Managing Febrile Neutropenia in Pediatric Oncology ...sites.utexas.edu/pharmacotherapy-rounds/files/2017/...oncology patients with febrile neutropenia. 2005. ii. Lehrnbecher T, et

Managing Febrile Neutropenia in Pediatric Oncology Patients: Wise Antimicrobial Stewardship

Lincy Varughese, Pharm.D. PGY1 Pharmacy Resident

The Children’s Hospital of San Antonio, San Antonio, Texas Division of Pharmacotherapy, The University of Texas at Austin College of Pharmacy

Pharmacotherapy Education and Research Center University of Texas Health Science Center at San Antonio

February 10, 2017

Learning Objectives

1. Understand the basics of febrile neutropenia2. Review the current guidelines and institutional practices on the management of febrile neutropenia3. Evaluate literature on the de-escalation of antimicrobial therapy in pediatric febrile neutropenia4. Given a patient case scenario, apply the literature, guidelines, and outside institutional resources to determine

appropriate therapy

Varughese 2

I. Febrile Neutropenia A. What is febrile neutropenia?

i. Background1 1. Febrile neutropenia is an oncologic emergency 2. Spectrum of infection is influenced by:

a. Nature and intensity of chemotherapy b. Antimicrobial prophylaxis c. Catheters and other medical devices

ii. Pathophysiology:2 1. Decreased production of white blood cells in the bone marrow 2. Chemotherapy causes disruption of the integrity of the gastrointestinal mucosa predisposing

patients to severe invasive infections a. Gut flora (also from mouth/esophagus/nose) translocates into the blood stream b. Reduced hematopoiesis leads to pancytopenia:

i. Anemia ii. Neutropenia

iii. Thrombocytopenia iii. Definition:3

1. Fever: a. Single oral temperature measurement of >38.3° C (101° F) or a temperature of

>38.0° C (100.4° F) sustained over a one hour period 2. Neutropenia:

a. Absolute neutrophil count (ANC) of <500 cells/mm3 or an ANC that is expected to decrease to <500 cells/mm3 during the next 48 hours

iv. Population at risk:4 1. It is an important cause of morbidity and mortality in oncology patients receiving cytotoxic

antineoplastic therapy a. With aggressive management in the last few decades mortality has fallen from 30–

40% to 1%. Table 1: Pathogens of concern5

II. Management of Febrile Neutropenia A. Guideline Based Management Recommendations:

i. Risk Stratification: 1. Factors to consider: patient, treatment, and episode specific (Table 2) 2. International risk stratification scoring systems have been included in the Journal of Clinical

Oncology guidelines (Table 3) a. Notable variations across countries and institutions

3. These are not necessarily adopted in all pediatric institutions and some institutions may have their own “scoring system” to stratify patients into high and low risk categories

Gram Positive Gram Negative Anaerobes Fungal species Coagulase-negative

Staphylococcus aureus (including MRSA)

Enterococcus species (including VRE)

Viridans group streptococci

Streptococcus pneumonia

Streptococcus pyogenes

Escherichia coli Klebsiella species Enterobacter species Pseudomonas aeruginosa Citrobacter species Acinetobacter species Stenotrophomonas

maltophilia

Bacteriodes fragelis Pneumocystis jiroveci Candida spp. Aspergillosis Zygomycosis

Varughese 3

Table 2: Risk Stratification Factors to Consider1

Table 3: Validated pediatric risk stratification strategies for low-risk patients6-11

ii. Initial Management of FN1

1. Low-risk FN: a. Appropriate oral agents for gram positive and gram negative coverage include:

i. Cefixime OR ii. Fluoroquinolone + amoxicillin-clavulanate

b. Consider initial or step-down outpatient management if infrastructure is in place to ensure careful monitoring and follow-up

c. Consider oral antimicrobial administration if child is able to tolerate this route of administration reliably

2. High-risk FN: a. Empiric monotherapy with an antipseudomonal β-lactam or carbapenem b. Reserve addition of a second Gram-negative agent or glycopeptide for patients

when: i. Clinically unstable OR

ii. Suspicion exists for a resistant infection OR iii. Located at a center with a high rate of resistant pathogens

Patient Specific Treatment Specific Episode Specific Age Malignancy type Disease status

Type of chemotherapy Timing of chemotherapy

Height of fever Hypotension Mucositis Blood counts/CRP

Strategy Factor Patient and disease related factors

Episode-specific factors Rule Formation

Demonstrated to be valid in these countries Low High

Rackoff et al. (1996) None AMC AMC ≥100

cells/uL HSCT United States

Alexander et al. (2002)

AML, Burkitt's lymphoma, Induction ALL, Progressive disease, Relapsed with marrow involvement

Hypotension, Tachypnea, Hypoxia<94%, New CXR Changes, AMS, Severe Mucositis, Vomiting or Abdominal Pain, Focal Infection

Absence of any risk factor

HSCT United Kingdom

Rondinelli et al. (2006) (Points)

CVC (2) ≤ 5 years old (1)

Site of infection (4.5) No URTI (2.5) HgB ≤ 70 g/L (1) Fever >38.5°C (1)

Total score <6

Total score >6 HSCT

Brazil

Santolaya et al. (2001)

Relapsed leukemia Chemotherapy within 7 days of episode

CRP ≥90mg/L Hypotension Platelets ≤50g/L

Zero risk factors, Only low platelets, or only < 7 days from chemo

All others

South America

Ammann et al. (2003)

Bone marrow involvement, central venous catheter, pre–B-cell leukemia

Absence of clinical signs of viral infection CRP>50 mg/L, WBC≤500 cells/uL, HgB>100 g/L

Three or fewer risk factors

HSCT Europe

Ammann et al. (2010)

(Points)

Chemo more intensive than ALL maintenance (4)

Hemoglobin ≥9 g/dL (5) WBC <300cells/uL (3) Platelets <50g/L (3)

Total score < 9

Total score > 9 HSCT

Europe;

Varughese 4

iii. Ongoing Management of FN: > 24 to 72 hours after initiation of empiric antibacterial treatment 1. Modification of High-risk FN:1

a. In children with persistent fever who become clinically unstable, escalate initial empiric antibacterial regimen to include coverage for resistant gram-negative, gram-positive, and anaerobic bacteria

b. Do not modify initial empiric antibacterial regimen based solely on persistent fever in children who are clinically stable

c. After 24 to 72 hours, if there is no specific microbiologic indication to continue combination therapy, discontinue empiric glycopeptide or agents added for double coverage of gram-negative bacteria

2. Cessation of Treatment:1,3 a. All patients:

i. Discontinue empiric antimicrobials if: 1. Negative blood cultures at 48 hours 2. Afebrile for at least 24 hours 3. Evidence of marrow recovery

b. Low-risk FN (IDSA data: predominantly in adults): i. Discontinue empiric antimicrobials without regard to marrow recovery in

patients at 72 hours if: 1. Negative blood cultures 2. Afebrile for at least 24 hours

Table 5: Summary of guideline recommendations:1,3

Guideline

What antimicrobials should be used empirically for in-patient management?

