Current operative approaches for necrotizing enterocolitis

Abstract

Necrotizing enterocolitis is an overwhelming, catastrophic gastrointestinal disease largely affecting premature infants. The disease is manifested by exaggerated intestinal inflammation, which can progress to necrosis of the bowel and systemic sepsis. Prognosis is poor, with unacceptably high mortality. There is currently no known cure, with management focused on supportive measures, including bowel rest, gastric decompression, and antibiotic administration. Surgical intervention is required for cases unresponsive to medical management. In this review, we describe the determination of the need for surgical intervention, the surgical procedures commonly employed, and novel perioperative adjuncts being developed to improve outcomes.

Introduction

Necrotizing enterocolitis (NEC) most commonly affects premature infants in the first weeks of life. The etiology is multifactorial, involving intestinal prematurity, intestinal microbial dysbiosis, heightened immunoreactivity, and dysregulated intestinal microvasculature.¹ The disease is characterized by an exaggerated inflammatory response, which proceeds to intestinal tissue necrosis. Most patients have received enteral feeds. While most commonly associated with prematurity and low birth weight (LBW) of <1500 g, the disease also occurs in term neonates with hemodynamic or pulmonary instability, congenital heart disease, or intrauterine drug exposure.¹ Mortality is approximately 7% for NEC managed medically (medical NEC), but rises to 20%–30% in infants requiring surgery (surgical NEC).¹ Survivors often develop significant sequelae, including intestinal strictures, nutritional deficiencies, intestinal failure secondary to short bowel syndrome (SBS), and poor neurodevelopmental outcomes.¹,³ SBS occurs following extensive bowel resection and leads to lifelong challenges with nutrient absorption and growth.⁴ Thoughtful surgical management, with an emphasis on bowel-sparing approaches, is required to minimize the development of SBS.

There are two main approaches for managing NEC surgically: exploratory laparotomy and peritoneal drainage (PD). Laparotomy, where the abdomen is explored and necrotic bowel resected, has been historically used to manage NEC.^{1 5} Following resection, either ostomy creation with delayed anastomosis or primary anastomosis is performed.⁶ PD was first described in 1977 by Ein and colleagues; five infants with NEC underwent this procedure, which entailed puncture into the abdomen under local anesthesia with drainage of intraperitoneal (IP) contents, saline lavage, and placement of a Penrose drain.⁷ The technique is less invasive compared with laparotomy and spares bowel while potentially attenuating peritonitis. In cases that continue to progress following PD, laparotomy with resection is required, often described as salvage or rescue laparotomy.^{8 9}

This review describes the principal procedures utilized in the surgical management of NEC and provides exposure to alternative bowel-preserving surgical techniques. The review also presents emerging adjunctive therapies and techniques that work concurrently with surgical management to improve mortality, spare bowel length, and diminish long-term sequelae in survivors of NEC.

Differentiating medical NEC from surgical NEC

Diagnosing patients with NEC requires assessment of clinical signs and symptoms and diagnostic imaging studies.^{10 11} Symptoms include abdominal distention, feeding intolerance, hematochezia, and, in advanced cases, signs of sepsis, including refractory hypotension, tachycardia, and temperature dysregulation. Primary imaging studies utilized include abdominal radiographs and abdominal ultrasound.¹¹²,¹⁴ Findings indicative of NEC include fixed and dilated loops of bowel, portal venous gas (PVG), pneumatosis intestinalis, and pneumoperitoneum.¹ Although PVG was previously considered an absolute indication for surgery, it is now considered a relative indication, although it should raise significant concern. Pneumoperitoneum indicates bowel perforation, but it is important to note that many cases of bowel perforation occur in the absence of pneumoperitoneum. Despite extensive research, there is currently no biomarker or laboratory study specific for the diagnosis of NEC. In lieu of this, a combination of laboratory tests including inflammatory markers (including C-reactive protein (CRP) and procalcitonin (PCT)), positive blood cultures, metabolic derangements (specifically metabolic acidosis), and thrombocytopenia are evaluated.¹⁴

Spontaneous intestinal perforation (SIP) presents with intestinal perforation in premature infants, but its pathophysiology differs from that of NEC.¹ Whereas NEC is characterized by exaggerated intestinal inflammation and necrosis, SIP is associated with minimal inflammation or necrosis. SIP is more commonly seen in the first few days of life, linked to postnatal steroids and/or indomethacin use, whereas NEC is more commonly seen after 1 week of life.¹¹⁵,¹⁷ The clinical differentiation between NEC and SIP can be challenging based on laboratory and imaging studies alone. Imaging features unique to NEC include pneumatosis intestinalis and PVG; however, NEC may be present in their absence.¹ To definitively distinguish between the two, visual evaluation of the intestine during surgical intervention is required.

The Bell’s Staging System, developed in 1978, consists of diagnostic criteria utilized in the diagnosis and stratification of NEC.^{10 14 18} Bell’s stage I (suspicion for NEC) is associated with mild feeding intolerance and abdominal distention, without major radiographic findings.^{1 14} Bell’s stage II (definite NEC) is characterized by marked abdominal distention, abdominal wall edema, and elevated intra-abdominal pressure. Patients with Bell’s stage II disease develop laboratory findings (e.g., thrombocytopenia) and imaging findings (e.g., pneumatosis) associated with the disease. Bell’s stage III (advanced NEC) includes patients with hypotension, examination findings consistent with peritonitis, and abdominal imaging concerning for necrosis or perforation (e.g., fixed loops of bowel, pneumoperitoneum). These patients are of greatest clinical concern as they can rapidly progress to multi-organ dysfunction and death, sometimes within hours. Generally, the staging system allows for decisions in medical and surgical management, with Stage I and II patients managed with supportive medical care and stage III patients requiring surgery.¹⁸

