Clinical Usefulness of Procalcitonin in Hematopoietic Stem Cell Transplantation

Infectious disease following hematopoietic stem cell transplantation (HSCT) is a major cause of transplantation-related mortality. The main manifestation of infectious diseases is fever. However, fever appears with several transplantation-related complications besides infectious diseases, such as acute graft-versus-host-disease (aGVHD), engraftment syndrome, thrombotic microangiopathy (TMA), veno-occlusive disease (VOD), and relapse of the underlying diseases. The more valuable markers to distinguish infections disease from non-infectious complications are needed.

A total of 28 consecutive patients who underwent HSCT in our institution between September and December 2010 were included in the study. Procalcitonin (PCT) and C-reactive protein (CRP) were measured periodically throughout the clinical course. The febrile episodes were classified in four groups: systemic bacterial or fungal infections, immunological reactions such as engraftment syndrome and aGVHD, intracellular infections such as viral infection and tuberculosis, and the other febrile episodes including VOD, TMA, mucositis, enterocolitis not ascribed to infectious microorganisms, relapse of the underlying disease, and unidentified causes.

The median follow-up period for survivors was 101 days (range: 11–178 days). The diagnoses of 103 febrile episodes were classified as follows: systemic bacterial or fungal infections (n=30), intracellular infections (n=7), immunological reactions (n=19), and other febrile episodes (n=47). PCT and CRP level on the first day of fever (O-PCT and O-CRP, respectively) significantly increased in systemic bacterial or fungal infection (p<0.001 and p<0.001, respectively). Although pair-wise comparisons corrected for multiple testing showed no significant difference in O-CRP levels between systemic bacterial or fungal infection and intracellular infection (p=0.621), there was a significant difference between the two groups in O-PCT levels (p=0.040). Both O-PCT and O-CRP were significantly higher in cases of systemic bacterial or fungal infections when compared with immune reactions (p=0.008 and p=0.001, respectively), and with other etiologies (p<0.001 and p=0.002, respectively). The area under receiver-operator characteristics curve for detection of bacterial or fungal infection was 0.82 for O-PCT and 0.76 for O-CRP. The optimal diagnostic cut-off levels were 0.5 ng/ml for PCT and 4.0 mg/dl for CRP. Combining O-PCT (over 0.5 ng/ml) and O-CRP (over 4.0 mg/dl) level improved the specificity from 79.5% to 89.0% and the PPV from 60.5% to 66.6%. When PCT levels did not increase over 0.25 ng/ml through the fifth day of fever, PCT yielded a specificity of 100.0% and a negative predictive value of 58.9% for the exclusion of bacterial or fungal infection. In cases of the systemic bacterial or fungal infection, there were 13 events of sepsis and 17 events of severe sepsis (including septic shock). Although O-CRP level was only marginally higher in severe sepsis than in non-severe sepsis (p=0.049), O-PCT values were explicitly different between the two categories (p=0.002). We next examined the maximum value of PCT and CRP during each febrile episode (M-PCT and M-CRP, respectively). The median days to M-PCT and M-CRP were both 2 days (range: 0–5 days and 0–7 days, respectively). M-PCT was apparently higher in severe sepsis than in normal sepsis (p=0.003), however, there were no significant differences in M-CRP levels between the two categories (p=0.116). The estimated OS according to Max-PCT during HSCT was significantly different between the patients whose the maximum level of PCT during a whole course of HSCT (Max-PCT) increased 2.0 ng/ml or more and less than 2.0 ng/ml (log-rank test, p=0.001). In a multivariate analysis, Max-PCT ≥2 ng/ml was independently associated with worse overall survival as post-transplant predictors (adjusted hazard ratio 6.42, p=0.035).

PCT levels provide additional information for discrimination between bacterial or fungal infection and other causes, estimating the severity of infection and predicting the patient’s prognosis after HSCT by periodical measurement throughout a febrile episode.

No relevant conflicts of interest to declare.