Comparison of 3 Explant-Based Prognostic Models as Predictors of Hepatocellular Carcinoma Recurrence After Liver Transplantation: Analysis of Our Experience.

Tumor load is often underdiagnosed on radiological examination previous to liver transplantation (LT) for hepatocarcinoma (CHC). Thus, post-liver transplant explant analysis is required following transplantation to assess the risk of the recurrence of CHC. The objectives were to compare the characteristics of CHC on pre-LT radiological examination and explant histology and validate three models for the prediction of recurrence based on data from a cohort of patients treated in our hospital. METHODS: A retrospective study was undertaken of 105 LTs for CHC performed in our unit between January 2006 and January 2015. The minimum follow-up was five years. The preoperative radiological tumor stage was compared to the explant-based histologic stage. Three prognostic models were validated using our cohort of patients. RESULTS: Following Milan's criteria, the tumor load was underdiagnosed on pre-LT radiological examination in 20 patients, which accounted for 19% of the total sample. The 5-year overall recurrence was 6.6% for scores <4 and 33.3% for scores ≥4 according to Decaens' model; 7% for scores ≤7 and 25% for scores >7 in the Up-to-Seven model; and 3.6% for PCRS ≤0, 27.8% for PCRS1-2, and 100% for PCRS≥3 according to Chan's model. The predictive model for 5-year recurrence after LT with the greatest area under the curve was Chan's model (0.813 [95% CI: 0.650-0.977]) versus Decaens' model (0.674 [95% CI: 0.483-0.866]) and the Up-to-Seven model (0.481 [95% CI: 0.296-0.667]). CONCLUSIONS: A pre-LT radiological examination leads to the underdiagnosis of tumor load, and the risk for recurrence must be recalculated following LT. In light of the results obtained, Chan's model is more accurate in predicting 5-year recurrence of CHC post-LT based on 3 levels of risk. New prognostic models are needed to optimize the prediction of recurrence after liver transplantation for hepatocarcinoma.

Combined Flush With Histidine-Tryptophan-Ketoglutarate and University of Wisconsin Solutions in Liver Transplantation: Preliminary Results.

INTRODUCTION: Ischemia reperfusion injury (IRI) is the main cause of early allograft dysfunction (EAD) and subsequent primary allograft failure (PAF). OBJECTIVES: The purpose of this study is to compare IRI, EAD, and PAF in liver transplantation in a cohort of patients perfused with histidine-tryptophan-ketoglutarate (HTK) solution and University of Wisconsin (UW) solution versus HTK alone. METHODS: A randomized trial was performed to compare outcomes in liver recipients who underwent transplantation surgery in the University Regional Hospital of Malaga, Spain. Forty patients were randomized to two groups. Primary endpoints included IRI, EAD, PAF, re-intervention, acute cellular rejection, retransplantation, arterial complications, and biliary complications at postoperative day 90. RESULTS: Postoperative glutamic oxaloacetic transaminase (1869.15 ± 1559.75 UI/L vs. 953.15 ± 777.27 UI/L; P = .004) and glutamic pyruvic transaminase (1333.60 ± 1115.49 U/L vs. 721.70 ± 725.02 U/L; P = .023) were significantly higher in patients perfused with HTK alone. A clear tendency was observed in recipients perfused with HTK alone to present moderate to severe IRI (7 patients in the HTK + UW solution group vs. 15 patients in the HTK-alone solution group; P = .06), EAD (0 patients in the HTK + UW solution group vs. 0 patients in the HTK-alone solution group; P = .76), and PAF (3 patients in the HTK + UW solution group vs. 8 patients in the HTK-alone solution group; P = .15). CONCLUSIONS: Initial perfusion with HTK solution followed by UW solution in liver transplantation improves early liver function as compared to perfusion with HTK alone.

Up-to-7 Criteria for Hepatocellular Carcinoma Liver Transplantation: A Retrospective Analysis of Experiences.