A challenging aspect of management is the identification of patients with progression of disease from medical NEC (Bell’s stage II) to surgical NEC (Bell’s stage III). Multiple studies have attempted to identify laboratory markers and risk factors associated with the need for surgery. Risk factors for increased NEC severity include LBW, lack of breast feeding, respiratory dysfunction, lack of maternal corticosteroid administration, and hemodynamically significant patent ductus arteriosus.¹⁹,²¹ Elevated CRP, interleukin 6, serum amyloid A, PCT, and intestinal fatty acid binding protein have been described as being associated with surgical NEC.²²,²⁶ The identification of a sensitive and specific biomarker useful for the diagnosis and prognosis of NEC is a topic of ongoing research.

In lieu of a specific biomarker for NEC, a constellation of clinical, laboratory, and radiographic findings have been used to develop scoring systems to stratify patients into higher and lower risk groups, to predict the need for surgery, and to predict mortality.^{20 27 28} Multiple scoring systems have been developed, with moderate to strong predictive capacity.^{1920 27},³⁰ An effective scoring system is the Neonatal Sequential Organ Failure Assessment (nSOFA). The nSOFA score is a neonatal variation of the adult and pediatric SOFA score, which is used in critical care settings to stratify the severity of sepsis. Scoring is based on respiratory status (intubation and PO₂/FiO₂ ratio), cardiovascular function (steroid requirement and/or vasopressor support), and hematologic status (thrombocytopenia). Scores range from 0 to 15, with higher scores indicating worse prognosis. The nSOFA score has been demonstrated to predict surgical NEC and NEC-related mortality.³⁰ Lewis et al. reported a retrospective multicenter study of infants with Bell’s stage II or greater NEC evaluated by nSOFA before and after the clinical diagnosis of NEC. Patients who died or required surgery had significantly higher nSOFA scores at diagnosis and 48 hours after diagnosis compared with patients with medical NEC (p<0.05).³⁰ Twelve hours after the diagnosis of NEC, nSOFA evaluation reliably identified patients at higher risk for mortality and need for surgery. Additionally, patients with nSOFA score ≥4 at initial evaluation had increased mortality. Nayak et al. evaluated a variant of the nSOFA score called the hyperglycemia, hyperkalemia, use of inotropes for hypotension during the prior week, acidemia, Neonatal Sequential Organ Failure Assessment (HASOFA) score, and its ability to predict mortality or need for surgery compared with nSOFA.²⁹ HASOFA was a better predictor of surgery or death within a week compared with nSOFA. Additionally, when HASOFA was combined with the presence of abdominal wall erythema or PVG, the predictive ability for surgical intervention or death was exceptionally strong. These scoring systems can be powerful tools for providers and can help to identify high-risk patients in order to allow for early intervention.

Peritoneal Drainage (PD)

PD has often been utilized as an initial treatment modality in premature infants who develop NEC or SIP when pneumoperitoneum and clinical deterioration are identified.^{7 31} Much debate exists regarding the best course of action (PD vs. laparotomy) for the initial treatment of NEC (figure 1).^{32 33} Several factors may guide surgeons to pursue PD over laparotomy, including weight <1000–1500 g and/or significant clinical instability.^{34 35} Although PD may only be a temporizing measure with delayed laparotomy eventually required,³⁶ PD may allow time for correction of clinical derangements in order to stabilize an infant for laparotomy.^{37 38}

Figure 1. Overview of current surgical interventions for NEC. The flowchart illustrates two main options for the surgical management of NEC: laparotomy and peritoneal drainage (PD). For patients undergoing laparotomy, the options for limited versus more extensive disease are shown. For patients undergoing PD, those that improve clinically undergo graduated drain removal, whereas those with clinical deterioration undergo rescue laparotomy. NEC, necrotizing enterocolitis. Created in BioRender. Jacobs, T. (2026) https://BioRender.com/yej4krw.

Procedural principles for PD

The technique for PD has been well-described,^{39 40} and the procedure can be performed rapidly in the neonatal intensive care unit. Since patients with surgical NEC are critically ill, they are usually intubated; therefore, a supplemental bolus of analgesic medication and sedation is typically all that is needed. The procedure begins with a small stab incision, typically in the right lower quadrant away from solid organ structures, and the abdominal musculature, fascia, and peritoneum opened. This will frequently result in a rush of air, pus, and/or fecal contents. Warm saline can be used to irrigate the abdomen. A Penrose drain is placed, carefully guided along the anterior abdominal wall, and sutured to the skin.^{39 40} The drain remains in place to facilitate continued drainage of enteric contents.³⁹ Timing and manner of drain removal varies, but in general, the drain remains in place until feculent drainage ceases, at which time the drain is backed out incrementally over ~2 days. Based on the clinical response, the patient may recover from primary PD alone or may require subsequent laparotomy for more definitive management.

PD can attenuate clinical decline by allowing for drainage of enteric contents while avoiding a highly invasive exploratory laparotomy. Moreover, in the clinical scenario of an extremely low birth weight (ELBW) premature infant with clinical decompensation, PD can be lifesaving,^{31 41} as relief of intra-abdominal pressure results in improved venous return with increased cardiac output, relief of upward tension on the diaphragm that may cause respiratory compromise, and correction of metabolic derangements.⁴² Risks of PD include those inherent to any surgery (infection, bleeding, need for additional operations, etc.), and failure to improve clinically requiring subsequent laparotomy and bowel resection.⁴⁰ Careful placement of the drain is imperative, since injury to the liver or spleen can result in rapid exsanguinating hemorrhage. In addition, the size of the insertion site must not be too large, or evisceration of bowel through the insertion site can occur, or hernias can occur after removal of the drain.

