Abstracts for research and audit cases
Clinical utility of a 104 gene panel for epilepsy: an audit of 654 patients
Catriona McArthur and Sameer Zuberi
Royal Hospital for Children, Glasgow, UK
Abstract
Objectives: Evaluate the clinical use of a gene panel in diagnosis and classification of unexplained epilepsy by looking at:
Diagnostic yield from the panel
Correlation between epilepsy syndrome and diagnostic yield
Correlations between age of seizure onset and diagnostic yield
Categorise the number of pathogenic variants detected in each gene
Rates of variants of uncertain significance (VUS) requiring clinical follow-up.
Methods: Patients were tested on a next generation sequencing (NGS) panel of 104 genes associated with epilepsy or a related neurological condition. The validity of variants detected was determined by geneticists and paediatric neurologists using population databases, pathogenicity predictions and other factors such as the location of the mutation. The variants were classified by pathogenicity using guidelines from American College of Genetics and Molecular Medicine. Clinical information was obtained from a proforma from the referring clinician.
Population: The cohort included 654 patients from across Scotland with unexplained epilepsy, with an age range from 2 days to 60 years. Sixty-six per cent of the patients had infantile-onset epilepsy under the age of two. The patients had a wide range of phenotypes, and 54 different epilepsy syndromes were reported. There was an almost equal distribution of patients with and without developmental delay.
Results: Of 654 patients, 70 were found to have a ‘likely pathogenic’ or ‘pathogenic’ variant (10.7%) in 34 of the genes tested. The gene with most significant variants was PRRT2 (n = 14), followed by CHD2 (n = 5), SCN2A (n = 5) and SCL2A1 (n = 5). Neonatal onset seizures had the highest rate of significant variants (25%), while 100% of cases of benign familial infantile epilepsy (n = 3) were found to have significant variants. A further 174 patients had variants of ‘uncertain significance’ (VUS) which were further investigated, most commonly in the genes DEPDC5 and CACNA1A.
Conclusion: The panel resulted in a diagnosis for 70 patients; which provides increased knowledge for patients and clinicians, and may potentially improve their epilepsy management. The use of an NGS panel allowed a wide spectrum of pathogenic variants to be detected and genes with poorly defined phenotypes to be tested; unlike targeted single-gene testing. The diagnostic yield was highest in patients with infantile-onset epilepsy, suggesting this group should be targeted for panel testing. High numbers of VUS were detected; most commonly in genes with poorly defined phenotypes, and correspondingly diagnostic yield was high in well-described epilepsy syndromes, reinforcing the importance of quality literature in this area.
Paeds 30:30 – a quality improvement challenge
Geetika Kumar1, Jackie Welsh2 and Christine Findlay1
1Paediatric Department, University Hospital Crosshouse, Kilmarnock, UK
2Quality Improvement Team, University Hospital Crosshouse, Kilmarnock, UK
Abstract
Background: The Royal College of Paediatrics and Child Health states that it is everyone’s duty to make any aspect of children’s care better. However, challenges exist in the promotion of quality improvement (QI). The ‘ED 30:30’ was successful in the local Emergency Department, initiating positive change. It comprised of teams developing a QI project for up to 30 min of each shift over a 30-day period.
Aims: 1. Promote QI by staff looking after children, encompassing the neonatal unit, children’s assessment unit, inpatient ward and community paediatrics by means of a 30:30 challenge. 2. Encourage team building and provide a platform for staff to voice and implement new ideas.
Methods: The challenge was advertised verbally, by email with a YouTube clip, with posters and via social media. Teams were created with volunteers within their workplace. Ideas for projects were self-generated or suggested by other healthcare workers. Weekly and overall prizes were offered as an incentive for groups.
Results: Projects covered topics including patient experience, staff wellness, medicines management, clinical efficiency, guideline review and staff learning. From responses to a survey:
– Sixteen per cent of participants had ideas for potential QI that they had not formally put forward
– Eighty-seven per cent felt the challenge was a good team building opportunity
– Ninety-two per cent felt that their opinions were valued
– Ninety-six per cent would be keen to take part in further QI projects.
