Robot-assisted vs Laparoscopic vs Open Radical Cholecystectomy for Gallbladder Cancer

Background:

The gallbladder is the most common primary cancer site among the biliary tracts and its incidence is increasing. Its prognosis is still poor with a 5-year survival of almost 20%.

Cholecystectomy is curative in patients with Tis and T1a. For patients with resectable T1b and above disease, radical cholecystectomy is advocated consisting of en bloc resection of the gallbladder, wedge resection of the liver or segment 4b, extrahepatic bile duct, and the regional lymph nodes.

Traditionally, the open approach was preferred over laparoscopic approach due to the difficulty in achieving adequate lymphadenectomy, the complexity of liver resection, and the risk of gallbladder rupture leading to peritoneal metastases. However, recent meta-analysis, including only few studies, showed that laparoscopy yielded similar long-term oncological outcomes and favorable short-term outcomes in comparison with open surgery. But, laparoscopy to treat this disease is a challenging operation and still entails some limitations, especially regarding the achievement of enough nodes.

In the last years, Robotic surgery is gaining more and more interest in HPB surgery. The robotic approach has advantages such as filtration of hand tremor, seven degrees of freedom of wrist articulation, 3-dimensional (3D) stereoscopic images, and elimination of counterintuitive "fulcrum effect" of conventional laparoscopic surgery. These advantages may facilitate precise dissection for lymphadenectomy and hepatic resection and thus make surgery safe.

Indeed, only a few studies have so far reported outcomes of resectable T1b and above disease patients operated on via a minimally invasive approach, specially including robotic approach and most of them with a small population.

The aim of this study is to compare patients who have undergone Robotic, Laparoscopic and Open liver resection with lymphadenectomy for >T1b gallbladder cancers in a case-matched analysis using propensity scores. The primary endpoints are intra- and postoperative outcomes, and the secondary endpoints long-term oncologic outcomes and feasibility and adequacy of minimally invasive versus traditional open approach.

Objectives: To evaluate operative, postoperative and oncologic outcomes of Robotic, Laparoscopic and Open approach in >T1b gallbladder cancers patients.

Study design: multicenter retrospective study based on a single database of all consecutive patients with >T1b gallbladder cancers who were submitted to Robot-assisted, Laparoscopic and Open liver resection (with or without en-block cholecystectomy, depending on previous surgery) and lymphadenectomy between January 2012 and December 2022.

Study population: Patients with >T1b gallbladder cancers

Data collection: Data will be extracted by a retrospective electronic database of each hospital. Patients demographics, medical and surgical history, pre-operative and intra-operative details, post-operative course and histopathological data will be retrieved.

A specific database will be created by the coordinating centres (Universitat Pompeu and University of Pisa) with the variables of interest, and participating centers will be contacted in case of requiring additional data.

Data items recorded are showed in the Supplement Excel data Sheet. The main outcomes will include main intra operative and post operative short term results and long term post operative results.

Patients will be coded by a numeric randomization code (anonymized) and the principal investigator will be the only one with access to it. The source data will be stored digitally and will be kept by the project leader for 5 years after the inclusion of the last patient.

For each hospital an identification unique number will be assigned and each case recorded will have unique identification number. Only the center that send data will have assigned to this unique number the name and surname of the case, in any case it is going to be shared with collaborators of the hospital.

Data will be recorded on a Excel program and shared only through the official hospital e-mails between each collaborator and promotor.

The entire data of the study will be provided to the collaborators only after a specific written query to the promotor.

Authorship will be based on international guidelines, with maximum 2 authorships per participating center

Involved centres This study is approved by the CRSA (Clinical Robotic Surgery Association) committee (https://clinicalrobotics.com) and all members are invited to participate after approval of each Ethical committee board.

Sample Size The sample size calculation is based on previous retrospective results. The primary endopoint will be the mean number of resected nodes and CCI score (Comprehensive Complication Index).

The sample size is calculated to achieve 80% power (1-β) in a per-protocol analysis with the following assumptions; 5% one-sided significance level (α), expected difference of the mean number of resected nodes and CCI score in the minimally invasive group and open group leads to a total number of patients of 30 for each group (Open vs Laparoscopic vs Robotic: 90 cases).

Data analyses:

According to the type of data, subgroup analyses will be done to compare the different surgical approaches. Student's t test will be used for comparison of normally distributed continuous variables, which will be reported as mean (standard deviation) values. Non-normally distributed variables will be presented as median (inter-quartile range) values and compared using the Mann-Whitney U test. Categorical variables will be reported as counts with proportions, and analysed with the X2 or Fisher's exact test, as appropriate.

To minimize the impact of selection bias, patients who have undergone Robotic, Laparoscopic and Open approach will be matched using propensity scoring method. Propensity scores will be based on variables known from literature associated with treatment assignment and included the baseline variables BMI (continuous), sex, ASA physical status, cT, cN, underlying liver disease. Matching without replacement will be done with a 1:1 ratio, based on nearest neighbours and with a caliper width of 0.01 standard deviation. Standardized mean differences (MDs) will be calculated for the assessment of distribution of baseline covariables between the three groups. The MD will be calculated only for baseline characteristics. A MD on or between 0.1 and 0.1 will be considered the optimal balance. After matching, normally distributed continuous data will be compared using the paired samples t test. For non-normally distributed continuous data, the Wilcoxon signed rank test will be used. Categorical data will be compared using McNemar's test. Additionally, to test whether potential confounders for the primary outcome not corrected for in the propensity score matching, a multivariable binary logistic regression analysis with backward selection will be performed on the unmatched cohort with previously described risk factors associated with major morbidity. The results will be reported as odds ratios (ORs) with 95 per cent confidence intervals. To investigate the selection criteria for robotic and laparoscopic approach, both univariable and multivariable binary logistic regression analyses with backward selection will be performed for baseline characteristics in the total cohort.

Ethical comments:

The study will be conducted in accordance with the current approved protocol. The Investigators will ensure that this study is conducted in accordance with relevant regulations and with Good Clinical Practice. The Investigators will ensure that this study is conducted in accordance with the principles of the Declaration of Helsinki. The protocol will be submitted to the local Research Ethics Committees and host institution(s) for written approval. The Investigators will submit and, where necessary, obtain approval from the above parties for all substantial amendments to the original approved documents.

A multicentre research ethics application will be made by the principal investigator. In addition, each participating site must supply documentary evidence of institutional review board approval in order that for General Data Protection Regulations (GDPR) it is feasible to transfer radiological images from one centre to another.

For this study, being a retrospective study of an aggressive malignancy (mean overall survival of almost 20 months), the informed consent is generally waived.

Data handling and collection Data collection will be performed using an electronic database system in accordance with the European Statement 679/2016/UE. Data from the participating centres will centralized at the leading centre (Hospital del Mar, Barcelona), responsible of the secure storage, data quality and analysis. Local investigators will be contacted and asked to complete the records when missing data will be identified. Each enrolled case will be identified by an alphanumeric code linked to the name of the patient with the enforcement of appropriate measures such as encryption or deletion to protect the identity of the patients and sensitive data.