Journal of Pediatric Surgery 50 (2015) 267–271
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
A randomized trial of laparoscopic versus open Nissen fundoplication in children under two years of age Dominic Papandria a, Seth D. Goldstein b,⁎, Jose H. Salazar b, Jacob T. Cox b, Kimberly McIltrot b, F. Dylan Stewart b, Meghan Arnold c, Fizan Abdullah b, Paul Colombani b a b c
Department of Surgery, St. Vincent Indianapolis Hospital, Indianapolis USA Division of Pediatric Surgery, Johns Hopkins University School of Medicine, Baltimore USA Section of Pediatric Surgery, C.S. Mott Children's Hospital, Ann Arbor USA
a r t i c l e
i n f o
Article history: Received 22 October 2014 Accepted 2 November 2014 Key words: GERD Nissen fundoplication Laparoscopic Randomized trial Pediatric
a b s t r a c t Aims: The surgery of gastroesophageal reflux disease (GERD) is common in modern pediatric surgical practice. Any differences in perioperative and long-term clinical outcomes following laparoscopic (LN) or open Nissen (ON) fundoplication have not been comprehensively described in young children. This randomized, prospective study examines outcomes following LN versus ON in children b 2 years of age. Methods: Four surgeons at a single institution enrolled patients under 2 years of age that required surgical management of GERD, who were then randomized to LN or ON between 2005 and 2012. A universal surgical dressing was employed for blinding. Analgesia and enteral feeding pathways were standardized. The primary outcome was postoperative length of stay. Perioperative outcomes and long-term follow up were collected as secondary outcomes and used to compare groups. Results: Of 39 enrolled patients, 21 were randomized to ON and 18 to LN. Length of postoperative hospital stay, time of advancement to full enteral feeds, and analgesic requirements were not significantly different between treatment cohorts. The LN group experienced longer median operating times (173 vs 91 min, P b 0.001) and higher surgical charges ($4450 vs $2722, P = 0.002). The incidence of post-discharge complications did not differ significantly between the groups at last follow-up (median 42 months). Conclusions: This randomized trial comparing postoperative outcomes following LN vs ON did not detect statistically significant differences in short- or long-term clinical outcomes between these approaches. LN was associated with longer surgical time and higher operating room costs. The benefits, risks, and costs of laparoscopy should be carefully considered in clinical pediatric surgical practice. © 2015 Elsevier Inc. All rights reserved.
The Nissen fundoplication is a common procedure for managing pediatric gastroesophageal reflux disease (GERD) [1,2]. In adult populations, laparoscopic Nissen fundoplication (LN) has been shown to be as physiologically effective as open Nissen laparotomy (ON) with lower perioperative and long-term morbidity [3–6]. However, the risk/benefit profile of a minimally invasive approach has been less well characterized in children, and practice patterns of pediatric surgeons vary widely based on preference and familiarity. There are limited prior data regarding LN versus ON in children, particularly with respect to long-term outcomes. Furthermore, infants and toddlers, a particularly high-risk population with distinct physiologic and technical challenges, have not been rigorously studied in previous trials. This study is a randomized trial designed to compare LN to ON in children b2 years of age with respect to postoperative
⁎ Corresponding author at: Division of Pediatric Surgery, Bloomberg Children's Center, 1800 Orleans Street, Room 7310, Baltimore, MD 21287-0005 USA. Tel.: +1 410 955 2717; fax: +1 410 502 5314. E-mail address: [email protected] (S.D. Goldstein). http://dx.doi.org/10.1016/j.jpedsurg.2014.11.014 0022-3468/© 2015 Elsevier Inc. All rights reserved.
hospital length of stay (LOS), postoperative pain, time to enteral feeds advancement, and long-term GERD outcomes. 1. Methods 1.1. Study design This was a single-institution, blinded clinical trial in which patients were randomly assigned to one of two treatment groups in a 1:1 ratio to receive Nissen fundoplication via either laparoscopic or open approach. Parents and nurses were masked to the treatment assignment. The study was approved by the Johns Hopkins Medicine Institutional Review Board (Application 72440) and unmasked data for patient safety and data quality were periodically reviewed by an independent Data and Safety Monitoring Committee. Eligible participants were children under two years of age who were referred for surgical management of GERD. Patients were excluded if any concomitant abdominal surgical procedure (apart from placement or revision of a feeding gastrostomy) was to be undertaken under the same anesthetic. The study took place at The Johns Hopkins Children’s Center between December 2005 and January
268
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
2012 and 4 pediatric surgeons participated in patient enrollment and operations following randomization. Sample size was based on the LOS reported in two previous studies and calculated to require 68 patients to be able to find a mean difference of 2 days of LOS with a power of 90% at an α = 0.05 level [7,8]. 1.2. Study procedures and data collection The study protocol specified a standardized pre-operative and postoperative care regimen that was used for all enrolled patients, except for those with clear contraindications. Epidural catheters were placed preoperatively to allow uniform postoperative pain management. Surgical dressings were applied uniformly to mask postoperative nurses and parents to surgical assignment (Fig. 1). Dressings were removed on postoperative day 2 at which time the operative approach was revealed. Enteral feeds were initiated at 24 h postoperatively and advanced according to an algorithm agreed upon by all participating surgeons (Fig. 2). Operative parameters collected included total operating time (exclusive of non-surgical anesthetic time) and total operating room time (time elapsed from patient entry to exit of the operating room). Postoperative parameters included analgesic measures (total days of epidural analgesia, total days requiring narcotic therapy) and hospitalization metrics (LOS, total hospital charges, readmissions within 30 days). Post-discharge parameters included continued use of antacid medications, recurrence of reflux-related symptoms, incidence of technical complications requiring reoperation, and all-cause mortality. 1.3. Interim analysis and trial termination At 71 months into enrollment, the DSMC convened to review the trial data, and ultimately recommended that the trial be truncated following review of ongoing accrual trends versus recruitment targets established in the study design using an a priori power analysis. Final data collection ceased at 24 months after the last patient was enrolled to ensure adequate long-term follow-up. 1.4. Statistical analysis Analysis was performed using GraphPad Prism, version 6 (GraphPad Software, Inc., La Jolla, CA). An unpaired t-test was used to compare continuous variables and Fisher’s exact test to compare categorical variables between groups. Comparison with a Mann–Whitney U test was used for non-parametric data. Comparator groups were defined
