ronk 11,2003 .4
Kongres

Ischemia and its limitation in liver surgery

Krawczyk, M., Korba, M.

Department of General, Transplant and Liver Surgery
Medical University of Warsaw, Poland

The prime concern of the hepato-biliary surgeon is to minimise blood loss during liver resection. It is necessary to remember that liver is highly vascularized and receives blood from hepatic artery (25%) and portal vein (75%). It is also important that blood flow in liver is very high and estimated for 1500 ml/min. that means a quarter of the cardiac output. Also when we analysed the blood supply before liver resection it is necessary remember that only in 50% we can find typical anatomy.
As I mentioned above the prime concern of the hepato-biliary surgeon is to minimise blood loss during liver resection. This may be achieved in several ways that range from segmental portal control to total vascular occlusion of the major liver vessels and each is joined with side effects. The most popular is the portal triad clamping (PTC) (so called Pringle manoeuvre), which was preformed first time 90 years ago (9). Pringle manoeuvre causes interrupting the arterial and venous inflow to the liver but has no effect on back flow bleeding from branches of the hepatic veins. Interruption of hepatic inflow to a normal liver under normothermin conditions is safe by Hugeut for to 60 min (5).
Clamping, moreover, may be continuous or intermittent with brief clamp release. Liver clamping has two main repercussions: splanchnic and systematic hemodynamic disturbances and visceral ischemia PTC causes 10% increase in mean arterial pressure, a 40% increase in systematic vascular resistance, a 5% decrease in pulmonary artery pressure, and a 10% decrease in cardiac index. The hemodynamic consequences are due to a reflex related to clamping of the hepatic pedicle.
It is necessary to remember that the pedicle clamping does not prevent air embolism. It may occur either during parenchyma transsection, if the central venous pressure is very low or when the blood is restored after Pringle maneuvour because it mobilizes air bubbles trapped in opened veins.
Unclamping of the hepatic pedicle is associated with a decrease in blood pressure due to deactivation of the reflex produced by clamping, as well as by ischemia reperfusion injury of the liver parenchyma (3).
The second way to control intraoperative bleeding is hemihepatic vascular clamping (so called Makuuchi's manoeuvre) (7). During this procedure vascular clamping selectively interrupts the arterial and venous inflow to the right or left hemiliver and therefore avoids both splanchnic blood stasis and ischemia or ischemia-reperfusion injury to the whole liver. The negative side of this procedure is that the bleeding of the resection plane cannot by completely avoided.
In patients with liver cirrhosis there is indication for segmental vascular clamping. Such procedure was introduced by Makuuchi (10). This manoeuvre produces selective intrahepatic occlusion of the segmental portal branch with occlusion of the right or left hepatic artery. Segmental vascular clamping visualising the margin of the portal unit in the liver allows the segmental resection while minimising the amount of blood loss and ischemic injury.
There is only one procedure, which combines total inflow and outflow vascular occlusion of the liver. It is called total hepatic vascular exclusion (THVE). It was introduced in 1966 by J.P.Heaney (4). This procedure isolates the liver from circulation and clamps are applied on the hepatoduodenal ligament, infrahepatic inferior vena cava and suprahepatic inferior vena cava. During THVE, the patient's body is divided into an upper vascular compartment with normal venous resistance and lower vascular compartment with increased resistance to venous return and increased blood volume. Prior to continuous THVE, a short THVE test of 3 min - 5 min should be done. If the mean arterial pressure decreases by more than 20 %, the clamps are taken off and possible causes of hemodynamic instability are corrected (intravascular volume, electrolyte imbalance, metabolic acidosis, coagulation etc.) Very often vascular overloading (500 ml of colloids) is useful before THVE (6).
This procedure is well tolerated by normal liver parenchyma for up 60 min (1). This operation provides to a 25% decrease in pulmonary artery pressure, a 40% decrease in cardiac index, an 80% increase in systematic vascular resistance, and stable mean arterial pressure. Despite careful hemodynamic monitoring and fluid expansion THVE is not tolerated in 10 -15% of patients. In this group during THVE it is necessary to use biopomp, which is used during liver transplantation or refuse the THVE.
It is possible to control inflow and outflow but only limited to major hepatic veins without occlusion of the caval blood flow. Such procedure occurs with hemihepatic vascular clamping and occlusion of the corresponding hepatic vein or clamping of the hepatic pedicle with occlusion of the three hepatic veins (SHVE - Selective Hepatic Vascular Exclusion). The advantage for this technique of reducing of bleeding is the prevention of backflow bleeding or air embolism without having to resort to caval blood flow interruption with its hemodynamic consequences. Extra benefit from that technique is good tolerance of the longer ischemic time of the liver. Disadvantages of this technique is the need of an extensive liver dissection and it is feasible in 90% of the cases (2). It is known that extraparenchymal control of a normal right hepatic vein can be easily achieved in almost all patients, whereas extraparenchymal control of the common trunk of the middle and left hepatic veins is more difficult and not always possible.
In the SHVE group the crucial component of the operation is the complete disconnection of the liver from the retrohepatic inferior vena cava (IVC) by dissecting and ligating the short hepatic veins. On the right side the procedure commenced from the infrahepatic IVC and advanced upward until the right hepatic vein is exposed. On the left side, the ligamentum venosum is transacted and the common trunk of the left and middle hepatic vein is dissected free from the surrounding tissues. At the completion of the dissection, the hepatic veins are encircled with vessel loops. Liver blood inflow is occluded as in the Pringle manoeuvre. Clamping with clamps separately the right hepatic vein and the common trunk of the middle occludes outflow and left hepatic veins at their origin from the IVC.
It has been reported that repetitive, short periods of ischemia, separated by intermittent of reperfusion render the liver more tolerant to subsequent, longer ischemic episodes, and attenuate the injury observed after ischemia-reperfusion. This protective effect has been called ischemic preconditioning. The mechanism of ischemic preconditioning is not exactly known, potentional mediators include nitric oxide, adenosine, prostacyclin or bradykinin. After using the ischemic preconditioning during liver resection it was observed no increasing of AST and LDH whereas ALT was increased. This observation needs new investigations (8).
There is also another possibility to control bleeding during liver resection. It is a central venous pressure. It is a commonly accepted way of reducing blood loss during liver parenchyma dissection due to reducing central venous pressure. For this purpose administration of a muscle relaxant and reduction of the tidal volume seems able to reduce the back bleeding from hepatic veins. It is also possible to eliminate air embolism by keeping the central venous pressure above 0 cm H2O.

