Background & aims: The inflammatory bowel diseases (IBD), Crohn‟s disease (CD) and ulcerative colitis (UC), are common causes of chronic gastrointestinal morbidity, affecting up to 1 in 250 of the general population in Northern Europe. Up to 25% of IBD is diagnosed during childhood or adolescence. The aims for this thesis were to study the epidemiology, natural history and novel genetic determinants of childhood onset IBD in Scotland. Methods: The existing repository of childhood onset and adult onset IBD patients, established at the Western General Hospital in Edinburgh, was used and expanded. Thus, anatomical location and behaviour of disease were assessed in 416 childhood onset (276 CD, 99 UC, 41 IBDU diagnosed before 17th birthday) and 1297 adult patients (596 CD, 701 UC) using the Montreal classification. Additional phenotypic (at diagnosis and at regular follow-up intervals) and epidemiological data were gathered. In this cohort, genotyping of germline variants in putative susceptibility genes (NOD1/CARD4, IL23R, ATG16L1, IRGM, FLG) was performed to enable single variant and haplotype-tagging association studies. Genotypic data of population-matched healthy controls were obtained locally (n=342) and from the Wellcome Trust Case Control Consortium (n=2937). Results: Compared with adults, childhood-onset CD was characterized by a more extensive, “panenteric” phenotype (ileocolonic plus upper GI; p<0.0001 OR23.3; 95% CI (13.4–40.6) with less isolated ileal (p<0.0001 OR 0.06 (0.03–0.1) or colonic disease (p<0.0001, OR 0.3 (0.2–0.5)). In 39%, the anatomic extent increased within 2 years. UC was also more extensive in children at diagnosis vs adults (p<0.0001 OR 5.1 (2.7–9.4)). In population-matched and age, sex and postcode-matched case-control analysis, childhood onset IBD and CD was associated with asthma (p<0.0001 OR 1.7 (1.3-2.1) and (p=0.005 OR 2.5 (1.3-4.8), respectively). Inherited variation of NOD1/CARD4 was not a strong determinant of disease susceptibility in the Scottish population (both in single marker and haplotype-tagging studies, all p>0.05 after Bonferroni correction). We found that the allelic frequency of rs11209026*A located within the IL23R gene, differed significantly between IBD/CD cases and controls (p=0.01 OR 0.51(0.3-0.9) and p=0.04 OR 0.5 (0.3-0.98)). Using a gene-wide haplotype-tagging strategy, we demonstrated that the multiple association signals of the IL23R locus are independent of rs11209026 in childhood onset IBD and CD. In Scottish children, the effect of germline variation of ATG16L1 and IRGM on CD susceptibility was relatively small (OR< 1.4), and appeared less than in adult disease. Genotype–phenotype analysis demonstrated an association of pure ileal disease with the ATG16L1 rs2241880G-allele (p=0.02 OR 1.3 (1.03–1.7)). Using binary logistic regression analysis, we confirmed the effect of rs2241880 genotype (GG) on ileal disease versus colonic disease (p=0.03 OR 2.4 (1.05–5.6)). Null alleles of the epithelial barrier protein FLG have no important effect on IBD susceptibility (p>0.4), but contribute to the high prevalence of atopy, notably co-existent eczema and food allergy (p=0.0003 OR 3.3 (1.7–6.6) and p=0.0001 OR 4.5 (2.0–10.0), respectively). Conclusion: Childhood onset IBD is characterised by extensive intestinal involvement and progression of disease after diagnosis. Genetic association studies in childhood and adult IBD have provided evidence for a large number of new genomic loci. These loci encode genes involved in a number of homeostatic mechanisms: innate pattern recognition receptors, the differentiation of Th17-lymphocytes, autophagy, maintenance of epithelial barrier integrity and the orchestration of the secondary immune response.

