Background: A preoperative neck ultrasound (US) is recommended for all patients with suspected thyroid cancer, to identify features potentially changing surgical extent. The extrathyroidal extension (ETE) is considered an indication for total thyroidectomy, but there is limited consensus on its US definition, and the interobserver reliability is low. This study aimed to evaluate the predictive value of neck US for ETE before surgery and to estimate the diagnostic performance of different US findings, evaluated during real-time examinations.
Methods: Patients referred to surgery between November 1, 2015, and May 31, 2019, for a suspicious thyroid cancer underwent a preoperative neck US, with systematic assessment for ETE. Three definitions were tested: very restrictive (capsular disruption with suspicious images of surrounding tissues invasion), restrictive (including also capsular abutment with evidence of capsular disruption), and nonrestrictive (capsular abutment is sufficient). Histopathology report of ETE involving at least soft tissues was considered positive. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated.
Results: The study cohort included 128 patients, with 102 (79.7%) confirmed malignancies, and 44 (43.1%) histological ETE. The nonrestrictive definition had good sensitivity (86.4%) but low specificity (29.8%), with an NPV of 80.6%; the restrictive definition had higher specificity (81%), while the very restrictive had specificity and PPV of 100%.
Conclusions: A more extensive surgical approach should not be based on US suspicion of ETE alone, with the possible exception of gross invasion appearance. The absence of any sign of ETE, on the other hand, has a substantial negative predictive value.