Exploratory laparotomy

Exploratory laparotomy is a standard surgical option for infants with NEC or SIP and can be chosen as the primary procedure or as a rescue procedure in patients who have failed PD. After resection of necrotic bowel, stoma creation versus primary anastomosis follow. For cases of extremely extensive or patchy disease, several techniques are available to address the remaining bowel, including “clip and drop,” damage control surgery (DCS) with planned second look laparotomy, diverting jejunostomy, “patch, drain, and wait,” and “shish kebob” techniques (figure 1).^{56 43},⁴⁸

Procedural principles for exploratory laparotomy

Exploratory laparotomy can be performed in the operating room (OR) under general anesthesia if the patient’s clinical status allows for transportation but is often performed at the bedside due to clinical instability.^{15 49} A transverse supraumbilical incision is used. If there is significant concern for hepatomegaly, an infraumbilical transverse incision can be utilized. The abdominal cavity is carefully examined for ascites, bilious content, or fecal matter. The full length of the bowel is evaluated for ischemia, perforation, or other abnormalities. In addition to drainage of spilled enteric contents, abdominal exploration allows for an accurate assessment of intestinal integrity, with excision of necrotic intestine. It is important to measure the length of resected intestine and remaining intestine, recording this in the operative note. If possible, the ileocecal valve should be preserved, but multiple retrospective case-series studies (Level 4 Evidence) have been mixed regarding its true significance.⁵⁰,⁵⁴ If marginal loops of bowel are left in place, it is important to clearly record their degree of injury and location.

Similar to PD, risks include those inherent to any surgery (infection, bleeding, need for additional operations, etc.), in addition to failure to improve clinically requiring repeat laparotomy and further bowel resection. Exploratory laparotomy is an invasive and stressful procedure which can further exacerbate clinically tenuous situations, leading to precipitous hemodynamic decompensation. Moreover, care must be taken to avoid liver or spleen injury, as subcapsular hematoma and hemorrhage are extremely difficult to control and associated with high mortality. In the case of intraoperative exsanguinating hemorrhage, application of hemostatic agents or abdominal packing may be required to control bleeding, with strict efforts to restore normothermia. Use of recombinant factor VIIa has been reported with anecdotal success in a case series of four patients who experienced spontaneous liver hemorrhage during laparotomy for NEC.⁵⁵ NEC is the leading cause of SBS in children, with resection causing irreversible loss of absorptive surface area;⁵⁶ thus, attempts should be made to preserve as much bowel length as possible. Wound complications including infection and dehiscence can occur post-operatively and should be carefully watched for.^{57 58}

Evidence for PD versus exploratory laparotomy as primary surgery

Historically, PD was suggested as a potential adjunct to the surgical management of critically ill extremely premature neonates with intestinal perforation, but was not intended to supplant laparotomy.⁵⁹ Much effort has gone into evaluating PD versus laparotomy in the treatment of NEC and SIP, particularly regarding the utility of PD as primary treatment.^{32 33} In the Moss et al. randomized controlled trial (RCT) (Level 2 Evidence) involving 15 pediatric surgical centers in the USA, there were no significant differences in 90-day mortality, with death of 19 of 55 infants in the PD group compared with 22 of 62 in the laparotomy group (34.5% vs. 35.5%, respectively; p=0.92).⁴¹ There were no significant differences in length of stay (LOS) or total parenteral nutrition (TPN) dependence between the two groups. An international multicenter RCT involving 31 institutions from 13 countries by Rees et al. (Level 2 Evidence) evaluated survival and outcomes for 68 ELBW patients diagnosed with perforated NEC or SIP treated with PD versus primary laparotomy.³⁶ Six-month survival was 51.4% for PD versus 63.6% for laparotomy (p=0.3). Delayed laparotomy was performed in 74% of the patients who underwent initial PD due to lack of improvement, with median time to laparotomy of 2.5 days. Patients who underwent PD with subsequent laparotomy did not have improved 6-month survival compared with primarylaparotomy (relative Risk (RR) of mortality 1.4; p=0.4; 95% CI 0.6 to 3.4). A recent systematic review and meta-analysis published by Solis-Garcia et al. (Level 1/2 Evidence) evaluated three RCTs (n=493) and 49 observational studies of PD vs. laparotomy (n=19 477) .⁶⁰ There were similar survival rates for both NEC and SIP treated with either PD or laparotomy in the RCTs, while analysis of the observational studies found that patients who received PD had lower survival compared with laparotomy. The comparisons of RCTs utilized high-quality data with low risk of bias, whereas the observational study data should be interpreted with caution as the overall certainty of evidence was graded as low or moderate with a moderate to serious risk of bias.

More recent multicenter studies have argued for laparotomy rather than PD in the surgical management of NEC. In a large RCT including 310 patients across 20 US centers (Level 2 Evidence), Blakely et al. evaluated laparotomy vs. PD as the primary surgical intervention for NEC and SIP.³³ This was the first RCT that included evaluation of neurodevelopmental outcomes in survivors of NEC. Primary study outcomes included death or neurodevelopmental impairment (NDI) at 18–22 months corrected age. When assessing NEC and SIP together, there were no significant differences in outcome for patients undergoing laparotomy versus PD (69% vs. 70%, respectively). However, for patients with a preoperative diagnosis of NEC, 69% undergoing laparotomy experienced death or NDI compared with 85% of patients undergoing PD (adjusted relative risk (aRR)=0.81; 95% CI 0.64 to 1.04). The Bayesian posterior probability that laparotomy was beneficial for patients diagnosed with NEC was 97%. On the other hand, in patients with a preoperative diagnosis of SIP, death or NDI occurred in 63% of patients who underwent PD compared with 69% of patients after laparotomy (aRR=1.11; 95% CI 0.95 to 1.31). In summary, recent evidence suggests that exploratory laparotomy may lead to improved mortality and long-term neurodevelopmental outcomes in patients with definitive NEC.

