ORIGINAL ARTICLE ANZJSurg.com
Risk factors, treatment and impact on outcomes of bile leakage after hemihepatectomy Si-Ming Zheng, Hong Li, Gen-Cong Li, Dan-Song Yu, Dong-Jian Ying, Bin Zhang, Cai-De Lu and Xin-Hua Zhou Department of General Surgery, Ningbo Li Huili Hospital, Medical School of Ningbo University, Ningbo, China
Key words bile leakage, complication, hepatectomy. Correspondence Dr Xin-Hua Zhou, Department of General Surgery, Ningbo Li Huili Hospital, Medical School of Ningbo University, No. 57 Xingning Road, Ningbo, Zhejiang Province 315040, China. Email: [email protected] S.-M. Zheng MD; H. Li MD; G.-C. Li MD; D.-S. Yu MD; D.-J. Ying MD; B. Zhang MD; C.-D. Lu MD, PhD; X.-H. Zhou MD. Accepted for publication 24 February 2015. doi: 10.1111/ans.13073
Abstract Background: Risk factors for bile leakage after hemihepatectomy are unknown. Methods: A prospectively maintained database review identified patients undergoing hemihepatectomy between 1 January 2009 and 30 September 2014. Patients were divided into B/C and non-B/C bile leakage groups. Risk factors for bile leakage were predicted and assessments of their impact on patients were made. Results: Bile leakage occurred in 91 of the 297 patients (30.6%); 64 cases were classified as grade B bile leakage (21.5%) and three cases as grade C bile leakage (1.0%). Multivariate analysis confirmed that elevated preoperative alanine transaminase (ALT), positive bile culture during surgery, hilar bile duct plasty, bilioenteric anastomosis and laparoscopic surgery were risk factors for B/C grade bile leakage (P < 0.05). Percutaneous transhepatic biliary drainage (PTBD) and endoscopic nasobiliary drainage (ENBD) were protective factors for B/C grade bile leakage (P < 0.05). PTBD, ENBD and Kehr’s T-tube drainage could reduce the drainage volume and duration of drainage after bile leakage (P < 0.05). The incidence of wound infection, abdominal infection, major complications and the Clavien classification system score in the B/C bile leakage group were higher than those in the non-B/C bile leakage group (P < 0.05). Patients in the B/C bile leakage group also required prolonged hospitalization (P < 0.05). The mortality of two groups was similar (P > 0.05). Conclusion: Patient with elevated preoperative ALT, positive bile cultures during surgery, hilar bile duct plasty, bilioenteric anastomosis and laparoscopic surgery are more likely to complicate bile leakage. We should use biliary drainage such as preoperative PTBD, ENBD or intraoperative Kehr’s T-tube drainage to reduce and treat bile leakage in patients with high risk of bile leakage.
Introduction Hepatectomy is the current main treatment for benign and malignant liver tumours and hepatolithiasis. As technology and surgical procedures advance, complications associated with hepatectomy and subsequent mortality rates have declined. However, hemihepatectomy is still a high-risk procedure because of the large number of liver resections and the complexity of the operation. Bile leakage is a major complication of liver resection. Previous studies have shown that the rate of bile leakage after liver resection is 3.6–7.4%.1–6 However, without a uniform standard for the diagnosis © 2015 Royal Australasian College of Surgeons
of bile leakage prior to 2011, it is difficult to compare bile leakage studies. In 2011, the International Study Group of Liver Surgery (ISGLS) developed a standard for diagnosing bile leakage.7 Due to the sensitivity of the standard, bile leakage rates after hepatic resection increased considerably to 16%, 27.2% and 35.7% in the recent three articles.7–9 Certain types of hepatectomy can significantly influence the probability of bile leakage, such as anterior segmentectomy, central bisegmentectomy and total caudate lobectomy.3,6 Hemihepatectomy is a single and fixed type of hepatectomy, which could avoid the influence on researching the bile leakage risk factors, so the study on the bile leakage of hemihepatectomy may be more meaningful. ANZ J Surg •• (2015) ••–••
2
Methods Patient cohort All patients who underwent hemihepatectomy between 1 January 2010 and 30 September 2014 were selected for analysis from a prospectively maintained institutional hepatectomy database. Preoperative patient’s information included: age, sex, diagnosis, previous liver or biliary tract surgery, liver cirrhosis, diabetes, white blood cell count, serum albumin, serum alanine transaminase (ALT) and serum total bilirubin. Post-operative pathological diagnosis was regarded as standard. Previous liver or biliary tract surgery included cholecystectomy, choledocholithotomy and hepatectomy. A diagnosis of post-operative cirrhosis was also regarded as standard for cirrhosis.
