Sepsis is systemic inflammation due to infection. It is the most common cause of death in intensive care units (1), with a mortality rate up to 50% depending on severity. Sepsis can be considered in individuals with systemic inflammatory response syndrome (SIRS) manifested by alterations in at least two of temperature, heart rate, respiratory rate, and/or white blood cell count. These SIRS criteria are non-specific for sepsis and may be present in alternative pathophysiologic states, both medical and surgical (1).
The earlier sepsis is identified and treated the better the prognosis (1). Currently diagnosis involves documentation of infection (by culture) in patients with SIRS. Cultures take at least 24 hours and > 50% may be false negative and 30% false positive. A biomarker that could identify sepsis early in a population of patients with overlapping clinical symptoms, i.e. patient’s with non-infectious SIRS would improve patient care.
Many suggest that Procalcitonin (PCT) is an ideal sepsis biomarker. PCT is the pro-hormone form of calcitonin and is produced by extra-thyroidal immune cells within 2-4 hours of a bacterial insult and/or inflammatory response. Elevated PCT is seen in septic patients and concentrations correlate with severity of disease. Increasing PCT over time is associated with poor prognosis, while decreasing concentrations correlate to good prognosis and/or response to antibiotic therapy. Thus, proponents suggest that the clinical utilities of PCT are: to differentiate patients with sepsis from those with non-infectious SIRS; to guide antibiotic therapy; and to predict prognosis of critically ill patients (2).
Procalcitonin: an early predictor of sepsis?
Thousands of studies have investigated the clinical utility of almost 200 biomarkers of sepsis (3). Among these, PCT and CRP are most often referenced. In comprehensive reviews comparing their diagnostic accuracies, PCT consistently performs better, but sensitivity and specificities to predict sepsis are <90% for both markers. Two meta-analyses recommend routine use of PCT for diagnosis of sepsis while two others discourage its use (3-6). Why the discrepancy? Lack of a gold standard for sepsis diagnosis. Diverse patient populations--PCT is elevated postsurgery, trauma, or in other instances of systemic inflammation limiting its diagnostic accuracy. Different PCT cut-offs are used. Sepsis is diagnosed retrospectively with knowledge of bacterial cultures and a full clinical picture in most studies. To date, few studies have addressed the utility of PCT to predict sepsis in “real time”. Until, large real time studies with well defined patient populations are completed, the utility of PCT to predict sepsis will remain controversial.
Does PCT have a role in tailoring antibiotic therapy?
Overuse of antibiotics in ICU settings has contributed to the increase in antibiotic resistant bacteria. Most randomized control trials demonstrate that using serial PCT based algorithms to guide continuation/cessation of antibiotic therapy reduces the number of days patients are on antibiotics without adverse effects. None has demonstrated a reduction in multidrug resistant (MDR) bacterial infections (7). In practice, there is limited utility for PCT to guide initiation of antimicrobial therapy because most ICU patients are taking antibiotics at admission. Further, these algorithms are not effective in patients with recent trauma, surgery or other global inflammation. Thus, PCT based algorithms for guiding antibiotic therapy are limited to guiding secession/continuation of therapy in non-surgical/trauma ICU patients. And even in these populations, serial PCT measurements may be cost prohibitive if no reduction in MDR infections is observed.
Can procalcitonin be used to predict prognosis?
Increasing PCT concentrations correlate with increasing severity of sepsis and poor outcome, while decreasing or low PCT predicts a good prognosis for ICU patients. Most studies exclude patients with a recent trauma, surgery or other inflammatory event (6).
Given the complicated pathobiology of sepsis and significant overlap in clinical symptoms with SIRS patients, it may be that no one biomarker works for sepsis. Studies looking at multi-biomarker panels have shown promise in the prediction and monitoring of sepsis in ICU and ED patients (3).
With today’s ED and ICU populations, is procalcitonin the answer to the sepsis dilemma?