Rodney Rouse, DVM, MBA, PhD, Research Veterinary Medical Officer and Acting Associate Director, Division of Applied Regulatory Science, Office of Translational Sciences, Center for Drug Evaluation and Research

Biomarkers are measurable indicators in the body that can signify the presence of disease or disease severity.  The identification of new, reliable and sensitive biomarkers remains a priority for the FDA.   In addition to measuring disease presence and severity, a biomarker may detect early tissue injury caused by a drug. These types of biomarkers may be used in safety studies to better characterize a drug’s risk profile during drug development.

One injury requiring more sensitive biomarkers is acute pancreatic injury. Post-market drug monitoring has revealed numerous drugs linked to pancreatic injury.  More recently, specific drugs used to treat type-2 diabetes were associated with acute inflammation of the pancreas, termed pancreatitis. This side effect was not evident in pre-approval, non-clinical safety studies and clinical trials that relied on the traditional pancreatic injury biomarkers, serum amylase and lipase.

Serum amylase and lipase are useful when their levels are greatly elevated in the serum, but they cannot reliably detect very early or low-level pancreatic injury, nor can they accurately measure disease severity or predict disease progression.  These traditional biomarkers are also non-specific, meaning other non-pancreatic problems can cause them to be elevated.  In drug development, more sensitive and specific biomarkers are required to detect and monitor the pancreatic injury potential of new drugs.  We also need new biomarkers that can accurately identify pancreatic injury from drugs earlier and with more specificity, be sensitive to the degree of injury, and provide information on prognosis.

Testing Candidate Biomarkers

Research conducted by the Division of Applied Regulatory Science suggested that microRNAs (miRNAs) may be good biomarker candidates. MiRNAs are short RNA molecules that do not help produce a protein in the body, but that have a key role in the regulation of genes by turning off—or repressing—gene expression. Many miRNAs are normally found in many types of biological fluids and can increase rapidly after tissue injury. MiRNAs are highly stable even if subjected to freezing and thawing. Importantly, certain miRNAs are tissue-enriched, meaning they are present primarily in specific organs or types of tissue.

Initially, we examined two miRNAs—miR-216a and miR-217—that had previously been identified as enriched in the pancreas of rats and mice. We aimed to determine whether these two miRNAs could detect the early onset of acute pancreatic injury, and potentially serve as reliable and sufficiently sensitive biomarkers.

In the lab, we induced acute pancreatic injury in rats and mice in three ways—by administering the chemical caerulein, by administering another chemical called l-arginine, and by surgical ligation of the pancreatic duct. We then measured levels of serum amylase, serum lipase, miR-216a, and miR-217 in each of the three conditions, and compared their relative accuracy in predicting the severity of microscopic injury in the pancreas.

Overall, we found that miR-217 and miR-216a were able to detect pancreatic injury earlier or at the same time as serum amylase and serum lipase. And unlike amylase and lipase, miR-217 and miR-216a were more specific, e.g., they were associated only with pancreatic injury and not with other types of disease or toxicities. Increases in miR-217 and miR-216a were detected within an hour or less of chemical or surgical treatment. Our research suggested that these miRNAs could potentially serve as biomarkers for pancreatic injury with greater specificity and sensitivity than amylase or lipase. Moreover, we found similar results with these two miRNAs and with three other pancreas-enriched miRNAs—miR-216b, miR-375, and miR-148a—in a canine caerulein model of pancreatic injury supporting the translatability or utility of these miRNAs across different species.  Subsequently, all of these miRNAs have been detected in humans with acute pancreatitis.  Additional data is needed to determine how these miRNAs perform under a wider range of drug-induced injury and diseases, and to determine their utility in patients with acute pancreatitis.

Informing Biomarker Development

Our work provides foundational support for the biomarker qualification package that is being developed by the Predictive Safety Testing Consortium (PSTC) of the Critical Path Institute. The PSTC works to identify new and improved safety testing methods and submits them to the FDA, the European Medicines Agency (EMA) and the Japanese Pharmaceutical and Medical Devices Agency (PMDA) for formal regulatory qualification. The PSTC brings together pharmaceutical companies to share and validate innovative safety testing methods with advice from the FDA, the EMA, and the PMDA, as well as from more than 250 scientists across industry and academia.

The development of new, more sensitive models and biomarkers for detecting and measuring drug-induced pancreatitis will improve our overall ability to predict drug safety. Moreover, once biomarkers are qualified, drug developers can confidently use them for the qualified context of use.