and based on intent to treat, using randomization allocation rather than the procedure actually performed. 2. Results A total of 39 patients were enrolled and randomized, with 18 (46%) allocated to LN and 21 (54%) to ON (CONSORT Flowsheet Fig. 3). Mean age and weight of the overall patient cohort were 6.4 months and 6.0 kg, respectively. A history of prior surgery or major procedure was present in 14 (36%) of patients, with 5 (13%) having had a prior abdominal surgery or procedure. Comparisons of patient parameters between LN and ON groups (Table 1) showed no significant differences with respect to estimated gestational age at birth (35.7 versus 33.4 weeks), birth weight (2.5 versus 2.2 kg), patient age (6.6 versus 6.3 months) or body weight (6.3 versus 5.8 kg) at time of surgery. Past history of surgery or major procedures showed no significant difference in total number of patients with prior procedures (5 versus 9) or abdominal operations (1 versus 4) within each group. Perioperative outcomes are shown in Table 2. Postoperative LOS, duration of epidural catheter, duration of narcotic use, and time to full enteral feeds were not significantly different between the two groups. Laparoscopy significantly lengthened the time of surgery (173 versus 91 min, P b 0.001) and the overall operating room duration (209 versus 165 min, P = 0.002). Within the LN group, four (22%) children required conversion from laparoscopic to open approach. Hospital charges were highly variable. Median operating room charges were significantly higher in the LN group ($4450 vs $2722, P = 0.002), but there was no statistically significant difference in total hospital charges due to the wide variability in room and other charges ($26,445 versus $13,906, P = 0.18). The median follow-up period was 42 (interquartile range 15–59) months. All children were seen again after hospital discharge, with no loss to follow-up. There was a 20.5% readmission rate within 30 days of hospital discharge, including in 5 (29%) of the LN patients and 3 (14%) of the ON patients. Six deaths occurred (LN n = 3, ON n = 3). None occurred perioperatively and none were related to surgery or related complications. All were attributed to co-morbid disease (Table 3). There were two complications requiring major reoperation in the LN group; one for intraabdominal leakage from the gastrostomy during the index hospitalization and one for a slipped wrap and hiatal hernia one year after fundoplication. One child underwent successful endoscopic dilatation of a tight wrap due to postoperative swallowing dysfunction. In the ON group, one child presented with fascial wound dehiscence after discharge requiring urgent surgical repair and another was diagnosed with a slipped wrap but has to date elected not to have it repaired. One patient in each group required surgical closure of persistent gastrocutaneous fistula following gastrostomy removal. With respect to GERD outcomes, one child in each group was noted to have recurrent symptomatic reflux, though a very high proportion (n = 28, 71.8%) were still on antacid medication at the time of last follow-up. Comparison of postoperative events revealed no statistically significant differences between groups. 3. Discussion
Fig. 1. Standardized surgical dressing blinding family and hospital providers not present in the operating room.
The scope of this study is of utmost relevance to contemporary pediatric surgical practice. It is widely recognized that infants requiring surgical management of GERD represent a distinct population, with a unique set of technical challenges and a co-morbidity profile that includes high incidence of neurological deficits and congenital abnormalities [9–11]. However, trends toward increasing use of LN have been largely based on studies outside of this population. Here, we report the results of a randomized controlled trial evaluating the efficacy of Nissen fundoplication in children b2 years of age and compare the open and laparoscopic approach.
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
269
Fig. 2. Enteral feeds advancement algorithm. CLD = clear liquid diet; hr = hour; NGT = nasogastric or gastrostomy tube; NPO = nil per os.
Our cohort shows that these data come from a particularly complicated patient population. These were generally infants under 1 year that had been born prematurely and remained underweight. The incidences of congenital anomalies, medical comorbidities, and prior surgeries were high. In this group, we found that laparoscopy added to OR time and cost, but had no significant effect on analgesic requirements, enteral advancement, hospital length of stay, or long-term GERD outcomes. For adults with GERD, LN has overtaken ON as the most common approach and is now generally considered the gold standard [12–14]. Randomized controlled trials have associated LN with reduced morbidity, shorter hospital length of stay, less postoperative pain, and superior durable outcomes [3–5,12,15–18]. Less rigorous data exist comparing surgical approaches for fundoplication in children. Meta-analyses and systematic reviews have consistently concluded that LN results in decreased hospital stay, more rapid return to enteral feeding, and less perioperative morbidity [19–21]. However, nonrandomized studies may offer unreliable comparisons between LN and ON due to selection bias. Patient size, comorbidities, and a history of prior abdominal surgery, among other factors, influence selection of surgical technique [22]. Furthermore,
long-term follow-up data are limited and there has been a suggestion that re-operation rates are higher after LN [20,23]. Two previously published randomized trials have similarities to ours. Notably, both contradict the findings of most retrospective LN versus ON studies in that they do find significant perioperative or postoperative benefit of LN over ON. McHoney et al. randomized 39 patients b13 years to LN versus ON and concluded that there is a reduced rate of retching in LN patients 22 months after surgery compared to ON patients (6% vs. 56%) but that there is no statistical difference in terms of hospital LOS, pain, or rates of dysphagia or recurrence [24]. Knatten et al. similarly randomized 88 children b16 years and concluded that LN is associated with longer duration of operation (150 min versus 89 min) but that there are no statistically significant differences in rates of other 30-day outcomes [25]. Our findings from a predominantly infant cohort agree with these studies, further suggesting that the benefits of laparoscopy are limited to nonexistent in this patient population specifically. Interestingly, the findings of all three randomized trials to date contest the body of nonrandomized studies that indicate LN has improved perioperative and postoperative outcomes, including shortened hospital length of stay, decreased duration of pain, lower rates of postoperative complications, and earlier resumption of feeding [19–21,26]. This discrepancy merits further investigation, particularly given that there are limitations to the McHoney and Knatten studies as well as our own. Unfortunately, the trials do not share a great deal of overlap regarding methodology. For example, Knatten et al. did not mask caregivers using standard surgical dressing, did not standardize enteral feeding advancement, and reported only 30-day outcome. Meanwhile, McHoney et al. focused on the endocrine hormone response in the perioperative period and offered limited follow-up information after median 22 months. Very recently, Pacilli et al. published four year follow-up to that series confirming the lower incidence of retching in Table 1 Patient characteristics.
