A puzzling microbiological and clinical discrepancy in the management of acute, severe skin-soft tissue and joint staphylococcal infection. In vitro antimicrobial susceptibility to glycopeptides, versus in vivo clinical efficacy of linezolid alone.

We present an intriguing case report of a septicemic post-elective surgical staphylococcal knee arthritis and cellulitis which did not respond to long-term courses of associated rifampicin and teicoplanin or vancomycin despite apparently favourable in vitro susceptibility assays, but rapidly resolved after i.v. followed by oral administration of linezolid. The lack of response to a two-week course of glycopeptides cannot be explained by the in vitro mimimum inhibitory concentrations (MIC90) of involved organisms, which showed full susceptibility of Staphylococcus aureus to vancomycin and teicoplanin, and sensitivity of an accompanying Staphylococcus epidermidis isolated from blood cultures to vancomycin and rifampicin, with borderline "intermediate" values found for teicoplanin. Since neither abscess formation nor bone involvement were of concern, effective glycopeptide and rifampicin penetration into infectious tissue should have been ensured. From a clinical point of view, only the introduction of a two-week i.v. linezolid followed by one more week of oral linezolid obtained a complete clinical and microbiological cure, and an unhoped-for functional success. When managing severe multiresistant gram-positive infections, in vitro activity should be carefully evaluated against expected drug penetration rates into the relevant infectious tissues.