Jul 14, 1988 - plasma and EDTA-thimerosal-aprotinin-treated plasma derived from blood collected from 10 leukemic and pyrexic patients. Aprotinin (100 x i03 ...
CLIN. CHEM. 35/3, 461-463 (1989)
Four ImmunochemicalMethods for Measunng C-Reactive Proteinin Plasma Compared Sven GrUtzmeier1 and Kenning von Schenck2 Four immunochemical methods for measuring C-reactive protein (CRP) in plasma were compared: radial immunodiffusion (RID), electroimmunoassay (EIA), immunoturbidimetry (IT), and laser nephelometry (LN). Close agreement was found between RID and EIA (EIA = 0.96 RID + 7.7 mg/L, r = 0.977, n = 100), ITand EIA(EIA = 1.11 IT - 15.8mg/L, r= 0.951, n = 100), IT and RID (RID = 1.10 IT 11.5 mglLr= 0.959, n = 94). In initial studies in 1983, LN showed good agreement with RID (LN = 0.98 RID + 11.9 mg/L, r = 0.973, n = 60). Since then, normal CAP values measured by LN have tended to increase and, at the time of the present study (1986), LN agreed poorly with other methods. The reason for this change is obscure. Having excluded the possibility of enzymatic degradation of CRP, we conclude that instrumentation problems are involved, and therefore we no longer use this method. -
AdditIonal Keyphraees: electroimmunoassay immunoturbidimetry laser nephelometry radialimmunodiffusion C-reactive protein (CRP) is the acute-phase reactant in serum that increases most rapidly and reaches very great concentrations upon infection or trauma.3 Its determination has thus been of great help in diagnosing infections or other causes of pyrexia, and is also valuable when one is deciding whether to prescribe an antibiotic. The traditional semiquantitative analysis (1) was replaced in 1969 by the quantitative but time-consuming technique of electroimmunoassay (EIA) developed by Kindmark (2). In studying CRP concentrations for several years in leukemia patients who had been undergoing chemotherapy, we used a rapid laser nephelometric assay (3) that gave results that agreed well with CRP measurements by radial iminunodiffusion (RID) (4) and by EIA. However, in 19851986 we noted in healthy subjects an unexplained increase in CRP exceeding our reference value of 10 mgfL as measured by laser nephelometry (LN). We therefore compared the four most commonly used methods to measure CRP to determine whether they yielded comparable results for both a reference population and a group of patients having high CRP values. We also investigated whether this alteration of measurable CRP values might be explained by increased enzymatic CRP degradation during either storage or incubation of the samples before measurement by LN.
MaterIals and Specimens CRP was quantified in plasma from 104 normal blood donors, by RID (4), EIA (2), immunoturbidimetry UT) (5), and LN (6). CRP values in plasma from 100 patients with leukemia
Medicine and 2 Clinical Chemistry, LinkOping, Sweden. 3Nonstandard abbreviations: CRP, C-reactiveprotein; RID, radial immunodiffusion; EIA, electroimmunoassay; IT, immunoturbidimetry; and LN, laser nephelometry. Received July 14, 1988; accepted December 21, 1988. Departments
of ‘Internal
University Hospital, 8-58185
and infections with high CRP values were determined by EIA, RID, and iT, and compared. Concentrations of CRP in plasma samples from an additional 100 leukemia patients with high CRP values were also quantified by if and LN and compared. Procedure Rabbit anti-human CRP antiserum (Dakopatts, Copenhagen, Denmark) was used in the LN, iT, and EIA determinations. To be assayable in the LN and if methods, lipemic sera were cleared by mixing 1 mL of serum with 1.5 mL of C2CI3F3 (“Lipoclean”; Behringwerke, Marburg, F.R.G.). For if determinations, we used a centrifugal analyzer (Multistat ifi; Instrumentation Laboratory, Lexington, MA). LN assays were done with a Laser Nephelometer PDQ (Hyland Diagnostics, Costa Mesa, CA) (6). We used: (a) a Behringwerke control that was in the range 75-85 mgfL; (b) plasma pooled (CRP, 120 mg/L) from patients with high CRP concentrations; and (c) a Bebringwerke standard diluted to make a standard curve with four points (the same as a). The standard curve, the pool, and the control were included with every batch of 18 samples. All standards, controls, and patients’ samples were analyzed with their respective “blanks.” We evaluated precision and analytical recovery once every month. CVs were between 5.68% and 10.6%. We did not change controls, standards, buffers, antisera, diluents, or laser beam source during the study period. For RID determinations we used commercially available plates (Norpartigen; Behringwerke). We performed the EIAs as described by Kindmark (2), using agarose plates in pH 8.6 barbital buffer, 70 pmol/L, containing 2 tmol of EDTA per liter. The electrophoresis time was 18 h, at 8 V/cm. To detect any effect of storage conditions on results for CRP, we analyzed 28 samples that had been centrifuged and then stored at -4#{176}C or +20#{176}C for 24,48, or 72 h. To assess the potential degradation of CRP, we determined CRP in serum, EDTA-treated plasma, EDTA-aprotinin-treated plasma and EDTA-thimerosal-aprotinin-treated plasma derived from blood collected from 10 leukemic and pyrexic patients. Aprotinin (100 x i03 arb. unitsfL) was immediately added to an aliquot of the blood sample before centrifugation to inhibit proteolysis, and 50 L of 0.1 mmolJL thimerosal was added per liter to prevent bacterial growth. Results Normal Subjects All 104 blood-donors’ samples analyzed by RID and EJA had CRP values below 10 mg/L and 11 mg/L, respectively. 01104 samples measured by IT, 97 had values below 10 mg/ L; values for the other seven ranged from 13 to 43 mg/L. All these sera were lipemic, and in all the CRP was
