Outcomes of Infants Undergoing Robot-Assisted Laparoscopic Pyeloplasty Compared to Open Repair

Transcription

1 Outcomes of Infants Undergoing Robot-Assisted Laparoscopic Pyeloplasty Compared to Open Repair Pankaj P. Dangle, James Kearns, Blake Anderson and Mohan S. Gundeti* From the Department of Surgery, Division of Urology, University of Chicago School of Medicine and Biological Sciences, and Comer Children s Hospital, Chicago, Illinois Purpose: Robotic surgery has evolved from simple extirpative surgery to complex reconstructions even in infants. Data are lacking comparing surgical and direct costs to open approaches. We describe the feasibility, salient tips and outcomes of robot-assisted laparoscopic pyeloplasty compared to an open approach. Materials and Methods: We evaluated patients undergoing open pyeloplasty or robot-assisted laparoscopic pyeloplasty. Ten patients in each group met inclusion criteria. Results: Mean patient age was 3.31 months in the open group and 7.3 months in the robotic group (p ¼ 0.02). Postoperative outcomes including length of stay (2.2 vs 2.1 days), estimated blood loss (6.5 vs 7.6 ml), days to regular diet (1 vs 1.1) and days to Foley catheter removal (1.3 vs 1.3) were similar between the open and robotic groups. Total operating time (199 vs 242 minutes) was significantly longer in the robotic group. Postoperative improvement in hydronephrosis was identical in both groups. Direct costs, excluding amortization, robotic cost, maintenance and depreciation, were $4,410 in the open group and $4,979 in the robotic group (p ¼ 0.10). Conclusions: In our preliminary experience robotic pyeloplasty in infants is feasible and safe. The immediate outcomes are similar to those of an open approach. The robotic technique in infants currently has the benefits of improved esthetic appearance, improved pain control and similar direct costs compared to the traditional open approach. Abbreviations and Acronyms CP-U ¼ cutaneous pyeloureteral stent MAG-3 ¼ mercaptoacetyltriglycine OP ¼ open pyeloplasty POD ¼ postoperative day RAL-P ¼ robotic pyeloplasty UPJO ¼ ureteropelvic junction obstruction Accepted for publication July 24, Study received institutional review board approval. * Financial interest and/or other relationship with Intuitive Surgical. Key Words: hydronephrosis, infant, kidney pelvis, robotics, urologic surgical procedures OPEN Anderson-Hynes dismembered pyeloplasty is the gold standard for treatment of ureteropelvic junction obstruction. However, a minimally invasive approach using laparoscopic pyeloplasty has become an effective modality at select centers, although intracorporeal suturing remains a crucial limitation to its wide acceptance. 1 Robotic technology has wide application with proved benefits of minimally invasive surgery, particularly for treating adults with UPJO. Although the feasibility of robotics has been shown, 2 the limited intra-abdominal working space and standardized instrumentation with minimal modification for pediatric use have remained major limitations to its wide acceptance. Kutikov et al first reported its feasibility and application in infants. 1 With increasing robotic experience, we have offered this approach in infants for correction of UPJO. We assessed surgical /13/ /0 THE JOURNAL OF UROLOGY 2013 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH,INC. Vol. 190, , December 2013 Printed in U.S.A. j 2221