Clinical decision-making for PD versus exploratory laparotomy for primary surgery

While evidence from RCTs reveals similar outcomes between the two procedures, findings from numerous observational studies support improved outcomes with exploratory laparotomy as the primary initial procedure for patients with definitive surgical NEC (table 1). Garg et al. supported performing exploratory laparotomy over PD in cases when patients are hemodynamically stable enough to tolerate the procedure.³⁷ PD may be indicated in ELBW patients who are unable to tolerate a significant surgical procedure due to hemodynamic instability. Providers should take into consideration the risk of poor neurologic outcomes after PD with a preoperative diagnosis of NEC, as noted in Blakely et al.’s trial,³³ recognizing that a preoperative diagnosis of NEC may be difficult to determine without direct visualization of the bowel. Overall, this should be a multidisciplinary discussion between medical and surgical providers, taking into account the patient’s complete clinical presentation.

Table 1. Comparison of current surgical interventions for NEC.

Surgical approach		Clinical scenario	Advantages	Disadvantages	Level of evidence		Cite
Exploratory laparotomy	Stoma and mucus fistula	Indicated in cases of surgical NEC where there is concern for non-viable bowel and patient is clinically stable enough to tolerate surgical intervention; but concern for ongoing hemodynamic instability and/or bowel viability is questionable	Allows for direct visualization and resection of necrotic bowel Lower risk of NDI Avoids risk of anastomotic breakdown Reduces stress on remaining bowel	Surgical risks Invasive procedure Risk of liver or splenic injury leading to hemorrhage Stoma prolapse Parastomal hernia Stenosis Electrolyte or fluid imbalances Necrosis of stoma Ostomy management Multiple procedures	Levels 1 and 2	Level 2, 3, and 4	6 61 to 65 33 36 41 60
	Primary anastomosis	Indicated in cases of surgical NEC where there is concern for non-viable bowel and the patient is clinically stable enough to tolerate surgical intervention; patients who are relatively hemodynamically stable with viable proximal and distal bowel which can be anastomosed without tension	Allows for direct visualization and resection of necrotic bowel Lower risk of NDI Increased nutrient absorption Negates additional procedures to restore intestinal continuity Avoids ostomy management	Surgical risks Invasive procedure Risk of liver or splenic injury leading to hemorrhage Anastomotic leakage Anastomotic stricture		Levels 2, 3, and 4
Peritoneal drainage		Indicated in cases of surgical NEC, often utilized in cases where exploratory laparotomy would be exceedingly high risk	Less invasive Can be performed with minimal sedation at bedside Quick and often effective temporizing measure	Surgical risks Risk of liver or splenic injury leading to hemorrhage Higher risk of NDI Incomplete drainage Inability to assess bowel and resect necrotic tissue	Levels 1 and 2		33 36 41 60
Damage control surgery		Indicated in cases with extensive intestinal involvement, where segments of bowel are of questionable viability and the clinical stability of the patient is tenuous	Allows for bowel reassessment Preservation of bowel length Allows for direct visualization and resection of necrotic bowel	Surgical risks Multiple invasive procedures Risk of liver or splenic injury leading to hemorrhage Delayed abdominal closure Limited clinical evidence for NEC	Level 4		46 66 67
“Clip and drop”		Indicated in cases of multifocal disease in which there are non-viable segments interspersed between viable segments of bowel	Allows for bowel reassessment Preservation of bowel length Allows for direct visualization and resection of necrotic bowel	Surgical risks Multiple invasive procedures Risk of liver or splenic injury leading to hemorrhage Delayed abdominal closure	Levels 3 and 4		43-45 68
Diverting jejunostomy		Indicated in cases of NEC-totalis, where almost all of the intestine is non-viable or of questionable viability	No bowel resection Potential bowel preservation	Surgical risks Multiple invasive procedures Risk of liver or splenic injury leading to hemorrhage Significantly limited clinical evidence for NEC	Level 4		47 69
“Patch drain and wait”		Indicated in cases of NEC-totalis, where almost all of the intestine is non-viable or of questionable viability	No bowel resection Potential bowel preservation	Surgical risks Invasive procedure Risk of continued bowel leakage requiring re-operation Risk of liver or splenic injury leading to hemorrhage Significantly limited clinical evidence for NEC	Level 4		48
Intraluminal stenting without anastomosis (“shish kebob”)		Indicated in cases with extensive multi-focal disease, in which there are many segments of non-viable bowel	Preservation of bowel length Allows for direct visualization and resection of necrotic bowel	Surgical risks Multiple invasive procedures Risk of anastomotic failure requiring re-operation Risk of liver or splenic injury leading to hemorrhage Limited clinical evidence for NEC	Level 4		70-72

The table provides a summary of the advantages and disadvantages of each approach, and the level of evidence available to support each technique.

Levels of evidence: Level 1, systemic review of RCTs; Level 2, randomized control trials, systematic review of cohort studies, or observational studies with dramatic effect; Level 3, non-randomized controlled cohort or follow-up studies, systematic review of case-control studies, systematic review of case-series studies, or systematic review of historically controlled studies; Level 4, case series, case-control studies, or historically controlled studies; Level 5, mechanism based reasoning.¹¹⁹

NDI, neurodevelopmental impairment; NEC, necrotizing enterocolitis; RCT, randomized control trial.

