Somatic mutations in the tyrosine kinase domain of the EGFR gene in lung cancers have generated great interest because they predict better clinical responses to tyrosine kinase inhibitors such as gefitinib (Iressa) and erlotinib (Tarceva).In the IPASS study, patients who had the EGFR mutation had longer progression-free survival with gefitinib treatment compared to those who received platinum-based chemotherapy (9.5 vs 6.3 months, p<0.0001). Patients without the EGFR mutation did better on platinum-based chemotherapy (1.5 vs 5.5 months <0.0001).1In the SATURN:BO18192 study progression-free survival was also longer in EGFR mutation positive patients treated with erlotinib compared to those who received placebo (12\u00b73 vs 11\u00b71 weeks p<0\u00b70001).2 Both drugs are generally less toxic and better tolerated than platinum-based chemotherapies.1,3,4Gefitinib and erlotinib are available in New Zealand and are funded by Pharmac. Gefitinib is funded as first-line treatment for patients with locally advanced, or metastatic, unresectable, non-squamous, NSCLC with the EGFR mutation.Erlotinib is funded as second-line treatment for advanced unresectable NSCLC in patients who have documented disease progression following treatment with first-line platinum-based chemotherapy.Pharmac subsidisation for gefitinib came into effect on 1 August 2012. From 1 January 2014, the funding criteria for erlotinib will be amended such that it will no longer be funded as a second-line treatment option for patients with NSCLC disease known to be negative for the EGFR mutation.5 Patients with advanced, non-squamous, NSCLC will need to undergo testing for the presence of the EGFR mutation in order to access funding for tyrosine kinase inhibitors.The EGFR mutation occurs in 20 to 40% of NSCLCs (Table 1).6-13 Clinicopathological characteristics associated with a higher prevalence of EGFR mutations are adenocarcinoma histology, South East Asian origin, female gender, and a history of never smoking.EGFR mutation testing has recently been introduced at CMDHB but the most cost-effective approach to selecting patients for testing is uncertain. The cost of testing in the Auckland Region is approximately $500, through LabPLUS at Auckland City Hospital. Often the amount of diagnostic tissue obtained from samples such as fine needle aspirates, bronchial washings and pleural fluid is insufficient for EGFR testing.In order to obtain sufficient tissue samples, better tissue sampling, for example with a CT-guided core biopsy, or repeat biopsy may be necessary. It is not currently known how often initial diagnostic procedures provide adequate tissue for EGFR testing.The aim of this study is to clarify the number of patients from CMDHB with non-squamous, NSCLC, and particularly those with advanced disease, who may be eligible for EGFR mutation testing.Methods We undertook a retrospective, observational study of all patients with lung cancer diagnosed at CMDHB between 01/07/2011 and 30/06/2012. Patients included were inpatients or outpatients with primary cancer of the trachea, bronchus or lung, or cancer suspicious of being primary cancer of the trachea, bronchus or lung. They were identified by: a search of the Delphic AP, CMDHB Pathology Laboratory, electronic database for relevant histology specimens and the Diagnostic Medlab Latte electronic database for relevant cytology specimens; review of the CMDHB Respiratory Cancer Nurses dataset of cases investigated for lung cancer; review of the CMDHB Respiratory Secretarys dataset of cases presented at the Thoracic Multidisciplinary Meeting (between CMDHB Respiratory Services and Cardiothoracic Surgical, Medical Oncology and Radiation Oncology Services at Auckland City Hospital); and review of cases identified by CMDHB Decision Support Services using relevant ICD-10 discharge and outpatient clinic codes. Cases were included if they fell within the 2004 WHO classification of invasive malignant epithelial lung tumours, which includes: squamous cell carcinoma, small cell carcinoma, adenocarcinoma, large cell carcinoma, adenosquamous carcinoma, sarcomatoid carcinoma, carcinoid tumour, or salivary gland tumour. Cases were included if their diagnostic specimen was collected within the study period. If they did not have a histological diagnosis they were included if their diagnosing CT scan fell within the study period. Patients referred for surgery without a histological diagnosis prior to surgery were included if their diagnosing CT scan fell within the study period. Patients with recurrence of lung cancer were included if they had been previously treated with curative intent. Patients with recurrence were excluded if they previously had radiotherapy or chemotherapy for palliation, without hoping to achieve a cure. The following data