2 2222 ROBOT-ASSISTED PYELOPLASTY IN INFANTS outcomes and costs of robotic pyeloplasty in infants compared to open repair. MATERIALS AND METHODS Following institutional review board approval, we retrospectively reviewed charts and conducted database analysis. We included infants undergoing dismembered Anderson-Hynes pyeloplasty with a confirmed diagnosis of UPJO based on ultrasound and/or MAG-3 scan, and excluded patients with concomitant vesicoureteral junction obstruction. All patients underwent ultrasound and/ or MAG-3 scan during followup. All cases were performed by the senior author (MSG), with various levels of resident involvement. Data collected included demographic data, length of hospital stay, estimated blood loss, complications, degree of hydronephrosis, change in hydronephrosis, pain management and cost between the groups. Total operating time (patient in and out of the operating room) and surgical time (from incision to closure) were recorded from the electronic privacy information center database of Wisconsin. All infant pyeloplasties were performed via an open approach until February With increasing robotic experience we applied robotic technology to infants beginning in December Ten patients in each group were eligible for study inclusion. To avoid selection bias, the last 10 patients from the open group were selected. All infants from our robotic database were included. Cost analysis was performed with the data provided by the institutional financial department. Direct cost entails cost per procedure per patient to the hospital. Our analysis excluded amortized, maintenance expenses or the depreciation cost of robot through time. Due to the complexity of analysis, we favored the direct cost analysis, which included operating room charges, operating room supplies, anesthesia, pharmacy, hospital stay and miscellaneous costs. All OP cases were performed with standard open 11th rib extrapleural, extraperitoneal, muscle splitting incision. RAL-P was standardized using 3 ports, ie a 12 mm umbilical camera port, an 8 mm left midline epigastric port and an 8 mm port at the ipsilateral (right) midclavicular line, as well as a 5 mm assistant suprapubic port. All ports are placed under vision by port-in-port technique, our recent modification of port placement in the midline along the camera line, and an assistant 5 mm port on the opposite side in the midclavicular line infraumbilically. Intraoperative orogastric tube and Foley catheter are placed for decompression. We avoid bowel preparation and cystoscopy with retrograde pyelogram, and the dilated colon in infants managed by placing a flatus tube. For OP epidural analgesia was administered for pain management. Nephroureteral stents were placed in patients undergoing OP, whereas an indwelling Double-JÒ stent (antegrade approach) or cutaneous pyeloureteral stent was inserted for those undergoing RAL-P. The CP-U stent was used after the initial 6 cases of antegrade Double-J stent. The Salle intraoperative pyeloplasty stent (CookÒ Medical) is prepared with minimal modifications (fig. 1). The distal end (bladder coil) is cut to 4 to 5 cm of ureteral arm, while the pelvic coil and distal part are left intact. The needle end is cut, and the Figure 1. CP-U stent used in robot-assisted pyeloplasty

3 ROBOT-ASSISTED PYELOPLASTY IN INFANTS 2223 guidewire is passed through the stent using the Seldinger technique. On POD 1 the CP-U or nephroureteral stent was capped, the Foley catheter was removed and patients were discharged home if they successfully tolerated a regular feeding regimen and had favorable pain control. Patients with CP-U and nephrocutaneous stents returned on POD 7 to 10 for stent removal, while those with an indwelling stent underwent cystoscopic stent removal 4 to 6 weeks postoperatively. Patients were followed with ultrasound at 1, 3 to 6, 12 and 24 months, and a MAG-3 scan at 6 to 12 months. Analysis was performed using Microsoft Excel. Fisher t-test was used to test the hypothesis of the categorical variables, and a p value of 0.05 was considered statistically significant. RESULTS All robotic cases were completed successfully without any open or laparoscopic conversion. Preoperative variables included prenatal diagnosis, gender, and hydronephrosis grade and laterality (table 1). Mean patient