Comment: Neuropharmacology.

readmittedto the hospital with lethargyand fever(38.2"C)of one day's duration. Alllaboratory parameterswerewithinnormal limits,excepta WBCcount of 3.1/mm' witha differentialof 2"70 segmented neutrophils, 0"70 band neutrophils, 31"70 lymphocytes, 18"70 atypical lymphocytes, 44"70 monocytes, 3"70 eosinophils, and 2"70 basophils. The patient had taken no other medications during the preceding six weeks. Upon hospitalization, vancomycin 500 mg iv q6hwasempirically administered for 19doses until the patient's absolute neutrophil count spontaneously rose to > 1000 (actual count 2064). All cultures werewithout growth. At time of discharge the patient was afebrile and not taking any antibiotics; one week later his WBC count was 4.7/mm' with a differential including27"70 segmented neutrophilsand 20"70 band neutrophils. All generations of the cephalosporins have been associated with a drug-induced neutropenia « 1000 neutrophils/mm'). Cephalosporin-induced neutropenia rarely occurs during the first week of treatment. Simultaneous eosinophilia, rash, fever, and/or thrombocytopenia may occur. Its incidence increases with large doses and prolonged therapy of beta-lactam antibiotics. The reaction has been reversible after withdrawal of the drug. '-3 The mechanism of cephalosporin-induced neutropenia remains unclear. Neftel et al. found that beta-lactam antibiotics inhibited in vitro granulopoiesis dose-dependently. Bone marrow aspirates in patients with beta-lactam neutropenia were characterized by a lack of well-differentiated myeloid elements in the presence of numerous granulocyte precursors. These findings were thought to be consistent with a nonimmunological drug-induced neutropenia.' Murphy et al. describes two patients with cephalosporin-induced neutropenia who developed immunoglobulin G cephalosporin antibodies against their neutrophils. This immune neutropenia involved either a drug absorption or immune complex mechanism." Neftel et al. also noted six cases of ceftriaxone-induced neutropenia from 17 hospitals in Switzerland and Germany. The mean time to develop neutropenia was 21 days (range 8-25 d) and the mean ± standard deviation of total and daily dose were 51 ±29 and 2.3 ± 1.0 g, respectively. Concomitant symptoms of neutropenia included eosinophilia, thrombocytopenia, reticulocytopenia, rash, and fever. In their literature review, these authors noted that large doses of any betalactam antibiotics induce neutropenia in 5 to > 15 percent of patients treated for ~ 10 days but in <0.1 percent of patients with a shorter duration of therapy.' Baumgartner and Glauser observed that 5 of 153 adult patients (3.3 percent) who received ceftriaxone developed neutropenia possibly related to the antibiotic. These cases appeared with a total ceftriaxone dose of 7-92 g within 2-25 days after initiating treatment and all were reversible within 2-12 days.' In the present case, neutropenia and fever appear to be associated with a high total dose (74 g) and long duration of therapy (37 d) of ceftriaxone. The low cumulative dose and short course of therapy of both vancomycin (1.5 g, < I d) and cefazolin (24 g, 7 d), as well as the long latency period between antibiotic administration and neutropenia, make them unlikely drug offenders. Also, cefazolin-induced neutropenia has been infrequently described in the Iiterature.s' Based on our experience, we suggest monitoring WBC count and differential regularly when a patient is receiving high doses and prolonged therapy with ceftriaxone.