Risk factors for cervical lymph node metastasis at different lateral levels in papillary thyroid cancer: level III as the central hub.

Abstract

Background: As the most prevalent histological type of thyroid cancer, the prognosis of papillary thyroid cancer (PTC) is closely related to cervical lymph node metastasis (LNM), particularly in the case of lateral LNM (LLNM). The patterns of cervical LNM in PTC remain inconclusive. We aim to assess the risk factors for LLNM and extranodal extension (ENE) at different lateral levels in N1b PTC patients.

Methods: A total of 486 PTC patients with pathologically diagnosed N1b, who underwent thyroidectomy with central and lateral neck lymph node dissection at our hospital between January 2019 and December 2021, were retrospectively retrieved. The metastatic status of different lateral levels was recorded. Further correlation analysis was conducted.

Results: Based on the correlation analysis, binary logistic regression analysis indicated that upper location [odds ratio (OR) =3.067; P<0.001; 95% confidence interval (CI): 2.075-4.534] and central LNM-ENE (CLNM-ENE) (OR =2.036; P=0.04; 95% CI: 1.044-3.970) are predictive factors for level II metastasis. Additionally, upper location (OR =2.106; P<0.001; 95% CI: 1.404-3.159) and CLNM (OR =2.664; P<0.001; 95% CI: 1.681-4.222) were found to be predictive factors for level III metastasis. LLNM at level IV was closely associated only with CLNM. Level III-ENE (OR =2.347; P=0.03; 95% CI: 1.065-5.176) was found to be highly correlated with level V metastasis. As for ENE, level III-ENE was found to be a risk factor for ENE in all other levels (II, IV, V), while ENE in the other lateral cervical regions also increased the risk for level III.

Conclusions: Despite some uncertainties, LLNM remains closely associated with the upper location of PTC. Meanwhile, level III functions as the hub of the lateral neck, with level III-ENE increasing the risk of LLNM in level V, and significantly raising the probability of ENE in other lateral cervical regions. LLNM have its inherent patterns, as evidenced in the study, but also have a degree of unpredictability, therefore an accurate preoperative assessment is essential.