age was 3.31 months (range 1 to 10) in the open group and 7.3 months (2 to 12) in the RAL-P group (p ¼ 0.02). Preoperative hydronephrosis grade was identical in both groups. Mean body weight was 7.9 kg (range 6.1 to 12.8) in the open group and 8.01 kg (5.1 to 13.1) in the robotic group (p ¼ 0.14). Total operating room time was 199 minutes (range 131 to 242) in the open group and 243 minutes (205 to 289) in the robotic group, and surgical time was minutes (90 to 157) in the open group and 160 minutes (138 to 216) in the robotic group. We initiated a separate log to record docking, undocking and actual procedure times. Based on our observations, our average docking and undocking time with port placement is around 20 minutes. Postoperative outcomes included length of stay, estimated blood loss, days to regular diet and days to Foley catheter removal, which were identical in both groups (table 2). Three patients (1 in the robotic group and 2 in the open group) had previously undergone open pyeloplasty on the contralateral Table 1. Open Robotic p Value No. prenatal diagnosis/total No. (%) 9/10 (90) 9/10 (90) 0.90 Mean mos age at surgery (range) 3.31 (1e10) 7.3 (2e12) 0.02 Mean kg wt (range) 7.9 (6.1e12.8) 8.01 (5.1e13.1) 0.14 No. males/total No. (%) 8/10 (80) 9/10 (90) No. females/total No. (%) 2/10 (20) 1/10 (10) Mean hydronephrosis grade No. prior surgery/total No. (%)* 2/10 (20) 1/10 (10) 0.56 No. laterality/total No. (%): 0.66 Rt 3/10 (30) 4/10 (40) Lt 7/10 (70) 6/10 (60) * Prior surgery consisted of contralateral open pyeloplasty. Table 2. Open Robotic p Value Mean ml estimated blood loss (range) 6.5 (5e15) 7.6 (5e15) 0.63 Mean mins surgical (90e157) 160 (138e216) 0.01 time (range) Mean mins TOR (range) 199 (131e242) 243 (205e289) 0.01 Mean days to general 1 (0e1) 1.1 (0e1.5) 0.07 diet (range) Mean days Foley catheter 1.3 (1e2) 1.3 (1e2) 1.00 indwelling (range) Mean days length of 2.2 (1e3) 2.1 (1e3) 0.79 stay (range) Mean direct costs $4,409.6 $4, No. immediate 0/10 (0) 1/10 (10) 0.15 complications/ total No. (%) Mean final hydronephrosis 2.6 (2e3) 2.8 (2e3) 0.63 grade (range) Mean change in hydronephrosis grade Mean mos followup (range) 18.1 (5.7e23.8) 9.01 (4e24) 0.14 side for bilateral UPJO. After the initial 5 cases the average decrease in operative time was 20.1 minutes (fig. 2). There were no intraoperative complications in either group. A trocar site omental hernia developed postoperatively in 1 patient in the RAL-P group (p ¼ 0.15). This complication was recognized on POD 1, and the patient underwent immediate repair. Since then, all ports, including the assistant port, have been closed under vision to avoid injury. Mean preoperative hydronephrosis grade was 3.5 in both groups. Postoperative improvement in grade was 1.4 in the OP group and 1.3 in the RAL-P group. Perioperative narcotic analgesic use in the form of morphine (0.32 mg vs 0.22 mg) was identical in both groups. Six patients in the OP group underwent epidural catheter and analgesia for pain control with a combination of 0.1% bupivacaine and fentanyl 2 mcg/ml at a rate of 1.2 ml per hour for a period of 1.2 days. Patients were discharged home with nonnarcotic oral pain medication. Since we have started the robotic approach, epidural analgesia has not been used. Postoperative renal function was assessed via MAG-3 scan. Seven patients (70%) in each group underwent MAG-3 scan and all exhibited stable or improved function with improved drainage based on the drainage curves. Direct costs, excluding professional fees but including room and board, anesthesia, pharmacy, supplies and miscellaneous costs, were $4,410 for the open group and $4,979 for the robotic group. The $500 difference was not statistically significant (p ¼ 0.10). For patients with a Double-J stent the direct costs of cystoscopy stent removal and anesthesia were $1,001 per patient.