What are appropriate antifungal agents to add-on?

When to narrow and/or discontinue antimicrobials?

Infe

ctio

us D

isea

se S

ocie

ty o

f Am

eric

a3

Low-risk Patients • Ciprofloxacin +

amoxicillin/clavulanate

High-risk Patients • Cefepime • Piperacillin/ tazobactam • Ceftazidime • Carbapenem

• Voriconazole • Echinocandin • AmphotericinB

preparations

If no evidence for a gram-positive infection after 2 days discontinue vancomycin Low-risk: all antimicrobials can be discontinued at 3 days if clinically stable, negative cultures, and no discernible infection Note: Pediatrics: support for stopping antimicrobial therapy at ANC <500 cells/mm3 with cultures negative at 48 hours and patient afebrile >24 hours

Jour

nal o

f Clin

ical

O

ncol

ogy1

Low-risk Patients • Cefixime • Fluoroquinolone • Fluoroquinolone + Amoxicllin-

Clavulanate

High-risk Patients • Cefepime or

Ceftazidime • Piperacillin-Tazobactam • Meropenem/ Imipenem

• Caspofungin • Liposomal amphotericinB

Discontinue empiric antimicrobials if cultures negative at 48 hours, afebrile >24 hours, and there is evidence of marrow recovery (Limited research but the panel suggests that an ANC >100 cells/mm3 post-nadir is reasonable)

B. Institution Based Management Recommendations

i. Detailed List-Serve Survey Responses in Appendix A (Table 1): 1. Institutions:

a. Connecticut Children’s Hospital (Hartford, CT) b. Children’s Minnesota (Minneapolis, MN) c. Arnold Palmer Medical Center (Orlando, FL)

Varughese 5

d. Children’s Mercy Hospital (Kansas City, MO) e. The Children’s Hospital of Philadelphia (Philadelphia, PA) f. Ann & Robert H. Lurie Children's Hospital of Chicago (Chicago, IL) g. Children’s Colorado (Aurora, CO) h. Nicklaus Children’s Hospital (Miami, FL) i. Norton Childrens (Louisville, KY) j. VCU Health (Richmond, VA) k. Seattle Children’s Hospital (Seattle, WA) l. Texas Children’s Hospital (Houston, TX) m. The Children’s Hospital of San Antonio (San Antonio, TX)

Table 6: Responses to Survey Questions

What antimicrobials does your institution use empirically?

Cefepime – 70% Other responses: Zosyn and Ceftazidime

At what point do you consider broadening coverage?

Evaluate previous history and duration of current antimicrobials, if unimproved may consider an additional agent Additional options: Vancomcyin, gentamicin, or an antifungal

What is your antifungal of choice?

Micafungin – 50% Other responses: Voriconazole and Caspofungin

When do you narrow and/or discontinue antimicrobials?

“Count recovery” Afebrile status: between 1 to 5 days Signs of infection: no local symptoms Cultures: negative at 48 hours Practitioner dependent

Do you have a febrile neutropenia pathway in place for your institution?

54% - Available 23% - In Progress 33% - No pathway available

III. The Dilemma

A. There are limitations to risk stratification when evaluating a pediatric febrile neutropenia patient B. There are minimal guideline recommendations on when and how to discontinue antimicrobials in this

population i. Limited research regarding appropriate cut-offs for ANC or afebrile period

C. The financial burden, social impact, and systemic concerns of keeping patients on antimicrobials unnecessarily can take a significant toll

D. Clinical question: What are the appropriate parameters to follow before discontinuation of antimicrobial therapy in pediatric febrile neutropenia?

IV. Literature Analysis A. Evaluation of literature on pediatric febrile neutropenia

i. Hodgson-Viden H, et al. Early discontinuation of intravenous antimicrobial therapy in pediatric oncology patients with febrile neutropenia. 2005.

ii. Lehrnbecher T, et al. Short courses of intravenous empirical antimicrobial treatment in selected febrile neutropenic children with cancer. 2002.

iii. Mueller EL, et al. Hospital Discharges for Fever and Neutropenia in Pediatric Cancer Patients: United States, 2009. 2015.

Varughese 6

Table 7: Review of Study by Hodgson-Viden et al.

Varughese 7

Early discontinuation of intravenous antimicrobial therapy in pediatric oncology patients with febrile neutropenia. Hodgson-Viden H, Grundy PE, Robinson JL. BMC Pediatr. 2005;5(1).

Background: FN is a common problem with pediatric oncology patients and although few have a MDI, those who do are at risk for overwhelming sepsis. Standard of practice is to start empiric antibiotic therapy if ANC <500-1000 cells/mm3. There are no standard criteria or guidelines on endpoints for stopping IVAMT if a bacterial source is not found. Traditionally the NACCP has initiated broad spectrum IVAMT until the patient was afebrile, blood cultures were negative at >48 hours and ANC>500 cells/mm3. This study looked to evaluate the outcome of febrile neutropenia patients in the face of a new trend to discontinue antibiotics at ANC < 500 cells/mm3. Methods:

1. Objective: a. To evaluate the outcome of a large cohort of children with FN who were initiated on IVAMT and

discontinued before ANC reached 500/mm3. 2. Study Design:

a. Retrospective chart review completed of patients in the North Alberta Children’s Cancer Program (Canda) with FN and no apparent clinical source of fever from June 1, 1997 to July 1, 2002.

i. Inpatient charts were reviewed for these patients at the Stollery Children’s Hospital in Edmont, Canada

ii. Outpatient charts were reviewed to determine if the fever recurred after hospital discharge b. Study Criteria

