Anatomic hepatectomy as a definitive treatment for hepatolithiasis: a cohort study

Abstract

Background

Treatment requirements in hepatolithiasis may vary and may involve a multidisciplinary approach. Surgical resection has been proposed as a definitive treatment.

Objectives

This study aimed to evaluate the clinical results of anatomic liver resection among Chilean patients with hepatolithiasis.

Methods

An historical cohort study was conducted. Patients who underwent hepatectomy as a definitive treatment for hepatolithiasis from January 1990 to December 2010 were included. Patients with a preoperative diagnosis of cholangiocarcinoma were excluded. Preoperative, operative and postoperative variables were evaluated.

Results

A total of 52 patients underwent hepatectomy for hepatolithiasis. The mean ± standard deviation patient age was 49.8 ± 11.8 years (range: 24–78 years); 65.4% of study subjects were female. A total of 75.0% of subjects had a history of previous cholecystectomy. The main presenting symptom was abdominal pain (82.7%). Hepatic involvement was noted in the left lobe in 57.7%, the right lobe in 34.6% and bilaterally in 7.7% of subjects. The rate of postoperative clearance of the biliary tree was 90.4%. Postoperative morbidity was 30.8% and there were no postoperative deaths. Three patients had recurrence of hepatolithiasis, which was associated with Caroli's disease in two of them. Overall 5-year survival was 94.5%.

Conclusions

Anatomic liver resection is an effective treatment in selected patients with hepatolithiasis and is associated with low morbidity and no mortality. At longterm follow-up, anatomic hepatectomy in these patients was associated with a lower rate of recurrence.

Introduction

Hepatolithiasis has been defined as the occurrence of stones in any intrahepatic bile duct proximal to the confluence of the right and left hepatic ducts, irrespective of the presence of stones in the main bile duct or the gallbladder.¹ It is prevalent in Japan and Southeast Asia² and uncommon in Western countries.^3,4

Hepatolithiasis is now more commonly detected with the regular use of diagnostic imaging methods, of which the most precise is cholangiographic assessment.⁵

Among patients in occidental countries, intrahepatic stones usually occur secondary to bile stasis resulting from an extrahepatic factor (such as common bile duct stones, or postoperative or inflammatory strictures of bile ducts), in congenital diseases (such as Caroli's disease), and in relation to the tendencies of some ethnic groups towards lithogenic bile.^6–8

Treatment options are diverse. The therapeutic purpose is to obtain complete clearance of the stone and to avoid recurrent episodes of cholangitis and subsequent hepatic fibrosis and secondary cholangiocarcinoma.¹ Recently, non-surgical procedures, such as percutaneous transhepatic or peroral cholangioscopic lithotripsy, have been applied to treat hepatolithiasis, but these non-surgical procedures are associated with recurrence rates of 30–60%.^9–11

Alternatively, liver resection has been proposed as a definitive treatment for segmental hepatolithiasis because it treats not only the stones, but also strictures and atrophic and inflammatory changes.¹²

The aim of this study was to assess the factors associated with hepatolithiasis in one centre and to evaluate the clinical results of its treatment by anatomic liver resection.

Materials and methods

All patients presenting from January 1990 to December 2010 at the Pontifical Catholic University of Chile Hospital were included in this study. Hepatolithiasis was defined as the presence of stones in the intrahepatic bile duct proximal to the right and left hepatic duct confluence. Hepatectomy involved any procedure in which excision of a part of the liver was performed using anatomic techniques according to Couinaud's segmentation and Brisbane 2000 terminology.¹³ Recurrence was deemed to have occurred when clinical and imaging evidence of new lithiasis was detected in the intrahepatic biliary duct above the biliary confluence.

Patients who were not candidates for hepatectomy and those with a preoperative diagnosis of cholangiocarcinoma were excluded from analysis.