4 2224 ROBOT-ASSISTED PYELOPLASTY IN INFANTS Figure 2. Surgical time (from incision to closure) DISCUSSION Robotic surgery in the pediatric population has been well accepted since its initial demonstration of feasibility, with recent application to more complex procedures.3e7 Cost-effectiveness and surgical outcomes have been evaluated with an emphasis on the benefits of minimally invasive surgery in favor of the robotic approach.3,8 Infants pose a challenge regarding space constraints, due to associated limitation of movements and difficulties with trocar sites and placement.9 Use of robotic technology in infants is well described in pediatric surgery.10 To date, there is a single reported series of infant robotic pyeloplasty using the open Hassan technique, with a mean patient age of 5.6 months (range 3 to 8) and a mean total operative time of minutes.3 The authors report improvement or resolution of hydronephrosis in 7 patients, and conclude that robotic pyeloplasty is safe and effective for treating infants. We prefer to use 8 mm rather than 5 mm ports, as suggested by previous authors, due to differences in the mechanics of the functional joint (fig. 3). The design of the 5 mm instruments with the specific gooseneck type joint requires greater clearance from the tissue for activation of the angle, making the actual functional space more limited. Although the functional length of the intra-abdominal trocar for the 5 and Figure 3. Comparison of 8 mm and 5 mm instruments

5 ROBOT-ASSISTED PYELOPLASTY IN INFANTS mm instruments is identical from the tip to the remote center (3.3 cm), the additional 1.9 cm functional length in the 5 mm trocar especially in infants could potentially limit freedom of movement. Instead of an anchoring stitch, we make a skin indentation with the empty trocar and incise across the diameter of the circle. Kutikov et al also describe routine use of conventional laparoscopy for manipulation of bowel. 3 We use the robot for colon mobilization without difficulty. We routinely burp the robotic and camera ports, which provides an additional working space by lifting the body wall. Our success in terms of hydronephrosis improvement is similar to previously reported longterm postoperative outcomes in pediatric RAL-P. Long-term outcome in 155 patients with a mean followup of 31.7 months revealed a primary success rate of 96%, with 85% exhibiting improved and 11% stable hydronephrosis. 8 Kutikov et al reported 78% resolution or improvement of hydronephrosis (7 of 9 patients), with 2 patients exhibiting unchanged hydronephrosis at 6 months of followup and no evidence of obstruction on followup renal scan. 3 In our experience all patients have shown an improvement in hydronephrosis grade, with the mean preoperative grade being 3.5 in both groups, and postoperative grade being 2.6 in the OP group and 2.8 in the RAL-P group. Average length of followup was 18.1 months (range 5.7 to 23.8) in the OP group and 9.01 months (4 to 24) in the robotic group. Seven patients in each group underwent followup MAG-3 scan, and all exhibited stable or improved function and improved drainage pattern. To define the learning curve, Sorensen et al compared the first 33 consecutive RAL-P cases with open cases performed by senior faculty. 11 There were no significant differences in length of stay, pain score or surgical success at a median followup of 16 months. The numbers of complications were identical and tended to be early and technical in the RAL-P group. Average operative time was 90 minutes longer (38%) for RAL-P (p <0.004). After chronological evaluation a learning curve was noted, and after 15 to 20 robotic cases overall operative times for RAL-P were within 1 SD of those for open pyeloplasty (p ¼ 0.23). In our experience after the first 5 cases the average decrease in operative time was 20.1 minutes, with 50% to 70% of console time per resident. At our center the difference in direct costs ($4,410 for open vs $4,979 for robotic surgery) is not statistically significant (p ¼ 0.10). Casella et al reported cost analysis comparing RAL-P and laparoscopic pyeloplasty based on the total costs by adding direct and indirect costs. 