Robot-assisted versus open hepatectomy for liver tumors: Systematic review and meta-analysis

1. INTRODUCTION

Hepatocellular carcinoma (HCC), including primary and secondary HCC, is one of the most common malignancies worldwide and has drawn interest from clinicians and researchers.1 Surgical resection is still the most effective treatment for liver tumors (LT).2,3 Owing to the special location of the liver, some lesions are not easily exposed. Moreover, its anatomical structure is complex, with numerous and closely distributed vital structures, and liver tissue is rich in blood vessels, rendering its operation difficult.4 For LT patients who are suitable for surgery, laparoscopic surgery has several advantages, including reduced trauma, quick recovery, and low complication rate.5,6 However, operating on the complex liver by laparoscopy has limited application development in resection because of the small degree of freedom of laparoscopic articular instrumentation, hand fibrillation amplification, long learning curve, and other inherent restrictions. Therefore, new technical equipment in clinical diagnosis and treatment is urgently needed to address the disadvantages associated with laparoscopy.

The introduction of the da Vinci surgical system compensates for the deficiencies of laparoscopy and even provides several advantages, such as flexibility and fine hand movement, enlarged three-dimensional (3D) imaging, minimally invasive technology and extremely low incidence of postoperative complications.7,8 The system has been widely used in general surgery, cardiac surgery, urology, gynecology, and other fields.9–12 General surgery, particularly hepatobiliary and pancreatic surgery, is widely applicable in minimally invasive surgery because of its professional branches, extensive organs involved, multiple anatomical changes, and various difficult operations. In 2002, Giulianotti et al performed the first robotic hepatectomy.13 The number of reported robotic hepatectomy cases has gradually increased, with the total number exceeding 600 cases. However, although the robot-assisted system has considerable advantages in all aspects, it still has many disadvantages. Compared with open surgery, robotic hepatectomy has no tactile feedback. The texture of the tissue, the strength of the intraoperative grip, and the toughness of the knot cannot be determined. Moreover, the original evidence supporting robotic hepatectomy mostly consists of single-center studies with small samples and no control group.14,15 Although reports of robot-assisted laparoscopic hepatectomy have increased significantly in recent years, few studies compare robot-assisted laparoscopic hepatectomy with open hepatectomy (OH). Therefore, whether robot-assisted laparoscopic hepatectomy is more beneficial to patients with HCC remains inconclusive. No meta-analysis has evaluated the effectiveness and safety of robot-assisted hepatectomy (RAH) versus OH for LT until recently. Therefore, the present meta-analysis was conducted to systematically review the published literature and evaluate the effectiveness and safety of this technique. The results can provide a reference in clinical practice.

2. METHODS 2. 1. Literature search

This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.16 Three authors independently performed a literature search of the PubMed, Embase, Cochrane Library, and Web of Science databases for eligible studies that compared RAH with OH in the treatment of LT. The search terms included the following: robot, open, laparotomy, hepatectomy, and liver resection. To find additional studies on this topic, the reference lists of the included articles were also reviewed. No language limitations were set for the literature search of trials dated up to July 2022.

2. 2. Inclusion and exclusion criteria

The inclusion criteria were as follows: (1) patients: patients of any sex, age, race or nationality who had undergone robot-assisted laparoscopy and OH. (2) Intervention measures: The experimental group received RAH, and the control group received OH. (3) Outcomes: At least one of the following outcomes is reported: operative time, blood loss, blood transfusion, resection margins, hospital stay, postoperative complications, and recurrence rate. (4) Study design: randomized controlled trial, retrospective cohort study or comparative studies. (5) No language limitations.

The exclusion criteria were as follows: (1) studies with no major outcome indicators in the literature; (2) case reports, abstracts, conference reports or experiments; and (3) studies whose samples had undergone surgery of other organs.

