Sinusoidal Perfusion in the Veno-Occlusive Region of Living Liver Donors Evaluated by Indocyanine Green and Near-Infrared Spectroscopy

Transcription

1 LIVER TRANSPLANTATION 14: , 2008 ORIGINAL ARTICLE Sinusoidal Perfusion in the Veno-Occlusive Region of Living Liver Donors Evaluated by Indocyanine Green and Near-Infrared Spectroscopy Takuya Hashimoto, Kenji Miki, Hiroshi Imamura, Keiji Sano, Shoichi Satou, Yasuhiko Sugawara, Norihiro Kokudo, and Masatoshi Makuuchi Artificial Organ and Transplantation Surgery, University of Tokyo, Hongo, Bunkyo-ku, Tokyo, Japan Split liver transplantation and living donor liver transplantation (LDLT) commonly use a right liver graft without the middle hepatic vein (MHV). Although tributaries of the MHV are not reconstructed in the majority of cases, the alterations of the microcirculation and its regional functions remain unknown. We addressed these issues by assessing liver tissue indocyanine green (ICG) uptake with near-infrared spectroscopy (NIRS) in 21 donors. After graft procurement, visual inspection (before and after hepatic arterial clamping) and Doppler examination of the veno-occlusive region were performed. Bolus ICG (100 g/kg) was then administered intravenously. Blood ICG at the finger tip was measured with pulse dye densitometry, whereas the liver ICG concentrations in the veno-occlusive and non veno-occlusive regions were simultaneously measured for 15 minutes by NIRS. We estimated the hepatic ICG uptake rate constants in the veno-occlusive region (Ku-oc) and non veno-occlusive region (Ku-non). Changes in sinusoidal perfusion in the veno-occlusive region were expressed by the ratio of Ku-oc to Ku-non (Roc/non). The median value of Roc/non was 0.47, although it ranged from 0.13 to Roc/non was related to the extent of liver surface discoloration before and after hepatic arterial clamping (P 0.03 and 0.01, respectively). In conclusion, sinusoidal perfusion was impaired in the veno-occlusive regions of living donor livers, but the magnitude of the effect varied greatly. Measurement of hepatic ICG uptake by NIRS could become a valuable tool for assessing the indication for venous reconstruction in LDLT and/or split donor liver transplantation. Liver Transpl 14: , AASLD. Received October 15, 2007; accepted January 18, In living donor liver transplantation (LDLT), the question of including the middle hepatic vein (MHV) in the graft or in the remnant donor liver presents a dilemma because the part of the liver without the MHV tends to become congested. Right liver grafts without MHV trunks are the most common in adult-to-adult LDLT. 1,2 Although deprivation of the MHV tributaries can cause congestion of a part of the right liver, the MHV tributaries in this type of graft were not reconstructed in early surgeries as it was assumed that the preexisting venous connections between the MHV and the right hepatic vein would develop as collateral vessels. 3 This assumption was based on the results of human autopsy studies in which vinyl polychloride was injected into the hepatic veins under relatively high pressure. 4-6 However, it is not clear whether these small connec- Abbreviations: B(t), blood indocyanine green concentration at time t; ELH, extended left hepatectomy; ERH, extended right hepatectomy; H(t), indocyanine green concentration of hepatocytes at time t; ICG, indocyanine green; ICG R15, indocyanine green retention rate at 15 minutes; Ku, hepatic indocyanine green uptake rate constant; Ku-non, hepatic indocyanine green uptake rate constant in the non veno-occlusive region; Ku-oc, hepatic indocyanine green uptake rate constant in the veno-occlusive region; L(t), indocyanine green concentration of liver tissue at time t; LDLT, living donor liver transplantation; MHV, middle hepatic vein; NIRS, near-infrared spectroscopy; Roc/non, ratio of the hepatic indocyanine green uptake rate constant in the veno-occlusive region to the hepatic indocyanine green uptake rate constant in the non veno-occlusive region; Vh, ratio of the volume of the hepatic blood pool to the total volume of liver tissue. Supported by a grant from the Kanae Foundation for Life & Socio-Medical Science and a grant-in-aid for scientific research from the Ministry of Education Science and Culture (Grant No ). Address reprint requests to Kenji Miki, M.D, Department of Surgery, Showa General Hospital, 2-450, Tenjin-cho, Kodaira, Tokyo , Japan. Telephone: ; FAX: ; miki-kenji@umin.ac.jp DOI /lt Published online in Wiley InterScience ( American Association for the Study of Liver Diseases.

