In the Literature

Adjunctive Rifampin for Staphylococcal Endocarditis

Riedel DJ, Weekes E, Forrest GN. Addition of rifampin to standard therapy for treatment of native valve endocarditis caused by Staphylococcus aureus. Antimicrob Agents Chemother 2008; 52:2463–7.

Despite a lack of adequate evidence to demonstrate benefit, rifampin is frequently added to treatment with either vancomycin or a β-lactam antibiotic for patients with endocarditis due to Staphylococcus aureus. Riedel and colleagues performed a retrospective, matched cohort study to assess this strategy in patients with native valve endocarditis. Only right-side valves were involved in 65% of 84 patients; 79% of the infections were caused by methicillin-resistant S. aureus (MRSA), and 83% of patients received vancomycin. Forty-two case patients also received rifampin, which was continued for 14–28 days (median duration, 20 days), having been added a median of 3 days (range, 0–19 days) after the initiation of therapy; 16 patients were still bacteremic at the time that rifampin was added to the treatment regimen. Eighty-one percent of case patients and only 17% of the 42 control subjects received gentamicin.

Case patients had a longer median duration of bacteremia than did control subjects (5.2 vs. 2.1 days). Bacteremia had persisted for a median of 4.0 days when rifampin was added in the 16 patients who still had positive blood culture results at the time of its initiation, and bacteremia persisted for an additional median of 4.0 days after initiation of combination therapy. Case patients were more likely than control subjects to require surgery. The survival rate was 79% among case patients and 95% among control subjects.

Resistance to rifampin emerged in 9 patients (21%) a median of 16 days after the addition of rifampin to the regimen; all 9 were still bacteremic at the time that rifampin was added. The serum transaminase level increased to >5 times the baseline level in 9 case patients (21%) and 1 control subject (2%); all 10 of these persons had chronic hepatitis C virus infection. Significant drug-drug pharmacokinetic interactions occurred with rifampin in 22 patients; the reactions predominantly involved methadone, warfarin, and HIV-1 protease inhibitors.

Current guidelines recommend that patients with prosthetic valve endocarditis due to either S. aureus or coagulase-negative staphylococci (both methicillin-susceptible and methicillin-resistant staphylococci) receive, in addition to either nafcillin (or oxacillin), both gentamicin and rifampin [1]. It should be noted that the addition of gentamicin and/or rifampin is based, at best, on minimal supportive evidence. Rifampin is not recommended for patients with native valve endocarditis, and gentamicin is an optional recommendation. In a study in which a total of 42 patients with (mostly right-side) native valve endocarditis due to MRSA were given vancomycin and were randomized to also receive rifampin or not, rifampin administration was associated with a longer duration of bacteremia (median duration, 9 vs. 7 days) [2].

Although rifampin superficially comes out looking terrible in this study, the picture is somewhat more complex. Despite the apparent comparability of the 2 patient groups at baseline, it is quite likely that patients who received rifampin (and frequently gentamicin as well) had evidence of more-severe illness that was not captured in the investigators' analysis. Although drug-drug interactions were frequent, these are largely predictable and can be appropriately managed. The frequent marked elevation of hepatic transaminase levels in rifampin recipients occurred only in patients with chronic hepatitis C virus infection. Nonetheless, in the absence of evidence of benefit, the addition of rifampin to baseline therapy for patients with native valve endocarditis would not appear to be warranted. Furthermore, despite the guideline recommendations [1], I also have doubts regarding its value in patients with prosthetic valve endocarditis.

Vancomycin “Just Can't Get No Respect”

Lodise TP, Graves J, Evans A, et al. Relationship between vancomycin MIC and failure among patients with methicillin-resistant Staphylococcus aureus bacteremia treated with vancomycin. Antimicrob Agents Chemother 2008; 52:3315–20.