2. 3. Study selection and data extraction

After the full text of the included literature was determined, three researchers independently extracted general information and relevant data. The extracted data were subsequently cross-checked. To obtain missing information, if any, the authors were contacted via e-mail. If the missing information, particularly the mean and standard deviation (SD), could not be obtained from the authors, they could be derived from the median, range, and sample size using a previously reported method.17,18

2. 4. Methodological quality assessment

The methodological quality of the included studies was assessed independently by three researchers using the Newcastle–Ottawa Scale. The scale has been widely used to assess the quality of nonrandomized studies in meta-analyses.19 Four items were selected by the study subjects, one item for intergroup comparability and three items for measurement results. Nine items were selected, each of which was assigned a score of one. The literature with a score lower than five on the basis of quality was excluded, with good quality assigned a score of five or higher. Disputes were discussed and resolved with third parties.

2. 5. Statistical analysis

The study-specific odds ratio (OR) for categorical variables, standardized mean difference (SMD) for continuous variation and their 95% confidence interval (CI) were combined to calculate the pooled value of each study by using the Cochrane Review Manager software (RevMan version 5.1) Cochran’s chi-square test and I2 were used to evaluate the heterogeneity among the effect estimates. Statistical heterogeneity among studies was defined as I2 statistic > 50%. If heterogeneity was present, the data were analyzed using a random-effects model. If heterogeneity was considered unimportant, a fixed-effects model was adopted. Sensitivity analysis was performed by removing one study at a time and then repeating the meta-analysis to assess whether the pooled estimates were significantly affected by any study. If the number of included studies exceeded 10, the potential publication bias was assessed by visually inspecting funnel plots, based on primary outcomes. The results were regarded as statistically significant at two-sided p < 0.05.

3. RESULTS 3. 1. Search results, characteristics, and quality of literature of the included studies

In accordance with the search strategy, 151 literature studies were retrieved. After the literature summaries were read, 137 substandard articles were excluded, based on the inclusion and exclusion criteria. After the full text of the remaining 14 articles were searched and the studies with no objective outcomes and lack of relevant data were excluded, eight articles were ultimately included.20–27 The flow chart of literature inclusion screening is presented in Fig. 1. A total of 1079 patients were included in the literature. The characteristics of each included study are listed in Table 1. Quality assessment outcomes of nonrandomized studies are summarized in Table 2. The included studies generally exhibited moderate to high methodological quality. Therefore, the methodological quality of the included studies was moderate to high.

Table 1 - Characteristics of included studies References Year Country Study design Age(y) Sample size Outcomes Total RAH OH Patriti et al20 2014 Italy RCS 62.6 ± 10.4/63.2 ± 11.5 88 19 69 ①②③⑤⑥ Han et al21 2016 Korea RCS 54.25 ± 11.07/54.70 ± 9.21 297 99 198 ①②③④⑤⑥ Sham et al22 2016 USA RCS NA 159 71 88 ①②⑤⑥ Croner et al23 2016 Germany RCS 64 (45–76)/62 (26–87) 63 10 53 ①④⑤⑥ Kingham et al24 2016 USA RCS 64 (40–91)/63 (29–85) 128 64 64 ①②③⑥ Morel et al25 2017 Switzerland RCS 60.37 ± 15.62/59.01 ± 17.57 32 16 16 ①④⑤⑥⑦ Wang et al26 2018 China RCS NA 240 63 177 ①②④⑤⑥⑦ Lee et al27 2020 China RCS 58.3 ± 11.3/58.7 ± 12.9 72 36 36 ①②③⑤⑥

NA = not available; OH = open hepatectomy; RAH = robot-assisted hepatectomy; RCS = retrospective cohort study.

① Operative time; ② blood loss; ③ blood transfusion; ④ resection margins; ⑤ hospital stay; ⑥ postoperative complications; ⑦recurrence rate.


