Elimination of Indocyanine Green in the Perioperative Evaluation of Donor Liver Function

Failure of the donor liver graft soon after transplantation is a life-threatening complication that usually requires emergent retransplantation. Patient mortality is increased, and the pool of scarce organs available to other potential transplant recipients is reduced. Acute failure of the donor liver graft is usually due to unrecognized parenchymal injury during the process of donation or technical difficulties that occur during vascular reconstruction, impairing blood flow to the graft (1). Rapid perioperative diagnosis of graft dysfunction would improve physician decision-making and, consequently, patient outcome. Although it is difficult to improve graft function caused by cellular injury or death, reversible ischemic injuries due to technical problems in vascular construction are potentially correctable if detected early. In this report, we describe how noninvasive indocyanine green (ICG) dye elimination identified a critical reduction in hepatic function. This observation led to a change in clinical management that significantly improved the outcome. Case Report Institutionally approved written, informed consent was obtained from a transplant recipient to participate in a larger study evaluating the perioperative recirculatory kinetics of ICG (2). A 46-yr-old man with hepatitis C, waiting for transplantation, was admitted to the intensive care unit (ICU) with ascites, renal failure, and progressive encephalopathy. A cadaveric liver became available on the ninth day after admission. In the operating room, ICG elimination was determined after the induction of general anesthesia, excision of the native liver, and reperfusion of the donor graft. Studies were repeated on the first postoperative day. Elimination slopes (K) were estimated from a noninvasive integrated pulse-spectrophotometric monitor (Nihon Kohden, Saitama, Japan) after a 10-mg bolus dose of ICG was injected into a central venous catheter. The transcutaneous probe of the monitor placed on the external nares emits wavelengths of 805 and 940 nm to measure arterial ICG concentration. The arterial hemoglobin concentration was measured before ICG injection. The device calculated kinetic estimates for cardiac output (CO), blood volume, and ICG K values by using standard equations from the dye dilution history. CO was estimated with the Stewart-Hamilton approach, and blood volume was estimated by dividing the ICG dose by the back-extrapolated concentration at Time 0. K was estimated by linear regression of the semilog concentration versus time data from t = 2.5 min to t = 5.5 min. After reperfusion of the donor liver, the portal vein appeared patent. There were no signs to suggest donor graft dysfunction, and the patient was hemodynamically stable, with minimal bleeding. However, we suspected significant graft dysfunction because the ICG elimination value of 0.012/min from the reperfused graft was similar to the anhepatic value of 0.009/min (Table 1). An intraoperative ultrasound was performed to independently assess hepatic blood flow, and a vein graft was inserted after we confirmed a critical reduction in portal venous blood flow.Table 1: Perioperative ICG Elimination Rate ConstantsThe ICG K improved significantly after reconstruction of the portal vein (Fig. 1, Table 1), and brisk portal venous flow was confirmed by ultrasound. A dye study conducted during the first postoperative day in the ICU showed progressive improvement in the K (Table 1), and ultrasonography showed hepatic blood flow to be unchanged. The patient had an uneventful recovery and was discharged from the hospital on postoperative Day 11.Figure 1: Arterial blood indocyanine green (ICG) concentration histories for Minutes 1 through 9 after rapid central venous injection during the five distinct surgical phases listed in Table 1 and described in the text. The reduction in ICG clearance during Phases I, II, and III is consistent with a lack of portal vein flow. There was dramatic improvement in the slope of the ICG concentration versus time curve once portal flow was restored. This improvement continued into the first postoperative day. Note that there was a significant motion artifact in the postoperative study when the patient was conscious. However, this did not affect the results.Discussion We identified donor liver dysfunction by using noninvasive ICG elimination data generated instantaneously at the bedside. The data suggested failure to reestablish adequate hepatic blood flow, a finding confirmed by ultrasonography and not previously reported for ICG testing during liver transplantation. This early diagnosis affected a change in clinical management that probably prevented subsequent graft failure and the need for retransplantation. ICG dye dilution is often used to measure liver function, blood volume, and CO (3,4). The dye is rapidly extracted from the blood only by the liver and is excreted in the bile unchanged (5,6); enterohepatic circulation and metabolism of the parent compound do not occur (7). Thus, only hepatic blood flow, extraction, and biliary transport determine clearance. Because ICG is extracted and eliminated according to first-order kinetics, ICG dye dilution is an excellent estimation of global liver function (8). Dye elimination correlates with donor graft function (9–11) and is a sensitive test capable of identifying reperfusion injury (9). However, standard testing is currently limited by the cumbersome technique of measuring the ICG concentration in serial blood samples. The extended time for standard analysis and reporting of results makes this a valuable test of hepatic function impractical in the operating room. Estimation of the same variables by the novel noninvasive technique compares favorably with standard techniques, and results are readily available at the bedside (12,13). Although dye elimination correlates well with postoperative liver function (10,11), investigators are unable to agree on a single clearance or elimination value that predicts outcome in all patients. Rather than use previously published values, we compared dye elimination during surgery with a baseline value measured after the induction of anesthesia. Although most measurements were made during surgery, ICG elimination is not affected by general anesthesia (4). The marked decline in dye elimination that our patient experienced after excision of the liver is easily explained by the near-complete interruption of hepatic blood flow. We were surprised that the dye-elimination rate did not increase after reperfusion, because previous studies have shown a rapid increase in ICG elimination curves within one hour of perfusion of the donor graft (12). It is unlikely that these changes were due to physiological perturbations associated with transplantation, because the CO and estimated blood volume remained relatively constant throughout surgery. The similarity of our anhepatic to reperfusion results suggested insufficient blood flow to the liver. We used ultrasonography to distinguish between hepatocellular dysfunction and impaired blood flow. Dye elimination improved significantly after surgical repair reestablishing portal blood flow. The continued improvement in ICG elimination observed in the ICU suggested improved donor graft function and is similar to findings from other studies (11). The diagnosis of hepatic dysfunction and insufficient blood flow is usually based on clinical judgment, because there are no quantitative tests that could provide rapid analysis of hepatic function in the operating room. Our findings are novel in that they are the first to describe the diagnosis of intraoperative donor liver dysfunction on the basis of objective information. In our case, early diagnosis led to a change in surgical management that improved liver function. We suggest that the technique of noninvasive ICG analysis may fill this diagnostic void by providing rapid data on hepatic function. The role of this tool as a multidimensional aid to evaluation of liver function during and after transplantation deserves further consideration.