Fever of Unknown Origin in Neutropenic Cancer Patients: A Patient-Friendly Guide to the 2017 German AGIHO Guidelines

Table of Contents

• Key Points
• Background: Why This Guideline Matters
• How These Guidelines Were Developed
• Understanding the Grading System (What the Letters and Numbers Mean)
• Key Definitions: Neutropenia and Fever
• Risk Stratification: Standard Risk vs. High Risk
• Criteria for Treating a Patient at Home (Outpatient Management)
• Which Germs Cause These Infections?
• Antibiotic Prophylaxis, Resistance, and Local Epidemiology
• Baseline Testing Before Chemotherapy Starts
• Screening for Fungal Infections in Patients Without Symptoms
• What to Do at the Onset of Fever
• Empirical Antibacterial Therapy: First-Line Choices
• Newer Antibacterial Agents and Allergy Alternatives
• Combination Antibacterial Therapy
• When to Add Antifungal Therapy
• Antimicrobial Stewardship: Stopping and Adjusting Treatment
• What This Means for Patients
• Limitations: What This Guideline Could Not Settle
• Actionable Recommendations at a Glance
• Frequently Asked Questions
• Source Information

Key Points

• Neutropenic fever is a medical emergency: empirical antibacterial therapy must start within 2 hours of fever onset, without waiting for microbiology results.
• Patients are grouped by expected neutropenia duration: standard risk up to 7 days, high risk 8 days or more, with MASCC criteria identifying standard-risk patients.
• Fever can be masked by prednisone, non-steroidal anti-inflammatory agents, and metamizole, so a lack of fever does not rule out serious infection.
• Home management is possible for some standard-risk patients meeting strict criteria, including not living alone and reaching a specialized clinic within 1 hour.
• If fever persists beyond 7 days of neutropenia without response to first-line antibacterials and no mold-active prophylaxis, therapy should also target fungi, especially Aspergillus.

Background: Why This Guideline Matters

Cancer patients whose white blood cell counts drop — a condition called neutropenia (dangerously low levels of the infection-fighting white blood cells called neutrophils) — face a high risk of serious infection. That risk depends on how low the count falls and how long it stays low. The risk grows further with additional cellular and/or humoral immunosuppression (weakened defenses from immune cells and from antibodies). The risk also grows with breaks in the skin and mucosal barriers that normally keep germs out.

The lowest point of the neutrophil count (the nadir) and the duration of neutropenia both directly track with how often fever and infection occur. These infections cause significant morbidity and mortality (illness and death), and they can also delay or compromise further chemotherapy — a real threat to curing the underlying cancer.

Two trends make this problem harder today than in the past. First, resistance among bacteria to widely used antibacterials — particularly beta-lactams and fluoroquinolones — has been rising steadily over the past decade. Second, only a small number of newer antimicrobial agents have become available, including tigecycline, linezolid, moxifloxacin, cefozopran, telavancin, oritavancin, ceftazidime-avibactam, ceftolozane-tazobactam, and micafungin. Data on using these drugs empirically (before the exact germ is known) in neutropenic patients are limited or simply do not exist.

This guideline is an update of a version published in 2003. This guideline focuses on risk-adapted diagnostic procedures for neutropenic cancer patients with fever of unknown origin (FUO). It also focuses on empirical antimicrobial treatment for these patients, matched to how likely each patient is to have a complicated course. This guideline complements other AGIHO guidelines covering sepsis, primary prevention of bacterial and fungal infections, and infection prevention after allogeneic hematopoietic stem cell transplantation (HSCT — transplant using donor stem cells). It also complements AGIHO guidelines on invasive fungal infections, pulmonary infiltrates, abdominal complications, and venous catheter-related infections. It further complements AGIHO guidelines on central nervous system infections, infections after autologous HSCT (using the patient's own stem cells), and community respiratory viral infections.

How These Guidelines Were Developed

A panel of hematologists, oncologists, and infectious disease specialists was assembled within AGIHO. The group performed a thorough literature search, and importantly, it included only full publications while excluding recipients of allogeneic hematopoietic stem cell transplantation. This means transplant patients receiving donor cells are not covered by these specific recommendations.

