Computed tomographic scans showing a huge tumor consisting of solid and cystic components. A, Computed tomographic scan showing a huge tumor involving the left lobe and the anterior segment of the right lobe. B, Tumor consisting of solid and cystic components compresses the right hepatic vein. C, Caudal part of the tumor is adjacent to the right portal pedicle.
Ishizaki Y, Yoshimoto J, Miwa K, Sugo H, Kawasaki S. Safety of Prolonged Intermittent Pringle Maneuver During Hepatic Resection. Arch Surg. 2006;141(7):649-653. doi:10.1001/archsurg.141.7.649
It has recently been demonstrated that the liver parenchyma is more tolerant to intermittent pedicular clamping than to continuous pedicular clamping. However, the possibility of increased blood loss during intermittent reperfusion is a major concern.
We retrospectively selected 34 cases in which the cumulative clamping time was 90 minutes or longer during hepatectomy and the intermittent Pringle maneuver was applied rather than continuous inflow occlusion.
Resections were performed for metastatic carcinoma in 19 patients, hepatocellular carcinoma in 7 patients, hilar bile duct carcinoma in 3 patients, intrahepatic cholangiocarcinoma in 1 patient, combined hepatocellular carcinoma and cholangiocarcinoma in 1 patient, undifferentiated embryonal sarcoma in 1 patient, carcinoid tumor in 1 patient, and benign mucinous cystic tumor in 1 patient. Patients were categorized on the basis of the cumulative clamping time, with 25 patients in group 1 (≤120 minutes) and 9 patients in group 2 (>120 minutes). In 2 patients in group 2, the cumulative clamping time exceeded 240 minutes. Twenty-eight patients had histologically normal underlying liver parenchyma; 6 patients had chronic hepatitis or cirrhosis.
No red blood cell transfusions were required in group 1; blood transfusions were needed for only 3 patients in group 2. There was no postoperative mortality or major complications. The rate of minor postoperative complications was 36% (9 patients) in group 1 and 22% (2 patients) in group 2.
Prolonged intermittent pedicular clamping is a useful maneuver in hepatectomy when resection is difficult or prolonged or when the liver parenchyma is abnormal. Such clamping can be used for cumulative periods exceeding 120 minutes without major intraoperative blood loss or complications.
Although most liver resections are performed with an inflow occlusion time of less than 90 minutes, longer periods of hepatic ischemia have sometimes been required to accomplish complicated and extensive hepatectomies for malignant neoplasms, such as multiple resections, caudate lobectomy, or hepatectomies with vascular reconstruction. The intermittent Pringle maneuver (IPM) has been applied to avoid ischemic injury induced by prolonged inflow occlusion.1,2 However, many Western surgeons are still reluctant to use intermittent pedicular clamping because of the associated risk of increased blood loss during the reperfusion period.3,4 A clinical controlled trial demonstrated that the liver parenchyma is more tolerant to intermittent pedicular clamping than to continuous pedicular clamping.5 In these studies1,2,5 of intermittent clamping, a total occlusion time of less than 90 minutes was used, and the safe upper time limit for total occlusion was not defined. The present study was performed to evaluate the course of patients who underwent hepatectomy with a cumulative intermittent inflow occlusion time of 90 minutes or longer to determine the benefits of such clamping and the tolerance of the liver parenchyma to clamping.
Between October 1, 2002, and December 30, 2004, 157 patients underwent hepatic resection using the IPM. We retrospectively selected 34 patients in whom the cumulative clamping time was 90 minutes or longer. There were 27 men and 7 women, with a mean age of 61.1 years (age range, 21-84 years). In the remaining 123 patients, the cumulative ischemic time was less than 90 minutes.
Resections were performed for malignant hepatic tumors in 33 patients and for a benign mucinous cystic tumor in 1 patient. The malignant tumors included 19 metastatic carcinomas, 7 hepatocellular carcinomas, 3 hilar bile duct carcinomas, 1 intrahepatic cholangiocarcinoma, 1 combined hepatocellular and cholangiocarcinoma, 1 undifferentiated embryonal sarcoma, and 1 carcinoid tumor.
The liver parenchyma was abnormal in 6 of 34 patients because of chronic hepatitis or cirrhosis associated with hepatitis B or C viral infection. The 3 patients with hilar bile duct carcinoma had obstructive jaundice and underwent biliary drainage before liver resection. Preoperative portal vein embolization was carried out in 2 patients whose ratio of the nontumorous parenchymal volume of the resected liver to that of the whole liver exceeded 60%. Major resections (≥1 sector) were performed in 27 patients (79.4%), multiple limited resections in 4 patients (11.8%), and single segmentectomy in 3 patients (8.8%) (Table 1). Twelve patients with small, bilateral, and multiple hepatic tumors required major hepatic resection with additional limited resection of the remnant liver.
