Febrile Neutropenia

Meet the patient

A 54-year-old woman is on day 10 after induction for acute myeloid leukaemia. The nurse records a single oral temperature of 38.5 °C. She looks a little washed-out but is talking to you, her blood pressure is normal, and her chest is clear. The full blood count returns an ANC of 0.2 ×10⁹/L.[8][3]

Nothing about her looks dramatic — and that is the trap. In profound neutropenia the inflammatory response is muted, so fever may be the only sign that bacteraemia has begun. From the moment that temperature registered, a clock started running: cultures, then piperacillin-tazobactam, inside one hour.[2][3]

Two numbers, one clock — 38.3 °C and an ANC under 0.5

Febrile neutropenia is a clinical syndrome, not a microbiological diagnosis — and it is one of the most dangerous complications of cytotoxic chemotherapy, radiation, and haematopoietic stem-cell transplantation (HSCT). It is the paradigm of a time-critical oncologic emergency: the neutropenic host cannot mount the inflammatory response that normally localises and signals infection, so fever may be the only manifestation of bacteraemia or early sepsis. Untreated FN kills more than half of patients; prompt empiric therapy brings in-hospital mortality down to roughly 5 to 12 percent.[3][1]

The definition in routine use has two parts, and both are required:[8]

• Fever — a single oral temperature of 38.3 °C or higher, OR two consecutive readings of 38.0 °C or higher taken one hour apart; AND
• Neutropenia — an absolute neutrophil count (ANC) under 0.5 ×10⁹/L (or expected to fall below 0.5 within 48 hours of chemotherapy).[8][10]

Two further terms are essential vocabulary. Profound neutropenia — an ANC under 0.1 ×10⁹/L — carries the highest infection risk and defines the "high-risk" patient. Functional neutropenia describes neutrophils that are present in normal numbers but blunted by chemotherapy, glucocorticoids, or the underlying disease (acute myeloid leukaemia is the classic), so the classical signs of infection — pus, induration, a radiographic infiltrate — are muted or absent.[1]

This biology is the reason for the central maxim of the field: fever in a neutropenic patient is infection until proven otherwise, and the absence of pus does not exclude a life-threatening bacteraemia. Hold that line and every management decision below follows from it.[1][3]

Three axes, one decision — risk, source, organism

FN is stratified along three complementary axes, and together they drive every management decision. They are risk of medical complication (inpatient IV versus oral outpatient care), microbiological certainty of the source (duration and escalation), and causative organism (antibiotic choice and de-escalation).[1][2]

The MASCC score — the bedside stratification tool

The MASCC risk index, built by Klastersky and colleagues and validated across thousands of episodes, is the bedside tool for finding the low-risk patient. A score of 21 or higher identifies low risk (positive predictive value about 91 percent, sensitivity about 71 percent, specificity 68 percent) — the candidate for oral step-down or carefully selected outpatient therapy.[1][2]

[1] [1] [1] [2]

IDSA risk categories — host and depth of neutropenia

The IDSA uses a complementary, host- and depth-oriented split: risk stratification by presenting signs and symptoms, underlying cancer, type of therapy, and medical comorbidities is the recommended starting point of the treatment algorithm.[3] Low-risk patients are candidates for oral step-down or outpatient care with an oral fluoroquinolone plus amoxicillin/clavulanate; high-risk patients need inpatient IV antipseudomonal therapy commenced promptly.[2][8]

Microbiological classification — drives escalation

• Microbiologically documented infection (MDI) — a pathogen isolated from a sterile site.
• Clinically documented infection (CDI) — a focal source identified (pneumonia, cellulitis, line-site) without a positive culture.
• Unexplained fever (FUO) — no source found after standard workup: the largest single group in trial endpoint taxonomy.[11]

How common, who, and why the nadir bites

FN complicates roughly 10 to 50 percent of solid-tumour chemotherapy cycles, and 80 percent or more of acute myeloid leukaemia induction and myeloablative HSCT conditioning. Overall in-hospital mortality of treated FN is 5 to 10 percent in contemporary cohorts, but it climbs to 30 to 50 percent in patients who develop septic shock or invasive fungal infection. Mortality is concentrated in the first 48 hours, which is why the door-to-antibiotic interval is the single most modifiable outcome.[3][1]

The risk of FN after any given cycle is set by an interplay of regimen intensity, host factors, and the integrity of the mucosal barrier. The principal determinants are:[1]

