RETRACTED: Long-Term Outcomes of Hepatectomy for Bilateral Hepatolithiasis with Three-Dimensional Reconstruction: A Propensity Score Matching Analysis

Abstract

Several months after the publication of the article entitled, “Long-Term Outcomes of Hepatectomy for Bilateral Hepatolithiasis with Three-Dimensional Reconstruction: A Propensity Score Matching Analysis” by Guan T, Fang C, Mo Z, Xiang N, Yang J, Zeng N. (J Laparoendosc Adv Surg Tech A. 2016;26(9):680–688. DOI: 10.1089/lap.2016.0165; PMID: 27420918), the lead author of the paper, Tianpei Guan, contacted the Editor of the Journal and requested a formal retraction of the article as the authors now believe that they used the wrong statistical methods and are concerned that the results might be inconclusive or wrong. Dr. Guan also stated— post-publication—that some of the patients underwent conventional hepatectomy in a different department than their own, and therefore, could present a potential conflict of interest.

After careful consideration of Dr. Guan's request, the Editor-in-Chief of Journal of Laparoscopic and Advanced Surgical Techniques agreed to officially retract the article from the literature based on these concerns.

Journal of Laparoscopic and Advanced Surgical Techniques

and its editorial leadership is committed to upholding the strictest standards of peer review and scientific publishing.

Introduction

Hepatolithiasis, characterized by the presence of stones in the intrahepatic duct, is prevalent in the Asia-Pacific region, where morbidity reaches 20%–50% in patients who underwent cholecystectomy.¹ Although it is rare in Europe and North America,^2,3 the increasing trend caused by immigration and westernization of diet has been described in recent publications.^3,4 Without appropriate treatment, the intrahepatic stones may lead to repeated cholangitis, biliary strictures, liver abscesses, cirrhosis, and even cholangiocarcinoma.^5,6 Bilateral hepatolithiasis has worse outcomes compared with unilateral hepatolithiasis, with higher rate of stone recurrence.⁷

Treatment for bilateral hepatolithiasis based on conventional radiological technology, such as percutaneous transhepatic cholangioscopic lithotripsy (PTCL), endoscopic retrograde cholangiopancreatography (ERCP), and hepatectomy, has been designed.⁸ However, as much as 15%–59% of the patients show residual and/or recurrent stones.^9–11 Completely removing stones, establishing effective drainage to the biliary tree, and eliminating underlying infection foci are widely accepted as the principles to prevent recurrent cholangitis and cholangiocarcinoma genesis.^9,12,13 Theoretically, surgical resection represents the most definitive treatment for hepatolithiasis as it can remove the stones, strictures, as well as atrophic and nonfunctioning segments, which harbor bacteria and serve as foci of potential infection simultaneously; in this case, the possibility of recurrent calculi and suppurative cholangitis is minimized.^14–16 However, the long-term outcome remains unsatisfactory.

Three-dimensional (3D) visualization technology developed rapidly and applied in clinical cases at present. It was reported that 3D visualization helped to display the morphology of intrahepatic bile ducts, especially for biliary stricture and calculi location, as well as the relationship between bile duct system and vessels.^17–19 Our pilot study indicated that Three-Divisional Visualization System (MI-3DVS) reduces stone residual rate in hepatolithiasis.²⁰ However, the specific impact of MI-3DVS on long-term outcomes of bilateral hepatolithiasis remains unknown.

In this retrospective study, we aimed to assess the effect of MI-3DVS on bilateral hepatolithiasis by comparing the perioperative and long-term outcomes of patients who underwent 3D technique-based hepatectomy with those treated with conventional liver resection. Meanwhile, potential risk factors associated with prognosis were explored preliminarily.

Materials and Methods

Patients

This study was approved by the Ethics Committee of Zhujiang Hospital, Southern Medical University (No.: 2014-GDYK-007) and exempted from requiring informed consent. Eligibility criteria were as follows: (1) age more than 18 years; (2) diagnosis for bilateral hepatolithiasis and liver resection treatment in our hospital; (3) absence of pathological or imaging evidence for cholangiocarcinoma; and (4) liver function classification Child-Pugh A or B. From January 2008 to October 2014, a total of 264 consecutive patients were diagnosed with bilateral hepatolithiasis and underwent liver resection as the primary treatment in Zhujiang Hospital. Of these, 6 cases were excluded because of cholangiocarcinoma, identified by preoperative imaging examination (4 cases) or pathological examination (2 cases), according to eligibility criteria. Of the 258 patients enrolled, 142 cases received MI-3DVS guided hepatectomy (group A), while the other 116 underwent conventional resection under the guidance of two-dimensional computed tomography imaging (CT) or magnetic resonance imaging (MRI) (group B). Demographic and clinical characteristics such as symptoms, underlying liver diseases, and serum biochemistries were recorded at the time of diagnosis.

