Sepsis is still nowadays the main cause of death in burn patients due to the impact of extensive burns in all organ systems, affecting homeostatic mechanisms, and to the greater susceptibility of this population to infection [1, 2], related to the loss of the cutaneous barrier, immunosuppression, use of invasive devices, nosocomial flora, etc. Survival is directly dependent on the institution of prompt and adequate antimicrobial therapy [3]. However, the gold standard for sepsis diagnosis still relies on the identification of microorganisms in blood cultures, which unfortunately are positive only in 20–30% of all confirmed bloodstream infections, and their results may take 48 to 72 h to reach the prescriber [4]. While more rapid methods of microbiological identification, such as polymerase-chain reaction (PCR) [5], matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS), gene expression profiling, aptamer panels, etc. [6], are not either widely available or fully developed, the use of early empirical often broad-spectrum antibiotic therapy is warranted. This empirical strategy increases the likelihood of cure of infection and survival but negatively impacts in terms of microbiome, leading to the selection and emergence of antimicrobial resistance. In this context, biochemical biomarkers, namely procalcitonin (PCT) alone [7, 8] or integrating a composite panel [9–12], and always coupled with thorough clinical examination, may be an important aid for the early suspicion of sepsis and rapid institution of therapy, which is strongly associated with improved outcomes [13, 14].

PCT is a 116-amino acid precursor of calcitonin, which synthesis and secretion, encoded by first calcitonin gene (CALC-I gene), and normally restricted to thyroid C cells and some neuroendrocrine cells of the lungs and gut, is upregulated by the presence in the blood of microbial toxins, necrotic body cells, and some proinflammatory cytokines (IL-1, IL-6, TNF-α, etc.), in a synergistic way, starting to be produced in great amounts by many other nonendocrine types of cells, including monocytes and adipocytes [15], reaching measurable levels in 2–4 h after onset of the infectious process, peaking at 24–30 h, and rapidly subsiding with recovery. PCT increment is less pronounced with fungal infection and is absent in viral disease, allegedly due to inhibition of its secretion by some cytokines released as a response to viral infection, like interferon-γ [16].

Besides its utility to help clinicians in the diagnosis of sepsis [17] including patients admitted to burn units [18], the magnitude and duration of PCT elevation seems to correlate with injury severity and outcome, and there are several published works analyzing its potential for the prognosis and for the monitoring of antimicrobial therapy, helping decisions on early antibiotic de-escalation or rescue therapy [19–21]. Most of these are focused in lower respiratory tract infections and/or intensive care patients, while papers on septic burn patients are scarce [22].

The purpose of this work is to evaluate the feasibility of PCT use to predict the outcome and to monitor the efficacy of antimicrobial therapy in a sample of severe adult burn patients.

The sample under analysis was composed by 101 burn patients, with 15% or more of total body surface area (TBSA) burned, admitted from January 2011 to December 2014 at Coimbra Burns Unit (CBU), a department of Coimbra Hospital and University Center (CHUC), in Portugal. Being a retrospective observational study of patients from a suitably anonymized dataset, involving only recording data from the medical record, the Ethics Committee from CHUC, according to the Declaration of Helsinki and Council for International Organizations of Medical Sciences (CIOMS) International Ethics Guidelines, waived the need of informed consent.

All the patients had a diagnosis of sepsis. This diagnosis was done according to the American Burn Association (ABA) criteria [23]: a clinical suspicion of infection coupled with the presence of three or more of the following parameters: temperature > 39 or < 36.5 °C; tachycardia > 110 beats per minute; tachypnea > 25 breaths per minute or minute ventilation > 12 L/min; thrombocytopenia < 100,000/μL; hyperglycemia (untreated plasma glucose > 200 mg/dL or intravenous glucose requirement > 7 U/h over 24 h); and enteral feeding intolerance: abdominal distension or gastric residuals more than two times feeding rate or diarrhea > 2500 mL/24 h.

PCT was measured with time-resolved amplified cryptate emission (TRACE) technology (Kryptor PCT; Brahms AG; Hennigsdorf, Germany). The sample was factored by survival (68 survivors and 33 non-survivors). The maximum value of PCT in each day of the study was used for statistical analysis and when samples were not collected in some days (till a maximum of 5 days), the missing values of the interval were calculated as the median value between the PCT determinations available.

Even acknowledging all advances in critical care, extensive burns are still associated with high morbidity and mortality mainly due to septic episodes [26, 27]. In the last years, diverse studies were published showing the utility of dosing PCT levels as an aid to the diagnosis of systemic infection in burn patients [28–34], particularly when a dynamic approach is used [35]. Notwithstanding the core decision should rely on the clinical features and never on a biomarker alone [36], PCT dosing may support the suspect of ongoing and uncontrolled systemic infection when its values keep rising, or at least does not subside in consecutive analysis, indicating that something must be done to control a probable septic process before it can lead to irreversible damage. Apart its potential to improve clinicians’ diagnostic capacity, PCT has been used with success at the emergency departments [37, 38], to predict the prognosis of suspected septic patients and to stratify them according to the risk of death and the necessity of admission in intensive care units (ICU) [39–41]. PCT levels at admission and, much more reliable [42], its evolution on subsequent days may give insights on the ultimate outcome, which is crucial to clinical management and may be of great importance to inform patient’s relatives and for judicial concerning [43–47]. This valuable predictive power was not found for C-reactive protein (CRP) or white blood cells counting, another currently employed blood biomarkers [48–53]. The prognostic power of PCT dosing has also been stated for burned patients by Kim et al. [54] who, in a prospective observational study with a cohort of 175 patients, showed a significant correlation between PCT levels and mortality rate. In this context, it is worth to note, as referred by Piroglu et al. [55], that clinical scoring systems used to predict mortality of intensive care patients, like Acute Physiology and Chronic Health Evaluation Score II (APACHE II), Simplified Acute Physiology Score (SAPS), Sequential Organ Failure Assessment (SOFA), and Pediatric Risk of Mortality (PRISM), do not include parameters specific for burn patients, and these authors showed that combination of the former score with PCT significantly increased its accuracy. A prospectively study of Lavrentieva et al. [31], including 145 patients, concluded that the maximum PCT level has prognostic value in burn patients, and Mokline et al. [32] found a close correlation of PCT levels with sepsis severity, showing that increasing values were linked with worse outcomes and vice versa.