• Chemotherapy intensity and regimen — high-dose cytarabine, anthracyclines, platinum agents, taxanes, and alkylating agents produce deeper and more prolonged nadirs than weekly fluoropyrimidines or low-dose oral agents. Acute leukaemia induction and HSCT conditioning carry the highest per-cycle risk.
• Expected nadir depth and duration — risk rises sharply once the ANC falls below 0.5 ×10⁹/L and climbs further when profound neutropenia (under 0.1 ×10⁹/L) is expected to last more than 7 days.
• Tumour type — haematological malignancies carry the highest risk; among solid tumours, small-cell lung cancer, germ-cell tumours, and aggressive lymphomas are particularly dangerous.
• Patient factors — age over 65, poor performance status (ECOG 2 or higher), hypoalbuminaemia, baseline renal or hepatic dysfunction, advanced-stage disease.
• Prior febrile neutropenia — a previous episode roughly doubles the risk of recurrence and is one of the strongest triggers for primary prophylaxis with G-CSF.
• Oral mucositis and mucosal barrier injury — severe mucositis roughly doubles the rate of Gram-negative and viridans-streptococcal bacteraemia.
• Indwelling devices — central venous catheters (tunnelled, PICC, port), urinary catheters, and prosthetic material are niches for coagulase-negative staphylococci, Staphylococcus aureus, and Candida biofilms.
• Corticosteroids and immunosuppression — prolonged steroids impair neutrophil function and T-cell immunity, raising the risk of Pneumocystis, viral reactivation, and mould infection.
• Prior colonisation with resistant organisms (MRSA, VRE, ESBL, carbapenem-resistant Pseudomonas or Acinetobacter) and prior invasive fungal infection.[5]

Five roads to bacteraemia — and why the signs go missing

Neutrophils are the principal effector of innate defence against bacteria and fungi. When the ANC falls below 0.5 ×10⁹/L — and especially below 0.1 ×10⁹/L — five overlapping mechanisms converge to produce clinical infection, and together they explain why the bedside can look deceptively quiet.[1]

The dominant cascade is mucosal barrier injury driving bacterial translocation. Cytotoxic chemotherapy is taken up by rapidly dividing cells — not only the haematopoietic marrow but also the oral, oesophageal, and intestinal epithelium. The mucosa ulcerates, loses its tight-junction integrity, and lets luminal organisms cross into the lamina propria and lymphatics. The nadir typically falls 7 to 14 days after a cytotoxic cycle and recovers over the following week, so the patient is most vulnerable precisely when both neutrophil number and mucosal integrity are at their lowest. The gastrointestinal tract is the single largest portal of entry.[1]

The five mechanisms, named for viva recall:[1]

1. Quantitative neutrophil deficiency — too few circulating neutrophils to contain the inoculum at any portal of entry: skin, oropharynx, lung alveoli, gut lumen, urinary tract.
2. Qualitative neutrophil dysfunction — chemotherapy, glucocorticoids, and the underlying haematological disease impair chemotaxis, phagocytosis, and the oxidative burst, so even the neutrophils that remain are blunted.
3. Mucosal barrier injury — chemotherapy-induced ulceration of the oral and intestinal mucosa, the principal route of Gram-negative bacteraemia and Candida translocation.
4. Indwelling central venous catheters — biofilm on the catheter hub and lumen, a persistent reservoir for coagulase-negative staphylococci, S. aureus, Candida, and Corynebacterium jeikeium.
5. Impaired adaptive surveillance — chemotherapy-induced lymphopenia blunts T-cell and humoral responses, permitting herpesvirus reactivation (HSV, VZV, CMV, EBV) and invasive mould infection.[1]

The organisms — Gram-positives first in frequency, Pseudomonas first in death

The microbiology of FN has shifted over thirty years, but a recognisable pattern remains — and the two sentences that earn marks are these: Gram-positives are the most frequent cause of bacteraemia, and Pseudomonas aeruginosa carries the highest mortality. That second sentence is the reason every empiric regimen must be antipseudomonal.[1][3]

Gram-positive organisms now account for over half of documented bloodstream infections: coagulase-negative staphylococci (line-related), Staphylococcus aureus (including MRSA), viridans streptococci (classically after high-dose cytarabine), enterococci (including VRE), and Streptococcus pneumoniae. The Gram-negatives that kill are Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, enterobacter, and the increasingly important carbapenem-resistant Enterobacteriaceae.[1][5]

Fever first, silence after — the muted exam

Fever is usually the first and only sign. Because the inflammatory response is blunted, a single rigour in a chemotherapy patient on day 10 post-cycle should trigger the full febrile-neutropenia pathway regardless of the measured temperature. Up to half of episodes never yield an identified source even after exhaustive investigation.[1]

• Fever meeting the definition — a single oral temperature over 38.3 °C, or two consecutive readings over 38.0 °C taken one hour apart.[8]
• May be the only symptom — roughly half of episodes have no identifiable localising source.
• Localising symptoms when present — cough, dyspnoea or pleuritic pain (pneumonia); sore throat and odynophagia (mucositis); dysuria (urinary tract infection); perianal pain (perianal abscess, fissure); right-lower-quadrant pain and diarrhoea (typhlitis, C. difficile colitis).
• Sepsis physiology — hypotension, tachycardia, tachypnoea, new confusion, oliguria, mottled or cool peripheries — mandates resuscitation alongside the first antibiotic dose.[18]
• Central-line site — erythema, tenderness, purulent exudate at the exit site, or tenderness along the tunnel.
• Skin — cellulitis, ecthyma gangrenosum (the classic haemorrhagic necrotic lesion of Pseudomonas septicaemia), petechiae, scattered pustules suggesting disseminated fungal infection, vesicles of HSV or VZV reactivation.[1]