Age, months; mean (SD) Weight at surgery, kg; mean (SD) Estimated gestational age at birth, weeks; mean (SD) Birth weight, kg; mean (SD) Prior surgery or major procedure, n (%) Prior abdominal surgery or major procedure, n (%) Fig. 3. CONSORT randomized trial patient flowchart.
LN (n = 17)
ON (n = 22)
6.6 (5.3) 6.3 (2.6) 35.7 (5.6) 2.5 (1.2) 5 (29%) 1 (6%)
6.3 (3.6) 5.8 (2.0) 33.4 (6.5) 2.2 (1.3) 9 (41%) 4 (18%)
Abbreviations: LN: laparoscopic Nissen fundoplication, ON: open Nissen fundoplication.
270
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
Table 2 Operative and perioperative results.
Conversion to open, n (%) Operating room time, min; median (IQR) Surgery length, min; median (IQR) Duration of epidural catheter, days; median (IQR) Duration of narcotic use, days; median (IQR) Time to full enteral feeds, d; median (IQR) Postoperative hospital length of stay, days; median (IQR) Operating room charges, USD; median (IQR) Total hospital charges, USD; median (IQR) 30-day readmissions, n (%)
LN (n = 17)
ON (n = 22)
P
4 (23%) 209 (180–248) 173 (140–182) 2 (1–3) 4 (3–5) 4 (2–5) 6 (4–9) $4450 ($3271–$5916) $26,445 ($12,784–$79,327) 5 (29.4%)
N/A 165 (139–184) 91 (75–110) 2 (1–2) 3 (3–4) 3 (3–4) 4 (4–7) $2722 ($2549–$3028) $13,906 ($11,463–$32,540) 3 (13.6%)
0.002 b0.001 0.78 0.26 0.91 0.08 0.002 0.18 0.26
Abbreviations: LN: laparoscopic Nissen fundoplication, ON: open Nissen fundoplication.
their LN group and documenting quality of life improvements in both groups. This study is the first to consider hospital charges in addition to the peri- and postoperative clinical outcomes. The strengths of this trial are methodologic and include randomization, blinding by way of a standardized surgical dressing, and adherence to perioperative protocol. Median long-term outcomes of over 3 years with no loss to follow-up are also a valuable addition to existing literature on the topic. A significant limitation of this study may be the lack of power to detect differences between the groups and the slow accrual over time. Also, the conversions to open during laparoscopy tended to be near the early part of the study period; thus, technical learning curve improvements may have confounded the group comparison. In addition, our results are from a single academic institution with a referral base for complex congenital anomalies and generalizability may be confined to similar large children’s hospitals. In conclusion, the consideration for offering open or laparoscopic Nissen fundoplication for GERD in infants and young children should include perioperative outcomes, long term outcomes, and cost. These have been incompletely studied in children. Our data suggest that the perceived benefits of laparoscopy – namely less perioperative morbidity – may not be realized in children with complex disease in an academic medical center. Appendix A. BAPS Discussion Presented by: Dr. Seth Goldstein Discussant: Mr. Bruce Jaffray (Newcastle) I congratulate you on doing a randomized control trial, but I am really disappointed with your outcomes. Do you really think that there is no difference in post-operative hospital stay in someone having a Laparoscopic Nissen procedure and someone having an Open mid-line incision? In my everyday experience it is chalk and cheese; one group is away home and one group is lying in pain with a morphine infusion stuck on the ward. Does this cohort represent your learning experience for Laparoscopic Nissen surgery? Were these childhood cases that the surgeons were taking on for the first time? Your operative durations again seem to be pretty wild with two and a half hours for your mean Laparoscopic Nissen surgery — this just doesn’t tally with your contemporary experience.
Table 3 Long-term postoperative results at median 42 months.
Mortality, n (%) Reoperation, n (%) Continued symptomatic reflux, n (%) Antacid use at last follow-up, n (%)
LN (n = 17)
ON (n = 22)
P
3 (18%) 2 (12%) 1 (6%) 13 (76%)
3 (14%) 1 (4 %) 1 (4%) 15 (68%)
0.99 0.57 0.99 0.72
LN: laparoscopic Nissen fundoplication, ON: open Nissen fundoplication.
Response: Dr. Seth Goldstein These are very perceptive points. The enrolment started in 2005 which was early in the experience at our institution and I think operative times reflect that. In regard to hospital stay, what we have noticed is that these are very ill children under the age of two. They are referred to us from the neonatal intensive care unit or the paediatric intensive care unit and surgical recovery is not their barrier to discharge. They are in the hospital long after they have recovered from surgery whether that is performed either open or laparoscopic and that is something to consider because it is their cardiac or pulmonary disease that keeps them in the hospital. Discussant: Bruce Jaffray (Newcastle) I will ask you the big question, you have produced data that would seem to suggest that open fundoplication is preferable by virtue of cost. When you go back to your home hospital and you are consenting the next patient for fundoplication are you going to push them towards open surgery? Response: Dr. Seth Goldstein That’s a provocative question! I will tell you that we performed this trial because unlike some centres, not all our Faculty agree, and we do have some surgeons with equipoise on the matter. Discussant: Bruce Jaffray (Newcastle) But what do you believe? Response: Dr. Seth Goldstein I have seen the open fundoplication done in under 45 min and that is good for children who have difficulty with insufflation due to their pulmonary status. I think there is a certain patient for whom I would recommend open Nissen fundoplication in order to decrease the operative time. Discussant: Dr. Warwick Teague (Adelaide) 23% conversion rate – we should never discourage conversion – I take it that you were on an intention to treat basis to compare the laparoscopic group with the converted cases? Response: Dr. Seth Goldstein We did indeed. Discussant: Dr. Warwick Teague (Adelaide) I would have done too, but it really annoys me because those patients have had an open fundoplication. Were you even interested enough to run that analysis in the less comfortable methodological way and put the converted cases in the open group? Response: Dr.Seth Goldstein We thought long and hard about this because with low numbers in each group we thought that could significantly affect our results. I ran them casually both ways and did not find a difference. Ultimately for methodology I would have to assume some confounder just in case that the conversions were due to some unrecognised factors. Discussant: Prof. Paul Losty (Liverpool) Just a comment. There have been two randomised control trials showing no difference in outcome. The first was presented at BAPS by the group from Norway, and second was, I believe, published by the group at Great Ormond Street in the Annals of Surgery. Again there was no actual difference in the outcomes following surgery. Is it all about the scar?
