, and 24 of 323 in Africa and the Middle East). Exon 18 mutations were identified in 1410 of cases (1%), and were most common in southern and northern Asia. Exon 20 and exon 21 L861Q mutations were the least commonly identified, with exon 20 mutations being most common in Africa and the Middle East and the exon 21 L861Q mutation being most common in northern Asia. Sample numbers for these rare mutations were too low for statistical analysis. Resistance Mutations EGFR-resistance mutations (S768I and T790M) were identified in 1375 of cases (1%) tested across all laboratories (Table 3). The distribution of cases across the various regions was significantly different (P,.001). Pairwise Fisher exact testing was performed using Bonferroni adjustment. Southern Asia laboratories detected resistance mutations (n ¼ 228 of 9337; 2.4%) at a significantly higher rate than laboratories in Europe (n ¼ 5 of 1030; 0.5%), North America (n ¼ 536 of ; 0.6%), South America (n ¼ 3 of 1439; 0.2%), and northern Asia (n ¼ 605 of ; 1.6%). Arch Pathol Lab Med Vol 142, February 2018 Detection Rates of EGFR Mutations in Lung Cancer Graham et al 165

4 Table 3. Distribution and Frequency of Cases Positive for the Resistant Mutations S768I and T790M Mutations No. of Cases Tested Region Country S768I T790M (% Mutation Detection Rate) Northern Asia China Japan South Korea Total (%) 75 (0.2) 527 (1.4) (1.6) North America Canada United States Total (%) 221 (0.2) 315 (0.4) (0.6) South America Brazil (%) 2 (0.1) 1 (0.07) 1439 (0.2) Europe Belgium France Portugal Germany Netherlands Total (%) 4 (0.4) 1 (0.1) 1030 (0.5) Africa and Middle East Israel Jordan Saudi Arabia South Africa Total (%) 0 (0) 1 (0.3) 323 (0.3) Southern Asia India Hong Kong Philippines Singapore Taiwan Total (%) 63 (0.7) 165 (1.8) 9337 (2.4) Total (% of all cases tested) 365 (0.3) 1010 (0.7) (1.0) S768I mutations were significantly more common in southern Asia (n ¼ 63 of 9337; 0.7%) compared with North America (n ¼ 221 of ; 0.2%) and northern Asia (n ¼ 75 of ; 0.2%). The very rare detection of S768I in other regions precluded statistical analysis. T790M mutations were significantly more common in southern Asia (n ¼ 165 of 9337; 1.8%) and northern Asia (n ¼ 527 of ; 1.4%) compared with Europe (n ¼ 1 of 1030; 0.1%) and North America (n ¼ 315 of ; 0.4%). DISCUSSION The CAP provides proficiency testing for an array of analytes for clinical laboratories, including EGFR mutations. As part of the CAP s proficiency testing program, the CAP surveys clinical laboratories to determine the frequency of positive results for various EGFR mutations as well as the volume of clinical tests for EGFR mutations also performed. In view of the global scope of this program covering 170 laboratories in 20 countries, this study allowed the determination of the frequency and distribution of EGFR mutations in various geographic locations across the world that used CAP accreditation. Landmark studies by Lynch and colleagues 1 and Paez et al 2 showed that response to anti-egfr therapy in lung adenocarcinoma corresponded to somatic activating EGFR mutations in the lung tumors. The therapeutic response observed with anti-egfr therapy is also superior to that of conventional chemotherapy. 8,9 Consequently, EGFR testing and anti-egfr therapy have become the standard of care for patients with advanced non small cell lung carcinoma. 10,11 Given the clinical significance of the detection of EGFR mutations and the need for conscientious use of testing resources, the CAP, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology have published guidelines for ALK and EGFR testing. 5 Currently, this multidisciplinary group recommends ancillary ALK and EGFR testing on all patients with advanced-stage lung carcinoma, regardless of age, sex, smoking history, or clinical risk factors. 5 Initial data indicated that there was an increased frequency of EGFR mutations in female nonsmokers of Asian ancestry. 