Supplemental Material
Robotic Splenectomy : What is the real benefit?
Dana-Elena Giza3, S. Tudor1, Raluca Roxana Purnichescu-Purtan4, C. Vasilescu1,2,
1. Department of General Surgery and Liver Transplantation, Fundeni Clinical Institute, Bucharest, Romania
2. “Carol Davila” University of Medicine and Pharmacy, Bucharest, Romania
3. Department of Hematology , Fundeni Clinical Institute, Bucharest, Romania
4. Faculty of Applied Sciences, Department of Mathematical Methods and Models, Politehnica University of Bucharest, Romania
Surgical techniques:
The first 10 cases of laparoscopic group were performed using an anterior approach, while in the remaining cases of laparoscopic and robotic groups a lateral approach with three or four-trocar technique was used.
Laparoscopic Total Splenectomy:
Under general anaesthesia and after placing an orogastric tube to decompress the stomach, the patients were placed in the right lateral decubitus at approximately 70° with an anti-Trendelenburg inclination of the table to improve exposure of the left upper quadrant. The surgeon and the first assistant stand on the right side of the table. The video monitor is placed on the left side of the table at the level of the patient’s shoulders. A three or four-trocar technique was used depending on the spleen size and splenectomy difficulty. The initial 10-mm port for the 30° laparoscope was placed on a line drawn from the umbilicus and perpendicular to anterior left costal margin. In cases of splenomegaly the first trocar was placed on this line, below the inferior margin of the spleen to avoid trocar injuries. Under visual control a 5-mm port in the epigastrium and a 12-mm port in the left flank were placed. The 12-mm port will be used for specimen retrieval. A limited diagnostic laparoscopy was performed in order to detect accessory spleens. After dividing the peritoneal attachments and the splenic ligaments using harmonic scalpel, Ligasure Atlas Sealer/Divider or monopolar electrocautery, the splenic vessels anatomy was identified (distribution pattern of splenic anatomy) and the time from the beginning of the laparoscopic procedure until full anatomy exposure was measured. Hilar vessels were divided using clips, sutures, Ligasure, Endo-GIA stapler or mixed techniques depending on specific anatomy and technology availability. After complete mobilization, the spleen was placed in an EndoCatch™ 15-mm specimen removal pouch and retrieved by kelly fragmentation. At the end of the procedure, one or two drains were placed in the splenic bed.
Laparoscopic Subtotal splenectomy:
The patient’s position and the first steps of the procedure were similar to total laparoscopic splenectomy. After dividing the peritoneal attachments and the splenic ligaments, the splenic vessels anatomy was identified (distribution pattern of splenic artery). In the first seven surgical procedures we choose to preserve the upper pole of the spleen and after dividing the hilar vessels and complete mobilization of the lower pole we preserved the last one or two short gastric vessels. At this point of the procedure a clear line of demarcation between vascularized and non-vascularized splenic tissue appeared on the splenic surface. A combination of monopolar electrocautery, blunt dissection, Harmonic scalpel or staplers was used to transect the splenic tissue, ensuring that a 5 mm rim of non-vascularized splenic tissue remained in situ. In cases with lower pole preservation (n=15) we used the same technique as in open approach. After dividing the hilar vessels and the short gastric vessels using Ligasure, a 10-15% splenic remnant was preserved, supplied by the ascending branch of left gastroepiploic artery.
Depending on its size, the specimen was retrieved in an EndoCatch™ specimen removal pouch through the accessory port incision (17 cases), or via a Pfannenstiel incision (5 cases). Haemostasis on the surface of the splenic remnant was completed in most cases with TachoSil® or TachoComb® haemostatic sponge. At the end of the procedure, one or two drains were placed in the splenic bed.