INTRODUCTION: The expansion of criteria for hepatocellular carcinoma (HCC) liver transplantation should produce satisfactory outcomes in terms of survival and recurrence. OBJECTIVES: To investigate if the up-to-7 criteria are applicable to liver transplantation for HCC. METHODS: A review of all liver transplantations performed at our unit between January 2002 and December 2010 was conducted (645 patients). The 91 patients of the sample who had HCC were divided into 3 groups: in Milan criteria (MC; n = 74), in up-to-7 criteria (UTSC; n = 12), and outside of up-to-7 criteria (OUTSC; n = 5). A descriptive retrospective study was carried out to analyze the characteristics of liver tumors and recipients and to estimate recurrence and survival rates for this population of patients. RESULTS: The characteristics of transplant recipients of the 3 groups were comparable. Statistically significant differences were observed in the number of tumors (1 ± 0.65 for MC, 3 ± 1.05 for UTSC, 6 ± 4.10 for OUTSC; P < .001), largest tumor size (2.47 ± 1.12 cm for MC, 3.78 ± 0.04 cm for UTSC, 4.04 ± 1.73 cm for OUTSC; P < .001), and recurrence (5.4% for MC; 33.3% for UTSC; 20% for OUTSC; P = .008). Survival rates (MC, UTSC, and OUTSC) at 3 and 5 years were 71.6%, 66.7%, and 60%, and 58.1%, 58.3%, and 40%, respectively, whereas tumor-free survival rates were 70.3%, 58.3%, and 60%, and 58.1%, 50%, and 40%, respectively. CONCLUSIONS: Survival in patients with HCC transplanted under up-to-7 criteria is acceptable. However, the expansion of criteria involves an increase in the number of patients included in the waiting list and a higher probability of relapse.

Acute Liver Failure Caused by Primary Non-Hodgkin's Lymphoma of the Liver.

INTRODUCTION: Acute liver failure (ALF) is a rare syndrome involving maximum liver dysfunction. This disease is characterized by a less than 26-week history of coagulopathy (INR ≥1.5) and hepatic encephalopathy and generally occurs in patients without any previously known disease. METHODS: We report the case of a healthy 25-year-old subject who presented with fulminant liver failure caused by a primary non-Hodgkin's lymphoma of the liver that required emergency liver transplantation. Diagnosis was based on pathologic confirmation of T-cell/histiocyte-rich large B-cell lymphoma and submassive hepatocyte necrosis. One year after surgery, the patient remains in complete remission. CONCLUSIONS: Fulminant liver failure is a sudden-onset severe disease that can be caused by a primary non-Hodgkin's lymphoma of the liver, which accounts for <1% of extranodal lymphomas. The diagnosis of this rare disease demands high diagnostic suspicion, and progression can be prevented through liver transplantation.

Results of Liver Transplantation With Donors Older than 75 Years: A Case-Control Study.

BACKGROUND: The inclusion of elderly donors can increase the pool of organs available for transplantation. The objective of this study was to compare clinical outcomes and survival rates of patients who received livers from donors aged ≥75 years versus younger donors. METHODS: We considered all liver transplantations performed in our unit from January 2006 to January 2015. Thirty-two patients received a liver from a cadaveric donor aged ≥75 years (study group), and their outcomes were compared with those of patients who received a liver from a younger donor (control group) immediately before and after each transplantation in the study group. This is a descriptive, retrospective, case-control study carried out to analyze the characteristics of donors and recipients as well as the clinical course and survival of recipients of older and younger donors. RESULTS: Statistically significant differences were observed according to donors' age (53.3 ± 13.6 vs 79 ± 3.4 years; P < .001). In total, 6.2% of the recipients of a liver from a donor aged <75 years required retransplantation versus 15.6% of recipients of donors ≥75 years. Patient survivals at 1, 3, and 5 years, respectively, were 89%, 78.6%, and 74.5% for recipients of donors <75 years versus 83.4%, 79.4%, and 59.6% for the study group. CONCLUSIONS: Livers from older donors can be safely used for transplantation with acceptable survival rates. However, survival rates are lower for recipients of livers from older donors compared with younger donors, and survival only increased with retransplantation.

Everolimus plus mycophenolate mofetil as initial immunosuppression in liver transplantation.

