Clinical and Molecular Genetic Testing Risk Stratification for Pediatric Papillary Thyroid Carcinoma Therapy.

BACKGROUND

Children and young adolescents with papillary thyroid cancer (CAYA-PTC) have high recurrence rates. Current guidelines advocate intensive treatment, yet existing risk stratification systems lack integration of genomic characteristics and provide limited precision in guiding surgical extent and postoperative radioactive iodine (RAI).

PATIENTS AND

METHODS

This retrospective, multicenter study analyzed 435 CAYA-PTC patients (ages 2-18 years) treated between 2006 and 2024. A clinicogenomic nomogram for disease-free survival (DFS) was developed using clinical data, postoperative pathology, and molecular testing of surgical tissue. Patients were stratified according to the nomogram score. Model performance was internally compared with the American Thyroid Association (ATA) pediatric risk system. Treatment associations were assessed by surgical extent (total thyroidectomy vs lobectomy) and receipt of postoperative RAI therapy within each risk stratification.

RESULTS

Five variables were independent predictors of worse DFS: age ≤15 years, tumor size >4 cm, lateral neck node involvement (N1b), coexisting Hashimoto thyroiditis, and RET/NTRK fusions (all P <.05). The nomogram stratified patients into low-, intermediate-, and high-risk groups with 7-year DFS rates of 92.4%, 83.1%, and 56.4% (P <.001). In internal comparison, nomogram-based risk stratification had higher concordance index (0.721 vs 0.656) and 3-year AUC (0.748 vs 0.651) values than ATA stratification (both P <.05). Total thyroidectomy was associated with better DFS in intermediate- and high-risk patients but not low-risk patients. Higher DFS with postoperative RAI was observed in the high-risk group (P =.028).

CONCLUSIONS

Using retrospective clinicogenomic features, our stratification suggests a risk-adapted framework for CAYA-PTC. However, non-significant differences, especially in small subgroups, do not establish treatment equivalence or confirm the safety of de-escalation. These findings represent associations that require future external validation.