Criticisms of the challenge from participants included: the Paeds 30:30 being introduced during a busy period of the year, and too many changes made over a short time Hurdles were met in promoting Paeds 30:30 across departments. Human factors were identified to be a challenge to ‘buying-in’ to the concept of Paeds 30:30. Sixty participants over 12 teams completed the challenge. The staff mix included: nurses, allied health professionals, pharmacists, administrators and doctors.
Conclusions: This was a bigger challenge than previous 30:30s due its spread across multiple clinical areas within the paediatric department. It has highlighted that paediatric staff are keen to improve the areas they are working in, given a suitable platform. Further work is required in considering the human factors that may provide a challenge for staff to take their ideas forward. This will be with the ultimate goal of developing a more accessible forum for people to voice their ideas and take part in QI to improve the quality of care we deliver.
Aetiology and accessibility to services for patients with West syndrome in India and Scotland
Priyanka Surana1, Romit Jain2, P Prabhar Shrivastava2, Vivek Jain2, Joseph Symonds1,3, Lucy Kinnel2 and Sameer Zuberi1,3
1Fraser of Allander Neurosciences Unit, Royal Hospital for Children, Glasgow, UK
2Department of Paediatric Neurology, Santokba Durlabhji Hospital, Jaipur, India
3School of Medicine, Glasgow University, Glasgow, UK
Abstract
Background: West syndrome is a severe infantile onset developmental and epileptic encephalopathy. This study explores the differences in aetiology and accessibility to Paediatric Neurology Services in a high resource service (Scottish NHS) and a developing healthcare system (Jaipur, India).
Population and methods: A retrospective review of West syndrome patients presenting over three years (May 2014–2017) to two tertiary hospitals in Glasgow (Scotland) and Jaipur (India). A structured clinical proforma was used in both centres with a minimal data set including onset to clinical presentation, electro-clinical features, aetiology and response to medication.
Results: A total of 132 patients were identified. In Jaipur (104 cases), acquired causes following perinatal events were the most common (68/104; 65%), with predominant male representation (63/68). Genetic (12/36 children) or unknown (24/36) causes comprised the rest. The majority (64%) of Glasgow patients had a genetic (18/28) or unknown cause (6/30). Four patients had an acquired cause with no major sex predisposition. In Jaipur, the median age of presentation to a paediatric neurologist was eight months (2–72 months), in contrast to Glasgow at three months (0–10 months). The time lag of starting specific treatment for spasms after diagnosis was one month (0–66 months) in Jaipur compared to a median of 1.2 days (0–2 months) in Glasgow. Fifty-four per cent (15/28) of Glaswegian patients and 32% (31/97) of Jaipur patients were resistant to drug treatment.
Conclusion: This study identifies differences in the causes and care for children with infantile spasms in Jaipur and Glasgow. Males are much more likely to be presented for medical care in Jaipur. The majority of Jaipur’s patients had preventable perinatal causes compared to predominantly genetic/unknown causes in Glasgow patients. The high percentage of treatment-resistant patients in the Glasgow cohort suggests that aetiology of infantile spasms may be a more crucial determinant of response to treatment than time lag to start therapy. This study highlights the importance of good perinatal care to prevent subsequent development of West syndrome, especially in a low resource developing healthcare system. More than 80% of the world's population lives in countries with low (normal) resource healthcare systems.
Continuous quality improvement to meet the gold standard of growth monitoring in acute admissions
Suzanne Milne, Katherine Jarman and Natalie Bee
Royal Hospital for Sick Children, Edinburgh, UK
Abstract
Background: When children are admitted acutely into hospital there is an easy opportunity to review growth and assess for faltering. Royal College of Paediatrics and Child Health (RCPCH) guidelines recommend that a child’s growth should be monitored if there is any concern about their general health. Missed growth assessment is known to feature in child protection serious case reviews. Electronic growth charts are available electronically and accessible to the multidisciplinary team to facilitate continuity. We hypothesised that these were not being fully utilised in the acute receiving unit (ARU) and not meeting the gold standard.
Aims:
1. To establish if current practice is in line with the gold standard from the RCPCH.
2. If not, establish continuous quality improvement initiative to achieve a higher standard.
Population and method: A series of monthly spot audits between September 2017 and February 2018 was performed. This was interspersed with team education between all cycles, including posters, inclusion of findings in weekly quality improvement email and presentation at a hospital-wide medical meeting. Ninety-four ARU ward patient’s electronic growth charts were reviewed in total. Patients had all been clerked and had a senior review at the time of audit. Measurements collected were height/length and weight, plus occipitofrontal circumference (OFC) in children under two years. Data were recorded on an anonymised Microsoft Excel spreadsheet.