Zheng et al.
Abdominal complications included wound infection, abdominal infections, intestinal obstruction, delayed gastric emptying, intraabdominal bleeding, renal failure and liver failure. Cardiovascular complications included arrhythmias, myocardial ischaemia and cardiac insufficiency. Pulmonary complications included pneumonia, atelectasis requiring thoracentesis drainage, acute pulmonary embolism and respiratory failure. Neurological complications include transient delirium and cerebrovascular events. The severity of complications was assessed using the Clavien classification system from 1 to 5.10 Using this algorithm, major complications represent those requiring endoscopic, radiologic or surgical intervention or greater and were defined as class 3 or higher.
Ethics Details of the procedures Preoperative percutaneous transhepatic biliary drainage (PTBD) or endoscopic nasobiliary drainage (ENBD) was indicated in the cases of total bilirubin over 171 μmol/L because of tumour obstruction or pressing bile duct and in some cases of acute suppurative cholangitis. One week after operation, it was confirmed by cholangiography that no bile leakage occurred so we removed the PTBD and ENBD tube. Experienced hepatobiliary surgeons performed all the surgeries. The number of hepatectomies performed per doctor exceeded 50 operations a year. Operation procedures included left or right hemihepatectomy, laparoscopic surgery, the pringle manoeuvre, segment 1 or 9 resection, hilar lymphadenectomy, hilar bile duct plasty and bilioenteric anastomosis. Open hepatic resections used clamp-crushing techniques or cavitational ultrasonic surgical aspirator (CUSA; Valleylab, Boulder, CO, USA) combined with diathermy coagulation. A harmonic scalpel (Ethicon Endo-Surgery, Cincinnati, OH, USA) was used to transect the liver parenchyma in cases of laparoscopic hepatectomy in combination with a Hem-o-lok clip (Weck Surgical Instruments, Durham, NC, USA). Hilar bile duct plasty, bilioenteric anastomosis, common bile duct and the bile duct stump were oversewn by 4-0 or 5-0 Vicryl (Ethicon, Cornelia, GA, USA). After liver resection, the liver raw surface was treated with argon beam coagulation (Valleylab). Small bile leakage sites were repaired by Z-suturing using 4-0 or 5-0 Prolene (Ethicon, Somerville, NJ, USA). After choledochus exploration of hepatolithiasis, Kehr’s T-tube (T-tube; Sewoon Medical, Seoul, Korea) drainage was indicated in the cases of cholangitis or residual calculi. In some cases of liver tumours, in order to avoid the stricture of bile duct or bilioenteric anastomosis, we used the T-tube to support the bile duct or bilioenteric anastomosis. T-tube would be removed 1.5–6 months after the operation. Drainage tubes were placed on the liver surface routinely and in the case of bilioenteric anastomosis or choledochotomy, we also put a drainage tube near the anastomosis or the common bile duct. The sample of bile culture during surgery was obtained from the gall bladder, common bile duct or left (right) hepatic duct.
Definition of complications Complications were defined as adverse events resulting in deviation from the normal postoperative course within 30 days after the operation. Bilirubin concentrations of the serum and drainage fluid were measured on post-operative days 1, 3, 5 and 7 if the drainage tubes were not removed. According to the ISGLS definition,7 bile leakage was diagnosed and graded into A, B and C levels based on severity.
The study protocol was approved by the ethics committee of Li Huili Hospital and written consent was obtained from all patients.