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
References [1] Fonkalsrud EW, Berquist W, Vargas J, et al. Surgical treatment of the gastroesophageal reflux syndrome in infants and children. Am J Surg 1987;154:11–8. [2] Turnage RH, Oldham KT, Coran AG, et al. Late results of fundoplication for gastroesophageal reflux in infants and children. Surgery 1989;105:457–64. [3] Salminen PTP, Hiekkanen HI, Rantala APT, et al. Comparison of long-term outcome of laparoscopic and conventional Nissen fundoplication: a prospective randomized study with an 11-year follow-up. Ann Surg 2007;246:201–6. [4] Salminen P, Hurme S, Ovaska J. Fifteen-year outcome of laparoscopic and open Nissen fundoplication: a randomized clinical trial. Ann Thorac Surg 2012;93:228–33. [5] Siddiqui MRS, Abdulaal Y, Nisar A, et al. Meta-analysis of outcomes after open and laparoscopic Nissen's fundoplication in the treatment for gastro-oesophageal reflux disease. Eur Surg Acta Chir Austriaca 2012;44:138–49. [6] Niebisch S, Fleming FJ, Galey KM, et al. Perioperative risk of laparoscopic fundoplication: safer than previously reported — analysis of the American College of Surgeons National Surgical Quality Improvement Program 2005 to 2009. J Am Coll Surg 2012;215:61–8. [7] Collins III JB, Georgeson KE, Vicente Y, et al. Comparison of open and laparoscopic gastrostomy and fundoplication in 120 patients. J Pediatr Surg 1995;30:1065–71. [8] Luks FI, Logan J, Breuer CK, et al. Cost-effectiveness of laparoscopy in children. Arch Pediatr Adolesc Med 1999;153:965–8. [9] Shah SR, Jegapragasan M, Fox MD, et al. A review of laparoscopic Nissen fundoplication in children weighing less than 5 kg. J Pediatr Surg 2010;45:1165–8. [10] Thatch KA, Yoo EY, Arthur III LG, et al. A comparison of laparoscopic and open Nissen fundoplication and gastrostomy placement in the neonatal intensive care unit population. J Pediatr Surg 2010;45:346–9. [11] Iwanaka T, Uchida H, Kawashima H, et al. Complications of laparoscopic surgery in neonates and small infants. J Pediatr Surg 2004;39:1838–41. [12] Aye RW, Swanstrom LL, Kapur S, et al. A randomized multiinstitution comparison of the laparoscopic Nissen and Hill repairs. Ann Thorac Surg 2012;94:951–8. [13] Tsuboi K, Gazallo J, Yano F, et al. Good training allows excellent results for laparoscopic Nissen fundoplication even early in the surgeon's experience. Surg Endosc Other Intervent Tech 2010;24:2723–9.
271
[14] Wehrli H. Laparoscopic fundoplication for gastroesophageal reflux disease (GERD). Ther Umsch 2005;62:96–9. [15] Sietses C, Wiezer MJ, Eijsbouts QAJ, et al. A prospective randomized study of the systemic immune response after laparoscopic and conventional Nissen fundoplication. Surgery 1999;126:5–9. [16] Laine S, Rantala A, Gullichsen R, et al. Laparoscopic vs conventional Nissen fundoplication: a prospective randomized study. Surg Endosc 1997;11:441–4. [17] Nilsson G, Wenner J, Larsson S, et al. Randomized clinical trial of laparoscopic versus open fundoplication for gastro-oesophageal reflux. Br J Surg 2004;91:552–9. [18] Ackroyd R, Watson DI, Majeed AW, et al. Randomized clinical trial of laparoscopic versus open fundoplication for gastro-oesophageal reflux disease. Br J Surg 2004; 91:975–82. [19] Siddiqui MRS, Abdulaal Y, Nisar A, et al. A meta-analysis of outcomes after open and laparoscopic Nissen's fundoplication for gastro-oesophageal reflux disease in children. Pediatr Surg Int 2011;27:359–66. [20] Peters MJ, Mukhtar A, Yunus RM, et al. Meta-analysis of randomized clinical trials comparing open and laparoscopic anti-reflux surgery. Am J Gastroenterol 2009; 104:1548–61. [21] Mauritz FA, van Herwaarden-Lindeboom MYA, Stomp W, et al. The effects and efficacy of antireflux surgery in children with gastroesophageal reflux disease: a systematic review. J Gastrointest Surg 2011;15:1872–8. [22] Rhee D, Zhang Y, Chang DC, et al. Population-based comparison of open vs laparoscopic esophagogastric fundoplication in children: application of the agency for healthcare research and quality pediatric quality indicators. J Pediatr Surg 2011; 46:648–54. [23] Kane TD. Laparoscopic Nissen fundoplication. Minerva Chir 2009;64:147–57. [24] McHoney M, Wade AM, Eaton S, et al. Clinical outcome of a randomized controlled blinded trial of open versus laparoscopic nissen fundoplication in infants and children. Ann Surg 2011;254:209–16. [25] Knatten CK, Fyhn TJ, Edwin B, et al. Thirty-day outcome in children randomized to open and laparoscopic Nissen fundoplication. J Pediatr Surg 2012;47:1990–6. [26] Mattioli G, Repetto P, Carlini C, et al. Laparoscopic vs open approach for the treatment of gastroesophageal reflux in children. Surg Endosc Other Intervent Tech 2002;16:750–2.