Table 2 - The quality of retrospective cohort studies with Newcastle–Ottawa scale scores Studies Selection Comparability Outcome Scores 1 2 3 4 5 6 7 8 Patriti et al20 ★ ★ … ★ ★★ ★ ★ … 7 Han et al21 ★ ★ ★ ★ ★ ★ … … 6 Sham et al22 ★ ★ ★ ★ ★ ★ ★ … 7 Croner et al23 ★ ★ ★ ★ ★ ★ ★ … 7 Kingham et al24 ★ ★ ★ ★ ★★ ★ … ★ 8 Morel et al25 ★ ★ ★ ★ ★ ★ ★ ★ 8 Wang et al26 ★ ★ ★ … ★ ★ ★ … 6 Lee et al27 ★ … ★ ★ ★ ★ ★ … 6

★ indicates one point. 1. Representativeness of exposed cohort. 2. Selection of nonexposed cohort. 3. Ascertainment of exposure. 4. Outcome of interest was not present at start of study. 5. Comparability of cohorts on the basis of the design or analysis. 6. Assessment of outcomes. 7. Follow-up long enough for outcomes to occur. 8. Adequacy of follow-up.


F1Fig. 1:

Flow diagram of literature selection.

3. 2. Meta-analysis of operative time (min)

Eight studies in the included literature reported on the operative time of RAH and OH for the treatment of LT, with significant heterogeneity among the studies (p < 0.01, I2 = 89%).20–27 A meta-analysis of the random effect model showed that the operative time in the RAH group was longer than that in the OH group and a significant difference was indicated (SMD = 70.55; 95% CI = 37.58–103.53; p < 0.001; Fig. 2). The robustness of the result was confirmed by sensitivity analysis.

F2Fig. 2:

The forest plot of the operative time.

3. 3. Meta-analysis of blood loss (mL)

Six studies in the included literature reported on blood loss in RAH and OH in the treatment of LT, with significant heterogeneity among studies (p < 0.01, I2 = 99%).20–22,24,26,27 A meta-analysis of the random-effects model showed that no significant difference was found between the two groups (SMD = −225.43; 95% CI = −463.67 to 12.81; p = 0.06; Fig. 3A). Moderate heterogeneity was determined (p = 0.03; I2 = 62%) and a change in pooled estimate (SMD = −152.52; 95% CI = −266.85 to 38.18; p = 0.009; Fig. 3B) was revealed by sensitivity analysis when one trial was excluded.22

F3Fig. 3:

The forest plot of the blood loss.

3. 4. Meta-analysis of blood transfusion

Five studies in the included literature reported on the blood transfusion of RAH and OH in the treatment of LT, moderate heterogeneity was determined among the studies (p = 0.04, I2 = 59%).20,21,24,25,27 A meta-analysis of the random effect model showed no significant difference in blood transfusion between the RAH group and the OH group (OR = 1.01; 95% CI = 0.33–3.07; p = 0.99; Fig. 4). The robustness of the result was confirmed by sensitivity analysis.

F4Fig. 4:

The forest plot of the blood transfusion.

3. 5. Meta-analysis of the resection margin (mm)

Four studies in the included literature reported on resection margin in RAH and OH in the treatment of LT, with significant heterogeneity among studies (p = 0.01, I2 = 73%).20,23,25,26 A meta-analysis of the random-effects model indicated no significant difference between the two groups (SMD = −0.30; 95% CI = −0.75 to 0.16; p = 0.20; Fig. 5A). A less homogenous result (p = 0.58, I2 = 0%) was determined by sensitivity analysis when one trial was excluded.23 However, no changes in the result were observed (Fig. 5B).

F5Fig. 5:

The forest plot of the resection margin.

3. 6. Meta-analysis of hospital stay (days)

Seven studies in the included literature reported on hospital stay in RAH and OH in the treatment of LT, with significant heterogeneity among studies (p < 0.01, I2 = 88%).20–26 A meta-analysis of the random-effects model showed that hospital stay was shorter in the RAH group than in the OH group and the difference was statistically significant (SMD = −2.79; 95% CI = −4.19 to −1.40; p < 0.001; Fig. 6). The robustness of the result was confirmed by sensitivity analysis.