Atypical presentations are deliberately tested at viva, because they are the ones that kill through delay. Elderly and diabetic patients may be afebrile or hypothermic in the face of overwhelming sepsis — a normal temperature in a sick-appearing neutropenic patient is itself a red flag. Patients on chronic steroids may not mount a fever. The profoundly neutropenic patient with pneumonia may have a normal chest examination and a near-normal chest X-ray early on, because there are no neutrophils to generate the infiltrate — chest radiography has limited sensitivity for pneumonia in FN.[24]

Infection until proven otherwise — the exclusions come later

Fever in a neutropenic patient is treated as infection until proven otherwise, and the non-infectious mimics are diagnoses of exclusion. Antibiotics come first; the differential is revisited only if fever persists despite appropriate therapy.[3]

• Tumour (neoplastic) fever — fever driven by the malignancy itself, considered only once other causes are excluded; in a sarcoma-series study, temperatures in most naproxen-treated patients with neoplastic fever normalised within 24 hours, and tumour resection produced prompt, complete lysis of fever.[16]
• Drug fever after chemotherapy — fever after cancer chemotherapy clusters on post-treatment days 3 and 4, with a second peak at days 10 to 14 that overlaps the neutropenic nadir; a non-infectious timing pattern worth recognising — but every such fever is still treated as infection first.[17]
• Deep-seated fungal infection — invasive aspergillosis, candidiasis, and mucormycosis are the classic causes of persistent fever despite broad-spectrum antibacterials; earlier detection has reshaped the debate between empirical and preemptive antifungal therapy.[3][13]

[16] [17] [3]

The bedside round — document the baseline

The first hour is structured, and the job of examination is to detect complications and document a baseline — not to make the diagnosis. In FN the trend in vital signs, not the absolute value, drives escalation, so a documented starting point is non-negotiable.[1]

• Vital signs and perfusion — temperature, heart rate, blood pressure, respiratory rate, oxygen saturation, urine output, capillary refill, mental state. Any organ dysfunction automatically up-classifies the patient to high risk.[18]
• Central venous catheter — inspect the exit site, palpate along the tunnel and over the port pocket for tenderness, check the connections.
• Oral cavity — mucosal erythema, ulceration, thrush, herpes vesicles, dental abscess.
• Skin head to toe — petechiae, cellulitis, ecthyma gangrenosum, pustules, VZV lesions, catheter-site infection.
• Perianal region — inspect for fissure, ulceration, or abscess, and avoid invasive instrumentation of the rectum in the profoundly neutropenic patient.
• Respiratory — auscultate for crackles, consolidation, effusion, but the examination may be entirely silent.
• Abdomen — distension, right-lower-quadrant tenderness (typhlitis), hepatosplenomegaly, rebound.
• Neurological and fundoscopy — mental state, and, if visual or neurological symptoms are present, fundoscopy for CMV retinitis, candida endophthalmitis, or Roth spots.[1]

Cultures, then the antibiotic — in that order, in that hour

The initial workup is fast and pre-antibiotic in principle — but the antibiotic is never delayed for investigations: the patient receives immediate basic diagnostic procedures and broad-spectrum antibiotics are initiated.[8][3]

• Full blood count and differential — confirm the ANC (under 0.5 ×10⁹/L defines the syndrome) and review the trend.[8]
• Blood cultures drawn immediately — central and peripheral cultures where a central venous catheter is in place, before the first antibiotic dose, labelled with site and time.[23]
• Renal and liver panels, coagulation — baseline organ function; evaluation is repeated on a daily basis to steer duration and de-escalation.[8]
• Lactate — a value over 2 mmol/L with vasopressor-dependent hypotension defines septic shock and mandates critical-care involvement.[18]
• Chest imaging — chest radiography has limited performance for detecting pneumonia in FN, so the imaging threshold should be low and cross-sectional imaging adds yield when symptoms or fever persist.[24]
• Biomarkers and targeted microbiology in prolonged fever — recommended biomarkers (such as galactomannan and beta-D-glucan) and radiologic investigations refine the evaluation of invasive fungal disease in prolonged FN; further cultures and viral testing follow symptoms and local epidemiology.[6][8]

The 60-minute door — what runs in parallel

The target is the first antibiotic dose within 1 hour of triage, and resuscitation, cultures, and the antibiotic run in parallel — not in sequence.[2]

1. Airway, Breathing, Circulation assessment and resuscitation; escalate to critical care for shock physiology or new organ dysfunction.[18]
2. Blood cultures drawn immediately — central and peripheral cultures where a central line sits — but do not delay the antibiotic; immediate treatment is the universal key of neutropenic fever management.[23][3]
3. Empirical broad-spectrum antipseudomonal β-lactam IV within 1 hour of triage — see the definitive ladder below.[2]
4. Vasopressors and fluids for shock — septic shock is identified by a vasopressor requirement to maintain mean arterial pressure of 65 mmHg or greater with lactate over 2 mmol/L; resuscitate in parallel.[18]
5. Gram-positive or combination additions are not routine — additions are reserved for resistant-pathogen risk factors or a complicated presentation.[5]
6. Daily re-evaluation and source control — evaluate daily; formally modify the regimen at 72 to 96 hours based on clinical course and microbiological results, and control the source (drain collections, remove infected lines) as clinically indicated.[8][5]