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
A randomized trial of laparoscopic versus open Nissen fundoplication in children under two years of age Dominic Papandria a, Seth D. Goldstein b,⁎, Jose H. Salazar b, Jacob T. Cox b, Kimberly McIltrot b, F. Dylan Stewart b, Meghan Arnold c, Fizan Abdullah b, Paul Colombani b a b c
Department of Surgery, St. Vincent Indianapolis Hospital, Indianapolis USA Division of Pediatric Surgery, Johns Hopkins University School of Medicine, Baltimore USA Section of Pediatric Surgery, C.S. Mott Children's Hospital, Ann Arbor USA
a r t i c l e
i n f o
Article history: Received 22 October 2014 Accepted 2 November 2014 Key words: GERD Nissen fundoplication Laparoscopic Randomized trial Pediatric
a b s t r a c t Aims: The surgery of gastroesophageal reflux disease (GERD) is common in modern pediatric surgical practice. Any differences in perioperative and long-term clinical outcomes following laparoscopic (LN) or open Nissen (ON) fundoplication have not been comprehensively described in young children. This randomized, prospective study examines outcomes following LN versus ON in children b 2 years of age. Methods: Four surgeons at a single institution enrolled patients under 2 years of age that required surgical management of GERD, who were then randomized to LN or ON between 2005 and 2012. A universal surgical dressing was employed for blinding. Analgesia and enteral feeding pathways were standardized. The primary outcome was postoperative length of stay. Perioperative outcomes and long-term follow up were collected as secondary outcomes and used to compare groups. Results: Of 39 enrolled patients, 21 were randomized to ON and 18 to LN. Length of postoperative hospital stay, time of advancement to full enteral feeds, and analgesic requirements were not significantly different between treatment cohorts. The LN group experienced longer median operating times (173 vs 91 min, P b 0.001) and higher surgical charges ($4450 vs $2722, P = 0.002). The incidence of post-discharge complications did not differ significantly between the groups at last follow-up (median 42 months). Conclusions: This randomized trial comparing postoperative outcomes following LN vs ON did not detect statistically significant differences in short- or long-term clinical outcomes between these approaches. LN was associated with longer surgical time and higher operating room costs. The benefits, risks, and costs of laparoscopy should be carefully considered in clinical pediatric surgical practice. © 2015 Elsevier Inc. All rights reserved.
The Nissen fundoplication is a common procedure for managing pediatric gastroesophageal reflux disease (GERD) [1,2]. In adult populations, laparoscopic Nissen fundoplication (LN) has been shown to be as physiologically effective as open Nissen laparotomy (ON) with lower perioperative and long-term morbidity [3–6]. However, the risk/benefit profile of a minimally invasive approach has been less well characterized in children, and practice patterns of pediatric surgeons vary widely based on preference and familiarity. There are limited prior data regarding LN versus ON in children, particularly with respect to long-term outcomes. Furthermore, infants and toddlers, a particularly high-risk population with distinct physiologic and technical challenges, have not been rigorously studied in previous trials. This study is a randomized trial designed to compare LN to ON in children b2 years of age with respect to postoperative
⁎ Corresponding author at: Division of Pediatric Surgery, Bloomberg Children's Center, 1800 Orleans Street, Room 7310, Baltimore, MD 21287-0005 USA. Tel.: +1 410 955 2717; fax: +1 410 502 5314. E-mail address: [email protected] (S.D. Goldstein). http://dx.doi.org/10.1016/j.jpedsurg.2014.11.014 0022-3468/© 2015 Elsevier Inc. All rights reserved.
hospital length of stay (LOS), postoperative pain, time to enteral feeds advancement, and long-term GERD outcomes. 1. Methods 1.1. Study design This was a single-institution, blinded clinical trial in which patients were randomly assigned to one of two treatment groups in a 1:1 ratio to receive Nissen fundoplication via either laparoscopic or open approach. Parents and nurses were masked to the treatment assignment. The study was approved by the Johns Hopkins Medicine Institutional Review Board (Application 72440) and unmasked data for patient safety and data quality were periodically reviewed by an independent Data and Safety Monitoring Committee. Eligible participants were children under two years of age who were referred for surgical management of GERD. Patients were excluded if any concomitant abdominal surgical procedure (apart from placement or revision of a feeding gastrostomy) was to be undertaken under the same anesthetic. The study took place at The Johns Hopkins Children’s Center between December 2005 and January
268
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
2012 and 4 pediatric surgeons participated in patient enrollment and operations following randomization. Sample size was based on the LOS reported in two previous studies and calculated to require 68 patients to be able to find a mean difference of 2 days of LOS with a power of 90% at an α = 0.05 level [7,8]. 1.2. Study procedures and data collection The study protocol specified a standardized pre-operative and postoperative care regimen that was used for all enrolled patients, except for those with clear contraindications. Epidural catheters were placed preoperatively to allow uniform postoperative pain management. Surgical dressings were applied uniformly to mask postoperative nurses and parents to surgical assignment (Fig. 1). Dressings were removed on postoperative day 2 at which time the operative approach was revealed. Enteral feeds were initiated at 24 h postoperatively and advanced according to an algorithm agreed upon by all participating surgeons (Fig. 2). Operative parameters collected included total operating time (exclusive of non-surgical anesthetic time) and total operating room time (time elapsed from patient entry to exit of the operating room). Postoperative parameters included analgesic measures (total days of epidural analgesia, total days requiring narcotic therapy) and hospitalization metrics (LOS, total hospital charges, readmissions within 30 days). Post-discharge parameters included continued use of antacid medications, recurrence of reflux-related symptoms, incidence of technical complications requiring reoperation, and all-cause mortality. 1.3. Interim analysis and trial termination At 71 months into enrollment, the DSMC convened to review the trial data, and ultimately recommended that the trial be truncated following review of ongoing accrual trends versus recruitment targets established in the study design using an a priori power analysis. Final data collection ceased at 24 months after the last patient was enrolled to ensure adequate long-term follow-up. 1.4. Statistical analysis Analysis was performed using GraphPad Prism, version 6 (GraphPad Software, Inc., La Jolla, CA). An unpaired t-test was used to compare continuous variables and Fisher’s exact test to compare categorical variables between groups. Comparison with a Mann–Whitney U test was used for non-parametric data. Comparator groups were defined