were collected: National Health Index (NHI) number, histological diagnosis, sample type and collection date, CT date, tumour node metastasis (TNM) staging system for NSCLC (7th edition), date of birth, gender, ethnicity, smoking history and ECOG performance status score at diagnosis. The ECOG performance status score is a score from zero to 5. Zero is asymptomatic and fully active; one is able to carry out work of a light or sedentary nature; 2 is ambulatory and unable to carry out any work activities but up and about more than 50% of waking hours; 3 is confined to bed or chair 50% or more of waking hours, not bed bound but capable of only limited self cares; 4 is bed-bound and unable to do any self-care; and 5 is death.14 Cases assessed by the managing clinician with a score between two values were categorised with the lower score e.g. a score between 1 and 2 were coded as an ECOG performance score of 1. Results 206 patients diagnosed at CMDHB with primary lung cancer were identified. 141 (68.4%) cases did not have squamous cell or small cell carcinoma histologies (see Table 2). Table 2. Histological diagnosis Histological diagnosis Number % Adenocarcinoma 87 42.2 Squamous cell carcinoma 52 25.3 Small cell carcinoma 13 6.3 Carcinoid tumour 4 1.9 Large cell carcinoma 3 1.5 Adenosquamous carcinoma 1 0.5 NSCLC not further differentiated 11 5.3 Poorly differentiated carcinomas 3 1.5 No histological diagnosis 32 15.5 Total 206 100% Of the 141 cases the mean (\u00b1SD) age was 69 \u00b113 years. 78 (55.3%) were European, 26 (18.4%) were Pacific Islanders, 22 (15.6%) were M\u0101ori, 15 (10.7%) were Asian, with nine of South East Asian origin. 68 (48.2%) patients were female. 28 (19.9%) cases had never smoked (Table 3). Table 3. Histological diagnosis by smoking history Diagnosis Never smoked Ex-smoker Current smoker Total Adenocarcinoma 22 43 22 87 Carcinoid tumour 1 1 2 4 Large cell carcinoma 0 0 3 3 Adenosquamous carcinoma 0 1 0 1 NSCLC not further differentiated 2 6 3 11 Poorly differentiated carcinomas 0 2 1 3 No histological diagnosis 3 17 12 32 Total (%) 28 (19.9%) 70 (49.6%) 43 (30.5%) 141 Advanced cases with unresectable malignancies (stage IIIA, IIIB or IV), accounted for 103 (73.0%) of the 141 cases. There were 38 (27%) early stage cases (stage IA, IB, IIA or IIB). Most of the 141 cases had a good performance status, with 79.4% having an ECOG performance score of 2 or less. 54 (38.3%) cases had a score of zero, 37 (26.2%) had a score of 1, 21 (14.9 %) had a score of 2, 19 (13.5%) had a score of 3 and 10 (7.1%) had a score of 4. No cases scored 5, as they were all alive at the time of diagnosis. Of the 103 non-squamous, NSCLC cases with unresectable malignancies, 7 died before a management plan could be made and 7 declined further investigation and treatment. One patient was deemed unfit for investigation, leaving 88 patients who may be eligible for testing. Of the 103 cases, 78 (75.6%) had an ECOG score of 2 or less. Of this 78, 5 died before a management plan could be made and 5 declined further investigation and treatment. Of this remaining 68, 53 had adenocarcinoma, 1 had a carcinoid tumour, 3 had large cell carcinoma, 1 had adenosquamous carcinoma, 5 had NSCLC not further differentiated, 1 had poorly differentiated carcinoma, and 3 had no histological diagnosis. Of the 53 adenocarcinoma cases, 3 were assumed to be adenocarcinoma as they had recurrence of a previously proven adenocarcinoma. Of the remaining 50 patients the diagnostic specimens used to confirm the diagnosis of adenocarcinoma are shown in Table 4. The cases had a mean (\u00b1SD) age of 69\u00b111 years. Table 4. Diagnostic specimens for patients with adenocarcinoma Diagnostic specimen n Brain biopsy 1 Bronchial wash 2 Bronchial biopsy 10 CT FNA 5 Transbronchial FNA 12 Liver biopsy 4 Pleural biopsy 3 CT core biopsy 5 *Lymph node FNA 3 Pleural fluid 4 Sputum cytology 1 Total 50 *Axilla or neck. There were 5 South East Asian female patients with adenocarcinoma who had never smoked. Four of these patients had stage IV disease and 1 had stage IB disease. They all had an ECOG performance score of less than 2. Discussion In a 1-year cohort of 206 primary lung cancer patients, 141 (68%) patients had non-squamous NSCLC and were potentially eligible for EGFR mutation testing. Of these 141 patients, 103 (73%) had advanced stage lung cancer; these patients comprised 50% of all lung cancer patients. 