Piperacillin-tazobactam first — additions only with a reason

After the first empirical dose, daily re-evaluation answers three questions: is there a documented organism, is the patient clinically improving, and is the source controlled? The day-0 antibiotics are then escalated, de-escalated, or continued on the strength of those answers.[5][8]

First-line — antipseudomonal monotherapy

Empiric antibiotic monotherapy is considered the standard initial treatment for febrile neutropenic patients with cancer, with combination regimens reserved for specific situations.[11]

• Piperacillin-tazobactam 4.5 g IV every 6 to 8 hours — the workhorse first-line monotherapy; both the every-6-hour and every-8-hour schedules at 4.5 g are established in randomised FN trial practice.[9][10]
• Cefepime 2 g IV every 8 hours (adults) — antipseudomonal fourth-generation cephalosporin monotherapy, tested head-to-head in randomised FN trials.[12]
• Carbapenems and combinations are not routine — an escalation approach that avoids empirical carbapenems and combinations is advised for patients without particular risk factors; a de-escalation approach (initial broad-spectrum therapy or combinations) applies with known prior colonisation or infection with resistant pathogens, a complicated presentation, or prevalent local resistance.[5]
• The initial regimen must cover both Gram-positive and Gram-negative pathogens broadly.[3]

Vancomycin is not routine — additions follow risk factors

Empirical additions to the baseline regimen — including glycopeptides — are not routine: they belong to patients with known prior colonisation or infection with resistant pathogens, a complicated presentation, or centres where resistant pathogens are prevalent, and any modification of the initial regimen is formally made at 72 to 96 hours based on the clinical course and microbiological results.[5][11] The baseline regimen itself is chosen to treat both Gram-positive and Gram-negative pathogens broadly, so a routine Gram-positive "top-up" adds nothing for the stable patient.[3]

Anaerobes and the right lower quadrant

For perianal or intra-abdominal sources — suspected typhlitis (neutropenic enterocolitis) or severe mucositis — antimicrobial therapy should cover the most frequent pathogens of that syndrome, taking into consideration the local epidemiology; an antipseudomonal β-lactam with anaerobic activity is a pragmatic choice.[15][8]

Escalation — persistent fever, shock, or a new belly

• Persistent fever at 48 to 72 hours without a source — continue daily evaluation; antibiotic rotation and/or antifungal therapy may be necessary, and the initial regimen is formally modified at 72 to 96 hours based on clinical course and microbiology.[8][5]
• New haemodynamic instability at any time — treat as septic shock physiology: resuscitate, broaden cover for resistant organisms where risk factors exist, and involve critical care.[18][5]
• Fever with new abdominal pain — think typhlitis: confirm with imaging (bowel-wall thickening on ultrasound or CT), cover the most frequent enteric pathogens per local epidemiology, rest the bowel, and watch for complications.[15]

Persistent fever — add an antifungal

For persistent fever despite 48 to 96 hours of broad-spectrum antibacterials in a high-risk host, empirical antifungal therapy is the classic addition — with ongoing debate between empirical and biomarker-driven preemptive strategies.[3][8]

• Caspofungin — in a randomised trial of 1095 persistently febrile neutropenic patients, caspofungin was as effective as liposomal amphotericin B as empirical therapy and generally better tolerated, with less nephrotoxicity (2.6 versus 11.5 percent).[13]
• Liposomal amphotericin B 3 mg/kg IV daily — the standard-dose regimen for invasive mould infection; a randomised trial found higher loading doses added no benefit.[14]
• Voriconazole — first-line for proven or probable invasive aspergillosis, with demonstrated efficacy and survival benefit over amphotericin B deoxycholate.[14]

De-escalation and duration — narrow when the bug is known

• If a single susceptible organism is identified and the source is controlled, narrow to the smallest-spectrum agent that covers it, and modify the initial regimen at 72 to 96 hours based on clinical course and microbiological results.[5]
• Unexplained fever with a stable course — in patients haemodynamically stable since presentation and afebrile for at least 48 hours, antibiotics can be discontinued after 72 hours or later, irrespective of the neutrophil count and expected duration of neutropenia.[5]
• Duration and daily evaluation — duration of treatment depends on the neutrophil count, the patient's symptoms, and fever, with evaluation on a daily basis; if fever continues after 48 to 72 hours, antibiotic rotation and/or antifungal therapy may be necessary.[8]

Low-risk step-down — an oral fluoroquinolone plus amoxicillin-clavulanate

For the carefully selected low-risk patient — identified by clinical judgment or a validated tool such as the MASCC risk index, with psychosocial and logistic considerations satisfied — oral step-down or even primary outpatient therapy is guideline-supported, building on pilot data showing home antibiotic therapy is feasible after brief inpatient observation:[2][4]

• An oral fluoroquinolone plus amoxicillin/clavulanate is the recommended empirical oral regimen — with clindamycin substituted for the penicillin-allergic patient — unless fluoroquinolone prophylaxis was already in use before the fever developed.[2]
• Patients being managed as outpatients are monitored for at least 4 hours before discharge, and any patient who has not defervesced after 2 to 3 days of initial empirical broad-spectrum therapy is re-evaluated and considered a candidate for inpatient treatment.[2]

