Febrile neutropenia: definition, incidence, mortality

Febrile neutropenia (FN) is a common and potentially life-threatening complication following chemotherapy. FN is defined as an oral temperature of > 38.3 °C or two consecutive readings of > 38.0 °C within 2 h and an absolute neutrophil count (ANC) of < 0.5 × 109/l, or expected to fall below 0.5 × 109/l. The incidence of FN is estimated at about 8/1000 patients receiving chemotherapy and it is associated with frequent complications, an in-hospital mortality rate of up to 10%, and a high financial burden [1, 2]. Therefore, understanding FN is of crucial importance to clinicians and patients.

Evidence-based guidelines on the prevention and treatment of FN have been established by the European Society for Medical Oncology (ESMO) [1], American Society of Clinical Oncology (ASCO) [3, 4], and National Comprehensive Cancer Network (NCCN) [5], which provide detailed recommendations and useful algorithms for clinical decision-making.

Initial workup and risk assessment in patients with FN

Initial workup of patients with suspected FN involves thorough history and physical examination as well as complete blood count (CBC) and biochemistry analysis to assess kidney and liver function, coagulation status, and level of inflammatory markers, such as C‑reactive protein (CRP). Peripheral and central blood cultures along with specimens from suspected sites of infection (e.g., urine or stool culture, sputum) should be taken before beginning empiric broad-spectrum antimicrobial therapy. Regarding imaging, chest radiographs are easily obtainable and can provide relevant information related to FN and/or present comorbidities as part of initial workup. Across guidelines, there is a universal consensus that antibiotics should be initiated within 1 h of admission.

Individual outcome risk of patients presenting with FN can be assessed using scoring systems such as the Multinational Association for Supportive Care in Cancer (MASCC) index [6] or the Clinical Index of Stable Febrile Neutropenia (CISNE) [7], shown in Tables 1 and 2, respectively. Both systems have shown themselves to be effective screening tools to help identify low-risk patients (who may be candidates for outpatient treatment) based on readily available clinical information [8]. Importantly, recent literature suggests superiority of CISNE, a three-tier scoring system, in the acute setting [9, 10].

Table 1 Multinational Association for Supportive Care in Cancer (MASCC) score: adapted from Klastersky et al. [6]
Full size table
Table 2 Clinical Index of Stable Febrile Neutropenia (CISNE) score: adapted from Carmona-Bayonas et al. [7]
Full size table

Etiology and optimal antimicrobial treatment in FN

In most patients with FN, an infectious focus or agent cannot be identified, with bacteremia present in only up to 30% of cases. Confirmed bacteremia, although providing useful information regarding choice of antibiotic treatment, is associated with higher mortality rates, especially in case of Gram-negative bacteremia [1, 11].

Common sources of bacteremia in FN are translocation of enteric bacteria into the bloodstream and catheter-related bloodstream infections (CRBSI) [12]; differential time to positivity (DTP) of ≥ 2 h between central and peripheral blood cultures is a sensitive and specific indicator of CRBSI and can therefore inform decisions regarding removal of indwelling catheters [1, 13]. However, FN is a diverse syndrome, and further diagnostic workup, such as imaging studies (e.g., computed tomography) to locate focus of infection or testing for fungal or viral etiologies is often necessary during the course of hospitalization in the event of persisting fever and/or worsening symptoms.

Choice of initial empiric antimicrobial treatment should be made based on patient-related factors such as outcome risk and allergies, as well as local epidemiological and antibiotic resistance data; in-hospital patients should receive broad-spectrum antibiotics with anti-pseudomonal activity, e.g., piperacillin–tazobactam [14]. While guidelines provide straight-forward recommendations regarding de-escalating and discontinuing antibiotic treatment upon recovery from FN, decisions about escalating or switching antimicrobial agents can prove to be challenging and should be discussed with infectious disease specialists, clinical microbiologists, or other specialists [1]. Implementation of antimicrobial stewardship programs has been shown to both improve clinical outcomes and lower use of carbapenems and glycopeptides [15].

Antibacterial prophylaxis and the role of G-CSF in prevention and treatment of FN

Chemoprophylaxis with fluoroquinolones after chemotherapy, although widely used in the past, is associated with the risk of antibiotic resistance [1]. The ASCO panel suggests chemoprophylaxis be used in situations where prolonged severe neutropenia (ANC ≤ 0.1 × 109/l for ≥ 7 days) can be expected. As this is unlikely to occur with current chemotherapy regimens for solid tumors, the panel recommends against the routine use of chemoprophylaxis in this setting, suggesting it should only be considered in patients with increased risk of death in the event of FN [3, 4].