15 patients died, declined investigation and treatment, or were deemed unfit for investigation, leaving 88 cases (43% of all lung cancer patients) with advanced lung cancer who may be eligible for EGFR mutation testing. No one dataset search resulted in a complete list of all lung cancer cases diagnosed at CMDHB during the study period. Each dataset contained a few cases not found elsewhere. Not all cases (175 of 206) were discussed at the Thoracic Multidisciplinary Meeting. The use of multiple datasets, however, ensured that we captured most of the cases diagnosed in the study period. Smoking status was recorded variably and came from multiple sources including discharge summaries, clinic letters, thoracic multidisciplinary meeting report forms, or written clinical records. Pack year history and how long an ex-smoker had quit for was poorly documented in the clinical records and were therefore not reported in this study. An ECOG performance score of 2 or less was chosen as a pragmatic way of identifying patients who were more likely to be offered treatment. There may, however, be some patients with an ECOG score of 3 who could be offered tyrosine kinase inhibitor treatment. Of the 19 patients with non-squamous NSCLC with an ECOG score of three, 16 were of advanced stage. Of these 16, 1 died before they could be offered treatment and 2 declined investigation and treatment. The EGFR mutation does not occur in small cell carcinomas and almost all squamous cell carcinomas. They are therefore excluded from Pharmacs funding criteria. Adenocarcinoma is associated with a higher prevalence of the EGFR mutation as shown in Table 1. The occurrence of other lung tumours (large cell, adenosquamous, and sarcomatoid carcinomas; carcinoid and salivary gland tumours of lung origin) is uncommon and the prevalence of the EGFR mutation in these tumours is uncertain..6,9,10,12 Rosell, however, reported that 33 of 287 large cell carcinoma samples tested positive for the EGFR mutation.11Targeting non-squamous NSCLC histologies for EGFR mutation testing appears to be appropriate. The selection of patient subgroups for EGFR mutation testing requires further clarification. Patients with adenocarcinoma who are female and never smokers should be tested. However, patients who are male, or are smokers or ex-smokers, may have the EGFR mutations and should not be excluded from testing. Targeting by ethnicity is not ideal as most studies have only been done in South East Asians or Europeans and the prevalence in Maori, Pacific Islanders and Indians is unknown. Of the patients who could potentially benefit from EGFR mutation testing, 50 were adenocarcinoma cases. The diagnostic tissue from many of these patients was from fluid or fine needle aspirate samples, which are

Somatic mutations in the tyrosine kinase domain of the EGFR gene in lung cancers have generated great interest because they predict better clinical responses to tyrosine kinase inhibitors such as gefitinib (Iressa) and erlotinib (Tarceva).In the IPASS study, patients who had the EGFR mutation had longer progression-free survival with gefitinib treatment compared to those who received platinum-based chemotherapy (9.5 vs 6.3 months, p<0.0001). Patients without the EGFR mutation did better on platinum-based chemotherapy (1.5 vs 5.5 months <0.0001).1In the SATURN:BO18192 study progression-free survival was also longer in EGFR mutation positive patients treated with erlotinib compared to those who received placebo (12\u00b73 vs 11\u00b71 weeks p<0\u00b70001).2 Both drugs are generally less toxic and better tolerated than platinum-based chemotherapies.1,3,4Gefitinib and erlotinib are available in New Zealand and are funded by Pharmac. Gefitinib is funded as first-line treatment for patients with locally advanced, or metastatic, unresectable, non-squamous, NSCLC with the EGFR mutation.Erlotinib is funded as second-line treatment for advanced unresectable NSCLC in patients who have documented disease progression following treatment with first-line platinum-based chemotherapy.Pharmac subsidisation for gefitinib came into effect on 1 August 2012. From 1 January 2014, the funding criteria for erlotinib will be amended such that it will no longer be funded as a second-line treatment option for patients with NSCLC disease known to be negative for the EGFR mutation.5 Patients with advanced, non-squamous, NSCLC will need to undergo testing for the presence of the EGFR mutation in order to access funding for tyrosine kinase inhibitors.The EGFR mutation occurs in 20 to 40% of NSCLCs (Table 1).6-13 Clinicopathological characteristics associated with a higher prevalence of EGFR mutations are adenocarcinoma histology, South East Asian origin, female gender, and a history of never smoking.EGFR mutation testing has recently been introduced at CMDHB but the most cost-effective approach to selecting patients for testing is uncertain. The cost of testing in the Auckland Region is approximately $500, through LabPLUS at Auckland City Hospital. Often the amount of diagnostic tissue obtained from samples such as fine needle aspirates, bronchial washings and pleural fluid is insufficient for EGFR