Anda di halaman 1dari 4

CONCISE REVIEWS OF PEDIATRIC INFECTIOUS DISEASES

CONTENTS
Antimicrobial Use in Febrile Neutropenia

EDITORIAL BOARD
Co-Editors for this Issue: Charles R. Woods, MD, and Geoffrey A. Weinberg, MD
Board Members
John S. Bradley, MD Charles T. Leach, MD Geoffrey A. Weinberg, MD
Margaret C. Fisher, MD Deborah Lewinsohn, MD Leonard Weiner, MD
Barbara Jantausch, MD Debra L. Palazzi, MD Charles R. Woods, MD
Leonard R. Krilov, MD Jennifer Read, MD Theoklis Zaoutis, MD

The Use of Antimicrobial Agents in Children With Fever During


Chemotherapy-induced Neutropenia
The Importance of Risk Stratification
Debra L. Palazzi, MD

Key Words: fever, antimicrobial agents, promptly receive empirical antimicrobial fection is the most common cause of micro-
neutropenia, cancer therapy. biologically documented FN. Bacteremia,
(Pediatr Infect Dis J 2011;30: 887 890) which often is catheter related, occurs in
DEFINITIONS approximately 10% to 20% of febrile neutro-
Fever, generally, is defined as a single penic patients after HSCT or aggressively
T he Infectious Diseases Society of Amer-
ica recently published updated clinical
practice guidelines for the management of
oral temperature of at least 38.3C (101F),
or a temperature of 38C (100.4F) sus-
treated acute leukemia or lymphoma.5,9 11
Other frequently involved sites of bacterial
tained over 1 hour. Axillary temperatures are infection include the respiratory tract, gastro-
neutropenic fever.1 Despite improvements in intestinal tract, and skin and soft tissues.
therapeutic strategies leading to better prog- discouraged and rectal temperatures are con-
traindicated in children with chemotherapy- Enterobacteriaceae are common
nosis for most pediatric patients with malig-
induced febrile neutropenia (FN). Neutrope- causes of documented infection in cancer-
nancy, infection remains a significant cause
nia is defined as an absolute neutrophil count related FN due to translocation of organisms
of morbidity and mortality in those with
neutropenia caused by chemotherapy.2 4 In (ANC) of 500 cells/mm3 or an ANC from the gastrointestinal tract into the blood-
neutropenic children with cancer and poten- 1000 cells/mm3, which is expected to de- stream. Depending on local epidemiology,
tially life-threatening infection, fever may be crease to 500 cells/mm3 in the next 48 some of these pathogens can be multiply
the only initial manifestation of illness be- hours. Profound neutropenia is defined as an drug-resistant (eg, extended-spectrum beta-
cause other signs of systemic disease can be ANC 100 cells/mm3, and prolonged neu- lactamase producers). However, although
attenuated. Therefore, it is recommended tropenia is that which lasts 7 days. gram-negative pathogens were the predomi-
that all children with fever and cancer-re- nant cause of infection in febrile neutropenic
lated neutropenia be assessed with a com- EPIDEMIOLOGY AND ETIOLOGY patients in past decades, gram-positive or-
plete history and physical examination, un- Fever occurs in approximately 35% ganisms now are equally, and in some stud-
dergo appropriate diagnostic studies, and (range: 10% 60%) of episodes of neutrope- ies more, responsible.5,8,12,13 The increased
nia in children, and is most common in incidence of gram-positive pathogens as eti-
patients after hematopoietic stem cell trans- ologic agents is the result of increased use of
From the Section of Infectious Diseases, Department plantation (HSCT) or aggressive myeloabla- invasive devices, such as central venous
of Pediatrics, Baylor College of Medicine and tive treatment for acute leukemia or lym- catheters, and extensive use of broad-spec-
Texas Childrens Hospital, Houston, TX. phoma.5 Documentation of an infectious trum antimicrobials.1,14
The author has no funding or conflicts of interest to Coagulase-negative staphylococci
disclose.
etiology or focus of infection occurs in only
Reprints: Debra L. Palazzi, MD, Department of Pedi- 10% to 45% of febrile episodes.5 8 How- account for the majority of gram-positive
atrics, Baylor College of Medicine, Texas Chil- ever, infection remains the most common organisms isolated, but Staphylococcus au-
drens Hospital, 1102 Bates, Suite 1120, Houston, potentially life-threatening complication of reus (methicillin-resistant and -suscepti-
TX 77030. E-mail: dpalazzi@bcm.edu.
Copyright 2011 by Lippincott Williams & Wilkins
cytotoxic chemotherapy.2,3 ble) and vancomycin-susceptible and van-
ISSN: 0891-3668/11/3010-0887 The majority of infections is caused comycin-resistant enterococci also occur.1
DOI: 10.1097/INF.0b013e3182311343 by bacterial pathogens, and bloodstream in- In patients with significant mucositis