12 They retrospectively reviewed 23 patients in each group and found that the robotic procedures were shorter than pure laparoscopic procedures (200 vs 265 minutes, p <0.001). Cost analysis did not demonstrate a significant difference between robotic and conventional laparoscopic cases ($15,337 vs $16,067). A subgroup analysis of robotic cases where the stent was placed in an antegrade fashion revealed decreased operative time (140 vs 265 minutes, p <0.001) and total costs ($11,949 vs $16, 067, p <0.001) in the RAL-P group. All of the previously reported cost analyses focus on surgical cost, with no focus on human cost savings for parents. Behan et al retrospectively reviewed 44 patients undergoing robotic (37) or open pyeloplasty (7) from 2008 to The human capital approach was used to calculate parental productivity. Based on their analysis, the average savings were $90.01 in lost parental wages and $ in hospitalization expenses per patient, after excluding the amortized robot cost. In the open group the instrument costs, processing and depreciation through time were not included in the cost analysis. There was no difference in hospitalization length between the groups, which may be attributed to the initial small sample size and the fact that patients were monitored for 48 hours to ensure safety, while they are now routinely discharged on POD 1. In our experience RAL-P is safe and feasible in infants, with a minimal complication rate. At our institution the technology was applied to infants after a significant experience in all other complex reconstructive procedures. Our senior author had performed 28 pyeloplasties and 60 other complex robotic procedures in older children before undertaking infant pyeloplasty. Based on the scientifically proved physiological and biochemical benefits of minimally invasive surgery in children, 14 Casale recommends a role for such surgery in small infants. 15 A robotic approach has also been shown to result in less collagen deposition in the wound compared to pure laparoscopy and more collagen deposition compared to an open approach. Significant inflammatory response via neutrophil derived oxidants secondary to stretching of the muscle is a well demonstrated phenomenon. 16,17 Casale suggests that minimally invasive surgery, particularly robotics, has equivalent outcomes, benefits of negligible external scar and less manipulation of tissue. 15 With increasing experience we have noted reduced operative times, with surgical outcomes comparable to the open approach. Direct costs are also comparable between the cohorts ($4,410 for open vs $4,979 for robotic surgery, p ¼ 0.10). Only the direct costs were analyzed. Indirect costs related to amortized cost, maintenance and depreciation,

6 2226 ROBOT-ASSISTED PYELOPLASTY IN INFANTS and robot cost were excluded due to inconsistent insurance claims and local hospital policies. Drawbacks of our study include that it is a single surgeon and single institution series. Also the study is retrospective in nature. To avoid selection bias, we chose the first 10 consecutive robotic and last 10 open cases. Despite significant robotic experience in a large pediatric population, there is a learning curve, especially in infants, due to the limited intracorporeal space and robotic miniaturized instruments. Although the number of patients in each arm was small, we compared the first consecutive 10 RAL-P procedures to include the learning curve and assess its impact on safety, feasibility and surgical outcome. Finally, there are shortcomings in the financial analysis due to the complexity of various factors involved, ie institutions, geographic locations, providers and payers. CONCLUSIONS RAL-P in infants is safe and feasible, and on shortterm followup has equivalent surgical outcomes to the traditional gold standard open approach. A strong robotic experience at a high volume center with extensive training is critical before undertaking infant robotic pyeloplasty. At this time, aside from improved esthetic appearance and pain control, application of the robotic technique in infants does not exhibit added benefits over the traditional open approach. A large prospective study with detailed financial analysis is required to compare these groups adequately. ACKNOWLEDGMENTS Anthony Navarra from the financial department provided financial and strategic analysis. REFERENCES 1. Kutikov A, Resnick M and Casale P: Laparoscopic pyeloplasty in the infant younger than 6 monthsdis it technically possible? J Urol 2006; 175: Monn MF, Bahler CD, Schneider EB et al: Emerging trends in robotic pyeloplasty for the management of ureteropelvic junction obstruction in adults. J Urol 2013; 189: Kutikov A, Nguyen M, Guzzo T et al: Robot assisted pyeloplasty in the infantelessons learned. J Urol 2006; 176: Olsen LH and Jorgensen TM: Computer assisted pyeloplasty in children: the retroperitoneal approach. J Urol 2004; 171: Lee RS, Retik AB, Borer JG et al: Pediatric robot assisted laparoscopic dismembered