Source control — pull the line when the bug demands it

• Remove the central line for Staphylococcus aureus, Pseudomonas aeruginosa, Candida, or mycobacterial bacteraemia; persistent bacteraemia after 72 hours of appropriate therapy; clinical tunnel infection; or septic emboli.
• Drainage of abscesses, empyema, septic joints; perianal abscess may need incision and drainage with platelet support.
• Surgical consultation for typhlitis with perforation, ischaemia, or uncontrolled bleeding.[15]

G-CSF — prophylaxis is the evidence-based use

The evidence-based use of G-CSF around febrile neutropenia is prophylactic, not therapeutic. ASCO recommends prophylactic colony-stimulating factors to reduce the risk of FN when the chemotherapy regimen carries an FN risk of approximately 20 percent or higher and no equally effective and safe regimen that does not require CSFs is available.[7]

• Pegfilgrastim — a single fixed 6 mg subcutaneous dose per chemotherapy cycle provided neutrophil support comparable to daily filgrastim injections in a randomised phase III trial, with a trend towards fewer febrile-neutropenia episodes across cycles.[21]
• Filgrastim — daily subcutaneous injections remain the alternative; the trial above found the two strategies comparable for efficacy and safety.[21]

The hour, the day, the week — a stepwise table

A practical MASCC-driven decision pathway, from the temperature that triggers the pathway to the day the antibiotic stops:[1][2][4]

| Step | Action | Target / threshold |[1] |---|---|---| | 1. Recognise | Oral temp over 38.3 °C + ANC under 0.5 | Immediate | | 2. Draw cultures | Central plus peripheral | Immediately, never delaying antibiotics | | 3. Stratify | Compute MASCC score | Within the first hour | | 4. First-dose antibiotic | Piperacillin-tazobactam 4.5 g IV | Within 60 min of triage | | 5. Reassess | Cultures, clinical course, source | Daily review | | 6. Escalate or de-escalate | Modify regimen at 72–96 h; add antifungal if persistent fever | Per clinical course | | 7. Step down | MASCC 21+ and afebrile and cultures negative → oral | Once stable | | 8. Discontinue | Stable and afebrile at least 48 h → stop irrespective of ANC | From 72 h |

[1] [5]

Typhlitis, lines, and the scenarios that change the plan

Most episodes follow the standard pathway, but a handful of scenarios force a specific deviation — and each is high-yield exam material.[8]

• Unexplained fever (FUO) — the most common category in trial endpoint taxonomy; treat empirically and re-evaluate daily, intensifying the search for fungal sources if fever persists.[11][8]
• Clinically documented infection (CDI) — a focal source is identified (pneumonia, cellulitis, line-site) but no organism is isolated; microbiologically documented infection (MDI) — a pathogen from a sterile site. These categories shape trial endpoints and de-escalation decisions.[11]
• Persistent fever at 48 to 96 hours — antibiotic rotation and/or empirical antifungal therapy may be necessary; modify the regimen on clinical course and cultures.[8][5]
• Empirical antifungal therapy — caspofungin (non-inferior to and better tolerated than liposomal amphotericin B) or liposomal amphotericin B 3 mg/kg daily; voriconazole if invasive aspergillosis is proven or probable.[13][14]
• Breakthrough bacteraemia on therapy — take a de-escalation approach with initial broad-spectrum cover when there is known prior colonisation or infection with resistant pathogens, a complicated presentation, or prevalent local resistance; reassess source control.[5]
• Typhlitis (neutropenic enterocolitis) — fever plus abdominal pain plus neutropenia, with the diagnosis resting on clinical features and bowel-wall thickening on ultrasound or CT. Treat with antimicrobials covering the most frequent pathogens per local epidemiology, bowel rest, and supportive care — the evidence base is largely expert opinion rather than randomised trials.[15]
• Paediatric febrile neutropenia — paediatric guidelines list a fourth-generation cephalosporin for empirical therapy in high-risk FN, refine risk stratification for invasive fungal disease, and make a weak recommendation to withhold empirical antifungal therapy in IFD-low-risk patients with prolonged fever.[6]
• Haematological malignancies — high-risk, profoundly neutropenic hosts: monotherapy remains standard, with the escalation-versus-de-escalation choice driven by local resistance epidemiology and patient risk factors.[11][5]

What kills them — shock, mould, and the missed dose

The complications are the causes of death, and most are preventable by the discipline of the first hour and the vigilance of the first week.[1]

• Septic shock and multiorgan dysfunction — the leading cause of early death; identified by vasopressor requirement to maintain MAP 65 mmHg or greater with lactate over 2 mmol/L, and carrying hospital mortality above 40 percent.[18]
• Invasive fungal superinfection — candidaemia, pulmonary and disseminated aspergillosis, mucormycosis; high mortality even with appropriate therapy.[13][14]
• Acute kidney injury and drug toxicity — including amphotericin-related nephrotoxicity, which was significantly less frequent with caspofungin in the empirical-therapy randomised trial (2.6 versus 11.5 percent).[13]
• Line loss — necessitating re-insertion in a thrombocytopenic patient.
• Death — overall in-hospital mortality 5 to 10 percent, rising steeply with septic shock (mortality above 40 percent by the Sepsis-3 definition).[18]