The Concise Reviews of Pediatric Infectious Diseases (CRPIDS) topics, authors, and contents are chosen and approved indepen-
dently by the Editorial Board of CRPIDS.

The Pediatric Infectious Disease Journal Volume 30, Number 10, October 2011 www.pidj.com | 887
Concise Reviews The Pediatric Infectious Disease Journal Volume 30, Number 10, October 2011

caused by aggressive chemotherapy, viri- Low Risk ommended as it offers no survival advantage
dans streptococci are important and poten- Patients are categorized as low risk for and fails to decrease the length of fever in
tially life-threatening pathogens.14 How- infection if they are clinically stable without most patients.27 However, in communities
ever, although gram-positive organisms medical comorbidities and have neutropenia with a high incidence of infections caused by
are equally common in patients with FN, expected to resolve within 7 days.1,20 Pa- methicillin-resistant staphylococci and in pa-
gram-negative pathogens, especially Pseu- tients with solid tumors generally have lower tients with hemodynamic instability, radio-
domonas aeruginosa, are associated with risk for infection than those with leukemia or graphically documented pneumonia, skin or
greater mortality.4 lymphoma. Any patient who does not clearly soft tissue infection, or concern for serious
Fungi (yeasts and molds) occur in less fulfill criteria for being at low risk of infec- catheter-related infection, addition of vanco-
than 5% of patients with FN but are impor- tion should be evaluated and treated accord- mycin to the empirical antimicrobial regimen
tant causes of morbidity and mortality.15,16 ing to high-risk guidelines.1 is reasonable. Its addition may not be neces-
These patients often have been treated with sary in patients with mucositis who are re-
prolonged courses of broad-spectrum antimi- ceiving monotherapy with piperacillin-tazo-
crobials and typically have neutropenic fever EMPIRIC ANTIBIOTIC THERAPY bactam, cefepime, or a carbapenem, since
that persists beyond 4 to 7 days or recurs these drugs provide adequate coverage
despite antibiotic therapy.16 Patients with High-risk Patients against viridans streptococci. In patients in
acute myelogenous leukemia (AML) and Febrile neutropenic patients at high whom vancomycin is warranted, if no patho-
those who have undergone HSCT are partic- risk of infection should be hospitalized and gens have been identified in cultures and no
ularly at risk. Candida species are the most administered broad-spectrum, empirical in- focus of infection is obvious, empirical ther-
common pathogens causing invasive fungal travenous antibiotic therapy.1,4 Choice of apy should be reassessed after 48 to 72
infections (IFI) in febrile neutropenic chil- initial antimicrobial therapy should be based hours. Vancomycin should be discontinued
dren.5,15 Invasive mold infections (aspergil- on history and physical examination findings if no obvious indication for ongoing treat-
losis, zygomycosis, fusariosis) are much less as well as local (and the patients own) ment is found.
common and occur most often in children epidemiology and antimicrobial susceptibil-
with profound neutropenia lasting more than ity data. Numerous clinical studies have not Low-risk Patients
10 days. established a single best therapeutic regi- A large meta-analysis in adults found
men for the empiric treatment of patients that oral quinolone therapy alone or in com-
with FN.1 Optimal empirical antibiotic bination with other antibiotics was an accept-
CLINICAL AND RISK ASSESSMENT therapy should include an agent(s) with able alternative to intravenous therapy in
Assessment of risk for severe infec- bactericidal activity, especially against select patients with FN.28 There was no dif-
tion is crucial in determining the appropriate Pseudomonas aeruginosa, and have a fa- ference in either mortality or treatment fail-
antimicrobial, route (intravenous vs. oral), vorable toxicity profile. ure rates in febrile neutropenic adults with
venue (inpatient vs. outpatient), and duration Monotherapy with an antipseudomo- cancer (other than acute leukemia) who were
