Author Affiliations: Department of Hepatobiliary Surgery, Beijing Tsinghua Changgung Hospital, Medical Center, Tsinghua University, Beijing 102218, China (Wang XD, Shi J and Dong JH); Hospital and Institute of Hepatobiliary Surgery, Chinese PLA General Hospital, Beijing 100853, China (Wang XD, Wang HG, Duan WD, Luo Y, Ji WB, Zhang N and Dong JH)

Corresponding Author: Jia-Hong Dong, MD, PhD, FACS, Department of Hepatobiliary Surgery, Beijing Tsinghua Changgung Hospital, Medical Center, Tsinghua University, China (Tel: +86-10-56118888; Fax: +86-10-56118500; Email: dongjiahong@mail.tsinghua.edu.cn)

 

© 2017, Hepatobiliary Pancreat Dis Int. All rights reserved.

doi: 10.1016/S1499-3872(17)60021-3

Published online May 18, 2017.

 

 

Acknowledgement: We thank Professor Li Fan, for her assistance in polishing the manuscript.

Contributors: WXD and DJH proposed the study. WXD, WHG and SJ performed the research, wrote the first draft, and contributed equally to this article. DWD, LY, JWB and ZN collected and analyzed the data. All authors contributed to the design and interpretation of the study and to further drafts. DJH is the guarantor.

Funding: This study was supported by grants from the China Postdoctoral Science Foundation (2014M562551), the National Key Technology R&D Program of China (2012BAI06B01) and the National S&T Major Project for Infectious Diseases of China (2012ZX10002-017).

Ethical approval: This study was approved by the Ethics Committees of Beijing Tsinghua Changgung Hospital and Chinese PLA General Hospital.

Competing interest: No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

 

 

BACKGROUND: Decision making and surgical planning are to achieve the precise balance of maximal removal of target lesion, maximal sparing of functional liver remnant volume, and minimal surgical invasiveness and therefore, crucial in liver surgery. The aim of this prospective study was to validate the accuracy and predictability of 3D interactive quantitative surgical planning approach (IQSP), and to evaluate the impact of IQSP on traditional surgical plans based on 2D images.

 

METHODS: A total of 305 consecutive patients undergoing hepatectomy were included in this study. Surgical plans were created by traditional 2D approach using picture archiving and communication system (PACS) and 3D approach using IQSP respectively by two groups of physicians who did not know the surgical plans of the other group. The two surgical plans were submitted to the chief surgeon for selection before operation. The specimens were weighed. The two surgical plans were compared and analyzed retrospectively based on the operation results.

 

RESULTS: The two surgical plans were successfully developed in all 305 patients and all the 3D IQSP surgical plans were selected as the final decision. Total 278 patients successfully underwent surgery, including 147 uncomplex hepatectomy and 131 complex hepatectomy. Twenty-seven patients were withdrawn from hepatectomy. In the uncomplex group, the two surgical plans were the same in all 147 patients and no statistically significant difference was found among 2D calculated resection volume (2D-RV), 3D IQSP calculated resection volume (IQSP-RV) and the specimen volume. In the complex group, the two surgical plans were different in 49 patients (49/131, 37.4%). According to the significance of differences, the 49 different patients were classified into three grades. No statistically significant difference was found between IQSP-RV and specimen volume. The coincidence rate of territory analysis of IQSP with operation was 92.1% (93/101) for 101 patients of anatomic hepatectomy.

 

CONCLUSIONS: The accuracy and predictability of 3D IQSP were validated. Compared with traditional surgical planning, 3D IQSP can provide more quantitative information of anatomic structure. With the assistance of 3D IQSP, traditional surgical plans were modified to be more radical and safe.

 

(Hepatobiliary Pancreat Dis Int 2017;16:271-278)

 

KEY WORDS: precision; quantitative; surgical planning; reconstruction; hepatectomy

To achieve the multi-objective optimization of maximizing the removal of the target lesion, maximizing the functional liver remnant and minimizing the surgical invasiveness (3M) of liver surgery, the surgical plan should be made with high certainty of precision.^[1] However, traditional surgical planning, mostly based on 2D images, may not provide enough information surgeons required, especially for some complex hepatectomies. On the contrary, 3D reconstruction images may provide more real and intuitive information.^[2-6] Between April 2010 and July 2016, we conducted this prospective study with the largest number of patients published to date to validate the accuracy and predictability of a 3D interactive quantitative surgical planning approach (IQSP), and evaluate the impact of IQSP on traditional surgical plans based on 2D images.