Robotic Total Splenectomy:
Patients were placed in the same position as for the laparoscopic approach. The place for the optical trocar depends on the spleen size and of the patient’s age. In cases of spleens palpable below the costal margin and in cases of small children, the optical port position is closer to the umbilicus in order to obtain a good position to approach the splenic hilum and to ensure enough space between the robotic arms. A 12 mmHg pneumoperitoneum was created using a Veres needle through a left paraumbilical incision, and an optical port (5 to 12-mm trocar) was introduced afterwards. Under visual control, two robotic 8-mm trocars were placed in the left hypochondriac region and in the epigastrium. An additional 5 to 12-mm accessory port was placed in the left lumbar region on the middle axillary line for the side-assistant surgeon. The surgical cart with the robotic arms was positioned on the patient’s left side at a 45° angle to the table’s longitudinal axis. Dissection was performed using the robotic Endowrist® Fenestrated Maryland Bipolar Cautery on the left hand and Harmonic™ Curved Shears on the right hand. After dividing the peritoneal attachments and the splenic ligaments using Harmonic™ Curved Shears splenic vessels anatomy was evaluated. Hilar vessels were divided using Endo-GIA stapler or by separately controlling the different branches of the splenic hilum with suture or Weck Hem-o-lok® polymer clips. After complete mobilization, the spleen was placed in an EndoCatch™ specimen removal pouch and retrieved by morcellation. At the end of the procedure, one or two drains were placed in the splenic bed.
Robotic subtotal splenectomy:
The patient’s position and trocar placement were similar as in robotic total splenectomy. After dividing the peritoneal attachments and the splenic ligaments using Harmonic™ Curved Shears splenic artery anatomy was identified and the origin of lower pole vascular supply was traced down to its origin. Hilar vessels and short gastric vessels were split using Ligasure, robotic clips or were ligated with intracorporeal knots using EndoWrist® Needle Holder. After complete mobilization, the splenic parenchyma was transected using robotic hook electrocautery, Harmonic™ Curved Shears, Endo GIA Roticulator™ blue cartridge stapler or Ligasure introduced by the side assistant surgeon through the accessory port. Afterward the procedure was carried out in the same fashion as the laparoscopic subtotal splenectomy procedure.
The operating time in the robotic group is the overall operating time, including the docking time of the robot. However, robotic set-up is not a significant source of time loss and has a low impact on overall surgical time. Robotic set-up and docking can be conducted time-effectively after better understanding of the technique and the development of a coordinated procedure. Acceptable time can be achieved after a steep learning curve.
Statistical Analysis:
Total operative time, conversion to open procedure, intra-operative blood loss, postoperative complications and therapeutic success were analyzed in order to highlight the most relevant parameters related to the learning process in laparoscopic surgery. Each of these outcome variables was monitored using a separate CUSUM chart [20, 21]. In order to construct the cumulative sum for the charts, all data were considered in chronological order (date of surgery). For continuous variables (total operative time and intra-operative blood loss) we used the simple moving average. This average was defined as:
Wherexi is the value of the i-th observation, Ci is the i-th CUSUM, T is the target value (set as the trimmed mean value) and C0 is defined to be 0.
For binary variables (conversion to open procedure, postoperative complications and therapeutic success), we used a standard non-risk adjusted CUSUM chart [18]. Definitions of long term surgical success rate and failure rate for the learning process were drawn up in such a way that they can be consistently interpreted. With each successive attempt of the surgical procedure, an increment is added to a cumulative score: positive for success (the calculated value s) and negative for failure (-(1-s)). The sizes of both types of increment are determined by ‘acceptable’ and ‘unacceptable’ failure rates, and by the probabilities of type I and II errors. We set the unacceptable failure rate at 2×the acceptable rate in order to clearly show changes in the learning curve slope.The parameters for this CUSUM chart were set as in Table 3.
The formula for the moving average in this case is defined as:
wherexi is the value of the i-th observation, Ci is the i-th CUSUM, s is the target value (see Table 2) and C0 is defined to be 0. All binary data considered in the analysis of the learning process received code 0 for a positive event (no conversion to open procedure, no postoperative complications and achievement of therapeutic success) and code 1 for the negative one. In the CUSUM chart, a positive slope indicates improvement of the process and a negative slope indicates a worsening of the process. Cut-off points might be easily identified around the main changes of the slope.
The comparisons within each group of the immediate outcome parameters (intra-operatory and post-operatory parameters) were made using Student t-test for continuous variables and Chi-square and Fisher's exact test with continuity correction for nonparametric variables. Correlations between parameters were made using Spearman’s correlation coefficient. All differences and associations were considered statistically significant if the 2-sided p-value was below 0.05 (95% Confidence Interval). Statistical analyses were performed using SPSS statistical software (version 17, Chicago, IL, USA).