of empirical antimicrobial therapy. The type nal beta-lactam agent has been shown to be hemodynamically stable without organ fail-
of malignancy and associated chemothera- as effective and as combination therapy in ure, pneumonia, catheter-related infection, or
pies, medical comorbidities, relevant history, patients who are hemodynamically stable with- severe soft-tissue infection. In adults, oral
and presenting signs and symptoms must be out evidence of skin or soft tissue infection, levofloxacin is the preferred fluoroquinolone
considered. In children, studies have sought pneumonia, or concern for catheter-related for monotherapy, because it has better activ-
to stratify risk for serious infection and its infection.2125 Current evidence supports the ity against gram-positive organisms than cip-
complications in febrile neutropenic patients, empirical use of piperacillin-tazobactam or rofloxacin.1 Selected low-risk adult patients,
with varying results.6,9,10,1719 cefepime in locations where antimicrobial who can be followed closely and in whom
susceptibility data do not warrant empirical prompt access to appropriate medical care is
High Risk carbapenem (meropenem or imipenem-cila- ensured, are candidates for outpatient man-
High-risk patients are those who have statin) use.23,25,26 Aztreonam can be used in agement of FN with oral antibiotic therapy.
an increased risk for severe infection or its children with a life-threatening beta-lactam Few studies exist in low-risk children
sequelae. Studies have found that children allergy but lacks activity against viridans with FN to support initial therapy with an
are at greater risk of life-threatening infec- streptococci (and thus should be used in oral agent.29 Considerable variation in prac-
tion if they have fever 39C, prolonged combination with an agent active against tice exists, but initial inpatient or emergency
neutropenia, recent intensive chemotherapy, these microbes in patients with severe mu- center management of low-risk patients with
underlying diagnosis of AML, bone marrow cositis). Routine initiation of an aminoglyco- FN remains routine in many (but not all)
involvement, relapsed malignancy or second side in combination with a beta-lactam agent centers. A prospective, randomized, single-
tumor, sick clinical appearance (hypoten- should be avoided because it offers no institution trial of 135 children with low-risk
sion, respiratory distress, hypoxemia, new- survival advantage and is associated with FN, most of whom had leukemia, found
onset abdominal pain, neurologic changes), increased toxicity and a higher treatment excellent clinical outcomes and no differ-
medical comorbidities (eg, renal or hepatic failure rate.21 Ceftazidime is no longer ence in treatment failure in children who
insufficiency), or evidence of focal infection recommended for empiric monotherapy in received oral ciprofloxacin versus intrave-
(mucositis, pneumonia, cellulitis, perianal high-risk patients with fever and neutrope- nous ceftriaxone after a single dose of
tenderness, presence of a central venous ac- nia due to its decreasing activity against ceftriaxone and amikacin.30 In the United
cess device).1,6,10,19 Laboratory abnormali- many gram-negative pathogens.26 Ceftazi- States, the use of oral fluoroquinolones
ties associated with a high risk of infection dime also offers only weak activity against empirically in low-risk children with FN
include profound neutropenia, a C-reactive viridans streptococci. and cancer is an off-label use.
protein 90 mg/L, platelet count 20,000 to Despite the prevalence of gram-posi- A recent study from Switzerland of
50,000 cells/mm3; these factors and an abso- tive organisms in the etiology of FN, routine more than 400 episodes of FN among 206
lute monocyte count 100 cells/mm3 also use of vancomycin or another agent active children reported that 8% of bacteremic ep-
predict a high risk for mortality.4,6,17 against gram-positive pathogens is not rec- isodes were identified only after inpatient