pyeloplasty: comparison with cohort of open surgery. J Urol 2006; 175: Gundeti MS, Eng MK, Reynolds WS et al: Pediatric robotic-assisted laparoscopic augmentation ileocystoplasty and Mitrofanoff appendicovesicostomy: complete intracorporeale initial case report. Urology 2008; 72: Nguyen HT, Passerotti CC, Penna FJ et al: Robotic assisted laparoscopic Mitrofanoff appendicovesicostomy: preliminary experience in a pediatric population. J Urol 2009; 182: Minnillo BJ, Cruz JA, Sayao RH et al: Long-term experience and outcomes of robotic assisted laparoscopic pyeloplasty in children and young adults. J Urol 2011; 185: Thakre AA, Bailly Y, Sun LW et al: Is smaller workspace a limitation for robot performance in laparoscopy? J Urol 2008; 179: Meehan JJ: Robotic surgery in small children: is there room for this? J Laparoendosc Adv Surg Tech A 2009; 19: Sorensen MD, Delostrinos C, Johnson MH et al: Comparison of the learning curve and outcomes of robotic assisted pediatric pyeloplasty. J Urol, suppl., 2011; 185: Casella DP, Fox JA, Schneck FX et al: Cost analysis of pediatric robot-assisted and laparoscopic pyeloplasty. J Urol 2013; 189: Behan JW, Kim SS, Dorey F et al: Human capital gains associated with robotic assisted laparoscopic pyeloplasty in children compared to open pyeloplasty. J Urol, suppl., 2011; 186: Blinman T: Incisions do not simply sum. Surg Endosc 2010; 24: Casale P: Minimally invasive survey in infants. Pro. J Urol 2012; 188: Kirman I, Cekic V, Poltaratskaia N et al: Plasma from patients undergoing major open surgery stimulates in vitro tumor growth: lower insulinlike growth factor binding protein 3 levels may, in part, account for this change. Surgery 2002; 132: Toumi H, F guyer S and Best TM: The role of neutrophils in injury and repair following muscle stretch. J Anat 2006; 208: 459. EDITORIAL COMMENT As health care reform continues to evolve in the United States, there appears to be increasing attention toward costs, outcomes and the impact of relatively expensive procedures such as robotic surgery. This single institution experience with robot-assisted minimally invasive pyeloplasty in infants adds to the expanding literature in this area. While undoubtedly expensive, robotic surgery has led to an expansion of the minimally invasive surgical option for pediatric urological reconstructive procedures, which previously had received only modest adoption in the United States since the description of laparoscopic pediatric pyeloplasty by Peters et al in Pediatric robotic pyeloplasty has also been associated with shorter operative times and similar perioperative parameters compared to conventional laparoscopy. 2 The authors acknowledge the limitations of this study, such as the small sample sizes for the comparative study and the limited followup that is

7 ROBOT-ASSISTED PYELOPLASTY IN INFANTS 2227 common with most single institution experiences. However, their experience with robotic instruments and their use of larger sized instruments highlight the challenging environment for pediatric medical device development and the adaptations that currently are made due to the paucity of pediatric oriented equipment. The future appears to be brighter, as the Food and Drug Administration has recently been encouraging pediatric medical device development through its grant consortia program, which has resulted in a tangible improvement in the currently challenged pediatric device development pipeline. 3 Chester J. Koh Director, Robotic Surgery Program Department of Pediatrics, Division of Urology Texas Children s Hospital/Baylor College of Medicine Houston, Texas Co-Director, Center for Technology and Innovation in Pediatrics Los Angeles, California REFERENCES 1. Peters CA, Schlussel RN and Retik AB: Pediatric laparoscopic dismembered pyeloplasty. J Urol 1995; 153: Riachy E, Cost NG, Defoor WR et al: Pediatric standard and robot-assisted laparoscopic pyeloplasty: a comparative single institution study. J Urol 2013; 189: Ulrich LC, Joseph FD, Lewis DY et al: FDA s pediatric device consortia: national program fosters pediatric medical device development. Pediatrics 2013; 131: 981. REPLY BY AUTHORS It is indeed true that we are far from miniaturizing the instruments as needed to meet the requirements (including costs) for neonatal and pediatric use. With the crucial financial aspect for manufacturers and consumers, market demand will drive the continued use of this innovation. Provided there is no risk to the patient, and considering that the indication for surgery does not change, we will continue moving forward into the digital age of surgery.