BACKGROUND: The purpose of this study was to assess the efficacy and safety of a de novo immunosuppressive regimen with everolimus (EVL) plus mycophenolate mofetil (MMF) without calcineurin inhibitors (CNI) for liver transplantation. The secondary purpose was to compare the renal function with a control group of patients treated with tacrolimus plus MMF. METHODS: Sixteen male and 4 female liver transplant patients received immunosuppression with EVL plus MMF without CNI, with induction with steroids and 16 with basiliximab also. In 10 cases it was indicated as induction immunosuppression without CNI as prevention against nephrotoxicity and neurotoxicity or recurrence of hepatocarcinoma in predisposed patients and in another 10 after withdrawing CNI during the immediate post-transplant period, before hospital discharge, as the result of toxicity, mainly nephrotoxicity and neurotoxicity or the presence of hepatocarcinoma with a high risk of recurrence. A control group comprising 31 patients taking tacrolimus plus MMF was included to compare the renal function. RESULTS: The mean follow-up time was 24 months. One patient had a recurrence of hepatocarcinoma at 8 months after transplant. The cases of nephrotoxicity and neurotoxicity resolved favorably. There were 7 rejections (35%); 2 evolved to chronic rejection with both needing retransplantation, 2 resolved with dose adjustment, and 3 required conversion to CNI. The side effects were hyperlipidemia (25%), wound dehiscence (10%), lymphedema (10%), cytomegalovirus infection (25%), myelotoxicity (25%) and proteinuria >1 g in 1 case (5%). No differences were found in renal function between the two groups. CONCLUSIONS: This regimen was proven to be efficient to prevent and treat nephrotoxicity and neurotoxicity with an acceptable tolerability profile. However, the high associated rejection rate indicates that great caution is required in its use during the immediate post-transplant period. It is advisable to associate the regimen with low doses of CNI and to have agile methods available to monitor EVL to enable rapid dose adjustment.

Liver transplantation without abdominal drainage.

This observational cohort compared 70 consecutive liver transplantations (OLT) with no intra-abdominal drain and 70 control subjects C with an intra-abdominal drain who were operated immediately prior to them. We sought to assess the impact of abdominal drainage on the diagnosis and prevention of early postoperative complications of hemoperitoneum, reinterventions, biliary leaks or percutaneous drainage. We assessed variables related to the recipient (age, indication, pretransplant ascites, body mass index, Model for End-stage Liver Disease score, and rejection episodes, to the donor (age, steatosis and, ischemia time) as well as intra- and postoperative factors (surgery time, blood product use, and coagulopathy). The endpoint was defined as the need for a reintervention, postoperative paracentesis, appearance/drainage of collections, as well as lengths of hospital and intensive care unit (ICU) stays. Postoperative ICU and in-hospital stay were similar between the groups (3.6 versus 3.7 days and 12 versus 14 days respectively). Six patients in the drainage group were reoperated due to hemoperitoneum, whereas it was one in the cohort without drainage. Three patients presented a biliary fistula, two in the group without drainage, and one in the drainage group. One patient in the drainage group required percutaneous drainage of an intra-abdominal collection. The need for postoperative paracentesis was greater among the group without drainage (30% versus 6%; P < .008) and among those with a preoperative ascites > 1000 mL (38%). Patients with drainage displayed a greater incidence of perihepatic hematomas upon ultrasound (50% versus 22%, P < .008) and required more postoperative blood products, especially plasma (P < .01). In conclusion, OLT without intra- abdominal drainage is safe and does not increase morbidity. It seems likely that drainage may be responsible for intra-abdominal hematomas and greater consumption of blood products.

Is liver transplantation without abdominal drainage safe?

UNLABELLED: This observational, analytical cohort consisted of 35 consecutive liver transplant (OLT) patients with no intra-abdominal drain and a control cohort of 35 subjects operated immediately before the former who had placement of an intra-abdominal drain. We sought to assess the impact of abdominal drainage on the diagnosis and prevention of early postoperative complications: hemoperitoneum, reinterventions, biliary leaks, or percutaneous drainage. We assessed variables related to the recipient (age, indication, pretransplant ascites, body mass index, Model for End-Stage Liver Disease score and rejection), the donor (age, steatosis, ischemia time) and intra- and postoperative factors (surgery time, blood product use, and coagulopathy). The end point was defined as the need for a reintervention, paracentesis, appearance, and drainage of collections as well as lengths of hospital and intensive care unit (ICU) stays. The postoperative ICU and in-hospital stays were similar between groups (3.7 vs 3.9 days and 12 vs 14 days, respectively). Two patients in the group with drainage were reoperated due to hemoperitoneum, whereas we did not reoperate any patients in the group without drainage. No patient from either group developed a biliary fistula or required drainage of an intra-abdominal collections. The need for paracentesis was greater among the group without drainage (23% vs 5.7%; P < .04) and among those with a prior history of severe ascites. Patients with drainage displayed a greater incidence of perihepatic hematomas by ultrasound (53% vs 21%; P < .08) and required more postoperative blood products, especially platelets (P > .04) and plasma (P < .01). CONCLUSION: OLT without intra-abdominal drainage is safe, not increasing morbidity. It seems likely that drainage may be responsible for intra-abdominal hematomas and greater consumption of blood products.

