Introduction

Ustekinumab (UST), a human monoclonal antibody against interleukin-12 and 23, is approved to treat adult patients with psoriasis or Crohn disease (CD). Outcomes data for off-label use in pediatric patients with CD are limited.

Aim

We conducted a retrospective cohort study to analyze the long-term efficacy of UST, including dose adjustments, in the treatment of pediatric patients with medically refractory CD. Adverse events were documented.

Methods

We identified 40 pediatric patients with CD treated with UST between January 1, 2016 and December 31, 2019. Electronic medical records were reviewed for demographics, Paris Classification, significant comorbidities, previous CD therapy, adverse events after initiation, and surveillance markers at the time of their first dose and most recent clinic visit. A validated abbreviated pediatric CD activity index (aPCDAI) was used to assess response to therapy.

Results

Thirty-eight pediatric patients with CD, including 34.2% with stricturing or penetrating disease, were analyzed after initiation of treatment with UST. Median age at diagnosis of CD was 12.5 years, and median age at UST induction was 17.2 years. No patients were anti-TNF-naive, and 34.2% were previously exposed to 2 or more anti-TNF agents. At time of last follow-up, 84.2% of patients remained on UST for a median duration on UST of 62.1 weeks, and 60.5% achieved clinical remission. Patients had significant improvement in aPCDAI scores, clinical remission rates, albumin, and hematocrit, and 89.5% of patients had no significant adverse events. Similar results were observed among those who required dose adjustment, including 61.1% achieving clinical remission, and among those with perianal disease, including 38.5% achieving clinical remission.

Conclusions

Our data suggest that, within our cohort of pediatric patients with CD, UST has long-term efficacy with no observed safety concerns. Dose adjustment may be helpful in achieving clinical remission.

Recent evidence suggests that hormones such as insulin and leptin act in the hypothalamus to regulate energy balance and glucose metabolism. Here we show that in leptin receptor-deficient Koletsky (fa(k)/fa(k)) rats, adenovirally induced expression of leptin receptors in the area of the hypothalamic arcuate nucleus improved peripheral insulin sensitivity via enhanced suppression of hepatic glucose production, with no change of insulin-stimulated glucose uptake or disposal. This effect was associated with increased insulin signal transduction via phosphatidylinositol-3-OH kinase (as measured by pY-insulin receptor substrate-1 and pS-PKB/Akt) in liver, but not skeletal muscle, and with reduced hepatic expression of the gluconeogenic genes, glucose-6-phosphatase and phosphoenolpyruvate kinase. Moreover, the beneficial effects of hypothalamic leptin signaling on hepatic insulin sensitivity were blocked by selective hepatic vagotomy. We conclude that hypothalamic leptin action increases peripheral insulin sensitivity primarily via effects on the liver and that the mechanism underlying this effect is dependent on the hepatic branch of the vagus nerve.

A20 is an anti-inflammatory protein that is strongly linked to human disease. Here, we find that mice expressing three distinct targeted mutations of A20's zinc finger 7 (ZF7) ubiquitin-binding motif uniformly developed digit arthritis with features common to psoriatic arthritis, while mice expressing point mutations in A20's OTU or ZF4 motifs did not exhibit this phenotype. Arthritis in A20^ZF7 mice required T cells and MyD88, was exquisitely sensitive to tumor necrosis factor and interleukin-17A, and persisted in germ-free conditions. A20^ZF7 cells exhibited prolonged IκB kinase activity that drove exaggerated transcription of late-phase nuclear factor-κB response genes in vitro and in prediseased mouse paws in vivo. In addition, mice expressing double-mutant A20 proteins in A20's ZF4 and ZF7 motifs died perinatally with multi-organ inflammation. Therefore, A20's ZF4 and ZF7 motifs synergistically prevent inflammatory disease in a non-catalytic manner.

Alternative splicing (AS) allows increased diversity and orthogonal regulation of the transcriptional products of mammalian genomes. To assess the distribution and variation of alternative splicing across cell lineages of the immune system, we comprehensively analyzed RNA sequencing and microarray data generated by the Immunological Genome Project Consortium. AS is pervasive: 60% of genes showed frequent AS isoforms in T or B lymphocytes, with 7,599 previously unreported isoforms. Distinct cell specificity was observed, with differential exon skipping in 5% of genes otherwise coexpressed in both B and T cells. The distribution of isoforms was mostly all or none, suggesting on/off switching as a frequent mode of AS regulation in lymphocytes. From the identification of differential exon use in the microarray data, clustering of exon inclusion/exclusion patterns across all Immunological Genome Project cell types showed that ∼70% of AS exons are distributed along a common pattern linked to lineage differentiation and cell cycling. Other AS events distinguished myeloid from lymphoid cells or affected only a small set of exons without clear lineage specificity (e.g., Ptprc). Computational analysis predicted specific associations between AS exons and splicing regulators, which were verified by detection of the hnRPLL/Ptprc connection.

To determine the breadth and underpinning of changes in immunocyte gene expression due to genetic variation in mice, we performed, as part of the Immunological Genome Project, gene expression profiling for CD4(+) T cells and neutrophils purified from 39 inbred strains of the Mouse Phenome Database. Considering both cell types, a large number of transcripts showed significant variation across the inbred strains, with 22% of the transcriptome varying by 2-fold or more. These included 119 loci with apparent complete loss of function, where the corresponding transcript was not expressed in some of the strains, representing a useful resource of "natural knockouts." We identified 1222 cis-expression quantitative trait loci (cis-eQTL) that control some of this variation. Most (60%) cis-eQTLs were shared between T cells and neutrophils, but a significant portion uniquely impacted one of the cell types, suggesting cell type-specific regulatory mechanisms. Using a conditional regression algorithm, we predicted regulatory interactions between transcription factors and potential targets, and we demonstrated that these predictions overlap with regulatory interactions inferred from transcriptional changes during immunocyte differentiation. Finally, comparison of these and parallel data from CD4(+) T cells of healthy humans demonstrated intriguing similarities in variability of a gene's expression: the most variable genes tended to be the same in both species, and there was an overlap in genes subject to strong cis-acting genetic variants. We speculate that this "conservation of variation" reflects a differential constraint on intraspecies variation in expression levels of different genes, either through lower pressure for some genes, or by favoring variability for others.