C-reactive protein in clinical practice with special regard to infectious diseases

Abstract

A new simple affinity purification method for human CRP using Arginine-Sepharose 4B was developed with a recovery of >90 % and containing < 5-10 % contaminating proteins. In a multicenter study including 607 consecutive patients in general practice, erythrocyte sedimentation rate (ESR) and serum CRP were compared. Comparable results were obtained in 71 % of the patients. In 20 % of the patients CRP was increased more than ESR. In patients with increase in serum CRP above 50 mg/l but with "normal" or only marginally increased ESR, the increased CRP concentration was often due to the presence of inflammatory diseases, mostly infections. Patients with an increased ESR but a normal or marginally increased CRP concentration were all more than 50 years of age and the high ESR was usually a consequence of an earlier infectious disease.

The serum CRP, ESR, white blood cell count (WBC), and polymorphonuclear granulocytes (PMN) were significantly more increased in patients with either septicaemia/endocarditis(N=176) or complicated influenza (N=22) than in patients with uncomplicated influenza (N=59). Serum CRP was more efficient than WBC, PMN, and ESR to differentiate between bacterial infection and pure viral infection. Five of 56 patients with uncomplicated influenza had serum CRP values >100 mg/l. However, 32 % of the patients with bacterial infections had WBC<9.0 x 109/1 and 30 % had PMN < 75 %.CRP in cerebrospinal fluid (CSF) was evaluated in 710 patients with suspected central nervous system (CNS) infection for the differential diagnosis of meningitis. In a subset of these patients (N=235) the ability of serum CRP to differentiate between acute bacterial and viral meningitis was evaluated. An overlap existed for both CSF and serum CRP between the different etiological groups. Even though CSF-CRP was the second best CSF test for the diagnosis of bacterial meningitis compared to CSF lactate CSF-CRP offered little or no additional information.

For patients 26 years of age a serum CRP level of 50 mg/l was empirically chosen as a decision limit between patients with bacterial and non-bacterial CNS-diseases. For children<6 years of age a decision limit for serum CRP as low as 20 mg/l could be used with a sensitivity of 97 % and a specificity of 86 %. Seven of the patients with bacterial meningitis had serum CRP values below 50 mg/l, four of them were children below the age of six years. Five of the seven patients had a duration of symptoms at admission of less than 12 hours. Patients with high serum CRP level (>100 mg/l) and viral meningitis often had signs of extensive tissue involvement. The median duration of illness prior to admission for 97 patients (mean age 69.6 years) with community-acquired pneumonia (CAP) was four days (range: <1-25 days). On admission 64 % of the patients were hypoalbuminemic (<37g/1). The serum level of albumin at admission did not correlate with the nutritional measurements, however, significant negative correlation was found for serum albumin with the acute phase proteins: CRP, haptoglobin, a-l-antitrypsin and oro-somucoid.

These results strongly indicate that inflammatory reaction is the main causes of the low serum albumin level in patients with CAP. The serum level of CRP was highest at admission (geometric mean: 154 mg/l) and began to decrease from day 3 indicating a very earlier response to the recovery of the patients. The median normalisation rate of CRP was 3.3 days. Albumin, transferrin, a-1anti-trypsin, haptoglobin, and orosomucoid were not useful for evaluation of response to therapy during the hospitalisation phase in patients with CAP. CRP was evaluated together with other plasma proteins as prognostic markers in patients with malignant melanoma (N=643). The serum levels of the different serum proteins had independent prognostic value in relation to the clinical stage in multivariate analyses. For each clinical stage the serum values below respectively above the different cut-off levels for albumin and transthyretin on one hand and the acute phase proteins CRP, haptoglobin, and orosomucoid on the other hand significantly influenced the survival rate and the relative hazard with respect to death.