2 SINUSOIDAL PERFUSION IN VENO-OCCLUSIVE LIVER 873 TABLE 1. Demographic Data for the 21 Donors ELH (n 13) ERH (n 8) Age (years) 35 (20 60) 38 (17 54) Male/female gender 11/2 3/5 Body height (cm) 168 ( ) 160 ( ) Body weight (kg) 65 (55 73) 45 (41 57) Total bilirubin (mg/dl) 0.6 ( ) 0.6 ( ) ICG R15 (%) 6.5 ( ) 4.5 ( ) Duration of operation (minutes) 495 ( ) 485 ( ) Amount of blood loss (g) 470 ( ) 410 ( ) NOTE: Data are expressed as median values (range). Abbreviations: ELH, extended left hepatectomy; ERH, extended right hepatectomy; ICG R15, indocyanine green retention rate at 15 minutes (normal 10%). tions are functional under physiological conditions when the venous drainage is interrupted. In the clinical setting, temporary occlusion of the hepatic vein reportedly results in the regurgitation of portal venous tributaries in the corresponding territory (that is, the venoocclusive region). This suggests that portal veins function as drainage vessels under these circumstances. 7,8 However, we previously encountered a case in which hepatectomy with concomitant MHV resection resulted in complete congestion of a part of the remnant liver. 9 Furthermore, Lee et al. 10 reported necrosis of a right liver graft due to congestion. With this in mind, in a previous study we used ultrasonography to investigate the changes in intrahepatic blood flow caused by interruption of the MHV tributaries in LDLT donors. 11 In summary, our results were as follows: after clamping of the MHV tributaries, although discoloration was not evident in any donor, intrahepatic communication between the right hepatic vein and the MHV was detected in only 8 of 34 donors (24%); in the majority of donors (26 of 34; 76%), regurgitating portal veins functioned as drainage vessels for the veno-occlusive region, receiving hepatic arterial flow as an afferent blood supply. Because of concern that these circulatory changes in the veno-occlusive region could lead to decreased graft function, workers in several institutions including ours have considered using right liver transplantation and reconstruction of the MHV tributaries with a recipient s autologous vein graft, a cryopreserved vein, or an artery graft However, the effects of these intrahepatic circulatory changes on sinusoidal perfusion are unknown, especially in regions in which the hepatic artery becomes the single afferent blood flow. Indocyanine green (ICG), which is a synthetic dye, has been used to assess hepatic function. ICG plasma clearance, measured from the initial distribution phase of the plasma concentration decay curve after bolus injection, is thought to represent the rate of hepatic ICG uptake from the blood. Recently, it was reported that the ICG concentration in deep liver tissue can be measured with near-infrared spectroscopy (NIRS) in animal models This method allowed us to measure directly the hepatic ICG uptake rate in arbitrary regions of the liver. The aim of the current study was to evaluate quantitatively the microcirculation in the veno-occlusive region of the liver during LDLT donor operations. To this end, we compared the regional hepatic ICG uptake rate in the veno-occlusive region with that in non venoocclusive region using NIRS after bolus injection of ICG. Accurate knowledge of the microcirculatory alterations in the veno-occlusive region should shed light on the indication for graft venous reconstruction in LDLT. PATIENTS AND METHODS Donors Between February 2004 and March 2005, 66 LDLTs were performed at our institution. The transplantation procedure and donor selection criteria have been described elsewhere. 19,20 The most common type of graft was right liver without MHV trunk (n 29), followed by left liver with MHV trunk with or without caudate lobe by extended left hepatectomy (ELH; n 19), right liver with MHV trunk by extended right hepatectomy (ERH; n 13), right lateral sector (n 3), and segments 2 3 (n 2). The base population of the study consisted of donors who underwent ELH and ERH, that is, donors who had a veno-occlusive region in the remnant liver after procurement of a liver graft with the MHV trunk. Theoretically, ELH results in a veno-occlusive region on the left side of the right paramedian sector, whereas ERH results in a veno-occlusive region on the right side of the left paramedian sector. This study protocol was approved by the ethics committee of our local authority. We were able to obtain informed consent before the operation from 25 of the 32 donors, 4 of whom were subsequently excluded from the study because of measurement failures. In total, 21 donors were entered into the study (n 13 for ELH and n 8 for ERH). Demographic data on the 21 donors are summarized in Table 1. Visual Evaluation of the Veno-Occlusive Region In accordance with our previous reports, 11,21 hepatic venous congestion was evaluated visually in 2 steps