and based on intent to treat, using randomization allocation rather than the procedure actually performed. 2. Results A total of 39 patients were enrolled and randomized, with 18 (46%) allocated to LN and 21 (54%) to ON (CONSORT Flowsheet Fig. 3). Mean age and weight of the overall patient cohort were 6.4 months and 6.0 kg, respectively. A history of prior surgery or major procedure was present in 14 (36%) of patients, with 5 (13%) having had a prior abdominal surgery or procedure. Comparisons of patient parameters between LN and ON groups (Table 1) showed no significant differences with respect to estimated gestational age at birth (35.7 versus 33.4 weeks), birth weight (2.5 versus 2.2 kg), patient age (6.6 versus 6.3 months) or body weight (6.3 versus 5.8 kg) at time of surgery. Past history of surgery or major procedures showed no significant difference in total number of patients with prior procedures (5 versus 9) or abdominal operations (1 versus 4) within each group. Perioperative outcomes are shown in Table 2. Postoperative LOS, duration of epidural catheter, duration of narcotic use, and time to full enteral feeds were not significantly different between the two groups. Laparoscopy significantly lengthened the time of surgery (173 versus 91 min, P b 0.001) and the overall operating room duration (209 versus 165 min, P = 0.002). Within the LN group, four (22%) children required conversion from laparoscopic to open approach. Hospital charges were highly variable. Median operating room charges were significantly higher in the LN group ($4450 vs $2722, P = 0.002), but there was no statistically significant difference in total hospital charges due to the wide variability in room and other charges ($26,445 versus $13,906, P = 0.18). The median follow-up period was 42 (interquartile range 15–59) months. All children were seen again after hospital discharge, with no loss to follow-up. There was a 20.5% readmission rate within 30 days of hospital discharge, including in 5 (29%) of the LN patients and 3 (14%) of the ON patients. Six deaths occurred (LN n = 3, ON n = 3). None occurred perioperatively and none were related to surgery or related complications. All were attributed to co-morbid disease (Table 3). There were two complications requiring major reoperation in the LN group; one for intraabdominal leakage from the gastrostomy during the index hospitalization and one for a slipped wrap and hiatal hernia one year after fundoplication. One child underwent successful endoscopic dilatation of a tight wrap due to postoperative swallowing dysfunction. In the ON group, one child presented with fascial wound dehiscence after discharge requiring urgent surgical repair and another was diagnosed with a slipped wrap but has to date elected not to have it repaired. One patient in each group required surgical closure of persistent gastrocutaneous fistula following gastrostomy removal. With respect to GERD outcomes, one child in each group was noted to have recurrent symptomatic reflux, though a very high proportion (n = 28, 71.8%) were still on antacid medication at the time of last follow-up. Comparison of postoperative events revealed no statistically significant differences between groups. 3. Discussion
Fig. 1. Standardized surgical dressing blinding family and hospital providers not present in the operating room.
The scope of this study is of utmost relevance to contemporary pediatric surgical practice. It is widely recognized that infants requiring surgical management of GERD represent a distinct population, with a unique set of technical challenges and a co-morbidity profile that includes high incidence of neurological deficits and congenital abnormalities [9–11]. However, trends toward increasing use of LN have been largely based on studies outside of this population. Here, we report the results of a randomized controlled trial evaluating the efficacy of Nissen fundoplication in children b2 years of age and compare the open and laparoscopic approach.
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
269
Fig. 2. Enteral feeds advancement algorithm. CLD = clear liquid diet; hr = hour; NGT = nasogastric or gastrostomy tube; NPO = nil per os.
Our cohort shows that these data come from a particularly complicated patient population. These were generally infants under 1 year that had been born prematurely and remained underweight. The incidences of congenital anomalies, medical comorbidities, and prior surgeries were high. In this group, we found that laparoscopy added to OR time and cost, but had no significant effect on analgesic requirements, enteral advancement, hospital length of stay, or long-term GERD outcomes. For adults with GERD, LN has overtaken ON as the most common approach and is now generally considered the gold standard [12–14]. Randomized controlled trials have associated LN with reduced morbidity, shorter hospital length of stay, less postoperative pain, and superior durable outcomes [3–5,12,15–18]. Less rigorous data exist comparing surgical approaches for fundoplication in children. Meta-analyses and systematic reviews have consistently concluded that LN results in decreased hospital stay, more rapid return to enteral feeding, and less perioperative morbidity [19–21]. However, nonrandomized studies may offer unreliable comparisons between LN and ON due to selection bias. Patient size, comorbidities, and a history of prior abdominal surgery, among other factors, influence selection of surgical technique [22]. Furthermore,
long-term follow-up data are limited and there has been a suggestion that re-operation rates are higher after LN [20,23]. Two previously published randomized trials have similarities to ours. Notably, both contradict the findings of most retrospective LN versus ON studies in that they do find significant perioperative or postoperative benefit of LN over ON. McHoney et al. randomized 39 patients b13 years to LN versus ON and concluded that there is a reduced rate of retching in LN patients 22 months after surgery compared to ON patients (6% vs. 56%) but that there is no statistical difference in terms of hospital LOS, pain, or rates of dysphagia or recurrence [24]. Knatten et al. similarly randomized 88 children b16 years and concluded that LN is associated with longer duration of operation (150 min versus 89 min) but that there are no statistically significant differences in rates of other 30-day outcomes [25]. Our findings from a predominantly infant cohort agree with these studies, further suggesting that the benefits of laparoscopy are limited to nonexistent in this patient population specifically. Interestingly, the findings of all three randomized trials to date contest the body of nonrandomized studies that indicate LN has improved perioperative and postoperative outcomes, including shortened hospital length of stay, decreased duration of pain, lower rates of postoperative complications, and earlier resumption of feeding [19–21,26]. This discrepancy merits further investigation, particularly given that there are limitations to the McHoney and Knatten studies as well as our own. Unfortunately, the trials do not share a great deal of overlap regarding methodology. For example, Knatten et al. did not mask caregivers using standard surgical dressing, did not standardize enteral feeding advancement, and reported only 30-day outcome. Meanwhile, McHoney et al. focused on the endocrine hormone response in the perioperative period and offered limited follow-up information after median 22 months. Very recently, Pacilli et al. published four year follow-up to that series confirming the lower incidence of retching in Table 1 Patient characteristics.