testing.In order to obtain sufficient tissue samples, better tissue sampling, for example with a CT-guided core biopsy, or repeat biopsy may be necessary. It is not currently known how often initial diagnostic procedures provide adequate tissue for EGFR testing.The aim of this study is to clarify the number of patients from CMDHB with non-squamous, NSCLC, and particularly those with advanced disease, who may be eligible for EGFR mutation testing.Methods We undertook a retrospective, observational study of all patients with lung cancer diagnosed at CMDHB between 01/07/2011 and 30/06/2012. Patients included were inpatients or outpatients with primary cancer of the trachea, bronchus or lung, or cancer suspicious of being primary cancer of the trachea, bronchus or lung. They were identified by: a search of the Delphic AP, CMDHB Pathology Laboratory, electronic database for relevant histology specimens and the Diagnostic Medlab Latte electronic database for relevant cytology specimens; review of the CMDHB Respiratory Cancer Nurses dataset of cases investigated for lung cancer; review of the CMDHB Respiratory Secretarys dataset of cases presented at the Thoracic Multidisciplinary Meeting (between CMDHB Respiratory Services and Cardiothoracic Surgical, Medical Oncology and Radiation Oncology Services at Auckland City Hospital); and review of cases identified by CMDHB Decision Support Services using relevant ICD-10 discharge and outpatient clinic codes. Cases were included if they fell within the 2004 WHO classification of invasive malignant epithelial lung tumours, which includes: squamous cell carcinoma, small cell carcinoma, adenocarcinoma, large cell carcinoma, adenosquamous carcinoma, sarcomatoid carcinoma, carcinoid tumour, or salivary gland tumour. Cases were included if their diagnostic specimen was collected within the study period. If they did not have a histological diagnosis they were included if their diagnosing CT scan fell within the study period. Patients referred for surgery without a histological diagnosis prior to surgery were included if their diagnosing CT scan fell within the study period. Patients with recurrence of lung cancer were included if they had been previously treated with curative intent. Patients with recurrence were excluded if they previously had radiotherapy or chemotherapy for palliation, without hoping to achieve a cure. The following data were collected: National Health Index (NHI) number, histological diagnosis, sample type and collection date, CT date, tumour node metastasis (TNM) staging system for NSCLC (7th edition), date of birth, gender, ethnicity, smoking history and ECOG performance status score at diagnosis. The ECOG performance status score is a score from zero to 5. Zero is asymptomatic and fully active; one is able to carry out work of a light or sedentary nature; 2 is ambulatory and unable to carry out any work activities but up and about more than 50% of waking hours; 3 is confined to bed or chair 50% or more of waking hours, not bed bound but capable of only limited self cares; 4 is bed-bound and unable to do any self-care; and 5 is death.14 Cases assessed by the managing clinician with a score between two values were categorised with the lower score e.g. a score between 1 and 2 were coded as an ECOG performance score of 1. Results 206 patients diagnosed at CMDHB with primary lung cancer were identified. 141 (68.4%) cases did not have squamous cell or small cell carcinoma histologies (see Table 2). Table 2. Histological diagnosis Histological diagnosis Number % Adenocarcinoma 87 42.2 Squamous cell carcinoma 52 25.3 Small cell carcinoma 13 6.3 Carcinoid tumour 4 1.9 Large cell carcinoma 3 1.5 Adenosquamous carcinoma 1 0.5 NSCLC not further differentiated 11 5.3 Poorly differentiated carcinomas 3 1.5 No histological diagnosis 32 15.5 Total 206 100% Of the 141 cases the mean (\u00b1SD) age was 69 \u00b113 years. 78 (55.3%) were European, 26 (18.4%) were Pacific Islanders, 22 (15.6%) were M\u0101ori, 15 (10.7%) were Asian, with nine of South East Asian origin. 68 (48.2%) patients were female. 