Own experience

Until 2003 in Department of General, Transplant and Liver Surgery 721 liver resections were performed in 686 patients. Reresection was done in 29 (4,0%) and third resection in 6 patients.
The malignant neoplastic diseases of the liver were indications for hepatic resections in 526 (78,9%) and 141 (21,1%) liver resections were done due to benign indications (Table 1).

  Number Morbidity Mortality
Total number 721 102 (14,1%) 35 (4,8%)
Malignant neoplastic liver lesions 526 (73,0%) 67 (12,7%) 25 (4,7%)
Benign liver lesions 141 (19,6%) 13 (9,2) 2(1,5%)
Trauma resections 54 (7,5%) 24 (44,4%) 8(14,8%)
Tab. No.1 - Morbidity and mortality of the patients who underwent the hepatic resections

The malignant neoplastic diseases included metastatic liver cancers in 421 (58,4%), HCC in 81 cases (11,2%) and cholangiokarcinoma in 24 (3,3%).
Metastatic tumours from colorectal cancer were indications for liver resections in 366 (86,9%) (table 2). In 55 patients metastases originated from carcinoid, kidney cancer, cancer of the breast.

  Number Morbidity Mortality
Without chemotherapy 261 (62,0%) 18 (6,7%) 5(2,2%)
With chemotherapy 160 (38,0) 20 (12,5%) 8 (5,0%)
Total number 421 38 (9,0%) 13 (3,1%)
Tab. No.2 - Morbidity and mortality of the patients who underwent hepatic resections due to metastatic lesions.