12 Several large studies followed that confirmed these clinical associations. One of the largest studies to date, 13 which included nonsmokers or former light smokers from multiple Asian countries, demonstrated an EGFR mutation rate of almost 60% in 437 patients. Another large study 14 that assayed 604 North American non-asian patients and 464 East Asian patients found a significantly lower EGFR mutation rate in the non-asian patients (25% versus 39%). Numerous smaller studies have confirmed the higher EGFR mutation rate in Asian patients. A review of the literature 5 demonstrated an EGFR mutation prevalence of 45% in the Pacific Asian population compared with 24% in the white population, though differences in study design and population selection were not taken into account in that review. Nonetheless, overall, the data are consistent in the enrichment for EGFR mutation in non small cell lung carcinoma in Asian patients. Based on our survey data, southern Asian laboratories indeed reported the highest EGFR mutation detection rate, whereas the North American and South American laboratories had the lowest mutation detection rate. This difference in mutation detection rate observed in this study is likely related to several factors, the first of which is the biology of lung carcinoma. The concordance of these data with previously published studies highlighting the increased frequency of activating EGFR mutations in Asian populations supports this interpretation. However, the authors recognize that the mutation detection rates presented in this study may also be impacted by testing patterns driven by 166 Arch Pathol Lab Med Vol 142, February 2018 Detection Rates of EGFR Mutations in Lung Cancer Graham et al

5 medical practice standards in different locales. For example, in resource-replete settings in the Western world, there is an ability to test more patients than in resource-limited settings. As such, this could lead to a decrease in mutation detection rate in North American laboratories. In fact, the authors are aware of laboratories where EGFR mutation testing is performed universally, 15 but a review of test use is beyond the scope of this study. Concerning the mutations, there was significant enrichment for detection of exon 18 mutations in southern and northern Asia. Similarly, exon 19 and L858R mutations were detected significantly more frequently in southern Asia, whereas L861Q mutations were more often detected in northern Asia. In contrast, exon 20 mutations were detected significantly more often in Africa (including the Middle East). These findings add to the observation of increased detection of EGFR mutations in Asia by highlighting that individual mutations and groups of mutations also show significant geographic differences. The detection rate and distribution of resistance mutations also showed significant variation when comparing southern and northern Asia with other regions. EGFR T790M was significantly more commonly detected in southern and northern Asia compared with other regions, whereas the detection rates in southern and northern Asia were similar. EGFR S768I mutations were significantly more common in southern Asia compared with North America and northern Asia. These findings are intuitive, as resistance mutations develop in response to treatment. As such, the populations with the highest EGFR-activating mutation rate will also display the greatest detection rate for resistance mutations. Taken together, the data indicate geographic and thus presumably ethnic differences in the distribution of somatic EGFR alterations in lung carcinoma. The factors responsible for this are unknown. It has been shown that the mutational spectrum in lung carcinomas arising in smokers is distinct from that in tumors affecting nonsmokers. 7,16,17 It is also known that EGFR mutations are negatively associated with smoking. 1,2,7 Consequently, alternate environmental and/or germline alterations may determine the differences in EGFR mutational distribution worldwide. Our study is limited by the reliance on survey-level data and the lack of information on platforms used. Although the survey was designed as a quick method of data collection from subscribers, it was not designed as a comprehensive instrument for assessment of molecular diagnostics practices. Additionally, it is possible that laboratories reporting zero positive results for certain mutations may not capture that mutation in their assays. Data clarifying which mutations are captured in each laboratory are not available in the survey results, and thus, the total number of specimens tested from these laboratories was used in the denominator for total number of tests performed. This survey is also limited in that only institutions that subscribe to this method of proficiency testing are included. We do not have data from many countries that use their own methods