Inclusion Exclusion Age < 17 years receiving therapy for a primary or recurrent malignancy

Children with leukemia who were not yet in remission

Fever > 38.0°C measured at home or in hospital β-hemolytic streptococci from throat culture ANC < 500cells/mm3 Clostridium difficile in stool Common skin organism isolates in blood

3. Data Collected: a. If cultures done during the first 48 hours of admission were positive for an organism that would account

for the fever, then no further data was collected b. Data collection on admissions where infectious etiology was not identified from cultures within 48 hours:

i. Choice and duration of IVAMT ii. Oral antibiotic use for FN treatment

iii. Systemic anti-fungal agents used iv. Microbiologic laboratory results v. Cell Counts: APC, ANC, AMC

vi. Length of stay in hospital attributable to FN vii. Duration of fever and recurrence of fever during or after the admission

1. Recurrent fever: new fever that occurred after the patient has been afebrile for > 24 hours, and before they reached an ANC >500 cells/mm3

4. Outcomes: a. Patients and episodes of FN b. Microbiologically documented infections c. Duration of fever d. Antimicrobial therapy e. Clinical course and outcome

5. Data Analysis: a. Descriptive statistics were used to summarize the patient cohort

Results: 6. Evaluation of outcomes

a. Patients and episodes of FN i. Appendix B – Table 1: Episodes of febrile neutropenia by diagnosis in 275 pediatric oncology

patients

Varughese 8

b. MDI

i. Appendix B – Table 2: Incidence of proven bacterial or viral infections 1. Identified within 1st 48 hours of admission in 21% of 276 episodes (76% bacterial, 24%

viral) 2. Identified after 48 hours of admission in 8 episodes

c. Duration of fever i. Documented only prior to hospitalization in 72/217 episodes of FN without proven infection

ii. Documented during hospitalization in 145 episodes and median duration of the initial fever was 2 days (range 1-33 days)

1. 39 of these episodes: fever recurred prior to discharge and persisted for a median of 3 days (range 2-19 days)

d. Antimicrobial therapy i. Most frequently prescribed IVAMT: piperacillin+tobramycin (75% of episodes)

ii. Median duration: 5 days (range 1-28 days) e. Clinical course and outcome

i. Appendix B – Table 3: Hematologic parameters when antibiotics were discontinued and at discharge in pediatric oncology patients with FN with negative cultures at admission

1. Median length of stay: 5 days (range 1-33 days) 2. Median ANC at discontinuation of IVAMT was 400 cells/mm3, and in 14% of patients

IVAMT was discontinued prior to ANC rising to 100 cells/mm3. In 7% IVAMT was discontinued when the APC was <100 cells/mm3.

3. Two patients died during their admission for FN: unrelated to infectious process 4. None of the patients required readmission related to clinical deterioration, suspected

infection, or recurrent fever prior to neutropenia resolution Discussion:

Strengths Limitations Fevers were not attributed to β hemolytic streptococci from throat culture

Exclusion of children with leukemia (unless in remission)

Fevers were not attributed Clostridium difficile in stool Utilization of antibiotics that are not preferred: piperacillin (without tazobactam) + tobramycin

Exclusion of patients with common skin organism isolates in single blood culture

Use of statistical median instead of mean

No re-admission criteria Authors’ Conclusions: There is limited consensus on the appropriate discontinuation of antibiotics in pediatric febrile neutropenia. In this retrospective review of patients’ charts, clinicians were able to select patients who could safely be discontinued from IVAMT even if ANC was <500 cells/mm3. These patients did not require readmission related to their initial fever and did not experience any adverse events from early discontinuation of IVAMT. Personal Conclusions: This retrospective analysis of 276 FN episodes reviewed the outcomes related to early discontinuation of IVAMT. Patients were able to be safely discharged after early discontinuation of IVAMT without re-admission. This study provided support to help move away from extended IVAMT therapy by safely analyzing patients based on fever status and notable marrow recovery in FN pediatric patients.

ANC = Absolute Neutrophil Count FN=Fever and Neutropenia

IVAMT= Intravenous Antimicrobial Therapy MDI=Microbiologically Documented Infection APC=Absolute Phagocytic Count AMC=Absolute Monocyte Count

275 FN Patients

148 Patients Not Admitted

127 Patients Admitted

276 Epsiodes of FN Admission

Varughese 9

Table 8: Review of Study by Lehrnbecher et al.

Short courses of intravenous empirical antibiotic treatment in selected febrile neutropenic children with cancer. Lehrnbecher T1, Stanescu A, Kühl J. Infection. 2002;30(1):17-21.

Background: In oncology patients, therapy-induced alterations of host defense influence the predisposition to infection, but neutropenia is the major risk factor for serious bacterial and fungal infections. The use of IV empiric broad-spectrum antibiotics has greatly reduced morbidity and mortality related to febrile neutropenia episodes. However the optimal duration of treatment is not clear, especially in patients who present with FUO. This retrospective review analyzed discontinuation of IV antibiotics and early hospital discharge in pediatric cancer patients with FUO regardless of ANC or evidence of bone marrow recovery. Discontinuation was allowed as long as patients were afebrile for at least 24 hours and had been treated with IV antibiotics for a minimum duration of 72 hours. Methods:

1. Objective: a. To evaluate the safety and efficacy of short courses of IV antibiotic treatment in selected pediatric cancer

patients admitted for fever and neutropenia. 2. Study Design:

a. Retrospective analysis of short courses of IV antibiotic therapy in pediatric cancer patients with FUO b. From January 1994 through June 1996 at the Children’s Hospital of the University of Würzburg, Germany

c. Study Criteria:

Inclusion Exclusion ANC: <500 cells/mm3 Antibiotic exposure in the last 72h prior to admission Fever: >38.5°C x 1 or >38-38.4°C x 2 within a 4 hour interval

3. Data Collected:

a. Clinical assessment at admission: i. Complete blood cell count with differential

ii. Blood cultures from lumen of central venous catheter, peripheral vein, urine culture, and any clinically suspicious sites of infection

b. New blood cultures were performed if the patient developed new temperature rises or remained febrile for more than 72 hours

c. Diagnostic criteria classified initial fever episode as FUO or fever attributed to clinically/microbiologically documented infection

d. Initial empiric antibiotic therapy i. January 1994 – April of 1995: Ceftazidime 150 mg/kg/day in three divided doses and teicoplanin