3 874 HASHIMOTO ET AL. after procurement of the liver graft; at this time, the MHV tributaries of the remnant liver had been ligated on the resection plane. First, we evaluated whether the surface of the veno-occlusive region of the remnant was discolored compared with that of the non veno-occlusive region (baseline color change). When discoloration was not observed, the remnant was classified as unchanged. When faint discoloration appeared, it was classified as faint. When discoloration compatible with complete congestion 9 appeared, it was classified as apparent. In the second step, the hepatic artery was clamped for 5 minutes, and the surface discoloration was re-evaluated. Usually, liver surface discoloration appeared after arterial clamping and was more marked than that observed without arterial clamping. We arbitrarily classified this discoloration into 2 groups: light and dark. Discoloration was classified by the agreement of 2 investigators (T.H. and K.M.). Evaluation of Intrahepatic Portal and Hepatic Venous Blood Flow by Intraoperative Doppler Ultrasonography After visual inspection of the color changes on the liver surface, the hepatic artery was declamped. After 2 minutes, intraoperative Doppler ultrasonography (SSD- 5500, Aloka, Tokyo, Japan) was performed on all of the donors to evaluate blood flow in the occluded tributaries of the MHV and in the portal branches of the corresponding region. When the hepatic venous flow had stagnated and the direction of portal flow of the corresponding region was hepatofugal, the region was finally considered to be veno-occlusive. By contrast, when reversed hepatic venous flow was detected and the portal flow of the same region was hepatopetal because of the presence of venous communications, the region was considered to be non veno-occlusive even though it was part of the MHV-drainage territory. Volumetric Measurement by Three- Dimensional Computed Tomography Volumetric estimation was carried out with a liver volume estimating program (Liver Segmentation Simulator, Hitachi Medico, Inc., Chiba, Japan) 22 using multidetector computed tomography images that had been taken before the operation. This software can calculate the segmental liver volumes perfused by the respective portal branches and those drained by the corresponding hepatic venous tributaries. The graft volume and the remnant liver volume were calculated as those perfused by the right and left portal veins according to the operation procedure. The volume of the veno-occlusive region was calculated as the sum of the volumes of the regions in which venous flow in the MHV tributaries was stagnant. Measurement of the Blood ICG Concentration After the visual and ultrasonic evaluation of the venoocclusive region, a bolus of ICG (Diagnogreen, Daiichi Pharmaceutical Co., Tokyo, Japan) of 100 g/kg was injected intravenously. The blood ICG concentration at time t [B(t); g/ml] was measured for each pulsation with pulse dye densitometry (DDG-2001, Nihon Kohden Industry Co., Ltd., Tokyo, Japan), which was applied to the right second finger. The initial ICG blood concentration was estimated by extrapolation of the blood ICG elimination curve to time zero; this value and the amount of administered ICG were used to calculate the blood volume (ml). 23 Measurement of the Absorption Intensity of the Liver After ICG Injection by NIRS The absorption intensity of the liver (in arbitrary units) was monitored simultaneously with the measurement of the blood ICG concentration for both veno-occlusive and non veno-occlusive regions by NIRS designed to measure tissue oxygenation (PSA-500, Biomedical Science Co., Ltd., Kanazawa, Japan). The sensor consisted of a light source (a diode emitting light at wavelengths of 750 and 830 nm) and a photodetector at a distance of 5 mm. Sensors were placed on the surfaces of the veno-occlusive and non veno-occlusive regions of the liver. Near-infrared light could penetrate the liver tissue, and the emerging light was measured by the photodetector. Absorption intensities were recorded every 2 seconds. ICG has a broad absorption spectrum in the near-infrared region ( nm, with a peak at 805 