Correlation of Different Serum Biomarkers with Prediction of Early Pancreatic Graft Dysfunction Following Simultaneous Pancreas and Kidney Transplantation

Abstract

Background: Despite recent advances and refinements in perioperative management of simultaneous pancreas−kidney transplantation (SPKT) early pancreatic graft dysfunction (ePGD) remains a critical problem with serious impairment of early and long-term graft function and outcome. Hence, we evaluated a panel of classical blood serum markers for their value in predicting early graft dysfunction in patients undergoing SPKT. Methods: From a prospectively collected database medical data of 105 patients undergoing SPKT between 1998 and 2018 at our center were retrospectively analyzed. The primary study outcome was the detection of occurrence of early pancreatic graft dysfunction (ePGD), the secondary study outcome was early renal graft dysfunction (eRGD) as well as all other outcome parameters associated with the graft function. In this context, ePGD was defined as pancreas graft-related complications including graft pancreatitis, pancreatic abscess/peritonitis, delayed graft function, graft thrombosis, bleeding, rejection and the consecutive need for re-laparotomy due to graft-related complications within 3 months. With regard to analyzing ePGD, serum levels of white blood cell count (WBC), C-reactive protein (CRP), procalcitonin (PCT), pancreatic lipase as well as neutrophil−lymphocyte ratio (NLR) and platelet−lymphocyte ratio (PLR) were measured preoperatively and at postoperative days (POD) 1, 2, 3 and 5. Further, peak serum levels of CRP and lipase during the first 72 h were evaluated. Receiver operating characteristics (ROC) curves were performed to assess their predictive value for ePGD and eRGD. Cut-off levels were calculated with the Youden index. Significant diagnostic biochemical cut-offs as well as other prognostic clinical factors were tested in a multivariate logistic regression model. Results: Of the 105 patients included, 43 patients (41%) and 28 patients (27%) developed ePGD and eRGD following SPKT, respectively. The mean WBC, PCT, NLR, PLR, CRP and lipase levels were significantly higher on most PODs in the ePGD group compared to the non-ePGD group. ROC analysis indicated that peak lipase (AUC: 0.82) and peak CRP levels (AUC: 0.89) were highly predictive for ePGD after SPKT. The combination of both achieved the highest AUC (0.92; p < 0.01) in predicting ePGD. Concerning eRGD, predictive accuracy of all analyzed serological markers was moderate (all AUC < 0.8). Additionally, multivariable analysis identified previous dialysis/no preemptive transplantation (OR 2.4 (95% CI: 1.41−4.01), p = 0.021), donor age (OR 1.07 (95% CI: 1.03−1.14), p < 0.010), donor body mass index (OR 1.32 (95% CI: 1.01−1.072), p = 0.04), donors cerebrovascular cause of death (OR 7.8 (95% CI: 2.21−26.9), p < 0.010), donor length of ICU stay (OR 1.27 (95% CI: 1.08−1.49), p < 0.010), as well as CIT pancreas (OR 1.07 (95% CI: 1.03−1.14), p < 0.010) as clinical relevant prognostic predictors for ePGD. Further, a peak of lipase (OR 1.04 (95% CI: 1.02−1.07), p < 0.010), peak of CRP levels (OR 1.12 (95% CI: 1.02−1.23), p < 0.010), pancreatic serum lipase concentration on POD 2 > 150 IU/L (OR 2.9 (95% CI: 1.2−7.13), p = 0.021) and CRP levels of ≥ 180 ng/mL on POD 2 (OR 3.6 (95% CI: 1.54−8.34), p < 0.01) and CRP levels > 150 ng/mL on POD 3 (OR 4.5 (95% CI: 1.7−11.4), p < 0.01) were revealed as independent biochemical predictive variables for ePGD after transplantation. Conclusions: In the current study, the combination of peak lipase and CRP levels were highly effective in predicting early pancreatic graft dysfunction development following SPKT. In contrast, for early renal graft dysfunction the predictive value of this parameter was less sensitive. Intensified monitoring of these parameters may be helpful for identifying patients at a higher risk of pancreatic ischemia reperfusion injury and various IRI- associated postoperative complications leading to ePGD and thus deteriorated outcome.

Keywords: C-reactive protein; biomarker; early pancreatic graft dysfunction; early renal graft dysfunction; graft function; graft outcome; ischemia reperfusion injury; lipase; procalcitonin; simultaneous pancreas–kidney transplantation.

Figure 1

Box plots showing the distribution of WBC (A), PCT (B), CRP (C), Lipase (D), NLR (E) and PLR (F) levels in the ePGD and non-ePGD group preoperative and on POD 1, 2, 3 and 5 following SPKT.

Figure 1

Box plots showing the distribution of WBC (A), PCT (B), CRP (C), Lipase (D), NLR (E) and PLR (F) levels in the ePGD and non-ePGD group preoperative and on POD 1, 2, 3 and 5 following SPKT.

Figure 2

Receiver operating characteristic curve for (A) peak CRP value, (B) peak lipase value as well as (C) combination peak CRP and lipase (Comb_CRPLIP) as predictors for early pancreatic graft dysfunction following SPKT.

Figure 2

Receiver operating characteristic curve for (A) peak CRP value, (B) peak lipase value as well as (C) combination peak CRP and lipase (Comb_CRPLIP) as predictors for early pancreatic graft dysfunction following SPKT.