Statistical analysis Depending on the severity of bile leakage, patients were divided into two groups: B/C bile leakage and non-B/C bile leakage. Data analysis was carried out using SPSS version 16.0 (IBM SPSS China Inc., Shanghai, China). Continuous variables were analysed using the test of normality. If the variables were in accordance with normal distribution, the variables were expressed in the form of x ± s and the means between groups were compared using the independent sample t-test. If the variables were not in accordance with the normal distribution, the variables were expressed in the form of M (range) and the medians between groups were compared using the rank test. Categorical variables were expressed in the form of n (%); non-ranked data were compared using the chi-square test, whereas ranked data were compared by the rank test. Variables that had relevant associations with grade B/C bile leakage on these analyses (P < 0.15) were included in binary logistic regression analysis. Collinearity analysis was performed in these variables. Significance was defined as twotailed P value less than 0.05.
Results Demography and surgical procedures A total of 297 patients underwent hemihepatectomy at the Li Huili Hospital between 1 January 2010 and 30 June 2014. There were 128 men and 169 women and the median age of the sample was 57.5 years (range 53 years). Pathological diagnoses included 137 cases of hepatolithiasis, 77 cases of cholangiocarcinoma, which included 13 cases of hepatolithiasis and cholangiocarcinoma, 53 cases of hepatocellular carcinoma, 31 cases of metastatic liver tumours and 12 cases of other benign liver disease. Surgical procedures are shown in Table 1. The number of left hemihepatectomy, right hemihepatectomy and laparoscopic surgery were 242, 55 and 72, respectively.
Morbidity and mortality In total, 91 cases of bile leakage occurred (30.6%), including grade A bile leakage in 24 cases (8.1%), 64 cases of grade B bile leakage (21.5%) and three cases of grade C bile leakage (1.0%). The non-bile © 2015 Royal Australasian College of Surgeons
Bile leakage after hemihepatectomy
3
Table 1 Demography, surgical procedures and univariate analysis for grade B/C bile leakage Variable
Age (years) Preoperative white blood cell (109/L) Preoperative albumin (g/L) Preoperative alanine transaminase (units/L) Preoperative total bilirubin (mmol/L) Operating time (min) Intraoperative blood loss (mL) Gender Male Female Hepatolithiasis Yes No Previous liver or biliary tract surgery Yes No Diabetes Yes No Liver cirrhosis Yes No Percutaneous transhepatic biliary drainage or endoscopic nasobiliary drainage Yes No Type of hepatic resection Left hemihepatectomy Right hemihepatectomy Laparoscopic surgery Yes No Segment 1 or 9 resection Yes No Hilar lymphadenectomy Yes No Hilar bile duct plasty Yes No Bilioenteric anastomosis Yes No Pringle maneuver Yes No Kehr’s T-tube drainage Yes No Intraoperative blood transfusion Yes No Positive bile culture during surgery Yes No
All patients (n = 297)
B/C group (n = 67)
Non-B/C group (n = 230)
t, z or χ2
P value
57.50 (53.00) 5.70 (24.60) 39.16 ± 5.10 29.00 (599.00) 14.40 (110.70) 225.0 (410.0) 400.0 (4700.0)
58.00 (49.00) 5.80 (17.70) 38.53 ± 5.06 46.00 (598.00) 16.90 (77.80) 250.00 (350.0) 500.0 (1900.0)
57.00 (52.00) 5.70 (24.60) 39.35 ± 5.11 28.00 (555.00) 14.10 (110.70) 210.00 (410.0) 400.0 (4700.0)
29 (22.7%) 38 (22.5%) — 43 (31.4%) 24 (15.0%) — 21 (34.4%) 46 (19.5%) — 6 (22.2%) 61 (22.6%) — 3 (12.0%) 64 (23.5%) —
99 (77.3%) 131 (77.5%) — 94 (68.6%) 136 (85.0%) — 40 (65.6%) 190 (80.5%) — 21 (77.8%) 209 (77.4%)
1.494 0.955 1.164 2.497 1.229 3.500 1.068 0.001 — — 11.345 — — 6.189 — — 0.002 — — 1.145 —
0.135 0.340 0.245† 0.013 0.219
Risk factors, treatment and impact on outcomes of bile leakage after hemihepatectomy Si-Ming Zheng, Hong Li, Gen-Cong Li, Dan-Song Yu, Dong-Jian Ying, Bin Zhang, Cai-De Lu and Xin-Hua Zhou Department of General Surgery, Ningbo Li Huili Hospital, Medical School of Ningbo University, Ningbo, China