888 | www.pidj.com 2011 Lippincott Williams & Wilkins


The Pediatric Infectious Disease Journal Volume 30, Number 10, October 2011 Concise Reviews

reassessment at 8 to 24 hours; prediction of neutropenia compared empiric caspofungin against Aspergillus is recommended in pa-
bacteremia at the reassessment was better and liposomal amphotericin B and found no tients with a prior history of IFI, neutropenia
than that at presentation of FN.11 Hakim et difference in safety, tolerability, and efficacy anticipated to last beyond 2 weeks, or pro-
al19 also highlighted the importance of both between the treatment groups.33 The use of 2 longed neutropenia prior to HSCT.35 Po-
the initial clinical assessment and 24 hours of antifungals from different classes of mold- saconazole prophylaxis is recommended spe-
close observation in detecting invasive bac- active agents (combination therapy) can be cifically for patients aged more than 12 years
terial infection or a clinical complication in considered in patients with documented IFI undergoing induction chemotherapy for
low-risk patients. Often, outpatient manage- who are clinically unstable or worsening, but AML or intensive chemotherapy for myelo-
ment of low-risk children with FN is not this approach is controversial and no sup- dysplastic syndrome.1 Posaconazole absorp-
feasible because of nonmedical barriers, in- portive efficacy data exist from well-con- tion is variable, and drug interactions with
cluding transportation issues, language bar- trolled trials. chemotherapeutic agents are a potential con-
riers, and lack of parent or physician comfort cern.1 The duration of antifungal prophylaxis
with care outside of the hospital.31 Low-risk Patients is uncertain; however, patients with acute
Empirical antifungal therapy is not leukemia generally are treated until myeloid
Duration of Antibiotic Therapy recommended for low-risk patients with che- reconstitution, and HSCT recipients through
In a child with a documented focus of motherapy-induced FN. day 75 posttransplantation or cessation of
infection, the duration of therapy is deter- immunosuppression.1
mined by the site of the infection, the clinical Duration of Antifungal Therapy
response (including resolution of fever) and The duration of antifungal therapy in a Antivirals
the pathogen identified. In general, directed child with a documented focus of infection is Patients who are seropositive for her-
antimicrobial therapy should continue at determined by the site of the infection, the pes simplex virus undergoing induction ther-
least until the ANC exceeds 500 cells/mm3 clinical response (including resolution of fe- apy for leukemia or allogeneic HSCT war-
(bone marrow recovery). When the source of ver and results of imaging studies and serum rant acyclovir prophylaxis. Prophylaxis
the fever remains unknown, the initial choice fungal assays) and the pathogen identified. In generally is continued until bone marrow
of antibiotic therapy should be continued general, directed antifungal therapy should recovery. However, in patients with recur-
until bone marrow production of granulo- continue at least until bone marrow recovery. rent herpes simplex virus infections or graft-
cytes has recovered. When the source of the fever is unknown, versus-host disease, prolonged antiviral pro-
antifungal therapy should continue until res- phylaxis may be necessary. All patients with
ANTIFUNGAL THERAPY olution of fever and bone marrow recovery. cancer and their household contacts should
Longer durations of therapy may be war- be immunized annually with influenza vac-
High-risk Patients ranted when there is evidence of persisting cine. Further, neutropenic patients exposed
Initiation of antifungal therapy is rec- fungal infection after marrow recovery and to influenza should receive prophylaxis with
