Blood-Based Biomarkers Prove Mettle in Liver Disease Diagnosis

Multiple noninvasive blood biomarkers for metabolic liver disease met accuracy assessments for diagnosing metabolic liver disease in the first step of an advanced qualification system to better utilize these noninvasive tools, a researcher said.

Top-line analyses of the Non-Invasive Biomarkers of Metabolic Liver Disease (NIMBLE) Stage 1 initiative found high sensitivity and specificity for the NIS-4, One-Way Lipidomics (OWLiver) Panel, Enhanced Liver Fibrosis (ELF) test, and PRO-C3 and FibroMeter-VCTE, reported Arun Sanyal, MD, of VCU School of Medicine in Richmond, Virginia.

“With the totality of all this data, we have made substantial progress towards biomarker qualification for a diagnostic for people who have NASH [non-alcoholic steatohepatitis] or significant fibrosis,” Sanyal said at a press conference at the American Association for the Study of Liver Diseases (AASLD) virtual meeting.

He explained how NIMBLE is a public-private initiative with the Foundation-NIH that hopes to fill in the gaps in scientific literature “on noninvasive tools that prevent their regulatory evaluation as diagnostic tests,” he said.

“The biggest implications for having noninvasive tests is every physician in their office can mark it on their lab sheet or click it in their Epic system” because it is a “blood-based test,” Sanyal told MedPage Today.

Stage 1 of NIMBLE involved evaluating the histology of the performance of five blood-based biomarker panels. The NIMBLE initiative obtained blood samples available within 90 days of liver biopsy from adults with histological evidence of non-alcoholic fatty liver disease (NAFLD) of varying severity. These included samples from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) biological repository, as well as various individual laboratories, Sanyal said.

The primary hypothesis was that the selected biomarkers would have at least a diagnostic accuracy of an area under the receiver operating curve (AUROC) of 0.7, and be statistically superior to an AUROC of 0.5 for their specific “fit for purpose” use.

They examined samples from 1,073 patients from stage 0 to stage 4 of fibrosis, selected for “a relatively even distribution across stages” in order to avoid spectrum bias, Sanyal noted. Patients were a mean age of 48-56, 30%-47% were men and 60%-86% were white. Not surprisingly, only 12% of patients in stage 0 had NASH versus 80%-100% in stages 1-4.

The following biomarkers fulfilled the primary hypothesis:

• NIS-4 for diagnosis of “at risk” NASH, NAS ≥4 and fibrosis stage 2 or higher
• ELF test for diagnosis of clinically relevant fibrosis
• PRO-C3 for diagnosis of clinically relevant fibrosis
• FibroMeter-VCTE for diagnosis of clinically relevant fibrosis

Sanyal noted that the OWLTest only had a yes/no score based on the probability of having NASH or NAS ≥4, and therefore could not generate an AUROC. However, the test identified “at risk” NASH with a sensitivity of 63.3% and a specificity of 75.4%.

He also discussed differential performance of certain diagnostics, such as the fact that ELF test performance improved progressively for diagnosis of more advanced fibrosis, while PRO-C3 declined when used to detect progressively advanced fibrosis stages, adding that “future combinatorial approaches could be used to further enhance diagnostic precision.”

In terms of next steps, Sanyal said these data set the stage for the next phase of NIMBLE, which involves identifying predictive values for these tests.

“First, we need to get a robust handle on sensitivity and specificity … but what really matters is predictive values. That is Stage 2, where we go to the intended use populations, knowing what the sensitivity and specificity is in a very rigorous way so that we can then attempt to identify predictive values.”

Planned secondary analyses include examining patients with type 2 diabetes. Sanyal told MedPage Today that since most of the patients who will die with liver disease have type 2 diabetes, the diabetes population should have an increased awareness of liver disease.

“It’s time to include liver as an organ associated with type 2 diabetes,” he said.