Infections in patients with haematological malignancies

Abstract

Infectious complications are an important cause of morbidity and mortality in patients with hematological diseases, especially during chemotherapy induced neutropenia. This study was undertaken with the ultimate goal to improve the clinical management of infections in patients with hematological diseases.

In a relatively large hematology centre where antibiotic prophylaxis was generally not used, the incidence of bacteremia and the spectrum of causative pathogens during the period 1988-2001, were determined. No tendency to a shift from gram-negative to gram-positive etiology was noted. Except for a significant increase in Enterococcus faecium bacteremia, the species distribution was essentially stable. The rates of antimicrobial resistance were low. The 7- and 30-day mortality rates were 6.3 and 15.6%, respectively.

Species distribution and triazole susceptibility of colonising yeasts in neutropenic patients treated for hematological malignancies were prospectively analysed. Previous exposure to fluconazole prophylaxis was not associated with a shift from Candida albicans to non-albicans Candida species. However, fluconazole MICs were significantly higher in yeast isolates from patients who had received prophylaxis, indicating that azole prophylaxis may contribute to the emergence of less susceptible yeast strains.

Cefepime monotherapy was, in a prospective randomised study, found to be as safe and as effective as imipenem-cilastatin in the management of febrile neutropenia in patients treated for hematological malignancies. In patients with neutropenia and fever of unknown origin (FUO), the early discontinuation of all antibiotic therapy 48 hours after fever defervescence was not associated with an increased rate of fever relapse or mortality. This therapeutic approach seems promising in patients with FUO. Prospective large randomised trials are warranted.

The Multinational Association of Supportive Care in Cancer (MASCC) risk-index score was validated in patients with hematological malignancies. The risk-index identified patients with febrile neutropenia who were at low risk for the development of serious medical complication with a specificity, sensitivity and positive predictive value of 87%, 58% and 84%, respectively. Accordingly, this risk-index was found to be a valuable tool for the identification of low-risk patients.

A substantial proportion (36%) of low-risk patients with febrile neutropenia were ineligible for oral antibiotic therapy; while the remaining 64% received oral antibiotic treatment following discharge from the hospital 24 hours after fever defervescence. Upon final evaluation 95% remained afebrile and there was no early mortality in this group. The strategy of early hospital discharge with oral antibiotic therapy of carefully selected low-risk patients offered a feasible, safe and costeffective alternative to the conventional management of these patients.

Splenectomy is associated with increased risk to pneumococcal infections. Antibody response to vaccination with 23-valent pneumococcal polysaccharide vaccine in splenectomised patients with hematological disorders was determined. A substantial proportion (28%) of these patients had a poor antibody response to vaccination. During the long follow-up period (median observation time 7.5 years) all recorded episodes of pneumococcal infections were confined to poor responders. Patients who had a poor response to vaccination did not benefit from revaccination. With the possible exception of age, no other clinical characteristics could predict a poor antibody response. Poor responders should be offered other prophylactic measures.