Primary post-resection options: stoma creation versus primary anastomosis

Stoma creation

Stoma creation versus primary anastomosis are the main procedures performed after resection of non-viable bowel. For stoma creation, the proximal end of intestine following resection is utilized to create a stoma. The stoma can be placed through the laparotomy incision to save time or through a separate incision to potentially avoid wound issues.^{15 49} Tension must be avoided to prevent retraction or necrosis. The stoma should be sutured to the fascia and does not need to be matured to the skin. The distal end of the intestine following resection is utilized to create a mucus fistula, allowing for an egress point for distal intestinal contents and gas. Additionally, creation of a mucous fistula allows for post-operative mucous fistula refeeding if desired. Adjacent placement of the functional stoma and mucous fistula simplifies subsequent stoma closure. Stoma creation allows for diversion of enteric contents, avoiding undue stress on the remaining intestine and avoids the risk of anastomotic breakdown. However, stoma formation is associated with many potential complications including dehiscence, prolapse, necrosis, parastomal hernia, stenosis, and electrolyte and fluid imbalances due to high stoma output.⁶¹,⁶⁴

Primary anastomosis

For primary anastomosis, one must ensure that the ends of the bowel are completely viable. Primary anastomosis is generally performed in patients with limited disease and relative hemodynamic stability.^{15 49 65} Following resection of diseased intestine, the proximal and distal ends of the intestine are anastomosed, usually end-to-end in a single layer, in a hand-sewn tension-free manner. Following primary anastomosis, patients will have bowel continuity with intact transit of enteric contents and increased absorption of nutrients. Primary anastomosis negates the need for an additional surgical intervention to restore intestinal continuity. Finally, primary anastomosis avoids the need for ostomy management and ostomy supplies, which can be stressful to patients and caregivers. However, significant complications, including anastomotic leakage or stricture, can occur.

Evidence

Stoma creation versus primary anastomosis has been a subject of debate for decades, with retrospective case series demonstrating comparable safety (Level 4 Evidence).⁶¹⁶⁶,⁶⁸ A 2024 retrospective study using the Pediatric Health Information System (PHIS) database demonstrated that 15.8% of patients underwent primary anastomosis versus ostomy creation, with no statistically significant differences in mortality or other complications. However, a potentially confounding factor in this study was that patients in the primary anastomosis group were statistically significantly older (Level 3 Evidence).⁶⁹ The 2024 STAT trial, an RCT across 12 hospitals in Europe, Canada, South America, Singapore, and New Zealand, randomized patients to either primary anastomosis or stoma creation (Level 2 Evidence).⁶ Patients were considered eligible if they had healthy bowel distal to the anastomosis or stoma. There were no significant differences in outcomes, indicating that primary anastomosis is safe, but these findings were limited to infants with viable remaining bowel, which is not always the case.⁶

Clinical decision-making

While studies have demonstrated that primary anastomosis following bowel resection is a safe reconstruction option, providers should consider the complete clinical picture when determining which technique should be utilized. In patients who are relatively hemodynamically stable, on little to no vasopressor support, and have grossly viable remaining bowel, primary anastomosis can be considered for reconstruction. However, if the patient is hemodynamically unstable and/or their bowel has questionable viability, primary anastomosis should be avoided, and stoma creation should be performed.

Resection and reconstruction techniques for extensive disease

At initial operation, some areas of bowel may be clearly non-viable, requiring resection, with other areas having borderline perfusion and injury without obvious necrosis. Additionally, there are instances where there is extensive disease, and total or near-total resection would be indicated but would ideally be avoided. This is where other techniques, including “clip and drop,” DCS, diverting jejunostomy, “patch drain and wait,” or “shish kebob” may be of use (table 1). If a second-look laparotomy is planned, the abdomen can be left open, using negative pressure dressings or placement of an abdominal silo to allow for direct visualization of the bowel and alleviation of abdominal compartment syndrome if present.^{15 70}

DCS

DCS originated from trauma and military surgical management to address deadly exsanguination, allowing for clinical stabilization before returning for definitive laparotomy,⁷¹ and has transferred to acute care surgery as well.⁷² DCS involves exploratory laparotomy and systematic evaluation of all four quadrants of the abdomen.⁷² Areas of bleeding are controlled with either ligation of bleeding vessels or temporary packing. Areas of enteric content spillage are controlled with intestinal resection or primary repair. Definitively non-viable bowel is resected, with intestinal segments left in discontinuity. The abdomen is irrigated, packing can be left in place, abdominal drains are placed, and a negative pressure wound vacuum assisted closure (VAC) system is utilized for temporary closure. The process is expedited, with the goal of getting patients out of the operating room and into the intensive care unit for continued resuscitation. The key is to prevent hypothermia, coagulopathy, acidosis, and hypocalcemia.^{73 74} Definitive surgery is undertaken as the patient improves clinically.⁷²