The panel created a set of core slides with statements and recommendations, then discussed them in face-to-face meetings, telephone conferences, and by electronic correspondence. The final version was approved at an AGIHO plenary meeting on 10 February 2017 and was reviewed by all co-authors. A detailed methodological report appears in the journal's electronic supplementary material.

One practical limitation is worth knowing up front. Randomized clinical trials — the strongest form of medical evidence — were mainly available for therapeutic decisions (how to treat), not for diagnostic decisions (how to test). New diagnostic procedures have entered clinical practice over the past decade without always being tested in rigorous head-to-head trials.

Understanding the Grading System (What the Letters and Numbers Mean)

This guideline adapted the grading system used by the European Society for Clinical Microbiology and Infectious Diseases (ESCMID). Two separate scales are used: one for how strongly the panel supports a recommendation, and one for how good the underlying evidence is. Knowing these codes helps you read exactly how confident the experts were.

Strength of recommendation:

• Grade A — Strongly supports a recommendation for use
• Grade B — Moderately supports a recommendation for use
• Grade C — Marginally supports a recommendation for use
• Grade D — Supports a recommendation against use

Quality of evidence:

• Level I — Evidence from at least 1 properly designed randomized, controlled trial
• Level IIa — Evidence from at least 1 well-designed clinical trial without randomization; from cohort or case-controlled analytic studies (preferably from at least 1 center); from multiple time series; or from dramatic results of uncontrolled experiences
• Level III — Evidence from opinions of respected authorities, based on clinical experience, descriptive case studies, or reports of expert committees

An added index letter refines the evidence type further. "r" stands for meta-analysis or systematic review of randomized controlled trials. "t" stands for transferred evidence, meaning results from different patient cohorts or similar immune-status situations. "h" means the comparator group is a historical control. "u" stands for uncontrolled trials.

Key Definitions: Neutropenia and Fever

There is no defined cut-off neutrophil count that cleanly separates patients with from those without an increased risk of infection and death. In line with most recommendations and with risk stratifications used in clinical trials, the panel defines neutropenia as a neutrophil count (segments and bands together) of less than 500 cells per microliter (/μl), or less than 1000/μl with a predicted decline to below 500/μl within the next 2 days.

Fever definitions vary across guidelines and trials, and body temperature can be measured in several ways. Generally, fever may be defined as an oral temperature of ≥38.3 °C measured once, or ≥38.0 °C lasting at least 1 hour, or a temperature measured twice within 12 hours using a method shown to be equivalent to these results.

There may be no definite non-infectious cause, such as a febrile reaction to cytokines, to cytotoxic drugs like cytarabine or bleomycin, or to a transfusion of blood products. If there is no definite non-infectious cause, then this fever must be regarded as a sign of an infectious complication. Critically, fever can be masked by antipyretic (fever-lowering) drugs used for pain relief or cancer treatment, including prednisone, non-steroidal anti-inflammatory agents, and metamizole (also called dipyrone). This means a patient on these medications might not mount the expected fever despite a serious infection.

Risk Stratification: Standard Risk vs. High Risk

It is widely accepted that infection rates in cancer patients track directly with the nadir and duration of neutropenia, but predicting this precisely in one individual is difficult. Clinical trials on treating fever in patients with short neutropenia — below 5 or 7 days — are limited. Some patients whose neutropenia lasts more than 5 days have still been safely enrolled in studies of oral therapy and outpatient care.

After reviewing the literature, the panel agreed on two risk groups:

• Standard risk: expected duration of neutropenia of up to 7 days
• High risk: expected duration of neutropenia of at least 8 days

All patients whose neutropenia lasts 8 or more days are considered high risk for a complicated course of a fever episode. However, some patients assigned to the standard-risk group show individual features that justify moving them into the high-risk group instead. The panel recommends using the Multinational Association of Supportive Care in Cancer (MASCC) criteria to identify these individuals, and these criteria have been repeatedly validated.

