A differentiated thyroid carcinoma-specific graded prognostic assessment for brain metastases: derivation in a...

Brain metastases from differentiated thyroid carcinoma (DTC) are rare but associated with significantly worse prognosis compared to extracranial metastatic sites. Existing prognostic models for brain metastases are not tailored to DTC, limiting risk stratification and therapeutic guidance in this population.

We conducted a retrospective cohort study of 10,306 patients with DTC treated at a single tertiary cancer center from 2011 to 2025. Patients with radiologically or pathologically confirmed brain metastases were included. Multivariable Cox regression was used to identify independent prognostic factors for overall survival (OS), defined from the time of Brain metastases diagnosis. Based on these factors, we developed the Differentiated Thyroid Cancer Graded Prognostic Assessment (DTC-GPA) and evaluated its discriminatory performance using Harrell's C-index. External validation was performed using individual-level data from published case series identified via systematic review.

Of 10,306 patients with DTC, 146 (1.4%) were found to have brain metastases. The median OS after Brain metastases diagnosis was 20.4 months. ECOG performance status ≥ 2 (HR, 2.67; 95% CI, 1.65-4.30; p < 0.001), presence of multiple brain metastases (HR, 2.30; 95% CI, 1.39-3.82; p = 0.002), and interval ≥ 3 years between primary DTC and Brain metastases diagnosis (HR, 1.84; 95% CI, 1.03-3.27; p = 0.03) were independently associated with worse OS. These three factors formed the DTC-GPA, assigning one point for each favorable variable (range: 0-3). Median OS by score group was 7.7 months (score 0), 22.9 months (score 1), and 61.3 months (score 2-3). The C-index was 0.70 in the derivation cohort and 0.68 in the external validation cohort of 83 patients.

The DTC-GPA is the first validated, disease-specific prognostic tool for patients with brain metastases from DTC. This simple, three-point score offers clinically meaningful risk stratification to guide treatment decisions and prognosis estimation in a rare and heterogeneous patient population.