Results: The percentage of patients with their weight plotted doubled initially from 32% in September to 60% in October following departmental education, poster display and inclusion of results in a weekly departmental ‘Learning Points’ email. There was a drop to 50% in November resulting in re-education and a hospital Grand Round presentation. Subsequently, results improved again to 71% in December. OFC followed a similar trend, improving overall from 33% in September to 62% in December. Height and length were consistently poorly documented (<10% patients, all months). Eight-one per cent of children with a measurement plotted were plotted on day 1 of their admission.
Conclusions: Results demonstrate that even a simple task such as plotting children’s growth can be lost in a busy hospital environment. There is an overall improving trend in documentation of weight and OFC (in patients <2 y) on admission to ARU since beginning this quality improvement initiative. The results demonstrate the need for continuous team education to maintain any increase in standard. This is the mainstay of continuous quality improvement and can be applied to any parameter. Work is ongoing in our department targeting the multidisciplinary team to ensure ongoing education and improvement.
Neonatal jaundice surveillance – are we winning?
Tayyaba Yasmeen and Allan Jackson
Princess Royal Maternity Hospital, Glasgow, UK
Abstract
Objectives: 1. To determine the incidence of babies presenting with serum bilirubin (SBR)>400 µmol/l in last five years (2013–2017). 2. To review the factors leading to admission of neonates with SBR>400 µmol/l in last two years (2016–2017).
Design: Retrospective observational study.
Setting: Three neonatal units across GGC Scotland including Princess Royal Maternity Hospital, Glasgow (PRM), Royal Hospital for Children, Glasgow (RHC) and Royal Alexandra Hospital, Paisley (RAH).
Method:
• Yearly data of neonates <28 days treated for very high SBR > 400 µmol/l across GGC between years 2013 and 2017 were collected from the laboratory database.
• Detailed retrospective review of all the neonates treated for SBR > 400 µmol/l in three NNU in 2016 and 2017 was done by case notes, midwifery record notes and online patient database (TrackCare, Clinical Portal).
• Variables including gestational age, age at presentation, gender, family history, type of feeding, postcode, serial bilirubin levels (transcutaneous bilrubinometre (TcB), blood) before and at the time of start of treatment were reviewed to determine the causes of presentation with high bilirubin levels in spite of having a jaundice surveillance system in place.
Results: The number of babies needing treatment due to high SBR ( > 400 µmol/l) has significantly declined from year 2013 to 2017 across GGC (p value = 0.006). A total of 18 patients with SBR > 400 µmol/l were admitted in 2016 (n = 10) and 2017 (n = 8). Lanarkshire had the largest number of babies admitted (n = 8, 44%), and PRM had the highest number of babies admitted (n = 10, 55%) overall. Majority of the babies were male (n = 10, 55%), commonest gestational age group was between 37 and 40w (n = 14, 78%), presenting between 5 and 10 days of life (n = 13, 72%) and referred by the community midwives (n = 17, 94%). TcB was most commonly used to assess SBR before referral to hospital (n = 11, 61%), majority of the babies received treatment for the first time (n = 16, 89%) and most of the babies were breast fed (n = 8, 44). Predisposing factors for jaundice included breast feeding (n = 10), cephalhematoma/bruising (n = 3), hereditary spherocytosis (n = 1) and G6PD deficiency (n = 1). Other factors included lack of close monitoring for jaundice in community (n = 8), discrepancy of correlation between TcB and SBR values (n = 6), inappropriate plotting on bilirubin charts (n = 2).
Conclusion: Incidence of babies presenting with SBR > 400 µmol/l has significantly reduced in last five years. This could be secondary to introduction of TcB in community and improvement in jaundice surveillance system. Although there is still margin for improvement, new cases are still presenting either due to inappropriate monitoring in community or discrepancy in co-relation of levels between TcB and SBR.