An alternative method of reconstruction of hepatic venous outflow in domino liver transplantation.

The double piggyback technique has been proposed for domino liver transplantation. To make this possible, it is necessary to reconstruct the venous outflow of the domino liver graft on the back table. We describe an alternative method of reconstruction of hepatic venous outflow, in which a neocaval segment is obtained using both common iliac veins from the cadaveric donor.

Liver transplantation for hepatocellular carcinoma: results of a multicenter study with common priorization criteria.

OBJECTIVE: To evaluate the results of liver transplantation (OLT) performed for hepatocellular carcinoma (HCC) among a multicenter cohort of patients with predefined common inclusion and priorization criteria. PATIENTS AND METHODS: Over a 5-year period (January 2002-December 2006), 199 HCC patients underwent OLT in four centers in Andalusia. The morphological (Milan) inclusion criteria were priorized in two consecutive periods, according to the Model for End-stage Liver Disease score: group I, 53 patients (HCC < 2 cm = 24 points; > or = 2 cm or multinodular = 29 points) and group II, 146 cases (HCC < 3 cm without priorization; HCC > or = 3 cm or multinodular = 18 points). RESULTS: Among the 199 HCCs, 186 (93.5%) subjects were transplanted and 13 (6.5%) were excluded. There were 18 cases (9.7%) where the diagnosis was incidental and 168 were known HCC cases; 144 (85.7%) complied with the Milan criteria (Milan+); 24 (14.3%) exceeded there criteria (Milan-). According to preoperative imaging, the number of nodules and tumor mean sizes among the excluded-Milan+ and Milan- groups-were 1.8/5.3 cm, 1.4/3.5 cm, and 2.3/6.7 cm, respectively (P < .001). Percutaneous treatment during listing was delivered to 55% of the excluded cases: 49% of Milan+ and 96% of Milan-. The median time on the list was 88 days for known HCC (53 days for group I, and 97 days for group II), and 172 days for the incidental HCCs. Staging (pTNM) was correct in 64% of cases: 23% were understaged and 13% were overstaged. Overall mortality within the first 90 days was 9%, and transplant patient survival at 5 years was 61%. No differences were observed in survival rates between both study periods, although there were differences between the Milan+ (65%) and Milan- (23%) groups (P < .04). In addition, the difference in the recurrence rates was also significant between the Milan+ (7%), Milan- (24%), and the incidental (25%) groups (P < .02). CONCLUSIONS: A common priorization policy of HCC for OLT based on morphological criteria results in a low exclusion rate on the waiting lists (6.5%). The Milan criteria are still a good cutoff to stratify the risk of recurrence, despite preoperative tumor staging being correct in only two-thirds of cases.

Impact of donor age on the results of liver transplantation in hepatitis C virus-positive recipients.

OBJECTIVE: Hepatitis C virus (HCV)-cirrhosis is the most frequent indication for orthotopic liver transplantation (OLT) among adults in most European and American transplant centers. The aim of this study was to analyze the impact of donor age on graft survival among HCV-positive cirrhotic transplant patients. MATERIALS AND METHODS: We performed an observational, retrospective study between March 1997 and December 2004, analyzing 340 liver transplantations. The patients were divided into 4 groups, considering whether the HCV infection was the indication for OLT and whether the age of the donor was older or younger than 48 years: group 1 (HCV, <48 years); group 2 (HCV, >48 years); group 3 (non-HCV, <48 years); and group 4 (non-HCV, >48 years). RESULTS: A univariate analysis showed that posttransplantation graft survival was clearly influenced by recipient HCV serologic status (P = .018). However, no graft survival differences were found when the analysis variable was age (>48 or <48 years). When both variables were studied, a positive HCV serology did not modify graft survival when the donor age was <48 years (P = .32), but had a statistically significant negative impact when the age was >48 years (P = .02). CONCLUSIONS: The use of older donors for HCV recipients resulted in worse graft and patient survivals in our study. This difference in survival was not present in non-HCV recipients or when grafts for HCV recipients were procured from younger donors. Donor age <30 years was a protective factor for graft survival among HCV recipients.