The MASCC score assigns points as follows:

• Burden of febrile neutropenia with no or mild symptoms — 5 points
• No hypotension (systolic blood pressure above 90 mmHg) — 5 points
• No chronic obstructive pulmonary disease — 4 points
• Solid tumor or hematologic malignancy with no previous fungal infection — 4 points
• No dehydration requiring intravenous (parenteral) fluids — 3 points
• Burden of febrile neutropenia with moderate symptoms — 3 points
• Outpatient status — 3 points
• Age under 60 years — 2 points

A score of ≥21 identifies a standard-risk patient. Note that the points attributed to the "burden of febrile neutropenia" variable are not cumulative. Mild and moderate symptom scores are not added together. The maximum theoretical score is therefore 26. Standard-risk patients with a MASCC score of ≥21 form a group with a high likelihood of an uncomplicated course of infection.

Criteria for Treating a Patient at Home (Outpatient Management)

Provided they meet all individual criteria, standard-risk patients can be managed primarily as outpatients (Grade B, Level IIr). The criteria fall into four categories.

General criteria:

• No signs of central nervous system infection, severe pneumonia, or venous catheter infection
• No signs of sepsis or shock
• None of the following: associated organ failure, pronounced abdominal pain (with or without diarrhea), dehydration, recurrent vomiting, need for intravenous supportive therapy. None of the following: need for permanent or close monitoring (for example, metabolic decompensation or high calcium levels).
• No new electrocardiogram abnormalities requiring treatment
• No new severe organ impairment

Oral antibiotic criteria:

• No fluoroquinolone prophylaxis or therapy within the last 7 days
• Oral medication is feasible
• Good compliance with oral medication is expected

Outpatient management criteria:

• Medical care is ensured, with different options available
• The patient does not live alone; the patient and helpers have a telephone; the patient can reach a clinic skilled in treating neutropenic patients within 1 hour
• The patient is conscious, knows the risks, and understands them

Which Germs Cause These Infections?

At the onset of fever, antibiotic therapy must start immediately. Because microbiological tests take time, treatment has to be empirical at the beginning. This is especially true in patients who show no suspected clinical focus of infection. In about half of patients with febrile neutropenia, the antibiotic therapy remains purely empirical, because no relevant pathogen or focus of infection can be identified over the following days.

Two things guide the choice of empirical antimicrobial agents. The first is results reported from prospective, randomized clinical studies. The second is the microorganisms identified in patients with microbiologically documented infections, used by analogy. The most frequent and relevant pathogens are Staphylococcus aureus, Streptococcus species, enterococci, coagulase-negative staphylococci, gram-negative enterobacteria, and Pseudomonas aeruginosa.

One important nuance: coagulase-negative staphylococci are numerically the most frequent microbial isolates in many institutions. However, a single blood culture positive for these commensal skin bacteria should be considered contamination rather than true infection. This applies in the absence of a matching clinical focus of infection. The same applies to other potential contaminants such as Corynebacterium, Bacillus cereus, Propionibacterium, and Micrococcus species.

Among fungal pathogens, Candida species and Aspergillus species predominate. Aspergillus is typically associated with a prolonged duration of neutropenia in high-risk patients.

The guideline also links typical clinical signs to the pathogens most often involved:

• Redness and/or pain at a venous access site — coagulase-negative staphylococci
• Mucosal ulcers — alpha-hemolytic streptococci, Candida species
• Single point-like skin lesions — gram-positive cocci, Candida species
• Necrotizing skin lesions — Pseudomonas aeruginosa, filamentous fungi
• Diarrhea and abdominal bloating (meteorism) — Clostridium difficile
• Enterocolitis and perianal lesions — polymicrobial infection, including anaerobes
• Lung infiltrates with or without sinusitis — filamentous fungi, Pneumocystis jirovecii
• Retinal infiltrates — candidemia (Candida in the bloodstream)

Antibiotic Prophylaxis, Resistance, and Local Epidemiology

A recent history of antibiotic prophylaxis or therapy raises the risk of infection with bacteria resistant to the antibiotic that was used. After ciprofloxacin prophylaxis, for example, a relative predominance of infections caused by gram-positive cocci compared with gram-negative bacteria has been observed.

