Retrohepatic vena cava replacement of hepatic malignancies without using total hepatic vascular exclusion or extracorporeal bypass

Total hepatic vascular exclusion and venovenous bypass are frequently used surgical procedures when concomitant resection of the inferior vena cava is required during surgery of liver cancer involving the retrohepatic inferior vena cava close to the hepatic veins. However, the duration of total hepatic vascular exclusion is limited due to the risk of hepatic ischemia. Three patients presented with severely compressed inferior vena cava and/or hepatic veins due to liver cancer. The surgical procedure involved initial taping of the inferior vena cava just below the hepatic veins by extrahepatic division and taping of the hepatic veins. After taping the inferior vena cava, hepatectomy with caval resection was performed by simply clamping the retrohepatic inferior vena cava, without the need for total hepatic vascular exclusion or venovenous bypass. In all patients the retrohepatic inferior vena cava were safely replaced with a prosthetic graft under stable hemodynamics. Duration of the inferior vena cava clamping was 31, 66, 75 minutes, respectively. No graft-related complications occurred, but 2 of the 3 patients showed temporal renal dysfunction associated with renal congestion postoperatively. The surgical procedure described herein is effective for the treatment of retrohepatic inferior vena cava in some patients. However, when the case is complicated by chronic nephropathy or simultaneous nephrectomy is required, venovenous bypass should be performed.

Long-term survival after resection for small cell carcinoma of the esophagus

We describe the rare case of a patient with esophageal small cell carcinoma who was completely cured. A 77-year-old man had small cell carcinoma of the esophagus with extensive lymph node metastases. Treatment comprised a subtotal esophagectomy and extended lymph node dissection. He has survived for more than 7 years with no evidence of recurrent disease. We suggest that radical operations should be considered for future patients if curative resection can be expected.

Effects of a vitamin E-modified dialysis membrane on neutrophil superoxide anion radical production

Activation of neutrophil by the dialysis membrane and peroxidative stress plays an important role on the pathogenesis of complications in hemodialysis (HD) patients. Vitamin E is one of the potent scavengers for reactive oxygen species. Recent studies suggest that a vitamin E-modified multilayer membrane (Excebrane, CL-EE dialyzer) has an inhibitory effect on serum lipids peroxidation in HD patients. To determine the effect of CL-EE on biocompatibility in clinical use, we measured the superoxide anion radical producing ability (SOPA) of polymorphonuclear leukocytes (PMNLs), the plasma hydroxyl radical producing ability (OHPA) and superoxide anion radical scavenging activity (SSA). SOPA was measured after stimulation of PMNLs with phorbol myristate acetate using electron paramagnetic resonance (EPR) method. Plasma OHPA and SSA were also determined using the EPR method. In addition, the plasma concentrations of malondialdehyde (MDA) and oxidized low-density lipoprotein (LDL), as the parameters for lipid peroxidation, were measured. SOPA was decreased in patients who used conventional filter membrane compared with healthy controls. In the patients using the CL-EE membrane, SOPA gradually increased and reached control levels after six months. However, no significant increase was observed in patients who used a conventional filter membrane. OHPA of HD patients was significantly decreased compared with controls. In the CL-EE membrane patient group, OHPA was significantly increased at six months. SSA was significantly higher in the conventional filter membrane group than controls. In the CL-EE membrane patient group, SSA gradually decreased at six months. Plasma MDA and oxidized LDL levels were significantly higher in HD patients compared with controls. These values slowly decreased, and significant differences were found after nine months of using the CL-EE membrane. These findings suggest that activation of PMNLs and plasma OHPA and SSA in HD patients is attenuated by antioxidant effects of the CL-EE.

Indications for portal vein embolization combined with major hepatic resection for advanced-stage hepatocellular carcinomas. A preliminary clinical study

BACKGROUND

Criteria for selection of patients for portal vein embolization (PVE) before major hepatectomy for advanced-stage hepatocellular carcinoma (HCC) have not been clarified in detail. This study was aimed at defining those benefiting from this therapy in a retrospective fashion.

PATIENTS AND METHODS

Firstly, to determine liver functional criteria for applying this approach 26 patients with stage III (17 patients) or IV (9 patients) disease, who underwent major hepatectomies after PVE, were divided into those without major complications (20 patients) and a postoperative liver failure group (6 patients). Clinical, analytical, and hemodynamic parameters obtained before and after PVE were compared between the groups. Secondly, to define the application of this approach with regard to tumor progression survival rates of patients were also obtained, taking into account factors which affect tumor development, i.e. lesion size, intrahepatic metastasis and vascular invasion.