Evidence

In 1986, Weber and Lewis compared outcomes of NEC patients who underwent second-look laparotomy due to initial marginally viable bowel compared with patients who did not undergo a planned second-look laparotomy due to normal-appearing bowel at the primary operation (Level 4 Evidence).⁴⁶ Survival was comparable in the second-look laparotomy group (71%) compared with the single operation group (66%). Additional perforations or necrotic bowel were identified in 8/14 cases in the second-look laparotomy group. While demonstrating the utility of second-look operations, the strength of evidence was limited by lack of statistical analyses, poorly controlled grouping, and the limited number of cases included. A 2015 UK retrospective case series study reported nine neonates with extensive NEC who underwent second-look laparotomy (Level 4 Evidence). Six survived to second-look laparotomy, with 44% perioperative mortality.⁷⁵ In a more recent 2019 UK retrospective study of 23 neonates with NEC who underwent DCS, there was 15% mortality, with all deaths due to multiorgan or respiratory failure (Level 4 Evidence).⁷⁶ In patients with extensive intestinal necrosis, or those with intestinal segments of questionable viability, DCS can be useful (table 1). It allows for expedited resection of diseased tissues, which can attenuate clinical decline, while the second-look laparotomy may allow questionable intestine the opportunity to recover, maximizing residual intestinal length. However, there is a paucity of well-controlled, well-powered, randomized trials to elucidate whether DCS has better outcomes compared with traditional single exploratory laparotomy.

“Clip and drop” procedure

The “clip and drop” technique was first described by Vaughan et al. in a case series that included three infants with NEC in 1996 (Level 4 Evidence). The procedure is a variation of DCS, where clearly necrotic bowel is resected, with the remaining bowel segments closed with suture, staples, or clips, and left in discontinuity, with return to the OR after clinical stabilization after 48–72 hours.⁷⁷

Evidence

In addition to the initial description by Vaughan et al., a retrospective case series study by Ron et al. reported 13 infants with multifocal NEC treated using this technique from 1998 to 2006, with six patients surviving and achieving full enteral feeds (Level 4 Evidence).⁴³ Another single-center retrospective case study of 16 patients that underwent “clip and drop” reported eight survivors, with a median time to full enteral feeds of 41 days (Level 4 Evidence).⁴⁴ A multicenter prospective observational cohort study of 236 infants in the UK or Ireland by Allin et al. found that the technique was used in only 4.5% of cases, and multivariate analysis revealed that the necessity of the “clip and drop” procedure was associated with a significantly higher rate of 28-day mortality (aOR=30, 95% CI 3.9 to 237) (Level 3 Evidence).⁴⁵ These studies highlight variability in the success of the technique. Consideration of the “Clip and Drop” technique should be reserved for patients with extensive multifocal disease in which significant bowel length is in jeopardy and bowel preservation is essential (table 1). Evidence demonstrating promising outcomes is largely limited to multiple case series, each of which has limited numbers of patients with survival ~50%.^{43 44 77} While the multicenter observational cohort study by Allin et al. is a more robust study and demonstrated poor outcomes in patients requiring the “Clip and Drop” procedure,⁴⁵ the study is still limited by infrequency of procedure performance (10/236 patients) and heterogeneity of patient presentation. Further studies are needed to assess the procedure with well-controlled comparisons between patients of comparable extent of disease and clinical presentation.

Diverting jejunostomy and “patch-drain-and-wait”

NEC totalis refers to extensive bowel necrosis with<20% of viable bowel remaining.¹⁵ In 2018, a US-based survey of 316 pediatric surgeons presented with cases of infants with varying degrees of disease severity, gestational ages, and comorbidities was conducted. Surgeons offering surgical intervention in cases where <30 cm of bowel remained varied between 11% and 41%, depending on gestational age. They cited issues with long-term intestinal and neurodevelopmental outcomes as reasons for recommending comfort care only.⁷⁸ Thus, resection may not be performed for infants with near-total bowel loss. However, diverting jejunostomy and “patch, drain and wait” techniques have been used to attempt possible patient salvage.^{47 48}

Diverting jejunostomy

The diverting or proximal jejunostomy technique involves the creation of a diverting ostomy above the most proximal extent of affected bowel within the jejunum, with the goal of promoting bowel rest while the patient is optimized medically, attempting to maximize salvage of functional bowel.

Evidence

The technique was first published in 1981 as a case series by Martin and Neblett discussing three infants with NEC and extensive bowel ischemia and necrosis, with one patient surviving to stoma closure (Level 4 Evidence).⁷⁹ Another case series in 2011 by Thyoka et al. evaluated 17 patients who underwent diverting jejunostomy for extensive NEC (Level 4 Evidence). There was 35% overall mortality, with pan-intestinal disease having a higher rate of mortality (67%) compared with multifocal disease (18%). Additionally, 63% of the surviving patients did not require intestinal resection distal to the diverting jejunostomy.⁴⁷ The evidence for this procedure in the management of extensive disease is significantly limited and warrants further study to determine how effective this technique is in comparison to other surgical management strategies (table 1). Diverting jejunostomy should be considered in patients where the extent of disease is near NEC totalis. Patients will need viable proximal jejunum in order to create a safe diverting jejunostomy. The evidence for this technique is limited and should not be a first-line surgical management plan. This approach should be implemented as a last resort surgical option, where the only alternative would be to transition care toward palliative care measures.

“Patch, drain, and wait”

Another technique to address extensive intestinal ischemia is the “patch, drain and wait” technique, where the concept is to “resect no gut and do no enterostomies.” The technique involves decompression of the abdomen via laparotomy, decompression of the intestine through existing perforations, and creation of a Stamm gastrotomy, followed by “patching” of intestinal perforations by loose suture approximation secured with omentum or another segment of intact intestine. The “draining” portion includes placement of Penrose drains from the subdiaphragmatic area coursing down the abdomen laterally and exiting the lower quadrants.⁴⁸

Evidence

In 2000, Moore reported a 15-year case series of 23 patients with NEC totalis using this technique (Level 4 Evidence). There was a 30% reoperation rate, and while most patients had auto-anastomosis of perforated intestine, patients who did not achieve spontaneous anastomosis required surgical intervention to restore intestinal continuity. There was no mortality within 60 days.⁴⁸ It should be noted that this technique and its associated outcomes lack reproducibility in literature. Therefore, while this may be a surgical management option for patients with extensive NEC, further research is needed before widespread implementation, and providers should approach this technique with skepticism as a final operative option.