Registered data

Demographic data as well as data on risk factors for hepatolithiasis, comorbidities, previous surgical therapies, clinical presentation, laboratory and imaging tests, perioperative data, postoperative morbidity, mortality and postoperative stone recurrence were recorded for all subjects. The extent of disease was assessed by imaging studies, including abdominal ultrasound (US), computed tomography (CT) and/or magnetic resonance cholangiopancreatography (MRCP). Some patients also underwent endoscopic retrograde cholangiopancreatography (ERCP) or percutaneous transhepatic cholangiography (PTC). A multidisciplinary team defined the type of liver resection based on imaging findings.

Clinical results and patient outcomes were classified in short- and longterm categories. Short-term results included postoperative morbidity (any complication occurring within 30 days of surgery). Longterm results included any morbidity or sequel detected during follow-up (any complication occurring >30 days after surgery), the rate of recurrence of stones and longterm survival.

Procedural details

Hepatectomy was considered to be indicated in all patients with symptomatic hepatolithiasis with hepatic or intrahepatic duct stenosis, or with recurrent pyogenic cholangitis and atrophy of the affected segments, with or without liver abscesses. An anatomic resection (with or without the Pringle manoeuvre) using an open or laparoscopic approach was performed for all segments affected by biliary stenosis and the affected bile duct drainage area. Hepaticojejunostomy was performed in patients with common bile duct stenosis and in those considered to be at high risk for recurrence. At each exploration of the common bile duct, routine cholangiography, with or without choledochoscopy, was performed to ensure clearing of the biliary system. When the biliary tract was not explored, cholangiography was achieved through the cystic duct and US was performed to detect residual stones. Drains were used routinely and removed postoperatively if there was no evidence of bile leakage or bleeding between postoperative days 2 and 5.

Postoperative management

Routine nasogastric drainage was not employed. Re-feeding was initiated 24 h after surgery if the patient showed signs of gastrointestinal transit resumption and had no nausea. Antibiotic prophylaxis was used for 24 h perioperatively. Standard doses of therapeutic antibiotics were used in patients with cholangitis or liver abscess. Routine liver tests were monitored after surgery on days 1, 3 and 5, and as required according to the patient's clinical status.

Follow-up

Recurrence screening was conducted using liver function tests. Ultrasound or CT were performed every 6 months for the first 2 years postoperatively and then annually or when the patient had symptoms of recurrence. To confirm the suspicion of recurrence, endoscopic or percutaneous cholangiography was carried out; from 2000 onwards MRCP was used.

Statistical analysis

Variables are presented using descriptive statistics with measures of central tendency and dispersion. Chi-squared and Fisher's tests were used for univariate subgroup and categorical variable analyses. Student's t-test and analysis of variance (anova) were used for continuous variables. A binary logistic regression model was generated for multivariate analysis. The Kaplan–Meier method was used for survival analysis. A P-value of <0.05 was considered to indicate statistical significance. All analyses were performed using spss Version 18.0 (SPSS, Inc., Chicago, IL, USA).

Report

This manuscript was developed according to the STROBE (strengthening the reporting of observational studies in epidemiology) criteria for the reporting of cohort studies.¹⁴

Results

From 1990 to 2010, a total of 52 patients underwent anatomic liver resection for treatment of hepatolithiasis. Twenty-six of these operations (50.0%) were performed between 2004 and 2010.

The mean ± standard deviation (SD) age of patients was 49.8 ± 11.8 years (range: 24–78 years). Overall, 34 (65.4%) patients were female, 39 (75.0%) had a prior history of cholecystectomy and 20 (38.5%) had associated comorbidities (Table 1). The main presenting symptom was abdominal pain. Acute cholangitis was diagnosed in 18 (34.6%) patients (Table 2).

Table 1.

History of bile pathology or biliary surgery in the present series (n= 52)

Bile pathology or biliary surgery	n	%
Patients with one or more record	42	80.8
Cholecystectomy	39	75.0
Choledocholithiasis	10	19.2
Caroli's disease	8	15.4
Acute biliary pancreatitis	2	3.8
Biliary bypass	1	3.8
Biliary resection	1	3.8

Table 2.

