febrile neutropaenia
Contents
Oncologic emergency defined by fever (single oral temp ≥38.3°C or ≥38.0°C sustained >1 hour) with ANC <0.5 × 10⁹/L (or ANC ≤1.0 × 10⁹/L with predicted decline to <0.5 within 48 hours). Overall mortality up to ~11%; reaches ~50% with septic shock. The immediate job: cultures, empirical antipseudomonal antibiotics within 60 minutes, risk stratification for site of care. Duration of therapy is shifting — newer evidence supports stopping after 72–96 hours of apyrexia regardless of ANC.
quick recognition
- febrile patient with recent chemotherapy (highest risk 1–2 weeks post-chemo, nadir ANC at days 12–14)
- screening question: “any chemotherapy or anti-cancer treatment in the last 6 weeks?” — if yes, assume neutropaenia until proven otherwise
- fever may be the only sign of infection — neutropaenic patients lack the inflammatory response to localise infection (no abscess formation, minimal CXR infiltrate, no pyuria)
- absence of localising signs does not mean absence of infection
when to suspect
severity of neutropaenia
| ANC | severity | infection risk |
|---|---|---|
| 1.0–1.5 × 10⁹/L | mild | low — usually safe |
| 0.5–1.0 × 10⁹/L | moderate | increasing risk |
| <0.5 × 10⁹/L | severe | high risk of inadequate total body neutrophils |
| <0.1 × 10⁹/L | profound | highest risk — prophylaxis indicated if >7 days expected |
Risk amplifiers: duration of neutropaenia (>7 days worse), concurrent immunosuppression (chemotherapy, rituximab, T-cell suppressants), mucosal barrier injury.
Duffy null associated neutropaenia
- genetic variant (Duffy antigen/receptor for chemokines) — neutrophils sequester in spleen rather than circulating peripherally → lower measured ANC but normal total body neutrophil count and immune function
- provides evolutionary advantage against Plasmodium vivax malaria
- prevalence: 80–100% in West Africa, ~67% of self-identified Black individuals in North America
- test for Duffy null phenotype when ANC 1.0–1.5 × 10⁹/L in patients of African or Middle Eastern ancestry
- these patients do not have increased infection risk and should not be treated as immunocompromised
diagnosis / investigations
first question: isolated neutropaenia or pancytopaenia?
- pancytopaenia → broader differential including marrow infiltration, aplastic anaemia, MDS → bone marrow biopsy usually indicated
- isolated neutropaenia → haematologic malignancies are uncommon causes; bone marrow biopsy usually NOT recommended
workup of neutropaenia
History drives the diagnosis more than labs in isolated neutropaenia:
- medications — most common cause (see below)
- comorbidities — autoimmune diseases, malabsorption (coeliac, gastric sleeve), nutritional deficiencies (B12, folate, copper)
- pattern — chronic neutropaenia often less concerning than acute
- severity — urgent haematology referral for any new ANC <0.5 × 10⁹/L
medication-induced neutropaenia
| type | mechanism | onset | management | examples |
|---|---|---|---|---|
| type 1 | dose/duration dependent | gradual | dose reduction may suffice | mycophenolate, azathioprine, ganciclovir, linezolid |
| type 2 | immune-mediated (antibody to drug-neutrophil complex) | abrupt, severe | must stop permanently — cannot rechallenge | clozapine, methimazole, PTU, dapsone, sulfasalazine |
febrile neutropaenia workup (within 15 minutes of triage)
- cultures: minimum 2 sets — 1 peripheral + 1 from each CVC lumen
- source assessment: oropharynx (mucositis), catheter sites, lungs, perianal inspection
- no DRE — risk of mucosal tear and bacterial translocation
- bloods: FBC, differential, electrolytes, creatinine, LDH, lactate, liver function
- imaging: CXR has low sensitivity in neutropaenia — low-dose chest CT preferred if respiratory symptoms
- urine: culture only if urinary symptoms or abnormal urinalysis — pyuria may be absent in neutropaenia but routine urine culture is no longer recommended
classification of neutropaenic fever syndromes
- microbiologically documented infection — identified pathogen with source (most useful for guiding targeted therapy)
- clinically documented infection — clear focus (e.g. pneumonia on imaging, cellulitis) without pathogen
- unexplained fever — most common scenario (~60–70%); SIRS without identified source
risk stratification
MASCC score
Determines site of care — inpatient vs. outpatient.
