Clinical impact of timing of surgery on outcomes in preterm infants with surgical necrotizing enterocolitis

Background: The clinical impact of the timing of surgery on outcomes in preterm infants with surgical necrotizing enterocolitis (NEC) is not well defined. Aim: We sought to investigate the impact of the different timing of surgery from the day of NEC diagnosis on clinical outcomes in preterm infants with surgical NEC. Study Design: Retrospective Cohort Study. Subjects: Preterm 75 infants admitted between January 2013 and December 31, 2018, with an NEC (Bell stage III) diagnosis. Outcomes: Comparison of clinical information by the timing of surgery at three different time points (less and more than 48 hours, 96 hours, and 168 hours) in preterm infants with surgical NEC. Results: 75 infants were included in the analysis. Those who received surgery after 48 hours (n= 29/75) had lower median gestational age, lower birth weight, had less pneumoperitoneum, were out born less frequently, had higher acute kidney injury, were intubated and ventilated more frequently, and had higher hemorrhagic and reparative lesions on histopathology than those receiving surgery after 48 hours. Infants receiving surgery after 96 hours had similar trends expect had significantly lower hematocrit and more prolonged parenteral nutrition dependence than less than 96 hours group. The infants receiving surgery after one week had significantly lower birth weight and had higher reparative changes and cholestasis than those receiving surgery < 1 week. There was no significant impact of surgery timing on the length of bowel loss, surgical morbidity, Bronchopulmonary dysplasia, white matter injury, and mortality. Conclusion: The infants receiving surgery later were young and smaller and received parenteral nutrition longer with no significant impact on morbidities and mortality. Our data point out that there are advantages of operating early with fewer morbidities which need further confirmation and evaluation in large multicentric prospective studies or clinical trials.


Introduction
Necrotizing enterocolitis (NEC) affects 3-10% of very low birth weight infants (1,2). NEC is a leading cause of morality and morbidities among preterm neonates, leading to increased health care costs. Those infants with NEC who progress to bowel perforation are managed with either peritoneal drainage (PD) or laparotomy (LAP). The published randomized control trials (3)(4)(5) comparing the Penrose drain and laparotomy did not report any duration of failed medical management, timing of surgery following NEC onset or the impact of intervention (PD/LAP) on the systemic morbidities such as bronchopulmonary dysplasia, acute kidney injury or white matter brain injury on brain MRI (5).
Most of the NEC infants receiving surgical care are initially managed with medical treatment. However, the duration of medical treatment varies among different surgical centers as directed by on-call surgeon. Patients not responding to the medical treatment have been associated with delayed surgery, leading to longer parenteral nutrition dependence and higher morality (6) indicating the impact of timing of surgical invention on clinical outcomes in preterm infants with NEC/SIP. Hence, the clinical impact of timing of surgery on outcomes in preterm infants with surgical necrotizing enterocolitis (NEC) are not well de ned and the optimal time of surgical intervention is still an open question.
In this report, we sought to investigate the impact of different timing of surgery from the day of NEC diagnosis on clinical outcomes in preterm infants with surgical NEC. Our primary hypothesis is that preterm infants with different timing of surgery would have different clinical outcomes. In our NEC cohort study, we did retrospective comparison of clinical information by the timing of surgery at three different time points (less and more than 48 hours, 96 hours and 168 hours) in preterm infants with surgical NEC.

Methods
The study was conducted at a level 4 NICU at the University of Mississippi Medical Center (UMMC) after the Institutional Review Board approval. Our unit admits around 1000 infants annually including referrals from the surrounding counties. The UMMC Institutional Review Board approved this retrospective study with a waiver of informed parental consent. All infants admitted between January 2013 and December 31, 2018 with an surgical NEC (Bell stage III)/SIP diagnosis were included in the study (7). In our study, we had 97 cases, including 75 cases of surgical NEC and 22 cases of SIP. Preterm infants diagnosed with congenital heart disease, medical NEC, and intestinal atresia were excluded. The cohort derivation is summarized in Fig. 1.

Clinical Information
Maternal information collected included chorioamnionitis, antenatal steroids, and pregnancy-induced hypertension (PIH). We collected demographic variables including birth weight, gestational age, Apgar scores, delivery mode, small for gestational age, sex, race, and the out born status We recorded clinical data including respiratory support, inotrope need 24 hours following NEC, patent ductus arteriosus (PDA), ibuprofen /indomethacin treatment, blood culture information sent at the time of NEC onset, antibiotic used, and the frequency of cholestasis (direct bilirubin ≥ 2 mg/dl) following NEC. NEC was de ned using Bell's criteria(7), Bell stage III/surgical NEC frequency was gathered (7). In addition, we recorded information on the age at NEC diagnosis and diagnosis of NEC was made on abdominal X-ray ndings.
SIP infants had pneumoperitoneum on Xray, or intestinal perforation noted on intestinal pathology and less than ve centimeters of bowel resected with no necrosis or in ammation on the histopathology. At our center, infants with birth weight less than 1 kg with pneumoperitoneum and critically sick were managed with PD rst and later received laparotomy if needed.