The efficacy of multidisciplinary inpatient and outpatient treatment for transitioning children with severe pediatric feeding disorders from gastrostomy tube dependency to oral nutrition was investigated utilizing caloric and fluid intakes as an outcome measure. The study involved 29 children ages 12 months to 5 years of age with gastrostomy tube dependency. Treatments were provided by speech therapists, occupational therapist, dietician and psychologist for a 30 day period. Four treatment groups were evaluated and average intakes compared at 4 observation periods including pretreatment, initiation of treatment, completion of treatment at 30 days and 4 month follow-up. Children receiving inpatient treatment for feeding disorders evidenced significant differences in oral caloric intake from pretreatment to discharge than outpatient treatment (p < .01) and wait list control group (p = .04). Oral caloric intake from discharge to 4 month follow up yielded no significant differences indicating treatment gains were maintained. Change in environment and caretaker showed a significant effect for the inpatient group (d = 1.89). Effects of treatment by age and weight at 4 month follow up were also analyzed.

L'opinion générale est que les ulcères sont rares pendant l'enfance, les lésions provoquées par Helicobacter pylori (H. pylori) ne se produisant que des décennies après l'acquisition de l'infection. L’infection par cette bactérie est en outre moins fréquente chez les enfants dans les pays développés par rapport aux adultes. Par ailleurs, l’usage chronique de médicaments gastro-toxiques est peu fréquent dans cette tranche d’âge. Cependant, plusieurs études ont montré qu’environ 1/10 des enfants référés pour des symptômes de dyspepsie en Europe et infectés par H. pylori présentent un ulcère gastrique ou duodénal, mais aussi que la fréquence de ces lésions chez les enfants non infectés n’est pas nulle.<p>Afin de déterminer la fréquence des ulcères gastriques et duodénaux et des érosions, nous avons commencé par réaliser une étude prospective avec la participation de 19 centres répartis dans 14 pays d'Europe. Tous les enfants référés pour une endoscopie haute ont été recrutés durant une brève période de 1 mois. Parmi les 694 enfants inclus, 56 (8,1%) avaient soit des ulcères (ulcère gastrique 17/56, 30% - ulcère duodénal 7/56, 13%) soit des érosions (érosions gastriques 21/56, 37% - érosions duodénales 9/56, 16% - érosions gastriques et duodénales 2/56, 4%). Cette étude a permis de confirmer que la fréquence des lésions augmente avec l’âge, les enfants atteints de lésions étant significativement plus âgés que les témoins. En effet, les lésions ont surtout été observées chez les enfants dans la deuxième décade de vie. Une infection par H. pylori était présente seulement chez 15 des 56 enfants (27%), un médicament gastro-toxique avait été utilisé chez 13/56 (23%), une maladie inflammatoire chronique de l’intestin était présente chez 7/56 (13%) et une polyarthrite juvénile chez 2/56 (4%, plus d'un facteur de risque présent dans la plupart des cas). Aucun facteur de risque n’a pu être démontré chez 24/56 enfants (43%), une proportion beaucoup plus élevée que celle initialement attendue.<p>Nous avons ensuite réalisé une étude cas-témoins prospective et multicentrique (12 centres participants). Tous les patients avec une lésion érosive ou ulcérée de la muqueuse gastroduodénale ont été inclus avec deux témoins appariés pour l’âge, le centre et la période. Sept cent trente-deux patients (244 cas dont 153 avec seulement des érosions et 91 avec un ou des ulcères, 488 témoins) ont été inclus. Les enfants qui avaient reçu un antibiotique, un inhibiteur de la pompe à proton ou un anti-H2 durant les 4 semaines précédant l’endoscopie ont été exclus de l’analyse statistique parce que ces médicaments influencent la détermination<p>7<p>du statut H. pylori et la gravité des lésions (42 cas et 98 témoins). Nos résultats montrent que, chez les enfants, l'infection à H. pylori est un facteur de risque pour les ulcères duodénaux et les érosions duodénales, mais pas pour les lésions gastriques. Le sexe masculin, la consommation d'AINS, les maladies rénales chroniques et le tabagisme sont d'autres facteurs de risque indépendants de lésions érosives ou d’ulcères gastroduodénaux. Cependant, aucun facteur de risque identifiable n’a été retrouvé dans une grande proportion d'enfants (97/202, 48.0%) ce qui confirme les résultats de notre première étude.