Empiric treatment of fever in acute leukaemia with tobramycin-cephalothin, and the escape clause provision of corticosteroids.

29 episodes of suspected septicaemia in patients with acute leukemia were treated empirically with tobramycin 180--240 mg/day intravenously together with cephalothin 12 g/day. Patients without documented infection who did not respond to antibiotics and whose fever developed after a course of cytotoxic drugs, were given the provision of high dose corticosteroid therapy. Infection was documented microbiologically or clinically in 13/29 episodes. Septicaemia was proven in 7, and 6 had pneumonia. Neutropenia was present in 18/29 episodes. A satisfactory response to initial therapy was achieved in 7/13 with documented infection and in 9/16 without proven infection. The overall good response was 55%, 5/7 cases with septicaemia, but only 2/6 with pneumonia responded well. The 2 septicaemia patients who did not respond had Pseudomonas aeruginosa sepsis. In 16 episodes without documented infection 7 did not respond to initial therapy. To 4 of them, who were subject to recent cytotoxic drug administration, high dose corticosteroid therapy was given, and 3 of them responded well. Of the remaining 3 non-responders, one became afebrile after cytostatic and one after prednisolone treatment. Serum assays of tobramycin were done on the 1st and 5th day of therapy and no difference in concentration was observed on these 2 occasions. Five patients developed renal failure, but this was attributed to antibiotic therapy only in 1, who initially had an elevated serum creatinine. It is concluded, that in hospitals where pseudomonas is not a dominating pathogen, tobramycin--cephalothin may be a good combination to start empiric therapy with. In patients without proven infection, who have recently been subjected to cytotoxic therapy, and who do not respond to the initial course of antibiotics, a high dose of corticosteroids may be tried, provided the patient is monitored for the hazard of bacterial infection.