Histopathological information
Hematoxylin & eosin-stained intestinal tissue sections were evaluated by a pathologist for in ammation, necrosis, in ammation, hemorrhage, and reparative changes and were classi ed as -grade 1 was limited to the mucosa, grade 2 changes extended to the submucosa, grade 3 to the muscularis, and grade 4 change was transmural.
SIP infants had pneumoperitoneum on Xray, or intestinal perforation noted on intestinal pathology and less than ve centimeters of bowel resected with no necrosis or in ammation on the histopathology. At our center, infants with birth weight less than 1 kg with pneumoperitoneum and critically sick were managed with PD rst and later received laparotomy if needed.
Bronchopulmonary dysplasia (BPD) data BPD was categorized as mild, moderate, and severe based on the oxygen requirement at assessment at 36 weeks of corrected gestational age (9). We also recorded information on the type of steroids used following the NEC diagnosis.

Kidney Function Data
The Modi ed Neonatal Staging Criteria described in Improving Global Outcomes (KDIGO) Clinical Practice Guideline for AKI was used to assess the kidney injury (10)(11)(12)(13)(14). We examined daily urine output and all the serum creatinine measurements the day before NEC diagnosis and up to seven days following NEC diagnosis.

Outcome Data
Post-operative information such as parenteral nutrition days, time to full feeds (≥ 120 ml/kg/day), post-operative ileus days, length of stay, surgical morbidities and the death were measured. We de ned mortality as death due to any reason before hospital discharge. Neonates receiving the parenteral nutrition for > 90 days were classi ed in the intestinal failure category. Surgical morbidity was classi ed as wound dehiscence, surgical site infections (including abscesses), strictures, stulas, adhesions, and perforations.

Brain Injury Information
Brain MRI without contrasts obtained at the term equivalent age and scored independently by two pediatric radiologists using a Woodward et al. scoring system of eight scales for white and gray matter injury (15).

Statistical Methods
Demographic and clinical factors in infants with surgical NEC and NEC/SIP cohorts were compared based on the timing of the surgical intervention.
Continuous data were summarized as median (1st quartile, 3rd quartile) with Mann-Whitney U or Kruskal-Wallis tests for differences. Categorical data were expressed as counts and percentages, with group differences assessed by Chi-squared or Fisher's exact tests. All analyses were performed in R statistical software (version 4.2.1), and a p-value < 0.05 was considered statistically signi cant.

NEC Cohort:
The data are summarized in Tables 1-4.

Discussion
Our study is one of the rst reports evaluating the impact of the different timing of surgery from the day of NEC diagnosis on clinical outcomes in preterm infants with surgical NEC comprehensively. Our data suggests infants receiving surgery at later time points following the disease onset were younger and smaller compared to those receiving surgery earlier. Infants receiving surgery after 48 hours were sicker, as evidenced by the higher hemorrhagic lesions on the intestinal histopathology and had AKI. Infants are most likely operated on later because surgeons give a full chance for medical treatment to sick infants before saying yes to surgical intervention. Those receiving surgery later were most likely to receive parenteral nutrition for a longer duration and had cholestasis more frequently than infants receiving surgery earlier, most likely due to delay in starting enteral feeds and reaching full feeds. Additionally, those infants operated on after 96 hours following surgical NEC had more adhesions following disease onset and later received another surgery. However, we did not observe any signi cant impact of the timing of surgery on the length of hospitalization, other systemic morbidities, including BPD, white matter injury on brain MRI and mortality.
Only a few studies report this aspect of surgical clinical care and outcomes; however, a recent population-based, prospective, population-based, observational study of all 27 pediatric surgical hospitals reported that infants not responding to the medical management lead to 30 hours delay in surgical invention and was most likely associated with longer duration of parenteral nutrition or mortality at 4 weeks following the NEC surgery(6). In our cohort, we observed infants operated on after 96 hours or four days following NEC onset received parenteral nutrition for more than 45 days or six weeks. We did not observe any signi cant difference in overall mortality in this cohort; however, the infants receiving surgery in less than 48 hours trended toward 12% lower mortality (28 .3% vs. 41.45, P = 0.36). However, infants operated on later had higher perioperative morbidity than at earlier time points.
The randomized control trials to date do not observe a signi cant difference in the duration of parenteral nutrition or mortality. However, the timing of the surgical intervention relative to the NEC diagnosis was not reported (3,4). Tepas et al. reported the relationship between the seven metabolic derangements (positive blood culture, acidosis, bandemia, thrombocytopenia, hyponatremia, hypotension, or neutropenia) following NEC onset and clinical decisionmaking of intervention and the observation (16). In their cohort, the appearance of any 3 of 7 metrics indicated timely operative intervention, and the mean time between the initial intervention (observation vs. intervention) and evaluation was two days.
Our study's strengths include a detailed and comprehensive analysis of impact of the timing of surgery of the clinical outcomes in preterm infants with surgical NEC.
The limitations of our study included, First, this was a single-center experience, reducing the study's generalizability. Most of the infants with NEC were African American in our cohort, partly due to race distribution in Mississippi. Second, sample size limits our power to detect associations between clinical risk factors and outcomes in preterm infants with surgical NEC and the multiple comparisons may result in type I errors.
In conclusion, the preterm infants receiving surgery later were smaller, younger, had higher hemorrhagic lesions, had acute kidney injury in the perioperative period, were most likely to receive parenteral nutrition for a longer duration, and had more frequent cholestasis than infants receiving surgery earlier.
However, However, we did not observe any signi cant impact of the timing of surgery on the length of hospitalization, other systemic morbidities, including BPD, white matter injury on brain MRI and mortality. Our data point out that there are advantages of operating early with fewer morbidities which need further con rmation and evaluation in large multicentric prospective studies or clinical trials.