Scandinavian Journal of Clinical and Laboratory

Investigation

ISSN: 0036-5513 (Print) 1502-7686 (Online) Journal homepage: http://www.tandfonline.com/loi/iclb20

Wide-range CRP versus high-sensitivity CRP on

Roche analyzers: focus on low-grade inflammation
ranges and high-sensitivity cardiac troponin T
levels

Denis Monneret, Fouzi Mestari, Shaedah Djiavoudine, Guillaume Bachelot,

Maxime Cloison, Françoise Imbert-Bismut, Maguy Bernard, Pierre Hausfater,
Jean-Marc Lacorte & Dominique Bonnefont-Rousselot

To cite this article: Denis Monneret, Fouzi Mestari, Shaedah Djiavoudine, Guillaume Bachelot,
Maxime Cloison, Françoise Imbert-Bismut, Maguy Bernard, Pierre Hausfater, Jean-Marc
Lacorte & Dominique Bonnefont-Rousselot (2018): Wide-range CRP versus high-sensitivity
CRP on Roche analyzers: focus on low-grade inflammation ranges and high-sensitivity
cardiac troponin T levels, Scandinavian Journal of Clinical and Laboratory Investigation, DOI:
10.1080/00365513.2018.1471618

To link to this article: https://doi.org/10.1080/00365513.2018.1471618

View supplementary material Published online: 15 May 2018.

Submit your article to this journal Article views: 42

View related articles View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=iclb20
SCANDINAVIAN JOURNAL OF CLINICAL AND LABORATORY INVESTIGATION
https://doi.org/10.1080/00365513.2018.1471618

ORIGINAL ARTICLE

Wide-range CRP versus high-sensitivity CRP on Roche analyzers: focus on

low-grade inflammation ranges and high-sensitivity cardiac troponin T levels
Denis Monnereta, Fouzi Mestaria, Shaedah Djiavoudinea, Guillaume Bachelota, Maxime Cloisona,
Françoise Imbert-Bismuta, Maguy Bernarda,b, Pierre Hausfaterc,e, Jean-Marc Lacorteb,d and
Dominique Bonnefont-Rousselota,f
a
Department of Metabolic Biochemistry, Piti
e Salp^etriere-Charles Foix University Hospital (AP-HP), Paris, France; bDepartment of Oncology
and Endocrine Biochemistry, Piti
e Salp^etriere-Charles Foix University Hospital (AP-HP), Paris, France; cSorbonne Universit
es, UPMC-Univ Paris
06, GRC-14 BIOSFAST, Paris, France; dSorbonne Universit
es, UPMC Univ-Paris 06; INSERM, UMR_S 1166, Institute of Cardiometabolism and
Nutrition, ICAN, Paris, France; eDepartment of Emergency, Piti
e Salp^etriere-Charles Foix University Hospital (AP-HP), Paris, France; fSorbonne
Paris Cit
e, Paris Descartes University, CNRS UMR8258-INSERM U1022, Faculty of Pharmacy, Paris, France