Quinolones have been linked to an increased rate of colonization by vancomycin-resistant enterococci (VRE) and methicillin-resistant S. aureus (MRSA), and with a higher prevalence of multidrug resistance among enterobacteria through extended-spectrum beta-lactamases (ESBL — enzymes that destroy many common antibiotics). Colonization by ESBL organisms, VRE, or MRSA has itself been associated with an increased rate of bloodstream infection with these same pathogens.

As a consequence, the panel states that the use of quinolones for interventional treatment in febrile neutropenic patients should be limited. Quinolones should be limited to microbiologically documented infections caused by microorganisms proven susceptible in the laboratory. Local epidemiology must be taken into account when choosing empirical antimicrobial therapy. Microbiological findings from patients treated in a given hematology-oncology institution should be reviewed at least once a year together with infection-control and antimicrobial stewardship experts (Grade B, Level III).

Baseline screening of newly admitted or re-admitted patients for multidrug-resistant pathogens should be considered: MRSA (Grade B, Level III), VRE (Grade B, Level III), and ESBL (Grade B, Level IIt).

Baseline Testing Before Chemotherapy Starts

Before myelosuppressive therapy (treatment that suppresses bone marrow function) begins, patients must be thoroughly evaluated. This evaluation is for any relevant previous or existing infection that could become important during treatment-induced neutropenia (Grade A, Level III). The clinical examination should pay special attention to the skin, mucosa, puncture sites, vascular catheter exit sites, paranasal sinuses, lungs, and the perianal region (Grade A, Level III).

In patients who report a penicillin allergy, skin testing is recommended (Grade B, Level IIt). This matters because a negative result is expected in the vast majority of cases. A negative result helps avoid unnecessary first-line use of carbapenems, aztreonam, or vancomycin, drugs that are broader, more toxic, or both.

Baseline laboratory tests include a blood count, liver enzymes (ASAT/SGPT, ALAT/SGOT, gGT), total bilirubin, alkaline phosphatase, LDH, creatinine, blood urea nitrogen. Baseline laboratory tests also include coagulation tests (INR, aPTT), C-reactive protein, and urinalysis (Grade B, Level III). Except for urinalysis, these tests should be repeated regularly — for example, twice a week — during prolonged neutropenia (Grade B, Level III).

Procalcitonin or cytokine levels such as interleukin-6 are not recommended for routine baseline diagnostics (Grade D, Level III).

Computed tomography (CT) or magnetic resonance imaging (MRI) of the chest and abdomen may not have been done for staging of the underlying disease. In that case, chest radiographs (two views) and abdominal ultrasound may be considered as baseline examinations before the first chemotherapy cycle. These examinations check for pre-existing abnormalities and make later comparisons easier (Grade C, Level III). Particularly in high-risk patients, a thoracic CT scan before chemotherapy appears desirable for documenting the baseline state; however, in the absence of prospective studies, no firm recommendation can be made. In patients with a history of an invasive infection, appropriate imaging is recommended even when there are no clinical symptoms of recurrence (Grade B, Level III).

Screening for Fungal Infections in Patients Without Symptoms

For high-risk patients with an expected duration of profound neutropenia exceeding 7 days, serial monitoring for Aspergillus galactomannan (a fungal cell wall component detectable in blood) in serum at least twice weekly has been recommended. Monitoring with 1,3-beta-D-glucan (another fungal marker) in blood samples is under discussion as an alternative but is rarely used because of its higher cost.

A sensitive, validated Aspergillus PCR (polymerase chain reaction — a test that detects fungal DNA) may also be helpful (Grade C, Level III) for screening blood samples in specific high-risk populations. However, the sensitivity of these tests is strongly reduced in patients already receiving systemic mold-active antifungals. False-positive results can be caused by beta-lactam antibiotics, parenteral nutrition, severe intestinal mucositis, or transfusion of blood products.

