A comparison of gamma-glutamyl transferase and alkaline phosphatase as prognostic markers in patients with coronary heart disease

Background and aims Whether gamma-glutamyl transferase (GGT) or alkaline phosphatase (ALP) is a better prognostic marker in patients with coronary heart disease (CHD) remains unknown. The aim of this study was to compare the prognostic value of GGT and ALP in patients with CHD. Methods and results This study included 3768 patients with CHD. The main study outcome was 3-year all-cause mortality. The median values of GGT and ALP were 36.2 U/L and 69.3 U/L. Patients were divided into subgroups according to GGT or ALP activity > or ≤median. Overall, there were 304 deaths: 195 deaths occurred in patients with GGT >median (n = 1882) and 109 deaths occurred in patients with GGT ≤median (n = 1886); Kaplan–Meier [KM] estimates of all-cause mortality were 11.9% and 6.4% (unadjusted hazard ratio [HR] = 1.85, 95% confidence interval [CI], 1.46 to 2.34]; P < 0.001). According to ALP activity, 186 deaths occurred in patients with ALP >median (n = 1883) and 118 deaths occurred in patients with ALP ≤median (n = 1885); KM estimates of all-cause mortality were 11.4% and 7.1% (unadjusted HR = 1.64 [1.30–2.06]; P < 0.001). After adjustment, GGT (adjusted HR = 1.32 [1.11–1.58]; P = 0.002) but not ALP (adjusted HR = 1.20 [1.00–1.43]; P = 0.051, with both HR calculated per 1 unit increment in logarithmic GGT or ALP scale) remained significantly associated with the risk for mortality. The C statistic of the mortality model with GGT was greater than the C statistic of the model with ALP (0.831 [0.802–0.859] vs. 0.826 [0.793–0.855]; P < 0.001). Conclusions In patients with CHD, GGT was a stronger correlate of all-cause mortality than ALP.