ABSTRACT ARTICLE HISTORY

Wide-range C-reactive protein (wr-CRP) has been proposed as an economical alternative to high-sensi- Received 4 November 2017
tivity C-reactive protein (hs-CRP) for the evaluation of low-grade inflammation-associated cardiovascular Revised 22 April 2018
risk (LGI-CVR). Concomitant values of serum hs-CRP and plasma wr-CRP
5 mg/L, and high-sensitivity Accepted 28 April 2018
cardiac troponin T (hs-cTnT), all assayed on Roche Diagnostics analyzers over a 1.8-year period, were
KEYWORDS
extracted from a hospital laboratory database. Hs-CRP and wr-CRP values were compared C-reactive protein;
(Bland–Altman method; Deming’s correlation), then separately classified into low (<1 mg/L), moderate inflammation; cardiovascular
(1–3 mg/L) and high (>3 mg/L) LGI-CVR ranges for agreement test (j), assessed before and after system; cardiovascular risk;
Deming’s regression-based adjustment of wr-CRP (Adj-wr-CRP). Wr-CRP and hs-CRP values were troponin T
strongly correlated, with linearity, whether below 5 mg/L (n ¼ 744; s ¼ 0.933; p < .001) or below 1 mg/L
(n ¼ 283; s ¼ 0.823; p < .001). Overall, wr-CRP values were lower than hs-CRP (mean bias:
–0.11 ± 0.17 mg/L). Agreement was good, with 8.1% of wr-CRP values misclassified compared to hs-CRP
(j: 0.874), and weakly improved after regression-based adjustment (7.7% reclassified values; j: 0.881).
Lowering the Adj-wr-CRP cutoff of the moderate LGI-CVR subrange from 1.0 to 0.9 mg/L resulted in an
almost perfect agreement (3.2% reclassified data; j: 0.950). Hs-cTnT concentration was positively associ-
ated with hs-CRP, wr-CRP, and Adj-wr-CRP (p < .001). Within each LGI-CVR subrange, hs-cTnT medians
were similar regardless of the hs-CRP, wr-CRP or Adj-wr-CRP used for risk classification. Based on hs-
cTnT, this study supports the use of wr-CRP as a low-cost alternative to hs-CRP for cardiovascular
risk evaluation.

Introduction
Since the mid-1990s, the role of chronic inflammation in the
atherosclerotic process and the development of cardiovascu-
lar diseases has been well-established [1–3]. In 1997, high-
sensitivity C-reactive protein (hs-CRP) has emerged as a
predictive biomarker of first-ever vascular events in healthy
subjects [4]. Thereafter, its increase has been confirmed as
being associated with a higher risk of coronary heart disease
following ranges: low risk <1 mg/L, moderate risk 1–3 mg/L,
and high risk >3 mg/L [9,12,13].
In parallel, many studies have demonstrated that the
‘regular’ wide-range CRP (wr-CRP) provides similar results
to hs-CRP in various clinical contexts, even at low concen-
trations, suggesting it could be an acceptable alternative
[14–17]. In particular, Ziv-Baran et al. [18] have recently
shown that wr-CRP assayed by immunoturbidimetry
(ADVIAV 2400; Siemens Healthcare Diagnostics, Tarrytown,
R

and cardiovascular morbidity [5–8], with a prognostic value NY) may be used as an economical alternative of hs-CRP
as important as that of cholesterol level or blood pressure measured by nephelemetry (BNIIV, Dade Behring, Eschborn,
R

[6,9]. Under statin therapy, hs-CRP concentrations
decreased in an low-density lipoprotein (LDL)-independent
manner [10,11], leading US guidelines to recommend its
assay in primary prevention (class IIb), when clinical deci-
sions for starting statin therapy are uncertain [9]. It is now
considered as the key biomarker of low-grade inflammation-
related cardiovascular risk (LGI-CVR), considering the
Germany) for evaluating cardiovascular risk in routine
annual checkups. More specifically, they demonstrated that
agreement between both methods is optimal after a linear
regression-based adjustment of wr-CRP and after lowering
the moderate-risk threshold value from 1.0 to 0.9 mg/L.
With the aim of testing whether such an alternative is pos-
sible on Roche Diagnostics analyzers, we compared wr-CRP

