Verification Of MicroRNA Mathematical Model To Identify Three Subtypes Of Lung Cancer In Bronchial Brushing Samples

Objective:To verify the diagnosis of lung cancer histologic subtype regression models established in the three histological subtypes of lung cancer brushing sampleS。Methods:Collection of Zhongshan Hospital affiliated to Fudan University from February to December2011, accepted bronchoscope brush biopsy and pathological diagnosis via biopsy or surgical lung brush biopsy specimens of207cases、Adenocarcinoma85cases、53cases of squamous cell carcinoma、69cases of Small cell lung cancer、The total RNA of brushing cells was extracted by Trizol method。MicroRNA diagnosis was performed on bronchial brushing samples of patients with lung cancer by real time quantitative RT-PCR method, then compared with pathological diagnosis。Result:1The differential expression of miR-375in the three types of lung cancer is significantly, P-value is far less than0.05。Differential expression of miR-205of SCLC/SQ and SQ/AC is obviously, While no significant difference between in the expression of SCLC/AC; miR-29a only express significant differences between the SQ/AC; the expression of miR-34b is no significant difference between three types of lung cancer. There are not high sensitivity and specificity in the identification of the three tumor types by a single microRNA; only the microRNA-375is up to90%in sensitivity of SCLC/SQ and SCLC/AC, Specificity failed to meet the90%wholly, AUC values with one exception, were less than0.9。 MicroRNA-205and microRNA-375differences are significant between SQ and SCLC, in addition to microRNA-375, Specificity and sensitivity of other microRNAs are not high and AUC values is less than0.9. In addition to microRNA-34a,the rests were expressed significantly different but the sensitivity and specificity is not high between SQ and AC, AUC values were less than0.9. Only microRNA-375significant differences between SCLC和AC. Specificity and sensitivity is not high and AUC values is less than0.9。2.The sensitivity and specificity of modal A were79%and94%in discriminating SCLC and NSCLC, the AUC value is0.947(0.907-0.973); The sensitivity and specificity of modal B were92%and88%in discriminating SCLC and NSCLC, the AUC value is0.962(0.919-0.986)。Conclusion:1.The diagnostic value of a single microRNA in discriminating lung cancer subtypes is limited, expression differences in lung cancer subtypes varies and the capability of discriminating is not high。2The two microRNA model has a high diagnostic value in brushings cells in discriminating three lung cancer subtypes, the AUC were about0.95and Sensitivity and specificity both are more than90%。...