nm). We measured the increments of absorption intensities at 750 nm to estimate the changes in the liver ICG content. In a preliminary study conducted with 3 donors, we confirmed that the baseline absorption intensities of the venoocclusive and non veno-occlusive regions remained stable for 15 minutes in the absence of ICG injection. Therefore, we concluded that the effects of other variables that might have influenced the baseline absorption intensity (such as hemoglobin and cytochrome) were negligible under the study conditions and that the change of absorption intensity was solely a function of the liver ICG content. Conversion from the Measured Absorption Intensity to the Liver ICG Concentration The measured absorption intensity change (arbitrary units) was converted to the liver ICG concentration at time t [L(t); g/cm 3 liver] as follows. In humans, the amount of ICG that is excreted from the liver into the bile during the first 10 minutes after bolus injection is negligible. 24 Therefore, the amount of ICG removed from the circulating blood can be assumed to be equal to that of the ICG contained in the liver at any arbitrary time point. A coefficient for converting the absorption intensity to the liver ICG concentration was calculated as follows: Coefficient ( g/arbitrary units/cm 3 liver) [Injected ICG ( g)

4 SINUSOIDAL PERFUSION IN VENO-OCCLUSIVE LIVER 875 where Ku is the tissue ICG uptake rate constant (ml/ minute/cm 3 liver). The integrated form of this equation is H t Ku 0 t B d (4) From Equations 2 and 4, it follows that L t Ku o t B d Vh B t (5) Therefore, L t /B t Ku o t B d /B t Vh (6) Figure 1. Schema of the Patlak plot model. Abbreviations: B(t), blood indocyanine green concentration at time t; H(t), indocyanine green concentration of hepatocytes at time t; Ku, hepatic indocyanine green uptake rate constant; L(t), indocyanine green concentration of liver tissue at time t. B(t) at 10 minutes ( g/ml) Blood volume (ml)]/[absorption intensity in the veno-occlusive region at 10 minutes (arbitrary units) Liver volume of veno-occlusive region (cm 3 ) Absorption intensity in the non venoocclusive region at 10 minutes (arbitrary units) Liver volume of the non veno-occlusive region (cm 3 )] (1) Estimation of ICG Uptake in the Veno- Occlusive and Non Veno-Occlusive Regions of the Liver In order to estimate the ICG uptake of the liver, we applied the Patlak plot technique. 25 The compartment model is illustrated in Fig. 1. Liver tissue has 2 components: hepatocyte and sinusoid. The amount of ICG contained in the liver tissue is the sum of the amount of intracellular ICG and the ICG in the hepatic blood pool (sinusoid). In addition, the ICG concentration of the hepatic blood pool is taken to be equal to B(t). Therefore, L t H t Vh B t (2) where H(t) is the intracellular ICG concentration at time t( g/cm 3 liver; that is, the amount of ICG in hepatocytes per unit of liver tissue) and Vh is the ratio of the volume of the hepatic blood pool to the total volume of liver tissue (ml/cm 3 liver). Assuming first-order kinetics, H(t) is described by the following equation: dh t /dt Ku B t (3) Because this is a first-degree equation, L(t)/B(t) can be plotted against B( )d /B(t) with the measured values for L(t) and B(t), and linear regression analysis yields Ku and Vh as the slope and the intercept, respectively. For the present analysis, we used the hepatic and blood ICG measurements from 2 to 10 minutes after ICG injection. We initially estimated the hepatic ICG uptake rate constants in the veno-occlusive region (Ku-oc) and the non veno-occlusive region (Ku-non) of the liver. Then, the ratio of Ku-oc to Ku-non (Roc/non) was calculated, and the relationship between the visual evaluation and Roc/non was investigated. The relationship between the postoperative serum aspartate and alanine aminotransferase levels and Roc/non was also investigated. Statistical Analysis Data analyses were performed with the Statview 5.0 software package (SAS, Cary, NC). The significance of differences was assessed with the Mann-Whitney U test for unpaired values and the Wilcoxon signed-ranks test for paired values. Differences of P 0.05 were considered to be statistically significant. RESULTS Visual Evaluation of the Veno-Occlusive Region It was generally difficult to identify the border of the area of discoloration in donors who underwent ERH because part of the left paramedian sector was covered with falciform and round ligament. Therefore, the results of the visual evaluation are presented only for the 13 donors who underwent ELH. A representative example of the baseline color change and discoloration after arterial clamping is shown in Fig. 2. Regarding the baseline color change, unchanged was observed in 8 donors, and faint was observed in 5 donors. Apparent discoloration was not seen in any of the donors. Five minutes after the right hepatic artery clamping, liver surface discoloration increased in all 13 donors. Of the 8 donors who were classified as unchanged,

