Large-scale scanning of serum proteins offers a potential method for noninvasive screening and monitoring of patients with nonalcoholic steatohepatitis (NASH), the technique’s developers claim.
By scanning for about 5,000 proteins in nearly 3,000 samples from patients enrolled in studies from the Clinical Research Network in NASH (NASH CRN), Rachel Ostroff, PhD, and colleagues from SomaLogic in Boulder, Colo., created four protein models that mimic results of the major pathologic findings in liver tissue biopsy.
“Concurrent positive results from the protein models had performance characteristics of ‘rule-out’ tests for pathologists’ diagnosis of NASH. These tests may assist in new drug development and medical intervention decisions,” they wrote in a late-breaking poster presented at the virtual annual meeting of the American Association for the Study of Liver Diseases.
“There is no single noninvasive method that can accurately and simultaneously capture steatosis, inflammation, hepatocyte ballooning and fibrosis, the four major pathologic components assessed by biopsy. Each of these is relevant to the multiple mechanisms targeted in drug development for NASH,” they wrote.
To see whether large-scale protemoics could serve as an alternative to invasive liver biopsy for use in clinical trials or in longitudinal studies of NASH, they used a modified aptamer proteomics platform to scan for liver-related proteins. Aptamers are olignonucleotide or peptide molecules designed to home in on a specific target.
They scanned for approximately 5,000 proteins in 2,852 serum samples from 638 patients in a NASH CRN natural history cohort, and in patients enrolled in two NASH treatment trials: the, which is evaluating pioglitazone versus vitamin E and placebo in nondiabetic patients, and the , which is comparing obeticholic acid with placebo. All of the patients in the natural history cohort and half of all patients in the clinical trial cohorts were included in the training sets, with the remaining half included in the validation set.
The accuracy of the models, as measured by the area under the curve (AUC) of receiver operating characteristics in the training and validation sets, respectively, were as follows:
- Fibrosis: AUC 0.92/0.85.
- Steatosis: AUC 0.95/0.79.
- Inflammation: AUC 0.83/0.72.
- Hepatocyte Ballooning: AUC 0.87/0.83.
“A concurrent positive score for steatosis, inflammation and ballooning predicted the biopsy diagnosis of NASH with an accuracy of 73%,” Ostroff and colleagues wrote.
They also found that model scores applied over time showed improvements in symptoms in the patients on active therapies in the clinical trials, compared with patients on placebo.
A specialist in liver pathology and nonalcoholic fatty liver disease who was not involved in the study said in an interview that she finds the results highly promising.
Elizabeth M. Brunt, MD, emeritus professor of pathology and immunology at Washington University in St. Louis, was a member of the NASH CRN when SomaLogic first proposed using the groups’ data for this study.
“I was impressed with them then, and I am very impressed with what they’re presenting here, and I can’t say that about all the noninvasive tests,” she said. “I think a lot of noninvasive tests are way over-simplifying what NASH is.”
She acknowledged that, although she spent much of her career performing liver biopsies, “you can’t biopsy every single patients who you suspect of having NASH, or certainly if you want to follow them over time – it’s unrealistic,” she said in an interview.
Although the protein scanning method cannot – and is not intended to – replace a well-conducted biopsy with the interpretation of a skilled pathologist, the proteins the company investigators identified can reflect the dynamic nature of liver disease and the liver’s ability to heal itself with a high degree of accuracy and hold promise for both screening patients and for monitoring responses to therapy, Dr. Brunt said.
The study was sponsored by SomaLogic. The authors are employees of the company. Dr. Brunt had no relevant disclosures.
SOURCE: Ostroff R et al. The Liver Disease Meeting Digital Experience,