Key words bile leakage, complication, hepatectomy. Correspondence Dr Xin-Hua Zhou, Department of General Surgery, Ningbo Li Huili Hospital, Medical School of Ningbo University, No. 57 Xingning Road, Ningbo, Zhejiang Province 315040, China. Email: [email protected] S.-M. Zheng MD; H. Li MD; G.-C. Li MD; D.-S. Yu MD; D.-J. Ying MD; B. Zhang MD; C.-D. Lu MD, PhD; X.-H. Zhou MD. Accepted for publication 24 February 2015. doi: 10.1111/ans.13073
Abstract Background: Risk factors for bile leakage after hemihepatectomy are unknown. Methods: A prospectively maintained database review identified patients undergoing hemihepatectomy between 1 January 2009 and 30 September 2014. Patients were divided into B/C and non-B/C bile leakage groups. Risk factors for bile leakage were predicted and assessments of their impact on patients were made. Results: Bile leakage occurred in 91 of the 297 patients (30.6%); 64 cases were classified as grade B bile leakage (21.5%) and three cases as grade C bile leakage (1.0%). Multivariate analysis confirmed that elevated preoperative alanine transaminase (ALT), positive bile culture during surgery, hilar bile duct plasty, bilioenteric anastomosis and laparoscopic surgery were risk factors for B/C grade bile leakage (P < 0.05). Percutaneous transhepatic biliary drainage (PTBD) and endoscopic nasobiliary drainage (ENBD) were protective factors for B/C grade bile leakage (P < 0.05). PTBD, ENBD and Kehr’s T-tube drainage could reduce the drainage volume and duration of drainage after bile leakage (P < 0.05). The incidence of wound infection, abdominal infection, major complications and the Clavien classification system score in the B/C bile leakage group were higher than those in the non-B/C bile leakage group (P < 0.05). Patients in the B/C bile leakage group also required prolonged hospitalization (P < 0.05). The mortality of two groups was similar (P > 0.05). Conclusion: Patient with elevated preoperative ALT, positive bile cultures during surgery, hilar bile duct plasty, bilioenteric anastomosis and laparoscopic surgery are more likely to complicate bile leakage. We should use biliary drainage such as preoperative PTBD, ENBD or intraoperative Kehr’s T-tube drainage to reduce and treat bile leakage in patients with high risk of bile leakage.
Introduction Hepatectomy is the current main treatment for benign and malignant liver tumours and hepatolithiasis. As technology and surgical procedures advance, complications associated with hepatectomy and subsequent mortality rates have declined. However, hemihepatectomy is still a high-risk procedure because of the large number of liver resections and the complexity of the operation. Bile leakage is a major complication of liver resection. Previous studies have shown that the rate of bile leakage after liver resection is 3.6–7.4%.1–6 However, without a uniform standard for the diagnosis © 2015 Royal Australasian College of Surgeons
of bile leakage prior to 2011, it is difficult to compare bile leakage studies. In 2011, the International Study Group of Liver Surgery (ISGLS) developed a standard for diagnosing bile leakage.7 Due to the sensitivity of the standard, bile leakage rates after hepatic resection increased considerably to 16%, 27.2% and 35.7% in the recent three articles.7–9 Certain types of hepatectomy can significantly influence the probability of bile leakage, such as anterior segmentectomy, central bisegmentectomy and total caudate lobectomy.3,6 Hemihepatectomy is a single and fixed type of hepatectomy, which could avoid the influence on researching the bile leakage risk factors, so the study on the bile leakage of hemihepatectomy may be more meaningful. ANZ J Surg •• (2015) ••–••
2
Methods Patient cohort All patients who underwent hemihepatectomy between 1 January 2010 and 30 September 2014 were selected for analysis from a prospectively maintained institutional hepatectomy database. Preoperative patient’s information included: age, sex, diagnosis, previous liver or biliary tract surgery, liver cirrhosis, diabetes, white blood cell count, serum albumin, serum alanine transaminase (ALT) and serum total bilirubin. Post-operative pathological diagnosis was regarded as standard. Previous liver or biliary tract surgery included cholecystectomy, choledocholithotomy and hepatectomy. A diagnosis of post-operative cirrhosis was also regarded as standard for cirrhosis.
