Necrotising enterocolitis (NEC) is a serious gut disease that can develop in infants born with cardiac defects. It causes inflammation and injury to the gut mucosa and may be life-threatening.1 The aetiology of NEC in cardiac infants is multifactorial, associated with poor cardiac outflow and poorly oxygenated systemic circulation, resulting in suboptimal blood flow to the intestines.2

Making an accurate and timely diagnosis of NEC is a significant clinical challenge.3 The clinical presentation of NEC is difficult to identify correctly in the early stages of the disease. Of concern, infants who are diagnosed with NEC at a later stage have worse disease.4 Due to the potentially life-threatening effects of NEC, clinicians who suspect a baby has NEC but are unable to confirm it, temporarily manage infants as if they have NEC as a precautionary measure.

This leads to infants without NEC having their milk feeds stopped for several days and being given unnecessary antibiotics. Compared to infants with confirmed NEC, the care of infants with suspected NEC is widely variable in terms of antibiotic regimen chosen, length of antibiotic treatment, and length of time nil-by-mouth - all these variabilities may adversely contribute to the length of hospital stay.5

We want to study a method to improve the accuracy of diagnosing NEC. We want to measure a protein marker found in babies' stools that, if combined with currently available tests (clinical features, lab tests and abdominal X-rays), 6,7 may improve the timeliness and accuracy of making a NEC diagnosis, crucially at an early stage of gut disease.

The protein marker is called calprotectin and monitoring levels in a baby's stools before and after heart surgery may help us to understand its relationship to NEC. Calprotectin is a biomarker that is released by white blood cells and is present in the gut when it is inflamed and can be measured in stool.8 The amount of calprotectin in the stool may indicate the level of gut inflammation and the presence of NEC in infants with heart defects.9

The addition of calprotectin measurement to the current routine parameters (clinical features, lab tests and abdominal X-rays) may improve the timeliness and accuracy of diagnosing NEC in infants with cardiac defects, hence improving their care pathway and outcomes.

Why is this research important in terms of improving the health of patients? Our study will ascertain whether there is a clear relationship between levels of faecal calprotectin and the true occurrence of necrotising enterocolitis in infants with congenital heart disease. This is important because infants with congenital heart disease are burdened with feeding difficulties and poor nutrition for a range of reasons, and they can be affected by necrotising enterocolitis because of abnormal circulation of the blood to all organs, including the gut. If necrotising enterocolitis could be identified correctly in a timely way, this would help clinicians manage the nutrition of these vulnerable infants, at the same time as helping clinicians to target necrotising enterocolitis treatment to babies who truly have this condition.

Considering the different disease processes for infants with congenital heart disease to other clinical populations, such as preterm infants, it is essential to investigate calprotectin concentration in this

1. An inaccurate diagnosis of NEC results in the withdrawal of an infant's milk feed, exacerbating parental stress in intensive care. Improved nutritional support has been associated with increased survival and a reduction in overall morbidity in infants with cardiac defects.10
2. An inaccurate diagnosis of NEC results in the unnecessary administration of antibiotics. Antibiotics cause antibiotic-related diarrhoea, which in turn affects expected growth.11,12
3. An inaccurate diagnosis of NEC results in the inappropriate use of intravenous nutrition as a substitute for enteral feeds during gut rest. Avoiding intravenous nutrition reduces the risk of line sepsis and liver disease.13
4. An accurate and timely diagnosis of NEC will reduce the length of hospital stay. Infants classified with suspected NEC have a longer hospital stay.5,9
5. The current per-patient marginal cost due to NEC is £13,900 and the cost due to feeding difficulties after surgery is £5,500,14, both of which could be reduced with a timely and accurate diagnosis of NEC, improving patient care.

3. THEORETICAL FRAMEWORK Review of existing evidence Importance to the NHS and patients Congenital heart disease (CHD) is diagnosed in approximately 1 in 100 births.15 Significant CHD that required intervention in early childhood affected 3.6 per 1000 live births in England, 2004-2016.16 The unadjusted crude 30-day mortality rate amongst 9,586 paediatric cardiac operations undertaken between 2019-2022 was 2.1% but was highest in the youngest babies.7

The number one priority for the 'Research priorities in children and adults with congenital heart disease: a James Lind Alliance Priority Setting Partnership': how can damage to the gut during heart surgery in children with heart defects be minimised to reduce complications?17

Another priority listed in the same Priority Setting Partnership is 'What are the best approaches to feeding in babies with heart defects, to optimise growth, improve outcomes, and reduce parental distress?'17

