Increased Serum Procalcitonin in Newborn Infants without Infection

High serum concentrations of procalcitonin, a stable 116-amino acid precursor molecule to the proteohormone calcitonin, have been reported in patients with severe systemic bacterial or fungal, but not viral, infections (1)(2). Serum procalcitonin concentrations correlated with the severity of bacterial infection and with the clinical course of the patients (3). In healthy volunteers, serum procalcitonin is barely detectable; however, it increases consistently after endotoxin injection (4). Therefore, serum procalcitonin is regarded by many authors as a useful marker of the presence, course, and prognosis of bacterial infection. To date we are aware of three reports describing the value of procalcitonin for diagnosis of neonatal infection (1)(5)(6). The reports agree that patients with bacterial invasion have high concentrations of serum procalcitonin. However, there are some discrepancies regarding the procalcitonin concentrations in sera of newborn infants without bacterial infection. In our study, procalcitonin concentrations were examined in 122 serum samples of 75 newborn infants. Neonates were included if surplus sample (20 μL) was available after routine determination of serum C-reactive protein (CRP); the study was conducted in accordance with the Helsinki Declaration of 1975/1983. Requests for routine CRP measurements were made by the neonatal ward of Oststadtkrankenhaus Hannover, according to the usual diagnostic protocol. This protocol includes analysis of serum CRP only in infants at increased risk of infection (premature rupture of membranes more than 12 h before delivery, discoloring of amniotic fluid, or clinical signs of maternal or neonatal infection). Thirteen infants received antibiotic treatment; the 22 samples obtained from these infants were excluded from further analysis. Blood was obtained from heel puncture; the sampling time was recorded diligently. If procalcitonin quantitation was not done the same day, the serum was stored at −20 °C. Procalcitonin was determined by an immunoluminometric assay (Lumitest® PCT, Brahms Diagnostica GmbH). For statistical comparisons, the Mann-Whitney test was used. In the study period, interassay CVs at low (≈1.3 μg/L) and high (≈59 μg/L) procalcitonin concentrations were <10% and <8%, respectively. Serum procalcitonin concentration was clearly related to the age of the infant (Table 1).In neonates <12 h of age, markedly increased procalcitonin concentrations (defined as >2 μg/L) were rarely observed; of 26 samples, only one, drawn 11 h after birth, showed a value of 3.8 μg/L. In contrast, procalcitonin >2 μg/L was found in a substantial proportion of samples obtained 12–47 h after birth. The highest values observed during 12-h periods were: 22.5, 8.6, and 5.7 μg/L at 12–23 h after birth; 11.7, 6.0, and 5.3 μg/L at 24–35 h; and 7.7 and 3.7 μg/L at 36–47 h. Mean procalcitonin peaked between 24 and 35 h after birth; the increase was highly significant (12–23 h vs 0–12 h, P <0.001; 24–35 h vs 12–23 h, P <0.05). Procalcitonin concentrations tended to decrease afterwards; the difference between the time periods 12–23 h and ≥48 h after birth was significant (P <0.001). Procalcitonin was clearly increased (2.4 μg/L) in one sample obtained more than 48 h after birth. Time course of serum procalcitonin concentration. Time course of serum procalcitonin concentration. Individual variations of serum procalcitonin concentrations were examined in 22 healthy neonates without antibiotic treatment. The general trends were uniformly reflected in individual time courses; serum procalcitonin always increased in the first 24 h after birth and always decreased in infants more than 36 h of age. In three infants, pronounced increases of procalcitonin concentrations (≥5 μg/L) were observed in the first 36 h after birth. There are conflicting data regarding serum procalcitonin in neonates without infection. Although procalcitonin concentrations >1 μg/L were not observed in the control neonates in two studies (1)(5), concentrations up to 15 μg/L were measured by other investigators at day 1 after birth (6). Our data confirm the latter findings and give a more detailed picture of the time course of serum procalcitonin in infants without infection. We studied neonates at increased risk of infection because we think that these patients represent the most relevant group for differential diagnosis of overt infection. We cannot rule out subclinical infection in some of our infants; however, it should be noted that none of them developed a significant increase of serum CRP, and they did well without antibiotic treatment. Our data suggest that procalcitonin concentrations peak between 24 and 36 h after birth; in this time period, markedly increased concentrations as well as clear increases in comparison to the first hours of life can be observed in the absence of relevant bacterial infection. Careful studies with regard to the age-dependent reference intervals of serum procalcitonin in infants should be conducted to maximize the use of this parameter in the diagnosis of neonatal sepsis.