A team of researchers led by Eleftherios P. Diamandis at the University of Toronto set out to discover novel serological markers for the early detection of pancreatic adenocarcinoma (PDAC). Their research hypothesis was that proteins expressed uniquely by the pancreas present stronger candidate biomarkers for pancreatic cancer.

The team used commercial immunoassay tests (ELISA) to analyze and compare serum samples from healthy donors with late-stage PDAC patients. These studies led to the discovery of a promising new pancreatic cancer biomarker, the protein CUB and zona pellucida-like domain 1 (CUZD1).

The results obtained by Diamandis et al. raised the expectations for a potential breakthrough in PDAC and opened the possibility to develop the first screening method for this deadly disease – based on CUZD1, a molecule indicating the presence of cancer at an early-stage if it is present in unusual amounts. As with practically every type of cancer, timing and the early diction are critical to significantly increase the five-year survival rate.

However, when Diamandis and his group investigated whether their ELISA assay truly recognized the protein of interest (something most researchers would never attempt to do with a commercial product), they found that the expensive ELISA kit did not detect CUZD1 at all. 

As it turned out, the detected antigen was actually the established cancer protein CA125 – readily explaining the high expression levels obtained with patient serum samples.

It took the Diamandis group two years and thousands of patient samples to realize the problem and around $500,000 for almost 100 CUZD1 ELISA kits, associated labor costs and all confirmatory experiments.

Importantly, if the scientists had not questioned their ‘convincing’ preclinical results, an obvious next step would have been to start clinical trials to test whether early detection with this novel biomarker could lead to a better patient outcome. However, in this case, detection of cancer would not have happened at an early stage but only at a later stage where the cancer protein CA125 is present – but most likely too late for most patients.

CUZD1 and CA125 share no molecular similarities that can explain this error. It is estimated that more than 300 companies offer more than 2 million antibodies for preclinical research (REF). When independently testing more than 6,000 commercial antibodies from 26 suppliers it was found that more than 75% of these antibodies were non-specific or did not work at all (REF). Furthermore, the consortium Human Protein Atlas examined more than 5,000 commercial antibodies, from which more than 50% could not be used in the anticipated application (REF).

Currently, there are no official standards for the quality of commercially available antibodies. To improve the quality of key research tools, manufacturers of antibodies need to fulfill more stringent quality standards, testing their antibodies with multiple methods. These quality standards should be established and evaluated by an independent third-party organization that would certify the antibodies prior to their usage. Funding organizations should demand this antibody quality standard prior to the start of a project and editors as well as reviewers of scientific journals should demand them to confirm antibody specificity and to control for cross-reactivity.