The operation with the ischemia of the liver is a routine operation in our department. We use three kinds of the liver ischemia. The most common is the portal triad clamping (PTC) - Pringle manoeuvre ( in 48.3% of resections). We perform it routinely during resection more than one segment in metastatic diseases.
In cirrhotic patients we performed intermittent PTC. After 15 min. clamping we back the blood inflow, and thereafter we continue the clamping.
In 17 cases we used total hepatic vascular exclusion (THVE). The localisation of the liver lesions needed the exclusion of the whole liver from circulation during surgery. All these operations were done without complications.
As it was mentioned above during liver resection due to metastatic tumor we performed routinely the PTC. The ischemia time was between 30 - 87 min. Evaluation of postresection complications and mortality of those patients who underwent liver resection due to metastatic tumor with previous chemotherapy indicated that complications occurred more frequently than in patients without chemotherapy. The chemical treatment consisted of intravenous administration of the folic acids (200 mg/m2/24 hours and FU ( 400 mg/m2/24 hours) on subsequent days. In the first group (patient without chemotherapy) the percentage of complications was 6,7% in the other group (patients with chemotherapy) - 12,5%. The mortality was 2,2% and 5,0% respectively.
The above differences motivated the functional studies of the liver in patients who underwent liver resection with the pedicle occlusion. We estimated it in two groups: first without previous chemotherapy and second after chemotherapy.
In all cases the parenchyma of the liver was healthy except focal conditions.
The goal of our studies was to estimate the biochemical consequences of the liver resection with pedicle continuous clamping. The following biochemical parameters were analysed:
1. Aspartate aminoitransferase (AST)
2. Alanine aminoitransferase (ALT)
3. Lactate dehydrogenase (LDH)
4. Bilirubin total
5. Alkaline phosphatase (Aph)
6. Gamma-glutamyltranspeptidase (GGTP)

The above listed enzymatic tests were done as follows : before the operation, 6 hours, 24 hours and 8 days after the procedure.
Among dynamic tests, lidocaine clearance test (MEGX) estimated microsomal function of the liver and a ketone body ratio estimating mitochondrial function were used.
The normal ketone body ratio was over 0.7 and MEGX over 90 g/l in all patients. It also referred to patients who had been previously treated chemically.
The dynamic tests were performed before and during liver surgery. In the latter case the tests were administered in the following manner: directly before reperfusion, 30 min after reperfusion and 8 days after surgery.
In both groups the outcome of the enzymatic studies in patients in with ischemic time below 30 min was unchanged even as late as 6 hours, 24 hours, 8 days after the surgery when compared to preoperative values. The results were not dependent on the size of resections.
The outcome of the enzymatic studies in patients with major liver resection indicates a higher level of AST, ALT, LDH and GGTP 6 hours and 24 hours after the procedure, when the ischemic time was over 60 min. There was no increase in the bilirubin and alkaline phosphatase levels observed. The abnormal values normalised 8 days after surgery. Considerable enzymatic changes were observed in resected patients after previous chemotherapy. The changes were already observed in resections of 2 segments after 30 min and 60 min of ischemic time although the differences were not significant statistically.
Table 3 and Table 4 show the results of the ketone body ratio and MEGX after liver resection in patients without previous chemotherapy (group I) and with chemotherapy (group II)
  Ischemic time > 30 min, resection < 2 segments Ischemic time > 30 min, resection > 2 segments Ischemic time > 60 min, resection < 2 segments Ischemic time > 60 min, resection > 2 segments
Normal > 0.7
Before reperfusion - group I 0.4 0.4 0.3 0.2
Before reperfusion - group II 0.3 0.3 0.2 0.2
30 min. after reperfusion - group I 0.6 0.6 0.4 0.4
30 min. after reperfusion - group II 0.6 0.6 0.4 0.4
8 days after reperfusion - group I 0.7 0.7 0.7 0.7
8 days after reperfusion - group II 0.7 0.7 0.7 0.7
Tab. No.3 - Ketone body ratio

The outcome of the dynamic tests was very significant and interesting.
  Ischemic time > 30 min, resection < 2 segments Ischemic time > 30 min, resection > 2 segments Ischemic time > 60 min, resection < 2 segments Ischemic time > 60 min, resection > 2 segments
Normal > 90 g/l
Before reperfusion - group I 50 g/l 50 g/l 30 g/l 20 g/l
Before reperfusion - group II 25 g/l 20 g/l 20 g/l 20 g/l
30 min. after reperfusion - group I 50 g/l 50 g/l 30 g/l 30 g/l
30 min. after reperfusion - group II 45 g/l 40 g/l 20 g/l 20 g/l
8 days after reperfusion - group I 90 g/l 90 g/l 90 g/l 90 g/l
8 days after reperfusion - group II 90 g/l 90 g/l 90 g/l 90 g/l
Tab. No.4 - MEGX

