Abstract
OBJECTIVE
Current recommendations for febrile neutropenia (FN) include initiation of broad-spectrum antibiotics without clear indications of when or how to de-escalate or target therapy, especially in those without microbiologically defined bloodstream infections (MD-BSIs). The purpose of this study is to characterize a pediatric FN population, FN management, and identify the proportion of patients with MD-BSI.
METHODS
This study was a single-center, retrospective chart review of patients admitted to the University of North Carolina Children's Hospital between January 1, 2016, and December 31, 2019, with a diagnosis of FN.
RESULTS
There were 81 unique encounters included in this study. MD-BSI was the etiology of fever in 8 FN episodes (9.9%). The most common empiric antibiotic regimen was cefepime (62%) followed by cefepime and vancomycin (25%). The most common de-escalation type was the discontinuation of vancomycin (83.3%), and the most common type of escalation was the addition of vancomycin (50%). The median antibiotic total duration in patients without MDI-BSI was 3 days (IQR, 5–9).
CONCLUSIONS
In this single-center, retrospective review, most FN episodes were not due to an MD-BSI. There were inconsistencies in practice of when discontinuation of antibiotic therapy occurred in patients without MD-BSI. De-escalation or cessation of antibiotic therapy before neutropenia resolution did not result in any documented complication. These data suggest a role for implementing an institutional guideline to improve consistency in antimicrobial use in pediatric patients with febrile neutropenia.
Keywords: antimicrobial stewardship, bloodstream infection, febrile neutropenia
Introduction
Febrile neutropenia (FN) represents one of the most common oncologic emergencies in pediatric patients receiving cytotoxic chemotherapy. It is defined as an absolute neutrophil count (ANC) of less than 500 cells/μL and a single temperature of at least 38.3°C, or persistent temperature of 38°C for >1 hour.1,2 It is estimated that 10% to 50% of patients with solid tumors, and >80% of those with hematologic malignancies will develop a fever during at least 1 chemotherapy cycle associated with neutropenia.3,4 This is especially concerning in patients who are expected to have profound neutropenia (ANC <100 cells/mm3) and neutropenia lasting ≥7 days, such as those receiving induction chemotherapy for acute leukemia.4 Fever may be the only sign of infection a patient exhibits and it is not always possible to identify a clear source or pathogen; however, FN warrants prompt recognition and treatment with antimicrobials to reduce the risk of mortality.5,6
Monotherapy with an antipseudomonal penicillin, fourth-generation cephalosporin, or carbapenem is currently recommended as empiric antibiotic therapy in pediatric FN pending microbiologic identification and susceptibility data.2 Antibiotic treatment for FN may be tailored to address commonly implicated organisms, such as Escherichia coli, Staphylococcus spp, Streptococcus spp, and Pseudomonas aeruginosa, yet many microbiologic investigations fail to identify a clear pathogen or source of infection.1,2,7 In the absence of this information, clinicians reach for empiric broad-spectrum antimicrobial regimens and use clinical markers, such as defervescence and ANC count recovery, to guide treatment duration. Continuing antimicrobial therapy until ANC ≥500 cells/mm3 represents a traditional treatment end point in current guidance for patients with clinically or microbiologically documented infections.1–3 In addition, this end point often defines the care of neutropenic patients without an identified microbiologically defined bloodstream infection (MD-BSI) because of the paucity of definitive guidelines or recommendations for discontinuing antibiotics if a source of fever is not found. Because ANC recovery may take upwards of 7 days in patients with prolonged neutropenia, broad-spectrum antimicrobials may be used long after a patient's defervesce.3 Unfortunately, long courses of treatment are not benign because of the risks of adverse effects, drug toxicity, and the development of antimicrobial resistance.
The purpose of this study is to characterize the management of the pediatric FN population at a children's hospital within a large academic medical center and identify the proportion of patients with MD-BSI. In addition, components of empiric antibiotic therapy, initial modification of empiric therapy, and duration of antibiotics were examined to evaluate the general treatment course of these patients.
