Background

Several European countries have their own guideline for the treatment of differentiated thyroid cancer (DTC), and these guidelines use a risk stratification system to predict disease outcome. No studies evaluated these guidelines in the same population. Therefore, we investigated the prognostic value of six different International guidelines and the TNM system with respect to predicting disease specific survival (DSS).

Methods

We retrospectively studied adult patients with DTC who were diagnosed and/or treated at a Dutch university hospital between January 2002 and December 2017. Patients were classified using the risk classification described in the latest British, Dutch, French, Italian, Polish, and Spanish guidelines, and according to the latest TNM system. DSS was analyzed using the Kaplan-Meier method. The statistical model performance was assessed using the C-index, AIC, BIC and PVE.

Results

We included 860 patients with DTC (79% PTC, 21% FTC) with mean age of 49 years. Median follow-up was 9 years. During follow-up 67 (7.8%) died due to DTC. For all guidelines, the high risk group had a significant worse survival than the other groups. The Dutch guideline had the worst statistical model performance, while the British guideline had the best. However this performance was still worse than the statistical performance of the latest TNM system.

Conclusion

In a European population of patients with DTC, harboring a large set of FTC patients, out of six International guidelines, the latest British guideline performs the best with respect to predicting DSS. However, the TNM system performs better in predicting DSS than this British guideline. Therefore, in clinical practice, these national guidelines should not be used to predict initial DSS.