Prescribing a biologic for people with Crohn’s disease is a complicated process that includes consideration of previous therapy, the severity of disease, cost, and other factors. Missing, however, has been the ability to accurately predict endoscopic response to a specific biologic agent to guide choice of therapy.
Although the biomarker panels are not yet clinically available, researchers demonstrated that they are accurate, valid, stable over time, and largely specific to each of the three biologic agents.
“Evidence over the last 10 years has shown a consistent difference in DNA methylation between people with IBD [inflammatory bowel disease] and healthy controls. Many of these studies suggest a role for DNA methylation for treatment response prediction,” Vincent Joustra, PhD, said when presenting results of the EPIC-CD trial at the annual congress of the European Crohn’s and Colitis Organisation.
After comparing endoscopic responders to nonresponders in different datasets. researchers found that “DNA methylation profiles are, in fact, associated with response to adalimumab, vedolizumab, and ustekinumab,” added Dr. Joustra, visiting fellow in the department of gastroenterology and hepatology at Amsterdam University Medical Centers.
DNA methylation – the presence or absence of a methyl group on a specific DNA location called a CpG – does not change a person’s genotype. Rather, the methylation process either activates or deactivates a gene’s expression. It can be used to predict treatment response.
Within the past 2 decades, “biologics have revolutionized care of IBD patients. Yet, despite their clinical efficacy, treatment choice is currently based on trial and error, which is suboptimal,” Dr. Joustra said.
Adding biomarkers to improve biologic medication selection is “urgently needed,” he added. “However, such biomarkers are not available for practice today.”
Methylation methodology
Dr. Joustra and colleagues prospectively studied DNA methylation in the peripheral blood samples of 184 adults with Crohn’s disease. They compared the biomarkers at baseline in people set to start biologic therapy and again at a median of 28 weeks following treatment with adalimumab (58 patients), vedolizumab (64 patients), and ustekinumab (62 patients).
Participants were divided into a discovery cohort to identify relevant biomarkers and a validation cohort to confirm the findings. Results were validated against a separate cohort of patients at Oxford (England) University.
Response was strictly defined as a decrease of at least 50% in a simple endoscopic score for Crohn’s disease, corticosteroid-free clinical response or remission using the Harvey Bradshaw Index, and/or biochemical response or remission.
Before patients were treated, the investigators created three epigenetic panels. The CpG loci of interest were identified using the Infinium MethylationEPIC BeadChip array, which measures over 850,000 CpG sites across the whole genome.
Key findings
One epigenetic panel featured 100 CpG loci relevant for adalimumab that correlated to an “endoscopic response with high accuracy,” with an area under the curve of 0.73 upon validation. A second panel, created for vedolizumab, included 22 CpG loci and had an AUC accuracy of 0.89. The third panel, specific to ustekinumab, had 68 CpG loci and an AUC accuracy of 0.94.
The markers are largely unique to each agent. Only two CpG loci overlapped between adalimumab and ustekinumab, Dr. Joustra said.
“Importantly, our model was able to predict response prior to treatment in a completely different set of patients from the Oxford validation cohort with an AUC of 0.75,” Dr. Joustra said.
A secondary analysis revealed no differences in the stability and robustness of the methylation markers between baseline and 28 weeks. This finding implies that the biomarkers are stable during the induction and maintenance phases of treatment.
“Of course, we need to clinically validate our findings in a clinical trial, which is ongoing,” Dr. Joustra said. This work will continue in the EPIC-CD study, as well as in the OMICROHN clinical trial.
Promising start
“These are really interesting findings that address an area of importance in treating patients with Crohn’s disease,” said ECCO session comoderator Tim Raine, PhD, who was not affiliated with the research.
“The team found a signature that appears to provide helpful prediction of response to specific treatments. Importantly, this signature appeared to be stable over time, to be specific to individual drugs, and could be validated in an external cohort of patients,” added Dr. Raine, consultant gastroenterologist at Cambridge (England) University Hospitals NHS Trust.
Although the technologies used in EPIC-CD are not yet routinely available in clinical practice, “the methodologies are well established, and with appropriate development in a validated laboratory, as well as further validation work, could form a useful test for gastroenterologists treating patients with Crohn’s disease,” Dr. Raine said.
The study was independently supported. Dr. Joustra and Dr. Raine reported no relevant financial relationships.
A version of this article first appeared on Medscape.com.