PRO: Procalcitonin has clinical utility in children with community-acquired pneumonia

Abstract

Procalcitonin (PCT) is a useful, albeit imperfect, diagnostic aid that can help clinicians make more informed decisions around antibiotic use in children with lower respiratory tract infections (LRTI), including community-acquired pneumonia (CAP). Recent data suggest that a very low PCT concentration has a high negative predictive value to identify a population of children at low risk of typical bacterial infections. Although the preponderance of data on the clinical utility of PCT in LRTI come from adult studies, the potential for benefit is likely greatest in paediatric CAP and other LRTIs where viral aetiologies predominate, yet antibiotics are frequently prescribed.

Community-acquired pneumonia (CAP) is one of the most pervasive paediatric infections and the leading infectious killer of young children worldwide.¹ In the USA, although studies suggest that viral aetiologies of CAP predominate in the post-pneumococcal vaccine era, more than 70% of outpatient children with CAP receive antibiotics.^2,3 Additionally, CAP contributes the leading number of antibiotic days by hospitalized children.^4,5 The management of CAP, including hospitalization rates, varies across institutions and providers, emphasizing the need for evidence-based tools to guide clinical decision-making.^6,7 Furthermore, clinical gestalt for prediction of severe CAP complications is only fair.⁸ These realities illustrate two fundamental challenges in the management of children with CAP: identifying children who require antibiotics (and limiting antibiotics in those less likely to benefit) and risk stratification to identify the proper site of care.

The biomarker procalcitonin (PCT) has received a great deal of attention as a potential solution to the challenges of targeting appropriate antibiotic use and risk stratification in children with CAP. PCT is not typically detectable in healthy individuals.^9–11 PCT expression and secretion are stimulated in inflammatory and infectious conditions, but circulating levels are particularly elevated in severe bacterial infection, sepsis and multiple organ dysfunction.^10,11 Procalcitonin has been studied as a marker of serious bacterial infection in children in several contexts, including the febrile neonate,¹² fever without source in young children,¹³ urinary tract infections,¹⁴ sickle cell disease¹⁵ and in those with central venous catheters.¹⁶

Several prior studies have highlighted associations between PCT and bacteraemic pneumonia.^17,18 The best evidence examining associations between PCT and bacterial CAP in children comes from the CDC Prevention Etiology of Pneumonia in the Community (EPIC) study, a prospective, population-based study that used comprehensive diagnostics to determine microbiological aetiology among more than 2000 children hospitalized with CAP in three US cities.² In a convenience sample of 532 children enrolled in EPIC, PCT was collected at the time of admission and evaluated to determine its association with microbiological aetiology using comprehensive molecular and culture-based diagnostics. Of these, 349 (66%) children had only viral pathogens detected, 82 (15%) had atypical bacterial pathogens (with or without a viral co-detection), and 54 (10%) had typical bacterial pathogens detected (with or without a viral co-detection).¹⁹ Median PCT levels were significantly higher in those with typical bacterial detections (median 6.10 ng/mL) compared with viral only (0.33 ng/mL; P < 0.001) and atypical bacterial detections (0.10 ng/mL; P < 0.001). None of the 120 children with very low PCT levels (<0.1 ng/mL) had typical bacteria detected (sensitivity 100%, 95% CI 92%–100%), while only 9 of the 247 (3.6%) with a PCT level <0.25 ng/mL had typical bacteria detected (sensitivity 85%, 95% CI 72%–93%; negative predictive value 96%, 95% CI 91%–98%). Importantly, while low and very low levels of PCT demonstrated high sensitivity, specificity was limited (45% and 20% for PCT <0.25 ng/mL and 0.1 ng/mL, respectively). This means PCT is most helpful for ruling out bacterial aetiologies but should not be used to rule in bacterial CAP. It is also important to note the lack of a true reference standard for bacterial CAP, a key limitation common to all investigations of pneumonia aetiology that must be kept front of mind. It is unlikely that new diagnostic tests will emerge in the near future to resolve this challenge, further underscoring the importance of clinical judgement as the ultimate decider.

In 2017, PCT was approved by the US FDA to guide antibiotic treatment decisions in acute lower respiratory tract infections (LRTI).²⁰ The most commonly studied algorithm categorizes PCT values into four groups (<0.1 ng/mL, 0.1 to <0.25 ng/mL, 0.25 to 0.5 ng/mL and >0.5 ng/mL) corresponding to likelihood of bacterial aetiology (very low, low, moderate and high) with higher values indicating a higher likelihood of bacterial disease. Algorithms that use PCT to guide antibiotic use decrease antibiotic exposure in adults with CAP. A meta-analysis of 26 randomized trials comparing PCT guidance to standard care that included 6708 adults with LRTI found lower mortality in the PCT guidance group as compared with controls [8.6% versus 10%, adjusted OR (aOR) 0.83, 95% CI 0.7–0.99]. PCT guidance was associated with 14.8% less antibiotic initiation (71.5% versus 86.3%, aOR 0.27, 95% CI 0.24–0.32), 2.4 fewer days of antibiotic exposure (5.7 days versus 8.1 days) and lower risk of antibiotic-related side effects (16.3% versus 22.1%, aOR 0.68, 95% CI 0.57–0.82).²¹ Treatment failure rates were similar between groups, as were length of hospital and intensive care stays. Even in individual studies where algorithms did not decrease antibiotic use, patients with lower PCT concentrations had the lowest antibiotic prescribing, hospitalization and intensive care use in both PCT guidance and standard care (i.e. where the PCT value was obtained but not provided to the clinician) groups.²²