CONTACT Denis Monneret dmonneret2@gmail.com Department of Metabolic Biochemistry, Piti
e Salp^etriere-Charles Foix University Hospital (AP-HP),
Paris, France
Supplemental data for this article can be accessed here.
ß 2018 Medisinsk Fysiologisk Forenings Forlag (MFFF)
2 D. MONNERET ET AL.
values <5 mg/L to those of hs-CRP, both assayed by latex-
enhanced immunoturbidimetric methods, using high-sensi-
tivity cardiac troponin T (hs-cTnT) as a reference biomarker
of cardiovascular risk.
Material and methods
Concomitant results of hs-CRP, wr-CRP, hs-cTnT and cre-
atinine assayed over a 1.8-year period (November 2015–July
2017) were extracted from the laboratory information system
(GLIMSV software, MIPS-CliniSys, Chertsey, UK) of the bio-
R
chemistry laboratory (Piti
e-Salp^etriere Hospital, Paris,
France). Hs-CRP was assayed on serum [Vacutainer CAT
tube, ref#368815, Becton-Dickinson (BD), Franklin Lakes,
NJ, USA], whereas wr-CRP and creatinine were assayed on
lithium-heparinized plasma (BD Vacutainer LH-PSTII,
ref#367374, with separator gel), as well as hs-cTnT (BD
Vacutainer LH, ref#368884, without separator gel). All tubes
were centrifuged on Modular Pre-AnalyticsV (Roche
R
Diagnostics, Mannheim, Germany) for 10 min at 1885 g
(temperature 17.3 ± 0.4  C).
High-sensitive CRP assay
V
R
Hs-CRP was measured on a Cobas c501 analyzer (Roche
Diagnostics, Germany) using latex particle-enhanced immu-
noturbidimetric assay [‘Cardiac C-Reactive Protein (Latex)
High Sensitive’ reagent kit; ref#04628918190], which is based
on the aggregation between serum CRP and latex particles
coated with monoclonal antibodies specific to human CRP,
resulting in an increase of the turbidity measured at 546 nm.
The analytical characteristics (French technical sheet) are the
following: limit of detection (LOD): 0.15 mg/L; functional
sensitivity (FSe): 0.3 mg/L (i.e. the lowest CRP concentration
corresponding to an inter-series coefficient of variation
<10%); limit of linearity (LOL): 185 mg/L; mean laboratory
inter-assay coefficient of variation (MLCVia) over the study
period: 3.97% (mean quality control: 4.31 mg/L) (CRP T
Control NV; Roche Diagnostics).
R
Wide-range CRP assay
Wr-CRP was measured on two ModularV P800 analyzers
R medians were assessed using a non-parametric unpaired test
(Roche Diagnostics, Germany), also using the latex-
enhanced immunoturbidimetric method (‘Tina-quant C-
Reactive Protein Gen.3’ reagent kit; ref#04956923190), with
a turbidity measurement at 570 nm. The analytical character-
istics (French technical sheet) are the following: limit of
blank (LOB): 0.2 mg/L; LOD: 0.3 mg/L; FSe: 0.6 mg/L (i.e.
the lowest CRP concentration corresponding to an inter-ser-
ies coefficient of variation <20%); LOL: 350 mg/L; reference
calibrator: CRM470; MLCVia: 1.59% (mean QC1: 7.51 mg/L)
and 2.17% (mean QC2: 37.0 mg/L) (PreciControl ClinChem
MultiV; Roche Diagnostics).
R
High-sensitive cardiac troponin T assay
Hs-cTnT was measured on two ModularV E170 analyzers
R
(Roche Diagnostics, Germany) by electrochemiluminescent
immunoassay (reagent kit; ref#05092744190), characterized
as follows (French technical sheet): LOB: 3 ng/L; LOD: 5 ng/
L; FSe: 13 ng/L (i.e. the lowest hs-cTnT concentration corre-
sponding to an inter-series coefficient of variation
10%);
LOL: 10,000 ng/L; MLCVia: 2.33% (mean QC1: 28.7 ng/L)
and 1.87% (mean QC2: 2046 ng/L) (PreciControl
TroponinV; Roche Diagnostics). Negative hs-cTnT results
R
(i.e. <3 ng/L) were changed to 3 ng/L and patients with hs-
cTnT >10,000 ng/L were discarded.
Creatinine assay
Creatinine was measured on two ModularV P800 analyzers
R
using the IDMS-standardized, rate-blanked compensated
kinetic Jaff
e method (reagent kit; ref#11875418216), charac-
terized as follows: LOD: 18 mmol/L; LOL: 2210 mmol/L;
MLCVia: 2.81% (mean QC1: 100 mmol/L) and 1.89% (mean
QC2: 371 mmol/L) (PreciControl ClinChem MultiV; Roche
R
Diagnostics). Patients with plasma creatinine concentrations
out of the analytical measuring range were discarded.
Calculations and statistics
Data were analyzed using Microsoft ExcelV and MedCalcV
R R
(MedCalc, Ostend, Belgium) software. Given their non-nor-
mal distribution (D’Agostino-Pearson test), concentrations
were expressed as median and interquartile range (IQR) or
10–90 percentile. Methods comparisons were assessed using
the Bland and Altman plot and Deming regression, while
groups were compared using a Mann–Whitney test.
Correlations were assessed using the Kendall’s rank correl-
ation test (coefficient s) with bootstrapping, which is known
to be less sensitive to tied values [19,20]. Cohen’s kappa
coefficient (j) was used to evaluate the agreement between
the different CRP-based risks ranges (low, moderate and
high LGI-CVR). Given the impact of age and renal impair-
ment on hs-cTnT levels [21–24], hs-cTnT concentrations
were adjusted (Adj-hs-cTnT) according to the estimated
glomerular filtration rate (eGFR), which was first calculated
using the age- and sex-specific adult CKD-EPI 2009 equa-
tions based on circulating creatinine [25,26]. Within-sub-
range inter-group comparisons of hs-cTnT and Adj-hs-cTnT
(Kruskal–Wallis test). Statistical significance was defined
as p < .05.
Results
Comparison of the different CRP
Over the 1.8-year period, a total of 744 biological files of
patients were collected, with concomitant values of hs-CRP
and wr-CRP
5 mg/L, along with hs-cTnT and creatinine
values. Median (IQR) for age was 64.6 (50.7–77.6) years,
and sex ratio (M/W, %) was 55/45. Overall, hs-CRP and wr-
 SCANDINAVIAN JOURNAL OF CLINICAL AND LABORATORY INVESTIGATION 3

