MANUSCRIPT CITATION
Eaton S, Ganji N, Thyoka M, Shahroor M, Zani A, Pleasants-Terashita H, Ghazzaoui AE, Sivaraj J, Loukogeorgakis S, De Coppi P, Montedonico S, Sindjic-Antunovic S, Lukac M, Hamill J, Choo CSC, Nah SA, Hulscher J, Emil S, Petersen A, Wijnen R, Sloots C, Sigalet D, Kiely E, Svensson JF, Wester T, Pierro A. STAT trial: stoma or intestinal anastomosis for necrotizing enterocolitis: a multicentre randomized controlled trial. Pediatr Surg Int. 2024 Oct 29;40(1):279. PMID: 39470842
REVIEWED BY
Caitlin Eason
University of Colorado, Department of Surgery
Caitlin.eason@cuanschutz.edu
S Christopher Derderian
Children’s Hospital Colorado Division of Pediatric Surgery, Department of Surgery
Sarkis.Derderian@childrenscolorado.org
TYPE OF INVESTIGATION
Treatment: stoma or intestinal anastomosis for necrotizing enterocolitis, an international, multicenter, randomized controlled trial
QUESTION
In infants with necrotizing enterocolitis (NEC) that required a laparotomy (population), did primary anastomosis after resection (intervention) improve intestinal recovery, compared to enterostomy/stoma formation (comparison), as indicated by shorter duration of parenteral nutrition, without an increase in mortality or other post-operative complications (outcome)?
METHODS
• Design: International, multicenter, randomized controlled trial
• Allocation: Randomized intra-operatively online, by the University of Aberdeen Health Services Research Unit, Scotland, UK, using weighted minimization. Criteria included: weight at enrollment, mechanical ventilation required, inotropic support required, extent of disease and intestine involved.
• Blinding: No blinding
• Follow-up period: three months and one year
• Setting: Covered 12 neonatal surgical intensive care units across Europe, Canada, South American, Singapore and New Zealand
• Patients:
o Inclusion criteria: suspected NEC, with need for laparotomy based on radiological signs of intestinal perforation or failure to improve with medical therapy. Enrollment and randomization occurred at the time of laparotomy, based on assessment of disease findings and extent. Only those deemed eligible for either operation by surgeons were then enrolled and randomized.
o Exclusion criteria: no evidence of NEC, focal intestinal perforation, extensive NEC precluding intestinal anastomosis, NEC affecting the colon that could not be completely assessed and intraoperative instability. Also, those who refused consent.
o There were 80 patients recruited between April 2010 and January 2019, but one was randomized twice, so the actual number was 79. Of these, 42 were randomized to the stoma group, with four lost to follow up, and 37 to the primary anastomosis group, with two lost to follow up.
o The research team analyzed 38 patients in the stoma group and 35 in the primary anastomosis group.
• Intervention:
o All patients underwent abdominal exploration through a transverse abdominal incision, regardless of their group.
o The stoma group underwent intestinal resection of the non-viable bowel and stoma at the proximal resection margin.
o The primary anastomosis group underwent intestinal resection of the non-viable bowel and anastomosis of the ends of the viable bowel. Maximum of two anastomoses.
• Outcomes:
o Primary outcome: duration of parenteral nutrition in days
o Secondary outcomes: mortality, unplanned surgical procedures, intestinal complications, including stricture of anastomosis/remaining intestine, confirmed by a contrast study and/or histology, anastomotic leak, prolapse of the stoma, stoma necrosis, intestinal obstruction, a high output stoma and recurrence of NEC
• Analysis and Sample Size: A total of 73 infants were analyzed. A power calculation suggested 66 patients, 33 per group, would be sufficient to detect a difference of 14 days in time to enteral nutrition, with a two-sided alpha level of 5% and power of 80%.
MAIN RESULTS
Baseline patient demographics, clinical characteristics and intraoperative laparotomy findings were similar between the intervention groups.
Primary outcome:
The primary anastomosis group spent less time on parenteral nutrition. Infants were on parenteral nutrition for a median of 51 [interquartile range 3-310] days in the stoma group and 30 [interquartile range 4-105] days in the primary anastomosis group, which was a statistically significant difference (p=0.036 Mann-Whitney test). This remained the case when they were compared by Cox regression analysis, with more than two times the odds (hazard ratio 2.38, 95% confidence interval 1.36–4.16, p = 0.004) of completing parenteral nutrition with primary anastomosis.
Secondary outcomes:
Mortality and most secondary outcomes were similar between the groups. These were single complications, stricture, wound infections, incisional hernias, wound dehiscence, anastomotic leaks and the need for additional abdominal operations, excluding stoma closure. There were significant differences in the number of patients in the stoma and primary anastomosis groups, regarding more than one intestinal complication (12/26 versus 5/31, p=0.02) and stoma complications (15/26 versus 1/31, p=0.0001).
