Infections in the compromised child.

Children receiving chemotherapy for malignant diseases show different patterns of infection depending on their underlying disease and its therapy. Granulocytopenia carries the risk of bacterial infection, and also, if prolonged, of fungal infection. Impairment of cell-mediated immunity predisposes to infections with Pneumocystis carinii and is thought to be responsible for severe primary infections with varicella and measles, as well as the severe cytomegalovirus infections seen after allogeneic bone marrow transplantation. Absence or impairment of splenic function predisposes to overwhelming septicaemia with encapsulated organisms, while defects in the normal mechanical barriers to infection provide routes for bacterial and fungal invasion. Despite the lack of physical signs of a normal inflammatory response, clinical evaluation may be critical to the localization of infection in the immunocompromised child. Blood culture and biopsy remain pivotal investigations in the achievement of a microbiological diagnosis. Empirical treatment with a combination of antibiotics has been shown in comparative studies to be effective in initial management of the febrile neutropenic patient: continuing studies are evaluating the role of monotherapy and of different antibiotic combinations, particularly in the light of changing patterns of bacterial infections. Empirical antifungal therapy has been shown to be necessary for persistent or recurrent fever, particularly as persistent fungal infection may compromise the outcome of continuing cytotoxic therapy. Continuing uncertainties over many aspects of management of the infected immunocompromised child provide scope for clinical trials in parallel with trials evaluating new anticancer regimens. The use of new diagnostic methods, the role of prophylaxis, the most appropriate empirical regimen, the evaluation of new antimicrobial agents, all require careful evaluation for efficacy and safety. Perhaps the greatest dilemma of all is how far results from trials in adults can be extrapolated to paediatric practice.