Background: Risk stratification in medullary thyroid cancer (MTC) has traditionally relied on standardized anatomic staging systems that, despite providing valuable prognostic information, do not adequately predict the risk of persistent or recurrent disease. As dynamic risk stratification has been demonstrated to be clinically valuable in nonmedullary thyroid cancer, we adapted our response to therapy definitions in order to apply them to MTC. In this study, we evaluate and compare the clinical utility of our previously proposed MTC response to therapy stratification with a traditional standardized anatomic staging system.

Methods: Both the Tumor, Node, Metastasis/American Joint Cancer Committee (TNM/AJCC) staging system and our previously proposed response to initial therapy staging system was evaluated in 287 MTC patients followed for a median of five years.

Results: The TNM/AJCC staging system provided adequate risk stratification with regard to disease-specific mortality and the likelihood of having no evidence of disease at final follow-up, but did not adequately stratify patients with regard to the likelihood of having structural persistent disease, biochemical persistent disease, or recurrence. However, the response to initial therapy risk stratification system provided clinically useful risk stratification with regard to disease-specific mortality, the likelihood of having no evidence of disease at final follow-up, the likelihood of having a biochemical persistent disease at final follow-up, and the likelihood of having structural persistent disease at final follow-up. Furthermore, the response to therapy risk stratification system demonstrated a higher proportion of variance explained (54.3%) than the TNM/AJCC system (23.9%).

Conclusion: Our data demonstrate that a dynamic risk stratification system that uses response to therapy variables to adjust risk estimates over time provides more useful clinical prognostic information than static initial anatomic staging in MTC thyroid cancer.