Randomized study of vancomycin versus teicoplanin for the treatment of gram-positive bacterial infections in immunocompromised hosts.

Seventy-four immunocompromised patients with severe infection due to gram-positive organisms were randomized to receive either vancomycin or teicoplanin. Extensive cancer was present in 71 patients, of whom 47 died within a month. The types of infections were 46 bacteremias (39 associated with central catheters), 24 skin and soft tissue infections (3 with bacteremia), and 7 others (mainly bronchopneumonia). The most frequent pathogen was Staphylococcus epidermidis, followed by Staphylococcus aureus. Microbiological eradication was obtained in 23 of 35 evaluable patients treated with vancomycin (65.7%) and 28 of 36 patients treated with teicoplanin (77.8%) (P = 0.4). Clinical cure and improvement were obtained in 26 of 35 patients (74.3%) and 27 of 36 patients (75.0%), respectively. No significant side effects were observed with teicoplanin, in contrast to reversible increases in serum creatinine (three patients) and skin rashes (four patients) with vancomycin. Superinfection was observed in five patients treated with vancomycin and two patients treated with teicoplanin. No relation was found between peak concentration in serum (at steady state) or bactericidal titers and outcome.