Age, months; mean (SD) Weight at surgery, kg; mean (SD) Estimated gestational age at birth, weeks; mean (SD) Birth weight, kg; mean (SD) Prior surgery or major procedure, n (%) Prior abdominal surgery or major procedure, n (%) Fig. 3. CONSORT randomized trial patient flowchart.
LN (n = 17)
ON (n = 22)
6.6 (5.3) 6.3 (2.6) 35.7 (5.6) 2.5 (1.2) 5 (29%) 1 (6%)
6.3 (3.6) 5.8 (2.0) 33.4 (6.5) 2.2 (1.3) 9 (41%) 4 (18%)
Abbreviations: LN: laparoscopic Nissen fundoplication, ON: open Nissen fundoplication.
270
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
Table 2 Operative and perioperative results.
Conversion to open, n (%) Operating room time, min; median (IQR) Surgery length, min; median (IQR) Duration of epidural catheter, days; median (IQR) Duration of narcotic use, days; median (IQR) Time to full enteral feeds, d; median (IQR) Postoperative hospital length of stay, days; median (IQR) Operating room charges, USD; median (IQR) Total hospital charges, USD; median (IQR) 30-day readmissions, n (%)
LN (n = 17)
ON (n = 22)
P
4 (23%) 209 (180–248) 173 (140–182) 2 (1–3) 4 (3–5) 4 (2–5) 6 (4–9) $4450 ($3271–$5916) $26,445 ($12,784–$79,327) 5 (29.4%)
N/A 165 (139–184) 91 (75–110) 2 (1–2) 3 (3–4) 3 (3–4) 4 (4–7) $2722 ($2549–$3028) $13,906 ($11,463–$32,540) 3 (13.6%)
0.002 b0.001 0.78 0.26 0.91 0.08 0.002 0.18 0.26
Abbreviations: LN: laparoscopic Nissen fundoplication, ON: open Nissen fundoplication.
their LN group and documenting quality of life improvements in both groups. This study is the first to consider hospital charges in addition to the peri- and postoperative clinical outcomes. The strengths of this trial are methodologic and include randomization, blinding by way of a standardized surgical dressing, and adherence to perioperative protocol. Median long-term outcomes of over 3 years with no loss to follow-up are also a valuable addition to existing literature on the topic. A significant limitation of this study may be the lack of power to detect differences between the groups and the slow accrual over time. Also, the conversions to open during laparoscopy tended to be near the early part of the study period; thus, technical learning curve improvements may have confounded the group comparison. In addition, our results are from a single academic institution with a referral base for complex congenital anomalies and generalizability may be confined to similar large children’s hospitals. In conclusion, the consideration for offering open or laparoscopic Nissen fundoplication for GERD in infants and young children should include perioperative outcomes, long term outcomes, and cost. These have been incompletely studied in children. Our data suggest that the perceived benefits of laparoscopy – namely less perioperative morbidity – may not be realized in children with complex disease in an academic medical center. Appendix A. BAPS Discussion Presented by: Dr. Seth Goldstein Discussant: Mr. Bruce Jaffray (Newcastle) I congratulate you on doing a randomized control trial, but I am really disappointed with your outcomes. Do you really think that there is no difference in post-operative hospital stay in someone having a Laparoscopic Nissen procedure and someone having an Open mid-line incision? In my everyday experience it is chalk and cheese; one group is away home and one group is lying in pain with a morphine infusion stuck on the ward. Does this cohort represent your learning experience for Laparoscopic Nissen surgery? Were these childhood cases that the surgeons were taking on for the first time? Your operative durations again seem to be pretty wild with two and a half hours for your mean Laparoscopic Nissen surgery — this just doesn’t tally with your contemporary experience.
Table 3 Long-term postoperative results at median 42 months.
Mortality, n (%) Reoperation, n (%) Continued symptomatic reflux, n (%) Antacid use at last follow-up, n (%)
LN (n = 17)
ON (n = 22)
P
3 (18%) 2 (12%) 1 (6%) 13 (76%)
3 (14%) 1 (4 %) 1 (4%) 15 (68%)
0.99 0.57 0.99 0.72
LN: laparoscopic Nissen fundoplication, ON: open Nissen fundoplication.
Response: Dr. Seth Goldstein These are very perceptive points. The enrolment started in 2005 which was early in the experience at our institution and I think operative times reflect that. In regard to hospital stay, what we have noticed is that these are very ill children under the age of two. They are referred to us from the neonatal intensive care unit or the paediatric intensive care unit and surgical recovery is not their barrier to discharge. They are in the hospital long after they have recovered from surgery whether that is performed either open or laparoscopic and that is something to consider because it is their cardiac or pulmonary disease that keeps them in the hospital. Discussant: Bruce Jaffray (Newcastle) I will ask you the big question, you have produced data that would seem to suggest that open fundoplication is preferable by virtue of cost. When you go back to your home hospital and you are consenting the next patient for fundoplication are you going to push them towards open surgery? Response: Dr. Seth Goldstein That’s a provocative question! I will tell you that we performed this trial because unlike some centres, not all our Faculty agree, and we do have some surgeons with equipoise on the matter. Discussant: Bruce Jaffray (Newcastle) But what do you believe? Response: Dr. Seth Goldstein I have seen the open fundoplication done in under 45 min and that is good for children who have difficulty with insufflation due to their pulmonary status. I think there is a certain patient for whom I would recommend open Nissen fundoplication in order to decrease the operative time. Discussant: Dr. Warwick Teague (Adelaide) 23% conversion rate – we should never discourage conversion – I take it that you were on an intention to treat basis to compare the laparoscopic group with the converted cases? Response: Dr. Seth Goldstein We did indeed. Discussant: Dr. Warwick Teague (Adelaide) I would have done too, but it really annoys me because those patients have had an open fundoplication. Were you even interested enough to run that analysis in the less comfortable methodological way and put the converted cases in the open group? Response: Dr.Seth Goldstein We thought long and hard about this because with low numbers in each group we thought that could significantly affect our results. I ran them casually both ways and did not find a difference. Ultimately for methodology I would have to assume some confounder just in case that the conversions were due to some unrecognised factors. Discussant: Prof. Paul Losty (Liverpool) Just a comment. There have been two randomised control trials showing no difference in outcome. The first was presented at BAPS by the group from Norway, and second was, I believe, published by the group at Great Ormond Street in the Annals of Surgery. Again there was no actual difference in the outcomes following surgery. Is it all about the scar?
