Response to comment on reevaluation of malignancy risk in nondiagnostic thyroid nodules with long‐term follow‐up using surgical resections or core needle biopsy: A retrospective study

Dear Editor, We sincerely thank Sengul and Sengul’s supportive comments on our study. Our study aimed to mitigate selection bias in estimating the upper limit of risk of malignancy (ROM) for nondiagnostic thyroid nodules.1 To address the overestimation risk in evaluating the upper limit of ROM based solely on surgical resection, we incorporated both surgical resection and core needle biopsy (CNB) follow-up in the ROM evaluation. The observed ROM ranged from 5.0% to 17.7%, which aligns with the 2023 Bethesda System's estimated range (5%–20%; mean: 13%).2 However, surgical follow-up alone yielded an overestimated upper limit of ROM up to 54.1%. This finding was consistent with our literature review (Table 6), where ROM in studies limited to surgical cohorts ranged from 7.1% to 57.1%, indicating selection bias inherent in such series. This substantial variance in the upper limit of ROM between surgical-only and integrated methods demonstrates that exclusive reliance on surgical outcomes introduces significant selection bias, as patients undergoing surgery clinically present with higher risk features. Our institution's previous study demonstrated significantly lower nondiagnostic rates with CNB compared to repeat fine-needle aspiration (1.1% vs. 40.0%).3 A meta-analysis supported this finding, demonstrating that CNB reduced the final nondiagnostic rate up to 6.4%, compared to 36.5% with repeat fine-needle aspiration.4 CNB provides comprehensive diagnostic information by preserving tissue architecture, which allows effective evaluation of both the interface between nodular lesions and surrounding nonlesional thyroid parenchyma and various secondary degeneration in nodular lesions. It enables supplementary immunohistochemical or molecular studies as a potential ancillary diagnostic tool in the evaluation of initially nondiagnostic thyroid nodules.5, 6 Furthermore, we suggested that extended surveillance periods may not substantially alter malignancy risk assessment through the observed stability of ROM estimates between 1- and 2-year follow-up thresholds (13.2% vs. 11.2%). We can expect individualized follow-up strategies based on patient-specific clinical risk factors rather than uniform time intervals would potentially improve patient compliance and reducing health care costs without compromising diagnostic accuracy. We totally agree that the potential attrition bias arising from excluding patients with incomplete ultrasonographic profiles may have influenced our observed risk profiles. Additionally, the single-center retrospective design may limit the generalizability of our findings across diverse clinical settings and patient populations. The observed ROM thresholds would require further validation studies in centers with different patient demographics, referral patterns, and institutional practices. We appreciate the opportunity to clarify these important aspects of our study to optimize diagnostic approaches for nondiagnostic thyroid nodules.