Karina Tanase1, Elena-Daphne Thies1,2, Uwe Mäder3, Christoph Reiners1, Frederik A Verburg1,4. 1. Department of Nuclear Medicine, University of Würzburg, Würzburg, Germany. 2. Department of Psychiatry, Psychosomatic Medicine and Psychotherapy, University of Würzburg, Würzburg, Germany. 3. Comprehensive Cancer Center, University of Würzburg, Würzburg, Germany. 4. Department of Nuclear Medicine, RWTH University Hospital Aachen, Aachen, Germany.
Abstract
OBJECTIVE: Many prognostic systems have been developed for differentiated thyroid cancer. It is unclear which one of these performs 'best'. Our aim was to compare staging systems applicable to our patient database to identify which best predicts DTC-related loss of life expectancy and DTC-specific mortality. DESIGN: Database study of patients with DTC treated in our centre between 1978 (earliest available data) up to and including 1 July 2014. All were staged in accordance with the AMES, Clinical Class, Memorial Sloan Kettering, Ohio State University, TNM versions 5 and 6/7, University of Alabama, University of Münster and qTNM systems. PATIENTS: A total of 2257 patients with differentiated thyroid cancer. MEASUREMENTS: Loss of life expectancy expressed as relative survival and thyroid cancer-specific mortality. Comparison was based on P values of univariate Cox regression analyses as well as analysis of the proportion of variance explained (PVE). RESULTS: Median available follow-up time was 7·2 years (range: 0-35·1 years). Three hundred and twenty-seven patients died, 149 of whom died of DTC. Version 7 of the TNM system was best for predicting DTC-related mortality (P = 7·1 × 10-52 ; PVE = 0·296), followed by TNM version 5 (P = 6·7 × 10-44 ; PVE = 0·255). For prediction of loss of life expectancy, version 7 of the TNM system was also best, closely followed by the Clinical Class system (P both < 2 × 10-16 ). CONCLUSIONS: The UICC/AJCC TNM system version 7 outperforms other prognostic classification systems based on extent of disease at the start of treatment both for prediction of differentiated thyroid cancer-related death and for prediction of loss life expectancy.
OBJECTIVE: Many prognostic systems have been developed for differentiated thyroid cancer. It is unclear which one of these performs 'best'. Our aim was to compare staging systems applicable to our patient database to identify which best predicts DTC-related loss of life expectancy and DTC-specific mortality. DESIGN: Database study of patients with DTC treated in our centre between 1978 (earliest available data) up to and including 1 July 2014. All were staged in accordance with the AMES, Clinical Class, Memorial Sloan Kettering, Ohio State University, TNM versions 5 and 6/7, University of Alabama, University of Münster and qTNM systems. PATIENTS: A total of 2257 patients with differentiated thyroid cancer. MEASUREMENTS: Loss of life expectancy expressed as relative survival and thyroid cancer-specific mortality. Comparison was based on P values of univariate Cox regression analyses as well as analysis of the proportion of variance explained (PVE). RESULTS: Median available follow-up time was 7·2 years (range: 0-35·1 years). Three hundred and twenty-seven patients died, 149 of whom died of DTC. Version 7 of the TNM system was best for predicting DTC-related mortality (P = 7·1 × 10-52 ; PVE = 0·296), followed by TNM version 5 (P = 6·7 × 10-44 ; PVE = 0·255). For prediction of loss of life expectancy, version 7 of the TNM system was also best, closely followed by the Clinical Class system (P both < 2 × 10-16 ). CONCLUSIONS: The UICC/AJCC TNM system version 7 outperforms other prognostic classification systems based on extent of disease at the start of treatment both for prediction of differentiated thyroid cancer-related death and for prediction of loss life expectancy.
Authors: Seul Gi Lee; Joon Ho; Jung Bum Choi; Tae Hyung Kim; Min Jhi Kim; Eun Jeong Ban; Cho Rok Lee; Sang-Wook Kang; Jong Ju Jeong; Kee-Hyun Nam; Sang Geun Jung; Young Suk Jo; Jandee Lee; Woong Youn Chung Journal: Medicine (Baltimore) Date: 2016-02 Impact factor: 1.889