Regional variation and factors associated with sonographic detectability of postoperative cervical nodal disease in papillary thyroid carcinoma

Abstract Purpose To quantify level-specific ultrasonographic detectability of postoperative cervical lymph node metastases in papillary thyroid carcinoma (PTC) and identify factors independently associated with detection. Materials and methods This retrospective study included 99 postoperative PTC patients with pathologically confirmed cervical nodal metastasis. Ultrasound (US) and surgical pathology were correlated at the patient-side-cervical-level unit because resected nodes were reported collectively by compartment. Among 397 surgically verified side-specific compartments, 257 were pathologically positive. The primary outcome was whether a pathologically positive compartment was detected by US. Multivariable generalized estimating equations (GEE) accounted for repeated compartments within patients. Results US detected 181 of 257 pathologically positive compartments and missed 76, yielding an overall sensitivity of 70.4%, specificity of 80.7%, and accuracy of 74.1%. Sensitivity was highest in levels IV (87.5%) and VI (84.6%) and lowest in levels V (44.0%), VII (40.0%), and II (50.0%). In the multivariable GEE model, levels III (adjusted odds ratio [aOR], 2.59), IV (aOR, 6.69), and VI (aOR, 3.31) were more likely to be detected than level II. In this retrospective analysis, each 10-percentage-point increase in the level-specific lymph node ratio derived from surgical pathology was associated with higher detectability (aOR, 1.12; 95% confidence interval [CI], 1.004–1.241; P = 0.041). No individual index-node feature robustly distinguished true-positive from false-positive US assessments after multiplicity correction. Conclusion In this surgically selected reoperative cohort, the compartment-level diagnostic performance of postoperative cervical US varied across neck levels. Cervical level and pathological nodal tumor burden, rather than any single conventional sonographic feature, were independently associated with detectability in this selected surgical cohort. These findings may inform level-specific US assessment before reoperation, with particular attention to levels II and V and with confirmatory evaluation of positive findings when clinically indicated.