Intraluminal stenting without anastomosis (“Shish kebob” method)

In cases of multifocal intestinal necrosis, bowel resection can be extensive. Patients with significantly reduced bowel length are at risk of intestinal failure.^{4 53} To balance adequate intestinal resection with maintenance of bowel length, transluminal stenting or “the shish kebob” method has been described.⁸⁰ First described in a case series by Hatch and Schaller for the management of multiple intestinal atresias (Level 4 Evidence),⁸¹ the technique has also been used for multifocal NEC.⁴⁸²,⁸⁵ The general principle is to perform multiple resections of non-viable intestine and link the remaining intestinal sections together over a transluminal tube,⁴ with as much viable bowel preserved as possible.^{80 82 86} A proximal diverting stoma and distal mucous fistula are created. The proximal stoma allows for enteral feeding, and the mucous fistula allows for decompression and possible mucous fistula refeeding.^{4 80 82 86} A catheter (silastic catheter, feeding tube, red rubber catheter, or broviac catheter) is threaded through the lumen of each remaining segment, taking care to maintain proper orientation of the bowel loops.^{4 80 82 86} Each segment is loosely connected with absorbable suture, with at least one stitch on the mesenteric side and one on the antimesenteric side of the junctions, with the two ends of the tube externalized through the abdominal wall.^{4 80 82 86} The segments of bowel are given time to heal and auto-anastomose to each other over ~3–6 weeks.^{80 82 86} Patients can then undergo removal of the tube, fistula takedown, and re-establishment of intestinal continuity.

Evidence

Three separate case series by Okeke et al., Blejter et al., and Lessin et al. evaluated the outcomes of a combined eight patients managed by the shish kebob technique (Level 4 Evidence)^{80 82 86} (table 1). Seven patients had restored intestinal continuity, were tolerating complete enteral nutrition, and survived beyond the first year of life. One patient had their stent tube removed, but was unable to tolerate enteral feeds, remained dependent on TPN, and died at 1 year of age from liver failure.

Adjuncts to the surgical management of NEC

Indocyanine green fluorescence angiography (ICG-FA) for NEC

ICG-FA is a near-infrared fluorescence optical imaging tool that allows real-time assessment of tissue perfusion and has been used for evaluation of tumor margins, tissue perfusion, and anastomoses, with utility in adult and pediatric hepatobiliary, urologic, thoracic, and gastrointestinal procedures.⁸⁷,⁹² The technique is being investigated as an adjunct to the surgical management of NEC to assess intestinal perfusion to questionably viable bowel segments and bowel intended to be anastomosed (figure 2).^{87 92} The technique involves intravenous infusion of ICG which binds to blood lipoproteins, maintaining infused ICG in the intravascular space with minimal tissue leakage.^{87 92} After ICG infusion, the tissue of concern is illuminated at 750–800 nm wavelength and ICG entering the tissue undergoes excitation. Emissions between 800 and 850 nm wavelengths are generated and assessed using a near-infrared camera.^{92 93} Areas of high emission demonstrate well-perfused tissue, whereas areas of low or zero emission are concerning for tissue ischemia.

Figure 2. Utilization of indocyanine green fluorescence angiography (ICG-FA) for NEC. The process is initiated intra-operatively by intravenous infusion of ICG. Shortly after infusion, the bowel in question is exposed to light at 750–800 nm to excite ICG present in the intestinal vasculature. An infrared camera is then used to detect ICG fluorescence emitted from the intestine at 800–850 nm. Intestine with high emission is considered viable, whereas intestine with poor emission is concerning for non-viability. NEC, necrotizing enterocolitis. Created in BioRender. Jacobs, T. (2026) https://BioRender.com/g32hsm7.

Evidence

Le-Nguyen et al. performed a systematic review of 64 pediatric surgical studies including 664 patients who had ICG-FA for surgical management of various conditions (Level 3 Evidence).⁹⁴ Most articles were case reports, with heterogeneity of study populations, interventions, and outcomes, and the calculated risk of bias was high. The analysis was therefore limited since quantitative meta-analysis was not performed and only qualitative analysis reported. There were no adverse events, and 89% of studies reported positive outcomes and benefits from ICG-FA use. Le-Nguyen et al. also performed a prospective mixed methods clinical trial of ICG-FA in 30 pediatric patients undergoing intestinal resection for various etiologies (Level 3 Evidence).⁹⁵ Acceptable fluorescence was achieved in <1 min, with no ICG-FA-related adverse events. Compared with standard visual assessment, viable resection margins based on ICG-FA vs visual inspection differed in 62% of cases (95% CI 0.41 to 0.82). When resection margins were contradictory, ICG-FA more frequently identified viable tissue. A prospective single center study by Pfennigs et al. evaluated the utility of ICG-FA in second look operations for six patients with NEC (Level 4 Evidence).⁹² In four patients, ICG-FA assessment was concordant with visual assessment, with extensive bowel necrosis seen in two of the concordant cases. Two patients had areas of bowel that appeared ischemic by ICG-FA but viable by standard inspection by a senior surgeon. These areas were resected and histopathology confirmed non-viability in both cases.