Materials and Methods
This study was a single-center, retrospective chart review of patients admitted to the University of North Carolina Children's Hospital between January 1, 2016, and December 31, 2019, with a diagnosis of FN. Patients were identified from a report generated from the electronic medical record. Inclusion criteria were: 1) admission diagnosis with FN secondary to underlying oncologic diagnosis, 2) age 0 to 21 years, 3) ANC <500 cells/mm3, 4) single temperature ≥38.3°C or persistent temperature of 38°C for >1 hour, and 5) blood cultures drawn before antibiotic initiation. Our institution routinely admits all pediatric patients presenting with FN. The primary outcome of this study was the proportion of MD-BSI among pediatric patients with FN. Secondary outcomes included empiric antibiotic therapy, initial modification of empiric therapy, and duration of antibiotic therapy in patients without MD-BSI.
Following the Infectious Disease Society of America (IDSA) and American Society of Clinical Oncology (ASCO) guidelines,1,2 our institutional practice is to initiate cefepime (50 mg/kg intravenously every 8 hours) as empiric therapy for patients presenting with FN. Blood cultures are primarily obtained from the patient's central venous catheter (CVC) only. It is standard practice that all patients receive Pneumocystis jirovecii prophylaxis during chemotherapy. The addition of an empiric glycopeptide is reserved for patients with suspected central line infection, skin and soft tissue infections, mucositis, or those presenting with clinical instability. Typical practice within our institution is to continue antibiotics until recovery of ANC >500 cells/mm3 and resolution of fever.
Baseline patient data collected included age, sex, weight, primary oncologic diagnosis, stage in therapy, ANC and body temperature on admission, and presence of a CVC. Additional clinical data collected included duration of fever, duration of neutropenia, and whether blood cultures were obtained from a central or peripheral catheter. Antibiotic administration information was also collected, including choice of empiric antibiotic, initial changes to antibiotic therapy (escalation, de-escalation, or no modification), and duration of antibiotic therapy. Escalation of therapy was defined as the addition of a glycopeptide, aminoglycoside, or a switch to an agent with an expanded spectrum of activity compared with the empiric therapy. De-escalation was defined as narrowing the spectrum of antibiotic therapy compared with the empiric regimen. No modification was defined as an empiric regimen remaining unchanged until the cessation of FN therapy.
Descriptive analyses were used to assess end points using Microsoft Excel. We presented descriptive statistics as median (IQR) for continuous variables and as proportion (%) for categoric variables.
Results
A total of 665 patient encounters were screened for inclusion. Of these, 569 encounters were excluded for not meeting body temperature or ANC criteria for FN based on IDSA and ASCO guidelines. There were 29 unique patient encounters excluded because the blood cultures were drawn after antibiotic administration. A total of 81 encounters of FN episodes met inclusion criteria. Of the 81 encounters, there were 64 unique patients consisting of 50 patients with only 1 encounter and 14 patients with at least 2 or more admission encounters. Baseline patient characteristics are summarized in Table 1.
Table 1.
Baseline Demographic and Clinical Characteristics *
| Variable | Result |
|---|---|
| Age, median (IQR), yr | 7 (4–12.5) |
| Weight, median (IQR), kg | 22.8 (16.4–39.6) |
| Sex, male, n (%) | 42 (51.8) |
| ANC, median (IQR), cells/μL | 100 (0–200) |
| Body temperature, median (IQR), Celsius | 38.7 (38.2–39.2) |
| Duration of fever, median (IQR), days | 2 (1–3) |
| Duration of neutropenia, median (IQR), days | 4 (2–7) |
| Presence of central venous catheter, n (%) | 79 (97.5) |
| Primary oncologic diagnosis, n (%) | |
| Liquid | 44 (54.3) |
| Acute lymphoblastic leukemia | 38 (46.9) |
| Hodgkin lymphoma | 3 (3.7) |
| Burkitt lymphoma | 2 (2.5) |
| Grey zone lymphoma | 1 (1.2) |
| Solid | 37 (45.7) |
| Ewing sarcoma | 10 (12.3) |
| Osteosarcoma | 7 (8.6) |
| Neuroblastoma | 6 (7.4) |
| Rhabdomyosarcoma | 4 (4.9) |
| Wilms tumor | 3 (3.7) |
| Other | 7 (8.6) |
| Stage in treatment, n (%) | |
| Induction | 15 (18.5) |
| Consolidation | 7 (8.6) |
| Delayed intensification | 8 (9.9) |
| Maintenance | 23 (28.4) |
| Staging not indicated | 28 (34.6) |
ANC, absolute neutrophil count
* Data derived from 81 patient encounters in 64 unique patients.