(A) 5 N = 744 (mg/L) based on the following Deming regression equation:

t = 0.933 Adj-wr-CRP=–0.1737 þ 1.034  hs-CRP. The standard error
y = – 0.1737 + 1.034 x and confidence intervals (95% CI) were 0.0089 mg/L
4
(–0.1912 to –0.1561) for the intercept, and 0.0059 mg/L
wr-CRP (mg/L)

3
(1.0223–1.0457) for the slope. Therefore, on all the data, the
intercept was significantly different from 0 and the slope
was significantly different from 1, giving lower wr-CRP
2
results compared to hs-CRP (mean bias: –0.11 mg/L, Figure
1(B), or –12.7%, Figure 1(C)). In order to known whether a
1
possible linearity bias at low level of CRP may affect the
results, we performed a dilution test of serum hs-CRP
0 (CobasV c501) and plasma wr-CRP (ModularV P800), pre-
R R

0 1 2 3 4 5 6 pared from pools of patients’ samples, adjusted at 5 mg/L,

hs-CRP (mg/L) and serially diluted at 1/2, 1/4, 1/8, 1/16, and 1/32 (i.e. up to
(B) 0.16 mg/L) in sodium chloride 0.9%. Over the dilution
0.8 range, results were linear for hs-CRP and wr-CRP
0.6
hs-CRP - wr-CRP (mg/L)

+1.96 SD (Supplementary Figure S1(A)), which were strongly corre-

0.4 0.45 lated (wr-CRP ¼ –0.130 þ 0.978  hs-CRP, R2 ¼ 0.999;
0.2 Mean Supplementary Figure S1(B)), and with wr-CRP concentra-
0.0 0.11 tions lower than those of hs-CRP, as observed on overall
-0.2 -1.96 SD patients’ results.
-0.23
-0.4
-0.6 Agreement between the different CRP
-0.8 As shown in Table 1(A), agreement between wr-CRP and
-1.0 hs-CRP was good, with 1.1 and 7.0% of wr-CRP values clas-
0 1 2 3 4 5 6
sified to a higher and lower LGI-CVR range, respectively
(j ¼ 0.874). Lowering the wr-CRP cutoff from 1.0 to 0.9 mg/
hs-CRP (mg/L)
(C)
L decreased the proportion of misclassified results compared
100 to hs-CRP (6.3%), improving the agreement (j ¼ 0.901;
(hs-CRP - wr-CRP) / hs-CRP %