5 876 HASHIMOTO ET AL. discolored at baseline and after arterial clamping. Hence, these regions were classified as veno-occlusive. In the other 19 donors, all of the MHV tributaries were stagnant, and the flow of the portal branches in the MHV-deprived territories was hepatofugal; hence, these territories were classified as veno-occlusive regions. Volumetric Measurement by Three- Dimensional Computed Tomography The results of the volumetric measurements are listed in Table 2. The median volume of the veno-occlusive region in the ELH donors was 272 cm 3 (range cm 3 ), which corresponded to 37% (range 14%- 44%) of the remnant liver volume. In the ERH donors, the median volume of the veno-occlusive region was 66 cm 3 (range cm 3 ), corresponding to 17% (range 10%-21%) of the remnant liver volume. Figure 2. Visual evaluation of the veno-occlusive region in donors who received an extended left hepatectomy. White arrows indicate the border of the discolored area. (A) Before hepatic arterial clamping (baseline color change), there was discoloration classified as faint. (B) After hepatic arterial clamping, the liver surface discoloration had increased and was classified as dark. light discoloration was seen in 3, and dark discoloration was seen in 5. All 5 donors who were classified as faint showed dark discoloration. Overall, 3 donors were classified as light, and the remaining 10 were classified as dark. Evaluation of Intrahepatic Portal and Hepatic Venous Blood Flow by Intraoperative Doppler Ultrasonography In 19 of the 21 donors, venous flow was stagnant in all of the MHV tributaries of the remnant liver. In each of the 2 donors who underwent ELH, one tributary draining the cranial part of segment 8 showed reversed flow, which entered the right hepatic vein via intrahepatic venous anastomoses. The portal flow of this region was hepatopetal. The liver surface of the territory corresponding to these reversed venous flows was not discolored upon temporary clamping of the arterial flow. Therefore, these regions were classified as non venoocclusive. In these 2 donors, the other MHV tributaries were stagnant, and the portal flow in the parts corresponding to these hepatic veins was hepatofugal. The liver surface corresponding to these hepatic veins was Measurement of the ICG Concentration of Blood and Liver Tissue Figure 3 shows representative ICG measurements. The blood ICG concentration had a sharp initial peak followed by exponential decay (Fig. 3A). The absorption intensity at 750 nm gradually increased after ICG injection and reached a plateau after about 15 minutes. The absorption intensity of the veno-occlusive area was always higher than that of the non veno-occlusive area (Fig. 3B). A coefficient ( g/arbitrary units/cm 3 liver) for converting absorption (arbitrary units) to concentration ( g/cm 3 liver) was calculated for the first 4 ELH donors and the first 4 ERH donors. Because the mean values ( standard error) of this coefficient in the ELH donors ( ) and the ERH donors ( ) were not significantly different, the mean value (10.4) was used as a conversion constant in the subsequent evaluations. Estimation of ICG Uptake in the Veno- Occlusive and Non Veno-Occlusive Regions of the Liver Figure 4 shows a representative graph produced by the Patlak plot analysis of ICG uptake in veno-occlusive and non veno-occlusive regions. The data for all subjects showed good fit to the regression lines, with high R 2 values ( in veno-occlusive regions and in non veno-occlusive regions). Figure 5 depicts the Ku-oc and Ku-non values in respective patients. The median value of Ku-non was 0.69 ml/minute/cm 3 liver (range ). The median value of Ku-oc was 0.33 ml/minute/cm 3 liver (range ). Ku-oc was lower than Ku-non (P 0.001) in all 21 donors. The median value of Roc/non was 0.47, although it had a relatively wide range ( ). Figure 6A,B shows the Roc/non values corresponding to the results of the visual evaluation in donors who underwent ELH. With respect to the baseline color change, the Roc/non values of the 5 donors who

