Scoring patients' risk in differentiated thyroid cancer - Validation of five scoring systems based on the therapeutic outcome of 406 patients treated in Bulgaria

Despite of excellent long-term results, there is a need for the selection of patients with differentiated thyroid cancer who experience an unfavorable outcome. Patients and Methods: Between 1980 and 1997, 4,06 patients underwent thyroidectomy, elective neck dissection if necessary, and radioiodine treatment and/or external beam radio-therapy according to different histologies and TNM stages. Long-term results were analyzed under the comparison of 5 classification systems (TNM, EORTC, AMES, MACIS, and NTCTCSR) to make a selection of patients with high risk and therefore to plead for aggressive treatment, Results: For 309 TNM stage I-II patients, selected in the low-risk group (LRG), a 20-year cancer-specific survival (CSS) of 96-99% was recorded. A lower CSS of 60-79% was observed for 97 cases (TNM stage III-IV) selected as the high-risk group (HRG) (p = 0.05). The persistence rate for the LRG was 0.9%, compared to 26% for the HRG. The salvage rate after recurrence in the LRG was also significantly higher (84%) than in the HRG (26%) (p = 0.003). The lowest ratio of death rates was scored for AMES (HRG:LRG = 1.2:1), and it is becoming even inverted for papillary cancer (1:2). A higher ratio of death rates (2.6:1) was scored for MACIS and an even higher ratio for EORTC and NTCTCSR (HRG and LRG both 6:1). An analysis of the correlation based on the parameter of cancer-specific mortality reveals that all systems except AMES correlate well with R2 = 0.97-1.0. Conclusion: The best validation for our patient cohort was found for classification criteria announced by EORTC, NTCTCSR and TNM, and they might be useful for optimization of the treatment.