Zheng et al.
Abdominal complications included wound infection, abdominal infections, intestinal obstruction, delayed gastric emptying, intraabdominal bleeding, renal failure and liver failure. Cardiovascular complications included arrhythmias, myocardial ischaemia and cardiac insufficiency. Pulmonary complications included pneumonia, atelectasis requiring thoracentesis drainage, acute pulmonary embolism and respiratory failure. Neurological complications include transient delirium and cerebrovascular events. The severity of complications was assessed using the Clavien classification system from 1 to 5.10 Using this algorithm, major complications represent those requiring endoscopic, radiologic or surgical intervention or greater and were defined as class 3 or higher.
Ethics Details of the procedures Preoperative percutaneous transhepatic biliary drainage (PTBD) or endoscopic nasobiliary drainage (ENBD) was indicated in the cases of total bilirubin over 171 μmol/L because of tumour obstruction or pressing bile duct and in some cases of acute suppurative cholangitis. One week after operation, it was confirmed by cholangiography that no bile leakage occurred so we removed the PTBD and ENBD tube. Experienced hepatobiliary surgeons performed all the surgeries. The number of hepatectomies performed per doctor exceeded 50 operations a year. Operation procedures included left or right hemihepatectomy, laparoscopic surgery, the pringle manoeuvre, segment 1 or 9 resection, hilar lymphadenectomy, hilar bile duct plasty and bilioenteric anastomosis. Open hepatic resections used clamp-crushing techniques or cavitational ultrasonic surgical aspirator (CUSA; Valleylab, Boulder, CO, USA) combined with diathermy coagulation. A harmonic scalpel (Ethicon Endo-Surgery, Cincinnati, OH, USA) was used to transect the liver parenchyma in cases of laparoscopic hepatectomy in combination with a Hem-o-lok clip (Weck Surgical Instruments, Durham, NC, USA). Hilar bile duct plasty, bilioenteric anastomosis, common bile duct and the bile duct stump were oversewn by 4-0 or 5-0 Vicryl (Ethicon, Cornelia, GA, USA). After liver resection, the liver raw surface was treated with argon beam coagulation (Valleylab). Small bile leakage sites were repaired by Z-suturing using 4-0 or 5-0 Prolene (Ethicon, Somerville, NJ, USA). After choledochus exploration of hepatolithiasis, Kehr’s T-tube (T-tube; Sewoon Medical, Seoul, Korea) drainage was indicated in the cases of cholangitis or residual calculi. In some cases of liver tumours, in order to avoid the stricture of bile duct or bilioenteric anastomosis, we used the T-tube to support the bile duct or bilioenteric anastomosis. T-tube would be removed 1.5–6 months after the operation. Drainage tubes were placed on the liver surface routinely and in the case of bilioenteric anastomosis or choledochotomy, we also put a drainage tube near the anastomosis or the common bile duct. The sample of bile culture during surgery was obtained from the gall bladder, common bile duct or left (right) hepatic duct.
Definition of complications Complications were defined as adverse events resulting in deviation from the normal postoperative course within 30 days after the operation. Bilirubin concentrations of the serum and drainage fluid were measured on post-operative days 1, 3, 5 and 7 if the drainage tubes were not removed. According to the ISGLS definition,7 bile leakage was diagnosed and graded into A, B and C levels based on severity.
The study protocol was approved by the ethics committee of Li Huili Hospital and written consent was obtained from all patients.