RESULTS

With regard to liver function 4 nonindications were obtained: (1) a portal pressure measured right after PVE >25 cm H(2)O; (2) post-PVE serum hyaluronate >200 ng/ml; (3) pre-PVE serum cholinesterase <150 U/l; (4) post-PVE serum cholinesterase <150 U/l. In view of the tumor progression in patients with HCCs featuring intrahepatic metastasis spread to more than 3 segments (IM3) 1-, 3- and 5-year survival rates were low (42.9, 28.6 and 0%) with a statistical significance, compared to those in patients with intrahepatic metastasis limited in the same lobe (76.9, 46.2 and 24.6%).

CONCLUSIONS

When laboratory data fulfill 3 or more of the criteria, the extent of hepatic resection may have to be carefully reconsidered. Patients with HCCs featuring IM3 intrahepatic metastasis may not benefit from the aggressive approach described here.

Effects of a vitamin E-modified dialysis membrane on neutrophil superoxide anion radical production

Activation of neutrophil by the dialysis membrane and peroxidative stress plays an important role on the pathogenesis of complications in hemodialysis (HD) patients. Vitamin E is one of the potent scavengers for reactive oxygen species. Recent studies suggest that a vitamin E-modified multilayer membrane (Excebrane, CL-EE dialyzer) has an inhibitory effect on serum lipids peroxidation in HD patients. To determine the effect of CL-EE on biocompatibility in clinical use, we measured the superoxide anion radical producing ability (SOPA) of polymorphonuclear leukocytes (PMNLs), the plasma hydroxyl radical producing ability (OHPA) and superoxide anion radical scavenging activity (SSA). SOPA was measured after stimulation of PMNLs with phorbol myristate acetate using electron paramagnetic resonance (EPR) method. Plasma OHPA and SSA were also determined using the EPR method. In addition, the plasma concentrations of malondialdehyde (MDA) and oxidized low-density lipoprotein (LDL), as the parameters for lipid peroxidation, were measured. SOPA was decreased in patients who used conventional filter membrane compared with healthy controls. In the patients using the CL-EE membrane, SOPA gradually increased and reached control levels after six months. However, no significant increase was observed in patients who used a conventional filter membrane. OHPA of HD patients was significantly decreased compared with controls. In the CL-EE membrane patient group, OHPA was significantly increased at six months. SSA was significantly higher in the conventional filter membrane group than controls. In the CL-EE membrane patient group, SSA gradually decreased at six months. Plasma MDA and oxidized LDL levels were significantly higher in HD patients compared with controls. These values slowly decreased, and significant differences were found after nine months of using the CL-EE membrane. These findings suggest that activation of PMNLs and plasma OHPA and SSA in HD patients is attenuated by antioxidant effects of the CL-EE.

[Outcomes of home anti-cancer chemotherapy--estimation of hepatic arterial infusion chemotherapy for patients with multiple liver metastases]

A total of 18 patients (13: colon cancer, 5: gastric cancer) with multiple liver metastases (H3) underwent hepatic arterial infusion chemotherapy (HAI) using an implanted arterial port with portable syringe pumps in our outpatient clinic. Clinical perspective: overall response rate was 22.2% (CR: 1 case, PR: 3 cases (1 case: hepatectomy after HAI), NC: 12 cases, PD: 2 cases), however, 7 of 12 cases of NC were long NC (more than 6 months). No major complications with HAI were experienced. Patient Perspective: After HAI in our outpatient clinic, the 50% survival was 341 days, 50% hospital free days were 319 days and home stay rate was 92.9%. Societal Perspective: cost and hospital stay days were significantly reduced. Home anti-cancer chemotherapy using HAI for gastrointestinal cancer patients with multiple liver metastases was safe and efficient from the viewpoint of medical outcomes.

