WASHINGTON – Researchers have identified six new biomarkers of drug-induced liver injury (DILI) that, when combined with traditional measurements, seemed to better predict the disease course, compared with traditional biomarkers alone, according to a presentation at the annual meeting of the American Association for the Study of Liver Diseases.
In addition, some of these biomarkers may provide a “liquid biopsy” to assess degree of inflammation and mode of hepatocyte death, said Rachel Church, PhD, of the University of North Carolina, Chapel Hill.
“The motivation behind this research is that the standard biomarkers for DILI have several shortcomings,” Dr. Church said. “They’re not entirely liver specific, they’re not mechanistically informative, and they’re not sufficiently predictive of outcome.”
The researchers found that elevated levels of these six candidate biomarkers were predictive for adverse outcome in DILI: total keratin18 (K18); caspase-cleaved K18 (ccK18); alpha-fetoprotein (AFP); osteopontin (OPN); fatty acid–binding protein 1 (FABP1); and macrophage colony-stimulating factor receptor (MCSFR) determined by immunoassay. “We believe that using some of these candidate biomarkers in combination with the standard tests may be the best way to identify individuals at risk for an adverse outcome,” Dr. Church said.
While their analysis found that the traditional international normalized ratio had the overall best predictive value, measured as area under the curve (AUC) of 0.922, the candidate biomarker OPN was second best with an AUC of 0.871, “and actually performed better than total bilirubin,” Dr. Church said.
The study evaluated mechanistic candidate biomarkers by obtaining biopsies in a cohort of 27 patients within 2 weeks of diagnosis, focusing on three physiological reactions: inflammation, necrosis, and apoptosis.
With regard to inflammation, Dr. Church said, “What we found was that MCSFR actually was significantly elevated in patients who had a high score for inflammation; however, there was no significant difference in OPN, although there was a slight elevation.”
They evaluated necrosis using a semiquantitative confluent coagulative necrosis score, and found no difference in the typical biomarkers of cell necrosis, such as alanine transminase, aspartate aminotransferase, and K18. “So we also looked at the regenerative biomarkers, OPN and AFP, and indeed, we observed that both were significantly elevated with high confluent coagulative necrosis scores,” she said.
To evaluate apoptosis, the researchers used the semiquantitative apoptosis score. “We found there was a small but significant elevation in ccK18 in individuals with a high apoptosis score,” she said. They then evaluated the ratio of ccK18 to K18. “The closer the score is to 1, the more apoptosis you have; and the closer the score is to 0, the more necrosis you have,” Dr. Church said.
They also developed a predictive model that combined the traditional biomarkers INR, total bilirubin, and aspartate aminotransferase with the candidate biomarkers OPN and K18, which had an AUC of 0.97. “Some analysis of candidate biomarkers in combination with tests such as MELD score [Model for End-Stage Liver Disease] and ‘Hy’s Law’ saw that incorporating candidate biomarkers was useful,” Dr. Church said.
Dr. Church reported having no financial disclosures.