 

 

The study was conducted according to the principles expressed in the Declaration of Helsinki. The study protocol was approved by the Institutional Ethics Committees of Beijing Tsinghua Changgung Hospital and Chinese PLA General Hospital. Traditional 2D surgical plans and 3D surgical plans were made using picture archiving and communication system (PACS) and IQSP respectively by two groups of surgeons. Each group had two HPB residents of R4 and the participants were in similar expertise including CT/MRI review and operation number performed, who did not know the surgical plan of the other group. Two surgical plans were submitted to the chief surgeon for selection and decision. The accuracy of IQSP were assessed with respect to volume calculation and territory analysis for hepatic vessels, and the predictability with respect to virtual resection including the functional liver remnant volume (FLRV), transaction plane, and the integrity of inflow and outflow vasculature of remnant liver. Based on operation results, the two surgical plans were compared and graded according to the significance of differences, validating the impact of IQSP on traditional surgical plans.

 

Between April 2010 and July 2016, 305 consecutive patients with malignant tumors and symptomatic benign tumors (163 males, 142 females; mean age 56±7 years) who were planned to undergo hepatectomy in our institution were included. Patients undergoing operations for hilar cholangiocarcinoma and graft harvesting for living donor liver transplantation were excluded.

 

Multislice CT scans were acquired using SOMATON Sensation64 (Siemens Medical Solutions, Erlangen, Germany) with the following protocol: 120 kV, 200-220 mA, collimation of 64×0.6 mm, pitch of 0.625, reconstruction interval of 1.172 mm, slice thickness of 1.5 mm after reconstruction. Arterial and venous phase scans were acquired 20-40 seconds and 70-100 seconds, respectively, after administration of contrast agent.

 

Enhanced magnetic resonance imaging (MRI) scans were acquired using Signa Excite 1.5T (GE) with the following protocol: 1.25-mm thickness, no skip, repeat time of 11 milliseconds, echo time of 2 milliseconds, and flip angle of 15°. Arterial and venous phase scans were acquired 20-30 seconds and 60-70 seconds, respectively, after administration of contrast agent.

 

Surgical plans were made to determine the following aspects: (i) the target lesion and region of obligatory hepatectomy; (ii) the essential functional liver volume and the obligatory extent of liver preservation; (iii) the FLRV; (iv) the optimal procedure for hepatectomy and the optimal transection plane; (v) the vasculature to be resected and reconstructed.

 

Laparoscopic approach was preferred for uncomplex hepatectomy. As for most complex hepatectomies, laparotomy was preferred due to the complexity of manipulation.

 

Anatomic hepatectomy is preferred in cases with lesions distributed segmentally, as well as cases where removal of the involved pedicle is required. Non-anatomic resection is more appropriate for benign tumors, carcinoma in situ, peripheral lesions that do not involve main vessels, or patients with marginal functional liver reserve which requires more FLRV.

 

The resectability of lesions was determined by comprehensive review of: (1) the ratio of FLRV and standard liver volume remnant liver (FLRV/SLV), SLV^[7]=706.2×BSA (body surface area, m²)+2.4, BSA^{[8, 9]}=W (weight, kg)^0.425×H (height, cm)^0.725×0.007184; (2) biological behavior of the lesions; (3) the integrity of structure and function of remnant liver.

 

Traditional surgical plans were made mainly through PACS. A line indicating the resection plane was drawn according to the anatomic landmarks (hepatic veins, portal veins, and the gallbladder). The resected volume (2D-RV) and liver remnant volume (2D-LRV) were calculated through sum of areas extracted in sequential layers manually.