10 mg/kg/day twice the first day then once daily ii. May 1995 – June 1996: Imipenem 50 mg/kg/day divided into four doses; if fever persisted then

teicoplanin 10 mg/kg/day was added e. Discharge criteria

i. Clinically stable ii. Negative blood culture results

iii. Absence of fever for at least 24 hours without antipyretics iv. IV antibiotic treatment for a minimum 72 hours v. ANC >500 cells/mm3 or evidence of bone marrow recovery were not a precondition for the

discontinuation of antibiotic therapy f. Evaluation of response

i. Success in FUO: defervescence during the antibacterial regimen and no recurrent fever until the recovery of neutrophils

4. Outcomes: a. Primary:

i. Modification of therapy based on fever status b. Secondary

56 FN Patients 106 FN Episode Admissions 84 Episodes of FUO

Varughese 10

i. Duration of fever ii. Duration of therapy

iii. Relapse of fever iv. Re-hospitalization v. Deaths

5. Data Analysis: a. Descriptive statistics were used to summarize the patient cohort

Results: 1. Characteristics of the Study Population:

2. Clinical Outcomes

a. Appendix C – Table 1: Clinical outcome in patients with FUO (n=84) i. Modification of therapy based on fever status

1. No modification required in 68% FUO patients 2. Modification required in 32%

ii. Duration of fever 1. Mean days: 3 (range: 1-17)

iii. Duration of therapy 1. Mean days: 5 (range: 3-18)

iv. Relapse of fever = 2 v. Re-hospitalization = 0

vi. Deaths = 0

Discussion: Strengths Limitations

Inclusion of patients with ANC drops within 72 hours of chemotherapy

Population size

Two treatment regimens with similar outcomes Antibiotic choices Re-admission criteria of 3 weeks

Authors’ Conclusions: In both treatment regimens, IV antibiotic discontinuation was achievable regardless of ANC or evidence of bone marrow recovery. In this review of outcomes this goal seems safe and effective in pediatric cancer patients with FUO if children are afebrile for at least 24 hours and are treated for a minimum of 72 hours. However, only the overall evaluation of inpatient treatments with short and long antibiotic courses and of outpatient management within the scope of a large randomized study will allow a final judgment about safety, efficacy and the reduction in cost of the different strategies. Personal Conclusions: This study supported the discontinuation of IV antibiotic therapy in patients classified as FUO regardless of ANC when patients were afebrile for >24 hours and were treated for a minimum of 72 hours with IV antibiotics. No death or major complication occurred and none of the patients were re-hospitalized for recurrent fever or infection. The treatment strategy outlined in this study potentially provided patients and their families with improved quality of life resulting from shorter hospital stays, decreased treatment costs, and the reduced impact of antibiotic overuse. The antibiotics chosen are not used in recent practice making the study less generalizable to current therapy. The use of fever and treatment duration as discharge criteria without evaluating bone marrow recovery could be controversial in most institutions however the positive outcomes of the study do provide support for this form of management.

FN = Febrile Neutropenia FUO=Fever of Unknown Origin ANC=Absolute Neutrophil Count IV=Intravenous

52 FN Episodes:

Ceftazidime + Teicoplanin

54 FN Episodes: Imipenem

106 FN Episodes:•Similar patient

characteristics•Similar outcomes

Varughese 11

Table 9: Review of Study by Mueller et al.

Hospital Discharges for Fever and Neutropenia in Pediatric Cancer Patients: United States Mueller EL, Walkovich KJ, Mody R, Gebremariam A, and Davis MM. BMC Cancer. 2015;15:388.

Background: Hospital utilization patterns for fever and neutropenia are poorly described. Despite the fact that hospitalizations for FN are common among pediatric cancer patients, recent data is limited regarding hospitalizations for pediatric FN at the national level. No aggregate evaluation of hospitalizations for FN in pediatric cancer patients across the United States has been published. A more comprehensive understanding of those patients with a short LOS may help inform future research aimed at decreasing the need for hospitalization of pediatric FN patients at low risk for serious complications. Methods:

1. Objective: a. The purpose of this study was to characterize discharge criteria for pediatric cancer patients admitted non-

electively and discharged with a diagnosis of fever and neutropenia, across the United States. b. The authors hypothesized that a substantial portion of pediatric cancer admissions would have a SLOS

with few serious infections and qualified encounters could lead to interventions for future out patient management.

2. Study Design: a. Retrospective chart review of the KID, an all-payer US hospital database, for 2009.

i. The KID is a nationally representative database compiled by the Agency for Healthcare Research and Quality.

Inclusion Exclusion

Age < 19 years Age > 19 years Admit type: urgent or emergent Non-emergent admit ICD-9 Codes: Fever& Neutropenia [or decreased White Blood Cell Count]

Inaccurate Codes

Non-transferred patients

3. Data Collected: a. Identified FN patient discharges associated with a relatively short, uneventful inpatient course and to

assess variables associated with these discharges. i. These discharges associated with a SLOS would be potential candidates for future outpatient

management strategies 4. Outcomes:

a. Primary: i. Frequency of discharges with a SLOS defined as < 3 days in order to capture all hospitalizations

that would have lasted 48 hours or less 1. This threshold was selected because of the clinical relevance of negative blood cultures at

48 hours b. Secondary:

i. Population-adjusted frequency of discharges ii. Mean LOS

iii. Mean and total hospital charges per discharge for FN 5. Data Analysis:

a. Descriptive statistics were used for gender, age, race, ethnicity, payer, hospital teaching status and location, and hospital census region

b. A weighted multivariate logistic regression model was used to estimate factors associated with SLOS for pediatric cancer patient encounters for FN

i. The authors hypothesized that socio-demographic and hospital level factors may influence admission and/or discharge decision making based on resources of the individual and within the community

Results: 6. Characteristics of the Study Population:

a. Appendix D (Table 1): Characteristics of discharges for non-transferred pediatric cancer patients: overall and for FN discharges

Varughese 12

i. Age: majority were 0-9 years of age ii. Sex: predominately male

iii. Ethnicity: non-Hispanic white race. iv. Insurance: most FN patients were insured

b. In 2009 there were a total of 7,370,203 weighted pediatric discharges in the United States i. 1.5% (n=110,967) were associated with a diagnosis of cancer

ii. 49% of pediatric patients were admitted non-electively and coded at either urgent or emergent iii. 19.3% of non-elective pediatric cancer discharges met criteria for FN

