Accuracy of Computed Tomography for Predicting Pathologic Nodal Extracapsular Extension in Patients With Head-and-Neck Cancer Undergoing Initial Surgical Resection

Abstract

Nodal extracapsular extension (ECE) in patients with head-and-neck cancer increases the loco-regional failure risk and is an indication for adjuvant chemoradiation therapy (CRT). To reduce the risk of requiring trimodality therapy, patients with head-and-neck cancer who are surgical candidates are often treated with definitive CRT when preoperative computed tomographic imaging suggests radiographic ECE. The purpose of this study was to assess the accuracy of preoperative CT imaging for predicting pathologic nodal ECE (pECE). The study population consisted of 432 consecutive patients with oral cavity or locally advanced/nonfunctional laryngeal cancer who underwent preoperative CT imaging before initial surgical resection and neck dissection. Specimens with pECE had the extent of ECE graded on a scale from 1 to4. Radiographic ECE was documented in 46 patients (10.6%), and pECE was observed in 87 (20.1%). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were 43.7%, 97.7%, 82.6%, and 87.3%, respectively. The sensitivity of radiographic ECE increased from 18.8% for grade 1 to 2 ECE, to 52.9% for grade 3, and 72.2% for grade 4. Radiographic ECE criteria of adjacent structure invasion was a better predictor than irregular borders/fat stranding for pECE. Radiographic ECE has poor sensitivity, but excellent specificity for pECE in patients who undergo initial surgical resection. PPV and NPV are reasonable for clinical decision making. The performance of preoperative CT imaging increased as pECE grade increased. Patients with resectable head-and-neck cancer with radiographic ECE based on adjacent structure invasion are at high risk for high-grade pECE requiring adjuvant CRT when treated with initial surgery; definitive CRT as an alternative should be considered where appropriate.