The impact of growth hormone therapy upon sleep disordered breathing in children with Prader–Willi syndrome in Scotland
Kerry Naismith and Philip Davies
Royal Hospital for Children, Glasgow, UK
Abstract
Objectives: Children with Prader–Willi syndrome (PWS) are predisposed to sleep disordered breathing (SDB) due to both obstructive and central causes. Growth hormone (GH) therapy has been shown to be beneficial in this population by improving muscle tone, muscle/fat body mass distribution and intellectual outcomes. However, GH may have adverse effects on airway obstruction and screening has been suggested for PWS patients being started on GH. We evaluated the impact of commencing GH on SDB on children in Scotland with PWS.
Population and methods: We retrospectively examined oximetry and cardiorespiratory sleep studies of all children commenced on GH in the RHC, Glasgow between May 2016 and January 2018. Information was gathered on studies prior to GH and 3–6 months post therapy. Any interventions required due to SDB were noted.
Results: Twenty-five patients with PWS were identified as having been commenced on growth hormone from 6 months to 10 years (mean 2.7 years). Prior to commencing growth hormone, 19 had oxygen saturation studies performed, 18 of these were cardiorespiratory studies, 15/19 were identified as having SDB (four mixed, three obstructive, eight central). Three patients went on to have adenotonsillectomy performed due to significant airway obstruction and saturation studies were rechecked prior to commencing growth hormone. Twenty-three patients had saturation studies performed after the initiation of growth hormone, 12 of these were full cardiorespiratory studies, 14/23 were identified as having SDB (nine mixed, two obstructive, three central). The two patients demonstrating obstruction had GH stopped and subsequently underwent adenotonsillectomy; one of these patients had exhibited no evidence of obstruction in previous investigations, the other had no previous studies performed. Another patient, who had previously had an adenotonsillectomy, developed mixed SDB and had their GH stopped. Nine patients had cardiorespiratory studies performed both before and after commencing GH. The mean apnoea hypopnoea index (AHI) prior to commencing GH was 12.5 and the mean AHI post GH was 9.7 (p = 0.0054). The mean oxygen dips >4%/h were 11.3 before GH and 8.2 post GH (p < 0.05). The mean oxygen dips >3%/h were 22.3 pre GH and 16.9 post GH (p < 0.05).
Conclusion: Sleep disorder breathing in patients with PWS is common particularly due to central events, although obstructive apnoeas/hypopnoeas are also found, which may require intervention. Most children developed no GH complications but a minority did develop significant airway obstruction requiring intervention. Screening for SDB both before and after GH therapy is indicated.
Abstracts for clinical presentation
Blue baby blues – a case report. Implications of maternal selective serotonin reuptake inhibitor use for sudden infant death syndrome
Alexander Simpson, Jonathan Coutts and Peter Mulholland
Royal Hospital for Children, Glasgow, UK
Abstract
Background: A girl was born at 38 + 2 weighing 3026 g. The pregnancy was uneventful. The mother took fluoxetine 20 mg daily. Baby was admitted on day 3 following incidents of cyanosis.
Investigations: Respiratory studies revealed significant hypoxia with episodes of hypoventilation and apnoea. Time spent below 94% saturation was 19%, 68 dips/h >4%, pCO2 was raised at 7 kPa. Causes of hypoventilation were excluded with a normal cranial magnetic resonance imaging and negative genetics testing for congenital central hypoventilation syndrome.
Outcome: Incremental increases in oxygen normalised her study. She was discharged on day 14 in 0.5 l/min of oxygen. Parents and other family members were taught basic life support and they were provided with an apnoea monitor. Follow-up at five months shows baby is thriving and the FiO2 was reduced to 0.3 l/min.
Discussion: The ‘triple-risk model’ for sudden infant deaths syndrome (SIDS) describes three important risk factors: a critical development period, an exogenous stressor and an underlying vulnerability. This underlying vulnerability may be accounted for by abnormalities of the serotonin system, as found in post-mortem analysis of SIDS infants. In utero, exposure to a selective serotonin reuptake inhibitors (SSRI) could superimpose this vulnerability onto healthy infants. Mouse models have demonstrated down-regulation of the respiratory response to acidosis in association with maternal fluoxetine use. A population-based study found exposure to an SSRI in utero increased rates of neonatal death, although a causal relationship could not be established. Our case has shown significant hypoventilation in an otherwise healthy infant exposed to maternal fluoxetine during pregnancy with no primary cause identified. This may represent an important link between maternal SSRI use and SIDS, and may have implications for future practice.