Because of these limitations, screening of patients who have no fever and no symptoms should be restricted to those not receiving systemic mold-active prophylaxis (Grade B, Level IIu). Importantly, these non-culture-based procedures do not replace clinical examination, imaging, endoscopy, or other microbiological diagnostics (Grade B, Level III).

What to Do at the Onset of Fever

Diagnostic measures at first fever in a neutropenic patient serve three purposes. The first is ruling out non-infectious causes of fever. The second is identifying a clinical focus and/or the causative pathogens. The third is assessing the severity of the inflammatory response so that patients needing intensive care can be recognized early.

A minimum of two separate pairs of blood cultures must be taken before antibiotic therapy starts (Grade A, Level III). There is no need to wait between sampling the cultures; separate sets can be obtained by drawing blood from both arms. If the patient has an indwelling central venous catheter (CVC), one pair should be drawn from a peripheral vein and at least one from the CVC. The diagnostic yield can be increased by taking a blood sample from each lumen of the CVC. The diagnostic yield can also be increased by taking three pairs of blood cultures, totaling 60 ml of blood (Grade B, Level IIt).

A "differential time to positivity" of 2 hours or more between the CVC and peripheral blood cultures may indicate a CVC-related infection. This gives reason for pre-emptive treatment described in a separate guideline (Grade B, Level IIu). Multiplex PCR-based methods do not replace standard microbiology (Grade C, Level IIu), although they may improve turnaround time, sensitivity, and specificity of pathogen detection. A reduction in illness or death among febrile neutropenic patients through PCR-based methods added to blood cultures has not yet been demonstrated.

Beyond repeat baseline laboratory tests, measurement of lactate, blood gas analysis, and coagulation tests should be considered to identify severe sepsis early (Grade B, Level III). Biomarkers such as procalcitonin or interleukin-6 are widely used to judge how severe inflammation is, but data on their prognostic or predictive value in adult febrile neutropenia are conflicting.

In high-risk patients who did receive systemic mold-active antifungal prophylaxis and were not screened for Aspergillus galactomannan, beta-D-glucan, or fungal PCR, such a test — preferably galactomannan — should be ordered at the time of fever to allow early detection of a breakthrough invasive fungal disease (Grade B, Level III).

Imaging plays a key role. At the onset of fever, a CT scan of the lungs is recommended if there are respiratory tract symptoms (Grade B, Level III). Conventional chest radiographs are actively discouraged (Grade D, Level IIt). Conventional chest radiographs show abnormalities in fewer than 2% of febrile neutropenic patients who have no clinical signs of lower respiratory tract infection. Nasal congestion or signs and symptoms of sinusitis should prompt a CT scan of the paranasal sinuses (Grade B, Level III).

Early data on PET-CT suggest a potential use for identifying the source of fever or infection, particularly abdominal foci. Despite these positive reports, an explicit recommendation for routine use cannot be given because systematic studies are lacking. Gastrointestinal complaints or laboratory abnormalities should prompt abdominal ultrasonography (Grade B, Level IIu). An abdominal CT scan is an alternative if neutropenic enterocolitis (inflammation of the bowel in the setting of low white counts) is suspected (Grade B, Level IIu).

Thorough clinical examination must be updated (Grade A, Level III) and repeated at least daily as long as a hospitalized patient is febrile (Grade A, Level III). This may reveal a presumed focus of infection and enable pre-emptive antimicrobial treatment targeting the typically involved pathogens, rather than purely empirical treatment.

Empirical Antibacterial Therapy: First-Line Choices

In high-risk patients, the spectrum of first-line antibacterial agents should cover gram-negative enterobacteria, P. aeruginosa, S. aureus, and streptococci (Grade A, Level I), while local epidemiology must always be taken into account. Treatment must be started within 2 hours after the onset of fever, without waiting for microbiology results (Grade A, Level IIt).

If oral fluoroquinolone prophylaxis has been given, it should be discontinued at the start of interventional antimicrobial therapy (Grade A, Level III). In a clinically unstable patient — who may present in the emergency room — prompt start of antimicrobial therapy is required (Grade A, Level I). Immediate referral to an intensive care unit must be considered.