80
60
40
20
0
-20
-40
0 1 2 3 4 5
hs-CRP (mg/L)
Figure 1. Deming regression (A) and Bland–Altman graphs (B,C) comparing cir-
culating wr-CRP and hs-CRP at concentrations
5 mg/L. Serum hs-CRP (assayed
on a CobasV c501 analyzer) and lithium-heparin plasma wr-CRP (assayed on
R (3.00–12.3) ng/L; p < .001); this difference; however, disap-
ModularV P800 analyzers) concentrations were strongly correlated (A). However,
R
wr-CRP results were lower than hs-CRP results, with a mean bias –0.11 mg/L (B)
or –12.7% (C). The horizontal short-dashed line of Bland–Altman graphs indi-
cates the mean bias, the horizontal long-dashed lines indicate two standard
deviations (SD) above and below the mean bias, and the inclined dashed line
indicates the regression line of differences. hs-CRP: high-sensitive C-reactive
protein; wr-CRP: wide-range C-reactive protein; s: Kendall’s tau correlation
coefficient.
CRP concentrations were strongly correlated (s ¼ 0.933;
p < .001; Figure 1(A)). The strong correlation with hs-CRP
was still observed considering wr-CRP values <1 mg/L
(n ¼ 283; s ¼ 0.823; p < .001). Considering all patients, the
wr-CRP results (mg/L) were adjusted on those of hs-CRP
Table 1(B)). The regression-based adjustment of wr-CRP
weakly improved the agreement (7.7% misclassification;
j ¼ 0.881; Table 1(C)). However, lowering the Adj-wr-CRP
+1.96 SD
48.2 cutoff of the moderate LGI-CVR range from 1.0 to 0.9 mg/L
resulted in an almost perfect agreement (3.2% misclassifica-
Mean tion; j ¼ 0.950; Table 1(D)).
12.7
-1.96 SD Troponin concentrations depending on CRPs-based risk
-22.9 classification
The regression equation for eGFR-based adjustment of hs-
6 cTnT was: Adj-hs-cTnT ¼ 10(1.7750–0.01043  eGFR). On the
whole group, hs-cTnT concentrations were significantly
higher in men than in women [8.69 (7.77–9.48) versus 6.12
peared after CKD-EPI-based eGFR adjustment [8.09
(5.91–12.0) versus 7.81 (5.70–11.2) ng/L; p ¼ .143]. Hs-CRP,
wr-CRP, and Adj-wr-CRP were positively associated with
hs-cTnT (Supplementary Figure S2(A–C)), and with Adj-hs-
cTnT (Supplementary Figure S2(D–F)) (p < .001 for all).
Regarding within-subrange comparisons (Supplementary
Figure S3), hs-cTnT medians did not differ whether the
LGI-CVR classification was based on hs-CRP, wr-CRP, or
Adj-wr-CRP [respective hs-cTnT medians: low risk: 5.78, 6.
05, 6.12 ng/L (p ¼ .954); moderate risk: 8.37, 8.62, 8.56 ng/L
(p ¼ .900); high risk: 11.1, 9.99, 11.1 ng/L (p ¼ .947)], nor did
Adj-hs-cTnT medians [respective Adj-hs-cTnT medians: low
risk: 7.36, 7.42, 7.31 ng/L (p ¼ .939); moderate risk: 8.38, 8.
4 D. MONNERET ET AL.

