Early diagnosis of primary immunodeficiency (PID) remains a significant clinical challenge, often delaying treatment and increasing the risk of severe infections. Accurate identification of antibody (Ab) deficiencies typically requires multiple diagnostic modalities, including flow cytometry (FACS), serum immunoglobulin measurements, and clinical criteria, each providing only partial insight into immune function. In this study, we evaluated whether PrimDx, a single whole-blood assay combining RNA sequencing with machine-learning algorithms, could provide a more integrated and reliable assessment of Ab deficiency and immune dysregulation.
Across 54 patients with clinically confirmed PID, including common variable immunodeficiency (CVID), specific antibody deficiency (SAD), X-linked agammaglobulinemia (XLA), activated PI3K δ syndrome, and others, and 16 non-PID controls (non-cystic fibrosis bronchiectasis without clinical immune deficiency), PrimDx showed strong concordance with established clinical tests. Individuals with higher PrimDx scores, reflecting greater immune dysfunction, demonstrated immune profiles consistent with Ab deficiency: reduced total, memory, and switched-memory B cells; lower natural killer (NK) and CD4 T cell counts; and a shift toward naïve-skewed B cell architecture. Transcriptomic patterns of immunoglobulin genes and B cell markers aligned closely with these cellular findings. Serum IgA, IgG, and IgM levels similarly tracked with PrimDx scores, reinforcing this multidimensional coherence.
A multivariate model integrating cellular parameters explained substantial variance in PrimDx scores (R2 ≈ 0.78), with naïve B cells, NK cells, and IgM contributing most strongly. Several patients also showed NK cell abnormalities, supporting PrimDx’s capacity to detect broader innate-adaptive dysregulation. Vaccine-response testing had been performed clinically, but formal results were available for only 14 PID (11 CVID; 2 SAD; 1 APDS) participants due to interpretability limitations. Among these, three demonstrated normal responses, ten had inadequate responses, and one result was uninterpretable due to high baseline antibody titers. PrimDx classified all 14 as PID positive, aligning with their clinical diagnoses and suggesting that it captures global immune impairment more comprehensively than vaccine-response results alone.
Together, these findings demonstrate that PrimDx integrates cellular, transcriptomic, and immunoglobulin information into a coherent assessment of immune dysfunction. As a single, minimally invasive assay reflecting B cell impairment, Ig reduction, and broader immune imbalance, PrimDx has the potential to streamline and enhance PID diagnostic workflows.