7. Outcomes a. Primary:

i. Frequency of discharges with a SLOS defined as < 3 days 1. 41% of patients were discharged at < 3 days 2. 33% were discharged within 4 to 7 days 3. 16% were discharged within 8 to 14 days 4. 7% were discharged within 15 to 30 days 5. 3% were discharged after 31 days

b. Secondary: i. Appendix D – Table 1: Population-adjusted frequency of discharges

1. Occurred at a rate of 13.4 per 100,000 US children each year using national Census data for 200916

ii. Appendix D – Table 2: Mean and total hospital charges per discharge for FN 1. Charges were adjusted from 2009 to 2014 using the Consumer Price Index17 2. The overall mean hospital charge for a pediatric cancer patient discharge for FN was

$52,160 and total costs was $587.3 million 3. FN discharges with a SLOS accounted for ~$66 million of the total costs

iii. Appendix D – Table 3: Comparison of proportion of infectious diagnoses by LOS category among pediatric cancer FN discharges

1. No infections were identified in ~76% of patients 2. Occurrences of pneumonia, blood stream, and urinary tract infections increased with LOS

iv. Appendix D – Table 4: Multivariate logistic regression to evaluate factors associated with a “Short LOS” (≤3 days) among pediatric cancer fever and neutropenia discharges

1. Age categories of “0 to 4” and “5 to 9” were associated with shorter lengths of stay 2. Acute otitis media and viral infections were also associated with shorter lengths of stay

Discussion: Strengths Limitations

Utilization of national database No data on antibiotic regimens used Population Size Categorization of “Shortened Length of Stay” Non-transferred patients Accuracy of ICD-9-CM codes Analysis of variables Admission information unavailable Updated with 2012 study

Summary of 2012 Study Update:15

Outcomes: a. Average charge for a SLOS was $17,437 b. Average LOS for FN discharges was 7.5 days and 39% of discharges had a SLOS (n=5230). c. SLOS discharges accounted for $91.2 million in charges for 2012

Authors’ Conclusions: Identification of the burden of FN is a key step in the process of developing awareness and motivation for a multi-center or national study to improve care of pediatric cancer patients experiencing FN. This study demonstrates the substantial impact on healthcare of pediatric oncology patients with discharges for FN. Discharges with a short LOS would be potential candidates for future outpatient management strategies. Personal Conclusions: The authors of this study provided significant data that can be used to limit/eliminate hospital length of stay for FN patients who are low-risk. There is clearly a financial burden associated with unnecessary hospital stay but also an increased risk for patients to develop hospital acquired infections. The results of this study were updated with a 2012 study that provided similar data with a continuing climb of healthcare costs related to FN LOS. Both studies suggest that hospitals and practitioners should objectively analyze and risk stratify their patients to determine if the patient requires hospital admission or if outpatient management may be more appropriate.

Varughese 13

SLOS = Shortened Length of Stay LOS = Length of Stay FN=Fever and Neutropenia

KID = Kids’ Inpatient Database

V. Conclusion

A. Application to Patient Case12,15 i. The continued documentation of the burden of FN will be key in expanding awareness and increasing

motivation for a larger study to improve the care of pediatric oncology patients. ii. Prospective randomized studies could determine if use of objective criteria in addition to clinical

judgment could increase the number of children who can safely have intravenous antimicrobial therapy discontinued while still neutropenic.

B. Summary i. Requiring admission for all patients with febrile neutropenia creates a significant impact in our

healthcare system. ii. By appropriately risk stratifying based on the patient, episode, and treatment factors we can reduce

the number of patients that are admitted and started on IV antimicrobials. iii. By revising our outlook on ANC goals for antimicrobial discontinuation/de-escalation we can

decrease the overuse of parenteral antimicrobials. iv. Benefits of limiting the duration of therapy:11

1. Decreasing the emotional, social, and financial burdens associated with prolonged hospitalization and extended antimicrobials regimens

2. Decreasing the risk of antimicrobial-related adverse effects and toxicities 3. Possibly decreasing the risk of secondary infection with organisms that are resistant to the

antimicrobials that were used for empiric treatment of febrile neutropenia C. Future Research

i. Appendix E: Stern et al. Intervention Protocol: Early discontinuation of antimicrobials for febrile neutropenia versus continuation until neutropenia resolution

1. Objective: To assess the safety of protocol-guided discontinuation of antimicrobials regardless of neutrophil count compared to continuation of antimicrobials until neutropenia resolution in cancer patients with fever and neutropenia, in terms of mortality and morbidity.

D. Recommendations for discontinuation/de-escalation of anti-microbial therapy i. Institutions should implement criteria for risk stratification to differentiate between high and low

risk patients. ii. Institutions should incorporate criteria for discontinuation/de-escalation of antimicrobial therapy

based on the following parameters 1. Afebrile period of 24 hours 2. Negative cultures at 48 hours 3. Absolute neutrophil count

a. Low risk population: ANC > 100 cells/microL b. High risk population: ANC > 200 cells/microL

4. No signs or symptoms of infection

Varughese 14

References

1. Lehrnbecher T, Phillips R, Alexander S, Alvaro F, Carlesse F, Fisher B, et al. Guideline for the management of fever and neutropenia in children with cancer and/or undergoing hematopoietic stem-cell transplantation. J ClinOncol. 2012;30(35):4427-38. doi: 10.1200/JCO.2012.42.7161.

2. Schwartzberg LS. Neutropenia: etiology and pathogenesis. Clin Cornerstone. 2006;8 Suppl 5:S5-11. doi:10.1016/S1098-3597(06)80053-0.

3. Freifeld AG, Bow EJ, Sepkowitz KA, Boeckh MJ, Ito JI, Mullen CA 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.

4. Ziino O, Tucci F, and Rossi MR. Outpatient management of febrile neutropenia in children with cancer. Pediatr Rep. 2011;3(1): e8. doi: 10.4081/pr.2011.e8.

5. Kanamaru A, Tatsumi Y. Microbiological data for patients with febrile neutropenia. Clin Infect Dis. 2004;39 Suppl 1:S7-S10.

