The Effect of Communication Between the Right and Left Liver on the Outcome of Surgical Drainage for Jaundice due to Malignant Obstruction at the Hilus of the Liver

Debate continues regarding the optimal management of irresectable malignant proximal biliary obstruction. Controversy exists concerning the ability of unilateral drainage to provide adequate biliary decompression with tumors that have occluded the communication between the right and left hepatic ductal systems. Between October 1986 and October 1989, 18 patients with malignant proximal biliary obstruction were treated by an intrahepatic biliary enteric bypass. Patients were divided into two groups based on the presence or absence of a communication between the right and left biliary systems. In Group I (n = 9), there was free communication; and in Group II (n = 9) there was no communication. There were two perioperative deaths (11%) one due to persistent cholangitis and the other to myocardial insufficiency both with one death in each group. The median survival (excluding perioperative deaths) was 5.6 months. Comparison of pre- and postoperative serum levels of bilirubin and alkaline phosphatase showed a significant decrease in each group, but no difference between the groups in the size of the reduction. Sixteen patients survived at least three months and the palliation was judged as excellent in eight, fair in five, and unchanged in three. These results demonstrate the effectiveness of biliary enteric bypass regardless of communication between the left and right biliary ductal systems.


INTRODUCTION
Obstructive jaundice resulting from either benign or malignant biliary structure may be relieved by biliary enteric bypass or transtumoral intubational methods 1-6. When such obstruction involves the confluence of the hepatic ducts at the liver hilus, the irresectable lesion may completely occlude the confluence, and prevent drainage between the right and left hepatic ductal systems. It has been suggested that adequate relief of jaundice requires decompression of both ductal systems 17-19. The aim of this study was to assess the outcome for patients with malignant biliary obstruction at the hilus treated by a single biliary enteric anastomosis which drained either one or both ductal systems.

PATIENTS AND METHODS
Between October 1986 and October 1989, 18  Clinical, biochemical, and radiological assessment was performed in all patients including sonography of the biliary tree, and computed tomography. Delineation of the biliary tract by percutaneous transhepatic cholangiography or endoscopic retrograde cholangiography was employed to provide precise information regarding the anatomy of the biliary obstruction.
All patients had a neoplastic process involving the proximal !/3 of the extrahepatic bile ducts with 9 patients having total occlusion at the confluence of the right and left hepatic ducts. The site of hilar occlusion and the presence of a communication between the right and left liver was confirmed at operation. An operative biliary enteric bypass to the segment III duct was the treatment of choice using a Roux-en-Y side to side anastomosis. The segment III bypass was originally described by Soupault and Couinaud 14 and has been modified by several authors 1'2'5'7'15'2'2. Anastomosis to the segment III duct was possible when preoperative studies demonstrated a dilated left hepatic duct in the absence of left lobar atrophy and without significant parenchymal involvement of the left liver. Consequently, 18 patients had a segment III bypass22. Two patients had only a segment V approach23-25.
The patients were retrospectively divided for assessment into two groups based on cholangiographic evidence of communication between the right and left ductal systems. In Group I (n 9), communication was present; and in Group II (n 9), there was complete obstruction.
The patients were reassessed at three months after operation to evaluate clinical symptoms, and by liver function tests. Assessment of symptomatic outcome was determined by asking whether their current level of activity was excellent, fair or poor. Sixteen patients were available for review (8 in each group). The serum bilirubin and alkaline phosphatase were compared to preoperative levels and between the two groups using the Wilcoxon sign rank test and the student's T test.