Recurrent spontaneous spinal epidural haematoma – a rare emergency in the paediatric ward
Katie Prentice1, Alaa Jalal1, Elizabeth Pilley1, Eric Ballantyne1, Matin Kirkpatrick1, Alice Jollands1, Gavin Main1, Peter Keston2, Drhoslav Sokol
1Tayside Children’s Hospital, Ninewells Hospital, Dundee, UK
2Western General Hospital, Edinburgh, UK
Abstract
Spontaneous spinal epidural haematomas (SSEH) are rare with an incidence of 0.1:100,000 in the adult population. Recurrence is even rarer with only four reported cases in children. No predisposing factors are identified in 40–50% and aetiology remains elusive in 46% of recurrent SSEH. Only one of the four reported paediatric cases had an identified cause.
We describe an 11-year-old girl with recurrent SSEH. She presented in December 2016 with sudden onset of severe back pain and lower limb paresis. Magnetic resonance imaging (MRI) of spine showed a posterior epidural haematoma from C7 to T2 with cord displacement but no signal change. A predisposing coagulopathy was excluded. The patient recovered spontaneously without surgical intervention. Repeat MRI at six weeks showed a resolving haematoma. Magentic resonance angiography (MRA) was normal. Risks and benefits of angiography were discussed but not performed in view of her recovery. She represented 10 months later with sudden onset of high thoracic back pain and urinary retention with reduced power and increased tone in her lower limbs and a sensory level at T8. MRI revealed a further epidural bleed at the same level. Spinal angiography was again considered but not done acutely. A coagulopathy was again excluded. Back pain worsened overnight. Vomiting and unequal pupils were noted but computed tomography of head was normal. Deteriorating power in her lower limbs necessitated a C7/T2 laminectomy and evacuation of the haematoma. Pathology was in keeping with clot formation. Repeat MRI two months later showed a focal area of abnormal signal in the anterior thoracic spinal cord consistent with gliosis. No vascular abnormality was noted. Subsequent formal spinal angiography was normal with no arterio-venous anomalies noted. SSEH are very rare in children and typically occur in the dorsal cervicothoracic region. Rapid surgical intervention is the treatment of choice in the context of spinal compression or worsening paralysis. Spontaneous recovery following conservative treatment is well documented but controversial. Surgical intervention is recommended following a recurrence. The risk for recurrence is however unknown in the absence of identifiable risk factors or aetiology.
Once called a seizure does not mean always a seizure
David Beattie, Dan Hufton and Kammath Tallur
Royal Hospital for Sick Children, Edinburgh, UK
Abstract
A term baby born was born without complication to a mother who is a senior geriatric medical registrar. On day 5 of life, her baby was brought to hospital with what her mother felt were a combination of fasciculations and potential abnormal movements. She was reassured and discharged home with outpatient follow-up. Over the following six-month period, she required acute care on 22 occasions including 14 admissions under the neurology service. On each occasion the problem appeared to be episodes of prolonged focal seizures with a secondary generalisation. A typical episode would involve a degree of eye deviation and facial twitching whilst remaining responsive. This would be followed by repetitive cycling movements of one arm and unresponsiveness. Episodes would last between 20 and 70 min before complete resolution followed by a period of sleep and swift return of normal behaviour. Such incidents would occur on average once per week. Developmentally she remained entirely appropriate for her age. This infant was extensively investigated with almost entirely normal results. Repeated EEG readings were normal including one which captured the entirety of such an episode. The frequency and duration of events were completely resistant to a multitude of trialled anticonvulsant medication. A channelopathy screen shed light on the reason such apparent seizures had proven so resistant to treatment when it concluded that these events were in fact representing alternating hemiplegia of childhood. The infant was identified as having a mutation within the ATP1A3 gene. This gene codes for a protein that forms a portion of Na+/K+ATPase within the brain. Exactly how this causes the clinical manifestations is unclear. The incidence of this condition is thought to be around 1 in 1,000,000 births. Alternating hemiplegia of childhood is characterised by spells of paralysis. This usually affects one side alternating between attacks, but it can alternate within a spell or affect both sides. Attacks can last for minutes, hours and occasionally days. Attacks are usually relieved by sleep. Alongside paralysis, attacks can also take the form of choreoathetosis, dystonia, ataxia, nystagmus and sensory loss. Developmental delay is variably reported. Paradoxically seizures can develop. Treatment involves trying to induce sleep with benzodiazepines, melatonin and chloral hydrate. Flunarizine (a calcium channel blocker) has reported to have some affects in reducing attack frequency or intensity. Raising awareness is important as this rare condition appeared to present as seizures that we see daily in paediatric practise.