Clinical presentation in the present series (n= 52)

Clinical presentation	n	%
Abdominal pain	43	82.7
Cholangitis	18	34.6
Jaundice	17	32.7
Fever	14	26.9
Choluria	11	21.2
Pruritus	3	5.8
Pancreatitis	3	5.8
Nausea	2	3.8
Acholia	2	3.8
Fatigue	2	3.8

Diagnoses were obtained through US in 40 (76.9%) patients. Computed tomography was performed in 26 (50.0%) patients. Twenty-eight (53.8%) patients had ERCP and 16 of these underwent papillotomy. Thirty (57.7%) patients underwent MRCP; after 2000, this diagnostic modality was used in all but one patient. Findings on MRCP indicated the presence of bile duct dilatation in 25 patients, associated choledocholithiasis in eight, liver abscess in three, stenosis of the intrahepatic bile duct in 27 patients, and liver segment atrophy in one patient.

Hepatic involvement was noted in the left lobe in 30 patients (57.7%), the right lobe in 18 (34.6%) and bilaterally in four (7.7%) patients.

Development of hepatolithiasis after cholecystectomy

Thirty-nine (75.0%) patients had a prior history of cholecystectomy. The mean period between cholecystectomy and the presentation of hepatolithiasis was 17.8 ± 10 years. No statistically significant differences emerged when this period was analysed according to which of the hepatic lobes were compromised.

Patients with right hepatic involvement (n= 17/18, 94.4%) were significantly more likely to have a history of previous cholecystectomy compared with those with left (n= 19/30, 63.3%) or bilateral (n= 2/4, 50.0%) liver involvement (P < 0.05) (Table 3).

Table 3.

Relationship between prior cholecystectomy, Caroli's disease and liver lobe involvement

	Liver lobe involvement			P-value
	Right liver	Left liver	Bilateral
Prior cholecystectomy	17 (94.4%)	19 (63.3%)	2 (50.0%)	0.035
Caroli's disease in biopsy specimen	3 (16.7%)	15 (50.0%)	3 (75.0%)	0.025

Caroli's disease

Evidence of Caroli's disease was significantly more frequent in surgical biopsy specimens of patients with left (n= 15/30, 50.0%) or bilateral (n= 3/4, 75.0%) liver involvement than in specimens of patients with right lobe involvement (n= 3/18, 16.7%) (P < 0.05). Of 17 patients with prior cholecystectomy who developed hepatolithiasis in the right liver, only two had Caroli's disease. Multivariate analysis considering the presence of Caroli's disease, sex, age and previous cholecystectomy showed that the latter was an independent risk factor for the development of right liver hepatolithiasis (odds ratio = 6, 95% confidence interval 1.2–30.5; P < 0.05) compared with left or bilateral hepatolithiasis (Table 3).

Surgical treatment

The most frequent procedure was left lateral sectionectomy, which was conducted in 42.3% of patients, three of whom underwent a laparoscopic procedure. This was followed by right posterior sectionectomy and left hepatectomy, conducted in 17.5% and 15.4% of patients, respectively (Table 4).

Table 4.

Type of liver resection in hepatolithiasis (Brisbane Terminology)¹³

Type of liver resection	n	%
Left lateral sectionectomy	22	42.3
Right posterior sectionectomy	9	17.5
Left hepatectomy	8	15.4
Right hepatectomy	6	11.5
Right anterior sectionectomy	2	3.9
Extended right hepatectomy	1	1.9
Extended left hepatectomy	1	1.9
Left lateral sectionectomy + segmentectomy VII	1	1.9
Right posterior sectionectomy + segmentectomy V	1	1.9
Segmentectomy VIII	1	1.9
Total	52	100

Intraoperative US was performed for anatomic delineation in 32 (61.5%) patients. This rate increased to 78.1% from the year 2000.

