Philip Castle

PhD, MPH

DCP/NCI/NIH/DHHS

Director/DCP and Senior Investigator/DCEG

PLCO (Learn more about this study)

2023-0078

Dec 19, 2023

Evaluation of a new rapid H. pylori serology test

There is an opportunity to reduced cancer health disparities for HP-related non-cardia gastric cancer by detecting and treating HP infection. However, new, robust, low-cost diagnostics are needed to reach populations at risk, including rural populations and those experiencing persistent poverty who may not have access to routine services. Lateral flow diagnostics, like those developed for rapid COVID antigen testing, for the detection of HP could be deployed into communities of underserved populations. One such test is the Pylori DuoTect (http://www.imevax.com/pyloriduotect) (Imevax, Munich, Germany), which requires only 40 uL of whole blood and takes 20 minutes to complete. Such a test, if proven to be accurate, could be used in a same-day screen-and-treat strategy of serology testing using Pylori DuoTect.
In a collaboration with University of Puerto Rico Cancer Center (UPRCC), DCP, and DCEG, we will evaluate Pylori DuoTect using stratified sampling of PLCO serum samples that have been previously tested for serum antibodies against numerous HP antigens using a well-validated DKFZ multiplex serology assay that is widely considered and used as a reference standard. Two studies tested in 1,774 serum specimens using the DKFZ assay, of which 1,743 still have available serum: 705 HP seropositive (HP+) (40.4%) and 1,038 HP seronegative (HP-) (59.6%).
We will test aliquots of all HP+ and an equal number of randomly selected HP- sera, as determined by the DKFZ test, by Pylori DuoTect at UPRCC (n=1,410, 705 HP+ and 705 HP-). We will also retest a stratified random sample of 20% H. pylori seropositive and 20% H. pylori seronegative for reproducibility. Thus, a total of 1,692 (1,410 for performance and 282 for reproducibility) 50 L aliquots will be requested from the PLCO biorepository, and 20 L serum (≈40 L whole blood) per aliquot will be tested (with the remainder is available for reproducibility testing and retesting as needed), masked to all other data. Results will be reported back to the NCI and linked to the DKFZ assay results, other enrollment data (e.g., age, race, and gender), and any relevant diagnoses of cancer. Clinical performance parameters (e.g., sensitivity and specificity) and agreement statistics (e.g., kappa values, and percent positive and negative agreement) will be calculated.
If we assume that 90% sensitivity and 90% specificity of the Pylori DuoTect vs. the DKFZ test results is the minimum acceptable performance, then this sample size achieves 91% power to detect a change in sensitivity from 0.9 to 0.86 using a two-sided binomial test and 91% power to detect a change in specificity from 0.9 to 0.86 using a two-sided binomial test.