6 SINUSOIDAL PERFUSION IN VENO-OCCLUSIVE LIVER 877 TABLE 2. Volumetric Measurements ELH (n 13) ERH (n 8) Total liver volume (cm 3 ) 1267 ( ) 1095 ( ) Remnant liver volume (cm 3 ) 788 ( ) 435 ( ) Remnant liver/total liver (%) 61 (57 66) 40 (36 49) Veno-occlusive liver volume (cm 3 ) 272 ( ) 66 (41 99) Veno-occlusive liver/remnant liver (%) 37 (14 44) 17 (10 21) NOTE: Data are expressed as median values (range). Abbreviations: ELH, extended left hepatectomy; ERH, extended right hepatectomy. Figure 3. Measurement of the indocyanine green (ICG) concentration of the blood and the absorption intensity of the liver surface by near-infrared spectroscopy in a donor who received an extended left hepatectomy. (A) Change of the blood ICG concentration. (B) Change of the absorption intensity of the veno-occlusive and non veno-occlusive regions of the liver after ICG injection. The thick line indicates the veno-occlusive region, and the thin line indicates the non veno-occlusive region. were classified as faint were significantly lower than those of the 8 donors who were classified as unchanged (P 0.03; Fig. 6A). As for the discoloration after arterial clamping, the Roc/non values of the 10 donors in whom dark discoloration appeared were significantly lower than those of the 3 donors in whom light discoloration was recognized (P 0.01; Fig. 6B). Figure 4. Diagram of the linear approximation of the Patlak plot analysis for a donor who underwent extended left hepatectomy. Open circles indicate the veno-occlusive region: y x , R The closed circles indicate the non veno-occlusive region: y x , R The slopes of individual regression lines were used for the hepatic indocyanine green uptake rate constant in the veno-occlusive region ( ml/minute/cm 3 liver) and for the hepatic indocyanine green uptake rate constant in the non veno-occlusive region (0.893 ml/minute/cm 3 liver). Abbreviations: B(t), blood indocyanine green concentration at time t; L(t), indocyanine green concentration of liver tissue at time t. No correlations could be confirmed between the maximum postoperative serum aspartate or alanine aminotransferase levels and Roc/non (correlation coefficients, and 0.076; P and , respectively). DISCUSSION The present study assessed microcirculatory disturbance in the parts of LDLT livers in which venous tributaries of the MHV were lacking. Assessment of hepatic ICG uptake using NIRS is thought to be the most appropriate method for this purpose. NIRS allows the he-

7 878 HASHIMOTO ET AL. Figure 5. The differences between hepatic indocyanine green uptake rates in veno-occlusive (Ku-oc) and non veno-occlusive livers (Ku-non). Open circles indicate extended right hepatectomy donors. Closed circles indicate extended left hepatectomy donors. The circles of the same donors are connected by black lines. Ku-oc was less than Ku-non for all donors (P 0.001, Wilcoxon rank test). patic ICG uptake to be assessed chronologically and, at the same time, in separate regions of the liver. Additionally, this parameter is thought to depend entirely on the regional sinusoidal perfusion in normal donor livers. Furthermore, as NIRS detects ICG taken up by hepatocytes, it measures the overall carrier-mediated transport of organic anions including sinusoidal perfusion. In the present study and in our previous work, 11 ultrasonography revealed hepatofugal portal flow throughout the venous-deprived region in the majority of donors, that is, 76%-90%. It should be noted that the change of liver surface coloration, both at the baseline and after hepatic artery clamping, varied considerably even within this population. This indicates that the Figure 6. Relationship between the visual evaluation and the difference in hepatic indocyanine green (ICG) uptake in venoocclusive and non veno-occlusive livers. Roc/non is the ratio of the hepatic ICG uptake rate in the veno-occlusive area to that in the non veno-occlusive area. (A) The relationship between the visual evaluation before hepatic artery clamping and Roc/non. Unchanged indicates the absence of discoloration. Faint indicates the presence of faint discoloration. The differences were statistically significant (P 0.03). (B) The relationship between the visual evaluation after hepatic artery clamping and Roc/non. Light and dark indicate the extent of discoloration. The Roc/non values of the 10 donors with dark discoloration were significantly lower than those of the other 3 donors with light discoloration (P 0.01).