To find and manage potential recurrent stones and/or residual stones in extrahepatic biliary ducts, patients were followed-up with ultrasonography, CT cholangiography, or magnetic resonance cholangiopancreatography (MRCP) every 2 to 3 months in the first year after operation. Then, follow-up was implemented annually, or whenever suggestive cholangitis recurred. For patients with stone recurrence during the follow-up period, rational treatment was offered according to the patient's condition and disease severity.

Diagnosis and definitions

Diagnosis of hepatolithiasis was mainly based on imaging examination. Bilateral hepatolithiasis was considered in patients with stones in bilateral intrahepatic bile ducts, regardless of extrahepatic stones and a history of biliary tract surgery. Disease classification was carried out according to the Guidelines for the Diagnosis and Management of hepatolithiasis recommended by the biliary tract surgery group of Chinese Medical Association (CMA).²¹ In this classification system, hepatolithiasis is divided into two types. Type I represents a localized disease with stones located in one or both lobes. Type II is a diffuse disease, which is divided into three subtypes as follows: type IIa involves no atrophy of the hepatic parenchyma or intrahepatic bile duct stricture; type IIb involves segmental atrophy and/or intrahepatic bile duct stricture; and type IIc involves biliary cirrhosis and portal hypertension. The severity of postoperative complications was estimated according to the Dindo–Clavien Classification system.²² Residual stones were considered when the calculi could not be removed totally at the first attempt (immediate residual rate), or cannot be removed completely after additional treatment such as extraction using T-tube tract, PTCL, or ERCP at least 3 months after surgery (final residual rate).Stone recurrence was diagnosed when there were stones on previously treated bile ducts, evidenced by ultrasonography, CT cholangiography, or MRCP.

Treatment

Preoperative preparation and three-divisional reconstruction

Stone locations, as well as biliary tree architectures, were evaluated preoperatively by ultrasonography, CT, ERCP, percutaneous transhepatic cholangiography (PTC), or magnetic resonance cholangiopancreatography (MRCP). For MI-3DVS group patients, 3D models were reconstructed on a Medical Imaging Three-Divisional Visualization platform allowing stone locations, strictures, atrophy segments, and individual anatomical features such as vessels and biliary ducts to be observed from different angles. Then, individualized hepatectomy was designed based on individualized segmentation, stone locations and strictures, and the volume of remnant liver. The process of 3D reconstruction and special requirements of the CT/MRI dataset by the system were described in our previous report.¹⁹ The types of this disease within the CMA classification were depicted intuitively in 3D visualization models (Fig. 1). The regions of interest, such as the bile duct, hepatic veins and arteries, and portal vein; number, size, and location of the stones; and location of strictures, were observed on 3D models. Then, an individualized hepatectomy was planned according to comprehensive anatomical structures, such as the stricture, distribution of calculi, variation of vessels, and atrophic segments.

FIG. 1.

Superiority of MI-3DVS in displaying intrahepatic structures and hepatolithiasis classification. Hepatolithiasis in two-dimensional computerized tomography images (A) and magnetic resonance images (B). In MI-3DVS (C), crucial anatomical structures such as hepatic vein (a), calculus (b), hepatic artery (c), and portal vein (d) were revealed clearly when the liver was hyalinized. In addition, hepatolithiasis classification was carried out with surface morphology of liver and portal hypertension symptoms. Type IIa without obvious pathological changes of liver parenchyma (D); (E) Type IIb with left lateral lobe atrophy, but portal hypertension (the portal vein was 11.5 mm in diameter); and (F) Type IIc, the liver parenchyma had diffuse atrophy with the splenic vein dilated (e).

Surgical resection procedures

Hepatectomy was carried out under general anesthesia, using a right subcostal incision with midline upward extension to the xiphoid process. During operation, the 3D model was displayed on a portable computer to guide the surgery indirectly, while traditional hepatectomy was performed based on the surgeon's clinical experience or imaging photographs. Intraoperative flexible choledochoscopy was routinely conducted to explore the biliary ducts and remove residual stones with forceps or mesh basket. Rigid choledochoscope pneumatic lithotripsy was used to crush and flush out large stones. For patients with biliary strictures in the remnant liver, reshaping surgery or internal stenting tube was conducted. Hepaticojejunostomy was performed for biliary drainage if needed. A T-tube was routinely placed through which postoperative cholangiography was performed about 2 weeks posttreatment. In case of residual stones, the T-tube was maintained at least 4 months before stone extraction through the tube tract, using a rigid choledochoscope or flexible choledochoscope. These operations were performed after infection was controlled with antibiotics in patients with cholangitis.

Propensity score matching analysis

To explore the association between treatment selection and prognosis in an observational nonrandomized study, potential variables related to treatment selection and associated with outcomes of liver resection, such as gender, age, Tbil, ALB, ALT, AST, HGB, platelet count, Child-Pugh class, presentations, and underlying diseases,^23–26 were comprehensively included in Propensity Score Matching analysis on the Stata 12.0 package. Probit regression with the selected variables was used to generate continuous propensity scores from 0 to 1. A 1:1 nearest-neighbor match between patients receiving 3D model guided hepatectomy and conventional liver resection was performed to decrease confounders in patient selection and screen out matched patients for subsequent analyses.

