Predicting Gram-Negative Bacteremia in Febrile Neutropenic Patients Using IL-8
Predicting Gram-Negative Bacteremia in Febrile Neutropenic Patients Using IL-8
Abstract & Commentary
Synopsis: Serum IL-8 levels below 2000 pg/mL at the onset of fever define a subgroup of patients who are at low risk of having Gram-negative bacteremia, and they can therefore safely be treated with monotherapy.
Source: Kern WV, et al. Prediction of Gram-negative bacteremia in patients with cancer and febrile neutropenia by means of interleukin-8 levels in serum: Targeting empirical monotherapy versus combination therapy. Clin Infect Dis. 2001;32:832-835.
This was a prospective observational study of 133 neutropenic episodes occurring in 98 patients with haematological malignancies (71 patients) or cancer who had also become febrile. A sample of venous blood was taken once fever was registered, and before any antibiotics were given, blood was centrifuged within an hour of collection and IL-8 was detected using an automated immunoluminescence analyzer capable of completing the assay in 40 minutes. The assay was available around the clock so that the results could be available within 2 hours of the onset of fever. Kern and colleagues preselected a cut-off of 2000 pg/mL based on earlier studies and estimated that concentrations in excess of this would predict Gram-negative bacteraemia with a sensitivity of at least 40% and a specificity of at least 80%. They also assumed an incidence of 10% and a risk of death within 7 days of 3%. Hence, at least 122 patients with levels of £ 2000 pg/mL would be needed to provide adequate discriminatory power and to confirm a risk of < 1% of early death for patients treated with monotherapy while attaining an alpha level of 0.05 and a power of 0.9. Patients with IL-8 levels exceeding 2000 pg/mL were treated with 2 g q8h cefepime plus 6-7.5 mg/kg/d netilmicin given as a single daily dose. All other patients received the cephalosporin alone.
Fever was explained in 97 cases (70%) of which 48 involved bacteraemia. Gram-negative bacilli were isolated alone in 11 episodes and together with other bacteria in 4 episodes. IL-8 levels ranged from < 18 pg/mL to > 7500 pg/mL. The highest level was associated with Gram-negative bacteraemia, but concentrations ranged from 41 upward. In total, 11 patients fell into the high IL-8 level group and 3 died of their initial infection despite combination therapy (2 polymicrobial Gram-negative bacteraemia, 1 pneumonia and culture-negative septic shock), while only 1 of the 122 patients with low levels of the cytokine died of the initial infection (acute abdomen and culture-negative sepsis; OR 45.4; P = .002). The odds for a Gram-negative bacillus infection when IL-8 levels were high were 2 to 1 whereas the odds against this when IL-8 levels were low were 17 to 1. Surprisingly, treatment was still modified in 63 (52%) episodes treated with monotherapy and 7 (63%) of those treated with the combination suggesting confidence in the initial regimen was less than perfect.
Comment by J. Peter Donnelly, PHD
IL-8 is the latest in a long line of naturally occurring substances including C-reactive protein, IL-6, and procalcitonin that have been investigated for the ability of high levels to distinguish Gram-negative infection from other causes of fever in neutropenic patients. However, Kern et al were much more explicit in what they expected of the test than hitherto and used the results directly to decide whether a patient should be treated immediately with a combination of cefepime and netilmicin or only with the cephalosporin. The results clearly show that the strategy worked. Whether this experience would be repeated in a full-fledged randomized controlled trial in which IL-8 is measured in one group and not the other remains to be seen. The approach has obvious appeal and merits such a formal trial especially as the thorny issue of whether monotherapy with a broad-spectrum beta-lactam such as cefepime (and ceftazidime, piperacillin-tazobactam, imipenem, and meropenem for that matter) is adequate for febrile neutropenic patients has never been resolved. Treating every febrile neutropenic patient empirically with combined therapy with a beta-lactam and an aminoglycoside is clearly unnecessary given the low incidence of Gram-negative infections, currently around 10%. Most physicians prefer not to take what they perceive as unwarranted risks and continue to institute an empirical combination regimen to be on the safe side judging the benefit to be greater than the risk. They may be right given that once-daily dosing regimens for aminoglycosides are both more effective and safer than earlier schemes. However, there is no denying that the ability to restrict combination regimens to only those who need them would minimize the risk of adverse drug reactions, reduce the need for assays, and lower drug acquisition costs.
It is interesting to note that sensitivity and specificity of IL8 levels > 2000 pg/mL are in the same order as found for surveillance cultures aimed at detecting potential pathogens in oral and fecal samples before the onset of fever in neutropenic patients, namely 95% specificity and between 38% and 54% depending on the species involved.1 Perhaps had the earlier studies of surveillance cultures adopted the same approach as Kern et al, their value might have been better appreciated. I wonder whether centers that still use regular surveillance cultures to detect colonization by Gram-negative bacilli will now feel more encouraged to press for this information to be used in the same way as Kern et al advocate for IL-8. Certainly, knowing both the identity and susceptibility of a colonizing Gram-negative bacillus would give more of an edge than knowing the IL-8 concentration. However, I suspect that the simpler logistics of determining IL-8 would prove too irresistible to ignore. Only time will tell.
Reference
1. Guiot HF, et al. Selective antimicrobial modulation of the intestinal microbial flora for infection prevention in patients with hematologic malignancies. Scand J Infect Dis. 1986;18:153-160.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.