In our national prospective study of NEC in 1819 infants who had cardiac surgery in 5 of the 10 UK specialist paediatric cardiac centres, we found 4% had NEC, with this proportion rising to 7.5% amongst neonates under 28 days of age.18

Necrotising Enterocolitis Infants with CHD are at high risk of developing NEC due to persistent diastolic flow reversal in the abdominal aortic, which reduces mesenteric blood flow.19 The aetiology of NEC is multifactorial with the interaction of cardiac surgery, gut hypoxia, poor mucosal integrity, and disruption of intestinal microbial dysbiosis.20,21

Management of NEC involves supportive care, empirical antibiotic therapy, parenteral nutrition, and bowel rest with gastric decompression initiated as soon as NEC is suspected.22 The exact duration of bowel rest in infants with nonsurgical NEC is based on disease severity and clinical judgment.23

Diagnostic Limitations of Necrotising Enterocolitis Interpretation of the clinical picture, including relevant plain abdominal radiographs in infants, is challenging and is one of the reasons for a wide variation in the reported incidence rate of NEC, ranging from 3 to 18%.5 Key markers of NEC, including intestinal dilation, ileus and pneumatosis intestinalis, are often absent on abdominal radiographs.25

The clinical presentation of NEC includes lethargy, temperature instability, recurrent apnoea, bradycardia, hypoglycaemia, and shock. Specific gastrointestinal symptoms include abdominal distention (70% to 98%), blood per rectum (79% to 86%), high gastric residuals after feeding (>70%), vomiting (>70%), and diarrhoea (4% to 26%).24

In an audit of 195 neonates at one cardiac centre (Great Ormond Street Hospital) 14% had feeds suspended during an evaluation for NEC for at least 24 hours but did not have a NEC diagnosis. This clinical path can involve inappropriate use of intravenous nutrition, unnecessary antibiotics administration and longer hospital stays.5,9 Of note, optimal nutritional support in infants requiring cardiac surgery has been associated with increased survival and a reduction in overall morbidity.10

Faecal Calprotectin- Biomarker Neutrophils are first-line responders of the innate immune system and generally combat infections by phagocytosis.26 Calprotectin is a neutrophil activation marker and is mainly exhibited in the cytoplasm of neutrophils and expressed on activated monocytes and macrophages.27 Calprotectin is secreted into the lumen in the early stages of intestinal inflammation; thus elevated faecal calprotectin concentrations are evident in infectious and inflammatory diseases such as inflammatory bowel disease.28 Faecal calprotectin is routinely used as a screening test in the evaluation of inflammatory gastrointestinal diseases and a faecal calprotectin value above 50μg/g should prompt review and a raised faecal calprotectin of greater than 200μg/g should prompt urgent assessment.29

Faecal calprotectin is used as a diagnostic marker for NEC in preterm infants.30 However, infants with CHD pose a unique clinical presentation in that their gut may be inflamed from birth due to compromised gut perfusion and therefore warrant specific investigation.5 Additionally, ischemia-reperfusion injury from cardiac surgery results in pronounced neutrophil activation followed by a hyperinflammation reaction that drives inflammation in NEC.31

Pilot Study Our published prospective single-centre observational pilot study monitored the relationship between faecal calprotectin concentration and gut inflammation in 30 infants admitted with cardiac defects requiring cardiac surgery.

Faecal calprotectin for infants who developed NEC (n=5) had a mean concentration of 3528μg/g (1004SD) compared with 1339μg/g (1008SD) in patients with suspected NEC (n=7) and 390μg/g (410SD) in those without NEC (n=18) (P=0.0001). Infants with suspected NEC had a significantly longer length of hospital stay, on average 18 days longer compared to infants without NEC (P=0.03).9

Research is warranted to establish whether faecal calprotectin can be a useful addition to the diagnostic algorithm for NEC in infants with heart defects.5,9 Important areas for further work include NEC prognostication, which includes earlier identification of the need for care escalation and the requirement for surgery.32 Considering the proposed different disease processes for cardiac infants, it is essential to investigate calprotectin levels independently from other diseases.5,9,31

Size of sample Total sample 300 infants Justification for sample size In a prior multi-centre observational study, the rate of true NEC in this age group was 7% and the rate of inconclusive suspected NEC was 14%. (Brown et al Incidence and risk factors for important early morbidities associated with paediatric cardiac surgery in a UK population, JTCVS 2019; 158;1185-1196)14 Therefore, to ensure faecal calprotectin is clinically useful, we define a minimum acceptable specificity of 80% with 5% precision, i.e. a threshold for the lower bound 95% CI of 75%.