F6Fig. 6:

The forest plot of the hospital stay.

3. 7. Meta-analysis of postoperative complication

Eight studies in the included literature reported on the postoperative complication of RAH and OH in the treatment of LT, with no significant heterogeneity found among studies (p = 0.41, I2 = 3%).20–27 A meta-analysis of the fixed-effect model showed that the occurrence of postoperative complications in the RAH group was less than that in the OH group and the difference was statistically significant (OR = 0.67; 95% CI = 0.47 to 0.95; p = 0.02; Fig. 7). The robustness of the result was verified by sensitivity analysis.

F7Fig. 7:

The forest plot of the postoperative complication.

3. 8. Meta-analysis of recurrence rate

Two studies in the included literature compared the recurrence rate of RAH and OH in the treatment of LT, with no significant heterogeneity determined among studies (p = 0.86, I2 = 0%).25,26 A meta-analysis of the fixed-effects model showed that the recurrence rate in the RAH group was lower than that in the OH group and the difference was statistically significant (OR = 0.42; 95% CI = 0.23–0.77; p = 0.005; Fig. 8). The robustness of the result was confirmed by sensitivity analysis.

F8Fig. 8:

The forest plot of the recurrence rate.

3. 9. Publication bias

The number of included studies with each outcome was <10. Thus, the funnel plot was not used to evaluate publication bias.

4. DISCUSSION

The clinical application of robotic surgical systems has been one of the major innovations in surgery in recent decades. The robotic surgical system is an intelligent minimally invasive surgical platform represented by the da Vinci surgical robot. Named after Leonardo da Vinci who designed a human-like machine in 1495, the system also features surgery as detailed as that of da Vinci’s. The system features a high-definition, well-defined spatial view that allows the surgeon to distinguish between the front and back tissues and facilitates the complex anatomy of the hilum.28 The endowrist instrument with seven degrees of freedom allows the smooth completion of the operation of the hepatic hilum, suture, hemostasis, and anastomosis of the endotracheal cavity, rendering the operation more accurate, thus overcoming the limitation of the traditional abdominal cavity. Robotic surgical systems have been used to operate on various benign and malignant LT. A 2015 systematic review by Ocuin reported that of the 437 cases of robotic hepatectomy, 72.0% were malignant tumors and the remaining 28% were benign lesions.29 HCC and liver metastasis of colorectal cancer were the main malignant tumors, comprising 33.5% and 23.0% of the cases. Robotic hepatectomy has thus far covered all indications for laparoscopic liver surgery. Hepatectomy and complete laparoscopic hepatectomy have completion rates of 93.0% and 49.0%, respectively.30 Among the limitations to robotic liver surgery are the lack of haptic feedback and the need for an assistant with extensive laparoscopic experience. The literature regarding the comparison between the RAH and OH for LT is scarce. Moreover, all previous studies were either case reports or small studies.25,31 Therefore, the feasibility, safety and efficacy of robotic hepatectomy have not been fully determined. This meta-analysis aimed to compare the effectiveness and safety of RAH versus OH in the treatment of LT.

To our knowledge, the current meta-analysis represents the first systematic review and meta-analysis comparing RAH versus OH in the treatment of LT. In this meta-analysis, only eight studies containing 1079 patients could be included. This study may thus far represent the largest body of information available for the comparison of RAH and OH in the treatment of LT. In accordance with the Newcastle–Ottawa scale for assessing the quality of studies, articles included in this meta-analysis were graded with eight, seven, or six, reflecting high quality in the selection of patients, comparability, and exposure measurement. The current meta-analysis showed the advantages of robotic hepatectomy over OH. Robotic surgery was associated with significant reduction in blood loss. Bleeding control during open surgery is usually difficult to achieve and unexpected bleeding may obscure the vision of the operator. The favorable outcomes with robotic surgery may be attributed to the decreased manipulation of organs and a magnified view. In the present study, robotic hepatectomy was associated with less intraoperative blood loss, a finding consistent with the results of other published studies.20,22,24,26,27 The length of hospital stay in the RAH group was also significantly shorter than that in the OH group. The reason may be that high-definition three-dimensional imaging provides operators with a better surgical field of view, facilitating more detailed positioning and instrument operation. In addition, the operation requires no open abdominal procedure and is associated with less trauma, shortening the length of hospital stay.