Another important use of PCT dosing is guiding antimicrobial therapy in septic ICU patients, which is becoming generally accepted [56], supported by several trials [57–62], systematic reviews, and meta-analysis [63–67]; however, some authors still consider that more studies on its safety and efficacy are needed yet [68, 69]. Once a clinical suspicion of sepsis is done, and in particular if corroborated by abnormally elevated PCT levels, empirical antimicrobial therapy, coupled with focus control when feasible, must be immediately started because survival is mostly depending on it and any delay, even hourly, is directly related with an increase in mortality [13, 70, 71]. On the other hand, there is an overwhelming acceptation that a lengthening of antimicrobial therapy beyond that strictly necessary to control the infectious process favors the development of microbial resistance, contributing to the soaring public health risk of having each time less sensitive microorganisms and lack of antimicrobials to combat them [72]. Many published works describe PCT kinetics as a mirror of the evolution of the infectious episode [73–75] as well as a trustable indicator of the antimicrobial therapy efficacy, allowing an early de-escalation and/or stopping of drug administration when its levels progress and consistently subside [76, 77]. When PCT levels keep elevated or even increasing, this is a sound indication that therapy is not working and/or that there are still infectious foci to clean, and if the situation is not rapidly controlled, a bad outcome is foreseeable.

Several authors have discussed in recent works this use of PCT, and a body of evidence is growing to support this approach. Jensen et al. in a trial (PASS Study) [78] published in 2011 advised against PCT-guided antimicrobial escalation, linking it to increased organ-related harm and length of stay at the ICU, without improvement in the outcomes. However, the sample analyzed came from just one developed country with antimicrobial restriction and a traditionally low microbial resistance. On the other hand, focus was not put on the possibility of using PCT levels to help decision on antibiotherapy discontinuation neither a subgroup analysis on burn patients was done. Nevertheless, and even if antimicrobial escalation may be somewhat controversial, PCT has proven to be very useful to monitor antimicrobials efficacy, with its levels paralleling clinical evolution, and to indicate when it is safe to stop it without prejudice to the patients [79]. Indeed, this methodology has proved to safely decrease antimicrobial consumption [80] by reducing days of antimicrobial therapy with strong potential to lower resistance development. This approach has already been validated for use in ICUs, with proven reduction of antimicrobial consumption without increase in morbidity or mortality [81]. Indeed, de Jong et al., in the largest prospective study in ICU patients published to date (SAPS Study) [61], were even able to show a significant reduction of mortality rate. The ever wider diffusion of PCT test, reducing its costs, and its efficacy in this setting, made also possible for some authors to consider it as probably cost-effective [82–86]. In a recent paper, Lavrentieva et al. [87] reported significantly shorter durations of antibiotic treatment in a PCT-guided group of burns patients compared to controls without differences in main outcome characteristics, including mortality rate, length of mechanical ventilation, and length of stay.

Among the limitations of this study are naturally its single-center, retrospective observational character as well as lacking of subgroup analysis according to concomitant pathologies. The definition of a precise cutoff of PCT levels for predicting outcomes or stopping antimicrobial therapy was also beyond the scope of this analysis and, as recognized in the literature, it will always be dependent on patient characteristics and facility features, and it is PCT kinetics that deserved authors attention, in spite of 100 ng/mL was often taken as an alert signal. On the positive aspects are the sample size and the strict use of ABA burn sepsis definitions for inclusion criteria. The strength of results from the present study would be largely enhanced by a desirable prospective multicentric trial.

The use of prognostic biomarkers in order to predict outcomes as well for guiding antimicrobial therapy in sepsis patients is nowadays a common practice in intensive care wards. As anytime more acknowledged in the literature, antimicrobial stewardship programs employing current available biomarkers or preferably, a panel of diverse ones, always associated with repeated clinical evaluation, may decisively improve patients’ stratification and antimicrobial use, optimizing patients outcome, reducing the spread of microbial resistance, and cutting financial burden [88–93]. Meanwhile more sophisticated and individualized system-based (integrating genomics, metabolomics, and proteomics) [94–96] data are not available to more accurately predict outcomes and tailor treatment options for burn victims, as well as other intensive care patients, PCT dosing will remain one of the more useful tools to help clinicians decisions.

In spite of its limitations, the close correlation between PCT levels and patients’ outcomes statistically demonstrated in the present work backs its use for prognosis determination in severe burn patients. Additionally, this study showed that the persistency of abnormally elevated PCT along the days of antimicrobial therapy was linked with poor outcomes in this set of patients, opposed to what happens when their levels fall in a consistent way, reflecting its efficacy.

Prospective multicentric studies would surely give more strength to the generalization of PCT use for prognosis and antimicrobial stewardship in burn patients and are much needed.