Extent of pathologic invasion of the inferior vena cava in resected liver cancer compared with possible caval invasion diagnosed by preoperative images

The extent of cancerous invasion of the inferior vena cava (IVC) determined from resected liver cancer was examined pathologically. Ten patients presenting with liver cancer (metastatic liver cancer, five patients; hepatocellular carcinoma, three; and cholangiocellular carcinoma, two) were diagnosed with positive IVC invasion using preoperative imaging techniques of extracorporeal ultrasonography, computed tomography, magnetic resonance imaging, and vena cavography. The diagnostic criterion for positive IVC invasion by preoperative imaging was longitudinal IVC compression measuring over 50 mm, or transverse IVC compression extending to more than half the circumference of the IVC, or the presence of lesions protruding into the IVC lumen, or the presence of developed collateral veins. All patients underwent combined resection of the IVC. However, pathology results revealed that four of the ten patients had no cancerous invasion of the IVC, and that the extent of invasion along both the longitudinal and transverse axes of the IVC was much smaller than the compression shown by imaging results. We believe that detailed preoperative assessment, using a more precise imaging technique, as well as further intraoperative examination, is required to predict the full pathological extent of cancerous invasion of the IVC.

Preventive effect of preoperative portal vein ligation on endotoxin-induced hepatic failure in hepatectomized rats is associated with reduced tumour necrosis factor alpha production

BACKGROUND

Preoperative portal vein embolization successfully reduces the incidence of postoperative hepatic failure in which endotoxin is postulated to be involved. To identify the mechanism of this preventive effect, the relationship of endotoxin-induced liver injury with tumour necrosis factor (TNF) alpha and nitric oxide production in the peripheral blood, liver and spleen of rats subjected to preoperative portal vein branch ligation (PVL) was compared with that in rats undergoing sham operation.

METHODS

Rats with PVL and those that underwent sham operation were subjected to resection of ligated liver lobes (PVL-Hx rats) and two-thirds hepatectomy (noPVL-Hx rats) respectively at day 5, followed by intravenous administration of endotoxin 200 microgram/kg body-weight at day 7. At various time intervals after endotoxin injection, the peripheral blood, liver and spleen tissues were harvested and analysed for TNF-alpha and nitric oxide production.

RESULTS

The survival rates of noPVL-Hx and PVL-Hx rats at 48 h after endotoxin administration were 40 and 100 per cent respectively. The former rats showed more extensive liver injury as represented by higher serum aminotransferase and hyaluronate levels than the latter. Plasma concentrations of TNF-alpha at 1.5 h after endotoxin treatment were significantly higher in noPVL-Hx rats (mean(s.e.m.) 22 125(2175) pg/ml; n = 6) than PVL-Hx rats (8344(4076) pg/ml; n = 6) (P < 0.01). Consistent with this, expression of TNF-alpha messenger RNA in the liver and spleen was suppressed in PVL-Hx rats. In two-thirds hepatectomized rats, plasma TNF-alpha concentrations after endotoxin administration at 1, 2 and 3 days (14 350(2186), 26 375(2478) and 23 000(3745) pg/ml respectively; n = 6 each) were significantly higher than that before operation (9067(1559) pg/ml; n = 6) (P < 0.05), whereas those at 5 and 7 days (10 102(3616) and 8580(1427) pg/ml respectively; n = 6 each) showed no significant increase. Furthermore, nitric oxide production in peripheral blood and liver was suppressed by preoperative PVL.

CONCLUSION

Prevention of endotoxin-induced liver failure by preoperative PVL is associated with reduced production of TNF-alpha in the later phase of liver regeneration.

Measurement of serum hyaluronate as a predictor of human liver failure after major hepatectomy

Serum hyaluronate can be used as an index of hepatic sinusoidal endothelial cell function. This study was designed to evaluate its application as a predictor of liver failure after major hepatectomy. Thirty-six patients who underwent right liver lobectomy after percutaneous transhepatic right branch portal vein embolization were divided into two groups based on their postoperative clinical course (groups 1 and 2, with and without postoperative liver failure, n = 6 and n = 30, respectively). We serially measured serum hyaluronate levels using a sandwich binding protein assay system before and after hepatectomy and determined relations with progression of the underlying chronic liver disorder, portal venous pressure, and liver growth of the left lobe after portal embolization. Serum hyaluronate levels were significantly elevated, in line with the degree of severity of the underlying chronic liver disorder, and correlated well with the portal venous pressure and the hypertrophic ratio of the left lobe subsequent to portal embolization. Serum hyaluronate levels in group 1 were significantly higher than those in group 2 before surgery and increased steeply during the early period after hepatectomy. These results suggest that the serum hyaluronate reflects the hepatic functional reserve, and serial measurement of this parameter after hepatectomy can serve as a simple indicator for early detection of posthepatectomy liver failure.