The MD-BSI was the etiology of fever in 9.9% (8 of 81) of FN episodes (Table 2), in all of which the patients presented with a previously placed CVC. Most patients had blood cultures drawn from the CVC only (76 of 81). A total of 7 of the 8 patients with MD-BSI had blood cultures drawn from the CVC. One patient had a blood culture drawn peripherally, resulting in a positive blood culture growing Staphylococcus capitis. Most of the patients with MD-BSI (46.9%) were those with a diagnosis of acute lymphoblastic leukemia (ALL). None of the organisms that caused MD-BSI required vancomycin therapy or expressed methicillin resistance. The most common empiric antibiotic regimen was cefepime (62%), followed by cefepime and vancomycin (25%). Other empiric therapies used are described in the Figure.
Table 2.
Proportion of Microbiologically Defined Bloodstream Infection and Species Isolated *
| Variable | Result, n (%) |
|---|---|
| Yes | 8 (9.9) |
| MSSA | 2 (25) |
| Citrobacter koseri | 1 (12.5) |
| Staphylococcus capitis | 1 (12.5) |
| Escherichia coli | 1 (12.5) |
| E coli and Bacillus cereus | 1 (12.5) |
| Neisseria spp | 1 (12.5) |
| Candida spp | 1 (12.5) |
| No | 73 (90.1) |
MSSA, methicillin-susceptible Staphylococcus aureus
* Data derived from 81 patient encounters in 64 unique patients.
Figure.
Initial empiric antibiotic regimens. Data derived from 81 patient encounters in 64 unique patients. Other: cefepime/vancomycin/metronidazole; cefepime/vancomycin/levofloxacin; cefepime/clindamycin; levofloxacin.
Most patients categorized as de-escalation or no modification had a fever duration of 3 days or less (14 of 17 and 36 of 39, respectively). Of the 24 patients categorized as escalation, 11 (45.8%) of them had fever of at least 4 days. A total of 17 patients had an initial modification of empiric antibiotics classified as de-escalation, and 6 of them required re-escalation of therapy. No re-escalations were due to a new MD-BSI. The most common de-escalation type was the discontinuation of vancomycin (83.3%), and the most common type of escalation was the addition of vancomycin (50%). Nearly half of the patients (48.1%) did not have any modification of their empiric antibiotic regimen. Other modifications to empiric antibiotic therapy are described in Table 3.
Table 3.
Classification of Initial Change of Empiric Antibiotic Therapy *
| Classification | Result, n (%) |
|---|---|
| De-escalation | 17 (21.0) |
| Discontinuation of vancomycin | 10 (58.8) |
| Switch to agent with narrower spectrum of activity | 7 (41.2) |
|
| |
| Escalation | 24 (29.6) |
| Addition of anaerobic coverage | 2 (8.3) |
| Addition of antifungal agent | 2 (8.3) |
| Addition of vancomycin | 12 (50) |
| Addition/switch to agent with expanded spectrum of activity | 8 (33.3) |
| Cefepime to meropenem | 6 (75) |
| Addition of other agent† | 2 (25) |
|
| |
| No modification | 39 (48.1) |
* Data derived from 81 patient encounters in 64 unique patients.
† Other agents = doxycycline, daptomycin.
