C. Pepe

Prevalence of obstructive sleep apnea–hypopnea in severe versus moderate asthma

BACKGROUND Previous studies have suggested a link between obstructive sleep apnea and poor asthma control, which may be mediated through airway inflammation, obesity, and other mechanisms. OBJECTIVE To test the hypothesis that the prevalence and severity of sleep apnea is greater among patients with severe compared with moderate asthma and controls without asthma. METHODS Complete overnight home polysomnography was performed in 26 patients with severe asthma consecutively recruited to a difficult asthma program, 26 patients with moderate asthma, and 26 controls without asthma of similar age and body mass index. Flow rates and Juniper asthma control and quality of life questionnaires were also obtained. RESULTS Obstructive sleep apnea-hypopnea, defined by an Apnea-Hypopnea Index > or = 15 events/h of sleep scored using Chicago criteria, was present in 23 of 26 (88%) patients with severe asthma, 15 of 26 (58%) patients with moderate asthma, and 8 of 26 (31%) controls without asthma (chi(2): P < .001). Using the more restrictive scoring criteria applied in the Wisconsin cohort study, Apnea-Hypopnea Index > or = 5/h was present in 50% (severe), 23% (moderate), and 12% (control) of subjects (P = .007). Mean nocturnal arterial oxygen saturation was significantly lower in patients with severe asthma versus controls, and apnea-hypopnea severity measures were significantly worse for both asthmatic groups compared with controls. Among subjects with asthma, no significant correlations were identified between the severity of sleep-disordered breathing and asthma severity or control measures (FEV(1), Juniper scores). CONCLUSIONS Obstructive sleep apnea-hypopnea was significantly more prevalent among patients with severe compared with moderate asthma, and more prevalent for both asthma groups than controls without asthma. These observations suggest potential pathophysiologic interactions between obstructive sleep apnea-hypopnea and asthma severity and control.

Are frailty markers useful for predicting treatment toxicity and mortality in older newly diagnosed cancer patients? Results from a prospective pilot study

INTRODUCTION The concept of frailty may be useful to characterize vulnerability. The aim of this pilot study was to explore the association between frailty/functional status and treatment toxicity at 3 months and mortality at 6 months. METHODS Patients aged ≥65 years referred to the Jewish General Hospital, Montreal, with a new cancer diagnosis. Seven frailty markers and 4 functional status measures were examined. Logistic regression was used to examine the association between frailty/functional status and toxicity, and Cox models for time to death. RESULTS 112 participated, median age 74.1, 31 had toxicity and 15 died. At baseline, 88% had ≥1 frailty marker. Low grip strength predicted toxicity (OR 8.47, 95%CI: 1.3-53.6), ECOG performance status and ADL disability predicted time to death. CONCLUSION The majority had ≥1 frailty marker. Low grip strength predicted toxicity, none of the functional measures did. Further researcher investigating the usefulness of frailty markers is needed.

Comparison of the Yield of Different Diagnostic Procedures for Cellular Differentiation and Genetic Profiling of Non–Small-Cell Lung Cancer

Introduction: As treatments for non–small-cell lung cancer (NSCLC) become personalized, cellular and molecular differentiation of the tumor is becoming the standard of care. Our objective is to compare the yield of different diagnostic procedures for cellular differentiation of NSCLC and analysis of epidermal growth factor receptor (EGFR) mutation. Methods: We evaluated all patients diagnosed with NSCLC from January 2004 to September 2010 at the Jewish General Hospital, Montreal. Diagnostic procedures included surgical biopsies, nonsurgical histologic biopsies (endobronchial and core needle), transbronchial needle aspirate (TBNA) and transthoracic needle aspirate (TTNA), bronchoalveolar lavage (BAL), and pleural fluid samples. Results: We included 702 subjects investigated for histopathologic differentiation of NSCLC. Of these, 269 were also investigated for EGFR mutation. Failure to ascertain the cellular subtype and EGFR mutation status was least likely with surgical specimens (0% and 1.8%, respectively); followed by TTNA (14% and 10%, respectively) and histologic biopsy (18% for both); and was frequent with TBNA (39% and 30%, respectively). Although BAL and pleural fluid specimens provided reasonable yield for cellular differentiation (20 % and 11%, respectively), their results were not accurate in 6% of their samples when compared with concurrent or subsequent surgical specimens (reference standard) performed in a subgroup of patients. Conclusion: Radiologically guided TTNA and histologic biopsies provided high yield for both molecular and histologic analyses. The yield of unguided TBNA was relatively low. Further studies are needed to assess the adequacy of BAL and pleural fluid samples for EGFR mutation analysis and accurate characterization of cellular subtypes of NSCLC.

Epidermal growth factor receptor (EGFR) mutations detected by denaturing high performance liquid chromatography (dHPLC) in non-small cell lung cancer (NSCLC): Correlation with disease stage and response to tyrosine kinase inhibitors (TKI) therapy

22118 Background: Somatic mutations of the EGFR gene predict sensitivity to EGFR-TKI in NSCLC patients. dHPLC is a sensitive method for detecting EGFR mutations in NSCLC tumours. The goal of this study was to investigate the relationship between EGFR status (mutation and type) and disease stage and response to TKI therapy. Methods: All patients (pts) diagnosed with NSCLC in 2005 were identified and genotyped for EGFR exons 19 and 21 mutations using dHPLC. Results were correlated with the stage of disease. In addition, all genotyped tumours were correlated with patient demographics, smoking status, tumor histology, disease stage, response to TKI, and overall survival. Results: 144 pts were diagnosed with NSCLC at our institution in 2005. 62 pts were genotyped, 82 are still pending. Our preliminary data show 34% (10/29) of pts with advanced disease (stage 3B/4) were mutation positive compared with 6% (2/33) of early stage disease (p=0.007). Since 2004 we have genotyped a total of 228 NSCLC pts. Using dHPLC,...