True impact of the indication of cirrhosis and the MELD on the results of liver transplantation.

UNLABELLED: Our Aim was to determine the impact of cirrhosis and the preoperative MELD score on the immediate postoperative mortality and hospital stay as well as survival at 1, 5, and 8 years in liver transplantation. MATERIALS AND METHODS: Transplanted cirrhotic patients were selected who did not display some of the main known risk factors affecting recipient. Donor and surgical technique were included in this analysis. These exclusion criteria for recipient factors were emergency transplants and retransplants; for donor factors, age over 60 years, ischemia time over 10 hours, and moderate or severe steatosis on back-bench biopsy; and for surgery, prior complex upper abdominal surgery (mainly derivative and gastroduodenal surgery). Among 340 total liver transplants including 16 retransplants performed from March 1997 to December 2005, 197 patients met the selection criteria. The mean age of the recipients was 52 years (17-67) and the donors, 39 years (11-60). The transplant indication was cirrhosis in all cases: HCV in 69 cases (35%); alcohol in 55 (28%); hepatocarcinoma in 38 (19%); HBV in 19 (10%); PBC in 8 (4%), and other etiologies in 8 cases (4%). The MELD scores were divided as group 1, <10 points (33 cases = 17%); group 2, 10 to 18 points (136 cases = 69%); and group 3, >18 points (28 cases = 14%). The statistical analysis was performed with SPSS 11.0. RESULTS: Postoperative mortality (up to 3 months) was 16 cases (8%). The median ICU and hospital stays were 3 and 13.5 days, respectively. Overall survivals at 1, 5, and 8 years were 89%, 80%, and 77%, respectively. The survival for the same periods according to MELD group was 97%, 97%, and 97% for group 1; 87%, 76%, and 72% for group 2; and 85%, 81%, and 81% for group 3 (P = NS). The survival according to the three main indications at 1, 5, and 8 years was: HCV, 91%, 80%, and 80%; alcohol, 87%, 80%, and 71%; and hepatocarcinoma, 84%, 80%, and 80% (P = NS). No significant differences were observed among early deaths between MELD groups or transplant indications. CONCLUSIONS: In a favorable liver transplant setting including acceptable donors, absence of prior complex abdominal surgery in the recipient, and nonemergency transplants, neither the cause of the cirrhosis nor its severity, as measured preoperatively by the MELD, were predictive of early postoperative death or long-term survival.

Transfusion requirements during liver transplantation: impact of a temporary portacaval shunt.

UNLABELLED: We evaluated the consumption of blood products during liver transplantation in cirrhotic patients association with the placement of a temporary portacaval shunt (TPCS). PATIENTS AND METHODS: We retrospectively divided 349 cirrhotic patients transplanted in our unit between March 1997 and October 2005 into two groups: transplants without a TPCS (group I, 189 cases) and those with a TPCS (group II, 160 cases). In all cases, we preserved the inferior vena cava (piggyback). The dependent variables were consumption of blood-derived products (banked red cells, recovered red cells, fresh frozen plasma, platelets), surgery time, kidney function, intensive care unit stay, and hospital stay. RESULTS: Consumption of blood products was significantly lower among patients who received a TPCS. In group II, no platelet transfusion was required in 54% of the patients, and no banked red cells in 12% compared with 18% and 3%, respectively, among group I patients (P < .005). The mean overall transplant procedure time was 74 minutes shorter in group II (361 minutes) compared with group I (435 minutes) (P < .001). The overall hospital stay was shorter among patients transplanted after TPCS. CONCLUSION: Liver transplantation with a TPCS was accompanied by a reduction in the intraoperative use of blood-derived products, especially platelet transfusion. Among other advantages, this reduction resulted in a shorter posttransplant hospital stay.

Liver transplant results for hepatocellular carcinoma applying strict preoperative selection criteria.