 

With assistance from an IQSP system (IQQA-Liver, EDDA Technology, USA), quantitative 3D surgical plans were created interactively in real-time in the following fashion: Step 1, analyze the liver by quantification of the liver parenchyma and individual anatomy with 3D reconstruction and rendering from original CT/MRI DICOM images, including the liver contour, hepatic lesions, portal vein, hepatic vein, hepatic artery, inferior vena cava; Step 2, define the 3D transection plane interactively, based on territory analysis for intrahepatic vessels and virtual resection. The FLRV could be calculated automatically; Step 3, compare and adjust the transection plane in Step 2 in real-time, for the balance between the maximal lesion removal and the maximal FLRV. Thus, the optimal procedure, the optimal transection plane, the vasculature to be resected and reconstructed, the resected volume (IQSP-RV), and the FLRV may be determined.

 

Two surgical plans were submitted to the chief surgeon for decision, in pursuit of the multi-objective optimization of maximizing the removal of the target lesion, maximizing the functional liver remnant and minimizing the surgical invasiveness.

 

The surgical procedure was carried out strictly according to the final decision. For anatomic hepatectomy, the resection boundary was identified through pedicle occlusion, vascular staining, intraoperative ultrasound, and/or anatomic landmarks. The resection boundary was compared with the results of preoperative territory analysis. For non-anatomic hepatectomy, the resection boundary was identified mainly by intraoperative ultrasound.

 

For malignant diseases, non-tumor principles and techniques were carefully carried out to prevent iatrogenic tumor-cell spread, including “no-touch” technique, en-bloc resection and wide tumor-free margin. For malignancies originating from cholangiocytes including cholangiocarcinoma and gallbladder carcinoma, regional lymph node dissection was routinely performed.

 

The specimens were weighed and the data (assuming 1 g=1 mL gravimetrically) was compared with preoperative calculated resection volume.

 

Based on operation results, the two surgical plans were compared and analyzed. The differences were graded according to the significance of modification on traditional surgical plans, including the resectability, operative procedure, and the resection extent.

 

The nomenclature of hepatectomy is defined according to the Brisbane 2000 terminology.^[10] Uncomplex hepatectomy includes left lateral sectionectomy and periphery local tumorectomy. Complex hepatectomy includes subsegmentectomy/segmentectomy, right posterior sectionectomy, left/right hemihepatectomy, extended left/right heptatectomy, left/right trisectionectomy, mesohepatectomy, polysegmentectomy, caudate lobectomy, large irregular hepatectomy and perihilar tumorectomy.

 

Quantitative variables were compared using Student’s t test, the Mann-Whitney U test or one-way ANOVA where appropriate. Data of quantitative variables were expressed as the mean±standard deviation (SD). All data were analyzed using the Statistical Package for the Social Science, version 16 (SPSS, Chicago, IL, USA). A value of P<0.05 was considered to be statistically significant.

 

 

Surgical plans were made in 272 patients, including 139 of anatomic hepatectomy and 133 of non-anatomic hepatectomy. Thirty-three patients were judged to be unresectable due to extensively distributed lesions or poor functional liver reserve. For 10 patients of hepatocellular carcinoma, the tumor-bearing portal veins and corresponding territory areas could not be identified in 2D images. Rather than segmentectomy or subsegmentectomy, surgical plans of 7 local tumorectomies and 3 major hepatectomies had to be made based on 2D images. In 3 patients of extended left hepatectomy, essential information on FLRV and dysfunctional liver volume caused by hepatic vein resection could not be acquired.

 

3D reconstruction, territory analysis and virtual resection were successfully performed in all 305 patients. The average process time for creation and analyzation of 3D models was 38±15 minutes in complex group and 27±7 minutes in uncomplex group. Surgical plans were made in 287 patients, including 146 of anatomic hepatctomy and 141 of non-anatomic hepatectomy. Eighteen patients were judged to be unresectable due to insufficient FLRV. IQSP could show the real spatial anatomic relationship between target lesions and its surrounding vasculature in an intuitive and interactive way. IQSP could also identify the tumor-bearing portal veins and corresponding territory areas through territory analysis, and automatically calculate the distance of surgical margins, the dysfunctional liver volume and FLRV through virtual resection.

