Variable ELISA Biomarker Test

Scientists say they have developed a new computational method to reduce variability in common research biomarker tests. They see their techniques as a promising step in improving the ability of biomedical researchers and basic scientists to reproduce data and facilitate more consistent results across labs and long-term projects.

Researchers from Boston Medical Center (BMC) and Boston University School of Medicine (BUSM) have named their new software ELISAtools, which provides a platform to compare test kit data for research use only and minimize variability over months or even years. The team's work ("A computational solution to improve biomarker reproducibility in long-term projects") appears online in PLOS One.

“Biomarkers are fundamental to basic and clinical research outcomes by signaling host responses and providing insight into disease pathophysiology. Measuring biomarkers with research-use ELISA kits is universal, but the lack of kit standardization and unexpected lot-to-lot variability present analytical challenges for long-term projects. During an ongoing two-year project measuring plasma biomarkers in cancer patients, control concentrations for one biomarker (PF) decreased significantly after changes in ELISA kit lots.

A comprehensive review of operations highlighted standard curve shifts with the new kits, an analytical variable that compromised data already collected on hundreds of patient samples. After ruling out other reasonable contributors to data variability, an IT solution was developed to provide a uniform platform for analyzing data across multiple lots of ELISA kits,” the investigators wrote.

“The solution (ELISAtools) was developed in open source R software in which variability between kits is treated as a batch effect. A defined best-fit reference standard curve is modeled, a unique "S" shift factor is calculated for each standard curve, and the data is fitted accordingly. The mean S-factors for PF ELISA kit lots #1-5 ranged from -0.086 to 0.735 and reduced the variability between control assays from 62.4% to <9%, within quality control limits.

Calculated S-factors for four other biomarkers provided a quantitative metric to monitor ELISAs over the 10-month study period for quality control purposes. Reproducible biomarker measurements are essential, especially for long-term projects with valuable patient samples. The use of ELISA kits for research use is ubiquitous and the judicious use of this computational solution maximizes the reproducibility of biomarkers.

ELISA (Enzyme-Linked Immunosorbent Assay) tests are used worldwide in clinical, biomedical and basic research to measure biomarkers in a range of media, including blood, plasma and urine.

Clinical ELISA test kits used in hospital settings are regulated to ensure strict quality control limits for accuracy and consistency. However, the hundreds of commercially available, research-only ELISA test kits are unregulated, often leading to noticeable variability in results over time, between test kits and between different labs, scientists say.

The BMC-BUSM research team unexpectedly encountered high variability in an ELISA test kit during a project for the National Cancer Institute measuring biomarkers of thrombosis and inflammation in the plasma of cancer subjects and healthy donors. After the first year of the project, they realized that the data changed significantly as they received different shipments of the kit from the manufacturer.

The researchers determined that differences in the ELISA kit caused the problem. They had data from over 400 patient samples that could not be compared due to these differences in ELISA kits. To address this issue, the team created ELISAtools software to reduce future variability in test results.

"After implementing this software, the variability of test results dropped from over 60% to less than 9%, well below our quality control limits," said Deborah J. Stearns-Kurosawa, PhD. , lead study author and associate professor of pathology. and doctor