The median total duration of antibiotics in patients without MD-BSI (73 of 81) was 3 days (IQR, 5–9) as listed in Table 4. A total of 22 patients (30.1%) without MD-BSI had antibiotics discontinued at ANC resolution. However, 20 patients (27.4%) without MD-BSI continued on antibiotic therapy beyond ANC recovery; 25% continued for 5 days or more. More than a third of patients (42.5%) had antibiotics discontinued before ANC recovery with no documented complications.
Table 4.
Duration of Antibiotic Therapy in Patients Without Microbiologically Defined Bloodstream Infection (MD-BSI) *
| Duration | Result |
|---|---|
| Antibiotic duration, median (IQR), days | 3 (5–9) |
| Antibiotics discontinued at ANC resolution, n (%) | 22 (30.1) |
| Antibiotics discontinued prior to ANC resolution, n (%) | 31 (42.5) |
| Discharged prior to ANC resolution, n (%) | 22 (70.9) |
| Antibiotics continued after ANC resolution, n (%) | 20 (27.4) |
| <5 days | 15 (75) |
| ≥5 days | 5 (25) |
ANC, absolute neutrophil count
* Data derived from 73 patient encounters in 64 unique patients in which the patient did have an MD-BSI.
Discussion
In this single-center, retrospective review, most FN episodes were not due to an MD-BSI. The proportion of MD-BSI among children admitted with FN between 2016 and 2019 was 9.9%, a slightly lower proportion compared with that reported in the literature of 17% to 30%.5,6,8–10 Most patients were initiated on appropriate empiric therapy for FN based on guidelines and institutional recommendations. De-escalation of antibiotic therapy, specifically vancomycin, or cessation of antibiotic therapy before neutropenia resolution did not result in any documented complications in patients without MD-BSI. Of the patients categorized as de-escalation, 64.7% (11 of 17) had a primary oncologic diagnosis of ALL. Those categorized as escalation were more evenly divided among liquid and solid tumors (13 of 24 vs 11 of 24, respectively).
Discontinuation of antibiotics in patients without MD-BSI varied at this institution in terms of occurring before ANC recovery, at ANC recovery, or after. None of the 28% of patients who had antibiotic therapy extended beyond ANC resolution had an MD-BSI; however, data regarding infections other than MD-BSI were not collected. The results of this study suggest that implementation of a guideline targeting safe de-escalation of empiric antimicrobials, earlier than neutrophil recovery alone, in patients without MD-BSI would have a significant effect.
Antimicrobial stewardship programs (ASPs) have emerged as a system of strategies for the optimization of antimicrobial management. Recent evidence has sought to investigate how ASP affects broad-spectrum antimicrobial use, treatment outcomes, and costs in pediatric hematology-oncology populations.11,12 One single-center study found that the implementation of an ASP reduced the use of broad-spectrum antimicrobials and produced annual cost savings of nearly $60,000 without compromising the average length of hospital stay, mortality rate, or appropriateness of empirical treatment for bacteremia in pediatric hematology-oncology patients.12 Additional chart reviews of small cohorts of pediatric patients with leukemia receiving care for FN also suggest that therapy de-escalation based on antimicrobial susceptibility data does not result in complications.8,13–16,4 In fact, many patients who underwent de-escalation received shorter courses of antimicrobials than those who did not.8 Although studies have been published on the de-escalation of antibiotics in pediatric FN patients, including those with MD-BSI and before ANC recovery, there is no consensus on how to appropriately de-escalate pediatric patients with FN and no identified infectious source.8,13–16 In addition to these studies, some organizations, like ASCO, have provided guidance on risk stratification and antibiotic de-escalation strategies; however, hesitancy among pediatric oncology providers remains a barrier for strategy implementation.17
In a systematic review assessing the safety of protocol-guided discontinuation of antibiotics regardless of ANC, compared with continuation of antibiotics until ANC resolution, found no difference in all-cause mortality between the 2 groups. However, no strong safety conclusions could be made from this data.13,15,16 Similarly, Santolaya et al14 found that cessation of antimicrobials in children with FN and known respiratory viral infection minimized antibiotic exposure without complications. The patient populations evaluated in these studies were similar to the mix of liquid and solid tumors in our study, with the exception of Reinecke et al8 who included patients with ALL or AML only.