8 SINUSOIDAL PERFUSION IN VENO-OCCLUSIVE LIVER 879 extent of the microcirculatory changes in the veno-occlusive region might differ considerably between individuals. With these observations in mind, we initially evaluated quantitatively liver microcirculatory changes by measuring the hepatic ICG uptake rate in veno-occlusive regions. The most striking result of the present study was that the extent of microcirculatory alteration as assessed by Roc/non showed large interindividual variability (ranging from 0.12 to 0.94: 0.5 in 14/21 cases and 0.7 in 4/21 cases), although conventional visual and/or ultrasonic evaluation has not detected this variability. Another important finding is that the ICG was still taken up in the veno-occlusive region, although at a slower rate than in the non veno-occlusive region. This observation suggests that effective sinusoidal perfusion is preserved in the veno-occlusive region in which the regurgitating portal veins function as drainage vessels. It remains unclear why there is large interindividual variability in the alterations of the microcirculation in the veno-occlusive region. One possible explanation is that there exist minute intrahepatic venovenous anastomoses that cannot be detected by Doppler ultrasonography, which function to different extents in patients even in the parts of the liver classified as venoocclusive regions. This idea is supported by the fact that, although intrahepatic venous anastomoses detectable by Doppler ultrasonography were found in a minority of subjects (10%-24%), they were seen in the majority of subjects (83%) when vinyl polychloride was forcibly injected into the hepatic veins in a human autopsy study. 6 In addition, in our previous study, hepatofugal portal flow in the veno-occlusive regions of remnant donor livers was changed to hepatopetal in 43% of cases 7 days after the operation, 11 and this points to the existence of subclinical venovenous anastomosis. Moreover, in the second step of the present visual inspection, the extent of liver surface discoloration after hepatic arterial clamping varied substantially, even though portal venous flow evaluated by Doppler ultrasonography appeared to be stagnant in all the patients. This heterogeneity could not be explained if the venous outflow was uniformly blocked in the veno-occlusive region. The result of this microcirculatory disturbance in the veno-occlusive region remains to be clarified. Our former study of donors of left liver grafts retaining the MHV showed that the larger the volume of the area drained by the MHV, the higher the maximum postoperative serum alanine aminotransferase level. 26 Similar results have also been reported in LDLT recipients with the veno-occlusive region in the implanted right liver graft. 27 These findings suggest that the microcirculatory disturbance due to deprivation of the MHV causes a small amount of liver injury, as reflected in the elevated aminotransferase values. However, in this study, no correlation could be confirmed between Roc/non and the maximum postoperative serum aminotransferase levels. Postoperative elevation of the serum aminotransferase levels seems to be influenced not only by the microcirculatory disturbances but also by several other factors. We previously reported that living donors who underwent left hemihepatectomy with MHV showed impaired liver regeneration of the right paramedian sector that included the veno-occlusive region. 28 Likewise, after LDLT using right liver grafts without MHV reconstruction, the veno-occlusive region reportedly showed impaired regeneration compared with that in the non veno-occlusive region 3 months after transplantation. 29 This finding agrees with the fact that the ligation and/or embolization of a portal venous branch leads to atrophy of the corresponding part of the liver, with compensatory hypertrophy of the contralateral part. 30,31 It remains controversial whether the MHV tributaries should be reconstructed when a right liver graft without the MHV trunk is used. Usually, the indication for venous reconstruction is decided on the basis of the calculation of the graft volume/recipient standard liver volume ratio, assuming that the veno-occlusive region does not function. If one could determine accurately the extent of microcirculatory alteration in the veno-occlusive region by, for example, the present NIRS examination, this would provide tailor-made information on the indication of venous reconstruction in each LDLT graft. In the current study, we evaluated quantitatively the regional microcirculation in the remnant donor liver by measuring hepatic ICG uptake using NIRS. The hepatic ICG uptake was reduced in the veno-occlusive region, but the magnitudes of microcirculatory alteration varied among the donors. Assessment of hepatic ICG uptake in the veno-occlusive region could become a useful tool for the indication of MHV tributary reconstruction in right liver transplantation. REFERENCES 1. Wachs ME, Bak TE, Karrer FM, Everson GT, Shrestha R, Trouillot TE, et al. Adult living donor liver transplantation