Multicenter Validation Study to Implement Plasma Epidermal Growth Factor Receptor T790M Testing in Clinical Laboratories.

PURPOSE: Plasma detection of EGFR T790M mutations is an emerging alternative to tumor rebiopsy in acquired epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor resistance. Validation of analytical sensitivity and clinical utility is required before routine diagnostic use in clinical laboratories. PATIENTS AND METHODS: Sixty-three patients with advanced EGFR-mutant lung cancer at 7 Canadian centers, who were being screened for the ASTRIS trial (ClinicalTrials.gov identifier: NCT02474355), participated in this companion study. Plasma T790M mutation was detected using droplet digital polymerase chain reaction, Cobas (Roche Diagnostics, Indianapolis, IN), or next-generation sequencing in 4 laboratories. T790M concordance was assessed between plasma and tumor samples. RESULTS: Assessment of T790M in tumor biopsy tissue was successful in 81% of patients; 49% had confirmed T790M results (tumor or plasma) for ASTRIS. Plasma testing in this companion study yielded T790M results in 97% of patients; 62% had T790M-positive results, 36% had negative results, and 2% had indeterminate results. Of 38 patients with negative or indeterminate biopsy results, 55% had positive plasma T790M results, increasing the proportion with T790M-positive results to 73%. Sensitivity of plasma T790M testing was 75%. Overall concordance between tissue and plasma was 64%, and concordance among laboratories was 90.3%. Response to osimertinib and duration of therapy were similar irrespective of testing method (overall response rate, 62.5% for tissue, 66.7% for plasma, and 70.6% for both). CONCLUSION: This multicenter validation study demonstrates that plasma EGFR T790M testing can identify significantly more patients than biopsy alone who may benefit from targeted therapy.

Impact of Asian ethnicity on outcome in metastatic EGFR-mutant non-small cell lung cancer.

AIM: To determine factors associated with survival in de novo stage IV, non-small cell lung cancer (NSCLC) patients possessing epidermal growth factor receptor mutations (EGFRmut+ ) receiving tyrosine kinase inhibitors (TKI) in the first-line setting. METHODS: The Glans-Look Lung Cancer Database was used to retrospectively review stage IV EGFRmut+ NSCLC patients diagnosed 2010-2016 receiving first-line TKI. Patients with overall survival times in the upper quartile (≥34 months) were designated "long-term survivors" (LTS), the remaining deemed "average-term survivors" and characteristics between these groups were compared in univariate analysis, and multivariable models constructed to determine predictors of outcome. RESULTS: Of 170 eligible patients, median overall survival was 21 months. LTS were significantly more likely to be of Asian ethnicity, be never-smokers and not possess brain or bone metastases at diagnosis. Asian and non-Asian patients were comparable, save for an increased propensity of Asian patients to be never smokers and have normal-range BMI. Multivariable analysis revealed Asian ethnicity [hazard ratio (HR) = 0.65; P = 0.016] and never-smoking history (HR = 0.65; P = 0.034) as indicators of improved outcome, and presence of brain metastasis at diagnosis an indicator of poor outcome (HR = 2.21; P < 0.001). CONCLUSIONS: Analysis of this population-based cohort identifies never-smoking history and absence of brain metastasis along with Asian ethnicity as an independent prognosticators of favorable outcome, and reveals Asian patients to be clinicopathologically similar to non-Asian patients. These findings suggest Asian patients represent a unique subpopulation within EGFRmut+ NSCLC who may possess different biological underpinnings of NSCLC.

Factors associated with early mortality in non-small cell lung cancer patients following systemic anti-cancer therapy: A 10 year population-based study.

OBJECTIVES: To investigate how clinical, demographic and treatment-related factors in non-small cell lung cancer (NSCLC) patients impact the risk of mortality in the 30 days following receipt of systemic anti-cancer therapies (SACT), and undertake a comprehensive review of the treatment decisions and experiences of a real-world population. MATERIALS AND METHODS: We reviewed NSCLC patients receiving SACT from 2005 to 2014, and captured in the Glans-Look Lung Cancer Database, which contains demographic, clinical, pathological, treatment and outcome data. The 30-day post-SACT mortality rate was calculated, and regimen changes in the last 14 days of life were identified. Univariate analysis and multivariate logistic regression were used to identify demographic, tumor and treatment-related factors that correlated with mortality risk. RESULTS: 1044 patients receiving ≥ 1 cycle of SACT in 2005-2014 were identified. 233 (22.3%) deaths occurred ≤ 30 days following SACT receipt; 32 (13.7%) of which had new SACT regimens ≤ 14 days prior to death. Risk of 30-day mortality and regimen changes at the end of life increased in association with being male [OR: 1.48 (1.12-1.95), p = 0.005], advanced disease at diagnosis [OR: 1.85 (1.19-2.88), p = 0.006], palliative-intent treatment [OR: 6.75 (3.88-11.77), p < 0.001], and use of EGFR-targeting agents [OR: 4.5 (3.27-6.18) p < 0.001]. Risk of early mortality decreased for never-smokers [OR: 0.62 (0.41-0.95), p = 0.028], and those receiving SACT in more recent years (2010-2014) [OR: 0.65 (0.49-0.86), p = 0.002]. CONCLUSION: Our findings identified several factors that affected the risk of early mortality in NSCLC patients following SACT. These results from a representative population provide insights regarding the benefits and risks of SACT and can serve to inform clinical and palliative best practices.

