Portal Venous Pressure in Adult Living Donor Liver Transplantation C.-D. Chang, Y.-F. Cheng, T.-Y. Chen, L.L.-C. Tsang, H.-Y. Ou, C.-Y. Yu, H.-W. Hsu, C.-L. Chen, A.M. Concejero, and T.-L. Huang ABSTRACT Objective. The relationship between portal pressure and small-for-size syndrome (SFSS) is unsettled. The purpose of this study was to evaluate the role of portal pressure in predicting SFSS. Methods. Thirty-four patients with end-stage liver disease who received adult-to-adult living-donor liver transplantation (ALDLT) were included. Recipients were grouped based on whether they received portal flow modulation or not. The intraoperative portal vein flow volume (PVFV) and portal venous pressure (PVP) between the 2 groups were compared. The relationship of PVP to PVFV, graft weight-to-recipient weight ratio (GRWR), and graft weight-to-recipient spleen size ratio (GRSSR) were analyzed. Results. Persistent portal hypertension was found after ALDLT. The PVP was linearly correlated with PVFV but not with GRWR or GRSSR. With the use of the following criteria, (1) PVFV >250 mL/min/100 g graft weight, (2) GRWR 250 mL/min/100 g graft weight) produces high shear stress in the vessel endothelium and sinusoids, where it causes destruction and denudation of vessels, and these have been found to be detrimental in liver regeneration [7]. Based on the above, modulation of the hyperkinetic portal flow by splenic artery ligation and splenectomy has proved to be effective in decreasing portal flow [8,9]. However, graft regeneration also relies on adequate blood supply, and interventions to reduce SFSS should strike a delicate balance between avoidance of hyperflow injury and stimulation of regeneration [5]. The post-modulation PVFV on postoperative day 2 in the present study showed no significant difference between the 2 groups, suggesting sufficient portal flow for regeneration. In a study by Ito et al, an elevated PVP (>20 mm Hg) in the early phase (0e4 days) after transplantation, was strongly associated with poor graft and patient survivals [1]. Poor prognosis is potentially mediated by infection. Our findings are not similar to those results. Persistent portal hypertension (>20 mm Hg) was found in 32 of our recipients after transplantation, with good survival rate. Therefore, pressure criteria alone can not predict survival outcome. In a report from Hong Kong, the graft portal flow was proportional to the portal pressure of the recipient before hepatectomy but not after graft implantation [10]. It is thought that a right lobe liver graft with the middle hepatic vein providing sufficient venous outflow is the main factor that contributes to the relief of the hyperdynamic portal flow even under high portal pressure. In our study, reconstruction of the middle venous territories was performed to

698

prevent anterior sector congestion in the graft with excellent results similar to the Hong Kong group. Portal hypertension is disadvantageous to the liver graft, but the influence of portal pressure in the graft varies. Splenic size is a main influential factor and source of the portal flow in patients with portal hypertension. Even with a new liver and with sufficient outflow, the hyperkinetic portal system may persist. In our previous experience, persistent splenomegaly can occur even after several years after transplantation. Three of the 6 high-risk recipients in our study group underwent early portal flow modulation. The 3 other patients did not undergo modulation, because they did not meet our preset criteria. In the 3 who underwent portal flow modulation, all their portal flows decreased to acceptable levels (20 mm Hg and even up to 35 mm Hg. By postoperative day 2, the PVFV also showed no significance difference between the modulation and nonmodulation groups. This suggests that portal flow modulation may be a very effective method to decrease excessive portal flow without any major influence that would negate graft regeneration. The PVP showed borderline significant difference between the 2 groups, making it a weak predictor. The receiver operating characteristic analysis showed that 23 mm Hg was the cutoff point for PVP. Although a PVP of 23 mm Hg had high sensitivity for predicting SFSS, it had low specificity. After analyzing the relationship of the 3 parameters with PVP, only PVFV showed a correlation with PVP. GRWR and GRSSR showed no significant correlation with PVP. The results suggest the unreliability of PVP alone to precisely predict the development of SFSS. For direct methods of measurement, the PVP can be measured only by catheterization into the portal vein, and there is difficulty in doing this during the postoperative follow-up. The PVFV can be measured by Doppler ultrasound which makes it easy to do and can be repeated by different operators. Therefore, PVFV, GRWR, and GRSSR are preferred to evaluate portal flow modulation. Our study has several limitations. First, we have a relatively small number of cases. Second, we presumed that the 3 patients who underwent modulation would develop immediate SFSS instead of waiting for SFSS to develop. The reason for this is that early modulation of portal flow may

CHANG, CHENG, CHEN ET AL

increase the chance of better graft outcome and recipient survival rate. Third, we measured the absolute PVP without taking into account the systemic arterial blood and venous pressures. Theoretically, the wedge hepatic venous pressure represents the real graft pressure, which can precisely predict graft function and outcome. Wedge hepatic pressure monitoring was not done in this study owing to the absence of invasive monitors to obtain such results. In summary, PVP is a weak parameter for assessing portal flow modulation in LDLT. It is sensitive but not specific to predicting SFSS. GRWR