Assuming the expected rate of NEC cases classified according to the gold standard is 7%, and the rate of no NEC cases is 93%, we would need 270 infants to be able to estimate the pre-defined specificity with the required precision. Given the observational nature of the study, lack of harm to the patient from participation and deferred consent, we expect few refusals and a low attrition rate (death), therefore assuming a loss to follow-up/drop-out rate of 10%, we plan to recruit 300 patients.

Buderer, N. M. F. (1996). Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity. Academic Emergency Medicine, 3(9), 895-900.37

10 SCREENING AND RECRUITMENT Children admitted to a participating centre will be screened for inclusion by the clinical care team in the cardiology wards and the paediatric intensive care units, as these babies will be admitted to either area depending on clinical need.

Children who meet inclusion and exclusion criteria will be recruited to the study as soon as possible after admission.

11 CONSENT

Consent Procedures As the passage of a first stool often occurs in the first 6-12 hours after admission to a specialist cardiac centre delaying collection of this stool would mean the loss of information critical to the study. Moreover, no harm will occur to participants related to their participation in this observational study nor from a sample of their stool being collected. We will commence data collection once the child's eligibility is confirmed. This is known as a deferred consent model ('research without prior consent'). This model, developed in line with the CONseNt methods in pediatric Emergency and Urgent Care Trials (CONNECT) study guidance,(38) has been found to be acceptable to parents/guardians and clinicians in several recent RCTs conducted in the UK pediatric critical care setting.( 39,40)

A trained member of the site research team will approach the parents/legal guardian to discuss the study as soon as practical and appropriate. This will usually be within 24-48 hours; however, this will depend on the patient's condition and will be left to the discretion of the clinical team. A Participant Information Sheet (PIS) will be provided. A modified consent procedure will be followed for 1) children in the care of the local authority, 2) where discharge occurs prior to consent being sought, or 3) the child dies before consent is sought. In cases where the local authority have full parental responsibility for a participant, the PIS and covering letter will be sent to the appropriate individual (e.g. the child's named social care worker or foster parent). If the local authority have shared responsibility for a participant, both parties with parental responsibility should be provided with the study information as described above. In the event where a participant is discharged from hospital before consent has been sought or confirmed, the most appropriate member of the site research team will attempt at least one phone call to the parents/legal guardians within five working days of hospital discharge to inform them of the participant's involvement in the study and provide more details. Following on from the call, as well as if there is no response to the call, the parents/legal guardians will be sent the covering letter for discharged patients and a copy of the PIS by post. If no objection is received within four weeks, then the participant's data will be included in the study. In the event where a participant has National Data Opt-out (NDOO) in place, only non-identifiable data will be retained and if there is no such decision recorded, all data will be included in the study.

In the situation where a participant dies before consent has been sought, a site research team member will obtain information from colleagues and bereavement counsellors to establish the most appropriate research team member to notify the parents/legal guardians of the involvement in the research study. Parents/legal guardians can be given the information prior to their departure from the hospital if this is deemed appropriate. If they are not approached prior to their departure from the hospital, then they will be sent a covering letter, and a copy of the bereaved parents/legal guardian PIS within four weeks. If no objection is received within four weeks, then the participant's data will be included in the study. If a participant has NDOO in place, only non-identifiable data will be retained and if there is no such decision recorded, all data will be included in the study.

A signed copy of the study informed consent form will be given to the participant's legal representative and a copy in the participant's medical records. An original signed form will be retained at the study site in a secure place.

A copy of the informed consent document will be given to the parent for their records. The original copy will be filed in the participant's medical notes and a further copy of the signed consent form will be filed in the Investigator Site File.

Participants who may have difficulties in an adequate understanding of English will be supported by GOSH Trust interpretation service. 'Thebigword' is the Trust's sole supplier of telephone interpreting, a service which will enable you to help any client who may have limited English language skills. 0800 694 5093 The consent will be recorded in a patient's electronic notes and in the study records.

12 STUDY PROCEDURES Aim 1: Assess the performance accuracy of faecal calprotectin to diagnose NEC in infants with cardiac defects.

Objective 1 methods:

12.1 Stool collection Stool samples - All participants will have at least two samples taken for measurement of calprotectin levels pre-surgery (sample 1) and post-cardiac surgery (sample 2). Calprotectin will also be collected and measured if the medical team believe NEC may be present (sample 3) 48 hours or more after sample 2 is taken.