Comparison of Clinical Characteristics and Outcomes in Relapsed Versus De Novo Metastatic Non-Small Cell Lung Cancer.

OBJECTIVES: To compare the clinical characteristics and outcomes between relapsed and de novo metastatic non-small cell lung cancer (NSCLC). MATERIALS AND METHODS: We reviewed all NSCLC diagnoses between January 1999 and December 2013 in the institutional Glans-Look Lung Cancer Database, which contains demographic, clinical, pathologic, treatment, and outcome information. Patients with distant metastasis at diagnosis (American Joint Committee on Cancer [AJCC] eighth edition, stage IV), the "de novo" cohort, were compared with the "relapsed" cohort, consisting of patients diagnosed with early stage disease (stage I/II) undergoing curative intent treatment and subsequently experiencing metastatic relapse. Survival analysis, along with univariate and multivariable analysis was performed. RESULTS: A total of 185 relapsed and 3039 de novo patients were identified. Significantly different patterns of smoking history, histology, systemic therapy use, and disease extent were observed between the relapsed and de novo cohorts. Median overall survival from time of metastasis was significantly longer in relapsed than in de novo disease (8.9 vs. 3.7 mo, P<0.001). Relapsed patients demonstrated significant improvements in outcomes over time. In multivariate analysis, de novo metastatic disease continued to bode a worse prognosis (adjusted hazard ratio [HR], 1.4) as did male sex (HR, 1.2), never-smoking history (HR, 1.2), and presence of extrapulmonary metastases (HR, 1.3). Systemic therapy receipt conferred better outcome (HR, 0.4), although the impact of relapsed versus de novo disease on outcomes persisted regardless of systemic therapy receipt. CONCLUSIONS: Relapsed and de novo patients represent significantly different subpopulations within metastatic NSCLC with the latter exhibiting poorer survival. This information facilitates discussions about prognosis with patients and supports screening initiatives aimed at reducing de novo disease.

Impact of number versus location of metastases on survival in stage IV M1b non-small cell lung cancer.

BACKGROUND: To assess the impact of location versus number of extra-pulmonary metastatic sites (EPMS) on survival in stage IV non-small cell lung cancer (NSCLC). METHODS AND MATERIALS: Retrospective analysis was conducted on patients diagnosed during 1999-2013 with stage IV, M1b (AJCC 7th edition) NSCLC using the large, institutional Glans-Look Database, which contains patient demographic, clinical, pathological, treatment, and outcome information. We assessed the impact of location and number of EPMS and identified correlates of overall survival using the Kaplan-Meier method and Cox regression. RESULTS: We identified a total of 2065 NSCLC patients with EPMS. Median age was 67 (IQR 58-75) years, 52% were men, and 78% were current or former smokers. 60% had one EPMS, and 40% had two or more EPMS. Among those with only one EPMS, most frequent organ involvement included bone (40%), brain (32%), and liver (13%). Median overall survival (mOS) was worst in those with liver metastasis and best in those with adrenal metastasis (2.0 vs. 5.2 months, p = 0.015). However, outcomes based on site of organ involvement were not significantly different in multivariable analysis. Compared to patients with one EPMS, individuals with two or more EPMS experienced worse outcomes (mOS ≤ 2.9 vs. 3.9 months, p < 0.001), and were associated with worse prognosis in Cox regression analysis (HR 1.5, 95% CI 1.3-1.7, p < 0.001). CONCLUSIONS: Number rather than location of EPMS is a prognostic factor in patients with stage IV M1b NSCLC. This information is relevant for accurate prognostication, stratification of participants in future clinical trials, and timely and appropriate advanced care planning.