With regard to safety, the RAH group experienced significantly fewer postoperative complications and significantly less recurrence, compared with the OH group. Possible reasons include the following: (1) The surgical field of the robotic surgery system is clear, three-dimensional and magnifiable. Moreover, intraoperative operation is more accurate, flexible and stable with less intraoperative damage to normal organs and tissues. (2) The leverage effect of laparoscopic technology is overcome and the eye-and-hand coordination of the operator is good. (3) Surgeons who can perform laparoscopic hepatectomy are generally highly skilled in laparoscopic techniques and have considerable surgical experience, which may greatly contribute to preventing damage to normal tissues and organs induced by surgery. However, this meta-analysis shows that operative time was longer in RAH than in OH, which may be attributed to the required equipment installation procedures and aseptic drape. In addition, more complex hepatectomy is performed in RAH. Tsung et al showed that operative time may be shortened in robot-assisted surgical systems for hepatectomy as surgical experience increases.31 In addition, this analysis also showed no significant difference in blood transfusion and resection margin between the RAH and OH groups. Further research needs to be conducted to confirm this result.

The studies exhibited heterogeneity in outcome, hence the adoption of the random-effects model in our analysis. Sensitivity analysis was then conducted by removing one study for each time and repeating the meta-analysis to assess whether any one study significantly affected the pooled estimates. Subsequently, the outcomes related to operative time, blood transfusion, resection margin and hospital stay were determined, and only slight changes were observed, strengthening the results of this meta-analysis. However, the heterogeneity and outcome related to blood loss completely changed when the study by Sham et al was excluded, suggesting that the aforementioned study was the source of heterogeneity.22 The heterogeneity might be attributable to design of retrospective cohort study, complex disease type, and extent of liver resection.

This systematic review has several limitations. First, most cases included in the study were both not randomized controlled trials and not blinded. Second, some important parameters, such as long-term follow-up outcomes and hospital costs, could not be evaluated because of insufficient data from the published literature. Third, publication bias was not determined because the number of included studies in the analysis was <10.

Therefore, using RAH in the treatment of LT is a safe and feasible approach. However, with clinical and methodological heterogeneity considered, multicenter randomized controlled studies with high quality, consistent intervention and large sample size need to be conducted to verify the findings of this study.