Blood tests were performed before surgery in all patients and included complete blood cell count, coagulation profile with prothrombin time, indocyanine green retention rate at 15 minutes, and routine liver biochemical tests, including the serum levels of total bilirubin, aspartate aminotransferase (AST), and alanine aminotransferase (ALT).
J-shaped and inverted T-shaped incisions were used. The operative field was kept open using Kent retractors (Takasago Co Ltd, Tokyo) throughout the operation. In some cases, the J-shaped incision was extended to the ninth intercostal space. Under intraoperative ultrasonographic guidance, the area to be resected was defined, the transection line was monitored, and the tumor was removed.
The IPM was applied at the time of liver transection and consisted of cross-clamping the hepatoduodenal ligament using a Satinsky clamp (and the aberrant left hepatic artery, if present) for 15 minutes and releasing the clamp for 5 minutes until the liver resection was completed. Cooling of the liver with topical refrigeration or hypothermic perfusion was not used. Liver transection was performed in all patients using the traditional clamp crushing method. A small amount of the hepatic parenchyma was divided by penetration using a Péan forceps. The remaining side of the parenchyma was ligated and divided using scissors; any bleeding or bile leakage points were fine sutured.
Postoperative variables of hepatocyte damage and recovery, including serum AST, ALT, and total bilirubin levels, were measured every day during the first week after the operation and twice during the second week. All accumulated abdominal fluids were drained percutaneously and sent for bacterial culture; the serum total bilirubin level was monitored to detect any biliary leakage. We diagnosed biliary leakage when the drainage fluid was found to contain a serum total bilirubin concentration of more than 3 times the upper normal limit in serum after surgery. The postoperative hospital stay was defined as the period between surgery and discharge.
The overall mean ± SD duration of cumulative clamping in the 34 study subjects was 120 ± 48 minutes (range, 90-325 minutes). These 34 patients were divided into 2 groups according to the cumulative period of occlusion. Patients in group 1 (n = 25) underwent inflow occlusion for 90 to 120 minutes (mean ± SD cumulative clamping time, 99.4 ± 8.4 minutes). Patients in group 2 (n = 9) underwent inflow occlusion for longer than 120 minutes (mean ± SD cumulative clamping time, 176.1 ± 68.0 minutes).
The 2 groups were compared using the bilateral t test for quantitative variables and the χ2 test or Fisher exact test for qualitative variables. Continuous variables were expressed as mean ± SD, and differences at P<.05 were considered statistically significant. Calculations were made using StatView computer software (SAS Institute Inc, Cary, NC).
Groups 1 and 2 were matched for age, sex, preoperative liver function test data, status of underlying liver disease, and number of viral hepatitis infections (Table 2). The 2 patients who underwent extended right lobectomy after preoperative portal vein embolization were fitted into group 1. In 2 patients, the cumulative clamping time exceeded 240 minutes. In the patient with the longest cumulative period of hepatic ischemia (325 minutes), the IPM was repeated 19 times to accomplish complete resection of a huge carcinoid tumor situated close to the hepatic veins and inferior vena cava (Figure). The intraoperative blood loss was 3820 mL during left trisegmentectomy, and the intraoperative blood transfusion volume was 1390 mL. The other patient, who had 11 bilobar hepatic metastatic tumors, underwent left hepatectomy and 8 limited resections. The IPM was repeated 16 times to accomplish complete resection of the multiple tumors, and the cumulative period of hepatic ischemia was 250 minutes. The intraoperative blood loss was 770 mL, and no blood transfusion was given. No obvious visceral congestion was recognized during the period of clamping in either of the patients. The postoperative course of both patients was uneventful.