Statistical analysis

Student t and chi-square tests were adopted for continuous and categorical data, respectively. Mann–Whitney U test was conducted for ranked data. The Kaplan–Meier method with log-rank test was used to generate survival curves and compare the 2 groups. Factors associated with stone recurrence were identified by both univariate and multivariate analyses in Cox Proportional Hazards Regression Models. Statistical analysis was performed using the SPSS 13.0 software for Windows (SPSS). P < .05 was considered statistically significant.

Results

Patient characteristics before and after propensity score matching

A total of 258 patients with bilateral hepatolithiasis were enrolled in this study. Among these, 142 subjects (55 males and 87 females) underwent MI-3DVS guided hepatectomy, and 116 participants (45 males and 71 females) received resection without MI-3DVS. Patient ages were 52.85 ± 11.63 and 53.28 ± 11.91 years in MI-3DVS guided hepatectomy and non-MI-3DVS guided resection groups, respectively. Serum ALT (84.83 ± 107.17 U/l versus 55.13 ± 39.41 U/l, P = .009) and AST (85.65 ± 204.62 U/l versus 42.01 ± 26.60 U/l, P = .042) were significantly higher in patients treated with MI-3DVS guided hepatectomy compared with those undergoing non-MI-3DVS resection. No significant differences were found in CMA class, additional class, history of biliary tract surgery, Child-Pugh class, presentations, and underlying diseases in the two groups. After propensity score matching (PSM), the variables with overt differences before propensity matching and associated with survival/postoperative complications such as ALT and AST were effectively balanced. A total of 148 subjects were matched for subsequent analysis. There were 74 patients in group A (29 men and 45 women; age, 52.50 ± 11.90 years) and 74 patients in group B (32 men and 42 women; age, 54.43 ± 11.78 years). There were 16 (group A) and 17 (group B) patients who had experienced previous bile duct surgery (Table 1).

Table 1.

Clinical Characteristics of Patients in the Two Groups Before and After Propensity Score Matching

	Group (before PSM)			Group (after PSM)
Variables	A (n = 142)	B (n = 116)	P	A (n = 74)	B (n = 74)	P
Gender (female/male)	87/55	71/45	.992	45/29	42/32	.739
Age (years)	52.85 ± 11.63	53.28 ± 11.91	.774	52.50 ± 11.90	54.43 ± 11.78	.323
CMA class^a			.777^b			.307^b
IIa	66 (46.5%)	59 (50.9%)		37 (50.0%)	46 (62.2%)
IIb	59 (40.8%)	44 (37.9%)		29 (39.2%)	23 (31.1%)
IIc	18 (12.7%)	13 (11.2%)		8 (10.8%)	5 (6.8%)
Additional class			.256			.123
None	17 (12.0%)	20 (17.2%)		6 (8.1%)	14 (18.9%)
Ea	93 (65.5%)	75 (64.7%)		52 (70.3%)	50 (67.6%)
Eb	27 (19.0%)	14 (12.1%)		14 (18.9%)	7 (9.5%)
Ec	5 (3.5%)	7(6.0%)		2 (2.7%)	3 (4.1%)
ALT (U/L)	84.83 ± 107.17	55.13 ± 39.41	.009	56.75 ± 50.81	51.59 ± 54.59	.552
AST (U/L)	85.65 ± 204.62	42.01 ± 26.60	.042	48.69 ± 43.77	48.91 ± 56.61	.979
Tbil (μmol/L)	38.48 ± 45.43	35.82 ± 45.35	.640	34.64 ± 46.85	42.12 ± 39.08	.696
ALB (g/L)	38.55 ± 6.28	38.85 ± 6.04	.696	38.99 ± 5.95	38.95 ± 6.18	.972
Hb (g/L)	118.97 ± 29.09	120.55 ± 15.96	.490	121.43 ± 14.85	121.45 ± 16.35	.955
PLT (≥10¹¹/<10¹¹)	95/47	90/26	.058	54/20	59/15	.333
Previous biliary tract surgery (yes/no)	28/114	23/93	.983	16/58	17/57	.843
Child-Pugh (A/B)	127/17	104/12	.681	69/5	66/8	.384
Presentations
Acute cholangitis	43 (30.3%)	26 (22.4%)	.156	21 (28.4%)	16 (21.6%)	.343
Acute pancreatitis	17 (12.0%)	14 (12.1%)	.981	11 (14.9%)	7 (9.5%)	.314
Liver abscess	10 (7.0%)	5 (4.3%)	.351	3 (4.1%)	4 (5.4%)	>.999^c
Jaundice	36 (25.4%)	42 (36.2%)	.334	20 (27.0%)	24 (32.4%)	.472
Others	42 (29.6%)	37 (31.7%)	.688	22 (29.9%)	26 (35.1%)	.482
Underlying diseases
Hypertension	14 (9.9%)	16 (13.8%)	.327	7 (9.5%)	9 (12.2%)	.597
Diabetes	11 (7.7%)	7 (6.0%)	.591	4 (5.4%)	5 (6.8%)	>.999^c
Liver fluke	3 (2.1)	1 (0.9%)	.419^c	0 (0/74)	1 (1.4%)	>.999^c

Evaluated according to the Chinese classification system recommended by biliary tract surgery group of Chinese Medical Association (CMA) in 2007.