Lidocaine-metabolizing activity after warm ischemia and reperfusion of the rat liver in vivo

The effect of warm ischemia on lidocaine-metabolizing activity was examined in vivo. Total liver ischemia was produced for 1 hr in Sprague-Dawley rats by clamping the portal vein and hepatic artery at the hilum. Livers were then reperfused, and liver microsomes were prepared before and 0, 2, 6, and 24 hr, and 3, 6, and 10 days after reperfusion. Microsomal lidocaine-metabolizing activity and cytochrome P-450 content were examined. Lidocaine N-deethylase activity was decreased from 2.25 +/- 0.33 to 0.97 +/- 0.21 nmol/mg protein/min (mean +/- SD) 24 hr after reperfusion. This inhibition was prolonged, and activity gradually recovered after 10 days. The cytochrome P-450 content showed the same tendency. On the other hand, serum levels of alanine aminotransferase increased significantly 2 hr after reperfusion and returned to control levels 3 days after reperfusion. Liver blood flow recovered rapidly after unclamping and reached baseline levels within 6 hr. Our results suggest that after warm ischemia, prolonged hepatic dysfunction in drug metabolism, which cannot be detected by evaluating serum enzymes or liver blood flow, exists at the microsomal level.

Recurrent mesenteric desmoid tumors with multiple peritoneal dissemination: a case report and review of desmoid in Japan

We report, herein, on the first case of a mesenteric desmoid tumor with multiple peritoneal dissemination. A 73 year-old Japanese woman, who had a history of uterine cancer that was treated with hysterectomy followed by a high dose of irradiation 25 years ago, had an unknown stenosis of the sigmoid colon, which was treated with partial resection of the stenosed colon 6 years ago, and then resulted in multiple small bowel obstructions due to the recurrence of mesenteric desmoids. The clinical behavior of this tumor is considered to be unpredictable. We emphasize that mesenteric desmoid tumors should be considered as one of the causes of stenosis of the colon and small bowel, and patients should receive careful follow-up after unknown stenosis.

Experimental study of the effect of intraportal prostaglandin E1 on hepatic blood flow during reperfusion after ischaemia and hepatectomy

BACKGROUND

Prostaglandin E1 (PGE1) has protective effects experimentally and clinically in individual models of hepatic ischaemia-reperfusion injury and of partial hepatectomy. The present study investigated the effects of intraportal administration of PGE1 on hepatic blood flow, systemic arterial pressure and long-term animal survival after 60 min of total liver ischaemia followed by 70 per cent partial hepatectomy in rats.

METHODS

Total liver ischaemia was induced by occluding the hepatoduodenal ligament for 60 min. PGE1 0.5 microg per kg per min was infused intraportally for 15 min before inducing ischaemia and for 120 min after ischaemia in the treatment group. Normal saline was infused in the control group. During ischaemia 70 per cent partial hepatectomy was performed. Portal venous flow (PVF), peripheral tissue blood flow (PTBF) and hepatic artery flow were measured before and after ischaemia. Serum biochemical analysis was carried out at 1, 3 and 24 h, and 7 and 14 days; and liver histology at 1 and 24 h, and 7 days after reperfusion. Survival was followed for 1 year.

RESULTS

Intraportal infusion of PGE1 significantly improved PVF and PTBF without affecting the systemic arterial pressure. Long-term survival was significantly higher in the PGE1 group. Serum aspartate aminotransferase, alanine aminotransferase and alkaline phosphatase levels decreased significantly, and 2-h bile flow was significantly improved, in the PGE1 group. Histological examination revealed significant portal venous congestion, sinusoidal congestion, fatty degeneration and tissue necrosis 24 h and 7 days after reperfusion in the control group.

CONCLUSION

PGE1 has a protective effect against liver damage when the liver is injured by warm ischaemia and reperfusion followed by partial resection.

[A case of Fabry's disease with chronic renal failure]