 

All the IQSP surgical plans were selected by the chief surgeon as the final decision. A total of 278 patients (278/287, 96.9%) successfully underwent surgery according to the virtual resection of IQSP surgical plans by the same surgical group, including 147 uncomplex hepatectomies and 131 complex hepatectomies. Nine patients were withdrawn from hepatectomy for the following reasons: 8 of extensive intrahepatic or distant metastasis, and 1 of intraoperative cardiac dysfunction.

 

The uncomplex group included 40 patients of laparoscopically stapled left lateral sectionectomy^[11] and 107 of local tumorectomy. The complex group included 101 patients of anatomic hepatectomy and 30 of non-anatomic hepatectomy (Table 1).

 

For 101 patients of anatomic hepatectomy in the complex group, the coincidence rate of the resection boundary with results of preoperative territory analysis and virtual resection was 92.1% (93/101) in contour. Mismatch occurred in 8 patients, including 5 of caudate lobectomy and 3 of cholelithiasis with atrophy/hypertrophy complex. No statistically significant difference was found between IQSP-RV calculated through territory analysis and the specimen volume (P>0.05)(Table 2).

 

For the uncomplex group, the 2D and 3D surgical plans were the same for all patients (147/147, 100%). No statistically significant difference was found between 2D-RV, IQSP-RV and the specimen volume, nor between 2D-LRV and FLRV (Table 2).

 

In the complex group, 3D IQSP surgical plans were different from traditional surgical plans in 49 patients (49/131, 37.4%). According to the significance of differences, the 49 different patients were classified into three grades: I. Lesions are determined to be unresectable in 2D surgical plan but resectable in 3D IQSP surgical plan (15 patients); II. 3D IQSP plan modifies the operation procedure of 2D surgical plan from non-anatomic hepatectomy to anatomic hepatectomy (7 patients); III. 3D IQSP plan modifies the resection extent of 2D plan (27 patients), including IIIa. extended resection (16); IIIb. reduced resection (8); IIIc. combined with vascular reconstruction (3) (Table 3). No statistically significant difference was found between IQSP-RV and the specimen volume (P>0.05) (Table 2).

 

There was no 30-day postoperative death in the 278 patients who underwent operation. The overall perioperative morbidity was 16.2% (45/278). The surgical and postoperative data of two groups are shown in Table 4. The clinical and pathological characteristics of the two groups are listed in Table 5. The surgical margins in 205 patients of malignant disease are negative (95.3%, 205/215).

 

 

Precision is essential in modern surgery, characterized by multi-objective optimization accommodating therapeutic effectiveness, surgical safety, and minimal invasiveness (3O).^{[1, 12, 13]} To achieve “precision” in liver surgery, surgical planning with high certainty is a prerequisite. However, traditional imaging tests, due to the inherited characteristic of 2D visual field, may not provide surgeons with enough operative information. This may bring about blindness during operation, and may even be misleading. In this study, traditional surgical plans based on 2D images and 3D surgical plans based on an IQSP system were made for each patient and compared with operation results. The accuracy and predictability of 3D IQSP were validated by the results. Compared with traditional surgical planning, a list of features provided by IQSP, including territory analysis, volume calculation, and virtual resection, enabled surgeons to make more precise surgical plans.

 

The comparison of volume calculations shown in Table 2 and the results of territory analysis for anatomic hepatectomy demonstrated the accuracy of IQSP. The consistence of operation results with the virtual resection of IQSP surgical plans reflected the predictability. The resection boundary did not match the preoperative territory analysis in 8 patients, including 5 of caudate lobectomy in which the vessels of caudate lobes were not visible in CT/MR images, and 3 of cholelithiasis in which the parenchyma was severely deformed with atrophy/hypertrophy complex. The perfusion or drainage area of hepatic vessels was determined by a simulation algorithm based on the diameter and distance of vascular structures,^[14] which explains the mismatching in the 8 patients.

 

The IQSP surgical plans were selected as the final decisions because they not only provided more quantitative information, but also improved the safety and radical effect of traditional 2D surgical plans. There were three main reasons accounting for the modifications of traditional 2D surgical plans.