Mann–Whitney U test.

Fisher^'s Exact Test.

ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Ea, combined with extrahepatic stones and the functioning of the Sphincter of Oddi is normal; Eb, combined with extrahepatic stones and the function of the Sphincter of Oddi is slack; Ec, combined with extrahepatic stones and the function of the Sphincter of Oddi is strictured; Hb, hemoglobin; PLT, platelet count; PSM, propensity score matching; Tbil, total bilirubin.

Procedures, intraoperative data, and postoperative complications in the PSM model

No procedure-related mortality was found in either group. The surgical procedures and intraoperative data are summarized in Table 2. Median (Range) volumes of blood loss (mL) in group A were significantly reduced compared with group B values (300 (30–1050) versus 400 (50–1080), P = .037). No significant differences were found in operation times.

Table 2.

Operative Procedures and Intraoperative Results of the Two Groups in Propensity Score Matching Model

	Group
	A (n = 74)	B (n = 74)	P
Variables
Left lateral segmentectomy	24 (32.4%)	19 (25.7%)	.365
Left hemihepatectomy	20 (27.0%)	22 (29.7%)	.715
Right hemihepatectomy	17 (23.0%)	15 (20.3%)	.690
Right posterior sectionectomy	7 (9.5%)	8 (10.8%)	.785
Nonanatomic hepatectomy	6 (8.1%)	10 (13.5%)	.290
Combined with choledochojejunostomy	15 (20.3%)	19 (25.7%)	.434
Combined with cholecystectomy	18 (24.3%)	12 (16.2%)	.220
Combined with CBDE	62 (83.8%)	52 (70.3%)	.051
Combined with splenectomy	5 (6.8%)	6 (8.1%)	.745
Duration, mean ± SD, minutes	322.57 ± 112.94	303.51 ± 104.19	.288
Blood loss, Median (Range), mL	300 (30–1050)	400 (50–1080)	.037
Blood transfusion, mean ± SD, mL	72.97 ± 201.59	95.95 ± 253.98	.543

CBDE, common bile duct exploration.

Hospital stay durations in groups A and B were 16.72 ± 8.04 and 16.80 ± 8.40 days, respectively. During this time, 29 (group A) and 33 (group B) patients had complications, indicating procedure-related adverse event incidence rates of 39% and 45%, respectively; 9 in group A and 13 in group B had more than 1 adverse event. There were 25 (group A) and 26 (group B) patients, who had mild to moderate (grade I–II) complications; the corresponding incidence rates for severe (grade III–IV) events were 5% (4/74) and 9% (7/74), respectively. Major complications in group A included cholangitis (2), liver failure (1), bile leakage (3), intra-abdominal infection (6), incision infection (15), postoperative obstruction (2), and pleural effusion (10). The corresponding complications in group B were cholangitis (5), liver failure (5), bile leakage (6), intra-abdominal infection (5), incision infection (10), and pleural effusion (18). Interestingly, rates of liver failure and pleural effusion in group A were lower compared with group B values, but with no statistically significant difference. Detailed data are provided in Table 3.

Table 3.

Postoperative Results and Long-Term Outcomes of the Two Groups in Propensity Score Matching Model

	Group
	A (n = 74)	B (n = 74)	P
Postoperative results
Hospital stays. Mean ± SD, days	16.72 ± 8.04	16.80 ± 8.40	.952
Immediate clearance, n (%)	57 (77)	39 (53)	.002
Final clearance, n (%)	70 (95)	59 (80)	.007
Stone clearance in patients with history of surgery
Immediate clearance, n/n (%)	16/16 (100)	9/17 (53)	.003^a
Final clearance, n/n (%)	16/16 (100)	13/17 (77)	.103^a
Postoperative complications (total), n (%)	29 (39.2)	33 (44.6)	.505
Cholangitis, n (%)	2 (2.7)	5 (6.8)	.442^a
Liver failure, n (%)	1 (1.4)	5 (6.8)	.095^a
Bile leakage, n (%)	3 (4.1)	6 (8.1)	.302^a
Intra-abdominal infection, n (%)	6 (8.1)	5 (6.8)	.754
Incision infection, n (%)	15 (20.3)	10 (13.5)	.273
Postoperative obstruction, n (%)	2 (2.7)	0 (0/74)	.497^a
Pleural effusion, n (%)	10 (13.5)	18 (24.3)	.093
Multiple complications, n (%)	9 (12.2)	17 (17.6)	.355
Dindo–Clavien Classification
Mild–moderate (grade I–II), n (%)	25 (33.8)	26 (35.1)	.863
Severe (grade III–IV), n (%)	4 (5.4)	7 (9.5)	.347
Long-term outcomes
Follow-up periods, Median (Range), months	55.0 (9–71)	51.0 (9–71)	.068^b
Stone recurrence, n (%)	15 (20)	19 (26)	.434
Cholangitis attacks, n (%)	1 (1.4)	4 (5.4)	.366^a
Carcinogenesis, n (%)	1 (1.4)	1 (1.4)	>.999^a

Fisher^'s exact test.

