A Bioinformatic Approach to Utilize a Patient's Antibody-Secreting Cells against Staphylococcus aureus to Detect Challenging Musculoskeletal Infections.

ABSTRACT

Noninvasive diagnostics for Staphylococcus aureus musculoskeletal infections (MSKI) remain challenging. Abs from newly activated, pathogen-specific plasmablasts in human blood, which emerge during an ongoing infection, can be used for diagnosing and tracking treatment response in diabetic foot infections. Using multianalyte immunoassays on medium enriched for newly synthesized Abs (MENSA) from Ab-secreting cells, we assessed anti-S. aureus IgG responses in 101 MSKI patients (63 culture-confirmed S. aureus, 38 S. aureus-negative) and 52 healthy controls. MENSA IgG levels were assessed for their ability to identify the presence and type of S. aureus MSKI using machine learning and multivariate receiver operating characteristic curves. Eleven S. aureus-infected patients were presented with prosthetic joint infections, 15 with fracture-related infections, 5 with native joint septic arthritis, 15 with diabetic foot infections, and 17 with suspected orthopedic infections in the soft tissue. Anti-S. aureus MENSA IgG levels in patients with non-S. aureus infections and healthy controls were 4-fold (***p = 0.0002) and 8-fold (****p < 0.0001) lower, respectively, compared with those with culture-confirmed S. aureus infections. Comparison of MENSA IgG responses among S. aureus culture-positive patients revealed Ags predictive of active MSKI (IsdB, SCIN, Gmd) and Ags predictive of MSKI type (IsdB, IsdH, Amd, Hla). When combined, IsdB, IsdH, Gmd, Amd, SCIN, and Hla were highly discriminatory of S. aureus MSKI (area under the ROC curve = 0.89 [95% confidence interval 0.82-0.93, p < 0.01]). Collectively, these results demonstrate the feasibility of a bioinformatic approach to use a patient's active immune proteome against S. aureus to diagnose challenging MSKI.