6. Rackoff WR, Gonin R, Robinson C, Kreissman SG, Breitfeld PB. Predicting the risk of bacteremia in children with fever and neutropenia. J Clin Oncol. 1996;14: 919– 924. doi: 10.1200/jco.1996.14.3.919. Abstract.

7. Alexander SW, Wade KC, Hibberd PL, Parsons SK. Evaluation of risk prediction criteria for episodes of febrile neutropenia in children with cancer. J Pediatr Hematol Oncol. 2002;24: 38– 42.

8. Rondinelli PI, Ribeiro Kde C, de Camargo B. A proposed score for predicting severe infection complications in children with chemotherapy-induced febrile neutropenia. J Pediatr Hematol Oncol. 2006;28: 665– 670. doi: 10.1097/01.mph.0000212996.94929.0b.

9. Santolaya ME, Alvarez AM, Becker A, et al. Prospective, multicenter evaluation of risk factors associated with invasive bacterial infection in children with cancer, neutropenia, and fever. J Clin Oncol. 2001;19: 3415– 3421. doi: 10.1200/JCO.2001.19.14.3415.

10. Ammann RA, Hirt A, Lüthy AR, Aebi C. Identification of children presenting with fever in chemotherapy-induced neutropenia at low risk for severe bacterial infection. Med Pediatr Oncol. 2003; 41: 436– 443. doi: 10.1002/mpo.10320.

11. Ammann RA, Bodmer N, Hirt A , et al. Predicting adverse events in children with fever and chemotherapy-induced neutropenia: The prospective multicenter SPOG 2003 FN study. J Clin Oncol. 2010; 28: 2008– 2014. doi: 10.1200/JCO.2009.25.8988.

12. 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). doi: 10.1186/1471-2431-5-10.

13. Lehrnbecher T1, Stanescu A, Kühl J. Short courses of intravenous empirical antibiotic treatment in selected febrile neutropenic children with cancer. Infection. 2002;30(1):17-21.

14. Mueller EL, Walkovich KJ, Mody R, Gebremariam A, and Davis MM. Hospital Discharges for Fever and Neutropenia in Pediatric Cancer Patients: United States, 2009. BMC Cancer. 2015;15:388. doi: 10.1186/s12885-015-1413-8.

15. Mueller EL, Croop J, Carroll AE. Fever and neutropenia hospital discharges in children with cancer: A 2012 update. Pediatr Hematol Oncol. 2016;33(1):39-48. doi: 10.3109/08880018.2015.1102998.

16. Annual Estimates of the Resident Population by Sex and Five-Year Age Groups for the United States: April 1, 2000 to July 1, 2009 (NC-EST2009-01) U.S. Census Bureau, Population Division Release Date: June 2010.

17. Consumer Price Index Inflation Calculator. [http://data.bls.gov/cgi-bin/cpicalc.pl]

Varughese 15

VI. Appendices

Table 1: Institution Based Management Recommendations:

What antimicrobials do you use empirically?

When do you broaden your coverage?

What is your antifungal of

choice?

When do you narrow and/or discontinue

antimicrobials?

Do you have a FN pathway?

Ann & Robert H. Lurie Children's Hospital of

Chicago (Chicago, IL)

Ceftazidime

Zosyn if presenting with abdominal symptoms

Ill appearing and CV stable with central line: consider ceftriaxone +

vancomycin

CV unstable: ceftazidime + tobramycin + vancomycin

Micafungin

Stop vancomycin or second gram stain

negative after 48 hours

Stop ceftazidime after count recovery

Available

Arnold Palmer Medical Center

(Orlando, FL)

Cefepime

Add vancomycin if patient has a high risk for MRSA or if patient has

persistent fever at 48-72 hours with no source

Add Micafungin if patient remains febrile after 3-5

days on cefepime/ vancomycin

Varies based on physician No

Children’s Colorado (Aurora, CO)

Cefepime

AML patients will also get empiric antifungal with

voriconazole

Vancomycin for patients with high risk for S. viridans (IE s/p high

dose Ara-C), add fungal coverage if febrile for >4/5 days, add flagyl if

concern for typhilitis

Depends on chemo regimen, usually

ends up being voriconazole or if we cannot get an azole due to drug interactions then

micafungin (issues with prophylaxis

breakthrough)

Afebrile

ANC >100 (>150 if AML)

In Progress

The Children’s Hospital of Philadelphia

(Philadelphia, PA)

AML/Relapsed ALL in induction/MRSA: Cefepime

+ vancomycin (24 hour rule-out) + gentamicin

If not improving then consider empiric antifungal therapy

Liposomal amphotericin B or

caspofungin

Continue cefepime until count recovery Available

The Children’s Hospital of San Antonio

(San Antonio, TX) Cefepime

May add vancomycin/gentamicin based on protocol criteria

If abdominal symptoms switch

from cefepime to Zosyn

If febrile at >96 hours add

micafungin therapy

Stop vancomycin at cultures negative x48

hours Afebrile X24 hours

ANC rising

In Progress

Children’s Mercy Hospital

(Kansas City, MO) Cefepime

Add vancomycin in 2-3 days (with a 48 hour rule-out infection)

Voriconzole (at 5-6 days) unless drug interactions; then either micafungin

or liposomal amphotericin B

Discharge to home if afebrile for 24 hours and cultures negative for 48

hours

No

Varughese 16

Children’s Minnesota (Minneapolis, MN) Cefepime See treatment guidelines Micafungin If clinically stable Available

Connecticut Children’s Hospital

(Hartford, CT) Zosyn Add vancomycin if concern for

MRSA or Streptococcus viridans

Micafungin (when no CNS concern); if CNS concern then

use ambisome

Practitioner dependent No

Nicklaus Children’s Hospital

(Miami, FL)

Cefepime +/- vancomycin (48hr rule-out)

Escalate cefepime to meropenem if remains febrile

Micafungin

Negative cultures Afebrile for 5 days

Clinically stable

In Progress

Norton Childrens (Louisville, KY)

Cefepime With mucositis add vancomycin Voriconazole/

caspofungin

Negative cultures at 48 hours

Afebrile for ≥24 hours Evidence of marrow

recovery

Available

Seattle Children’s Hospital

(Seattle, WA)

Ceftazidime If AML post high-dose ARA-

C then Cefepime

If hemodynamically unstable then add vancomycin and gentamicin

Based on fungal infection guidelines

Discontinue once ANC >200 and patient is

afebrile for 24 hours and blood cultures are

negative at 48 hours

Available

Texas Children’s Hospital (Houston, TX)