Table 1. Classification of patients into groups of low-grade inflammation and cardiovascular risk according to the hs-CRP and wr-CRP assays.
Hs-CRP (mg/L)
<1 (1–3) (3–5)
Low risk Moderate risk High risk Total
A. Before wr-CRP adjustment
Wr-CRP (mg/L) <1 Low risk 248 34 1 283
(1–3) Moderate risk 4 290 17 311
(3–5) High risk 0 4 146 150
Total 252 328 164 744
B. Before wr-CRP adjustment but lowering the cutoff from 1.0 to 0.9 mg/L
Wr-CRP (mg/L) <0.9 Low risk 236 9 1 246
(0.9–3) Moderate risk 16 315 17 348
(3–5) High risk 0 4 146 150
Total 252 328 164 744
C. After wr-CRP adjustment
Adj-wr-CRP (mg/L) <1 Low risk 252 48 1 301
(1–3) Moderate risk 0 280 8 288
(3–5) High risk 0 0 155 155
Total 252 328 164 744
D. After wr-CRP adjustment and lowering the cutoff from 1.0 to 0.9 mg/L
Adj-wr-CRP (mg/L) <0.9 Low risk 252 15 1 268
(0.9–3) Moderate risk 0 313 8 321
(3–5) High risk 0 0 155 155
Total 252 328 164 744
Classification of patients into risk groups according to the hs-CRP and wr-CRP values. Light grey cells indicate patients who were reclassified toward a lower risk
group, dark grey cells indicate participants who were reclassified toward a higher risk group and black cells indicate those whose risk classification didn’t change.
A: The classification based on initial results led to 8.1% of wr-CRP results misclassified compared to hs-CRP (j ¼ 0.874, very good agreement). B: Lowering the
wr-CRP cutoff from 1.0 to 0.9 mg/L decreased the proportion of misclassified results compared to hs-CRP (6.3%), improving the agreement (j ¼ 0.901). C: The
classification according to the initial hs-CRP values and the wr-CRP values adjusted based on Deming’s regression equation (Adj-wr-CRP¼–0.1737 þ 1.034  hs-
CRP) led to 7.7% reclassified values (j ¼ 0.881). D: The classification according to the initial hs-CRP results and lowering the Adj-wr-CRP cutoff of the moderate
range from 1.0 to 0.9 mg/L led to 3.2% reclassified data (j ¼ 0.950, almost perfect agreement).

Table 2. Comparisons of wr-CRP, Adj-wr-CRP and hs-CRP concentrations according to the tertiles of hs-cTnT (A)(A) and eGFR-adjusted hs-
cTnT (B).
Tertile 1 Tertile 2 Tertile 3
(n ¼ 248) (n ¼ 248) (n ¼ 248)
Median (10–90 P) Median (10–90 P) Median (10–90 P) p
(A) Hs-cTnT (ng/L) 3.00 (3.00–4.55) 7.52 (5.30–10.7) 19.3 (11.9–47.5) <.001
Wr-CRP (mg/L) 1.00 (0.30–3.57) 1.30 (0.40–3.77) 1.90 (0.40–3.90) <.001
Adj-wr-CRP (mg/L) 0.97 (0.28–3.53) 1.28 (0.36–3.57) 1.87 (0.44–3.90) <.001
Hs-CRP (mg/L) 1.11 (0.44–3.58) 1.41 (0.51–3.62) 1.98 (0.59–3.94) <.001
(B) Adj-hs-cTnT (ng/L) 5.15 (3.49–6.31) 7.93 (6.75–9.40) 13.8 (10.6–24.0) <.001
Wr-CRP (mg/L) 1.10 (0.30–3.57) 1.40 (0.30–3.90) 1.70 (0.50–3.80) <.001
Adj-wr-CRP (mg/L) 1.05 (0.29–3.59) 1.38 (0.30–3.87) 1.66 (0.48–3.75) <.001
Hs-CRP (mg/L) 1.18 (0.45–3.64) 1.51 (0.46–3.91) 1.78 (0.63–3.79) <.001
Categorizing hs-cTnT (A) and Adj-hs-cTnT results into tertiles (B), the medians of wr-CRP, Adj-wr-CRP, and hs-CRP increased significantly with the
troponin-based cardiac severity (Kruskal–Wallis rank test).
Adj-hs-cTnT: hs-cTnT adjusted according to the CKD-EPI-based estimated glomerular filtration rate; Adj-wr-CRP: wr-CRP adjusted based on hs-CRP;
CRP: C-reactive protein; hs-cTnT: high-sensitive cardiac troponin T; wr-CRP: wide-range CRP.

