Detection of acute phase response and infection. The role of procalcitoninand C-reactive protein

Objective: Established parameters, e.g. C-reactive protein (CRP), do not di fferentiate specifically enough between patients developing an infection an d those exhibiting an acute phase response following cardiac surgery. The o bjective of this prospective study was to investigate if procalcitonin (PCT ) ist more helpful than CRP. Methods: During a 1-year period, seven out of 563 patients (1.2 %) develope d systemic infections (group A) after cardiac operations with cardiopulmona ry bypass (CPB), and additional eight patients (1.4%) had local wound infec tions requiring surgical therapy (group B). Blood samples for PCT and CRP m easurements were taken preoperatively, at the onset of infection (d1), as w ell as on the third day (d3), fifth day (d5), and seventh day (d7) followin g diagnosis of infection. Forty-four randomly selected patients undergoing cardiac surgery with CPB without clinical signs of infection, additional in tensive care unit (ICU) management or additional antibiotic treatment serve d as control (group C) to assess the PCT and CRP contribution to acute phas e response. PCT and CRP levels were measured preoperatively, on the first ( dl), third (d3) and fifth day (d5) after operation. Results: At the onset of infection, PCT levels (median interquartile range 25%-75%) increased significantly in group A as compared to baseline values (10.86 (3.28-15.13) ng/ml vs. 0.12 (0.08-0.21) ng/ml), and decreased during treatment to still significantly elevated values on d5 (0.56 (0.51-0.85)ng /ml). CRP levels were significantly elevated on all days investigated with no trend towards normalisation (dl: 164.5 (137-223) mg/l) vs. 1.95 (1.1-2.8 ) mg/l preoperatively, d5: 181.1 (134-189.6) mg/l. In group B, no increase in PCT levels, but a significant increase of CRP from dl (165.9 (96.6-181.6 ) mg/l vs. 3.7 (2-4.3) mg/l preopratively) until d5 98 (92.8-226.2) mg/l wa s detected. In group C, postoperative PCT levels increased slightly but sig nificantly in the absence of infection on dl (0.46 (0.26-0.77) ng/ml vs. 0. 13 (0.08-0.19) ng/ml preoperatively, and d3 (0.37 (0.2-0.65)ng/ml and reach ed baseline on d5 (0.24 (0.11-0.51) ng/ml)). CRP levels were significantly elevated as compared to baseline on all postoperative days investigated (ba seline: 1.75 (0.6-2.9) mg/l, dl: 97.5 (74.5-120) mg/l), d3: 114(83.05-168.5 ) mg/l, d5: 51.4 (27.4-99.8) mg/l)). PCT showed a significant correlation t o CRP in group A (r = 0.48, p<0.001), a weak correlation in group C (r = 0. 27, p = 0.002) and no correlation in group B. Intergroup comparison reveale d a significant difference for PCT between all groups (A>C>B) and significa ntly higher CRP levels in group A vs. C and in group B vs. C. Thus, the pat tern high PCT/high CRP appears to indicate a systemic infection, while low PCT/high CRP indicates either acute phase response or local wound problems but no systemic infection. The cost for PCT measurements was 5.6-fold highe r as compared to CRP. Conclusion: Due to the significant differences in the degree of increase, P CT appears to be useful in discriminating between acute phase response foll owing cardiac surgery with CPB or local problems and systemic infections, w ith additional CRP-measurement increasing the specifity.