Data regarding the impact of PCT testing in children, although less abundant than adult data, are encouraging. One paediatric trial found a 14% absolute decrease in antibiotic use in the PCT-guided group of hospitalized children with CAP with PCT < 0.25 ng/mL.²³ In this study, antibiotic duration was also shorter in the PCT group (5.4 versus 11 days), and there were fewer antibiotic-related adverse effects (3.9% versus 25.2%). Another randomized trial found no difference in antibiotic initiation between PCT-guided and standard-of-care groups (62% versus 56%, OR 1.26, 95% CI 0.81–1.95), but a decrease in antibiotic duration with PCT guidance (5.7 versus 9.1 days).²⁴ A retrospective study from China found a 30% lower antibiotic prescribing rate (54.6% versus 83.9%) with no clinically or statistically significant differences in hospital length of stay (10 versus 10.6 days) or adverse effects (23% versus 29.3%).²⁵ These data suggest that at low concentrations, particularly in a well-appearing child, PCT is useful in decreasing antibiotic use without an uptick in adverse outcomes.

Procalcitonin has also been evaluated as a marker of severity in paediatric CAP. In the CDC EPIC study, PCT was higher among children admitted to intensive care (median 0.61 ng/mL versus 0.24 ng/mL; P < 0.001) and in those with pneumonia complicated by empyema (median 2.94 ng/mL versus 0.27 ng/mL; P < 0.001) as compared with those without these severe features.¹⁹ A study of 100 children with pneumonia demonstrated higher PCT levels among those requiring hospitalization versus outpatients (median 17.8 ng/mL versus 0.72 ng/mL).²⁶ In another study of 488 children with pneumonia ranging in severity from mild outpatient pneumonia to very severe disease requiring invasive respiratory or vasopressor support, increasing PCT was associated with increasing severity (range of aOR 1.03–1.25).²⁷ Only those with very severe disease (median 5.06 ng/mL) demonstrated PCT values substantially higher than other groups (medians for mild, moderate and severe disease of 0.21 ng/mL, 0.29 ng/mL and 0.38 ng/mL, respectively). A similar study of 477 children with LRTI also demonstrated associations between PCT and an ordinal severity outcome (aOR 1.08), although associations were again driven by higher PCT in those with the most severe outcomes.²⁸ These studies suggest only modest associations between PCT and LRTI severity in children and the utility of PCT for predicting disease severity may be limited, although additional studies are needed.

Procalcitonin is not the ‘silver bullet’ in the management of paediatric CAP. However, to dismiss its utility altogether would be throwing out the baby with the bathwater. Existing and emerging evidence suggests that PCT is most helpful identifying children with CAP who may not require antibiotics due to low risk of bacterial disease and to shorten duration of antibiotics in those who initially required them. For example, when the perceived likelihood of bacterial pneumonia is equivocal, clinicians very often prescribe antibiotics. In this scenario, a very low PCT may provide added reassurance that antibiotics have little benefit. Given the predominance of viral illness and the large number of children with CAP who receive antibiotics empirically, there is potential to substantially decrease antibiotic use in children whose PCT concentration is low, where viral illness is most likely. Similarly, serial PCT measurements may help guide antibiotic duration in children recovering from confirmed bacterial CAP. Due to its modest specificity, PCT should not be used to identify children who do require antibiotics. At this time, there is insufficient evidence to suggest that PCT be used to predict disease severity or risk stratification as a standalone marker. Importantly, in all cases, PCT should always be used in conjunction with, and not as a replacement for, clinical judgement.

Procalcitonin is an FDA-approved biomarker that can support clinical decision making in paediatric CAP. As highlighted above, the most promising clinical use of PCT in this population is in limiting antibiotic use in those with viral aetiologies of disease. Given the widespread nature of paediatric CAP, along with the high burden of viral aetiologies and widespread antibiotic use that is often unnecessary, the potential stewardship benefits of PCT when used as intended are monumental. At the same time, we fully acknowledge the potential consequences of diagnostic overuse that often accompany the marketing of new diagnostic tests. Thus, while we encourage use of PCT to inform antibiotic decisions when the risk of bacterial disease is equivocal, we do so with caution and a call for continued study. In particular, trials evaluating clinical outcomes in children with low PCT and treated without antibiotics, and pragmatic randomized trials embedded within clinical care and across diverse patient populations are needed to evaluate the real-world clinical utility of PCT against clinical judgement alone.

Funding

This work is supported in part by the National Institutes of Health under award R01AI125642 to Dr Williams and awards R03AI147112 and R34HL153474 to Dr Florin.

Transparency declarations

Dr Williams reports receiving previous in-kind support from BioMérieux, a maker of procalcitonin assays. Drs Florin and Williams have no conflicts of interest to declare. The views in this report are solely those of the authors and do not represent the views of the National Institutes of Health or BioMérieux.