RESULTS
There were two 30 day postoperative deaths (11%), with one death from each group. The first was due to persistent cholangitis in a patient with a diffuse multicentric cholangiocarcinoma with intrahepatic infiltration. The second patient had gallbladder carcinoma and died of progressive heart failure. Six patients developed complications (33%). Five patients had a pulmonary infiltrate on chest X-ray, and were treated for pneumonia. One patient required reoperation for an early postoperative small bowel obstruction in group I.
The serum bilirubin in Group I before surgery was a median of 303 mmol/1 (normal < 25; range 16-545), and at three months after operation, the bilirubin was a median of 101 mmol/1 (range 9-334). In Group II, the median serum bilirubin preoperatively was 419 mmol/1 (range 84-560), and in three months, had decreased to 110mmol/1 (range 18-253) (Figure 1). In each individual group, the reduction of serum bilirubin was statis tically significant (p < 0.05), but there was not a significant difference in the degree of the i'eduction between the groups. The fall in the serum level of alkaline phosphatase paralleled that of the bilirubin (Figure 2). There was a significant reduction within each group (P < 0.05), and no significant difference in the amount of reduc-tion between Groups I and II. In Group I, the median preoperative alkaline phosphatase value was 783 U/1 (range 288-1440; normal < 115), and decreased to 274 U/1 (range 46-538). In group II, the median level fell from a median of 534U/1 (range 185-1087) to 316 U/1 (range 74-892).
At three months after operation, the quality of palliation was assessed by patient interview. The self described level of activity by the patient was excellent in eight patients, fair in five, and no improvement after operation in three. The median hospital stay for patients discharged from hospital was 25days (range 12-111 days). The median survival for the 18   there is no difference between the two groups (p 0. l).

DISCUSSION
The treatment of malignant proximal biliary obstruction can be difficult due to technical operative problems and the hematological and immunological sequelae of longstanding jaundice15'26'27'28. Furthermore, the palliative nature of the treatment, and the availability of transhepatic and endoscopic intubational techniques has resulted in considerable debate regarding the best management of these patients 17 '18'19'22'23. Much of the controversy has centered on the ability of unilateral drainage of the right or left biliary system to provide adequate biliary decompression and subsequent clinical palliation. Indeed, although there has been well documented surgical experience of unilateral drainage, the arguments have persisted, mainly among endoscopists and interventional radiologists, as to whether bilateral drainage is more effective or mandatory. Contributing to the confusion is the lack of documentation in the radiological literature of the presence of lobar liver atrophy in the patients selected for unilateral or bilateral drainage. The efficacy of palliation by operative biliary enteric bypass has been reconfirmed in the current series. The mortality of 11% is comparable to series of patients treated by operative and nonoperative methods 3 Two studies have considered bilateral drainage of the biliary system by endoscopic or the percutaneous route 3 and have failed to show an advantage in either mortality or morbidity when compared with unilateral drainage. These are in contrast to the results of Deviere et al. 3 which showed prolonged survival and a low early and late incidence of cholangitis in bilaterally drained patients compared with those in whom only unilateral drainage was possible. They note, however, the poor survival and early cholangitis in undrained segments may have been precipitated by the ERCP itselfintroducing infection into an undrained or poorly drained segment. Problems with non-operative management of hilar malignancy may be as high as 30% 30 but allows early discharge from hospital. Operative procedures of intrahepatic biliary bypass, while offering lower mortality 6% to 13.6/o 6-5 carry a prolonged hospitalization and associated cost, but usually lower the number of readmissions.
Unilateral lobar atrophy when present, may be associated with an ipsilateral dilated hepatic duct32. Biliary enteric bypass, or stent placement to the duct draining an atrophied lobe may seem technically inviting, but caution should be exercised since the parenchyma of the atrophied lobe may not have the functional capacity to relievejaundice3'22'33. Furthermore anastomosis to the duct of a nonfunctional atrophied lobe may result in an increased incidence of early and late infectious complications, and limits the effectiveness of palliation6.
Comparison between the two groups of patients with drainage of a single versus both sides of the liver is illustrative of what can be expected with transtumoral stent placement. The reductions in serum bilirubin and alkaline phosphatase were similar, and there was no difference in mean survival, or in the quality of palliation if hilar communication is present or absent. The demonstration that unilateral drainage produces similar results to bilateral drainage suggests that an increased number of such patients can be offered the option of palliative operative bypass. Complete hilar obstruction is not a contraindication to unilateral operative biliary decompression.