The classic pitfalls are the recurring trainee errors, and examiners test every one: (1) delaying antibiotics while waiting for the "perfect" culture; (2) omitting central-line cultures; (3) performing invasive rectal instrumentation of the profoundly neutropenic patient; (4) adding routine combination or Gram-positive cover to every patient; (5) failing to escalate to an antifungal when fever persists at 48 to 96 hours; (6) premature outpatient discharge of a borderline MASCC patient; (7) missing typhlitis as the cause of right-lower-quadrant pain; (8) forgetting that G-CSF regimens themselves can colour the fever picture through drug fever.[5][17]

Three variables decide the outcome

Prognosis is driven by three variables: the depth and duration of neutropenia, the patient's haemodynamic reserve, and the time-to-antibiotic.[1]

• Favourable — MASCC low risk (score 21 or higher), prompt response to the first dose. Low-risk mortality is under 5 percent.[1]
• Unfavourable — prolonged profound neutropenia, allogeneic HSCT, refractory or relapsed leukaemia, septic shock, invasive mould disease, multidrug-resistant bacteraemia.[1][5]

Disposition:[1]

• High-risk (MASCC under 21) — inpatient IV therapy; discontinue only when the patient has been haemodynamically stable since presentation and afebrile for at least 48 hours.[5]
• Low-risk (MASCC 21 or higher) — oral step-down or outpatient therapy is feasible in carefully selected patients, with monitoring for at least 4 hours before any discharge.[2][1]
• ICU — any organ dysfunction: shock physiology, hypoxia, acute kidney injury, encephalopathy.[18]

The child, the pregnant, and the β-lactam-allergic

Special populations change the thresholds, the drugs, or both — name the adjustment before you name the dose.[8]

• Paediatric — dedicated paediatric guidelines apply: a fourth-generation cephalosporin is listed for empirical therapy in high-risk FN, risk stratification is refined for invasive fungal disease, and empirical antifungal therapy may be withheld (weak recommendation) in IFD-low-risk patients with prolonged fever.[6]
• Penicillin-allergic patients — for low-risk oral therapy, the ASCO/IDSA outpatient guideline substitutes clindamycin for amoxicillin/clavulanate alongside the fluoroquinolone; inpatient choices follow local allergy policy and resistance epidemiology.[2]
• Centres with prevalent resistance — the escalation-versus-de-escalation choice at presentation depends on the local resistance epidemiology and the patient's risk factors for resistant bacteria and for a complicated clinical course.[5]

Where the guidelines agree, and where they diverge

Landmark evidence

• Klastersky et al. (2000) — established and validated the MASCC risk index, defining the score of 21 as the low-risk threshold (PPV 91 percent) and enabling safe outpatient selection.[1]
• Freifeld et al. (IDSA 2010/2011) — the foundational guideline: risk stratification as the starting point and immediate, broad empirical therapy as the unchanging indication.[3]
• Taplitz et al. (ASCO/IDSA 2018) — outpatient management: the 1-hour initial-dose rule, at least 4 hours of pre-discharge monitoring, and the fluoroquinolone plus amoxicillin/clavulanate oral regimen.[2]
• Averbuch et al. (ECIL 2013) — European guidance on escalation versus de-escalation in the era of growing resistance, with modification at 72–96 hours.[5]
• Talcott et al. (1994) — the original pilot showing home therapy was feasible for low-risk patients, the foundation of modern outpatient pathways.[4]
• Smith/ASCO (2015) — G-CSF guidance: primary prophylaxis when FN risk is approximately 20 percent or higher.[7]
• Lehrnbecher et al. (2017) — the paediatric FN guideline update, including empirical antifungal recommendations.[6]
• Walsh et al. (2004) — the randomised trial establishing caspofungin as effective and better tolerated than liposomal amphotericin B for empirical antifungal therapy in persistent fever and neutropenia.[13]

Stop the nadir before it starts — prevention

Prevention is the highest-yield intervention in FN and is more effective than treatment. The strategy is layered: G-CSF prophylaxis, antimicrobial prophylaxis, and non-pharmacological measures.[7][19]

G-CSF primary prophylaxis

For chemotherapy regimens carrying an FN risk of approximately 20 percent or higher, primary prophylactic G-CSF is recommended, informed by age, medical history, disease characteristics, and the myelotoxicity of the regimen:[7]

• Pegfilgrastim 6 mg subcutaneously as a single fixed dose per chemotherapy cycle — comparable neutrophil support to daily filgrastim in a randomised phase III trial.[21]
• Filgrastim as daily subcutaneous injections after chemotherapy remains the alternative, with comparable efficacy and safety in the same trial.[21]

