Detection of Helicobacter pylori infection by saliva IgG testing

Most currently available tests for the detection of Helicobacter pylori are invasive, time consuming, or impractical. We examined the test performance of a simple and rapidly administered salivary IgG assay kit in the diagnosis of Helicobacter pylori infection.

Patients referred to a tertiary care setting for upper gastrointestinal endoscopy were included in a prospective evaluation of the test performance of the Helisal Kit which uses an ELISA technique to determine IgG antibodies in saliva. The results of the salivary IgG assay were compared to those of the Helisal Serum Kit, and to gastric histology. Two by two contingency table analyses were performed, and 95% confidence intervals (CI) were determined.

Upper gastrointestinal endoscopy was performed on 106 patients over a 3-month period. A statistically significant correlation was found between the blood and saliva IgG results (r = 0.60, p = 0.0001). When compared to serum IgG, the salivary assay test performance was: sensitivity 84% (CI: 70-93%), specificity 81% (CI: 69-90%), positive predictive value 76% (CI: 61-87%), negative predictive value 88% (CI: 76-95%), and diagnostic accuracy 88% (CI: 76-95%). Compared with gastric histology, the test performance of the salivary IgG assay decreased to: sensitivity 66% (CI: 52-79%), specificity 74% (CI: 60-85%), positive predictive value 71% (CI: 57-83%), negative predictive value 68% (CI: 55-80%), and diagnostic accuracy 70% (CI: 60-78%). More specifically, the salivary assay gave false-negative results in nine of 17 patients with duodenal ulcers. Results did not vary significantly when outcomes of the salivary and serum assays were combined. The incremental information obtained in the salivary test was greatest in the patient population exhibiting an intermediate pretest probability (30-70%) of being infected with Helicobacter.

The salivary IgG results correlated significantly with the serum IgG titers but exhibited only modest test performance, compared with the results of gastric histology. This salivary test may be most useful in certain patient subpopulations or in specific clinical contexts.