[The Value of Circulating Tumor Cells Detected by Chromosome Centromere Probe Identification in Diagnosis of Non-Small Cell Lung Cancer].

Q Shen,L S Shen,Q Chen,J Y Zhou

DOI: https://doi.org/10.3760/cma.j.issn.1001-0939.2018.10.005

2018-01-01

Abstract:Objective: This study explored the value of circulating tumor cells (CTC) detected by chromosome 8 centromere probe in diagnosis of non-small cell lung cancer (NSCLC) as well as correlation between CTC counts and clinical pathological characteristics. Methods: We collected 136 patients with newly diagnosed NSCLC (101 males and 35 females, age ranged from 34 to 79 years), 149 patients with non-malignant pulmonary diseases (103 males and 46 females, age ranged from 24 to 80 years) and 32 healthy volunteers (5 males and 27 females, age ranged from 24 to 42 years). Detection was performed using an epithelial cell adhesion molecule-independent strategy that combined immunocytochemistry staining (ICC) of CD(45) and fluorescence in situ hybridization detection (FISH) with chromosome 8 centromere probe (CEP8). CTC was defined as 4',6-diamidino-2-phenylindole (DAPI) positive, CD(45) negative and CEP8 more than 2 positive points. Quantitative data were reported as x±s and Mann Whitney test was used to compare them. Measurement data were analyzed as contingency tables and Pearson chi-squared test was used. P values of less than 0.05 were considered statistically significant. Results: CTCs were detected in 114 patients (83.8%) with NSCLC, 35 patients (23.5%) with non-malignant pulmonary diseases (P<0.000 1) and 5 volunteers (15.6%). CTC counts in NSCLC patients were (5.98±0.64) per 3.2 ml and (0.60±0.13) per 3.2 ml (P<0.000 1). A receiver operating characteristic curve (ROC) analysis showed similar capability of CTC count, with CEA, in discriminating NSCLC and non-malignant diseases with an area under ROC curve of 0.854 (95% confidence interval: 0.808-0.900, P<0.001). Cutoff of 2 circulating tumor cells per 3.2 ml peripheral blood gave the highest Youden Index of 0.614 in diagnosis of NSCLC with sensitivity of 72.1% (98/136) and specificity of 89.3% (133/149). When patients with CTC≥2/3.2 ml peripheral blood or serum CEA≥5 ng/ml were considered as having NSCLC, sensitivity and specificity of this combined test were 87.5% (119/136) and 85.9% (128/149), with a higher Youden Index of 0.734. No correlation was observed between positive rate of CTC (rate of patients with CTC≥2/3.2 ml peripheral blood) and age, gender, smoking status, pathologic types, and clinical stages. Conclusions: CTC counts detected by CD(45)-FISH is higher in NSCLC patients than in non-malignant disease patients. CTC≥2/3.2 ml peripheral blood is valuable in discriminating NSCLC from nonmalignant diseases, which can be more accurate when combined with CEA.