REFERENCES 1. Yang JD, Roberts LR. Epidemiology and management of hepatocellular carcinoma. Infect Dis Clin North Am 2010;24:899–919. 2. Altendorf-Hofmann A, Scheele J. A critical review of the major indicators of prognosis after resection of hepatic metastases from colorectal carcinoma. Surg Oncol Clin N Am 2003;12:165–92. 3. Benson AB, Bekaii-Saab T, Chan E, Chen YJ, Choti MA, Cooper HS, et al. Metastatic colon cancer, version 3.2013: featured updates to the NCCN Guidelines. J Natl Compr Canc Netw 2013;11:141–52. 4. Viganò L, Torzilli G, Aldrighetti L, Ferrero A, Troisi R, Figueras J, et al. Stratification of major hepatectomies according to their outcome: analysis of 2212 consecutive open resections in patients without cirrhosis. Ann Surg 2020;272:827–33. 5. Biertho L, Waage A, Gagner M. Laparoscopic hepatectomy. Ann Chir 2002;27:164–70. 6. Hong SK, Tan MY, Worakitti L, Lee JM, Cho JH, Yi NJ, et al. Pure laparoscopic versus open right hepatectomy in live liver donors: a propensity score-matched analysis. Ann Surg 2022;275:e206–e12. 7. Hao R, Özgüner O, Çavuşoğlu MC. Vision-based surgical tool pose estimation for the da vinci® robotic surgical system. Rep US 2018;2018:1298–305. 8. Di Benedetto F, Petrowsky H, Magistri P, Halazun KJ. Robotic liver resection: hurdles and beyond. Int J Surg 2020;82:155–62. 9. Niclauss N, Morel P, Jung MK, Hagen ME. A comparison of the da Vinci Xi vs. the da Vinci Si surgical system for Roux-En-Y gastric bypass. Langenbecks Arch Surg 2019;404:615–20. 10. Billah MS, Stifelman M, Munver R, Tsui J, Lovallo G, Ahmed M. Single port robotic assisted reconstructive urologic surgery-with the da Vinci SP surgical system. Transl Androl Urol 2020;9:870–8. 11. Shin HJ, Yoo HK, Lee JH, Lee SR, Jeong K, Moon HS. Robotic single-port surgery using the da Vinci SP® surgical system for benign gynecologic disease: a preliminary report. Taiwan J Obstet Gynecol 2020;59:243–7. 12. Pu F, Zhang Z, Chen Z, Cai K, Wang B, Wu Q, et al. Application of the da Vinci surgical robot system in presacral nerve sheath tumor treatment. Oncol Lett 2020;20:1125–1. 13. Giulianotti PC, Coratti A, Sbrana F, Addeo P, Bianco FM, Buchs NC, et al. Robotic liver surgery: results for 70 resections. Surgery 2011;149:29–39. 14. Vasile S, Sgarbură O, Tomulescu V, Popescu I. The robotic-assisted left lateral hepatic segmentectomy: the next step. Chirurgia (Bucur) 2008;103:401–5. 15. Choi SB, Park JS, Kim JK, Hyung WJ, Kim KS, Yoon DS, et al. Early experiences of robotic-assisted laparoscopic liver resection. Yonsei Med J 2008;49:632–8. 16. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg 2010;8:336–41. 17. Hou XW, Shi JP, Chen X. How to estimate the mean and standard deviation based on the median, range and sample size when conducting meta-analysis. Chin J Evid Based Med 2015;15:484–7. 18. Hozo SP, Djubegovic B, Hozo I. Estimating the mean and variance from the median, range and the size of a simple. BMC Med Res Methodol 2005;5:13. 19. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603–5. 20. Patriti A, Cipriani F, Ratti F, Bartoli A, Ceccarelli G, Casciola L, et al. Robot-assisted versus open liver resection in the right posterior section. JSLS 2014; 18: e2014.00040. 21. Han DH, Choi SH, Park EJ, Kang DR, Choi GH, Choi JS, et al. Surgical outcomes after laparoscopic or robotic liver resection in hepatocellular carcinoma: a propensity-score matched analysis with conventional open liver resection. Int J Med Robot 2016;12:735–42. 22. Sham JG, Richards MK, Seo YD, Pillarisetty VG, Yeung RS, Park JO. Efficacy and cost of robotic hepatectomy: is the robot cost-prohibitive? J Robot Surg 2016;10:307–13. 23. Croner RS, Perrakis A, Hohenberger W, Brunner M. Robotic liver surgery for minor hepatic resections: a comparison with laparoscopic and open standard procedures. Langenbecks Arch Surg 2016;401:707–14. 24. Kingham TP, Leung U, Kuk D, Gönen M, D’Angelica MI, Allen PJ, et al. Robotic liver resection: a case-matched comparison. World J Surg 2016;40:1422–8. 25. Morel P, Jung M, Cornateanu S, Buehler L, Majno P, Toso C, et al. Robotic versus open liver resections: a case-matched comparison. Int J Med Robot 2017;13:1. 26. Wang WH, Kuo KK, Wang

留言 (0)

沒有登入
gif