IBD De-escalation


The goal of ulcerative colitis (UC) care is to achieve and maintain remission, defined as the normalization of inflammatory markers, symptoms, and endoscopic inflammation. Lifelong treatment is often advised to maintain remission, since stopping treatment leads to high rates of recurrence. However, it is also evident in practice that some patients can reduce (‘de-escalate’) or even stop treatment with durable long-term remission. There is strong patient interest in de-escalation and given the high cost of medications, there are substantial patient and societal economic cots that could be relieved with successful de-escalation strategies. Herein arises a key clinical question, who can de-escalate and how do we assess their risk for flare? Our inability to answer these questions has limited de-escalation strategies.


  • Jonathan Golob
  • Arvind Rao
  • Hongzhe Lee
  • Shrinivas Bishu


We are employing a multi-omics approach to establish the mechanisms by which the gut microbiome combined with or independent of the transcriptional state of the host epithelium distinguishes between active versus controlled UC, as well as if these same markers can predict an imminent flare. To do so we are combining previously published data on the microbiome in UC with a well-curated SPARC-IBD dataset.

Findings so far

This a recently launched project, but our preliminary efforts have successfully tokenized all of the trancsriptional and microbiome data, harmonized a large complementary set of data from prior studies, and generated preliminary predictive models that are reproducing across datasets to distinguish between active and quiescent UC, or healthy controls.

Next steps

We plan to combine this retrospective analysis (enabled by software developed by the Golob Lab) with a planned prospective study of UC patients, getting closer to our long-term goal of developing a reliable biomarker to support de-escalation of biologic therapy in controlled UC.

Jonathan Golob, M.D., Ph.D.
Assistant Professor

A physician scientist interested in applying microbiome-trained AI/ML models to improve human hea