Recurrence of Pneumococcal Sepsis Caused by an Intermediately Penicillin-Resistant Organism Treated With Loracarbef

A number of reports have documented failure of treatment with third generation cephalosporins in patients with penicillin-resistant pneumococcal isolates.1 Most of these reports have been in patients with infections caused by highly resistant isolates (e.g. MIC ≥ 2.0 μg penicillin/ml), particularly in those patients with meningitis.2-4 We report a case of a child with recurrent pneumococcal bacteremia caused by an intermediately resistant strain, who relapsed after treatment with an oral carabacepham, loracarbef. Materials and methods. Isolates of Streptococcus pneumoniae, recovered from blood culture, were serotyped by the New Mexico State Department of Health Laboratories. Pulsed field electrophoresis typing (CHEF DRII®; Bio-Rad) of the isolates were performed in the University of New Mexico Molecular Epidemiology Laboratory by modifications of methods previous published.5 Susceptibility of both isolates was tested by the E-Test® (AB Biodisk North America, Culver City, CA) in the University of New Mexico Hospital Microbiology Laboratory.6 Case report. A previously healthy 18-month-old boy became ill with otitis media and was treated with trimethoprim/sulfamethoxazole suspension. He failed to improve, and on the fifth day of treatment a chest radiograph demonstrated a right upper lobe pneumonia. Additional study revealed a white blood cell count (WBC) of 15 700 mm3 with 86% neutrophils. Treatment was changed to lorcarbef suspension (30 mg/kg/day in two divided doses). Blood culture subsequently grew S. pneumoniae reported to be intermediately resistant to penicillin and susceptible to cefotaxime. At the time of the report (second day of treatment), the WBC had fallen to 10 200 mm3 with 69% neutrophils, treatment with loracarbef was continued, and the child became afebrile for the remainder of the 10-day course of antibiotics. Three days after completing the oral antibiotic he returned with a complaints of fever and the WBC had risen to 16 600 mm3 with 81% neutrophils. Chest radiographs showed improvement of the right upper lobe pneumonia, but a repeat blood culture again grew S. pneumoniae with a susceptibility pattern similar to that of the initial isolate. Clinical evaluation, including radiographs of the paranasal sinuses, bone scan and a radioisotope-labeled WBC scan, failed to identify an additional site of infection. Therapy was changed to clindamycin (40 mg/kg/day divided every 8 h iv). He became afebrile by 48 h and intravenous therapy was changed to oral therapy and administered for an additional 10 days. On the seventh day of treatment he experienced a single day of fever associated with a presumptive viral illness. He became afebrile the next day and remained well thereafter. Results. Pneumococcal isolates for the first and second episode were both serotype 14. Both had identical patterns on Sma I digests submitted to CHEF pulsed field electrophoresis. Susceptibility testing results of both isolates were identical by E-Test®: penicillin, 0.50 μg/ml; and cefotaxime, 0.125 μg/ml. Subsequent testing of the organism revealed MICs of >2.0 μg/ml of erythromycin and 0.12 μg/ml of clindamycin (Pasco MIC Supplemental Panels®; Difco Laboratories, Detroit, MI). Discussion. In this child both isolates in the primary and recurrent sepsis episodes were caused by S. pneumoniae type 14 with identical pulsed field and susceptibility patterns to penicillin and cefotaxime. Susceptibility to loracarbef was not specifically tested, but in vitro susceptibility to a third generation cephalosporin, coupled with his initial response to therapy suggested that continued treatment with loracarbef would be successful. Clindamycin was empirically started at the onset of his second episode because of recent data that suggested that American isolates are frequently susceptible to this antibiotic.7 The reason for persistence and subsequent relapse of this child's pneumococcal infection is unknown. Initial studies did not suggest meningitis and clinical evaluation at the time of the recurrence did not reveal osteomyelitis, sinusitis, mastoiditis, persistent pneumonia or endocarditis as a possible factor in his relapse. This child's infection was caused by an intermediately penicillin-resistant pneumococcal organism. Recent data have suggested that newer oral cephalosporins may not be active against intermediately susceptible strains.7 The MIC50 for loracarbef and oral cephalosporin (cefaclor, cefuroxime, cefixime) ranged from 16 to 32 μg/ml with the percent of susceptible organisms of intermediately susceptible strains ranging from 8 to 100%. Cefuroxime had the best activity of newer oral cephalosporins. However, clindamycin may have the most promising activity with 95% of strains being susceptible at an MIC90 of 0.03 μg/ml. Resistance to beta-lactam compounds is associated with production of altered penicillin-binding proteins.8, 9 Although resistance to penicillin does not negate the activity of third generation parenterally administered cephalosporins (e.g. cefotaxime), it does tend to predict resistance to many of the oral cephalosporins. The use of parenteral cephalosporins is often successful against systemic infections caused by intermediately penicillin resistant pneumococci.10 However, the study of our patient demonstrates that extended spectrum oral cephalosporins may not successfully eradicate pneumococci intermediately susceptible to penicillin, even if otitis media or pneumonia is the primary infection, and that these agents should not be relied upon for clinical treatment. Currently, clindamycin or the newer macrolides may offer the best potential for therapeutic success. Laurence Shandler, M.D.; Janis Gonzales, M.D.; Gary D. Overturf, M.D. Santa Fe Pediatric Associates, P.C.