All infants recruited to the study will have stool collection (100mg) 1) at baseline (hospital admission), and 2) first stool post-surgery. Patients who have subsequent later concerns for NEC timed after their first post-operative stool sample, will have a further sample (3) collected at this time point.

Stool samples will be collected and stored in plastic test tube containers and directly sent to the hospital clinical laboratory. The calprotectin will be measured using the commercially available enzyme-linked immunosorbent assay (ELISA) kit of Bühlmann Laboratories AG (Switzerland) according to the manufacturer's instructions. In brief, monoclonal capture antibodies highly specific to the calprotectin heterodimeric and polymeric complex are coated on the microliter plate when calprotectin was assayed in a single measurement using DS2 automated ELISA processing system (extended range 30-1,800 μg/g).

A limitation identified from our pilot study is the unpredictability of the infant's bowel opening, therefore for each time point, we will get the first available stool with the possibility of delays in stool arrival.

Index test: Faecal calprotectin concentration will be measured for stool samples 1,2 and if applicable 3.

12.2 Clinical data collection Clinical data will be collected specific to the study limited to the minimum needed to deliver the study objectives.

Data collection will be undertaken daily for up to 30 days after the day of admission to the cardiac centre.

Baseline data - patient diagnosis (cardiac and non-cardiac conditions), weight, gestational age, sex.

Daily clinical data - interventions received by the infant (surgery type and invasive medical treatments), cardio-pulmonary bypass time (min); Dietetic data - Type of feed (expressed human milk, whole protein, hydrolysed or amino acid), actual intake - feed volume per day (ml/kg), total fluid allowance (ml/ kg) and reason for nil by mouth, weight (kg), at birth, admission, and discharge, Z-scores for weight-for-age and length-for-age.

Serial clinical assessment for NEC- all participants will be evaluated using a NEC assessment checklist capturing key features that will be required for the 'gold standard evaluation' for NEC, once per day. This evaluation is part of the routine assessment of these complex patients and includes clinical features (abdominal examination, any blood from gastric or colonic sites, feed tolerance); lab findings (white blood cell and platelet counts, C reactive protein level) and any abdominal radiography (if undertaken then the findings including bowel gas pattern and any signs of NEC such as pneumatosis). 7,

13 DATA ANALYSIS METHODS

13.1 The reference test (gold standard)

The reference test (gold standard) will be undertaken by a blinded endpoint committee at the end of patient participation at which point all serial patient data will be available. The information that this committee will have access to will be - clinical assessments, laboratory tests and abdominal radiography. They will use the definition developed for use in a national audit by a multi-disciplinary group).6,7 Reference test: NEC diagnosis is the identification of a clinical syndrome involving features from 1) plus either 2) or 3) or both 2) and 3) as outlined below:

1. Systemic signs - Include temperature instability, apnoea, bradycardia, raised inflammatory markers, thrombocytopenia, and shock features. In NEC these are present with abdominal and or radiological signs stated below.
2. Abdominal - include significant abdominal distension, reduced or absent bowel sounds, larger than normal gastric aspirates, gastric bleeding, rectal bleeding, abdominal tenderness, or cellulitis.
3. Radiological signs - The abdominal radiograph will be performed in the supine anteroposterior position and will have NEC features of Dilated bowel loops (often asymmetrical in distribution), pneumatosis intestinalis, pneumoperitoneum.

Grading classification for NEC:

Medium severity systemic and abdominal signs and the child recovers after 5-7 days of conservative treatment.

Severe - a more serious picture with perforation and peritonitis.

Adjudication of NEC diagnosis by 'blind endpoint review committee': At the end of the patient's participation, the clinical data on all babies, including the checklist collected for NEC features, and de-identified radiography images, will be independently adjudicated by the 'blind endpoint review committee'.

The committee consists of a paediatric surgeon and cardiac intensivist not involved in patient care, referring to a consensus-based definition (outlined above).6,7 They will be blinded to faecal calprotectin values. The blind endpoint review committee will independently adjudicate a diagnosis informed by serial data points, as either: No NEC or NEC. They will time stamp a NEC diagnosis referenced to patient admission and surgery. If there is disagreement about the final assignment, cases will be reviewed and adjudicated in conjunction with a third paediatric surgeon member. Prof Nigel Hall will lead the committee as one of the adjudicators, assigned as a co-applicant. The second adjudicator will be listed as a collaborator. The committee will divide the no NEC group into patients who had zero features of NEC (we estimate this will be 213 of 270 infants) and patients who had suspected NEC and were then ruled out (we estimate this will be 38 of 270 infants). The committee will identify babies with confirmed NEC to standard (we estimate this will be 19 of 270 infants).