The guideline also notes that emergency treatment algorithms for this clinical situation, along with supervision or audits, are part of good practice. A clinical treatment algorithm for high-risk patients is provided in the original publication (Fig. 1).

Newer Antibacterial Agents and Allergy Alternatives

For patients with a documented penicillin hypersensitivity reaction, aztreonam may be used as a less well-studied alternative (Grade C, Level IIu). In this setting, adding vancomycin or teicoplanin to aztreonam may be considered, because aztreonam has no activity against gram-positive bacteria (Grade C, Level III).

For newer broad-spectrum antibacterial agents, the evidence in febrile neutropenia is thin. These include ertapenem, which has insufficient activity against Pseudomonas species; doripenem; ceftazidime-avibactam; ceftolozane-tazobactam; and cefozopran. There are very limited data on their safety and efficacy for empirical treatment in adult febrile neutropenic cancer patients.

Tigecycline, given in combination with an antipseudomonal beta-lactam, has shown benefit in terms of a lesser need for treatment modification. This benefit is in institutions with excessive rates of multidrug-resistant pathogens, and for second- or third-line treatments.

Combination Antibacterial Therapy

In high-risk patients, there is no evidence that combining antibacterial agents is more effective than monotherapy (Grade A, Level IIr). Combining antibacterial agents in this situation has also not been shown to prevent the development of resistance.

A combination might be useful in institutions with a high prevalence of multidrug-resistant bacteria (Grade A, Level IIr). When a combination is used, an antipseudomonal beta-lactam should always be included. The partner drug should be an aminoglycoside or a fluoroquinolone such as levofloxacin or ciprofloxacin (Grade A, Level IIt). For standard-risk patients without critically impaired kidney function, combining an aminoglycoside with a third- or fourth-generation cephalosporin can be considered (Grade A, Level I).

When aminoglycoside antibiotics are given, therapeutic drug monitoring (measuring drug levels in the blood to keep them safe and effective) is mandatory (Grade A, Level IIu). Once-daily dosing is appropriate (Grade A, Level IIr).

Combination therapy that includes vancomycin or teicoplanin (Grade D, Level IIr) or linezolid (Grade D, Level III) is generally discouraged for empirical first-line therapy. It might be considered (Grade C, Level III) in cases of severe mucositis (painful mouth and gut ulcers), skin or soft tissue infection, foreign body infection. It might also be considered in cases of documented colonization of a patient with MRSA. Note that vancomycin use carries an increased risk of kidney toxicity, which should be minimized through therapeutic drug monitoring.

When to Add Antifungal Therapy

In patients whose expected duration of neutropenia is more than 7 days and who do not respond to first-line antibacterial treatment, further therapy should also be directed against fungi — in particular Aspergillus species. This recommendation specifically applies when there is no mold-active antifungal prophylaxis already in place. In other words, persistent fever despite appropriate antibiotics in a deeply and durably neutropenic patient is a signal to think about molds, not just bacteria.

Antimicrobial Stewardship: Stopping and Adjusting Treatment

Antimicrobial stewardship is the practice of using antibiotics and antifungals as effectively and responsibly as possible. With regard to stewardship, two points stand out in this guideline.

1. Treatment duration after defervescence (when the fever breaks) in persistently neutropenic patients must be critically reconsidered.
2. The choice of anti-infective agents should be adjusted to local epidemiology — that is, to the resistance patterns and organisms actually seen in that hospital.

This guideline updates the recommendations for diagnosis and empirical therapy of fever of unknown origin in adult neutropenic cancer patients in light of these stewardship challenges.

What This Means for Patients

For patients and families, several practical messages emerge from this guideline. If you are receiving chemotherapy that lowers your white blood cell count and you develop a fever at home, treat it as an emergency and contact your cancer team immediately. Antibiotics should be started within 2 hours once fever is recognized.