INTRODUCTION: Liver transplantation is currently the best therapeutic option for small hepatocellular carcinoma (HC) in selected cirrhotic patients. The main aim of this study was to analyze the results of a recent series of liver transplant cirrhotic patients with small HC applying strict preoperative selection criteria. PATIENTS AND METHODS: During a period of 6 years we performed 53 liver transplants with a final diagnosis of HC on cirrhosis. The selection criteria for liver transplantation (LT) by modern imaging techniques were the Milan criteria (TNM I and II of the modified classification). RESULTS: Of the 53 patients, 44 (83%) were transplanted with preoperatively known HC, and 9 (17%) with incidental HC. The mean time on the waiting list was 74 +/- 62 days. Despite using strict selection criteria, 23 patients (43%) exceeded the Milan criteria in the specimen and 17 (32%) even exceeded the extended criteria of the UCSF. With a mean follow-up of 2 years, only two patients have developed recurrences. The overall survival at 1, 3, and 5 years was 80%, 70%, and 70%, respectively. The survival of patients that exceeded the Milan or USF criteria at 1, 3, and 5 years was 72% and 76%; 67% and 69%; 67% and 69%, respectively. CONCLUSIONS: The results of liver transplantation for HC are excellent when applying strict preoperative selection criteria. The current imaging methods lead to a considerable infrastaging percentage (30% to 40%), extending the indications for liver transplant due to HC beyond the scope that clinical reports would justify.

[Pancreatic cystic metastasis from pulmonary carcinoma. Report of a case]. / Metastasi cistica pancreatica da carcinoma polmonare. Presentazione di un caso.

We report the case of a patients with a metachronous cystic pancreatic metastasis from an undifferentiated large cell lung carcinoma two years after the primary tumor had been surgically removed. Clinically, he presented with epigastric pain, fever, weakness and anorexia. The patient was operated and a palliative cystogastrostomy was performed after an intraoperative biopsy had been informed as positive for carcinoma. Six months later the patient died. Pancreatic metastases from lung carcinoma are found in approximately 7-9% of patients deceased of this neoplasm. Clinical and radiological findings simulate primary pancreatic tumors, being epigastric pain, jaundice and upper digestive bleeding the most frequent symptoms. They represent stages of advanced systemic disseminated tumoral disease, and because of this reason total or partial surgical curative resections will only be performed in a few cases of patients with isolated metastasis, criteria of resectability and without evidence of extended disease to other organs or systems. In the most of the cases, the treatment will only be palliative, even medical or surgical.

Use of the breast implant for liver graft malposition.

Transplantation of a small liver into a large patient may cause problems with correct reperfusion of the graft because of torsion of the hepatic pedicle, leading to malfunction of the transplanted organ. We describe the case of a 60-year-old man with alcoholic cirrhosis and hepatitis B virus who received a small-sized liver graft. Owing to the lack of adequate reperfusion of the transplanted organ arising from pedicular kinking caused by disparate sizes, a breast implant was placed behind the graft as a means of support, thereby resolving the problem. The use of prosthetic materials for the correct placement of grafts with size incongruity is also discussed, and the literature is reviewed.

[Shunt surgery versus disconnection in portal hypertension. A comparative study]. / Cirugía derivativa versus de desconexión en hipertensión portal. Estudio comparativo.

Controversy remains about the best treatment for patients with esophagogastric variceal bleeding. In spite of different therapeutic alternatives and recent progress, such as liver transplantation of TIPS, many patients will finally be treated by standard surgery. With the aim to know the results of surgery in shunting vs non-shunting procedures, we have analyzed a recent and consecutive series of 68 cirrhotics patients operated on for variceal hemorrhage. According to the surgical technique there were three groups: I) 30 patients underwent a total portacaval shunt; II) 18 cases with an azygosportal disconnection; III) 20 patients with a partial portacaval shunt (8-10 mm H-portacaval PTFE graft). All groups were homogeneous considering age, sex etiology, Child-Pugh grade and timing of surgery (elective vs urgent). The complications and mortality rates were similar for the three groups. The overall operative mortality was 10%, and 5% in selected cases (Child-Pugh A-B, non urgent cases). With a follow-up for I, II and III group of 47, 44 and 27 months respectively, chronic encephalopathy have been seen in 61%, 15% and 15% respectively (p < .05). No patient in group I has rebled, and only one case in the II and III groups had a recurrence of hemorrhage. The actuarial three years survival was 76%, 86% and 85% (p NS), and the five year survival was 41% vs 86% for the groups I and II respectively (p < .05). In conclusion, in selected cases, surgery gives excellent immediate and long term results. In patients with variceal bleeding and failure of first line treatments, such as sclerosis or pharmacology, the partial H-portacaval shunt and non-shunting procedures are good alternatives, with low incidence of rebleeding and chronic encephalopathy.