| characteristic | weight |
|---|---|
| burden of illness: no/mild symptoms | 5 |
| burden of illness: moderate symptoms | 3 |
| no hypotension (SBP ≥90 mmHg) | 5 |
| no COPD | 4 |
| solid tumour or no prior fungal infection | 4 |
| no dehydration | 3 |
| outpatient at onset of fever | 3 |
| age <60 years | 2 |
- ≥21: low risk — outpatient management candidate (PO ciprofloxacin + amoxicillin-clavulanate)
- <21: high risk — inpatient admission required
CISNE score is an alternative validated tool for identifying low-risk patients in the ED.
who is high risk?
- haematologic malignancy (leukaemia, lymphoma) — “double hit” to marrow from cancer + chemotherapy → prolonged, severe neutropaenia
- expected neutropaenia >7 days
- significant comorbidities, haemodynamic instability
- solid tumour patients receiving cyclical chemotherapy are generally lower risk — functional marrow, shorter neutropaenia duration
management
step 1: empirical antibiotics within 60 minutes
Mortality increases with each hour of delay. Initiate antibiotics within 60 minutes of presentation — do not wait for culture results.
Standard first-line — antipseudomonal monotherapy:
| agent | dose | notes |
|---|---|---|
| imipenem-cilastatin | 500 mg IV q6h | lowers seizure threshold — caution in CNS disease |
| Piperacillin-Tazobactam | 4.5 g IV q6h | 3.375 g is insufficient for FN |
| cefepime | 2 g IV q8h | common first-line at many Canadian centres |
| meropenem | 1 g IV q8h | reserve for ESBL history or deterioration |
Protocols vary by institution based on local resistance patterns. Always check local antibiogram and guidelines.
Severe penicillin/cephalosporin allergy:
- Vancomycin + aztreonam (preferred) or vancomycin + ciprofloxacin 400 mg IV q8h — consult ID/allergy
step 2: when to add vancomycin
Vancomycin is not standard first-line. Add only if:
- haemodynamic instability / septic shock
- known MRSA colonisation and infection source likely MRSA (gram-positive cocci in blood, pneumonia, soft tissue, severe mucositis)
- clinically apparent line infection (inflamed tunnel/exit site)
- gram-positive cocci on preliminary blood culture
ASCO/IDSA 2018 recommend against routine upfront vancomycin — no mortality benefit and drives resistance.
step 3: resistant organism considerations
- VRE history → add linezolid or daptomycin
- ESBL history → early escalation to carbapenems
- KPC (carbapenemase) → consult ID (ceftazidime-avibactam, cefiderocol, or colistin)
step 4: antifungal coverage
Trigger: persistent fever ≥96 hours (4–7 days) despite broad-spectrum antibiotics in high-risk patients (ANC ≤0.1 × 10⁹/L for ≥7 days, haematologic malignancies especially AML).
Two strategies:
- pre-emptive (preferred if rapid diagnostics available): galactomannan + β-D-glucan + chest CT → initiate antifungal only if positive
- empiric: start antifungal if no rapid diagnostics
Agents (mould-active):
- liposomal amphotericin B 3 mg/kg/day
- voriconazole (if aspergillosis suspected)
- caspofungin 70 mg load → 50 mg daily
prophylaxis in prolonged neutropaenia
Indicated when ANC ≤0.1 × 10⁹/L expected for >7 days (typically haematologic malignancy patients on chemotherapy):
- antibacterial: fluoroquinolone (ciprofloxacin or levofloxacin — pseudomonal coverage)
- antifungal: posaconazole or isavuconazole (yeast + mould coverage)
reassessment and escalation (72–96 hours)
persistent fever, stable patient
- do not switch antibiotics solely for persistent fever if clinically stable
- repeat blood cultures only if clinical deterioration or new fever spike — routine repeats discouraged
- consider chest/abdominal CT to rule out occult source
clinical deterioration
- escalate to meropenem (if on cefepime/pip-taz) ± vancomycin
- reassess for invasive fungal infection
- consider ICU transfer
de-escalation and duration
ECIL-10 (2025): empirical antibiotics can be discontinued after 72–96 hours of apyrexia and clinical recovery irrespective of neutrophil count — the strongest guideline endorsement of this approach.