Overall, the mean ± SD intraoperative blood loss was 1022 ± 687 mL (range, 175-3820 mL). The mean ± SD intraoperative blood loss in group 1 (883 ± 461 mL) was significantly lower than that in group 2 (1409 ± 1039 mL). We used the hematocrit as an indicator of transfusion requirements and carried out transfusion when the hematocrit was lower than about 25%. Red blood cell transfusion was necessary in only 3 patients (33.3%) in group 2. No patient in group 1 required red blood cell transfusion. No patient required postoperative red blood cell transfusions in either group. The mean ± SD duration of surgery in group 1 (179 ± 49 minutes) was significantly shorter than that in group 2 (665 ± 237 minutes). Group 2 included 3 patients with hilar bile duct carcinoma who underwent extended lobectomy. These patients required concomitant biliary reconstruction. In the 123 patients in the original cohort with a cumulative clamping time of less than 90 minutes, the mean ± SD blood loss during surgery was 450 ± 336 mL, and 3 (2.4%) of these patients had blood transfusion. Although 4 patients in group 1 and 2 patients in group 2 had chronic liver disease associated with viral hepatitis (their mean ± SD indocyanine green retention rate at 15 minutes was significantly higher than that in the 28 patients with normal liver; 16.5% ± 2.5% vs 9.9% ± 4.8%, P = .003), their intraoperative blood loss was not significantly higher than that in the 28 patients with a normal liver (987 ± 560 vs 1030 ± 720 mL, P = .89). Only 1 patient with cirrhosis in group 2 required red blood cell transfusion.
The postoperative maximum levels of AST and ALT in the patients in group 2 were significantly higher than those in the patients in group 1. There was no significant difference in the maximum serum total bilirubin levels between the groups (Table 3). The AST level in group 1 returned to the preoperative value at a mean ± SD of 6.4 ± 2.7 days after the operation, whereas in group 2 it did so at 9.7 ± 7.6 days after the operation (P = .07). The ALT level in group 1 returned to the preoperative value at a mean ± SD of 9.5 ± 3.9 days after surgery, whereas in group 2 it did so at 13.7 ± 4.3 days after the operation (P = .01).
There was no postoperative mortality among the 34 patients who underwent prolonged multiple intermittent portal triad clamping. The overall morbidity was 32% (11 patients). No major complications (including postoperative liver failure) occurred, and no patient required reoperation. The morbidity in group 1 was 36% (9 patients), whereas that in group 2 was 22% (2 patients) (P = .73). There were 17 minor complications in 11 patients, and there were no potentially fatal complications. There were 7 cases of bile leakage, all of which resolved spontaneously within 3 weeks. Five patients developed wound infection that was diagnosed by positive culture of microorganisms from the wound, 2 patients developed right-sided pleural effusion, 1 patient experienced prolonged lymphorrhea, 1 patient had delayed gastric emptying, and 1 patient experienced diarrhea. All complications were cured with conservative treatment. The postoperative hospital stay was similar in both groups. Only 1 of 6 patients with chronic liver disease developed wound infection, which was cured by drainage and administration of antibiotics. The overall 1-year survival rates after hepatectomy were virtually the same in the 2 groups: 89% (23 patients) in group 1 and 88% (8 patients) in group 2.
Uncontrollable massive hemorrhage can lead to deterioration of liver function and increased postoperative mortality and morbidity. At the beginning of the 20th century, Pringle6 demonstrated that inflow vascular occlusion could reduce the degree of liver bleeding. His observation that vascular control could enable major hepatic resection without excessive blood loss predated the progressive decrease in mortality and morbidity associated with liver resection that has occurred during the past 50 years, largely as a result of lower intraoperative blood loss.3,4 Hannoun et al7 reported that continuous vascular occlusion during major hepatic surgery is a useful maneuver that may be performed safely on normal hepatic parenchyma for up to 90 minutes. Kim et al8 demonstrated that prolonged continuous hepatic inflow occlusion could be used during hepatectomy for up to 75 minutes without serious complications in selected patients with active chronic liver disease. Despite the widespread use of continuous vascular clamping during major hepatic resection, concern arose that ischemia followed by reperfusion associated with vascular clamping could cause injury to the liver.9
To improve hepatic parenchymal tolerance to ischemia, intermittent inflow occlusion has gained wide popularity, particularly in Asia.1 Various protocols have been recommended using periods of inflow occlusion for 10 to 30 minutes followed by 5 to 15 minutes of reperfusion. Elias et al10 demonstrated that the IPM results in less blood loss and better preservation of liver function and permits a total ischemia time of up to 120 minutes. A randomized clinical study of patients undergoing liver resection by Belghiti et al5 showed that intermittent clamping using multiple cycles of 15 minutes of ischemia and 5 minutes of reperfusion was associated with decreased injury compared with similar periods of continuous inflow occlusion. The previously reported maximum limit of normothermic cumulative hepatic ischemia using intermittent clamping was 322 minutes.11 Our clinical data show that the safe upper limit of cumulative hepatic ischemia in normal liver can be extended to 325 minutes without major complications.
