Bacteremic pneumococcal infections in immunocompromised patients without AIDS: the impact of beta-lactam resistance on mortality.

BACKGROUND

Streptococcus pneumoniae is the leading cause of community-acquired pneumonia in the elderly, and in recent years it has arisen as an important pathogen in HIV-infected patients. However, there is a scarcity of information on clinical and therapeutic problems associated with pneumococcal infections in other immuno-compromised patients. The objective of this study was to assess the most relevant epidemiologic aspects, clinical features and prognostic factors of pneumococcal bacteremia in immunocompromised hosts without AIDS.

METHODS

This was a retrospective analysis of patients with pneumococcemia, carried out in a 600-bed, university-affiliated hospital in Madrid, Spain. Two-hundred and sixty patients were evaluated retrospectively; 69 (26.5%) immunocompromised patients based on strict case definitions were compared with a group composed of 191 non-immunocompromised hosts with a variety of chronic conditions. Conventional management of pneumococcal bacteremia according to clinical standards was assessed. The MICs of penicillin and other beta-lactam antibiotics, and related mortality and hospital mortality at 30 days, were measured.

RESULTS

A comparison of clinical manifestations of pneumococcemia between immunocompromised patients and non-immunocompromised patients did not show differences in the presence of fever, obtundation, type of lung involvement, frequency of primary bacteremia, or meningitis. Hospital-acquired pneumococcemia was significantly more frequent in immunocompromised patients (34.7% versus 6.8%, P<0.0001), and resistance to penicillin was also more common in pneumococcal strains isolated from these patients (37.5% versus 20%, P=0.0009). Septic shock occurred more frequently in immunocompromised patients, although the overall and related mortality were not significantly different from those found in non-immunocompromised patients (33.3% versus 22.5%, P=0.07, and 28.9% versus 20.9%, P=0.7 respectively). In the multivariate analysis, multilobar pneumonia (odds ratio (OR) 15.7; 95% CI 6.00-41.30; P<0.001), inadequate treatment (OR 12.20; 95% CI 4.10-37.20; P<0.001), obtundation (OR 5.80; 95% CI 2.20-15.00; P<0.001) and hospital-acquired bacteremia (OR 4.80; 95% CI 1.00-14.60; P<0.006) were associated with an increased risk of mortality in patients with pneumococcemia. Only multilobar pneumonia (OR 7.90; 95% CI 4.10-15.35; P<0.001) was significantly associated with an increased risk of mortality in immunocompromised patients. Patients with acute leukemia and lymphoma had a greater mortality rate than non-immunocompromised patients (53.8% related mortality, P=0.05). Analysis of these patients showed frequent inadequate empirical therapy with ceftazidime plus amikacin in the presence of beta-lactam resistance.

CONCLUSIONS

Much of the burden of pneumococcal bacteremia was attributable to immunosuppressive diseases. In immunocompromised patients, pneumococcemia was frequently acquired within the hospital during the treatment of the underlying condition, and resistance to penicillin was common. Patients with acute leukemia and lymphoma who develop fever and pneumonia should be treated with drugs active against beta-lactam-resistant pneumococci, irrespective of the setting in which the infection develops.