Evaluation of Liver Function Tests to Predict Operative Risk in Liver Surgery

Despite numerous studies in the past it is not possible yet to predict postoperative liver failure and safe limits for hepatectomy. In this study the following liver function tests ICG-ER (indocyaninegreen elimination rate), GEC (galactose elimination capacity) and MEGX-F (monoethylglycinexylidid formation) are examined with regard to loss of liver tissue and prediction of operative risk. Liver function tests were assessed in 20 patients prior to liver resection and on the 10th. postoperative day. Liver and tumor volume were measured by ultrasound and pathologic specimen and the parenchymal resection rate was calculated. In patients without cirrhosis (n = 10) ICG-ER and MEGX-F remained unchanged after resection, GEC was reduced but did not correspond to the resection rate. Patients with cirrhosis (n = 10) had a significantly lower ICG-ER and GEC before resection than patients without cirrhosis. After resection these tests were unchanged. Patients with liver related complications and cirrhosis (n = 5) had lower ICG-ER and GEC than patients with cirrhosis and no complications. In the postoperative course all liver function tests in these patients were significantly lower compared to preoperative results. Comparing liver function tests ICG serves best to indicate postoperative liver failure. Liver function tests do not correspond with loss of liver tissue.


INTRODUCTION
Liver resection is the only method to cure primary and metastatic liver tumors. Standardization of the operation has led to a decrease in operative mortality, but the mortality rate is still near 20%. In approximately 70% HCC is combined with cirrhosis (Table 1). These persons bear a high risk of developing liver insufficiency, since liver function and liver regeneration is impaired 6,9. Despite numerous studies in the past it is not yet possible to predict postoperative liver failure and safe limits for hepatectomy. Some authors prefer the determination ofprotein synthesis like coagulation Address  analysis or urea-nitrogen-synthesis-rate to estimate preoperative liver function in cirrhotic patients but a common standard does not exist yet. Quantitative liver function tests are well known in hepatology but since the test procedure is elaborate and therapeutic consequences are lacking their use is limited to special questions. Progress in surgery and transplantation induced a reestimation of these tests3,4,1.
The tests commonly used now and employed in our study are the ICG, GEC and MEGX tests.
ICG is a dye which is excreted by the liver dependant on liver blood flow. The elimination rate normally amounts 16-20 %/minll. Galactose is converted to galactose-l-phosphate by the enzyme galactokinase. The elimination capacity should reflect the metabolically active liver tissue. The normal value is about 6, 5 mg/min kg12. MEGX is a metabolite of lidocaine. The MEGX--formation is controlled by the cytochrome P 450 system in the liver and is supposed to indicate the active liver tissue as well1.

13
In our study we raised the following issues: 1. Are the tests capable of recognizing a loss of liver tissue by hepatectomy since they are called "quantitative"?
3. Can we predict the operative risk of hepatectomy with the help of liver function tests?

Methods
In 34 patients undergoing surgery for hepatectomy, cholecystectomy or hydatid cyst resection liver function tests were assessed prior to operation and on the 10th. postoperative day. Liver volume and tumor volume were determined by ultrasound at the same time in order to calculate the parenchymal resection rate described by Okamoto1. The resected liver specimen served as a control for the calculation of the volume resected. The incidence of cirrhosis was documented. The postoperative course was recorded with special referrence to signs of liver insufficiency like intractable ascites, coma or deficiant protein synthesis. Liver resection was performed in 20 patients, 6 of them suffered from HCC. 10 Patients had liver cirrhosis proven by histological examination. Patients  with gall stone disease and hydatid cysts served as a control group to show the effect of anaesthesia and operation trauma on liver function. Informed consent was obtained from each patient.

Liver Function Tests
The ICG test was performed by the bolus injection technique described by Paumgartner1. The extinction of the dye was measured in 7 serum samples, 3 to 21 min after injection of 0,5 mg/kg ICG. The elimination rate was calculated by least square log linear-regression analysis of the descending portion oflCG extinction as a function of time.
Liver and tumor volume was determined pre-and postoperatively by ultrasound1. The parenehymal resection rate was calculated according to the method by Okamoto1.
Supposing a normal distribution of the test values the student-t-test was applied to calculate the significance from mean values and standard deviation.