Antimicrobial prophylaxis

• Antibiotic (quinolone) prophylaxis in afebrile neutropenic patients reduces all-cause mortality (relative risk 0.67 in a meta-analysis of 95 trials) and infection incidence, at the price of more adverse events (relative risk 1.69).[19] Its ecological cost is real: in a centre using levofloxacin prophylaxis in AML, 78.9 percent of isolated E. coli were ciprofloxacin-resistant — local microbiology surveillance is mandatory.[22]
• Antifungal prophylaxis — posaconazole showed superior clinical efficacy over fluconazole or itraconazole in preventing invasive fungal disease among neutropenic patients with AML or MDS.[20]

Non-pharmacological

• Hand hygiene — the single most effective infection-control measure.
• Avoid fresh flowers and plants (soil-borne Pseudomonas, Aspergillus).
• Avoid crowds and sick contacts; protective isolation with high-efficiency particulate (HEPA) filtration for the highest-risk patients.
• A neutropenic diet (avoiding raw foods, soft cheeses, unpasteurised dairy) is traditional but controversial — modern evidence does not show a clear reduction in infection, and most centres now recommend standard food-safety advice rather than a strict "neutropenic" diet.
• Vaccination of household contacts (influenza, COVID-19) and of the patient once recovered; live vaccines are contraindicated during active chemotherapy.[1]

Ward-round test

Stem 1 — the day-10 patient with a single temperature

A 54-year-old on day 10 of AML induction has a single oral temperature of 38.5 °C and an ANC of 0.2 ×10⁹/L. She looks well, her blood pressure is normal, and her chest is clear. The registrar wants to examine the rectum to "look for a perianal source" and to wait for the blood cultures before giving anything. What do you do, in what order, and within what time?[8][3]

Answer

This is febrile neutropenia — a single oral temperature over 38.3 °C with an ANC under 0.5 ×10⁹/L. The clock is already running. Draw blood cultures immediately — central plus peripheral where a central line sits — and give an empirical antipseudomonal β-lactam: piperacillin-tazobactam 4.5 g IV, with the first dose within 1 hour of triage.[8][23][9][2] Do not wait for the cultures before the antibiotic — if access is hard, give the antibiotic first and try the cultures again; immediate treatment is the universal key in neutropenic fever.[3] Inspect the perianal skin rather than performing invasive rectal instrumentation in the profoundly neutropenic patient. MASCC will guide later step-down; it does not delay the first dose.[1]

Stem 2 — day 5 and still febrile

The same patient is still febrile on day 5 of piperacillin-tazobactam, with no organism grown and a normal repeat CT chest. She remains profoundly neutropenic. What is the next drug class, and why?[3][13]

Answer

Persistent fever after several days of broad-spectrum antibacterials in a high-risk neutropenic host is a fungus until proven otherwise — add an empirical antifungal. The randomised evidence: caspofungin was as effective as and generally better tolerated than liposomal amphotericin B as empirical antifungal therapy in 1095 persistently febrile neutropenic patients, with significantly less nephrotoxicity (2.6 versus 11.5 percent); liposomal amphotericin B 3 mg/kg daily is the standard-dose mould-active alternative, and voriconazole carries an efficacy and survival benefit over amphotericin B deoxycholate when invasive aspergillosis is proven or probable.[13][14] Biomarkers (galactomannan, beta-D-glucan) and radiologic investigations refine the evaluation of invasive fungal disease in prolonged FN.[6]

Stem 3 — the borderline MASCC score

A 60-year-old on chemotherapy for a germ-cell tumour scores MASCC 20, is afebrile at 48 hours with negative cultures, and the registrar wants to discharge him home on oral ciprofloxacin plus co-amoxiclav. What is the correct disposition, and why?[1][2]

Answer

Keep him as an inpatient on IV antipseudomonal therapy. A MASCC score under 21 does not meet the validated low-risk threshold — 21 or higher — so he remains in the high-risk pathway.[1] Outpatient management is reserved for patients judged low-risk by clinical criteria or the validated MASCC tool, with psychosocial and logistic considerations satisfied, at least 4 hours of pre-discharge monitoring, and re-evaluation with consideration of admission if fever has not settled within 2 to 3 days of empirical therapy.[2] The oral regimen (a fluoroquinolone plus amoxicillin/clavulanate) is appropriate only once the low-risk criteria are genuinely met.[2]

Stem 4 — the right lower quadrant and the drug to add

A patient on day 12 of induction develops fever, right-lower-quadrant pain and tenderness, and diarrhoea, with ANC 0.1. CT shows caecal wall thickening with pneumatosis. Name the syndrome and the management bundle.[15]

Answer

Typhlitis (neutropenic enterocolitis) — the diagnosis rests on fever and abdominal pain in a neutropenic patient together with bowel-wall thickening on ultrasound or CT.[15] Management is antimicrobial therapy with a regimen that covers the most frequent pathogens, taking into consideration the local epidemiology, together with bowel rest and supportive care. Note that the evidence base is largely expert opinion rather than randomised trials, and complications — bleeding, perforation, ischaemia — are what force surgical involvement. Do not delay the antibiotics for imaging if the patient is unstable.[15]