Fabry's disease is a genetic disorder caused by the absence of alpha-galactosidase (alpha-Gal), the gene of which is carried on the long arm of the X chromosome. This enzymatic defect leads to an accumulation of glycosphingolipids in the plasma and lysosomes of endothelial, perithelial, and smooth muscle cells, especially involving those of the cardiovascular, renal and cerebrovascular systems. We report one male case of Fabry's disease with renal deterioration. A 36-year-old man who was a classic case with acroparesthesia, angiokeratoma, and hypohidrosis from 10 years of age, was diagnosed to be a hemizygote of Fabry's disease at 27 years as a result of severe decreased alpha-Gal activity of his peripheral white blood cells. This patient was found to have a point mutation of a G to A transition in exon 1. In May, 1989, he was reported to have proteinuria with normal renal function and admitted to our hospital due to renal deterioration in September, 1993. Laboratory examinations revealed a serum urea nitrogen of 65 mg/dl and creatinine value of 6.9 mg/dl. Urinary protein excretion was 3.9 g/day and urinary sugar was negative. On the renal biopsy specimens, light microscopic examinations revealed multiple sclerosing and collaptic lesions in glomeruli without severe tubulo-interstitial damage, but with stenotic change of the small arteries and arterioles. Electron microscopic examinations revealed a large number of electron dense deposits in the tubules. We diagnosed this case as Fabry's disease with chronic renal failure, however the pathogenesis of this renal progressive deterioration remained obscure. In this case, degenerative changes in the renal vessels due to Fabry's disease may be associated with rapid deterioration in renal function.

Castleman disease of the pararenal retroperitoneum: report of a case

We describe herein the case of a 21-year-old woman in whom Castleman disease of the pararenal retroperitoneum was successfully resected. The patient was referred to our hospital from another hospital for investigation of a retroperitoneal mass in the right middle abdomen. Ultrasonography, computed tomography, and magnetic resonance imaging demonstrated a large retroperitoneal mass with heterogeneous imaging characteristics. An aortogram showed arterial feeding to this mass from a few lumbar arteries. Although a definitive preoperative diagnosis could not be made, surgical excision was performed and histopathological examination confirmed a diagnosis of the hyaline type of Castleman disease. The patient had an uneventful postoperative course and was discharged 14 days after her operation. She now leads an active social life without any signs of sequelae or recurrence 14 months later. To the best of our knowledge, only 2% (6/315) of all reported cases of Castleman disease have been located in the pararenal and retroperitoneal area.

The effects of intraportal administration of prostaglandin E1 on liver ischemia and hepatectomy in rats

The effects of intraportal administration of prostaglandin E1 (PGE1) on portal venous flow, hepatic arterial flow, peripheral tissue blood flow, and systemic arterial flow before and after 60 min total liver ischemia followed by 70% partial hepatectomy in rats were investigated. Total liver ischemia was induced by occluding the hepatoduodenal ligament for 60 min. PGE1 at a dose of 0.5 microg/kg/min was infused intraportally for 15 min before inducing hepatic ischemia (preischemic period) and for 60 min after ischemia (postischemic reperfusion period) in the treatment group. Normal saline was infused in the control group. Seventy percent partial hepatectomy was performed during ischemia. Serum biochemical analysis and liver tissue histology were carried out 1, 3, and 24 h, and 1 and 24 h after reperfusion respectively. One-week survival of the PGE1 group was improved to 70% compared to that of the control group of 30%. Postischemia reperfusion values of portal and peripheral tissue blood flows in the PGE1 group were 6.33 +/- 0.600 ml/min and 27.2 +/- 23.5 (arbitrary), and were significantly different from those of the control group of 4.34 +/- 0.400 ml/min and 23.5 +/- 5.54 (arbitrary), respectively. There was no significant difference in hepatic arterial flow between the two groups. Serum alkaline phosphatase decreased significantly in the prostaglandin group. Histological examination revealed a significant portal venous congestion in the control group 1 and 24 h after reperfusion. The extent of the sinusoidal congestion was also severe in the control group 24 h after reperfusion. It was concluded that PGE1 has a protective effect against liver damage when the liver was injured by warm ischemia and reperfusion followed by partial resection.

Modified hepatoduodenal ligamentectomy for advanced carcinoma of the biliary tract: the importance of preservation of the replaced left hepatic artery

Hepatoduodenal ligamentectomy (ligamentectomy) is the ultimate surgery for biliary tract carcinoma involving perioperative difficulties such as total hepatic ischemia during revascularization of the hepatic artery and the portal vein, patency of the reconstructed hepatic artery, and high incidence of related operative mortality. In the present study, modified ligamentectomies with extended right hepatic lobectomy, including resection of the caudate lobe, were performed on three patients with advanced biliary tract carcinoma in whom the left hepatic artery had been replaced and the original artery was preserved. In all patients, postoperative courses were uneventful: success of the resection was confirmed by histological examination. This procedure enabled en bloc resection of hepatoduodenal ligament with positive cancer invasion to take place. It was carried out safely without concern for the difficulties described above. In our view, ligamentectomy should be performed in all such cases.