Statistical analysis Depending on the severity of bile leakage, patients were divided into two groups: B/C bile leakage and non-B/C bile leakage. Data analysis was carried out using SPSS version 16.0 (IBM SPSS China Inc., Shanghai, China). Continuous variables were analysed using the test of normality. If the variables were in accordance with normal distribution, the variables were expressed in the form of x ± s and the means between groups were compared using the independent sample t-test. If the variables were not in accordance with the normal distribution, the variables were expressed in the form of M (range) and the medians between groups were compared using the rank test. Categorical variables were expressed in the form of n (%); non-ranked data were compared using the chi-square test, whereas ranked data were compared by the rank test. Variables that had relevant associations with grade B/C bile leakage on these analyses (P < 0.15) were included in binary logistic regression analysis. Collinearity analysis was performed in these variables. Significance was defined as twotailed P value less than 0.05.
Results Demography and surgical procedures A total of 297 patients underwent hemihepatectomy at the Li Huili Hospital between 1 January 2010 and 30 June 2014. There were 128 men and 169 women and the median age of the sample was 57.5 years (range 53 years). Pathological diagnoses included 137 cases of hepatolithiasis, 77 cases of cholangiocarcinoma, which included 13 cases of hepatolithiasis and cholangiocarcinoma, 53 cases of hepatocellular carcinoma, 31 cases of metastatic liver tumours and 12 cases of other benign liver disease. Surgical procedures are shown in Table 1. The number of left hemihepatectomy, right hemihepatectomy and laparoscopic surgery were 242, 55 and 72, respectively.
Morbidity and mortality In total, 91 cases of bile leakage occurred (30.6%), including grade A bile leakage in 24 cases (8.1%), 64 cases of grade B bile leakage (21.5%) and three cases of grade C bile leakage (1.0%). The non-bile © 2015 Royal Australasian College of Surgeons
Bile leakage after hemihepatectomy
3
Table 1 Demography, surgical procedures and univariate analysis for grade B/C bile leakage Variable
Age (years) Preoperative white blood cell (109/L) Preoperative albumin (g/L) Preoperative alanine transaminase (units/L) Preoperative total bilirubin (mmol/L) Operating time (min) Intraoperative blood loss (mL) Gender Male Female Hepatolithiasis Yes No Previous liver or biliary tract surgery Yes No Diabetes Yes No Liver cirrhosis Yes No Percutaneous transhepatic biliary drainage or endoscopic nasobiliary drainage Yes No Type of hepatic resection Left hemihepatectomy Right hemihepatectomy Laparoscopic surgery Yes No Segment 1 or 9 resection Yes No Hilar lymphadenectomy Yes No Hilar bile duct plasty Yes No Bilioenteric anastomosis Yes No Pringle maneuver Yes No Kehr’s T-tube drainage Yes No Intraoperative blood transfusion Yes No Positive bile culture during surgery Yes No
All patients (n = 297)
B/C group (n = 67)
Non-B/C group (n = 230)
t, z or χ2
P value
57.50 (53.00) 5.70 (24.60) 39.16 ± 5.10 29.00 (599.00) 14.40 (110.70) 225.0 (410.0) 400.0 (4700.0)
58.00 (49.00) 5.80 (17.70) 38.53 ± 5.06 46.00 (598.00) 16.90 (77.80) 250.00 (350.0) 500.0 (1900.0)
57.00 (52.00) 5.70 (24.60) 39.35 ± 5.11 28.00 (555.00) 14.10 (110.70) 210.00 (410.0) 400.0 (4700.0)
29 (22.7%) 38 (22.5%) — 43 (31.4%) 24 (15.0%) — 21 (34.4%) 46 (19.5%) — 6 (22.2%) 61 (22.6%) — 3 (12.0%) 64 (23.5%) —
99 (77.3%) 131 (77.5%) — 94 (68.6%) 136 (85.0%) — 40 (65.6%) 190 (80.5%) — 21 (77.8%) 209 (77.4%)
1.494 0.955 1.164 2.497 1.229 3.500 1.068 0.001 — — 11.345 — — 6.189 — — 0.002 — — 1.145 —
0.135 0.340 0.245† 0.013 0.219