<p>Chez les adultes également la proportion d’ulcères sans infection à H. pylori et sans prise d’AINS est en augmentation ces dernières années tout en restant plus faible que chez l’enfant. La fréquence des ulcères gastriques et duodénaux avec un diamètre d’au moins 5 mm a été comparée, dans notre centre et dans un centre d’endoscopie adulte situé dans la même région de Bruxelles, sur une période de deux ans. Ces données montrent que les ulcères sont moins fréquents chez les enfants que chez les adultes (20/1279 enfants avec endoscopie haute - 1,6% vs adultes 58/1010 - 5,7%, OR 0,30, 95%CI 0,10-0.86, p = 0,02) et surtout moins fréquemment associés à une infection par H. pylori (8/20 vs 40/58, OR 0,26, 95%CI 0,16- 0.78, p <0,0001).<p>Comme l’activation de la réponse immunitaire locale est inefficace pour éliminer l’infection par H. pylori et serait plutôt impliquée dans la pathogenèse des lésions de la muqueuse, nous avons comparé la réponse immunitaire muqueuse des lymphocytes T et les réponses naïves chez les enfants et chez les adultes infectés par H. pylori ainsi que chez des témoins non infectés appariés pour l’âge.<p>Dans une première étude, nous avons obtenu des biopsies de la muqueuse antrale chez 43 patients dyspeptiques (12 enfants, 31 adultes). Les concentrations de cytokines libérées dans le milieu de culture et la densité de cellules CD3+, CD25+ et CD69+ ont été évaluées par cytométrie en flux. Le nombre de cellules sécrétant de l’interféron-γ (IFN-γ), de l’interleukine-4 (IL-4) et de l’IL-10 a été mesuré par ELISPOT. Les données obtenues montrent que l’augmentation de la sécrétion d'IFN-γ et l’élévation du nombre de cellules secrétant de l’IFN-γ au niveau de la muqueuse antrale lors d’une infection par H. pylori sont plus faibles chez les enfants que chez les adultes.<p>8<p>Dans une seconde étude, nous avons comparé l’infiltrat inflammatoire de la muqueuse antrale dans différents groupes d’âge (moins de 8 ans, 8 à 17 ans, 18 à 55 ans) de patients successifs infectés par H. pylori et des témoins appariés pour l’âge. Nous avons montré une corrélation entre l'âge et la densité de neutrophiles, de cellules CD3+ et de CD8+, mais pas de cellules CD20+. Le recrutement des neutrophiles dans la muqueuse antrale est plus faible chez les enfants et apparaît corrélé avec une plus faible activation du facteur de transcription NF-kB (déterminé par immunohistochimie et par EMSA) dans cette même muqueuse. L’infiltrat inflammatoire et l’activation du NF-kB sont légèrement (mais non significativement) plus intenses en cas d’infection par une souche plus virulente (facteur de virulence cagA). Ces souches cagA+ sont retrouvées en proportion équivalente dans les différents groupes d’âge. Par contre, la charge bactérienne, mesurée par un score semi-quantitatif en histologie, n’influence pas l’intensité de l’infiltrat inflammatoire.<p>En conclusion :H. pylori reste un facteur étiologique majeur pour les ulcères et les érosions duodénales chez l’enfant, mais pas pour les lésions gastriques dans les pays à faible prévalence de l'infection et la proportion de lésions associées à une infection est plus faible que chez les adultes. Aucun facteur d’exposition connu ne peut être associé aux lésions endoscopiques dans la moitié des cas, ce qui justifiera des études ultérieures pour identifier d’autres causes exogènes ou endogènes à ces lésions.<p>La réponse immunitaire de l’hôte est impliquée dans la pathogenèse des lésions gastroduodénales associées à une infection par H. pylori. Or il a été démontré dans les travaux faisant l’objet de cette thèse que cette réponse immunitaire est plus faible chez l’enfant que chez l’adulte pour certains facteurs (cytokines Th1, immunité humorale, recrutement des polynucléaires et des lymphocytes au niveau muqueux, activation du facteur de transcription NF-κB). D’autres études confirment la plus faible réponse humorale et Th1, mais également Th17 ainsi qu’une activation plus intense des Treg. Les cytokines ou les voies de signalisation responsables de cette réponse immunitaire plus faible restent inconnues, ce qui ouvre la voie à d’autres investigations. / Doctorat en Sciences médicales / info:eu-repo/semantics/nonPublished