Mann–Whitney U test.

Stone clearance and long-term outcome

Median (Range) follow-up times in the 2 groups were 55 (9–71) months and 51 (9–71) months, respectively. At the first attempt, stones in 57 patients (group A) and 39 patients (group B) were successfully removed, respectively, indicating immediate clearance rates of 77% and 53%, respectively. After additional treatment, rates of complete removal increased to 95% and 80% in groups A and B, respectively. Both immediate and final clearance rates were significantly higher in group A than group B (all P < .05). There were 16 and 17 patients with a history of bile duct surgery in groups A and B, respectively. All the stones in group A were completely removed at the first attempt. However, immediate and final clearance rates in group B were 53% (9/17) and 77% (13/17), respectively. No significant difference was found in final clearance rate between both subgroups.

During the follow-up period, stone recurrence was found in 15 and 21 individuals in groups A and B, respectively, indicating recurrence rates of 20% and 26%, respectively. One and 4 patients in groups A and B had cholangitis, respectively, that is, incidence rates of 1.4% and 5.4%, respectively. There is 1 patient in each group diagnosed with carcinogenesis, indicating incidence rates of 1.4% and 1.4% in groups A and B, respectively. No statistical differences were found in cholangitis and carcinogenesis incidence (all P > .05) between the 2 groups (Table 3).

Recurrence-free survival and prognostic predictors

The 1,2,3,4,5,6-year recurrence-free survival rates for groups A and B were 98.6%, 94.5%, 88.1%, 82.9%, 75.5%, 75.5% and 98.6%, 86.0%, 82.9%, 74.7%, 68.9%, 59.0%, respectively, indicating no significant differences between the two groups (P = .207) (Fig. 2).

FIG. 2.

Recurrence-free survival of patients. The patients treated with three-dimensional guided hepatectomy experienced a slightly improved 6-year survival without stone recurrence compared with the conventional hepatectomy group (75.5% versus 59.0%), although the difference was not statistically significant (P = .207).

The risk factors associated with stone recurrence were non-MI-3DVS and CMA class IIc, with Hazard Ratios of 2.00 (95% confidence interval [CI] = 1.02–3.95, P = .045) and 10.60 (95% CI = 4.45–25.25, P < .001), respectively. Detailed information regarding these parameters is found in Table 4.

Table 4.

Univariate and Multivariate Analysis for Recurrence-Free Survival

	Univariate analysis		Multivariate analysis
Variable	Hazard Ratio (95% CI)	P	Hazard Ratio (95% CI)	P
Intervention (non-3D versus 3D)	0.66 (0.79–2.96)	.221	2.00 (1.02–3.95)	.045
Gender (male versus female)	0.88 (0.46–1.71)	.771
Age (years)	1.01 (0.98–1.04)	.433
ALT (U/L)	1.00 (0.99–1.00)	.386
AST (U/L)	1.00 (0.99–1.00)	.328
Tbil (μmol/L)	0.99 (0.98–1.00)	.156
ALB (g/L)	0.97 (0.98–1.03)	.233
Hb (g/L)	1.01 (0.98–1.03)	.608
PLT (≥10¹¹ versus <10¹¹)	1.89 (0.74–4.87)	.186
Previous biliary tract surgery (yes versus no)	1.32 (0.60–2.90)	.492
Child-Pugh (A/B)	1.03 (0.31–3.34)	.968
CMA class^a
IIa			—	—
IIb (versus IIa)	0.11 (0.05–0.27)	<.001	1.89 (0.87–4.12)	.107
IIc (versus IIa)	0.20 (0.09–0.45)	<.001	10.60 (4.45–25.25)	<.001
Additional class (versus none)
None
Ea	0.66 (0.29–1.53)	.341
Eb	0.33 (0.08–1.27)	.105
Ec	0.82 (0.10–6.67)	.853

Evaluated according to the Chinese classification system recommended by biliary tract surgery group of Chinese Medical Association (CMA) in 2007.

ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Hb, hemoglobin; PLT, platelet count; Tbil, total bilirubin.

