Abstract
OBJECTIVE: Validation of a U.S.-derived nomogram for individual prediction of disease-specific gastric cancer survival at a European institution. SUMMARY BACKGROUND DATA: One major issue of modern cancer treatment is the individualization of therapy. For gastric cancer, Kattan et al, at Memorial Sloan-Kettering Cancer Center, New York, NY, developed a nomogram, allowing to predict individual patient risk of tumor-related death after R0 resection from basic patient-related variables. The validity of the nomogram has not yet been shown in patients from other institutions. The accuracy of the nomogram when applied to patients after having undergone R0 gastric cancer resection at a European high-volume center was investigated. METHODS: Clinical data from patients who underwent R0 gastric cancer resection at Klinikum rechts der Isar, Technical University of Munich, Germany and fitted the respective derivation criteria were used for external validation (n = 862). Nomogram predictions for 60- and 108-month disease-specific survival were calculated for each patient and compared with actual survival. The concordance index was used as an accuracy measure. RESULTS: The bootstrap-corrected concordance index was 0.77 and was superior when compared with the predictive ability of International Union Against Cancer tumor stage (P < 0.008). Nomogram calibration was excellent for 60-month disease-specific survival. Nomogram predictions showed the trend to underestimate survival in stage II/III disease of the MRI patients. CONCLUSIONS: The use of the nomogram created by Kattan et al is not only confined to the institution where it was created, but it can be adopted by other institutions with similar surgical strategies.