Low Risk: Ceftazidime

High Risk: Vancomycin+ Piperacillin/tazobactam +

Gentamicin

If febrile > 48 hrs Add vancomycin

Or consider changing medications if already on triple antibiotics

If fever continues on antibiotics for >

5 days begin antifungal agents

Afebrile for ≥24 hr Negative blood culture

for 48hrs No signs of focal

infection ANC > 100/mm³ and

climbing

Available

VCU Health (Richmond, VA)

Ceftriaxone at presentation

Cefepime if neutropenic

Add vancomycin if mucositis or cellulitis or history of high-dose

cytarabine

If septic: cefepime + vancomycin + gentamicin

OR Meropenem + vancomycin

Not answered

Discontinue vancomycin and gentamicin if

cultures are negative after 48 hours

Discontinue all

antibiotics if cultures are negative to date,

afebrile, and ANC is trending upward

Available

Varughese 17

Appendix B: Table 1: Episodes of febrile neutropenia by diagnosis in 275 pediatric oncology patients11 (Abbreviated)

Diagnosis Number of patients (%)

Number of patients with FN episodes

(%)

Mean number of episodes per patient

Range of number of episodes per patient

Osteogenic sarcoma 8 (3%) 7 (87%) 3.13 0–7 Ewings sarcoma 5 (2%) 4 (80%) 2.20 0–5 Rhabdomyosarcoma 15 (5%) 11 (79%) 1.93 0–6 Hepatoblastoma 9 (3%) 5 (56%) 1.22 0–5 Acute myeloid leukemia (AML)

16 (6%) 7 (44%) 1.00 0–5

Neuroblastoma 23 (8%) 11 (48%) 0.83 0–5 Non-Hodgkin's lymphoma

24 (9%) 14 (58%) 0.75 0–4

Acute lymphoblastic leukemia (ALL)

85 (31%) 41 (48%) 0.73 0–7

Others * 22 (8%) 9 (41%) 0.41 0–3 Hodgkin's lymphoma 25 (9%) 8 (32%) 0.36 0–5 Central nervous system tumors

26 (9%) 7 (27%) 0.23 0–3

Wilms tumor 17 (6%) 3 (18%) 0.12 0–1 Total 275 127 (46%)

*adrenocortical carcinoma, epithelial tumor, juvenile myelomonocytic leukemia, Langerhans cell histiocytosis, small cell tumor, teratoma

Table 2: Incidence of proven bacterial or viral infections11

Type of infection Age 0–1 year, n = 40 Age 2–4 years, n = 91 Age ≥ 5 years, n = 145 Viral infection 2 (5%) 6 (7%) 7 (5%) Bacterial infection 9 (23%) 17 (19%) 18 (12%) Cultures negative 29 (73%) 68 (75%) 120 (83%)

n = number of episodes of FN in this age range

Table 3: Hematologic parameters when antibiotics were discontinued and at discharge in pediatric oncology patients with FN with negative cultures at admission11

Median Range Number (%) <100 cells/mm3

Number (%) 100–500 cells/

cells/mm3

Number (%) > 500 cells/ cells/mm3

ANC when antibiotics discontinued (n = 199)

400 0–34900 28 (14%) 84 (42%) 87 (44%)

Monocyte count when antibiotics discontinued (n = 199)

400 0–4400 20 (10%) 117 (59%) 62 (31%)

APC when antibiotics discontinued (n = 199)

800 0–39300 13 (7%) 56 (28%) 129 (65%)

ANC at discharge (n = 194)

500 0–42200

n = number of episodes of FN; ANC – absolute neutrophil count; APC – absolute phagocyte count

Varughese 18

Appendix C: Table 1: Clinical outcome in patients with fever of unexplained origin (FUO)12

Initial therapy Imipenem(n = 41) Ceftazidime/teico- planin group (n = 43)

Total group (n = 84)

No modification required 27 30 57 Afebrile within first 72 hours

24 23 47

Afebrile > 72 hours 3 7 10 Modification required* 14 13 27 Subsequent antibiotic regimen

14 12 26

Other antibiotic modification 1 1 2 Addition of antifungal agents 5 5 10 Addition of antiviral agents 0 1 1 Duration of fever, mean (range) days

3 (1–10) 3 (1–17) 3 (1–17)

Duration of therapy, mean (range) days

5 (3–13) 5 (3–18) 5 (3–18)

Relapse of fever 1 1 2 Re-hospitalization 0 0 0 Deaths 0 0 0 * Some patients had more than one change of therapy (e.g. change of antibiotics, addition of antifungal or antiviral agent)

Appendix D: Table 1: Characteristics of discharges for non-transferred pediatric cancer patients: overall and for fever and neutropenia (FN) discharges – United States, 200913 (Adapted)

Proportion of overall pediatric cancer discharges

Proportion of pediatric FN discharges

% (95 % CI) Patient characteristics Gender Female 45.0 (44.2–45.8) 46.5 (45.0–48.0) Age 0–4 years 28.9 (27.8–30.0) 36.7 (34.6–38.8) 5–9 years 22.0 (21.2–22.7) 27.9 (26.6–29.3) 10–14 years 21.3 (20.6–22.0) 18.2 (17.0–19.5) 15–19 years 27.8 (26.6–29.1) 17.2 (15.8–18.7) Race/Ethnicity White 48.4 (44.2–52.7) 55.7 (50.5–60.9) Black 9.7 (8.4–11.2) 7.1 (5.7–8.7) Hispanic 21.2 (17.7–25.3) 17.1 (13.6–21.3) Asian/pacific islander 7.9 (6.7–9.4) 7.6 (6.0–9.4) Type of cancer ALL 24.6 (23.5–25.8) 44.3 (41.6–47.0) Bone cancer 12.8 (12.0–13.6) 10.4 (9.1–12.0) CNS tumor 9.6 (8.8–10.4) 6.2 (5.1–7.5) AML 5.9 (5.5–6.3) 7.4 (5.9–9.1) Soft tissue sarcoma 5.1 (4.7–5.6) 4.6 (3.9–5.4) Neuroblastoma 4.5 (3.9–5.1) 4.7 (3.7–6.0) Hodgkin lymphoma 3.2 (2.9–3.5) 2.9 (2.5–3.4) Wilms tumor 2.7 (2.5–3.0) 3.0 (2.5–3.6) Non–Hodgkin lymphoma 2.6 (2.3–2.9) 3.6 (3.1–4.3) Ovarian or testicular tumor 1.7 (1.5–1.9) 0.7 (0.5–1.0)