Cholecystectomy was performed as an additional surgical procedure in 13 (25.0%) patients, choledochostomy in 16 (30.8%) patients (13 patients prior to 2000) and hepaticojejunostomy in three (5.8%) patients. The mean ± SD operative time was 229.7 ± 84.0 min (range: 110–450 min). Mean ± SD intraoperative bleeding amounted to 437 ± 209 ml (range: 100–2000 ml). Hepatic pedicle clamping (Pringle manoeuvre) was performed in 48.1% of patients, with a mean clamping time of 32.0 ± 14.5 min (range: 10–100 min).

Short-term results

The mean hospital stay was 11.1 ± 8.8 days (range: 3–55 days).

Complete clearance of the biliary system was achieved in 47 (90.4%) patients. Early complications occurred in 16 (30.8%) patients (Table 5), nine of whom had biliary complications. There were no significant differences in the incidence of biliary complications according to type of surgery or segment affected. Three of the five patients with bile leakage had Caroli's disease that was confirmed by histology. Three patients with bile leakage were treated with drainage and the leakage resolved within 30 days of surgery. Of the four patients with residual stones, two had postoperative cholangitis and were treated with ERCP, and the other two were treated percutaneously with choledochal drainage. The final rate of clearance of the biliary tree was 98.1%. There was no postoperative liver failure and no inpatient mortality (Table 5).

Table 5.

Early complications in patients who underwent hepatectomy for hepatolithiasis (n= 16/52)

Type of complication	n	%
Wound infection	2	12.5
Residual stones	4	25.0
Acute cholangitis	2	12.5
Intra-abdominal collection	3	18.8
Liver abscess	2	12.5
Atelectasis	3	18.8
Pleural effusion	2	12.5
Bile leakage	5	31.3
Acalculous obstructive jaundice	1	6.3
Hepatic infarction	1	6.3

Surgical specimen evaluation (histopathological study)

Biopsy results showed that 21 (40.4%) patients had Caroli's disease. Biopsy findings also showed that 46 (88.4%) patients had bile duct dilatation and periductal fibrosis, 21 (40.4%) had chronic cholangiohepatitis, and nine (17.3%) had purulent cholangitis. Liver atrophy was found in two patients, micronodular cirrhosis in another two patients, and intrahepatic bile duct tubular adenocarcinoma in one.

Longterm results

The median follow-up was 63.7 months (range: 3–274 months).

Late complications occurred in 12 (23.1%) patients and were of biliary origin in nine (17.3%) (Table 6). Biliary complications included choledocholithiasis, hepatolithiasis and common hepatic duct lithiasis in patients with biliodigestive anastomosis. One patient had choledocholithiasis plus hepatolithiasis. All patients with recurrence of lithiasis in the main biliary duct were successfully treated by endoscopic or percutaneous procedures.

Table 6.

Late complications in patients who underwent hepatectomy for hepatolithiasis (n= 12/52)

Type of complication	n	%
Choledocholithiasis	3	25.0
Recurrent hepatolithiasis	3	25.0
Remnant bile duct lithiasis	2	16.7
Biliodigestive anastomosis stenosis	2	16.7
Bile leakage for >30 days	2	16.7
Intestinal obstruction	1	8.3
Incisional hernia	1	8.3
Abdominal collection	1	8.3
Pulmonary embolism	1	8.3
Hepatic infarction	1	8.3

Three patients developed recurrent hepatolithiasis. Two of these patients have Caroli's disease and are currently undergoing treatment with ursodeoxycholic acid and have not developed complications from their disease. The remaining patient with recurrent hepatolithiasis developed stenosis of the biliodigestive anastomosis, requiring multiple interventions. The patient is currently listed for liver transplantation because of diffuse hepatolithiasis.

In two of the five patients who developed bile leakage as an early complication, fistulae remained unresolved for >30 days. One of these patients had a common bile duct stenosis, required a biliodigestive anastomosis and thereafter had recurrent stones in the common hepatic duct and required endoscopic balloon dilatation with percutaneous extraction of bile duct stones. Some patients had more than one complication (Table 6).

Overall survival at 5 years was 94.5%. Of the two patients who died within 5 years of surgery, one had developed cholangiocarcinoma 6 months after surgery and the other demonstrated chronic renal failure 53 months after surgery. Overall survival at 10 years was 90.5%; a third patient died of acute pneumonia within 10 years of surgery. Three other patients died after the 10-year follow-up, all for reasons not attributable to surgery or biliary stones.