28 (19.9%) cases had never smoked (Table 3). Table 3. Histological diagnosis by smoking history Diagnosis Never smoked Ex-smoker Current smoker Total Adenocarcinoma 22 43 22 87 Carcinoid tumour 1 1 2 4 Large cell carcinoma 0 0 3 3 Adenosquamous carcinoma 0 1 0 1 NSCLC not further differentiated 2 6 3 11 Poorly differentiated carcinomas 0 2 1 3 No histological diagnosis 3 17 12 32 Total (%) 28 (19.9%) 70 (49.6%) 43 (30.5%) 141 Advanced cases with unresectable malignancies (stage IIIA, IIIB or IV), accounted for 103 (73.0%) of the 141 cases. There were 38 (27%) early stage cases (stage IA, IB, IIA or IIB). Most of the 141 cases had a good performance status, with 79.4% having an ECOG performance score of 2 or less. 54 (38.3%) cases had a score of zero, 37 (26.2%) had a score of 1, 21 (14.9 %) had a score of 2, 19 (13.5%) had a score of 3 and 10 (7.1%) had a score of 4. No cases scored 5, as they were all alive at the time of diagnosis. Of the 103 non-squamous, NSCLC cases with unresectable malignancies, 7 died before a management plan could be made and 7 declined further investigation and treatment. One patient was deemed unfit for investigation, leaving 88 patients who may be eligible for testing. Of the 103 cases, 78 (75.6%) had an ECOG score of 2 or less. Of this 78, 5 died before a management plan could be made and 5 declined further investigation and treatment. Of this remaining 68, 53 had adenocarcinoma, 1 had a carcinoid tumour, 3 had large cell carcinoma, 1 had adenosquamous carcinoma, 5 had NSCLC not further differentiated, 1 had poorly differentiated carcinoma, and 3 had no histological diagnosis. Of the 53 adenocarcinoma cases, 3 were assumed to be adenocarcinoma as they had recurrence of a previously proven adenocarcinoma. Of the remaining 50 patients the diagnostic specimens used to confirm the diagnosis of adenocarcinoma are shown in Table 4. The cases had a mean (\u00b1SD) age of 69\u00b111 years. Table 4. Diagnostic specimens for patients with adenocarcinoma Diagnostic specimen n Brain biopsy 1 Bronchial wash 2 Bronchial biopsy 10 CT FNA 5 Transbronchial FNA 12 Liver biopsy 4 Pleural biopsy 3 CT core biopsy 5 *Lymph node FNA 3 Pleural fluid 4 Sputum cytology 1 Total 50 *Axilla or neck. There were 5 South East Asian female patients with adenocarcinoma who had never smoked. Four of these patients had stage IV disease and 1 had stage IB disease. They all had an ECOG performance score of less than 2. Discussion In a 1-year cohort of 206 primary lung cancer patients, 141 (68%) patients had non-squamous NSCLC and were potentially eligible for EGFR mutation testing. Of these 141 patients, 103 (73%) had advanced stage lung cancer; these patients comprised 50% of all lung cancer patients. 15 patients died, declined investigation and treatment, or were deemed unfit for investigation, leaving 88 cases (43% of all lung cancer patients) with advanced lung cancer who may be eligible for EGFR mutation testing. No one dataset search resulted in a complete list of all lung cancer cases diagnosed at CMDHB during the study period. Each dataset contained a few cases not found elsewhere. Not all cases (175 of 206) were discussed at the Thoracic Multidisciplinary Meeting. The use of multiple datasets, however, ensured that we captured most of the cases diagnosed in the study period. Smoking status was recorded variably and came from multiple sources including discharge summaries, clinic letters, thoracic multidisciplinary meeting report forms, or written clinical records. Pack year history and how long an ex-smoker had quit for was poorly documented in the clinical records and were therefore not reported in this study. An ECOG performance score of 2 or less was chosen as a pragmatic way of identifying patients who were more likely to be offered treatment. There may, however, be some patients with an ECOG score of 3 who could be offered tyrosine kinase inhibitor treatment. Of the 19 patients with non-squamous NSCLC with an ECOG score of three, 16 were of advanced stage. Of these 16, 1 died before they could be offered treatment and 2 declined investigation and treatment. The EGFR mutation does not occur in small cell carcinomas and almost all squamous cell carcinomas. They are therefore excluded from Pharmacs funding criteria. Adenocarcinoma is associated with a higher prevalence of the EGFR mutation as shown in Table 1. The occurrence of other lung tumours (large cell, adenosquamous, and sarcomatoid carcinomas; carcinoid and salivary gland tumours of lung origin) is uncommon and the prevalence of the EGFR mutation in these tumours is uncertain..6,9,10,12 Rosell, however, reported that 33 of 287 large cell carcinoma samples tested positive for the EGFR mutation.11Targeting non-squamous NSCLC histologies for EGFR mutation testing appears to be appropriate. The selection of patient subgroups for EGFR mutation testing requires further clarification. Patients with adenocarcinoma who are female and never smokers should be tested. However, patients who are male, or are smokers or ex-smokers, may have the EGFR mutations and should not be excluded from testing. Targeting by ethnicity is not ideal as most studies have only been done in South East Asians or Europeans and the prevalence in Maori, Pacific Islanders and Indians is unknown. Of the patients who could potentially benefit from EGFR mutation testing, 50 were adenocarcinoma cases. The diagnostic tissue from many of these patients was from fluid or fine needle aspirate samples, which are