 

Firstly, territory analysis, the most important function of IQSP, can identify the tumor-bearing portal veins and the perfusion areas, which are crucial for anatomic segmentectomy/subsegmentectomy. For hepatocellular carcinoma, the complete removal of the territory area perfused by the tumor-bearing portal branches contributes to the radical effect, because of the possible intrahepatic dissemination through the portal venous system. However, the tumor-bearing portal veins, which may be more than one branch, could not be easily identified on 2D images. In this study, anatomic hepatectomy was preferred for hepatocellular carcinoma from 2 to 5 cm,^[15-17] and 12 patients of traditional surgical plans were modified to anatomic segmentectomy (Fig. 1) for this reason, including 7 of grade II, 2 of grade IIIa, and 3 of grade IIIb.

 

Secondly, the function of calculation embedded in IQSP can be combined with territory analysis and virtual resection to automatically provide surgeons with a series of quantitative information, including resection volume, FLRV, dysfunctional volume and distance of surgical margin. Obvious ischemia or congestion would result in postoperative biliary fistulas, liver necrosis, or even liver failure.^[18-20] Therefore, the integrity of the inflow and outflow of liver remnant should be guaranteed as much as possible to avoid dysfunctional parenchyma.

 

In the study, 11 patients of traditional 2D surgical plans were modified by IQSP due to insufficient FLRV, including 8 of reduced resection (IIIb) and 3 of combined hepatic vein reconstruction (IIIc). The decision to perform vascular reconstruction mainly depended on the Chinese consensus of safety limit for liver resection, which considers the presence of cirrhosis, Child-Pugh functional classification and ICGR15. The most meaningful modification was that 15 patients which were previously determined to be unresectable by traditional 2D surgical planning due to extensively distributed lesions, successfully underwent complex hepatectomy after IQSP surgical planning (I), including 8 patients of large irregular hepatectomy, 4 segmentectomies of S2-S8+Roux-Y anastomosis (caudate lobe-sparing subtotal hepatectomies) (Fig. 2),^[21] 1 segmentectomy of S4, S7, S8+right hepatic vein reconstruction, 1 segmentectomy of S1, S2, S3, S6, S7+Roux-Y anastomosis, and 1 segmentectomy of S1, S4a, S7, S8 (Table 3). This is attributed to the virtual resection and automatic calculation of IQSP. With the assistance of IQSP, the resection rate of complex hepatic lesions was increased by 11.5% (15/131).

 

For malignant tumors, a safe surgical margin should be secured^[22-25] to maximally guarantee the radical effect and reduce the recurrences. In the study, 1 cm was identified as the safety margin distance for malignant tumors and 16 patients of traditional surgical plans were modified by IQSP to extended resections (IIIa) due to insufficient surgical margins.

 

Lastly, the intuitive observation of IQSP helps surgeons evaluate the spatial variation of intrahepatic vasculature more easily. Variation of intrahepatic vasculature is ever-present and may at times change the surgical planning to a great extent,^{[4, 26]} even for anatomic hepatectomy following natural anatomical landmarks. Identification of vascular variations in advance may prompt surgeons to take measurements to avoid intraoperative injury. In this study, 4 patients of traditional surgical plans were modified due to atypical vascular variations which were omitted in 2D images. Two patients were extended from left/right hemihepatectomy to left/right trisectionectomy (IIIa) due to the existence of a left medial portal vein originating from the right umbilical portion and a right anterior portal vein originating from the left umbilical portion, respectively. The other two patients were reduced from left trisectionectomy and mesohepatectomy to left hemihepatectomy and dorsal right anterior section-sparing mesohepatectomy respectively (IIIb), because S7 was perfused by dP8 while the right posterior portal vein only perfused S6.