RESULTS
No operative death occurred in 20 cases of liver resection but there were 5 late complications and 3 patients died because of liver failure. The majority of these complications was induced by preexisting liver cirrhosis. It seems that the extent of resection is not only necessarily a cause of liver failure. This may focus our interest to the question whether the degree of cirrhosis and consequently the risk of postoperative liver insufficiency can be determined prior to operation. Therefore we allocated our patients to 3 categories: The results are shown separately for each test employed. (Fig. 1) The ICG-elimination rate apparently is not influenced by liver resection in normal livers and cirrhotic livers. Postoperative complications lead to a significant decrease of the elimination rate (p 0,004 vs. preop in the same group). The difference between the preoperative values of noncirrhotic and cirrhotic livers is significant (p 0,001). There is a further drop in the preoperative elimination rate of patients who will develop liver failure but this is not significant.
The galactose elimination capacity (Fig. 2) shows a significant decrease after resection in normal livers (p 0,003). The regression equation however fails to show a correlation between the resection rate and the decrease of the elimination capacity (r 0,452174).
The difference between cirrhotic and non cirrhotic livers is significant (p 0,001). Similar to the ICG elimination there is a further decrease of GEC in patients with liver failure. It cannot be explained why GEC remains unchanged after resection in cirrhotic patients. Postoperative complications result in a significant decrease of the GEC (p 0,036 vs. preop).
Regarding the MEGX test (Fig. 3) we find a postoperative rise mainly in the control group and with decreasing extent in the resection groups. Only persons with complicated postoperative course have a lower MEGX value 15 min after lidocaine injection.

DISCUSSION
Generally the operative risk in liver surgery is estimated on the basis of serum chemical liver profile and coagulation studies. Under the compensatory condition of the diseased liver these data are usually within or near normal limits. Therefore it is difficult to predict liver function following hepatectomy preoperatively. In consequence it is desirable to apply a test which is    Res. and Res. and C. without Cirrhosis Liver with liver F:ailure failure Figure 3 Monoethylglycinexylidide formation before and after cholecystectomy, liver resection in unaffected livers and cirrhotic livers. Patients with cirrhosis are allocated to a group with uneventful postoperative course or to a group with liver failure. capable of determining liver function and to estimate the limit of safe hepatectomy in those patients. Quantitative liver function tests are supposed to indicate the metabolically active liver volume and liver function. Several studies affirmed this statement. ICG is reduced after functional hepatectomy in dogs and induction of cirrhosis14. A positive correlation between liver volume and GEC has been found in aging man15. In our study only GEC indicates loss of liver tissue but the decrease of test level does not correlate with the resection rate.
It has been shown that liver resection induces an additional capacity for metabolizing galactose16. This may serve as an explanation for our results, that liver resection is not followed by an adequate reduction of elimination capacity. The ICG elimination rate performed by the bolus injection technique reflects the liver perfusion 1 and depends on intrahepatic shunting to be found in cirrhotic and tumor bearing livers17. Therefore it is not a valuable parameter of the active liver volume. As we know other application techniques are apt to show a volume dependant ICG elimination 8 but application form and calculation are complicated and not suitable for clinical use. The MEGX formation interferes with many drugs 9 and maintains a wide interindividual range of test values. In our opinion the test procedure which consists of only one blood sample after 15, 30 or 60 minutes, is not sufficient for analyzing liver function.
The value of the ICG test, GEC and MEGX test in indicating presence of cirrhosis is not doubted ,2,21. Since we are aware of liver cirrhosis by routine histological examination done preoperatively the superiority of the tests to simple methods like determination of the Child index is questioned 22,23. The results of our examination present a significant difference between test values of cirrhotic and non cirrhotic livers only for ICG and GEC. In contrast to ICG and GEC the MEGX test fails to recognize cirrhotic livers.
The ICG elimination is diminished postoperatively in patients suffering from liver related complications24. This statement is confirmed by our results. Further information is awaited to select patients with impending liver insufficiency by preoperative investigations to enhance the effect of liver function tests. Okamoto , Mizumoto 8 and Yamanaka 5 have developed sophisticated score systems including ICG which are able to recognize high risk patients. Yet clinical application is not widely accepted. Although patients with cirrhosis and postoperative complications have lower ICG and GEC values in our study than patients with cirrhosis and uneventful postoperative course this difference is not significant presumably due to the small number of patients.
As the test procedure for determination of GEC is complicated and time consuming the ICG test in our opinion serves best to distinguish high risk patients. The score system proposed by Yamanaka2s, which includes the ICG elimination, is more accurate than the ICG test but it is difficult to perform in daily clinical routine.
Although quantitative liver function tests examined in our study do not correspond to the actual liver volume, they seem to be a valuable parameter to select patients with severe cirrhosis who are likely to suffer from postoperative liver insufficiency. Those patients should be treated by a limited resection.