Our study identified several opportunities for quality improvement and guideline design within this institution regarding the management of pediatric patients with FN. Given the differences in practice regarding antibiotic discontinuation in patients without MD-BSI, a guideline may provide standardization of timely discontinuation of antibiotics upon neutrophil recovery, or prior to recovery, if appropriate. Additionally, based on the characterization of this FN population, and in accordance with the ASCO and IDSA guidelines, the results support the discontinuation of vancomycin as empiric therapy in FN patients without MD-BSI after 48 hours. Although these data suggest appropriate de-escalation of antibiotics, the risk of infection-related complications in neutropenic patients contributes greatly to physician hesitancy. Collaboration on guideline development and discussion of both generalized and local data are imperative to obtaining physician acceptance and successful implementation.
There were expected limitations to this study based on the retrospective and single-center design, including less external validity of the results. Additionally, this review focused solely on antibacterial agents and bacterial causes of bloodstream infections specifically. It is well known that patients receiving cytotoxic chemotherapy are at increased risk for fungal and viral infections and data for those infections were not collected during this study. Despite these limitations, these results may be used as guidance to minimize unnecessary antibiotic exposure in patients with prolonged neutropenia without MD-BSI.
Conclusion
In conclusion, this retrospective review demonstrates a slightly lower proportion of MD-BSI in our pediatric patients presenting with FN compared with reported literature. Additionally, there were inconsistencies in practice of the timing of discontinuation of antibiotic therapy in patients without MD-BSI. These results highlight an opportunity for developing a guideline to support the discontinuation of antibiotics before ANC recovery.
Acknowledgments
Pediatric Clinical Pharmacist Specialists at University of North Carolina Children's Hospital for their assistance. This was a resident project presentation at the virtual Pediatric Pharmacy Association Annual Conference on April 23, 2021.
ABBREVIATIONS
- ANC
absolute neutrophil count;
- ALL
acute lymphoblastic leukemia;
- ASCO
American Society of Clinical Oncology;
- ASP
antimicrobial stewardship programs;
- CVC
central venous catheter;
- FN
febrile neutropenia;
- IDSA
Infectious Diseases Society of America;
- IV
intravenous;
- MD-BSI
microbiologically defined bloodstream infection
Footnotes
Disclosures. The authors declare no conflicts or financial interest in any product or service mentioned in the manuscript, including grants, equipment, medications, employment, gifts, and honoraria. The authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Ethical Approval and Informed Consent. This retrospective evaluation was reviewed and approved through the University of North Carolina Institutional Review Board as Non-Human Subjects Research.