On the other hand, there is always a risk of ischemic damage to hepatocytes, especially in patients with chronic liver disease, the degree of which is likely to be accentuated by a prolonged period of vascular inflow occlusion.1,3 Within tolerable periods of ischemia, the liver can recover from operative and ischemic injury. Otherwise, deterioration of liver function will ensue. The present study revealed that prolonged intermittent pedicular clamping does not increase operative mortality and morbidity even in patients with chronic liver disease. Selected patients with cirrhotic liver withstood prolonged intermittent ischemia for up to 142 minutes, a period within which any difficult hepatectomy can be safely accomplished without excessive bleeding.
The concept of ischemic preconditioning has been applied recently to hepatic surgery and is based on the biological principle that tissue primed by various types of sublethal stress develops tolerance to subsequent lethal injury.12 Clavian et al13 provided evidence that a preconditioning period of 10 minutes followed by 10 minutes of reperfusion confers protection against prolonged ischemic insults in patients undergoing liver resection. Ischemic preconditioning is superior for ischemic periods of up to 75 minutes because it is not associated with blood loss during transection of the liver. Multiple short ischemia-reperfusion cycles were less deleterious than 1 continuous period of the same duration, suggesting that some of the benefit of intermittent pedicle clamping may actually result from the effect of the first clamp-unclamp sequence as a preconditioning treatment. Further studies in this area are encouraged.
Pedicle inflow occlusion induces liver ischemia and subsequent changes in the serum aminotransferase levels.5,14 The present study confirmed that there was significant correlation between the maximum postoperative aminotransferase levels and the duration of ischemia (AST r = 0.681 and ALT r = 0.694, P<.001 for both), but changes in the serum enzyme levels were always transient, with a tendency to return toward the preoperative level within a week. On the other hand, liver ischemia is not the only factor responsible for aminotransferase release. Surgical trauma probably contributes considerably to this, as suggested by the 8-fold increase in the serum aminotransferase levels seen after hepatectomy performed without vascular occlusion.15 The amount of liver resected has been linked to the rise of aminotransferase after hepatectomy.1,15,16 During left or right hepatectomy, the branches of the hepatic artery and portal vein of the resected lobe are divided initially. In cases of limited resection, the glissonian triad that feeds the resected area is ligated first. Aminotransferase release into the systemic circulation from these ischemic tissues continues until the vein draining the resected area is ligated. Therefore, the postoperative serum aminotransferase levels after hepatectomy may reflect not only the degree of ischemic damage to the residual liver but also the aminotransferase released from the resected liver.
Drawbacks associated with intermittent occlusion include increased duration of surgery and blood loss from the transected surface of the liver during the successive periods of reperfusion.5,17,18 The minimization of intraoperative blood loss and the consequent lack of necessity for blood transfusion is a widely accepted goal, because these factors impair patient outcome and prognosis. In the 123 patients in the original cohort with a cumulative clamping time of less than 90 minutes, the blood loss during surgery was low and only 3 patients required blood transfusion; most of the 34 patients with a cumulative ischemic time of 90 minutes or longer also had no significant blood loss and only 3 patients required blood transfusion. Because a transfusion rate of up to 30% is generally considered acceptable,17,18 our findings in the present series of difficult hepatectomies demonstrate that good results can be achieved with intermittent pedicular clamping. To obtain a bloodless operative field, liver dissection was accomplished using Péan forceps.19 Careful millimeter-by-millimeter progression is mandatory, with ligation of every thin vascular or biliary structure. The minimization of backbleeding from accidentally injured hepatic veins is another way of reducing the need for blood transfusion; to achieve this, reducing the inferior vena caval pressure by administering a muscle relaxant and reducing the tidal volume seem to be adequate. Moreover, to prevent liver congestion, the division of the glissonian triad of the partial liver to be resected preceded that of the hepatic vein. The hepatic veins were ligated on the nonresected side and were clipped on the specimen side.
Our retrospective study of these 34 cases in which intermittent portal triad clamping of 90 minutes or longer was used showed no postoperative mortality and acceptable minor complications. This study highlights the usefulness of intermittent pedicular clamping in terms of parenchymal tolerance to ischemia. Intermittent pedicular clamping allows an increase in the cumulative ischemic time and the performance of difficult hepatectomies without increasing the amount of blood loss or significant postoperative liver dysfunction, even in patients with histologically abnormal liver parenchyma.
Correspondence: Yoichi Ishizaki, MD, Department of Hepatobiliary-Pancreatic Surgery, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan (firstname.lastname@example.org).
Accepted for Publication: June 14, 2005.