D. Papandria et al. / Journal of Pediatric Surgery 50 (2015) 267–271
References [1] Fonkalsrud EW, Berquist W, Vargas J, et al. Surgical treatment of the gastroesophageal reflux syndrome in infants and children. Am J Surg 1987;154:11–8. [2] Turnage RH, Oldham KT, Coran AG, et al. Late results of fundoplication for gastroesophageal reflux in infants and children. Surgery 1989;105:457–64. [3] Salminen PTP, Hiekkanen HI, Rantala APT, et al. Comparison of long-term outcome of laparoscopic and conventional Nissen fundoplication: a prospective randomized study with an 11-year follow-up. Ann Surg 2007;246:201–6. [4] Salminen P, Hurme S, Ovaska J. Fifteen-year outcome of laparoscopic and open Nissen fundoplication: a randomized clinical trial. Ann Thorac Surg 2012;93:228–33. [5] Siddiqui MRS, Abdulaal Y, Nisar A, et al. Meta-analysis of outcomes after open and laparoscopic Nissen's fundoplication in the treatment for gastro-oesophageal reflux disease. Eur Surg Acta Chir Austriaca 2012;44:138–49. [6] Niebisch S, Fleming FJ, Galey KM, et al. Perioperative risk of laparoscopic fundoplication: safer than previously reported — analysis of the American College of Surgeons National Surgical Quality Improvement Program 2005 to 2009. J Am Coll Surg 2012;215:61–8. [7] Collins III JB, Georgeson KE, Vicente Y, et al. Comparison of open and laparoscopic gastrostomy and fundoplication in 120 patients. J Pediatr Surg 1995;30:1065–71. [8] Luks FI, Logan J, Breuer CK, et al. Cost-effectiveness of laparoscopy in children. Arch Pediatr Adolesc Med 1999;153:965–8. [9] Shah SR, Jegapragasan M, Fox MD, et al. A review of laparoscopic Nissen fundoplication in children weighing less than 5 kg. J Pediatr Surg 2010;45:1165–8. [10] Thatch KA, Yoo EY, Arthur III LG, et al. A comparison of laparoscopic and open Nissen fundoplication and gastrostomy placement in the neonatal intensive care unit population. J Pediatr Surg 2010;45:346–9. [11] Iwanaka T, Uchida H, Kawashima H, et al. Complications of laparoscopic surgery in neonates and small infants. J Pediatr Surg 2004;39:1838–41. [12] Aye RW, Swanstrom LL, Kapur S, et al. A randomized multiinstitution comparison of the laparoscopic Nissen and Hill repairs. Ann Thorac Surg 2012;94:951–8. [13] Tsuboi K, Gazallo J, Yano F, et al. Good training allows excellent results for laparoscopic Nissen fundoplication even early in the surgeon's experience. Surg Endosc Other Intervent Tech 2010;24:2723–9.
271
[14] Wehrli H. Laparoscopic fundoplication for gastroesophageal reflux disease (GERD). Ther Umsch 2005;62:96–9. [15] Sietses C, Wiezer MJ, Eijsbouts QAJ, et al. A prospective randomized study of the systemic immune response after laparoscopic and conventional Nissen fundoplication. Surgery 1999;126:5–9. [16] Laine S, Rantala A, Gullichsen R, et al. Laparoscopic vs conventional Nissen fundoplication: a prospective randomized study. Surg Endosc 1997;11:441–4. [17] Nilsson G, Wenner J, Larsson S, et al. Randomized clinical trial of laparoscopic versus open fundoplication for gastro-oesophageal reflux. Br J Surg 2004;91:552–9. [18] Ackroyd R, Watson DI, Majeed AW, et al. Randomized clinical trial of laparoscopic versus open fundoplication for gastro-oesophageal reflux disease. Br J Surg 2004; 91:975–82. [19] Siddiqui MRS, Abdulaal Y, Nisar A, et al. A meta-analysis of outcomes after open and laparoscopic Nissen's fundoplication for gastro-oesophageal reflux disease in children. Pediatr Surg Int 2011;27:359–66. [20] Peters MJ, Mukhtar A, Yunus RM, et al. Meta-analysis of randomized clinical trials comparing open and laparoscopic anti-reflux surgery. Am J Gastroenterol 2009; 104:1548–61. [21] Mauritz FA, van Herwaarden-Lindeboom MYA, Stomp W, et al. The effects and efficacy of antireflux surgery in children with gastroesophageal reflux disease: a systematic review. J Gastrointest Surg 2011;15:1872–8. [22] Rhee D, Zhang Y, Chang DC, et al. Population-based comparison of open vs laparoscopic esophagogastric fundoplication in children: application of the agency for healthcare research and quality pediatric quality indicators. J Pediatr Surg 2011; 46:648–54. [23] Kane TD. Laparoscopic Nissen fundoplication. Minerva Chir 2009;64:147–57. [24] McHoney M, Wade AM, Eaton S, et al. Clinical outcome of a randomized controlled blinded trial of open versus laparoscopic nissen fundoplication in infants and children. Ann Surg 2011;254:209–16. [25] Knatten CK, Fyhn TJ, Edwin B, et al. Thirty-day outcome in children randomized to open and laparoscopic Nissen fundoplication. J Pediatr Surg 2012;47:1990–6. [26] Mattioli G, Repetto P, Carlini C, et al. Laparoscopic vs open approach for the treatment of gastroesophageal reflux in children. Surg Endosc Other Intervent Tech 2002;16:750–2.