Do not assume that a lack of fever means a lack of infection. This is especially important if you take steroids such as prednisone, non-steroidal anti-inflammatory drugs, or metamizole (dipyrone) for pain or cancer treatment. These can hide a fever. Report any new symptoms, including mouth ulcers, skin changes, diarrhea, abdominal pain, or nasal congestion, because these point toward particular types of infection.

If your cancer team offers outpatient (home-based) management, this is only appropriate if you meet strict criteria. These criteria include not living alone, having a telephone, being able to reach a specialized clinic within 1 hour, and being able to take oral medication reliably. It also matters whether you have taken a fluoroquinolone antibiotic in the previous 7 days.

You may also be offered screening tests before chemotherapy, such as skin testing if you report a penicillin allergy, blood tests, or imaging. If you are at high risk, your team may monitor you twice weekly for fungal markers such as Aspergillus galactomannan. If you have had an invasive infection in the past, imaging may be recommended even when you feel well.

Limitations: What This Guideline Could Not Settle

Honest medicine includes acknowledging uncertainty, and this guideline does so in several places.

• Randomized clinical trials were mainly available for therapeutic decisions, not for diagnostic decisions, so many diagnostic recommendations rest on weaker evidence (Level II or III) or expert opinion.
• Recipients of allogeneic hematopoietic stem cell transplantation were excluded from the literature review, so these recommendations do not directly apply to that group.
• No recommendation could be made for routine thoracic CT before chemotherapy in high-risk patients, because prospective studies are lacking, even though such imaging appears desirable.
• No explicit recommendation can be given for routine PET-CT use to find the source of fever, despite encouraging early reports, because systematic studies are missing.
• Data on newer broad-spectrum antibacterial agents — ertapenem, doripenem, ceftazidime-avibactam, ceftolozane-tazobactam, and cefozopran — are very limited regarding safety and efficacy in empirical treatment of febrile neutropenia.
• A reduction in illness or death from using PCR-based methods alongside blood cultures has not yet been shown.
• Data on prognostically useful biomarkers such as procalcitonin and interleukin-6 in adult febrile neutropenia are conflicting.

Actionable Recommendations at a Glance

1. Before myelosuppressive therapy, thoroughly evaluate patients for previous or existing infection, with careful examination of skin, mucosa, puncture sites, catheter exit sites, sinuses, lungs, and the perianal region (AIII).
2. Offer penicillin allergy skin testing to patients reporting penicillin allergy (BIIt).
3. Obtain baseline blood count, liver enzymes, bilirubin, alkaline phosphatase, LDH, creatinine, blood urea nitrogen, coagulation tests, CRP, and urinalysis — and repeat most of these twice weekly during prolonged neutropenia (BIII).
4. In high-risk patients with expected profound neutropenia beyond 7 days, monitor serum Aspergillus galactomannan at least twice weekly (restricting this to patients not on systemic mold-active prophylaxis) (BIIu).
5. Take at least two separate pairs of blood cultures before starting antibiotics, one pair peripherally and at least one from the central venous catheter; three pairs (60 ml) increase yield (BIIt).
6. Start empirical antibacterial therapy within 2 hours of fever onset, without waiting for microbiology results (AIIt), and discontinue any oral fluoroquinolone prophylaxis (AIII).
7. Cover gram-negative enterobacteria, P. aeruginosa, S. aureus, and streptococci with first-line therapy in high-risk patients (AI).
8. Use monotherapy in high-risk patients unless local multidrug resistance rates are high; perform therapeutic drug monitoring for aminoglycosides and use once-daily dosing (AIIu, AIIr).
9. Avoid routine empirical vancomycin, teicoplanin, or linezolid as first-line therapy (DIIr, DIII), reserving them for severe mucositis, skin or soft tissue infection, foreign body infection, or MRSA colonization (CIII).
10. If fever persists beyond 7 days of neutropenia without response to first-line antibacterials — and no mold-active prophylaxis has been given — direct therapy against fungi, especially Aspergillus.
11. Critically reconsider how long treatment continues after the fever resolves in persistently neutropenic patients, and adjust drug choices to local epidemiology.

