• febrile neutropenia;
  • infection;
  • interleukin-5;
  • interleukin-8

Paediatric oncology patients with febrile neutropenia are usually hospitalised and given empirical broad-spectrum antibiotic therapy as a number will later prove ‘culture positive’ (Freifeld et al, 1999). In ‘low risk’ febrile neutropenia, oral antibiotics or outpatient intravenous antibiotics are safe and effective (Mustafa et al, 1996; Abrahamsson et al, 1997) allowing less frequent hospitalisations.

However, there is currently no method to rapidly identify bacteraemia in these ‘low risk’ patients to allow stratification of treatment. Previous studies have investigated several inflammatory plasma cytokines in an attempt to define suitable markers; however, the predictive value of individual cytokines has generally been too low to influence management (Engel et al, 1998). We have previously successfully applied cytometric bead arrays (CBA) to screen multiple plasma cytokines to predict bacteraemia in neonatal infection (Hodge et al, 2004) and hypothesised such an approach may also provide a rapid and sensitive method of determining bacteraemia in febrile neutropenic patients.

Blood for C-reactive protein (CRP), haematological parameters, CBA analysis and blood cultures was collected from 31 paediatric oncology patients presenting with febrile neutropenia (absolute neutrophil counts <1·0 × 109/l and either a single temperature >38·5°C or a sustained temperature >38·0°C). Patients were divided into those with patients with positive blood cultures and those with negative cultures.

T-helper type 1 (Th1) and Th2 cytokines [interleukin (IL)-2, IL-4, IL-5, IL-10, interferon-γ and tumour necrosis factor (TNF)-α] and inflammatory cytokines (IL-12, TNFα, IL-10, IL-6, IL-1β and IL-8) were simultaneously determined as previously reported (Hodge et al, 2004) using CBA kits and software provided by BD Biosciences (San Jose, CA, USA). Reference ranges for plasma cytokine levels were established using plasma samples from the culture-negative patients. Positive blood cultures were obtained in eight of the 31 episodes of febrile neutropenia. A variety of organisms were isolated from this patient group (Table I). There was a significant increase in IL-5 in four of eight patients with proven infection (Table I). Overall, there was a significant increase in IL-5 levels in the culture-positive group (39 ± 29·5 pg/ml) compared with the culture-negative group (8·5 ± 4 pg/ml) (P = 0·047 Mann–Whitney). There was a significant increase in IL-8 in the remaining four of eight culture-positive patients compared with the culture-negative group. Plasma IL-12 levels were significantly reduced in the culture-positive compared with the culture-negative group (Table I) (P = 0·047). There was no significant difference with any other cytokines between these two groups. Three of 23 culture-negative patients had IL-5 and/or IL-8 levels outside the 95% confidence limits (mean ± 2 SD) for this patient group. There was no correlation between the culture-positive group and CRP levels (Table I) or haematological parameters (absolute neutrophil count, white blood cell count, differential count or morphology) (data not shown). These novel findings need confirmation with a larger study cohort.

Table I.  Plasma cytokine (pg/ml) and CRP (mg/l) levels for culture-negative and culture-positive patient groups (significantly elevated results in bold).
Patient infected group
  1. CRP, C-reactive protein; ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia.

Pseudomonas aeruginosaNeuroblastoma3230042
Enterobacter cloacaeRhabdomyosarcoma6296012
Staphylococcus aureusNeuroblastoma18510118
Streptococcus vestibularisMetastatic ependymoma176158064
Streptococcus vestibularisALL (relapsed)34700<5
Bacteroides distasonisAML112310240
Escherichia coliNeuroblastoma11333010
Brevibacterium caseiAML247205
Patient control group: reference range<17<220<27<138


  1. Top of page
  2. References
  • Abrahamsson, J., Pahlman, M. & Mellander, L. (1997) Interleukin 6, but not tumour necrosis factor-alpha, is a good predictor of severe infection in febrile neutropenic and non-neutropenic children with malignancy. Acta Paediatrica, 86, 10591064.
  • Engel, A., Mack, E., Kern, P. & Ker, WV. (1998) An analysis of interleukin-8, interleukin-6 and C-reactive protein serum concentrations to predict fever, gram-negative bacteremia and complicated infection in neutropenic cancer patients. Infection, 126, 213221.
  • Freifeld, A., Marchigiani, D., Walsh, T., Chanock, S., Lewis, L., Hiemenz, J., Hiemenz, S., Hicks, J. E., Gill, V., Steinberg, S.M. & Pizzo, P.A. (1999) A double blind comparison of empirical oral and intravenous antibiotic therapy for low risk febrile patients with neutropenia during cancer chemotherapy. New England Journal of Medicine, 341, 305311.
  • Hodge, G., Hodge, S., Haslam, R., McPhee, A., Sepulveda, H., Morgan, E., Nicholson, I. & Zola, H. (2004) Rapid detection of neonatal sepsis by simultaneous measurement of multiple cytokines using 100 μL sample volumes. Clinical and Experimental Immunology, 137, 402407.
  • Mustafa, M.M., Aquino, V.M., Pappo, A., Tkaczewski, I. & Buchanan, G.R. (1996) A pilot study of outpatient management of febrile neutropenic children with cancer at low risk of bacteremia. Journal of Pediatrics, 128, 847849.