Optimizing the indeterminate zone for the DiaSorin LIAISON H. Pylori stool antigen test

Abstract

Introduction

Helicobacter pylori (H. pylori) colonization increases the risk of upper gastrointestinal disorders and can be detected by various tests, including the stool antigen test (HpSAT). DiaSorin recommends an HpSAT equivocal/indeterminate zone of 0.90-<1.10, but high variability observed in our laboratory prompted clinical implementation of a broader zone (0.60-<1.80). This study aimed to define an optimal HpSAT indeterminate zone using molecular as reference and urea breath test (UBT) for confirmation.

Materials and Methods

HpSAT and stool molecular results were available from 379 patients, of which 52 had follow-up UBTs. HpSAT was analyzed by the LIAISON HpSAT assay (DiaSorin), UBT by isotope ratio mass spectrometry, and molecular testing by qPCR targeting H. pylori DNA. Logistic regression modeled HpSAT index values against PCR positivity to define an optimal indeterminate zone, supported by clinical performance and flagging rates analyses.

Results

Logistic regression determined an HpSAT index of 0.79 (95 % CI: 0.34–1.14) had 90 % probability of a negative PCR result, and 4.99 (4.09–6.63) had 90 % probability of a positive result, rounded to an indeterminate zone of 0.80-<5.00. Lower thresholds assessed all had ≤ 2 % false negatives, while upper thresholds exhibited decreased false positive rates (22 % to 6 %) as thresholds increased (1.10 to 5.00), with minimal improvement beyond 3.00 (9 %). A modified zone of 0.80–<3.00 offered high accuracy with 12.9 % indeterminate results (DiaSorin’s threshold: 2.8 %; laboratory’s current threshold: 13.1 %).

Conclusions

Our findings show that DiaSorin’s HpSAT indeterminate zone is too narrow to reliably distinguish true positive and negative results in clinical practice. A modified broader zone (0.80–<3.00), derived via logistic regression using PCR as reference, improves diagnostic accuracy while minimizing indeterminate results.