Thin-slice MDCT of the neck: impact on cancer staging.

Abstract

The objective of this study was to compare thin-slice multiplanar evaluation and conventional 3-mm axial evaluation of head and neck MDCT in tumor staging. Ninety-six patients with histologically proven squamous cell carcinoma were evaluated independently, once using 3-mm axial images and once using 1-mm interactive multiplanar reformation (MPR) images. Tumor stage was assessed with both methods; histology served as the reference. Thirty-seven patients with hypopharyngeal and laryngeal tumors had en bloc resection, allowing direct comparison of tumor infiltration into designated anatomic structures. Two examiners independently assessed the data sets. Interobserver agreement was tested with a modified kappa test. The Wilcoxon signed rank test with continuity correction was applied to test the null hypothesis, which postulates the equality of both methods. The chi-square test was applied to compare the number of correctly classified tumors for the two methods and readers. Interobserver agreement was high (kappa = 0.88-0.91). Both methods allowed accurate tumor staging, and no significant differences between the two methods were found (reader A, p = 0.61; reader B, p = 1). With MPR assessment, more anatomic structures were rated positive for tumor infiltration, but diagnostic accuracy did not differ significantly in the subgroup of patients with histologic correlation from en bloc resection. Conventional 3-mm axial evaluation of head and neck MDCT proved to be sufficient in tumor staging.