Single-cell map links childhood Crohn’s to treatment response

A biopsy can hold more than a diagnosis. In the small intestine of 14 untreated children with Crohn’s disease, researchers mapped the activity of individual cells and found patterns associated with disease severity and later response to anti-TNFα treatment. The work, led by scientists including Kyle Kimler and Leslie Kean at Boston Children’s Hospital, produced a cellular atlas containing 94,451 cells.

The team used single-cell RNA sequencing, a technique that measures gene activity cell by cell, rather than averaging signals across an entire tissue sample. It also developed ARBOL, a publicly available bioinformatics tool that repeatedly splits data into smaller groups to build trees of cellular states. The researchers compared the Crohn’s samples with 13 biopsies from children with noninflammatory gastrointestinal disorders and examined repeat biopsies from eight Crohn’s patients after treatment.

The pattern was not one rogue cell type. More severe disease showed increases in pro-inflammatory immune cells, including T cells, cytotoxic natural killer cells, and subsets of monocytes and macrophages. At the same time, metabolically specialized cells lining the normal intestine became less prominent. These diagnostic cell patterns were linked to the clinical severity that helped determine whether children received anti-TNFα agents, and to whether treatment produced a complete or partial response.

The atlas also points to a warning about treatment resistance. When the pediatric data were combined with a single-cell atlas of adults with Crohn’s disease receiving treatment, anti-TNF treatment was associated with a shift in the childhood cellular ecosystem toward the more severe, treatment-resistant state often found in adults. That finding describes a cellular pattern; it does not establish that the treatment causes resistance.

For children and clinicians, the practical promise is earlier sorting of risk: a biopsy taken at diagnosis could eventually help indicate who may need anti-TNF treatment, who may respond only partly, and who may be unlikely to benefit. The present study does not establish a clinical test or a treatment decision rule. Its untreated cohort was limited by strict enrollment criteria, follow-up length and the difficulty of capturing patients early, and represented the demographics of inflammatory and functional gastrointestinal disorders in Seattle. The authors also could not directly connect the results to genetic or environmental factors such as the gut microbiome, infections or diet.