After 30 min of pedicle occlusion, the ketone body ratio decreased to 0.4 directly before reperfusion. At the same time, the values of the lignocaine test lowered to 50 g/l, which indicated an insignificant functional defect of the liver parenchyma. The value was much lower in patients with previous chemotherapy (average 0.3 ketone body ratio and 25 g/l MEGX). These changes were observed regardless the size of resection.
Thirty minutes after reperfusion, a small increase of the ketone body ratio values was observed. A moderate normalisation was correlated with the size of the liver resection. In patient who had undergone more than 2 segment resection, the normalisation of the ketone body test results was not noticed in the 30th minute after the reperfusion. This referred both to patients with and without previous chemotherapy. Some of these patients had ascites and a high bilirubin level.
The results of the both dynamic tests normalised on the 8th day after liver resection in all patients.
After a 60 min pedicle occlusion, the ketone body ratio decreased to 0.3 directly before reperfusion. At the same time, the values of the lignocaine test decreased to 30 g/l , which indicated a significant functional defect of the liver parenchyma. The value was significantly lower in patients with previous chemotherapy (average 0.2 ketone body ratio and 20 g/l MEGX).
The results of the dynamic tests obtained 30 min after reperfusion were slightly decreased and normalised not sooner than 8 days after the procedure.
Eight days after the liver parenchyma resection of two or less segments, then results of the lignocaine test and ketone body ratio showed no changes.
In patients who had undergone four or more segments resection, the mitochondrial tests normalised during 8 days after the operation, while then lignocaine tests still indicated impaired functioning of the liver.

Conclusions

1. Liver resections with occlusion of the hepatoduodenal ligament impair the function of the liver during a few days.
2. The lignocaine test was more precise in estimating the liver parenchyma function.
3. The pedicle occlusion impair the parenchyma function more intensively in patients with previous chemotherapy.
4. The degree of functional impairment depends on the duration of pedicle occlusion and size of the liver resection.
5. Although functional changes were observed, the occlusion of portal and arterial blood inflow longer than 60 min does not cause irreversible changes in the liver parenchyma.
6. The liver resections should be carried out under intermittent warm ischemia.

Literature

  1. Bernay T. at al : Total vascular exclusion of the liver for the resection of lesions in contact with the vena cava or the hepatic veins. Br. J. Surg., 1998, 85, 485 - 488
  2. Elias D. et al : Intermittent complete vascular eclusion of the liver during hepatectomy: technique and indications. Hepato-Gastroenterology, 1998, 45, 389 - 395.
  3. Hasselgren P.O.M.: Prevention and treatment of ischemia of the liver. Surg. Gynecol. Obstet., 1987, 164, 187 - 196.
  4. Heaney J.P. et al : An improved technique for vascular isolation of the liver: Experimental study and case reports. Ann. Surg., 1966, 173, 237 - 241.
  5. Huguet C. et all : Liver ischemia for hepatic resection: Where is the limit ? Surgery, 1992, 11, 251 - 259.
  6. Huguet C. et al : Total vascular exclusion for liver resection. Hepato-Gastroenterology, 1998, 45, 368 - 369.
  7. Makuuchi M. et al : Safety of hemihepatic vascular occlusion during resection of the liver. Surg. Gynecol. Obstet., 1987, 164, 155 - 158.
  8. Nakano A. at al : Ischemic preconditioning. From basic mechanisms to clinical applications. Pharmacology & Therapeutics, 2000, 86, 263 - 275.
  9. Pringle J.H. : Notes on the arrest of hepatic haemorrhage due to trauma. Ann. Surg. , 1908, 48, 541 - 549.
  10. Takayama T. at al : Selective and unselective clamping in cirrhotic liver. Hepato-Gastroenterology, 1998, 45, 376 - 380.

Address for correspondence:

Professor Krawczyk, Marek
Department of General, Transplant and Liver Surgery
Medical University of Warsaw
1a Banacha Str., 02 - 097 Warsaw, Poland