Serum-based biomarker test yields positive diagnostic accuracy for Crohn’s disease
SAN DIEGO — The Mucosal Healing Index, a serum-based biomarker assay, demonstrated efficacy in identifying mucosal inflammation and similar accuracy to fecal calprotectin in patients with Crohn’s disease, according to a study presented at Digestive Disease Week.
“Biomarker-based management has been shown to improve outcomes [in Crohn’s disease],” Robert Battat, MD, told Healio Gastroenterology.
Battat performed this research as an IBD fellow at the University of California San Diego; he is currently an assistant professor of medicine at Weill Cornell Medicine/New York Presbyterian Hospital and the Jill Roberts Center for Inflammatory Bowel Disease.
“Currently, the only biomarkers that are being used in clinical practice are CRP, which has issues with test characteristics, and fecal calprotectin, for which data shows that there is poor uptake by physicians and patients.”
The 13-biomarker serum-based Mucosal Healing Index (MHI) is a new test that intends to identify mucosal inflammation in patients with Crohn’s disease, Battat said.
Battat and colleagues conducted a study to determine the diagnostic accuracy of the MHI for endoscopic and histologic outcomes among 211 patients with Crohn’s disease with a total of 290 endoscopic assessments. They collected clinical, endoscopic and histologic data on serum, stool and blinded disease activity from a single-center biobank. The area under the receiving operating curve (AUROC) served as a marker of the MHI’s diagnostic accuracy to determine endoscopic remission and mucosal healing. The diagnostic accuracy was compared with fecal calprotectin in patients with stool samples.
Responsiveness, as measured by Spearman’s correlations, examined for changes in MHI, fecal calprotectin and symptoms in comparison with changes in the Simple Endoscopic Score for Crohn’s Disease (SES-CD).
Diagnostic misclassification rates were elevated with symptom-based guides.
The researchers found that AUROC estimations for MHI evaluating endoscopic remission was 0.7 (95% CI, 0.64-0.76). The AUROC for MHI evaluating mucosal healing (defined as achieving endoscopic and histologic remission) was 0.69 (95% CI, 0.59-0.79).
The MHI diagnostic accuracy for endoscopic remission was comparable among patients with isolated colonic disease and those with small bowel involvement (area under the curve, 0.7 vs. 0.69; P = .88). However, patients with isolated colonic disease had numerically higher MHI diagnostic accuracy for mucosal healing than those with small bowel involvement (AUC, 0.83 vs. 0.65).
There were similar AUC approximations for MHI when controlling for the following conditions: current treatment with biologic therapy, prior bowel surgery, assessing only samples of serum that were attained on the day of endoscopic evaluation, discounting analyses that showed endoscopic but not histologic inflammation and using single assessments per patient.
At an MHI cutoff of above 50, there was 100% specificity for mucosal inflammation that was found on endoscopy and histology and 91% specificity for inflammation found on endoscopy only (positive likelihood ratio, 3.08; 95% CI, 1.67-5.68).
At an MHI cutoff of 20, there was 83% sensitivity for the exclusion of mucosal inflammation on both endoscopy and histology (negative likelihood ratio, 0.42; 95% CI, 0.24-0.75) and endoscopy alone (negative likelihood ratio, 0.49; 95% CI, 0.32-0.74).
Therefore, a cutoff above 50 was effective at ruling in active disease and a cutoff below 20 was effective at ruling out active disease, Battat said. There was a gray area between a cutoff of 21 and 49, he added.
There was no significant difference in diagnostic accuracy between MHI and fecal calprotectin for endoscopic remission (AUROC, 0.78 vs. 0.84) and mucosal healing (AUROC, 0.76 vs. 0.8) when using matched samples (119 assessments). Comparable findings were observed in sensitivity analyses.
In a subset of 51 patients who had more than one visit and with unique SES-CD scores, changes in the MHI was associated with changes in SES-CD (r = 0.34; P = .01). However, changes in SES-CD scores were not associated with changes in symptoms (r = 0.18) or fecal calprotectin (r = 0.17) in these patients.
“We have started using [MHI] in our routine practice,” Battat said. “We found that patients prefer to have a serum-based test that assesses disease activity. Although this may reduce the frequency of endoscopies used for disease activity assessments, endoscopy is still an essential component of our practice.”
The assay is available for use now, he said, although research is needed to confirm these findings. – by Alaina Tedesco
Battat R, et al. Su1907. Presented at: Digestive Disease Week; May 18-21, 2019; San Diego.
Disclosure: Battat reports no relevant financial disclosures. Please see the abstract for all other authors’ relevant financial disclosures.