In a retrospective cohort study, Lodise and colleagues examined the relationship of MICs of vancomycin, as determined by Etest (AB Biodisk), to outcomes of vancomycin treatment in nonneutropenic adults with MRSA bacteremia at the Albany Medical Center. Treatment failure could be the result of microbiologic failure (defined as the persistence of bacteremia after at least 10 days of vancomycin therapy), death due to any cause ⩽30 days after the first positive blood culture result, or recurrence of bacteremia ⩽60 days after discontinuation of vancomycin therapy.

Ninety-two patients who received the first dose of vancomycin within 24 h after the first positive blood culture result and who survived for at least 24 h after receipt of that dose were evaluated. Classification and regression tree analysis identified an MIC of vancomycin ⩾1.5 µg/mL as associated with an increased risk of treatment failure, which occurred in 28 patients (30.4%); 66 patients (71.7%) were infected with an isolate that had an MIC ⩾1.5 µg/mL (“high MIC”). No isolates with MICs of vancomycin >2.0 µg/mL were identified. The rate of treatment failure was 36.4% among patients with high-MIC strains and only 15.4% among those whose isolates yielded MICs ⩽1.0 µg/mL (P=.049). Among patients who were not treated in the intensive care unit, failure occurred in 37.8% of persons infected with high-MIC strains and only 12.5% of those infected with low-MIC strains (P=.03). Vancomycin trough levels were not associated with outcome in the 53 patients in whom it was measured. Poisson regression identified MICs of vancomycin ⩾1.5 µg/mL, APACHE II scores ⩾20, infective endocarditis, and weight ⩾112 kg as being independently associated with treatment failure.

By current US Food and Drug Administration and Clinical and Laboratory Standards Institute standards, isolates of S. aureus with a vancomycin MIC ⩽2.0 µg/mL are considered to be susceptible. Several retrospective studies, however, have found that vancomycin treatment of patients with bloodstream MRSA isolates with MICs at the upper range of “susceptibility” results in treatment failure (variously defined) at significantly higher rates than occurs in patients whose isolates have lower MICs. The most recently published of these studies (prior to that by Lodise and colleagues) found that an MIC of vancomycin of 2.0 µg/mL was an independent risk factor for mortality in patients with MRSA bacteremia treated with this glycopeptide [1]. Although each and every study can be criticized, the published evidence is overwhelmingly consistent with the conclusion that vancomycin, as generally administered, is ineffective therapy for patients with bloodstream infection due to MRSA with vancomycin MICs at the high end of the range currently considered to indicate susceptibility.

Vancomycin, metronidazole, Clostridium difficile and Vancomycin-Resistant Enterococci (VRE)

Al-Nassir WN, Sethi AK, Li Y, et al. Both oral metronidazole and oral vancomycin promote persistent overgrowth of vancomycin-resistant enterococci during treatment of Clostridium difficile—associated disease. Antimicrob Agents Chemother 2008; 52:2403–6.

Reluctance to recommend the use of oral vancomycin for treatment of most cases of C. difficile—associated diarrhea (CDAD), despite accumulating evidence that it may be superior to metronidazole, seems to emanate from 2 factors: the greater acquisition cost of vancomycin, and the fear that its use will lead to increased prevalence of gastrointestinal colonization with VRE. In a prospective observational study, Al-Nassir and colleagues examined the effect of treatment of CDAD with either of these 2 agents on the emergence of VRE in stool specimens.

VRE were detected before the initiation of 56 treatment courses; in these patients, the density of VRE colonization persisted largely unchanged during treatment with each antibiotic and then decreased significantly ∼2 weeks after the end of each treatment regimen. VRE was not detected in stool specimens that had been obtained before 34 treatment courses using a method in which the lower limit of detection is ∼2.5 log>10 cfu/g. In this group, VRE was subsequently detected after receipt of 3 (14%) of 22 metronidazole courses and 1 (8%) of 12 vancomycin courses. Thus, it appears that there may be no significant difference in overgrowth of VRE in association with treatment of CDAD with either metronidazole or vancomcyin. This suggests that the only reason to give what may be an inferior therapy is acquisition cost, an issue that can be greatly ameliorated by the use of the intravenous formulation of vancomycin for oral administration.