ommended in high-risk children with chemo- neutropenia is likely to recur soon due to the a neuraminidase inhibitor (eg, zanamivir, os-
therapy-induced neutropenia when fever per- next chemotherapy cycle. eltamivir). Neutropenia is not a risk factor
sists or recurs at 4 to 7 days.1 Candida for reactivation of other herpes viruses such
(yeast) and Aspergillus (mold) are the most as cytomegalovirus and human herpes virus
ANTIMICROBIAL PROPHYLAXIS
common pathogens causing IFI, although in- 6. No other antiviral prophylaxis regimens
fection due to mold is most common in Antibiotics have been proven effective for patients with
patients with profound neutropenia and fever The Infectious Diseases Society of chemotherapy-induced FN.
lasting beyond 7 to 10 days.1,32 America recommends considering fluoro-
Limited data are available comparing quinolone prophylaxis for high-risk afebrile SUMMARY
antifungal agents in children with suspected adult patients with anticipated prolonged and Children with fever and chemotherapy-
IFI.33,34 Important considerations in choos- profound neutropenia, although some con- induced or cancer-associated neutropenia
ing an antifungal agent include type of anti- troversy remains regarding precisely which should be assessed with complete history and
fungal prophylaxis (if any) and local epide- patients are most appropriate for prophy- physical examinations, undergo appropriate
miology of Candida isolates. In patients who laxis.1 However, no clear recommendation is diagnostic studies, and promptly receive
have not been given fluconazole prophylaxis found for children since high-quality clinical broad-spectrum empirical antimicrobial ther-
and lack risk factors for mold infection, trials assessing the safety and efficacy of apy. Assessment of risk for severe infection
fluconazole may be an appropriate empiric prophylaxis are lacking, and concern regard- is crucial in determining the appropriate an-
therapy. For those in whom there is con- ing increases in infections due to fluoroquin- timicrobial, route, venue, and duration of
cern for IFI caused by fluconazole-resistant olone-resistant pathogens exists. In general, empirical antimicrobial therapy and need for
Candida species (Candida krusei or Can- antibiotic prophylaxis (with fluoroquinolo- prophylactic antimicrobial agents.
dida glabrata) or mold, amphotericin B, nes) in children is considered only in very
echinocandins (eg, caspofungin), and newer REFERENCES
high-risk situations (eg, allogeneic HSCT,
azoles (voriconazole or posaconazole) are aggressive myeloablative therapy for leuke- 1. Freifeld AG, et al. Clinical practice guideline for
the use of antimicrobial agents in neutropenic
options. One should consider choosing an mia). patients with cancer. Clin Infect Dis. 2011;52:
empirical antifungal (preferably intravenous) e56 e93.
agent that is in a different class from that Antifungals 2. Walsh TJ, et al. Infectious complications in pe-
used for prophylaxis. Antifungal prophylaxis (eg, flucona- diatric cancer patients. In: Pizzo PA, Poplack DG,
Children and adults are less likely to zole, itraconazole, voriconazole, posacona- eds. Principles and Practice of Pediatric Oncol-
develop nephrotoxicity or infusion-related zole, caspofungin) against Candida species ogy. 5th ed. Philadelphia, PA: Lippincott Wil-
liams & Wilkins; 2006:1269 1329.
effects with liposomal amphotericin B than is recommended in those patients most at
3. Paganini HR, et al. A prospective, multicentric
other amphotericin formulations.32,34 One risk for IFI, such as allogeneic HSCT recip- scoring system to predict mortality in febrile
randomized, multicenter study of 82 children ients and patients undergoing intensive che- neutropenic children with cancer. Cancer. 2007;
aged 2 to 17 years with persistent fever and motherapy for acute leukemia. Prophylaxis 109:25722579.