 

Although many advantages of 3D IQSP has been described above, there are still some points to be discussed. First, we need to acknowledge that the main weakness of the paper is the study design. All patients underwent surgery according to 3D IQSP planning based on the decision of the chief surgeon. As such, it is difficult to evaluate the appropriateness of traditional 2D planning and we actually did not compare the outcomes of the two plans. Second, the validation of the accuracy of “territory analysis” is not so rigorous. Volume measurement of the resection specimen is easy, but the comparison of the resection boundary is a surgeon-dependent subjective observation. Furthermore, the hemodynamic role of collateral vessels is still unclear, which may theoretically influence the accuracy of territory analysis. In addition, the selection of 3D surgical plans as the final decision by the chief surgeon would lead to bias. However, 3D surgical plans were based on traditional 2D images and could provide more quantitative information, which helps surgeons make precise surgical plans towards the multi-objective optimization of maximizing the removal of the target lesion, maximizing the functional liver remnant and minimizing the surgical invasiveness. The success of operations could also reflect the superiority of 3D surgical plans. At last, the effectiveness of 3D planning is largely influenced by the image quality of 2D images. Strict phase time of CT/MR images and slice thickness ≤2 mm^[27] are the premise of satisfactory 3D visualization.

 

In this study, the accuracy and predictability of IQSP were validated. Compared to traditional 2D surgical planning, 3D IQSP can provide more quantitative information with high predictability, and modify traditional surgical plans for complex hepatectomy to be more radical and safe. With the assistance of 3D IQSP, traditional surgical plans were modified to be more radical and safe. Therefore, we recommend 3D IQSP as a routine surgical planning approach for complex hepatectomy.

 

 

1 Dong J, Yang S, Zeng J, Cai S, Ji W, Duan W, et al. Precision in liver surgery. Semin Liver Dis 2013;33:189-203. PMID: 23943100

2 Radtke A, Sotiropoulos GC, Molmenti EP, Schroeder T, Peitgen HO, Frilling A, et al. Computer-assisted surgery planning for complex liver resections: when is it helpful? A single-center experience over an 8-year period. Ann Surg 2010;252:876-883. PMID: 21037445

3 Kumamoto K, Mizuno S, Kuriyama N, Ohsawa I, Kishiwada M, Hamada T, et al. Postoperative liver dysfunction in living donors after left-sided graft hepatectomy: portal venous occlusion of the medial segment after lateral segmentectomy and hepatic venous congestion after left lobe hepatectomy. Transplant Proc 2012;44:332-337. PMID: 22410009

4 Warmann SW, Schenk A, Schaefer JF, Ebinger M, Blumenstock G, Tsiflikas I, et al. Computer-assisted surgery planning in children with complex liver tumors identifies variability of the classical Couinaud classification. J Pediatr Surg 2016;51:1801-1806. PMID: 27289416

5 Zygomalas A, Karavias D, Koutsouris D, Maroulis I, Karavias DD, Giokas K, et al. Computer-assisted liver tumor surgery using a novel semiautomatic and a hybrid semiautomatic segmentation algorithm. Med Biol Eng Comput 2016;54:711-721. PMID: 26307199

6 Radtke A, Sgourakis G, Molmenti EP, Beckebaum S, Cicinnati V, Broelsch CE, et al. Computer-assisted surgical planning in adult-to-adult live donor liver transplantation: how much does it help? A single center experience. Transplantation 2012;94:1138-1144. PMID: 23222737

7 Schroeder T, Nadalin S, Stattaus J, Debatin JF, Malagó M, Ruehm SG. Potential living liver donors: evaluation with an all-in-one protocol with multi-detector row CT. Radiology 2002;224:586-591. PMID: 12147860

8 Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. 1916. Nutrition 1989;5:303-313. PMID: 2520314

9 Urata K, Kawasaki S, Matsunami H, Hashikura Y, Ikegami T, Ishizone S, et al. Calculation of child and adult standard liver volume for liver transplantation. Hepatology 1995;21:1317-1321. PMID: 7737637

10 Strasberg SM. Nomenclature of hepatic anatomy and resections: a review of the Brisbane 2000 system. J Hepatobiliary Pancreat Surg 2005;12:351-355. PMID: 16258801

11 Wang X, Li J, Wang H, Luo Y, Ji W, Duan W, et al. Validation of the laparoscopically stapled approach as a standard technique for left lateral segment liver resection. World J Surg 2013;37:806-811. PMID: 23329421

12 Dong JH, Yang SZ, Duan WD, Ji WB, Cai SW, Wang J, et al. Clinical application of precise liver resection techniques in patients with complicated liver space-occupying lesions. Zhonghua Wai Ke Za Zhi 2009;47:1610-1615. PMID: 20137393