Use of granulocyte colony-stimulating factor (G-CSF) for prophylaxis of FN has been shown to have a significant impact on reducing incidence, duration, and severity of FN [1, 5]. Current guidelines provide algorithms on primary FN prophylaxis in chemotherapy patients. In general, risk factors for development of FN can be defined as chemotherapy regimen-related, on the one hand, and patient-related, on the other hand.

There are clear recommendations to give or withhold G‑CSF in chemotherapy regimens associated with high (> 20%) and low (< 10%) FN risk, respectively [1]. However, regimens with intermediate (10–20%) FN risk require case-to-case consideration of patient-related risk factors such as age, comorbidities, and performance status [1, 5, 16]. This leads to variable real-world use of G‑CSF within this heterogeneous group, with an overall trend of reduced FN incidence in patients who receive G‑CSF prophylaxis [17].

Therapeutic use of G‑CSF in manifest FN is a controversial topic due to scarce evidence. A large meta-analysis of 14 randomized clinical trials showed shorter time to neutrophil recovery and shorter duration of hospital stay with use of G‑CSF in patients with FN; however, overall mortality was not affected [18]. The NCCN panel distinguishes between patients with FN who have already received prophylactic G‑CSF versus those who have not, and recommends considering short-acting G‑CSF in patients at high risk of infection-related complications or death [5]. Overall, more data on the use of G‑CSF in the setting of FN are required before specific recommendations can be defined.

Role of genetic factors in predicting and reducing risk of FN

Aside from all clinical patient-related factors discussed above, predictive biomarkers for chemotherapy-induced neutropenia have been identified. For example, dihydropyrimidine dehydrogenase (DPYD) gene variants causing decreased metabolism of fluoropyrimidines (e.g., 5-fluoruracyl, capecitabine) are associated with higher toxicity of these agents, including shorter time to neutropenia [19]. Consequently, testing for these gene variants and dose-reduction when appropriate has become widespread clinical practice [20]. Further genetic variants which appear to have predictive value for neutropenia, such as single nucleotide polymorphisms (SNPs) in the hyaluronan-mediated motility receptor (HMMR) gene, have been investigated [21]. Identifying new potentially relevant genetic factors could help to predict and reduce FN risk in chemotherapy patients [16].

Discussion

Adequate management of FN, a classic oncologic emergency, requires sufficient clinical experience and in-depth knowledge of current guidelines. However, implementing guidelines in everyday practice may prove challenging. Initial workup must often be carried out under stressful conditions and time constraints; the universal consensus of beginning antibiotic therapy within 1 h of admission is often not feasible in the setting of emergency care [22].

As discussed above, individual risk assessment is of crucial importance. Unfortunately, established outcome risk scores such as MASCC, although possessing high predictive value, are rarely used in the real-world setting [23]. Their integration into everyday practice could be of significant benefit in clinical decision-making.

Another challenging aspect in FN is the need to re-evaluate patients on a regular basis according to clinical presentation and diagnostic findings; cancer patients are often frail, of advanced age, and have multiple comorbidities. Infections can drastically worsen pre-existing conditions, leading to complex clinical situations. Therefore, treatment of FN patients (especially high-risk individuals) sometimes requires a multidisciplinary approach.

Regarding early detection of FN, patient education likely plays the most important role in reducing time to administration of first antibiotic dose. Real-world data suggest incidence of FN is highest after the first cycle of chemotherapy [24, 25], emphasizing the importance of providing patients with up-front and clear instructions on monitoring symptoms and body temperature, and contacting the appropriate service when needed [1].

Due to factors such as inaccurate measurement of body temperature (e.g., axillary or rectal instead of oral temperature) and use of certain medications, such as corticosteroids or nonsteroidal anti-inflammatory drugs (NSAIDs), which may “blunt” the natural fever response to infection, patients should be encouraged to contact medical care providers even if their temperature does not exactly match the definition criteria for FN.

Current challenges in the field of FN are summarized in Fig. 1. Future improvements in FN prevention lie in better understanding individual patient risk with regard to G‑CSF use and identifying underlying predictive biomarkers.

Fig. 1
figure 1

Challenges in management of febrile neutropenia (FN). G‑CSF granulocyte colony-stimulating factor

Full size image