PhD (Microbiology) / Department of Microbiology / Diarrhoea is still a major problem worldwide. Enteric pathogens such as Enteroaggregative E. coli (EAEC), Enteropathogenic E. coli (EPEC) and Enterotoxigenic E. coli (ETEC) have been reported to cause diarrhoea in children under the age of 5 years. The incidences of these pathogens are due to factors such as poor water quality, sanitation and hygiene practices. Infections with these pathogens result in diarrhoea and have been reported to result in severe disease outcomes more especially in children under 2 years of age. EPEC infections have been well studied using in vitro analyses, with studies highlighting the adherence traits, proteins and virulence genes involved in pathogenesis and inflammatory responses. EPEC is characterized by localized adherence with microcolony formation at the site of infection. In vivo studies have reported on human EPEC infection. However, the current animal models have not been able to replicate clinical outcomes (such as diarrhoea and weigh loss) of EPEC infection similar to humans. Therefore, there is still a need for a suitable small animal model that mimic clinical outcomes of human EPEC infections in vivo. Children living in poor environmental conditions are more susceptible to diarrhoeal pathogens. Furthermore, the incidences of children being exposed to co-infections (more than one pathogen at the same time) is relatively high. The EAEC/EPEC (A/P) and EPEC/ETEC (P/T) co-infections have been increasingly detected in children with and without diarrhoea. It has been suggested that patients infected with these co-infections might result in severe disease outcome than those infected with single pathogens. Pathogens are constantly evolving and the microbe-microbe interaction in the host can result in these pathogens competing for the same niche and thus result in increased virulence. Interaction of co-infections can lead to increased inflammatory responses thus affecting the infected host. The first objective of this study was to develop an EPEC murine model using weaned C57BL/6 mice that have been pretreated with antibiotic cocktail. Mice were orally infected with wild-type (WT) typical EPEC, bfp- and escN mutant strains. The WT had transient weight loss and wet stools with mucous; and the bfp- infected mice also had transient weight loss and bloody stool appearance. Increase in inflammatory biomarkers MPO, LCN-2, CRP, IL-6 and SAA were observed in the WT and bfp- infected mice. The mice infected with escN mutant did not exhibit any weight changes and the stools were similar to the uninfected mice. Furthermore, no inflammatory biomarkers were observed in mice infected with the escN mutant. Metabolic perturbations were observed in WT EPEC infected mice at day 3 post infection with the TCA cycle metabolites (reduced succinate, citrate, fumarate, cis-aconitate) being excreted at lower quantities indicating that the energy production in the infected mice was greatly affected. The second objective of this study was to determine the interaction between the P/T coinfections using in vitro and in vivo analyses. In vitro, human colorectal tumour 8 (HCT-8) cells were infected with single strains of ETEC, EPEC and both the pathogens and incubated for 3 hours. After infection the cells were analysed for bacterial adherence using real-time PCR. The single strains adhered at the same rate similar to the P/T coinfected cells. IL-8, as a marker of inflammatory response, was measured using ELISA. The results indicated that the P/T co-infected cells had a significant increase in IL-8 response higher than the single infections. The P/T co-infections were further analysed in vivo using the EPEC murine model developed in this study. Interestingly, mice infected with P/T co-infections developed severe diarrhoea accompanied with significant increased weight loss and some mice died during the 3-day infection period. The inflammatory responses MPO, LCN-2 and SAA were higher in the co-infected