Discussion

In the past 10 years, the introduction of new imaging technology in clinical practice has improved the diagnosis of hepatolithiasis. Magnetic resonance imaging (MRI) and MRCP are non-invasive tests that not only allow the clear diagnosis of hepatolithiasis, but also facilitate the complete evaluation of the biliary system and the degree of liver atrophy. This allows for better planning of treatment for patients and clarifies indications for surgery.⁷ In the present series, US continued to play a leading role in the diagnosis of hepatolithiasis, but since the advent of MRCP, almost all patients have been evaluated using MRI or MRCP as the main diagnostic tool.

There are many treatment modalities for intrahepatic stones. The most frequently used are endoscopic and percutaneous treatments, with or without lithotripsy, by choledochoscopy and surgery (hepatectomy or biliary bypass). Currently, there is no consensus on which treatment best achieves adequate clearing of the biliary tree and lower rates of residual stones, recurrence, morbidity and mortality. The main objectives in the management of this disease are to extract all intra- and extrahepatic stones, and to remove any bile duct stenosis, affected bile duct drainage areas and all atrophic segments. The chronic inflammatory process that causes stones in the intrahepatic bile ducts is a well-known risk factor for cholangiocarcinoma and the only treatment to eliminate this bile duct pathology is liver resection.¹⁵

In the present series, the rate of stone clearance at 30 days prior to surgery was 91.4% and all failures occurred at an extrahepatic level. After percutaneous or endoscopic treatment, the clearance rate rose to 98.1% postoperatively. These results are comparable with those in a series reported by Uenishi et al.,¹⁶ in which 86 patients underwent hepatectomy for hepatolithiasis, achieving an initial clearance rate of 88%, which increased after postoperative choledochoscopy with lithotripsy. These clearance rates are considerably higher than those achieved by a single percutaneous treatment (44%) or single endoscopic treatment (57%) in a comparative study by Cheon et al.¹⁷ Treatment with ERCP alone reported clearance rates that did not exceed 70%.^18,19 The development of more effective lithotripsy techniques, such as laser and percutaneous electrohydraulic lithotripsy, have resulted in bile duct clearance rates of 73–96%, but recurrence rates are high.^9,20–24

The present series recorded no postoperative mortality, but did show high morbidity, most of which concerned infections. This increased rate may reflect the high number of patients who underwent surgery with recurrent cholangitis and/or medically treated episodes of acute cholangitis. The high rate of bile leakage after elective hepatectomy is probably associated with the substantial proportion (40.4%) of patients with Caroli's disease, as similar associations have been previously reported.^25–27

In the present study, the recurrence rate was 13.5% and the median follow-up was 63.7 months. Recurrences occurred mainly in the extrahepatic bile ducts and were successfully treated with ERCP. The rate of recurrence in the intrahepatic bile ducts was 5.8%, which is lower than comparative rates reported in other studies using similar treatment.^12,28 The higher rates of recurrence probably reflect the development of biliary stenosis, which occurs in 42–96% of patients with hepatolithiasis and is the main factor in recurrence. Consequently, when treatment does not involve removing the stenotic bile duct, recurrence is likely to occur relatively soon.^9,29–32

In the present study, an analysis of causes associated with hepatolithiasis revealed that 50.0% of patients showed evidence of Caroli's disease in biopsy specimens. Similar results have been reported by other authors in occidental countries,^19,33 where patients have a different pathophysiology.³⁴

