Procalcitonin as a prognostic marker for infectious complications in liver transplant recipients in an intensive care unit

Abstract

Clinically significant infections (CSIs) are life-threatening but difficult to diagnose after liver transplantation (LTx). This study investigates the value of procalcitonin (PCT) in addition to c-reactive protein (CRP) and the leukocyte count (LC) as a prognostic marker for CSIs in LTx recipients. The clinical course of 135 LTx recipients was prospectively studied. CSIs were defined as pulmonary, bloodstream, or intra-abdominal infections. Independent risk factors for CSIs were determined by Cox proportional hazard analysis. The concordance statistics (c-statistics) were used to assess the discrimination effect of PCT. Thirty recipients (22%) experienced a CSI. They had significantly higher peak PCT (27.2 versus 12.7 ng/mL, P = 0.014) and peak CRP (13.7 versus 9.9 mg/dL, P < 0.001) and a tendency toward a higher peak LC (19.3 versus 14.2 cells/nL, P = 0.051) in comparison with recipients without CSIs. Independent risk factors for CSIs were male sex [hazard ratio (HR) = 6.4], a body mass index (BMI) < 20 kg/m(2) (versus a BMI > 25 kg/m(2), HR = 13.8), acute liver failure as an indication for LTx (HR = 7.1), a cold ischemic time > 420 minutes (HR = 3.5), and peak CRP (HR = 1.1) but not peak PCT. The addition of peak PCT marginally improved the c-statistic from 0.815 to 0.827. In conclusion, although peak PCT differed significantly between recipients with and without CSIs, it was not an independent risk factor for CSIs and added little prognostic accuracy. Interestingly, the parameters peak CRP, male sex, low BMI, acute liver failure, and long cold ischemic time were independent risk factors for CSIs. They could serve as risk stratifiers directing medical therapy in clinical practice.