13 Su L, Zhou XJ, Dong Q, Zhang H, Shen F, Chen YJ, et al. Application value of computer assisted surgery system in precision surgeries for pediatric complex liver tumors. Int J Clin Exp Med 2015;8:18406-18412. PMID: 26770445

14 Saito S, Yamanaka J, Miura K, Nakao N, Nagao T, Sugimoto T, et al. A novel 3D hepatectomy simulation based on liver circulation: application to liver resection and transplantation. Hepatology 2005;41:1297-1304. PMID: 15846773

15 Eguchi S, Kanematsu T, Arii S, Okazaki M, Okita K, Omata M, et al. Comparison of the outcomes between an anatomical subsegmentectomy and a non-anatomical minor hepatectomy for single hepatocellular carcinomas based on a Japanese nationwide survey. Surgery 2008;143:469-475. PMID: 18374043

16 Wakai T, Shirai Y, Sakata J, Kaneko K, Cruz PV, Akazawa K, et al. Anatomic resection independently improves long-term survival in patients with T1-T2 hepatocellular carcinoma. Ann Surg Oncol 2007;14:1356-1365. PMID: 17252289

17 Chen J, Huang K, Wu J, Zhu H, Shi Y, Wang Y, et al. Survival after anatomic resection versus nonanatomic resection for hepatocellular carcinoma: a meta-analysis. Dig Dis Sci 2011;56:1626-1633. PMID: 21082347

18 Lang H, Radtke A, Hindennach M, Schroeder T, Frühauf NR, Malagó M, et al. Impact of virtual tumor resection and computer-assisted risk analysis on operation planning and intraoperative strategy in major hepatic resection. Arch Surg 2005;140:629-638. PMID: 16027326

19 Gertsch P, Vandoni RE, Pelloni A, Krpo A, Alerci M. Localized hepatic ischemia after liver resection: a prospective evaluation. Ann Surg 2007;246:958-965. PMID: 18043097

20 Fischer L, Cardenas C, Thorn M, Benner A, Grenacher L, Vetter M, et al. Limits of Couinaud’s liver segment classification: a quantitative computer-based three-dimensional analysis. J Comput Assist Tomogr 2002;26:962-967. PMID: 12488744

21 Dong J, Lau WY, Lu W, Zhang W, Wang J, Ji W. Caudate lobe-sparing subtotal hepatectomy for primary hepatolithiasis. Br J Surg 2012;99:1423-1428. PMID: 22961524

22 Shimada K, Sakamoto Y, Esaki M, Kosuge T. Role of the width of the surgical margin in a hepatectomy for small hepatocellular carcinomas eligible for percutaneous local ablative therapy. Am J Surg 2008;195:775-781. PMID: 18440487

23 Shi M, Guo RP, Lin XJ, Zhang YQ, Chen MS, Zhang CQ, et al. Partial hepatectomy with wide versus narrow resection margin for solitary hepatocellular carcinoma: a prospective randomized trial. Ann Surg 2007;245:36-43. PMID: 17197963

24 Ikai I, Arii S, Kojiro M, Ichida T, Makuuchi M, Matsuyama Y, et al. Reevaluation of prognostic factors for survival after liver resection in patients with hepatocellular carcinoma in a Japanese nationwide survey. Cancer 2004;101:796-802. PMID: 15305412

25 Dong S, Wang Z, Wu L, Qu Z. Effect of surgical margin in R0 hepatectomy on recurrence-free survival of patients with solitary hepatocellular carcinomas without macroscopic vascular invasion. Medicine (Baltimore) 2016;95:e5251. PMID: 27858885

26 Nakamura T, Tanaka K, Kiuchi T, Kasahara M, Oike F, Ueda M, et al. Anatomical variations and surgical strategies in right lobe living donor liver transplantation: lessons from 120 cases. Transplantation 2002;73:1896-1903. PMID: 12131684

27 Cash DM, Miga MI, Glasgow SC, Dawant BM, Clements LW, Cao Z, et al. Concepts and preliminary data toward the realization of image-guided liver surgery. J Gastrointest Surg 2007;11:844-859. PMID: 17458587