The left liver lobe emerged as the most frequent location of hepatolithiasis in the present series, which is consistent with findings reported by other authors.^1,6,7,35 This can be explained by the acute angle taken by the left hepatic duct when it reaches the confluence, which results in more bile stasis when associated with stenosis of the biliary system. However, the frequency of left hepatolithiasis in the present series (57.7%) was lower than those reported by Chen et al.³⁶ (77%) in a series of 103 patients and Jiang et al.³⁷ (72%) in 106 patients. This difference reflects the higher frequency of right hepatolithiasis in the present series (32.7%). This higher frequency of right side involvement, especially in the posterior segments, may be explained by the presence of anatomic variations in which the posterior segment artery comes from the common hepatic artery or is divided so that a branch passes through the gallbladder fossa, which can be injured during cholecystectomy, resulting in ischaemia of the bile ducts in that area. It is the present authors’ belief that the history of previous cholecystectomy in patients with iatrogenic vascular injuries during cholecystectomy is related to right side involvement; however, further study is required to substantiate this effect. It is noteworthy that the average period between cholecystectomy and the diagnosis of hepatolithiasis was 18 years. Balandraud et al.³⁸ found anatomic variations in 14 patients with right hepatolithiasis, whereby a cranial shift of the right sectorial ducts proximal to the hepatic confluence, present in six of the patients, was an independent risk factor for hepatolithiasis and explained the occurrence of bile stasis and formation of hepatolithiasis.

The surgical protocol used in the present institution favours anatomic over non-anatomic liver resection as the main approach. This is supported by studies that have shown the former to have benefit in terms of reducing the incidence of residual lithiasis, bile leakage and recurrence.^37,39

In the present series, morbidity reached 30.8% and most complications were mild and rapidly resolved; these results are comparable with the international standard.⁴⁰ No postoperative liver failure occurred in the present series, although two patients had liver cirrhosis.

The treatment of patients with bilateral hepatolithiasis is controversial and patients with bilateral disease show higher rates of recurrence. Therapeutic approaches range from hepatectomy to complex bilioenteric anastomosis using a subcutaneous loop to remove residual or recurrent lithiasis.^41–43 In the present series, only four patients had bilateral hepatolithiasis, three of whom underwent hepatectomy for severe involvement in segments on both sides. The remaining patient underwent a left lateral sectionectomy in addition to the extraction of the contralateral lithiasis by intraoperative choledochoscopy. To date, none of these patients have demonstrated residual stones or recurrence.

The association between hepatolithiasis and cholangiocarcinoma is well known; hepatolithiasis has been reported to occur in 2–10% of cholangiocarcinoma patients.¹⁵ Patients with a preoperative diagnosis of cholangiocarcinoma were excluded from the present series. However, in one patient, an intraoperative incidental diagnosis was confirmed by frozen biopsy. The patient underwent radical oncological surgery with lymph node dissection and achieved survival of 6 months. None of the patients in the present series developed cholangiocarcinoma during follow-up. However, Chen et al.⁴⁴ observed that 9% of patients with and 3% of those without residual stones developed cholangiocarcinoma.

The prevention of hepatolithiasis and bile stasis at an early stage avoids the end result of progressive liver failure and cirrhosis. However, in patients with diffuse bilateral hepatolithiasis, disease recurrence can lead to repeated episodes of cholangitis, liver atrophy and biliary cirrhosis associated with portal hypertension and liver failure. Liver transplantation has become the best treatment option for patients with end-stage liver disease.⁴⁵ Patients with diffuse hepatolithiasis and secondary biliary cirrhosis should consider liver transplantation, which eliminates the possibility of developing cholangiocarcinoma, as an alternative treatment.

The present authors believe that hepatectomy is the best treatment when hepatolithiasis is located in a segment or lobe associated with biliary stenosis or atrophy of a segment or liver lobe. Hepatectomy should also be considered as the treatment of choice when there is suspicion or certainty of cholangiocarcinoma. Endoscopic or percutaneous treatment should be considered as a first approach in patients with a poor performance status and in patients with postoperative residual or recurrent lithiasis.

Conclusions

This study showed that anatomic hepatectomy is an effective and safe procedure for hepatolithiasis associated with intrahepatic biliary stenosis, liver abscess or liver atrophy, and that it has low morbidity and no mortality. Longterm follow-up showed hepatectomy in these patients to be associated with a low rate of recurrence.

Conflicts of interest

None declared.