In a new Frontiers in Cellular and Infection Microbiology study, researchers from Karolinska Institutet, in collaboration with Aplex Bio, used Hyperplex PCR (hpPCR) to profile periodontal pathogens in saliva, working toward non-invasive biomarkers for periodontitis.

Periodontitis is a common, largely silent condition that becomes irreversible once established, making early and reliable detection a persistent clinical challenge. Existing diagnostic methods often fall short on sensitivity and specificity, and saliva offers a promising but underused source of diagnostic information.

In this study, hpPCR was used to simultaneously quantify 11 periodontal bacterial taxa plus 2 internal controls targeting the 341F-518R region of bacterial 16s rRNA in saliva from 57 participants, comparing periodontally healthy individuals with those with periodontitis stage III/IV. The resulting bacterial profiles were then analyzed together with host inflammatory marker data using machine learning to explore which biomarkers best distinguished disease from health.

Why hpPCR was the enabler:

True multiplex from a single saliva sample with single base pair specificity

All 11 bacterial targets and 2 controls were measured in one reaction, avoiding the need to split limited saliva volume across separate assays and preserving material for parallel inflammatory profiling. Probes were designed against 16S rRNA sequences allowing accurate and versatile discrimination of bacterial genera and species by balancing single nucleotide specificity and mixed bases to modulate coverage. 

Digital molecular counting

Molecular counting enabled precise relative quantification of bacteria present at low abundance, which was essential for detecting the subtle sub-2-fold shifts that distinguish periodontal health from disease.

A practical path to clinical translation

The standardized, single-amplicon design and streamlined workflow position hpPCR as a practical, clinically translatable approach to targeted microbiome profiling in a diagnostic context, rather than a research-only tool.

The analysis identified Fretibacterium spp. as the strongest individual discriminator between health and disease (AUC = 0.82), a result that held up better than combinations of multiple markers. The findings mark a step toward saliva-based, non-invasive diagnostics for periodontitis, and point to further validation in larger cohorts as the next stage of this research.

Read the full publication here: https://www.frontiersin.org/journals/cellular-and-infection-microbiology/articles/10.3389/fcimb.2026.1846125/full