Back

Discovery of a simple two-gene expression biomarker in whole blood predictive of the need for treatment escalation in inflammatory bowel disease

Nowak, J. K.; Kalla, R.; Adams, A. T.; IBD Character Consortium, ; Halfvarson, J.; Satsangi, J.

2021-07-10 gastroenterology
10.1101/2021.07.09.21259804 medRxiv
Show abstract

Background and aimsThe IBD-Character consortium has recruited large internationally based inception cohorts of treatment-naive inflammatory bowel disease (IBD) patients, providing a unique resource to derive a simple transcriptome signature in the field of prognostication. MethodsThe discovery cohort (n=160) was recruited in Norway, Sweden and Spain. The replication inception cohort from the United Kingdom (n=97) was followed-up for a mean ({+/-}SD) of 350{+/-}228 days. Treatment escalation was formally defined as the need for a biologic agent, ciclosporin and/or surgery, instituted for disease flare after initial remission, or colectomy during the index admission for ulcerative colitis. Whole blood RNA was subject to paired-end sequencing. In the discovery cohort a simple procedure was applied, which exploited differences of transcript ratios. The ten top performing ratios were tested using Cox regression models in the validation cohort. ResultsNewly diagnosed IBD patients with high CACNA1E/LRRC42 expression ratio had an increased risk of treatment intensification (validation cohort: HR=19.3, 95%CI 2.6-143.9, p=0.000005; AUC 0.76, 95%CI 0.66-0.86). In 51 patients with CRP < 3.5 mg/L, CACNA1E/LRRC42 still predicted escalation (HR=10.4; 95%CI 1.2-86.5, p=0.007). The second best performing transcript ratio was CACNA1E/CEACAM21 yielding a HR of 10.9 (95%CI 2.5-46.7, p=0.00002) and an AUC of 0.76 (95%CI 0.65-0.86) in the validation cohort. ConclusionTranscriptomic profiling of an IBD inception cohort identified gene expression ratios CACNA1E/LRRC42 and CACNA1E/CEACAM21 as prognostic biomarkers. These were validated in a replication cohort as strongly associated with short- and long-term risk of treatment intensification and may provide valuable information in clinical decision-making.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.