The notion that inflammation parameters such as CRP and albumin might have an influence on outcome in malignant disease is gaining interest. Serum CRP- and albumin levels are mathematically combined in the simple, yet elegant Glasgow Prognostic Score (GPS). This study aimed to evaluate whether the GPS and its variants are able to predict mortality in live donor and deceased donor liver transplantation for hepatocellular carcinoma. Data of 29 live- and 319 deceased-donor transplantations from two German transplant centers was analysed. The Glasgow Prognostic Score (GPS), the modified (mGPS), hepatic (hGPS) and Abe-GPS were investigated. Receiver operating characteristic-curve analysis was performed to calculate the sensitivity, specificity, and overall model correctness of the investigated scores as a predictive model. Study endpoints were 1-year, 3-year and long-term mortality. A 1-year mortality of 19.1% (n=61), 3 year mortality of 26.3% (n=84), and overall mortality of 37.3% (n=119) was observed. All investigated scores failed to predict outcome in deceased-donor liver transplantation (AUROCs <0.700), whereas GPS, hGPS, mGPS and the Abe-score reached AUROCs >0.700 for the prediction of 1 year-mortality in live-donor transplantation. GPS and Abe-score were also able to predict 3 year-mortality. None of the investigated scores was a reliable predictor of long-term mortality. In conclusion, results of previous studies on the prognostic relevance of the GPS variants in living donor transplantation can be confirmed. Furthermore, the most recently developed Abe-score could be successfully externally validated in the current study for the first time. The mGPS with an AUROC of 0.913 for the prediction of 1-year mortality and the Abe-score with an AUROC of 0.842 for the prediction of 3-year mortality had highest prognostic relevance and thus should be applied in living donor transplantation. Further studies should focus on this specific subgroup and would probably benefit from modification of the systemic inflammatory-based scores with donor-specific data to develop a reliable model for the prediction of survival in deceased donor liver transplantation for HCC.

CITATION INFORMATION: Kaltenborn A, Gwiasda J, Heits N, Braun F, Schrem H. The Glasgow Prognostic Score and Its Variants Predict Mortality in Living Donor but Not in Deceased Donor Liver Transplantation for Hepatocellular Carcinoma: A Double-Center Validation Study. Am J Transplant. 2017;17 (suppl 3).

« Back to 2017 American Transplant Congress