Data usage in paediatric-onset inflammatory bowel disease epidemiology: from clinical observations to nationwide data linkage

Abstract

BACKGROUND: Inflammatory bowel disease (IBD) is a chronic, immune-mediated inflammatory disorder which comprises Crohn’s disease (CD), ulcerative colitis (UC) and IBD unclassified (IBDU). When diagnosed in childhood or adolescence it is classified as paediatric-onset IBD (PIBD), with this phenotype shown to be more extensive and more aggressive than adult-onset IBD. The incidence and prevalence of PIBD continue to climb globally, with robust, population-based epidemiological data vital to organising current health care provision and planning future service design. Unique longitudinal data captured within Scotland has enabled detailed exploration of the epidemiology of PIBD, providing the ability to recognise rare phenotypes, calculate nationwide incidence and prevalence rates, describe the landscape of biologic medicine use, and employ administrative health data to link PIBD with potential perinatal risk factors as well as morbidity and mortality outcomes.

AIMS: This thesis aims to use local, national and administrative epidemiological data on Scottish PIBD patients to investigate the following: 1.Local case series of rare PIBD phenotypes including stricturing duodenal CD and fissuring perianal CD – to calculate incidence rates, describe clinical cases and summarise treatment options. 2.The nationwide incidence and prevalence of PIBD in Scottish children. 3.The nationwide incidence and prevalence of biologic use within Scottish PIBD services, including demographics and outcomes for those patients who required escalation beyond anti-TNFα. 4.Whether mode of delivery, gestational age at birth or type of infant feeding contribute to the development of PIBD in a nationwide cohort of Scottish children. 5.Whether PIBD patients are at increased risk of ischaemic heart disease, stroke, cancer and death in early adulthood.

METHODS: To describe local case series of rare PIBD phenotypes; a regional cohort of prospectively acquired incident cases of paediatric CD diagnosed <16 years of age in South-East Scotland was captured prospectively over a 19-year period (1999-2018). Retrospective case note review was conducted on the medical records of all patients, together with a review of the available literature and consensus guidelines. Incidence rates for all CD and for duodenal stricturing CD and complex fissuring perianal CD were calculated.

To determine nationwide incidence, prevalence and use of biologics; a database of incident and prevalent PIBD cases was prospectively recorded by the three Scottish regional paediatric gastroenterology networks covering all paediatric units nationwide. Incidence rate, plus point (30th June each year) and period prevalence (calendar year) were calculated against age-specific population data for PIBD cases diagnosed <16 years of age within Scotland. All patients in PIBD services diagnosed <17 years of age and received biologic medicines during the study period were retrospectively audited through case note review.

To investigate potential perinatal risk factors and determine morbidity and mortality outcomes in young adulthood; administrative health data for PIBD patients diagnosed <16 years of age between 1981-2017 (and relevant controls) was identified, with validation performed within an entire Scottish health board. Utilising this database; a retrospective birth cohort study was performed with data-linkage to determine mode of delivery, gestational age at birth, and type of infant feeding; a second cohort study was also performed to identify any diagnosis of ischaemic heart disease, stroke, cancer or mortality age 17-39 years. Hazard ratios (HR) were calculated for each exposure.

RESULTS: Within the regional South-East Scotland PIBD cohort, 247 new cases of paediatric CD were identified over the 19-year the study period. The overall paediatric CD incidence rate was 5.70/100,000/year. Two incident cases of stricturing duodenal CD occurred, giving a specific duodenal B2 phenotype disease incidence rate of 0.05/100,000/year (representing 0.8% of incident cases at diagnosis) within this regional cohort. Four patients developed complicated perianal fissuring disease with a cumulative incidence of 1.6% within this regional cohort.

Nationally, 438 patients with PIBD were diagnosed in Scotland 2015-2018 providing an overall incidence of 12.0/100,000/year within Scotland. The number of prevalent patients per year ranged from 518-538 with the highest period prevalence 58.5/100,000/year (01.01.2016 - 31.12.2016) and the highest point prevalence 45.9/100,000/year (30.06.17). 495 incident new-start biological therapies were commenced on 403 PIBD patients: 295 (60%) infliximab, 161 (32%) adalimumab, 24 (5%) vedolizumab, 15 (3%) ustekinumab. The proportion of new-start biologics changed with infliximab initiation rates decreasing (87% to 54%) while adalimumab (13% to 31%), vedolizumab (0% to 9%) and ustekinumab (0% to 6%) all increased. The incidence rate (1st dose of new biologic not including biosimilar switch) increased from 6.9% to 8.1% over the study period. Point prevalent rates of current biologic therapy increased from 20.2% on 30.06.2015 to 43.5% on 30.06.2019 within Scotland. Biosimilar penetration of new-start anti-tumour necrosis factor alpha monoclonal antibodies increased from 3% to 91%.

Administrative health data identified a cohort of 2,013,851 children born in Scotland 1981-2017, including 1,721 PIBD cases. Children delivered vaginally did not have an altered risk of developing PIBD compared to those delivered by caesarean section; adjusted HR 0.95 [95% CI 0.84-1.08] (p=0.46). Compared to children born at term (≥37 weeks), children born prematurely did not have an altered risk of developing PIBD; 24-31 weeks gestation HR 0.99 [95% CI 0.57-1.71] (p=0.97); 32-36 weeks gestation HR 0.96 (95% CI 0.76-1.20] (p=0.71). Compared to children exclusively breast fed at age 6 weeks, children exclusively formula fed did not have an altered risk of developing PIBD; adjusted HR 0.97 [95% CI 0.81-1.15] (p=0.69). A cohort study comprising 1,907 children with PIBD and 5,720 matched controls with 127,322 total person-years follow-up demonstrated few episodes of IHD (≤5) or stroke (≤5) for PIBD cases in early adulthood. PIBD cases did have a significantly higher risk of any cancer diagnosis (54 cases: 80 controls) HR 2.07 (p<0.001). PIBD cases had a greater risk of mortality in early adulthood (60 cases: 51 controls) HR 3.61 (p<0.001), including a greater risk of suicide which was not statistically significant; HR 2.27 (p=0.13).

CONCLUSION: Scotland’s regional and national population-based epidemiological data for PIBD is invaluable in its ability to assess for rare phenotypes, demonstrate sustained epidemiological trends over time and provide a nationwide overview of the treatment landscape. The incidence of PIBD in Scotland continues to increase, in keeping with global trends, and the prevalence rate is higher than for any other current published cohort (aged <16yrs). Complete accrual of Scottish biologic usage within paediatric services demonstrates a rapidly changing, inexorably increasing PIBD biologics landscape. Ongoing health workforce planning is therefore essential to maintain high quality and timely care for this complex patient group. The use of linked administrative health data to investigate PIBD risk factors has demonstrated no association between mode of delivery, gestational age, or exclusive formula feeding at 6 weeks with the development of PIBD. When investigating outcomes, cancer and mortality were found to be meaningful, serious early adult life adverse sequelae of PIBD. Ongoing utilisation and linkage of administrative health data will allow continued expansion of this epidemiological cohort and encourage further investigation of a diverse range of clinical questions.