Frequently Asked Questions

What exactly counts as neutropenia and fever?

Neutropenia means a neutrophil count below 500 cells per microliter, or below 1000/μl if it is predicted to fall under 500/μl within two days. Fever may be an oral temperature of 38.3 °C measured once, 38.0 °C lasting at least one hour, or a temperature measured twice within 12 hours using an equivalent method.

Why must antibiotics start so quickly when I have a fever?

In neutropenic cancer patients, fever is treated as a medical emergency because infection can be serious. Empirical antibacterial therapy must begin within 2 hours after fever starts, without waiting for microbiology results. If you develop a fever at home, contact your cancer team immediately rather than waiting to see whether it settles.

Can I be treated at home instead of in hospital?

Some standard-risk patients can be managed mainly as outpatients if they meet strict criteria. These include no sepsis or shock, no central nervous system, severe pneumonia or catheter infection, no dehydration or recurrent vomiting, no fluoroquinolone in the past 7 days. These also include reliable oral medication, not living alone, a telephone, and reaching a clinic skilled in neutropenic care within 1 hour.

What does a MASCC score of 21 or more mean for me?

The MASCC score helps identify standard-risk patients likely to have an uncomplicated course of infection. Points are given for mild symptoms, no low blood pressure, no chronic obstructive pulmonary disease, no previous fungal infection, no dehydration needing intravenous fluids, outpatient status, and age under 60. A score of 21 or above identifies a standard-risk patient.

I take steroids and painkillers — could they hide a fever?

Yes. Fever can be masked by antipyretic drugs used for pain relief or cancer treatment, including prednisone, non-steroidal anti-inflammatory agents, and metamizole, also called dipyrone. This means you might not develop the expected fever even during a serious infection. Do not assume that a lack of fever means a lack of infection, and report any new symptoms to your team.

When might antifungal treatment be added?

If your expected neutropenia lasts more than 7 days and fever does not respond to first-line antibacterial treatment, therapy should also be directed against fungi, particularly Aspergillus species. This applies specifically when you are not already receiving mold-active antifungal prophylaxis. Persistent fever despite appropriate antibiotics in deep, lasting neutropenia is a signal to consider molds.

What tests happen when I first develop a fever?

At least two separate pairs of blood cultures must be taken before antibiotics start. One pair must be from a peripheral vein and at least one from a central venous catheter if you have one. Three pairs, about 60 ml, increase yield. Lactate, blood gas analysis and coagulation tests may be considered to detect severe sepsis early, and imaging may be arranged depending on your symptoms.

If I have neutropenic fever during chemotherapy, when should I get a second opinion on my treatment plan?

A second opinion can be useful when the risk group or treatment plan is unclear. Fever with a neutrophil count below 500/μl, or below 1000/μl with a predicted decline within two days, is an emergency: antibacterial therapy must start within 2 hours. Patients expected to have neutropenia for 8 or more days are high risk, and those not improving may also need antifungal therapy aimed at Aspergillus. A second opinion can review whether risk stratification, blood cultures, imaging, and drug choices fit the recommendations. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

Original article title: Diagnosis and empirical treatment of fever of unknown origin (FUO) in adult neutropenic patients

Authors: W. J. Heinz, D. Buchheidt, M. Christopeit, M. von Lilienfeld-Toal, O. A. Cornely, H. Einsele, M. Karthaus, H. Link, R. Mahlberg, S. Neumann, H. Ostermann, O. Penack, M. Ruhnke, M. Sandherr, X. Schiel, J. J. Vehreschild, F. Weissinger, and G. Maschmeyer

Publication details: Annals of Hematology (2017) 96:1775–1792. DOI 10.1007/s00277-017-3098-3. Received 3 May 2017; accepted 6 August 2017; published online 30 August 2017. This is an open access publication. Correspondence: G. Maschmeyer, Department of Hematology, Oncology and Palliative Care, Klinikum Ernst von Bergmann, Potsdam, Germany. Electronic supplementary material is available with the original article.

This patient-friendly article is based on peer-reviewed research.