A 62-year-old Japanese female was referred to Nagasaki University Hospital to discuss an LDLT for recurrent hepatocellular carcinoma (HCC) and end-stage liver disease due to non-B, non-C liver cirrhosis. Two years earlier, she underwent trans-arterial chemoembolization. During follow-up, her liver function gradually worsened with refractory ascites. She then repeatedly received a cell-free and concentrated ascites reinfusion therapy for severe ascites. A new HCC lesion with a 1.5-cm nodule located in segment 7 was also detected by enhanced computed tomography (CT). The Child–Pugh score was 9 (grade B), and the Model for End-stage Liver Disease score was 16. Her dismal status indicated LDLT, and after an appropriate workup, an extended left lobe graft from her 35-year-old healthy daughter was voluntarily donated. Preoperative imaging findings of the donor did not exhibit calcification of the arteries or anatomical variation. Potential causes of failure of hepatic arterial reconstruction, such as a median arcuate ligament syndrome, were not present, and potential vessel injuries by previous trans-arterial chemoembolization had not occurred in the recipient. Donor surgery was performed without any difficulty in the procurement of the extended left lobe graft. The graft’s two hepatic arteries—the left hepatic artery (LHA) and the middle hepatic artery (MHA)—were anastomosed with the recipient’s counterpart microscopically, but it led to thrombosis at numerous attempts. After confirming blood flow in the anastomosed hepatic artery, we performed duct-to-duct biliary reconstruction. We noticed afterwards that we could not confirm blood flow in the hepatic artery with the Doppler US probe. Anastomosis using the recipient’s middle colic artery to the graft’s LHA microscopically through a right radial artery jump graft was attempted with no arterial flow on Doppler control. Thereafter, the arterial oxygen supply was abandoned, and an arteriovenous shunt was created between the selected ileocecal artery and the vein to obtain PVA. We used mesenteric blood vessels to perform selective portal arterialization, in reference to Hayashi et al.’s report [2]. Because the operative field of this procedure was separated from the hepatic hilum, it was easy to approach not only when we created the partial portal arterialization but also when we closed the shunt to prevent symptoms of portal hypertension, after confirmation of collateral arterial blood flow [2]. The mesoileum was opened, and the selected ileocecal artery and vein were taped (Fig. 1a). The ileocecal artery was anastomosed to the vein microscopically in an end-to-side fashion with 9-0 Nylon (Fig. 1b). After anastomosis, ultrasonography confirmed favorable blood flow in the portal and hepatic veins. Portal vein pressure was not measured, and we did not perform any modulation intraoperatively. The operative time was 1737 min, and the blood loss was 23,314 g. The time from portal reperfusion to recovery of arterial oxygenation was 1240 min.
Laboratory analysis revealed that during the period of no oxygenation after portal reperfusion, the AST level was 100 U/L and the ALT level was 118 U/L; on postoperative day (POD) 1, these levels were 118 U/L and 89 U/L, respectively; on POD 2, 72 U/L and 149 U/L, respectively; on POD 3, 71 U/L and 145 U/L, respectively; on POD 4, 108 U/L and 159 U/L, respectively; and on POD 4, 78 U/L and 134 U/L, respectively. After POD 5, both levels decreased gradually. The patient’s serum total bilirubin levels were elevated on POD 1 but decreased gradually after the LDLT. Angiography of the superior mesenteric artery was performed on POD 3, showing a patent shunt and a clearly visible portal vein (Fig. 2). After transplantation, continuous hemodiafiltration was needed until POD 10 because of preoperative hepatorenal syndrome. On POD 13, this was switched to hemodialysis, which was withdrawn on POD 21. The amount of ascites was not severe after LDLT: approximately 2000 mL on POD 7 and approximately 300 mL on POD 14. We removed the drainage tube completely on POD 23. We checked portal flow with the Doppler US probe twice a day for 1 week after LDLT; however, we could not confirm an arterial waveform in portal flow, and in fact, we had never seen high flow volume of the portal vein on US study. Thereafter, the patient’s postoperative course was uneventful. On POD 67, she was discharged. During postoperative month (POM) 3, a CT scan was performed, showing that the portal vein was enhanced through the ileal arteriovenous shunt at the arterial phase (Fig. 3a). Then, during POM 12, another CT scan was performed, revealing that a thin hepatic artery, which might indicate improved arterialization through biliary anastomosis or via intrahepatic biliary plexus, was visible during the early arterial contrast phase (Fig. 3b). On the late arterial phase, the portal vein was enhanced through the ileal arteriovenous shunt (Fig. 3c). We measured the patient’s spleen volume with a three-dimensional image analysis system (SYNAPSE VINCENT; Fujifilm Corporation, Tokyo, Japan) before surgery, on POD 7, and during POM 1, POM 3, and POM 12. The spleen volume was 292 mL preoperatively, 386 mL on POD 7, 337 mL during POM 1, 339 mL during POM 3, and 261 mL during POM 12. The patient’s spleen volume was elevated on POD 7 but decreased gradually after the LDLT. During POM 12, we confirmed, with upper gastrointestinal endoscopy, that the esophageal varices had not progressed. No signs of portal hypertension, progression of esophageal varices, hepatic abscess, biliary complication, or fibrosis were noted after 18 months of follow-up. One year after the LDLT, no ischemic change of the bile duct, biliary necrosis, inflammation around the portal tract, or acute cellular rejection was detected on liver biopsy.