using a right hepatic lobe. Transplantation 1998;66: Marcos A, Fisher RA, Ham JM, Shiffman ML, Sanyal AJ, Luketic VA, et al. Right lobe living donor liver transplantation. Transplantation 1999;68: Strasberg SM, Lowell JA, Howard TK. Reducing the shortage of donor livers: what would it take to reliably split livers for transplantation into two adult recipients? Liver Transpl Surg 1999;5: Elias H, Petty D. Gross anatomy of the blood vessels and ducts within the human liver. Am J Anat 1952;90: Lasinski W, Zientarski B. The anastomotic system of subhepatic veins in man. Bull Assoc Anat (Nancy) 1976;60: Couinaud C, Nogueira C. Les veines sus-hepatiques chez l homme. Acta Anat 1958;34: Rappaport AM. Hepatic venography. Acta Radiol 1951;36: Murata S, Itai Y, Asato M, Kobayashi H, Nakajima K, Eguchi N, et al. Effect of temporary occlusion of the hepatic vein on dual blood supply in the liver: evaluation with spiral CT. Radiology 1995;197: Kakazu T, Makuuchi M, Kawasaki S, Miyagawa S, Nakazawa Y, Kubota T, et al. Reconstruction of the middle

9 880 HASHIMOTO ET AL. hepatic vein tributary during right anterior segmentectomy. Surgery 1995;117: Lee S, Park K, Hwang S, Lee Y, Choi D, Kim K, et al. Congestion of right liver graft in living donor liver transplantation. Transplantation 2001;71: Sano K, Makuuchi M, Miki K, Maema A, Sugawara Y, Imamura H, et al. Evaluation of hepatic venous congestion: proposed indication criteria for hepatic vein reconstruction. Ann Surg 2002;236: Cattral MS, Greig PD, Muradali D, Grant D. Reconstruction of middle hepatic vein of a living-donor right lobe liver graft with recipient left portal vein. Transplantation 2001; 71: Gyu Lee S, Min Park K, Hwang S, Hun Kim K, Nak Choi D, Hyung Joo S, et al. Modified right liver graft from a living donor to prevent congestion. Transplantation 2002;74: Sugawara Y, Makuuchi M, Sano K, Imamura H, Kaneko J, Ohkubo T, et al. Vein reconstruction in modified right liver graft for living donor liver transplantation. Ann Surg 2003; 237: Hwang S, Lee SG, Ahn CS, Park KM, Kim KH, Moon DB, et al. Cryopreserved iliac artery is indispensable interposition graft material for middle hepatic vein reconstruction of right liver grafts. Liver Transpl 2005;11: Shinohara H, Tanaka A, Kitai T, Yanabu N, Inomoto T, Satoh S, et al. Direct measurement of hepatic indocyanine green clearance with near-infrared spectroscopy: separate evaluation of uptake and removal. Hepatology 1996;23: El-Desoky A, Seifalian AM, Cope M, Delpy DT, Davidson BR. Experimental study of liver dysfunction evaluated by direct indocyanine green clearance using near infrared spectroscopy. Br J Surg 1999;86: Jiao LR, El-Desoky AA, Seifalian AM, Habib N, Davidson BR. Effect of liver blood flow and function on hepatic indocyanine green clearance measured directly in a cirrhotic animal model. Br J Surg 2000;87: Sugawara Y, Makuuchi M, Takayama T, Imamura H, Kaneko J, Ohkubo T. Safe donor hepatectomy for living related liver transplantation. Liver Transpl 2002;8: Dulundu E, Sugawara Y, Makuuchi M. Revolution and refinement of surgical techniques for living donor partial liver transplantation. Yonsei Med J 2004;45: Sano K, Makuuchi M, Takayama T, Sugawara Y, Imamura H, Kawarasaki H. Technical dilemma in living donor or split-liver transplant. Hepatogastroenterology 2000;47: Saito S, Yamanaka J, Miura K, Nakao N, Nagao T, Sugimoto T, et al. A novel 3D hepatectomy simulation based on liver circulation: application to liver resection and transplantation. Hepatology 2005;41: Iijima T, Aoyagi T, Iwao Y, Masuda J, Fuse M, Kobayashi N, et al. Cardiac output and circulating blood volume analysis by pulse dye-densitometry. J Clin Monit 1997;13: Meijer DK, Weert B, Vermeer GA. Pharmacokinetics of biliary excretion in man. VI. Indocyanine green. Eur J Clin Pharmacol 1988;35: Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab 1983; 3: Kaneko J, Sugawara Y, Matsui Y, Akamatsu N, Kishi Y, Makuuchi M. Correlation between drainage area volume of middle hepatic vein and liver function. Hepatogastroenterology 2005;52: Mitsuta H, Ohdan H, Fudaba Y, Irei T, Tashiro H, Itamoto T, et al. Near-infrared spectroscopic analysis of hemodynamics and mitochondrial redox in right lobe grafts in living-donor liver transplantation. Am J Transplant 2006; 6: Maema A, Imamura H, Takayama T, Sano K, Hui AM, Sugawara Y, et al. Impaired volume regeneration of split livers with partial venous disruption: a latent problem in partial liver transplantation. Transplantation 2002;73: Akamatsu N, Sugawara Y, Kaneko J, Sano K, Imamura H, Kokudo N, et al. Effects of middle hepatic vein reconstruction on right liver graft regeneration. Transplantation 2003;76: Honjo I, Suzuki T, Ozawa K, Takasan H, Kitamura O. Ligation of a branch of the portal vein for carcinoma of the liver. Am J Surg 1975;130: Makuuchi M, Takayasu K, Takuma T, Yamazaki S, Hasegawa H, Nishiura S, et al. Preoperative transcatheter embolization of the portal venous branch for patients receiving extended lobectomy due to the bile duct carcinoma. J Jpn Soc Clin Surg 1984;45:14-20.