CONCLUSION
Primary anastomosis should be considered when a laparotomy is carried out to investigate NEC. It improves intestinal recovery, with a shorter duration on parenteral nutrition, and reduces the risk of multiple intestinal complications, without increasing adverse outcomes.
COMMENTARY
An estimated 42% of infants with NEC will require surgery for necrotic bowels. It was unknown whether resection with stoma creation or primary anastomosis resulted in better outcomes.1 The STAT trial compared the efficiency of intestinal recovery between these operative interventions and the primary outcome was parenteral nutrition (PN).
Primary anastomosis following a resection remains low (15.8%) in patients with NEC, despite evidence that it is a suitable surgical intervention.2-5 However, it is difficult to evaluate retrospective studies due to inherent bias, because those receiving a primary anastomosis may be less sick than those requiring enterostomy. This trial aimed to minimize this bias by randomizing infants with NEC who could have reasonably had either operation.
The STAT trial methodology highlights the challenges of comparing primary anastomosis to enterostomy, because the inclusion criteria were based on the subjective determination that either operation was appropriate. This compromised generalizability, because it depended on the surgeons’ judgement. Moreover, it is unclear why a surgeon would opt for a stoma when either operation was appropriate. It would be beneficial to obtain information about the surgeons’ decisions to perhaps establish protocols for such an ambiguous decision-making process. Furthermore, it took nearly a decade to recruit 79 patients across 10 centers. Details about those judged ineligible would have been valuable, but the baseline characteristics were similar between the groups. However, more patients in the stoma group required mechanical ventilation.
The primary outcome, of PN duration, was predictably shorter in the primary anastomosis group, as intestinal continuity was maintained. PN is associated with several risk factors, including liver disease, electrolyte imbalances and central line-associated blood stream infections.6 These factors, and infant growth, were not evaluated as additional outcomes and may have provided clinically relevant support for anastomosis than simply PN duration.
The main differences in secondary outcomes between the primary anastomosis and stoma groups were multiple intestinal complications and those related to the stoma. As expected, the stoma group had more stoma-related issues. Multiple intestinal complications were not explicit, making this difference rather obscure. We assumed that this higher rate of multiple complications indicated a more complex postoperative course. However, it was difficult to tell without knowing the combination of complications and the differences in additional abdominal operations. In addition, the study was probably underpowered to detect significant differences in secondary outcomes.
Differences in intestinal complications and PN duration, in conjunction with similar mortality, support the need to evaluate length of stay. This could hypothetically be longer for the enterostomy group, due to the need for more healthcare resources for multiple complications and PN.
The STAT trial findings support primary anastomosis in infants with NEC that require a laparotomy, if surgeons judge this to be safe and feasible. Such operations would probably reduce PN duration and decrease risks of multiple postoperative intestinal complications. Future research should focus on the parameters and protocols involved in these surgical decisions and evaluate the clinically significant results of such outcomes including infant growth, PN risks and length of stay.
REFERENCES
1. Han SM, Hong CR, Knell J, et al. Trends in incidence and outcomes of necrotizing enterocolitis over the last 12 years: A multicenter cohort analysis. J Pediatr Surg. 2020;55(6):998-1001. doi:10.1016/j.jpedsurg.2020.02.046
2. Goldfarb M, Choi PM, Gollin G. Primary Anastomosis Versus Stoma for Surgical Necrotizing Enterocolitis in US Children’s Hospitals. J Surg Res. 2024;295:296-301. doi:10.1016/j.jss.2023.11.005
3. Ade-Ajayi N, Kiely E, Drake D, Wheeler R, Spitz L. Resection and primary anastomosis in necrotizing enterocolitis. J R Soc Med. 1996;89(7):385-388. doi:10.1177/014107689608900708
4. Hall NJ, Curry J, Drake DP, Spitz L, Kiely EM, Pierro A. Resection and primary anastomosis is a valid surgical option for infants with necrotizing enterocolitis who weigh less than 1000 g. Arch Surg. 2005;140(12):1149-1151. doi:10.1001/archsurg.140.12.1149
5. Haricharan RN, Gallimore JP, Nasr A. Primary anastomosis or ostomy in necrotizing enterocolitis?. Pediatr Surg Int. 2017;33(11):1139-1145. doi:10.1007/s00383-017-4126-z
6. Calkins KL, Venick RS, Devaskar SU. Complications associated with parenteral nutrition in the neonate. Clin Perinatol. 2014;41(2):331-345. doi:10.1016/j.clp.2014.02.006