Microsomal cytochrome P-450 monooxygenase system and its drug-metabolizing activity after partial portal vein ligation in the Rat

Percutaneous transhepatic portal vein embolization (PTPE) has been used to decrease the risk of hepatic failure after hepatectomy in patients with poor liver function. The effect of PTPE on hepatic drug-metabolizing activities is not clear. Therefore we examined the effect of portal vein branch ligation, a model of PTPE, on hepatic drug-metabolizing activities in Sprague-Dawley rats. Ligated and nonligated lobes were harvested separately. Drug-metabolizing activities and concentrations of components of the microsomal cytochrome P-450 monooxygenase system were examined. In ligated lobes, drug-metabolizing activities (lidocaine and aminopyrine) and enzymatic concentrations of the microsomal cytochrome P-450 monooxygenase system gradually decreased over 10 days. In nonligated lobes these functions were depressed rapidly to 60% of those before PBL but then recovered 10 days after PBL. From the viewpoint of drug metabolism, hepatic dysfunction occurred in both ligated and nonligated lobes.

Wound healing of intestinal anastomosis after digestive surgery under septic conditions: participation of local interleukin-6 expression

This study aimed to evaluate the integrity of anastomotic wound healing after digestive surgery under septic conditions and to observe local interleukin-6 (IL-6) expression around the anastomotic segment. Experimental animals were separated into lipopolysaccharide (LPS) and control groups. Each was injected with LPS or normal saline solution into the peritoneal cavity 24 hours before transection and anastomosis of the colon. The anastomotic bursting pressure (ABP) and tissue hydroxyproline concentration (HP) were measured as indicators of wound healing. Immunohistochemical staining for IL-6 was performed on tissue samples obtained from the anastomotic segment, lung, liver, and kidney. The reactive cells were counted by light microscopy. The ABP and HP were significantly lower in the LPS group than the control group 7 days after the surgery. In the LPS group, IL-6 expression around the anastomotic segment was enhanced 1 and 6 hours after surgery but suppressed 24 hours afterward. In contrast, IL-6 expression in lung, liver, and kidney was enhanced in the LPS group 24 hours after surgery but not in the control group. It is suggested that anastomotic wound healing is impaired after digestive tract surgery under septic conditions, and local IL-6 expression participates in wound healing.

Morphological and functional alterations to sinusoidal endothelial cells in the early phase of endotoxin-induced liver failure after partial hepatectomy in rats

Liver failure following major hepatectomy is characterized pathologically by massive hepatic necrosis, which is thought to begin with injury of sinusoidal endothelial cells (SECs). To examine the early events of SECs leading to hepatic damage, we performed time-course analyses of the morphological and functional perturbation of SECs after endotoxin administration to hepatectomized rats. At 1.5 h after endotoxin injection, when hepatocellular damage was not yet evident, SECs showed augmented expression of intercellular adhesion molecule-1, with frequent adherence of infiltrating leucocytes and ultrastructural features of defenestration and hypertrophied cytoplasm enriched with cell organelles. The serum level of hyaluronate, as an indicator of the functional state of SECs, was significantly elevated. At 3 h, SECs underwent necrosis and disruption, accompanied by fibrin deposits with concomitant hepatocellular necrosis. The morphological and functional alterations of SECs precede necrotic changes in hepatocytes and SECs in endotoxin-induced liver failure after partial hepatectomy.

Treatment of recurrent hepatocellular carcinoma by hepatectomy with right and middle hepatic vein reconstruction using total vascular exclusion with extracorporeal bypass and hypothermic hepatic perfusion: report of a case

We report herein the case of a patient who had previously undergone a lateral segmentectomy for hepatocellular carcinoma (HCC) in whom recurrent HCC invading the trunk of the right and middle hepatic veins in a damaged liver was treated by reconstruction of both hepatic veins, using total vascular exclusion with extracorporeal bypass and hypothermic hepatic perfusion. Reconstruction was performed using a graft taken from the left external iliac vein and divided into two pieces. Hepatic ischemia lasted for 91 min during the procedure and the intrahepatic temperature, as monitored by inserting a needle-type thermometer, was decreased to 11 degrees C throughout the procedure. The peak levels of serum glutamic pyruvic transaminase, lactate dehydrogenase, and total bilirubin were 363 IU/l, 1198 IU/ml, and 2.8 mg/dl, respectively, on postoperative day (POD) 2. The patient's postoperative course was uneventful except for mild, temporary swelling of the left leg. Postoperative computed tomography and magnetic resonance imaging examinations disclosed no obstruction of either graft, and the patient was discharged on POD 40.