for quality control, including some of northern and eastern Europe, Russia, Australia/New Zealand, and Scandinavia. Nevertheless, this assessment provides detection rates of EGFR mutations in a large worldwide sampling of countries and regions. There are a number of potential ways to assign countries and regions in a study of this nature. For example, the assignment of countries may be based on geographic proximity, or alternatively the assignment might be based on the predominant ethnicity within the country under consideration. Both approaches have their limitations. First, grouping by predominant ethnicity may, by definition, result in the aggregation of individuals from different ethnic backgrounds within a given region. This would be difficult to avoid given today s commonplace migration of individuals. Alternatively, grouping by geographic proximity may result in the separation of individuals from a given ethnicity across regions. Importantly, the data in this study provide the specific data from each involved country, allowing for follow-up study of hypotheses related to geography and/or ethnicity. In conclusion, we provide results from the CAP proficiency testing on EGFR mutation analysis that indicate the relative detection rate of EGFR mutations. Although not truly representative of the population across the world, we believe this work represents the largest and broadest study of EGFR mutation frequency. It will be interesting to revisit this data source in the future to determine the stability and reproducibility of the observations made in this work. References 1. 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(21): Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science. 2004;304(5676): Liang W, Zhang Y, Kang S, et al. Impact of EGFR mutation status on tumor response and progression free survival after first-line chemotherapy in patients with advanced non-small-cell lung cancer: a meta-analysis. J Thorac Dis. 2014; 6(9): Giaccone G. Targeted therapy in non-small cell lung cancer. Lung Cancer. 2002;38(suppl 2):S29 S Lindeman NI, Cagle PT, Beasley MB, et al. Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. J Mol Diagn. 2013;15(4): Shi Y, Au JS, Thongprasert S, et al. A prospective, molecular epidemiology study of EGFR mutations in Asian patients with advanced non-small-cell lung cancer of adenocarcinoma histology (PIONEER). J Thorac Oncol. 2014;9(2): Zhang Y, Sun Y, Pan Y, et al. Frequency of driver mutations in lung adenocarcinoma from female never-smokers varies with histologic subtypes and age at diagnosis. Clin Cancer Res. 2012;18(7): Fong T, Morgensztern D, Govindan R. EGFR inhibitors as first-line therapy in advanced non-small cell lung cancer. J Thorac Oncol. 2008;3(3): Neal JW, Sequist LV. First-line use of EGFR tyrosine kinase inhibitors in patients with NSCLC containing EGFR mutations. Clin Adv Hematol Oncol. 2010;8(2): Rosell R, Moran T, Queralt C, et al. Screening for epidermal growth factor receptor mutations in lung cancer. N Engl J Med. 2009;361(10): Haaland B, Tan PS, de Castro G Jr, Lopes G. Meta-analysis of first-line therapies in advanced non-small-cell lung cancer harboring EGFR-activating mutations. J Thorac Oncol. 2014;9(6): Hsieh RK, Lim KH, Kuo HT, Tzen CY, Huang MJ. Female sex and bronchioloalveolar pathologic subtype predict EGFR mutations in non-small cell lung cancer. Chest. 2005;128(1): Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361(10): Girard N, Sima CS, Jackman DM, et al. Nomogram to predict the presence of EGFR activating mutation in lung adenocarcinoma. Eur Respir J. 2012;39(2): Sauter JL, Butnor KJ. Clinical and cost implications of universal versus locally advanced-stage and advanced-stage-only molecular testing for epidermal growth factor receptor mutations and anaplastic lymphoma kinase rearrangements in non-small cell lung carcinoma: a tertiary academic institution experience. Arch Pathol Lab Med. 2016;140(4): Arcila ME, Drilon A, Sylvester BE, et al. MAP2K1 (MEK1) mutations define a distinct subset of lung adenocarcinoma associated with smoking. Clin Cancer Res. 2014; 21(8): Li H, Pan Y, Li Y, et al. Frequency of well-identified oncogenic driver mutations in lung adenocarcinoma of smokers varies with histological subtypes and graduated smoking dose. Lung Cancer. 2013;79(1):8 13. Arch Pathol Lab Med Vol 142, February 2018 Detection Rates of EGFR Mutations in Lung Cancer Graham et al 167