References

1. [1]Klastersky J, Paesmans M, Rubenstein EB, et al. The Multinational Association for Supportive Care in Cancer risk index: A multinational scoring system for identifying low-risk febrile neutropenic cancer patients J Clin Oncol, 2000.PMID 10944139
2. [2]Taplitz RA, Kennedy EB, Bow EJ, et al. Outpatient Management of Fever and Neutropenia in Adults Treated for Malignancy: American Society of Clinical Oncology and Infectious Diseases Society of America Clinical Practice Guideline Update J Clin Oncol, 2018.PMID 29461916
3. [3]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.PMID 21258094
4. [4]Talcott JA, Whalen A, Clark J, Rieker PD, Finberg R Home antibiotic therapy for low-risk cancer patients with fever and neutropenia: a pilot study of 30 patients based on a validated prediction rule J Clin Oncol, 1994.PMID 8270967
5. [5]Averbuch D, Orasch C, Cordonnier C, et al. European guidelines for empirical antibacterial therapy for febrile neutropenic patients in the era of growing resistance: summary of the 2011 4th European Conference on Infections in Leukemia Haematologica, 2013.PMID 24323983
6. [6]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.PMID 28459614
7. [7]Smith TJ, Bohlke K, Lyman GH, et al. Recommendations for the Use of WBC Growth Factors: American Society of Clinical Oncology Clinical Practice Guideline Update J Clin Oncol, 2015.PMID 26169616
8. [8]Conen K, Joerger M, Omlin A, et al. [Fever with chemotherapy induced neutropenia] Ther Umsch, 2014.PMID 24394205
9. [9]Sime FB, Udy AA, Roberts JA, et al. Can therapeutic drug monitoring optimize exposure to piperacillin in febrile neutropenic patients with haematological malignancies? A randomized controlled trial J Antimicrob Chemother, 2015.PMID 25953805
10. [10]Aynioglu A, Caglasin M, Baran T, et al. A comparison of the efficacy of piperacillin-tazobactam and cefoperazone-sulbactam therapies in the empirical treatment of patients with febrile neutropenia Rev Esp Quimioter, 2016.PMID 26888338
11. [11]Bucaneve G, Micozzi A, Menichetti F, et al. Results of a multicenter, controlled, randomized clinical trial evaluating the combination of piperacillin/tazobactam and tigecycline in high-risk hematologic patients with cancer with febrile neutropenia J Clin Oncol, 2014.PMID 24733807
12. [12]Ponraj M, Alexander SP, Anandan S, et al. Cefepime vs. cefoperazone/sulbactam in combination with amikacin as empirical antibiotic therapy in febrile neutropenia Support Care Cancer, 2018.PMID 29774477
13. [13]Walsh TJ, Teppler H, Donowitz GR, et al. Caspofungin versus liposomal amphotericin B for empirical antifungal therapy in patients with persistent fever and neutropenia N Engl J Med, 2004.PMID 15459300
14. [14]Cornely OA, Maertens J, Bresnik M, et al. Liposomal amphotericin B as initial therapy for invasive mold infection: a randomized trial comparing a high-loading dose regimen with standard dosing (AmBiLoad trial) Clin Infect Dis, 2007.PMID 17443465
15. [15]Portugal R, Nucci M Typhlitis (neutropenic enterocolitis) in patients with acute leukemia: a review Expert Rev Hematol, 2017.PMID 28075196
16. [16]Nakamura T, Kuranishi R, Maeda T, et al. Neoplastic fever in patients with bone and soft tissue sarcoma Mol Clin Oncol, 2016.PMID 27900101
17. [17]Ogawara D, Noda Y, Kadowaki S, et al. Drug fever after cancer chemotherapy is most commonly observed on posttreatment days 3 and 4 Support Care Cancer, 2016.PMID 26108172
18. [18]Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) JAMA, 2016.PMID 26903338
19. [19]Gafter-Gvili A, Fraser A, Paul M, et al. Meta-analysis: antibiotic prophylaxis reduces mortality in neutropenic patients Ann Intern Med, 2005.PMID 15968013
20. [20]Sung AH, Walker RM, Kanda J, et al. An update to the cost-effectiveness of posaconazole vs fluconazole or itraconazole in the prevention of invasive fungal disease among neutropenic patients in the United States J Med Econ, 2015.PMID 25524741
21. [21]Green MD, Hudis C, Kornbluth N, et al. A randomized double-blind multicenter phase III study of fixed-dose single-administration pegfilgrastim versus daily filgrastim in patients receiving myelosuppressive chemotherapy Ann Oncol, 2003.PMID 12488289
22. [22]De Rosa FG, Miletich F, Isabella G, et al. Epidemiology of bloodstream infections in patients with acute myeloid leukemia undergoing levofloxacin prophylaxis BMC Infect Dis, 2013.PMID 24289496
23. [23]Barriga FJ, Cofán F, Nubiola A, et al. Efficacy of a vancomycin solution to prevent bacteremia associated with an indwelling central venous catheter in neutropenic and non-neutropenic cancer patients Med Pediatr Oncol, 1997.PMID 9024516
24. [24]Hwang EJ, Park S, Cho YJ, et al. Deep learning computer-aided detection system for pneumonia in febrile neutropenia patients: a diagnostic cohort study BMC Pulm Med, 2021.PMID 34876075