Discussion

To date, ultrasonography, CT, and MRCP remain the major imaging techniques used for intrahepatic stones²⁷; these modalities do not allow a comprehensive assessment of anatomical characteristics and spatial relationships of an individual patient. In contrast, ERCP and PTC are invasive modalities and not suitable for surgery planning and stimulation.²⁸ Excitingly, MI-3DVS–based diagnosis and classification of hepatolithiasis are sensitive and reliable.²⁹

Propensity Score Matching analysis is an effective method to balance the covariates in treatment and control groups; therefore, differences, which may lead to biased estimates of treatment effects in observational studies, are reduced.^30,31 In the present study, serum AST and ALT levels, which predict liver disease severity³² and are associated with survival and postoperative complications,^23–26 were effectively balanced by Propensity Score Matching analysis.

Intraoperative blood loss and postoperative complications are important parameters in comparing surgical options. This study demonstrated that intraoperative bleeding was slightly lower in patients who underwent MI-3DVS guided hepatectomy than in those administered conventional liver resection for intrahepatic stones (Median [Range], mL, 300 [30–1050] versus 400 [50–1800], P = .037). This difference could be explained by the optimized hepatectomy plan and surgical approach, with directions of various vessels intuitively observed. In addition, the incidence rates of postoperative complications (39.2% versus 44.6%, P = .505) and severe events were similar between the 2 groups (5.4% versus 9.5%, P = .347), similar to recent values reported for hepatectomy complications in bilateral intrahepatic stone (about 46%).¹³

Residual stones remain one of the most important challenges of hepatolithiasis treatment,^8,33 although their incidence has been markedly reduced in the past decades.^13,34 In the present study, both immediate and final stone clearance rates were significantly higher in group A compared with group B values. These differences may be attributed to the application of 3D visualization technique, which may compensate for the loss of tactile sensation in open hepatectomy; indeed, this method offers superior assessment over tactile sensation in detecting and locating stones in the deep liver parenchyma.²⁰ In contrast with recent publications, immediate clearance rates in the 2 groups were consistent to published data on bilateral or complicated hepatolithiasis.^13,34,35 However, final clearance rates in group A were significantly higher than those of group B and as high as values obtained with laparoscopic or open hemihepatectomy for left lobe hepatolithiasis (about 94%).³⁶ The advantage of MI-3DVS for remarkable final clearance could be explained by an ideal approach to extract the knotty stones, which otherwise could not be appropriately removed.

Stone recurrence is a common postoperative adverse event; similar incidence rates were obtained between groups in this work (20% versus 26%, P = .434). Previous reports suggested that stone recurrence rates and recurrent cholangitis events in patients with partial hepatectomy for complicated hepatolithiasis are about 12%–34%.^8,13 Thus, MI-3DVS is limited in reducing stone recurrence, although more stricture and dilatation, which are primary causes for stone recurrence,^37,38 could be diagnosed and managed with the guidance of reconstruction system.^16,19 However, the incidence of stricture-associated recurrent cholangitis (1.4%) and cholangiocarcinoma (1.4%) was lower in this study compared with recently published data (10%–32% and 2.4%–12%, respectively).^8,13,39

Risk factors associated with stone recurrence were reported to be bile duct stricture, dilatation, liver atrophy, residual stones, and Sphincter of Oddi dysfunction.^8,37,40 In the present study, Cox regression analysis identified intervention (Hazard Ratio [HR] = 2.00, 95% CI = 1.02–3.95, P = .045) and CMA class IIc (HR = 10.60, 95% CI = 4.45–25.25, P < .001) as risk factors for recurrence-free survival. The definition of CMA classification system, which is helpful for treatment selection, resection optimization, and improved progression,²¹ could explain the differences between our data and previous findings.^8,37,40–42 Indeed, the CMA pathological evolution-based classification system, recommended by the biliary tract surgery group of CMA in 2007, considers strictures, location of stones, liver atrophy, and cirrhosis synthetically. Therefore, we did not include these variables separately in the study design.

A few limitations of this study should be mentioned. First, its retrospective nature; however, application of Propensity Score Matching analysis could reduce treatment selection bias effectively.³¹ A randomized controlled trial is needed to confirm our findings. Then, this study found no impact of MI-3DVS on some long-term outcomes such as carcinogenesis, and longer follow-up is needed. Finally, hepatectomy and rigid choledochoscopy are experience-dependent approaches, and the present study might be not repeatable by other specialists.

In conclusion, 3D visualization technique–based hepatectomy for bilateral hepatolithiasis provides comparable results with less blood loss and excellent, immediate, final clearance rates. Therefore, MI-3DVS is a feasible and reliable tool for the management of bilateral hepatolithiasis, although it could not decrease postoperative complications and prolong recurrence-free survival.

Footnotes

Acknowledgments

This work was supported by The National High Technology Research and Development Program of China (863 Program) (Grant No.: 2012AA021105) and the United Fund of National Natural Science Foundation of China and Government of Guangdong Province (Grant No.: U1401254).

Disclosure Statement

No competing financial interests exist.

References

Catena

, Aldrighetti

, Finazzi

, Arzu

, Arru

, Pulitano

, Ferla

. Treatment of non-endemic hepatolithiasis in a Western country. The role of hepatic resection. Ann R Coll Surg Engl, 2006; 88:383–389.

