Use of vancomycin or first-generation cephalosporins for the treatment of hemodialysis-dependent patients with methicillin-susceptible Staphylococcus aureus bacteremia

Stryjewski ME, Szczech LA, Benjamin DK Jr, Inrig JK, Kanafani ZA, Engemann JJ, Chu VH, Joyce MJ, Reller LB, Corey GR, Fowler VG Jr. Clin Infect Dis 2007;44:190–196.

Summary:

The authors conducted a cohort study of 123 consecutive haemodialysis patients hospitalised between 1994 and 2001 in a tertiary care medical centre in Durham, North Carolina, USA who met two criteria: (1) methicillin-susceptible Staphylococcus aureus (MSSA) organisms were obtained from at least one culture of blood and (2) the treatment given for this was either vancomycin (n=77) or cefazolin (a first-generation parenteral cephalosporin) (n=46). Ten (8%) of these patients died. The endpoint of the study — treatment failure — was defined as either death or recurrent infection during a 12-week follow-up period. More patients treated with vancomycin experienced treatment failure than did those treated with cefazolin (31.2% vs 13.0%; p=0.02).

Other findings of the study:

• This was not a randomised trial; the choice of drug (and the dose and duration of treatment) was determined by the patient’s physician. However, the two groups of patients were comparable in terms of comorbidities and APACHE II score (an indicator of severity of illness), source of infection, type of haemodialysis access and several other factors that might affect outcome. The groups were not comparable in two factors (age and presence of metastatic infections) that might influence outcome; however, at this centre, patients who received cefazolin as their principal treatment fared better, despite the fact that they tended to be older and had more metastatic infections at presentation than did the patients who were treated with vancomycin.
• There was no significant difference in serum levels of vancomycin between patients whose treatment failed or those who were cured at 12 weeks; therefore, the difference in treatment outcome for the two drugs was not due to inadequate dosing in the vancomycin group. The authors noted that “data correlating serum levels of vancomycin with clinical outcome are limited and routine monitoring of vancomycin levels remains controversial”.
• There was no significant difference in minimum inhibitory concentration (MIC) of vancomycin between isolates from patients whose treatment with vancomycin failed and isolates from patients who were cured with vancomycin treatment; therefore, the difference in treatment outcome for the two drugs was not predicted by vancomycin MIC.
• In a multivariable logistic regression model controlling for other potentially important variables, the only two factors found to be significantly associated with treatment failure were a failure to remove the haemodialysis access catheter and the use of vancomycin. In this model, the odds of treatment failure when vancomycin was used were 3.5 times higher than for treatment with cefazolin.
• These data reflect the various practice patterns at one academic centre; how well these findings generalise to other settings is not clear.

AIM core principles supported:

• Antibiotic choice: If appropriate, change antibiotic dosage or therapy based on resistance and pathogen information.
• Infection control: Ensure adequate containment of the infection source by removing contaminated devices and draining/debriding infectious tissue.

Learning points:

• In serious infections such as those described here, once the responsible organism is known to be an S. aureus strain susceptible to methicillin (and thus to oxacillin and other β-lactam agents), these agents should be used in preference to vancomycin.
• Here, the independent association between treatment failure and a lack of removal of the vascular access — regardless of the antibiotic therapy given — echoes the findings of other studies (listed in the paper), indicating that removal of contaminated vascular access devices is a key treatment factor.