Varughese 19

Table 2: Mean charges for pediatric cancer fever and neutropenia discharges: overall and by length of stay category - United States, 200913

Mean charges Total charges Overall $52,160 $587,398,210 Length of stay category ≤3 days – “Short LOS” $14,549 $67,296,971 4–7 days $33,423 $125,499,734 8–14 days $72,552 $128,683,775 15–30 days $166,316 $128,267,528 >30 days $412,730 $138,805,796

Table 3: Comparison of proportion of infectious diagnoses by LOS category among pediatric cancer FN discharges13

Overall LOS category “Short LOS” 4-7 days 8-14 days 15-30 days 31+ days ≤3 days

Proportion (%) Proportion of FN DCs

41 33 16 7 3

No infection identified 75.9 82.7 77.3 66.9 62.7 44.4 Type of infection Upper respiratory infection 5.4 6.0 5.6 4.0 4.3 5.3 Acute otitis media 2.9 3.7 2.3 2.3 2.2 3.3 Bloodstream infection 10.4 3.1 9.5 20.8 23.9 35.6 Viral infection 2.3 3.1 2.3 1.6 0.2 0 Urinary tract infection 1.9 1.1 2.4 2.2 2.5 4.6 Pneumonia 1.2 0.3 0.6 2.3 4.1 6.8

Table 4: Multivariate logistic regression to evaluate factors associated with a “Short LOS” (≤3 days) among pediatric cancer fever and neutropenia discharges13

Factors Adjusted odds ratio (OR) 95% CI P-value Gender Female 1.06 1.02–1.10 0.006 Age 15–19 years Ref

10–14 years 1.06 0.99–1.13 0.076 5–9 years 1.20 1.11–1.30 <0.001 0–4 years 1.08 1.01–1.15 0.029 Type of infection Upper respiratory infection 1.07 0.97–1.18 0.165 Acute otitis media 1.26 1.10–1.44 0.001 Bloodstream infection 0.21 0.18–0.24 <0.001 Viral infection 1.82 1.55–2.15 <0.001 Urinary tract infection 0.44 0.39–0.50 <0.001 Pneumonia 0.27 0.22–0.32 <0.001 Type of cancer ALL 0.84 0.78–0.89 <0.001 Bone cancer 0.77 0.71–0.85 <0.001 Central nervous system tumor 0.99 0.92–1.07 0.787 AML 0.40 0.36–0.45 <0.001 Soft tissue sarcoma 1.25 1.13–1.37 <0.001 Neuroblastoma 0.56 0.48–0.66 <0.001 Hodgkin lymphoma 1.66 1.45–1.91 <0.001 Wilms tumor 0.59 0.52–0.68 <0.001 Non-Hodgkin lymphoma 0.46 0.39–0.54 <0.001 Ovarian or testicular tumor 0.63 0.52–0.76 <0.001

Varughese 20

Controlled for Race; CI = Confidence Interval, ALL = Acute Lymphoblastic Leukemia; AML = Acute Myelogenous Leukemia; Type of Infection and Type of Cancer are presented as dichotomous variables

Appendix E

Intervention Protocol: Early discontinuation of antibiotics for febrile neutropenia versus continuation until neutropenia resolution.

Stern A, Carrara E, Yahav D, Leibovici L, et al. Cochrane Database of Systematic Reviews. 2016. Objective To assess the safety of protocol-guided discontinuation of antibiotics regardless of neutrophil count, compared to continuation of antibiotics until neutropenia resolution in cancer patients with fever and neutropenia, in terms of mortality and morbidity. To assess the emergence of resistant bacteria in patients treated with short courses of antibiotic therapy compared with patients treated until resolution of neutropenia. Currently there is no consensus between different practice guidelines with regards to the duration of antibiotic therapy. Recommendations are variable and sometimes imprecise, leaving clinicians in doubt regarding the appropriate course of management. Methods Criteria for considering studies for this review:

1. Studies: a. Randomized control trials

2. Participants Inclusion Adults and children with fever and neutropenia caused by cancer-related chemotherapy, treated with any antibiotic regimen Fever of unknown origin: single oral temperature higher than 38.3°C or higher than 38.0°C sustained >1 hour Neutropenia: ANC <500 cells/mm3 Clinically documented infections Microbiologically documented infections

3. Outcome Measures a. Primary: Thirty-day all-cause mortality; if all-cause mortality is not available at 30 days, we will use the time

point reported by the study and document it b. Secondary:

i. Fever days, defined as the number of consecutive febrile days (starting from the beginning of the febrile neutropenia episode or as defined by study)

ii. Total antibiotic days during follow-up/hospitalization, including initial antibiotic course and any subsequent antibiotic therapy initiated after randomization

iii. Clinical failure, as defined in study. iv. Any bacteremia developing after the time point defined for the short-course antibiotic arm or after

randomization and during the follow-up period v. Any documented infection diagnosed after the time point defined for the short-course antibiotic arm

or after randomization and during the follow-up period vi. Any invasive fungal infection diagnosed after randomization (including invasive candida infections,

any mold infection and others as defined by study) vii. Clostridium difficile infections diagnosed after randomization

viii. MDR acquisition, defined as any isolation of a resistant pathogen after randomization, including methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococci (VRE), and multidrug-resistant Gram-negative (MDR GN), as defined in study

ix. Secondary infections caused by bacteria resistant to the initial antibiotic treatment, as defined in each trial

x. Number of hospitalization days xi. Any need for chemotherapy delay as reported in study

4. Subgroup analysis and investigation of heterogeneity a. Adult patients versus children (adults defined as 18+ years). b. Documented infection (clinically or microbiologically) versus fever of unknown etiology. c. High-risk patients versus low-risk patients, preferably defined by MASCC score, but we will accept and

document the study definitions. d. Solid tumor patients versus hematologic malignancy patients

Varughese 21

e. Severity of neutropenia (ANC < 100 versus 100 < ANC < 500), referring to the lowest ANC count documented (nadir of neutropenia).