DNA image cytometry test for primary screening of esophageal cancer: a population-based multi-center study in high-risk areas in China

Abstract

Objective: To evaluate the feasibility of DNA image cytometry (DNA-ICM) as a primary screening method for esophageal squamous cell cancer (ESCC).

Methods: A total of 5,382 local residents aged 40-69 years from three high-risk areas in China (Linzhou in Henan province, Feicheng in Shandong province and Cixian in Hebei province) from 2008 to 2011 were recruited in this population-based screening study. And 2,526 subjects declined to receive endoscopic biopsy examination with Lugol's iodine staining, while 9 and 815 subjects were excluded from liquid-based cytology and DNA-ICM test respectively due to slide quality. Finally, 2,856, 5,373 and 4,567 subjects were enrolled in the analysis for endoscopic biopsy examination, liquid-based cytology and DNA-ICM test, respectively. Sensitivity (SE), specificity (SP), negative predictive values (NPV) and positive predictive values (PPV) as well as their 95% confidence intervals (95% CI) for DNA-ICM, liquid-based cytology and the combination of the two methods were calculated. Receiver operating characteristic (ROC) curves were applied to determine the cutoff point of DNA-ICM for esophageal cancer.

Results: DNA-ICM results were significantly correlative with esophageal cancer and precancer lesions (χ(2)=18.016, P<0.001). The cutoff points were 5,802, 5,803 and 8,002 based on dissimilar pathological types of low grade intraepithelial neoplasia (LGIN), high grade intraepithelial neoplasia (HGIN), and ESCC, respectively, and 5,803 was chosen in this study considering the SE and SP. The SE, SP, PPV, NPV of DNA-ICM test (cutoff point 5,803) combined with liquid-based cytology [threshold atypical squamous cells of undetermined significance (ASCUS)] were separately 72.1% (95% CI: 70.3%-73.9%), 43.3% (95% CI: 41.3%-45.3%), 22.8% (95% CI: 21.1%-24.5%) and 87.0% (95% CI: 85.7%-88.3%) for LGIN, 85.7% (95% CI: 84.3%-87.1%), 41.3% (95% CI: 39.3%-43.3%), 4.6% (95% CI: 3.8%-5.4%) and 98.9% (95% CI: 98.5%-99.3%) for HGIN, and 96.0% (95% CI: 95.2%-96.8%), 40.8% (95% CI: 38.8%-42.8%), 1.7% (95% CI: 1.2%-2.2%) and 99.9% (95% CI: 99.8%-100.0%) for ESCC.

Conclusions: It is possible to use DNA-ICM test as a primary screening method before endoscopic screening for esophageal cancer.

Keywords: DNA image cytometry; cutoff point; esophageal cancer.

1

Trial profiles of enrollment and screening.

2

ROC curves on the basis of various pathologic types. (A) LGIN; (B) HGIN; (C) ESCC.