2011 Lippincott Williams & Wilkins www.pidj.com | 889


Concise Reviews The Pediatric Infectious Disease Journal Volume 30, Number 10, October 2011

4. Santolaya ME, et al. Admission clinical and lab- 15. Castagnola E, et al. Fungal infections in children 26. Paul M, et al. Anti-pseudomonal beta-lactams for
oratory factors associated with death during a with cancer. Pediatr Infect Dis J. 2006;25:634 the initial, empirical, treatment of febrile neutro-
febrile neutropenic episode. Pediatr Infect Dis J. 639. penia. Cochrane Database Syst Rev. 2010;11:
2007;26:794 798. 16. Villarroel M, et al. Risk factors associated with CD005197.
5. Castagnola E, et al. A prospective study on the invasive fungal disease in children with cancer 27. Paul M, et al. Additional anti-gram-positive anti-
epidemiology of febrile episodes during chemo- and febrile neutropenia. Pediatr Infect Dis J. biotic treatment for febrile neutropenic cancer
therapy-induced neutropenia in children. Clin In- 2010;29:816 821. patients. Cochrane Database Syst Rev. 2005:
fect Dis. 2007;45:1296 1304. 17. Santolaya ME, et al. Prospective, multicenter CD003914.
6. Badiei Z, et al. Risk factors associated with life- evaluation of risk factors associated with invasive 28. Vidal L, et al. Oral versus intravenous antibiotic
threatening infections in children with febrile bacterial infection in children with cancer, neu- treatment for febrile neutropenia in cancer pa-
tropenia and fever. J Clin Oncol. 2001;19:3415 tients. Cochrane Database Syst Rev. 2004:
neutropenia J Pediatr Hematol Oncol. 2011;
3421. CD003992.
33:e9 e12.
18. Santolaya ME, et al. Prospective evaluation of 29. Gupta A, et al. Randomized controlled trial com-
7. Bakhshi S, et al. Infections in childhood acute prediction of invasive bacterial infection risk paring oral amoxicillin-clavulanate and ofloxacin
lymphoblastic leukemia. Pediatr Hematol Oncol. among children with cancer, fever and neutrope- with intravenous ceftriaxone and amikacin. J Pe-
2008;25:385392. nia. Clin Infect Dis. 2002;35:678 683. diatr Hematol Oncol. 2009;31:635 641.
8. Klastersky J, et al. Bacteraemia in febrile neutro- 19. Hakim H, et al. Risk prediction in pediatric can- 30. Paganini H, et al. Outpatient, sequential, paren-
penic cancer patients. Int J Antimicrob Agents. cer patients with fever and neutropenia. Pediatr teral-oral antibiotic therapy for lower risk febrile
2007;30(suppl 1):S51S59. Infect Dis J. 2010;29:5359. neutropenia in children. Cancer. 2003;97:1775
9. Baorto EP, et al. Clinical parameters associated 20. Klaassen RJ, et al. Low-risk prediction rule for 1780.
with low bacteremia risk in 1100 pediatric oncol- pediatric oncology patients presenting with fever 31. Quezada G, et al. Medical and non-medical
ogy patients. Cancer. 2001;92:909 913. and neutropenia. J Clin Oncol. 2000;18:1012 barriers to outpatient treatment of fever and
10. Wicki S, et al. Risk prediction of fever in neutro- 1019. neutropenia in children with cancer. Pediatr
penia in children with cancer. Pediatr Blood Can- 21. Paul M, et al. Beta-lactam versus beta-lactam- Blood Cancer. 2007;48:273277.
cer. 2008;51:778 783. aminoglycoside combination therapy in cancer 32. Wiley JM, et al. Efficacy and safety of ampho-
patients with neutropenia. Cochrane Database tericin B lipid complex in 548 children and ado-
11. Agyeman P, et al. Predicting bacteremia in chil- Syst Rev. 2003:CD003038.
dren with cancer and fever in. neutropenia. Pedi- lescents with invasive fungal infections. Pediatr
atr Infect Dis J. 2011;30:e114 e119. 22. Uygun V, et al. Piperacillin/tazobactam versus Infect Dis J. 2005;24:167174.
cefepime for the empirical treatment of pediatric 33. Maertens JA, et al. A randomized, double-
12. Greenberg D, et al. Microbiological spectrum and neutropenia and fever . Pediatr Blood Can- blind, multicenter study of caspofungin versus
susceptibility of pathogens causing bacteraemia cer. 2009;53:610 614. liposomal amphotericin B for empirical therapy
in paediatric febrile neutropenic oncology pa- in pediatric patients with persistent fever and
23. Simon A, et al. Piperacillin-tazobactam in pedi-
tients . Int J Antimicrob Agents. 2005;25:469 neutropenia. Pediatr Infect Dis J. 2010;29:
atric cancer patients. Eur J Clin Microbiol Infect
473. Dis. 2007;26:801 806. 415 420.
13. El-Mahallawy HA, et al. Antibiotic resistance is 24. Corapcioglu F, et al. Monotherapy with pipera- 34. Blyth CC, et al. Antifungal therapy in infants and
associated with worse outcome in febrile neu- cillin/tazobactam versus cefepime as empirical children with proven, probable or suspected in-
tropenic children with cancer. Pediatr Blood therapy for febrile neutropenia. Pediatr Hematol vasive fungal infections. Cochrane Database Syst
Cancer. 2011;57:283288. Oncol. 2006;23:177186. Rev. 2010:CD0006343.
14. Danilatou V, et al. Three cases of viridans group 25. Viscoli C, et al. Piperacillin-tazobactam mono- 35. Segal BH, et al. Prevention and early treatment of
streptococcal bacteremia in children with febrile therapy in high-risk febrile and neutropenic can- invasive fungal infection in patients with cancer
neutropenia. Scand J Infect Dis. 2003;35:873 cer patients. Clin Microbiol Infect. 2006;12:212 and neutropenia and in stem cell transplant recip-
876. 216. ients. Clin Infect Dis. 2007;44:402 409.

890 | www.pidj.com 2011 Lippincott Williams & Wilkins

Anda mungkin juga menyukai