Recognising renal cysts and diabetes
Laura Kane and Ian Craigie
Royal Hospital for Children, Glasgow, UK
Abstract
Objective: To appropriately diagnose children presenting with atypical diabetes.
Population: An 81-kg (99.6th centile), 14-year-old female referred by GP with two months of lethargy, osmotic symptoms and a medical history of recurrent UTI, proteinuria, nocturnal enuresis and single left cystic kidney, currently on nitrofurantoin prophylaxis and solifenacin. Her father was diagnosed with T1DM at 25 years, although describes a 10-year history of non-ketotic osmotic symptoms, and at 42 years has hypertension and gastroparesis. Her mother has polycystic ovary syndrome. She was referred for investigation of diabetes with random blood glucose of 10.5 mmol/L and HbA1c of 45 mmol/mol.
Methods: Following normal systemic examination (abdominal striae only, no acanthosis nigricans), normal urinalysis (no glycosuria/ketonuria), post-prandial blood glucose of 7.3 mmol/L and a non-acidotic gas (H+ 34.1, pCO2 5.6, Bic 30.6, BE 6.2, Glu 7.1, Lac 1.9), she was discharged from the emergency department with GP follow-up. One week later, with a fasting blood glucose of 6.5 mmol/L, she was referred to the diabetes service for further investigation, due to ongoing symptoms.
Results: Fasting blood glucose 6.4 mmol/L; OGTT – T0 = 6.4 mmol/L, T30 = 14.8 mmol/L, T60 = 18.7 mmol/L, t90 = 17.7 mmol/L, t120 = haemolysed; HbA1c 46 mmol/mol (20–42); urinalysis – 2+ protein only; urea and electrolytes (U&Es), liver function tests (LFTs) and thyroid function tests (TFTs) normal. Insulin dependent diabetes was diagnosed and subcutaneous injections commenced. Weeks later, outstanding bloods were reported: insulin – 24.6 µ/L (<13.0) during OGTT; C-peptide – 4.01 nmol/L (0.36–1.12); autoantibodies negative – anti-GAD < 5.0 U/mL (0.0–5.0), anti-IA2 < 7.5 U/mL (0.0–7.5); genetics – HNF1β whole gene deletion detected. Hypomagnesaemia and hyperuricaemia were confirmed – Mg2+ 0.6 mmol/L (0.7–1.0), urate 547 µmol/L (140–360). Insulin therapy was continued, HbA1c normalised (35 mmol/mol) and family members were referred for genetic counselling/testing.
Conclusions: Genetic testing detected HNF1β gene deletion, confirming renal cysts and diabetes (RCAD) monogenic diabetes. Maturity onset diabetes of the young (MODY) affects 2% of childhood onset diabetes. With a heterogeneous clinical presentation, 80% of cases of MODY are undiagnosed or misdiagnosed. This case highlights the importance of not being falsely reassured by normal fasting blood glucose and the need to formally investigate all patients with osmotic symptoms. Correctly diagnosing MODY optimises patient management, informing pharmacotherapy (many forms of MODY do not require insulin), prognosis and facilitates genetic counselling for family members. Monogenic diabetes should be considered when a family member has been diagnosed before 25 years and the patient has extra pancreatic features (such as renal cysts), mild fasting hyperglycaemia, negative autoantibodies, detectable C-peptide post-remission phase or low insulin requirement (<0.5 unit/kg/day).