13.2 Statistical Methods Objective 1 Statistical Analysis: We will describe the demographic and clinical characteristics of our sample using mean, standard deviation or median, interquartile range for continuous data as appropriate frequency, and proportion for categorical data. We will check the representativeness of our sample by describing the available characteristics of those patients who dropped out.

Exploratory analysis: Faecal calprotectin concentration pre and post-heart surgery will be described and compared using paired t-tests. Changes in concentration levels across the 2 (or 3) sample values will be derived and described. Regression analysis will be used to explore the association between changes in concentration levels and the timing of the samples, e.g. time of the second sample relative to surgery time. Linear and logistic regression analysis will be performed to assess the relationship between faecal calprotectin concentration and other confounding factors such as weight for age Z-score, length of cardiac surgery time (cardiopulmonary bypass), pre-surgery use of prostaglandin and antibiotics.

Establishing calprotectin cut-off values to diagnose NEC: We will primarily use the calprotectin level from stool sample 2 for this evaluation. We estimate that the timing of this sample will be day 1-3 post-op, as babies are commonly fasted for surgery, this is the median time to the first stool being passed. Based on prospective data collection, the median time to NEC occurrence post-operation was 6 (IQR 4,12) days. In any baby who develops signs of NEC after day 6 post-operation, we will use stool sample 3 instead of sample 2.

The calprotectin measurement from the selected stool sample from all participants will be used to establish calprotectin cut-off values to diagnose NEC (referenced to the gold standard as determined by the blinded endpoint committee). We will explore the data using receiver operating characteristic (ROC) analysis and estimate area under curve (AUC), sensitivity and specificity, along with 95% CIs to distinguish between NEC and No NEC. A ROC curve is a plot of sensitivity against (1-specficity) and shows the test's performance. The better the test, the closer the ROC curve approaches the top left-hand corner which, represents the combination of 100% sensitivity and 100% specificity. The ROC analysis will provide two outcomes: the diagnostic accuracy of faecal calprotectin and the optimal cut-point value for faecal calprotectin. A cut-point will be referred to as optimal when the point classifies most of the individuals correctly, area under curve (AUC).

Progression criterion: we will use the following to judge the success of calprotectin assay - a well-defined optimal cut-off level that has a sensitivity and negative predictive value approaching 1.00, and specificity and positive predictive value >0.8.36

Aim 2. Assess the clinical burden of infants who are treated for NEC but who do not have a confirmed NEC diagnosis.

We will undertake a retrospective secondary analysis to review and describe the impact of adding ROC calprotectin cut-off values to the current diagnostic algorithm for infants diagnosed with NEC at the time they had the index test.

Outcome measures to assess patient burden if re-classified as No-NEC from the suspected NEC group and will include the number of days nil by mouth, anthropometrics, length of hospital stay, additional biochemistry (inflammatory markers), abdominal X-rays and intravenous nutrition.

Objective 2 Statistical Analysis: We will describe with mean, standard deviation or median, interquartile range for continuous data as appropriate frequency, and proportion for categorical data. We will compare clinical characteristics and treatments over the post-op phase, such as the number of days nil by mouth, antibiotic use and percentage feed volume target, for the NEC and No NEC groups of patients using chi-square tests and t-tests as appropriate.

Inclusion Criteria:

• Infants aged from birth up to 3 months old, of any gestational age who require admission to a specialist cardiac centre for treatment of major congenital heart disease.

Included patients are those who are judged as likely to require early cardiac surgery or interventional catheterisation, including babies who are treated with invasive medical therapies of respiratory support or intravenous vasoactive medications and their pathway will include a cardiac intervention of any type.

Exclusion Criteria:

• Patients will be excluded if they are older than 3 months, do not have major heart disease and do have any gastrointestinal abnormality, including gastroschisis or imperforate anus.

Also, patients with neutropenia will be excluded since calprotectin is not activated when there is neutropenia.

Infants with congenital heart disease

Specialist children's hospitals undertaking cardiac surgery in England:

1. Great Ormond Street Children's Hospital
2. Alder Hey
3. Birmingham
4. Southampton hospitals

No: There is not a plan to make IPD available