mice indicating a synergistic effect. The bfp and eltA virulence genes were significantly increased in the P/T co-infections. The third objective of this study was to determine the interaction between A/P coinfections using in vitro and in vivo analyses. HeLa cells and HCT-8 cells were infected with EAEC, EPEC and both the pathogens at the same time in order to determine adherence and inflammatory responses. EAEC adherence was higher than EPEC and A/P co-infections adherence. A/P co-infections did not have increased IL-8 response in HCT-8 cells when compared to EAEC alone. The virulence genes involved in EPEC adherence and Type 3 Secretion System (bfp, eae, tir, ler, per, espB and espA) were significantly reduced in A/P co-infected cells. An interesting adherence trait was observed between the A/P co-infections in HeLA cells, EAEC was found to adhere around EPEC altering the localized adherence pattern. The A/P co-infections were further analysed using the EPEC murine model developed in this study. The A/P infected mice had diminished weight changes and EAEC shedding was enhanced when EPEC was present. Faecal inflammatory biomarkers MPO and LCN-2 in A/P infected mice did not have any additive effect. The findings of this study contributed significantly to the knowledge of human EPEC infection in weaned C57BL/6 mice, highlighting clinical outcomes, inflammatory responses and metabolic perturbations. Furthermore, this study also highlighted the interaction of P/T and A/P co-infections using in vitro and in vivo analyses in order to determine the disease severity and outcomes. It was observed in this study that coinfections can result in either synergistic or antagonistic effects. Further studies are therefore, required in order to understand the underlying mechanisms that are involved during co-infections and this can further assist in the development of therapeutic interventions. / NRF

Aims An increasing number of children and adolescents worldwide suffer from inflammatory bowel disease (IBD) such as Crohn’s disease (CD) and ulcerative colitis (UC). The present work aims to investigate the incidence, prevalence and future trends of IBD in children and adolescents in Saxony, Germany. Methods The Saxon Pediatric IBD Registry collected data on patients up to 15 years of age from all 31 pediatric hospitals and pediatric gastroenterologists in Saxony over a 15-year period (2000–2014). In 2019, an independent survey estimated a registry completeness of 95.7%. Age-standardized incidence rates (ASR) per 100,000 person-years (PY) and prevalence per 100,000 children and adolescents were calculated. Evaluation was also been performed in sex and age subgroups. Joinpoint and Poisson regression were used for trend analyses and projections. Results 532 patients with confirmed IBD during 2000–2014 were included in the epidemiological evaluation. 63.5% (n = 338) patients had CD, 33.1% (n = 176) had UC and 3.4% (n = 18) had unclassified IBD (IBD-U). The 15-year IBD prevalence was 111.8 [95%-CI: 102.3–121.3] per 100,000. The incidence ASR of IBD per 100,000 PY over the whole observation period was 7.5 [6.9–8.1]. ASR for the subtypes were 4.8 [4.3–5.3] for CD, 2.5 [2.1–2.9] for UC and 0.3 [0.1–0.4] for IBD-U. The trend analysis of ASR using the joinpoint regression confirmed a significant increase for incidence of IBD as well as CD. For IBD, the ASR per 100,000 PY increased from 4.6 [2.8–6.3] in 2000 to 8.2 [7.5–13.6] in 2014; projected incidence rates for IBD in Germany are 12.9 [6.5–25.5] in the year 2025 and 14.9 [6.7–32.8] in 2030, respectively. Thus, the number of new IBD diagnoses in Germany would more than triple (325%) in 2030 compared to 2000. The increase is expected to be faster in CD than UC, and be more in males than in females. The expected number of newly diagnosed children with IBD in Germany is projected to rise to about 1,584 [1,512–1,655] in 2025, and to about 1,918 [1,807–2,29] in 2030. Conclusion The incidence of IBD in children and adolescents in Saxony increased at a similar rate as in other developed countries during the observation period. Given this trend, the health care system must provide adequate resources for the care of these young patients in the future.