Tsui

, Chan

, Wong

, Lo

, Yeung

, Lee

. Hepatolithiasis and the syndrome of recurrent pyogenic cholangitis: Clinical, radiologic, and pathologic features. Semin Liver Dis, 2011; 31:33–48.

Tabrizian

, Jibara

, Shrager

, Schwartz

, Roayaie

. Hepatic resection for primary hepatolithiasis: A single-center Western experience. J Am Coll Surg, 2012; 215:622–626.

Tsui

, Lam

, Lee

, Chan

. Primary hepatolithiasis, recurrent pyogenic cholangitis, and oriental cholangiohepatitis: A tale of 3 countries. Adv Anat Pathol, 2011; 18:318–328.

Endo

, Gonen

, Yopp

, Dalal

, Zhou

, Klimstra

, D'Angelica

, DeMatteo

, Fong

, Schwartz

, Kemeny

, O'Reilly

, Abou-Alfa

, Shimada

, Blumgart

, Jarnagin

. Intrahepatic cholangiocarcinoma: Rising frequency, improved survival, and determinants of outcome after resection. Ann Surg, 2008; 248:84–96.

Tyson

, El-Serag

. Risk factors for cholangiocarcinoma. Hepatology, 2011; 54:173–184.

Lin

, Lin

, Ko

, Chen

, Kuo

. Hepatic resection for bilateral hepatolithiasis: A 20-year experience. ANZ J Surg, 2013; 83:978–984.

Tan

, Tan

, Chen

, Zhu

, Leng

, Dong

. Endoscopic or laparoscopic approach for hepatolithiasis in the era of endoscopy in China. Surg Endosc, 2015; 29:154–162.

Kim

, Kim

, Suh

, Lee

, Park

, Lee

, Kim

, Bak

. Cholangiocarcinoma risk as long-term outcome after hepatic resection in the hepatolithiasis patients. World J Surg, 2015; 39:1537–1542.

10.

Uchiyama

, Kawai

, Ueno

, Ozawa

, Tani

, Yamaue

. Reducing residual and recurrent stones by hepatectomy for hepatolithiasis. J Gastrointest Surg, 2007; 11:626–630.

11.

Otani

, Shimizu

, Chijiiwa

, Ogawa

, Morisaki

, Sugitani

, Yamaguchi

, Tanaka

. Comparison of treatments for hepatolithiasis: Hepatic resection versus cholangioscopic lithotomy. J Am Coll Surg, 1999; 189:177–182.

12.

Kusano

, Isa

, Ohtsubo

, Yasaka

, Furukawa

. Natural progression of untreated hepatolithiasis that shows no clinical signs at its initial presentation. J Clin Gastroenterol, 2001; 33:114–117.

13.

Yang

, Lau

, Lai

, Yang

, Zhang

, Yang

, Lu

, Wu

. Hepatectomy for bilateral primary hepatolithiasis: A cohort study. Ann Surg, 2010; 251:84–90.

14.

Vetrone

, Ercolani

, Grazi

, Ramacciato

, Ravaioli

, Cescon

, Varotti

, Del Gaudio

, Quintini

, Pinna

. Surgical therapy for hepatolithiasis: A Western experience. J Am Coll Surg, 2006; 202:306–312.

15.

Sakpal

, Babel

, Chamberlain

. Surgical management of hepatolithiasis. HPB (Oxford), 2009; 11:194–202.

16.

Xie

, Fang

, Huang

, Fan

, Pan

, Peng

. Application of three-dimensional reconstruction and visible simulation technique in reoperation of hepatolithiasis. J Gastroenterol Hepatol, 2013; 28:248–254.

17.

Dong

, Fang

, Wu

, Cai

, Fan

, Yang

, Zeng

, Zhong

, Bao

. Management complicated hepatolithiasis with operative rigid choledochoscope guided by computed tomography (CT)-based 3D reconstruction technique. Hepatogastroenterology, 2014; 61:1556–1562.

18.

, Fang

, Liu

, Cai

, Fan

, Yang

, Zeng

, Zhong

, Bao

. 3D reconstruction aids surgery for complicated hepatolithiasis. Hepatogastroenterology, 2014; 61:613–622.

19.

Fang

, Li

, Wang

, Fan

, Zhong

. Computer-aided rigid choledochoscopy lithotripsy for hepatolithiasis. J Surg Res, 2015; 195:105–112.

20.

Fang

, Liu

, Fan

, Yang

, Xiang

, Zeng

. Outcomes of hepatectomy for hepatolithiasis based on 3-dimensional reconstruction technique. J Am Coll Surg, 2013; 217:280–288.

21.

Feng

, Zheng

, Xia

, Ma

, Wang

, Bie

, Dong

. Classification and management of hepatolithiasis: A high-volume, single-center's experience. Intractable Rare Dis Res, 2012; 1:151–156.

22.

