The usefulness of infection biomarkers in patients with febrile neutropenia in the Emergency Department

Dear Editor,

We have read with great interest the recent study published in your journal by Aguado et al.¹ in relation to the management of infection and febrile neutropenia in patients with solid cancer. We congratulate the authors on their great work and capacity for synthesis, which is sure to facilitate the management of these patients at all levels of care and, of course, in emergency departments (ED). This is the area where we work and it is where a very significant number of the febrile neutropenia episodes are initially treated. In fact, in the last 10 years there has been an increase in the incidence of infections in EDs, and very significantly (p < 0.001) in patients with neoplasia (from 3.6% to 9.3%) and neutropenia (from 1.3% to 4.6%), as well as the clinical severity² and mortality rates in these patient subgroups.³ For these objective reasons and in order to adapt to and overcome situations in real clinical practice (“day-to-day”) that often saturate or collapse our EDs,⁴ tools of objective assistance have been sought, such as automatic alarm systems that prioritize the triage or first evaluation of the patient, to suspect and detect situations of severe sepsis⁵ and, thus, perform in an early manner both the clinical assessment of the patient and obtain the lab work that includes lactate and procalcitonin,⁶ especially in the patients most susceptible to infection and who have a more severe prognosis and evolution, such as, undoubtedly, patients with febrile neutropenia. In this regard, it has recently been published that both, and in combination, are helpful in predicting the severity and/or possibility of bacteraemia in these patients.⁷ In fact, it has been published and strongly recommended that adult patients with infection in which lactate reaches concentrations ≥2–2.5 mmol/L and/or procalcitonin is ≥1 ng/ml should not be discharged (they should at least be kept under observation). And if the initial values of lactate are ≥4 and/or of procalcitonin ≥10 ng/ml they should be admitted to Intensive Care due to being considered serious patients (although they do not have hypotension) in need of resuscitation and intensive haemodynamic support.^6–8 Therefore, it is striking that the joint evaluation of lactate and procalcitonin in the initial algorithm for treatment of patients with neutropenia in the ED is not included in the work of Aguado et al.,¹ since it is a common recommendation for all patients with severe infection^6,8 and, therefore, especially indicated in the elderly, neonates and neutropenic patients, where the clinical manifestations may be more variable and subtle, resulting in a diagnostic-treatment delay, which may be fatal to the latter.

In recent years, various reviews^6,9 have confirmed the usefulness of these biomarkers for the detection of bacterial infection and severity in patients with febrile neutropenia, as well as their value in monitoring clinical evolution and treatment effectiveness. These are assertions that, specifically, Aguado et al.¹ quite rightly point out in their article and with which we agree. But these useful evaluations of initial stratification and subsequent monitoring cannot be carried out without the initial evaluation in the ED and the inclusion of these biomarkers in the algorithm or initial treatment recommendations in the ED.

Finally, we would like to make another observation. Although it is well known that C-reactive protein has, for patients with cancer and neutropenia, a false positive rate in >50% of cases (being higher without infection) and a significantly lower sensitivity and specificity than that offered by procalcitonin (for example, yielding areas under the ROC curve of 0.94, sensitivity of 91% and specificity of 89% to predict the presence of bacteraemia in neutropenic patients),^6,7,9 it has been proven that in the initial care of the patient with fever in the ED (and especially in immunosuppressed patients and patients with neutropenia) C-reactive protein is requested for the initial evaluation in more than 90% of cases, although later it is not useful, but lactate and procalcitonin are determined in only 10–20%,¹⁰ which are really recommended. Therefore, we believe that their inclusion in the evaluation protocols and algorithms in EDs would promote their effective and efficient use. But, of course, on another important note, lactate and procalcitonin should only be requested if they will be useful in decision-making and patient assessment, and can never substitute clinical examination, microbiological testing or the clinical judgement of the doctor.

Funding

No grant or assistance was received.

Conflicts of interest

The authors declare that they have no conflicts of interest.