ICG-FA is a useful tool that can aid in surgical decision making in NEC, but well-controlled and well-powered clinical studies are needed to reliably characterize its utility.⁹⁴ The primary limitations of the technology include high cost and limited equipment accessibility.⁹⁴ ICG is a non-ionic and non-toxic chemical, and there have been no reports of mortality from its use.⁹² Adverse events are rare. Hope-Ross et al. reported an adverse event rate of 0.4%,⁹⁶ while others report events occurring at a rate of 1 out of 42 000 patients,⁹⁵ with the majority of events involving inadvertent subcutaneous injection, allergic flush discoloration of the skin, or anaphylaxis in patients with iodine allergy.^{91 92 94 96} ICG-FA is therefore a safe and effective tool that may aid in the surgical management of patients with NEC.

Direct peritoneal resuscitation (DPR)

DPR is most commonly implemented in the context of damage control laparotomy with significant bleeding or intra-abdominal trauma.^{97 98} Patients in shock shunt blood to vital organs and away from the intestine or liver (“Central Steal”). Even after adequate resuscitation, the intestinal microvasculature continues to experience vasoconstriction, leading to end organ dysfunction, metabolic acidosis, increased intestinal barrier permeability, intestinal edema, and release of inflammatory cytokines. DPR is used to combat this process. DPR typically involves placement of protective coverings over the bowel, placement of an IP catheter for infusion of dialysate, and use of either an improvised or commercially available abdominal wound VAC system to allow for drainage. In some cases, a second IP catheter can be utilized for drainage of dialysate instead of utilizing the abdominal wound VAC system.⁹⁹ Low calcium 1.5% or 2.5% dextrose solutions are used to create an osmotic gradient. As fluid is instilled, it is simultaneously suctioned to prevent abdominal distention and increased intra-abdominal pressure. The gradient decreases edema and washes out inflammatory cytokines and vasoconstrictive factors.

Evidence

In hemorrhagic shock and ischemia or reperfusion rat models, DPR leads to improvements in vasodilation of intestinal arterioles, splanchnic tissue perfusion, and intestinal barrier function.¹⁰⁰,¹⁰⁵ DPR has been well studied in adult trauma patients, with decreased intra-abdominal complications, rates of abdominal hernias, improved sepsis scores (APACHE II, SOFA, and SAPS II), and primary fascial closure rates.¹⁰⁶,¹⁰⁸ Animal studies have evaluated DPR in NEC. In studies of rat models of NEC, Walker et al. and Downard et al. both demonstrated improved intestinal microvascular blood flow with peritoneal dialysate flushing, with increased intestinal arteriolar vasodilation and improved ileal blood flow (Level 5 Evidence).^{109 110} Hopkins et al. reported instillation of peritoneal dialysate into the abdomen of two premature infants with NEC (Level 4 Evidence).⁹⁹ Both patients underwent drainage of intra-abdominal collections under local anesthesia, with placement of two intra-abdominal 10-french catheters to allow for installation and drainage of dialysate. The first patient underwent stoma creation as well. Both had 2.5% dextrose-based peritoneal dialysate instilled every 12 hours, with the dialysate drained by gravity. Both patients had improved skin coloration, urine output, and spontaneous movements. The first patient experienced improvements in acidosis, improved platelet counts, and decreased dopamine support, whereas the second patient did not demonstrate these additional improvements. Both patients had inadequate drainage leading to intra-abdominal fluid retention. Both patients were eventually discharged. These cases demonstrate the potential benefit of DPR in the management of severe NEC. While data support DPR in adult trauma patients and animal models, there is a paucity of support for its use in human neonates. The lack of clinical evidence and standardized protocols is a significant limiting factor. Well-controlled and well-powered studies are needed to assess its efficacy in NEC. Technique optimization and protocols for effective implementation in premature neonates are needed.

Mucous fistula refeeding

In patients with an ostomy and a distal mucous fistula, “mucous fistula refeeding” can be employed by instilling the proximal stoma output into the distal mucous fistula to maximize intestinal absorptive capacity and fluid and nutrient retention.¹¹¹,¹¹⁴ This is beneficial in decreasing fluid losses, improving electrolyte balance, enhancing bowel adaptation, stimulating bowel growth, preventing distal bowel atrophy, decreasing time to goal feeds after restoring continuity, decreasing time to discontinuing parenteral nutrition, and decreasing hepatobiliary complications.¹¹¹,^{113115 116} Mucous fistula refeeding is performed by placement of a catheter into the mucous fistula with instillation of stoma output via an automated pump.¹¹⁴ Care should be taken to monitor for signs of ischemia, perforation, or bleeding, as this could quickly lead to patient deterioration.^{111 117}

Evidence

Solís-García et al. performed a systematic review of 16 observational studies of mucous fistula refeeding, including 623 patients (Level 2 Evidence).¹¹⁸ The refeeding group had significantly fewer days of TPN, shorter time to full feeds, shorter LOS, and decreased cholecystitis. However, the certainty of evidence in this study was graded as low to very low due to study designs and the calculated risk of bias.

Conclusion

NEC is a devastating condition which primarily affects one of the most vulnerable patient populations: premature and LBW neonates. There is no cure for NEC to date, and when supportive care fails, surgical intervention is key. This review has discussed strategies to identify cases of surgical NEC, has critically reviewed the literature for primary surgical principles and techniques designed for extensive disease, and has highlighted adjunctive therapies which have the potential to improve patient outcomes. While the overall goal is to save a child’s life, it is just as important to optimize long-term health and quality of life. Tailored surgical care and perioperative adjunct measures must be utilized to optimize outcomes in situations with often bleak prognoses. The ultimate goal for NEC is to find a cure that will render the use of a scalpel in the treatment of NEC a thing of the past.

Data availability statement

Data sharing not applicable as no datasets generated and/or analyzed for this study.