Lower invasiveness of papillary thyroid carcinoma associated with Hashimoto's thyroiditis: evidence from AJCC 8th-edition stratification and a dose-response relationship with autoantibody titers.

Background: Hashimoto's thyroiditis (HT) and papillary thyroid carcinoma (PTC) frequently coexist, but the precise impact of HT on PTC invasiveness - particularly across the strict American Joint Committee on Cancer (AJCC) 8th-edition extrathyroidal extension (ETE) categories - remains incompletely characterized. Methods: We retrospectively analyzed 266 consecutive patients with histopathologically confirmed PTC (69 with concurrent HT; 197 without) who underwent thyroidectomy with routine prophylactic central neck dissection at our institution between March 2020 and December 2022. ETE was stratified per strict AJCC 8th-edition criteria into three tiers: No ETE (including pure thyroid capsule invasion), Microscopic ETE (perithyroidal fat invasion), and Gross ETE (T3b strap-muscle or T4 organ invasion). The primary multivariable models adjusted for age, sex, tumor diameter, and multifocality but deliberately excluded BRAF V600E mutation (treated as a mediator on the HT→aggressiveness pathway) to avoid overadjustment bias; secondary models including BRAF are also presented. Sensitivity analyses for nodal metrics included binary logistic regression for any CLNM and negative binomial regression for positive node count with log(nodal yield) offset. Results: Under strict AJCC 8 reclassification, gross ETE was observed in 10.7% of HT(-) versus 4.3% of HT(+) patients (adjusted OR = 0.21, 95% CI 0.06-0.81, P = 0.024). HT was an independent protective factor for ETE severity in ordinal regression (adjusted OR = 0.41, 95% CI 0.19-0.88, P = 0.023) and for BRAF V600E mutation (adjusted OR = 0.47, 95% CI 0.25-0.88, P = 0.018). High CLNM ratio (≥0.4) was less frequent in HT(+) (adjusted OR = 0.47, 95% CI 0.23-0.93, P = 0.029), and yield-adjusted negative binomial regression confirmed HT independently reduced the positive node count (IRR = 0.56, 95% CI 0.40-0.78, P < 0.001). The proportion with any CLNM did not differ between groups (62.9% vs. 55.1%, P = 0.313), nor did Ki-67 ≥5% (27.4% vs. 26.9%, P = 1.000). The proportional-odds assumption for ordinal regression was supported (Brant-type tests, all P > 0.30). Exploratory analyses across TPOAb strata suggested a stepwise gradient with BRAF V600E (Ptrend = 0.020) but were limited by very small subgroup sizes. Conclusions: In this Chinese single-center cohort, HT was independently associated with a less aggressive PTC phenotype - reduced gross ETE, lower BRAF V600E mutation, and reduced nodal metastatic burden - after applying strict AJCC 8 criteria and a causal-inference-informed analytic strategy. These observations strengthen the biological hypothesis that HT modulates PTC behavior, but require prospective multi-center validation with recurrence and survival endpoints before any clinical management changes.