Immune deficiency hiding in the bowels
Eilidh Thomson, Louise Gannon and Buddhi Gunaratne
Tayside Children’s Hospital, Ninewells Hospital, NHS Tayside, Dundee, UK
Abstract
A 10-year-old girl presented to the paediatric immunology service with recurrent sinusitis, persistent respiratory infections and lethargy. Initial investigations highlighted elevated IgG and undetectable IgA levels but normal lymphocyte subsets and vaccine responses. Within four weeks of presentation she developed intermittent abdominal pain, increased stool frequency and vomiting. Her IgG anti-endomysial antibody test was positive and she was referred to paediatric gastroenterology for investigation of possible coeliac disease. Upper and lower gastrointestinal endoscopy revealed mucosal inflammation with epithelioid granuloma within both oesophagus and colon. Incomplete clinical correlation of endoscopy findings with intermittent, self-resolving gastrointestinal symptoms and concern regarding immune status prompted further investigation. A neutrophil oxidative burst dihydrorhodamine (DHR) test was requested and found to be abnormal with a pattern suggestive of deficiency in p47phox. This was confirmed by protein analysis and a diagnosis of chronic granulomatous disease (CGD) was made 18 months after initial presentation. CGD is a rare immunodeficiency (prevalence 1:250,000) caused by defects in protein components within the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex. Seventy per cent of CGD cases are the result of X linked inheritance conferring deficiency in gp91phox protein and result in serious bacterial and fungal infections. P47phox deficiency is of autosomal recessive inheritance and represents 25% of CGD patients with a milder phenotype. Whilst GI symptoms are common in CGD, literature suggests only a small number of patients initially present with these symptoms. Interestingly, following an uneventful and successful 12 months of prophylactic antibacterial and anti-fungal treatment, the patients gastrointestinal symptoms returned with a markedly raised faecal calprotectin and likely CGD associated colitis. In addition, she developed persistent thrombocytopenia, prompting expedited preparation and planning for bone marrow transplant. As this case highlights, a high index of suspicion and clinical vigilance are required in cases of possible inflammatory bowel disease where even mild infectious symptoms are present, as use of immunosuppressant medication in these patients can have severe infectious consequences. Management requires a collaboration between paediatric gastroenterologists and immunologists and patients require dual follow-up.
A case of irreversible posterior reversible encephalopathy syndrome ecstasy ingestion?
Laura Combe, Elizabeth Pilley and Alice Jollands
Neonatal Unit, Tayside Children’s Hospital, Ninewells Hospital, Dundee, UK
Abstract
Previously, healthy 14-year-old boy presented with serotonin syndrome following ecstasy ingestion. Admitted to adult ICU and developed life-threatening multi-organ failure requiring intubation, ventilation, cooling, inotropic support and renal dialysis. Discharged to paediatric High dependancy unit (HDU) on D12 with ongoing renal failure, sedation withdrawal, LRTI and hypertension. On D22 had prolonged generalised seizure. Computed tomography imaging showed posterior reversible encephalopathy syndrome (PRES). Required re-admission to ICU and blood pressure management with labetolol and hydralazine. No further seizures following this and making a steady recovery. PRES is a syndrome of clinical and radiological findings. More commonly described in adults, it is increasingly recognised in the paediatric population with reported incidences of 0.4% within paediatric critical care, 4.7% following liver transplant and 5.3% post-stem cell transplant. Clinical features include seizures, headache, visual disturbances and encephalopathy. Case reports suggest multiple associations including hypertension, drug ingestion (recreational and prescribed), renal disease, organ transplant, hematopoietic disease and anaemia. Typical CT/MRI imaging appearances are abnormalities of grey and white matter with predominance, but not exclusivity, in the parietal–occipital regions, likely secondary to vasogenic oedema. Areas on MRI of restricted diffusion can be associated with irreversible injury leading to incomplete recovery. Intra-cranial haemorrhages have also been described. The pathophysiology is incompletely understood but is thought to represent endothelial dysfunction causing hyperperfusion (or hypoperfusion) resulting in breakdown of the blood brain barrier and vasogenic oedema. Postulated causes of endothelial dysfunction include changes in blood pressure overcoming cerebral autoregulation or the direct effect of circulating cytokines. There is no specific treatment for PRES. Management involves identification and treatment of precipitating factors and seizure control. Many will require ICU admission. Most children will fully recover within a week, including resolution of CT/MRI changes. Although defined as ‘reversible’, some studies have reported deaths or ongoing neurological sequelae. Early recognition and management may reduce risk of this. Our patient had clear risk factors for PRES (hypertension and drug ingestion). He was discharged on D32 with ongoing renal and neurology follow-up. Repeat MRI imaging has shown resolution of initial oedema but a small focus of cortical laminar necrosis in the right occipital lobe. He has ongoing issues with memory impairment and is awaiting neuropsychology assessment at six months post event.