References
- 1.Freifeld AG, Bow EJ, Sepkowitz KA et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis . 2011;52(4):e56–93. doi: 10.1093/cid/cir073. [DOI] [PubMed] [Google Scholar]
- 2.Lehrnbecher T, Robinson P, Fisher B et al. Guideline for the management of fever and neutropenia in children with cancer and hematopoietic stem-cell transplantation recipients: 2017 update. J Clin Oncol . 2017;35(18):2082–2094. doi: 10.1200/JCO.2016.71.7017. [DOI] [PubMed] [Google Scholar]
- 3.Klastersky J. Management of fever in neutropenic patients with different risks of complications. Clin Infect Dis . 2004;39(suppl 1):S32–S37. doi: 10.1086/383050. [DOI] [PubMed] [Google Scholar]
- 4.Orasch C, Weisser M, Mertz D et al. Comparison of infectious complications during induction/consolidation chemotherapy versus allogeneic hematopoietic stem cell transplantation. Bone Marrow Transplant . 2010;45(3):521–526. doi: 10.1038/bmt.2009.187. [DOI] [PubMed] [Google Scholar]
- 5.Hakim H, Flynn PM, Knapp KM et al. Etiology and clinical course of febrile neutropenia in children with cancer. J Pediatr Hematol Oncol . 2009;31(9):623–629. doi: 10.1097/MPH.0b013e3181b1edc6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bochennek K, Hassler A, Perner C et al. Infectious complications in children with acute myeloid leukemia: decreased mortality in multicenter trial AML-BFM 2004. Blood Cancer J . 2016;6(1):e382. doi: 10.1038/bcj.2015.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gustinetti G, Mikulska M. Bloodstream infections in neutropenic cancer patients: a practical update. Virulence . 2016;7(3):280–297. doi: 10.1080/21505594.2016.1156821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Reinecke J, Lowas S, Snowden J, Neemann K. Blood stream infections and antibiotic utilization in pediatric leukemia patients with febrile neutropenia. J Pediatr Hematol Oncol . 2019;41(4):251–255. doi: 10.1097/MPH.0000000000001279. [DOI] [PubMed] [Google Scholar]
- 9.Calton EA, Le Doaré K, Appleby G et al. Invasive bacterial and fungal infections in paediatric patients with cancer: incidence, risk factors, aetiology and outcomes in a UK regional cohort 2009-2011. Pediatr Blood Cancer . 2014;61(7):1239–1245. doi: 10.1002/pbc.24995. [DOI] [PubMed] [Google Scholar]
- 10.Agyeman P, Kontny U, Nadal D et al. A prospective multicenter study of microbiologically defined infections in pediatric cancer patients with fever and neutropenia: Swiss Pediatric Oncology Group 2003 fever and neutropenia study. Pediatr Infect Dis J . 2014;33(9):e219–e225. doi: 10.1097/INF.0000000000000326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wolf J, Sun Y, Tang L et al. Antimicrobial stewardship barriers and goals in pediatric oncology and bone marrow transplantation: a survey of antimicrobial stewardship practitioners. Infect Control Hosp Epidemiol . 2016;37(3):343–347. doi: 10.1017/ice.2015.295. [DOI] [PubMed] [Google Scholar]
- 12.Horikoshi Y, Kaneko T, Morikawa Y et al. The north wind and the sun: pediatric antimicrobial stewardship program combining restrictive and persuasive approaches in hematology-oncology ward and hematopoietic stem cell transplant Unit. Pediatr Infect Dis J . 2018;37(2):164–168. doi: 10.1097/INF.0000000000001746. [DOI] [PubMed] [Google Scholar]
- 13.Hodgson-viden H, Grundy PE, Robinson JL. Early discontinuation of intravenous antimicrobial therapy in pediatric oncology patients with febrile neutropenia. BMC Pediatr . 2005;5(1):10. doi: 10.1186/1471-2431-5-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Santolaya ME, Alvarez AM, Acuña M et al. Efficacy and safety of withholding antimicrobial treatment in children with cancer, fever and neutropenia, with a demonstrated viral respiratory infection: a randomized clinical trial. Clin Microbiol Infect . 2017;23(3):173–178. doi: 10.1016/j.cmi.2016.11.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lehrnbecher T, Stanescu A, Kühl J. Short courses of intravenous empirical antibiotic treatment in selected febrile neutropenic children with cancer. Infection . 2002;30(1):17–21. doi: 10.1007/s15010-002-2094-1. [DOI] [PubMed] [Google Scholar]
- 16.Stern A, Carrara E, Bitterman R et al. Early discontinuation of antibiotics for febrile neutropenia versus continuation until neutropenia resolution in people with cancer. Cochrane Database Syst Rev . 2019;1(1):CD012184. doi: 10.1002/14651858.CD012184.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zimmer AJ, Freifeld AG. Optimal management of neutropenic fever in patients with cancer. J Oncol Pract . 2019;15:19–24. doi: 10.1200/JOP.18.00269. [DOI] [PubMed] [Google Scholar]