NCCN 2026: presents discontinuation, de-escalation, or continuation as options for afebrile patients with ANC <0.5 × 10⁹/L — not a blanket recommendation for early stopping.
Traditional approach (ASCO 2018): continue until ANC >0.5 × 10⁹/L.
HOW LONG (2017) randomised patients to stop antibiotics after 72 hours of apyrexia regardless of ANC — no increase in adverse outcomes, but trial design variability limits widespread adoption. ~50% of centres still follow the traditional rule.
- documented infection: treat for standard duration of the specific infection (e.g. pseudomonal bacteraemia per bacteraemia guidelines, not as unexplained FN)
- unexplained fever (FUO): stop after 3–5 days afebrile + stable
- most patients are on antibiotics for 7–10 days total (median time to defervescence ~5 days + additional days afebrile)
- gut mucosal recovery from cytotoxic therapy coincides with this timeline — mucositis → bacterial translocation risk resolves over ~10 days
neutrophil recovery timelines
| cause | expected recovery | notes |
|---|---|---|
| viral infection | 10–14 days | self-limited |
| immune-mediated drug (type 2) | 10–14 days after stopping | rapid once drug cleared — depends on drug half-life |
| dose-dependent drug (type 1) | 3–4 weeks | nadir at 14–21 days post-exposure |
| autoimmune disease | variable | paradox: immunosuppressive treatment for the underlying disease may improve neutropaenia |
| MDS | variable/may not recover | even with treatment |
It takes 7–10 days to produce a mature neutrophil from haematopoietic stem cell — sets the floor for recovery from any cause.
special populations
catheter-related infections
| organism | line management |
|---|---|
| coagulase-negative staph (CoNS) | retain line, treat systemically |
| S. aureus, Candida, Pseudomonas, mycobacteria | prompt removal recommended |
| tunnel or pocket infection | remove line |
In severe thrombocytopaenia or coagulopathy, removal may be deferred until haemostatically safe — but source control is critical for these organisms.
viral considerations
In patients with respiratory symptoms/infiltrates — send multiplex respiratory PCR. Bronchoscopy (BAL) is mandatory if infiltrates suspected to be invasive fungal infection.
G-CSF in febrile neutropaenia
- not routinely recommended for treatment of established FN
- meta-analysis of RCTs: G-CSF decreased hospitalisation time and neutropaenia duration but no improvement in overall mortality or infection-related mortality
- consider in high-risk patients: prolonged neutropaenia >10 days, profound neutropaenia <0.1 × 10⁹/L, age >65, severe sepsis, invasive fungal disease
what NOT to do
- delay antibiotics waiting for culture results or ANC confirmation — treat empirically within 60 minutes
- add vancomycin routinely — no mortality benefit, drives resistance
- “double-cover” Pseudomonas (β-lactam + aminoglycoside) routinely — not indicated unless septic shock or high local resistance
- switch antibiotics solely for persistent fever in a stable patient
- repeat blood cultures routinely without clinical deterioration
- perform DRE — mucosal tear → translocation risk
- use 3.375 g pip-taz for FN — insufficient dose; use 4.5 g q6h
- give G-CSF routinely for FN — no mortality benefit
key trials summary
| trial / source | year | finding |
|---|---|---|
| Klastersky, JCO. 2000 | 2000 | MASCC score validated for identifying low-risk FN patients suitable for outpatient management |
| HOW LONG (2017) | 2017 | antibiotics discontinued after 72h apyrexia regardless of ANC — no increase in adverse outcomes |