Clavien

, Barkun

, de Oliveira

, Vauthey

, Dindo

, Schulick

, de Santibanes

, Pekolj

, Slankamenac

, Bassi

, Graf

, Vonlanthen

, Padbury

, Cameron

, Makuuchi

. The Clavien-Dindo classification of surgical complications: Five-year experience. Ann Surg, 2009; 250:187–196.

23.

Jin

, Zhu

, Liao

, Li

, Liao

, He

. Elevated preoperative aspartate aminotransferase to lymphocyte ratio index as an independent prognostic factor for patients with hepatocellular carcinoma after hepatic resection. Oncotarget, 2015; 6:19217–19227.

24.

Zhang

, Zhao

, Liu

, Mao

, Bai

. Factors affecting the recurrence and survival of hepatocellular carcinoma after hepatectomy: A retrospective study of 601 Chinese patients. Clin Transl Oncol, 2015.

25.

Zheng

, Li

, Yu

, Ying

, Zhang

, Lu

, Zhou

. Risk factors, treatment and impact on outcomes of bile leakage after hemihepatectomy. ANZ J Surg, 2015 [Epub ahead of print]; DOI: 10.1111/ans.

26.

Xie

, Yao

, Chen

, Lu

, Lv

, Wang

, Zhang

, Lu

. Reduced red blood cell count predicts poor survival after surgery in patients with primary liver cancer. Medicine (Baltimore), 2015; 94:e577.

27.

Ananthakrishnan

, Saeian

. Caroli's disease: Identification and treatment strategy. Curr Gastroenterol Rep, 2007; 9:151–155.

28.

Nagorney

. The impact of hepatic venous anatomy on the hepatic remnant: Need for assessment?. Surgery, 2010; 147:811.

29.

, Fang

, Fan

, Wu

, Zhong

. A comparative study of the diagnostic accuracy of the medical image three-dimensional visualization system, MRCP, CT and US in hepatolithiasis. Hepatogastroenterology, 2014; 61:1901–1907.

30.

D'Agostino

Jr.

Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med, 1998; 17:2265–2281.

31.

Austin

. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat, 2011; 10:150–161.

32.

Boleslawski

, Vibert

, Pruvot

, Le Treut

, Scatton

, Laurent

, Mabrut

, Regimbeau

, Adham

, Cosse

, Farges

. Relevance of postoperative peak transaminase after elective hepatectomy. Ann Surg, 2014; 260:815–820; discussion 820–811.

33.

Chen

, Huang

, Yang

, Yeh

, Wu

, Chou

, Yueh

, Nien

. Reappraisal of percutaneous transhepatic cholangioscopic lithotomy for primary hepatolithiasis. Surg Endosc, 2005; 19:505–509.

34.

Cheng

, Lee

, Sheen-Chen

, Huang

, Chen

. Treatment of complicated hepatolithiasis with intrahepatic biliary stricture by ductal dilatation and stenting: Long-term results. World J Surg, 2000; 24:712–716.

35.

Pan

, Xu

, Fang

, Deng

, Xu

. Surgical treatment of complicated hepatolithiasis using the ultrasound-guided fiberoptic choledochoscope. Surg Endosc, 2011; 25:497–502.

36.

, Ni

, Zhou

, Xie

, Hong

. Laparoscopic versus open left hemihepatectomy for hepatolithiasis. J Surg Res, 2015; 199:402–406.

37.

Suzuki

, Mori

, Yokoyama

, Nakazato

, Abe

, Nakanuma

, Tsubouchi

, Sugiyama

. Hepatolithiasis: Analysis of Japanese nationwide surveys over a period of 40 years. J Hepatobiliary Pancreat Sci, 2014; 21:617–622.

38.

Chen

, Tung-Ping Poon

, Liu

, Fan

, Wong

. Immediate and long-term outcomes of hepatectomy for hepatolithiasis. Surgery, 2004; 135:386–393.

39.

Park

, Jeong

, Lee

, Bang

, Lee

, Kwon

, Kim

, Shin

, Kim

, Park

. Risk factors for long-term outcomes after initial treatment in hepatolithiasis. J Korean Med Sci, 2013; 28:1627–1631.

40.

Tsuyuguchi

, Miyakawa

, Sugiyama

, Sakai

, Nishikawa

, Sakamoto

, Nakamura

, Yasui

, Mikata

, Yokosuka

. Ten-year long-term results after non-surgical management of hepatolithiasis, including cases with choledochoenterostomy. J Hepatobiliary Pancreat Sci, 2014; 21:795–800.

41.

Cheon

, Cho

, Moon

, Lee

, Shim

. Evaluation of long-term results and recurrent factors after operative and nonoperative treatment for hepatolithiasis. Surgery, 2009; 146:843–853.

42.

Zhang

, Lin

, Qian

, Zhou

. Identification of risk factors for intraoperative hemobilia and its correlation with early postoperative complications in patients with hepatolithiasis. Am J Surg, 2015; 209:260–267.