The clinical value of procalcitonin in the prediction of infected necrosisin acute pancreatitis

Citation
B. Rau et al., The clinical value of procalcitonin in the prediction of infected necrosisin acute pancreatitis, INTEN CAR M, 26, 2000, pp. S159-S164
Citations number
11
Categorie Soggetti
Aneshtesia & Intensive Care
Journal title
INTENSIVE CARE MEDICINE
ISSN journal
03424642 → ACNP
Volume
26
Year of publication
2000
Supplement
2
Pages
S159 - S164
Database
ISI
SICI code
0342-4642(2000)26:<S159:TCVOPI>2.0.ZU;2-C
Abstract
Objective: Infection of pancreatic necrosis (IN) has a major impact on mana gement and outcome in acute pancreatitis (AP). Currently, guided fine-needl e aspiration (FNA) is the only means for an accurate diagnosis of IN. Pro-c alcitonin (PCT), a 116 amino acid pro-peptide of calcitonin has been found in high concentrations in patients with sepsis. In the present study we ana lyzed the clinical value of serum PCT for predicting IN in AP and compared the results to guided FNA. Design: Clinical study. Setting: A collaborative study between the Departments of General Surgery a nd Clincal Chemistry/ Pathobiochemistry of the University of Ulm, Germany. Patients: 61 patients with AP entered this study and were stratified into t hree groups according to morphological and bacteriological data: I. 22 pati ents with edematous pancreatitis (AIP), II. 18 patients with sterile necros is (SN), III. 21 patients with IN. Measurements and Results: During an obse rvation period of 14 days PCT was measured by immunoluminometry, CRP was de termined by lasernephelometry on a routine base. In patients with IN overal l PCT concentrations were significantly higher than in those with SN, where as CRP levels did not differ in both groups. In contrast, only low concentr ations of both parameters were found in patients with AIP. By ROC analysis the best PCT cut-off level for predicting IN or persisting pancreatic sepsi s was obtained at greater than or equal to 1.8 ng/ml. If this cut-off was r eached on at least two consecutive days, IN could be predicted with a sensi tivity of 95%, a specificity, of 88%, and an accuracy of 90%. Guided FNA ac hieved a sensitivity, specificity, and accuracy of 91%, 79%, and 84% in dif ferentiating IN from SN, respectively. After surgical treatment of IN media n PCT values continued to be significantly higher in patients with persisti ng pancreatic sepsis (n = 12) compared to those with an uneventful postoper ative course (n = 7). Our results demonstrate that monitoring of serum PCT could serve as a noninvasive and accurate method to predict IN in AP as wel l as to select patients with persisting septic complications after surgical debridement.