44, 8.58 ng/L (p ¼ .880); high risk: 8.41, 8.10, 8.47 ng/L
(p ¼ .922)].
Categorizing hs-cTnT and Adj-hs-cTnT results into ter-
tiles, the medians of wr-CRP, Adj-wr-CRP, and hs-CRP
increased significantly with the troponin-based cardiac
severity (p < .001; Table 2).
were substantially lower than those of hs-CRP (mean bias
–0.11 mg/L, SD ±0.17). Our findings are close to those of
Ziv-Baran et al. [18], who showed a strong correlation
between wr-CRP assayed on ADVIAV 2400 and hs-CRP
R
assayed on BNIIV (r ¼ 0.98, p < .001), with lower wr-CRP
R
values (mean bias –0.15 mg/L, SD ±0.29). Our results are
also close to those of Arbel et al. [27], which displayed a
strong correlation between wr-CRP assayed on ADVIAV
R

Discussion
1650 and hs-CRP measured on BNIIV, with a mean bias at
R

Our study confirmed that, using immunoturbidimetry on
Roche analyzers, wr-CRP remains strongly correlated to hs-
CRP at low-grade inflammation levels, even below the
threshold cutoff of moderate cardiovascular risk range at
1 mg/L. However, despite a good linearity, wr-CRP results
–0.17 mg/L for wr-CRP values below 3.76 mg/L. Several dis-
tinctive characteristics in Roche immunotubidimetric meth-
ods could potentially explain our mean bias of –0.11 mg/L
observed for wr-CRP values. Differences include the reagent
volumes (wr-CRP: R1 ¼ 250 mL, R2 ¼ 120 mL; hs-CRP:
 SCANDINAVIAN JOURNAL OF CLINICAL AND LABORATORY INVESTIGATION
R1 ¼ 82 mL, R2 ¼ 28 mL), which are not in the same propor-
tions as sample volumes (wr-CRP: 3 mL; hs-CRP: 6 mL), or
the type of calibration (wr-CRP: end-point measurement,
spline calibration; hs-CRP: kinetic measurement, linear cali-
bration), or even the main wavelength (wr-CRP: 570 nm; hs-
CRP: 546 nm) (data from French package insert: wr-CRP:
version 9.0, 2015; hs-CRP: version 10.0, 2017).
Using hs-CRP as a reference and the same LGI-CVR
ranges, Ziv-Baran et al. [18] found 14.3% subjects reclassi-
fied with wr-CRP (substantial agreement j ¼ 0.766) [18],
whereas we found a reclassification proportion of 8.1% with
a good agreement (j ¼ 0.874). This difference in reclassifica-
tion rates may be due to their high number of subjects
(15,780 versus 744 for us), thus resulting in a higher disper-
sion of values (mean bias SD: ±0.29 versus ±0.17 mg/L for
us), which could furthermore explain why, unlike us, their
agreement was substantially improved after linear adjust-
ment (reclassification rate change: 14.3–8.4% versus
8.1–7.7% for us; j ¼ 0.766–0.861 versus 0.874–0.881 for us).
As expected given the similar mean biases, both our studies
showed an almost perfect agreement when lowering the cut-
off of Adj-wr-CRP from 1 to 0.9 mg/L (Ziv-Baran et al. [18]:
7.6% reclassification rate with a lower cutoff versus 3.2% for
present study; j ¼ 0.874 versus 0.950, respectively).
However, the regression-based adjustment of wr-CRP may
be fastidious for laboratories since it requires a comparison
with hs-CRP and a new complete validation of method. By
simply lowering the LGI-CVR cutoff of wr-CRP from 1 to
0.9 mg/L–thus without regression-based adjustment—the
agreement with hs-CRP remained very good (6.3% misclas-
sified values; j ¼ 0.901), which could be an acceptable com-
promise from a diagnostic standpoint.
However, lowering the cutoff in such a way would
require to know the uncertainty of measurement (UM) at
0.9 mg/L CRP, which should be below 10% to differentiate
such a result from 1 mg/L at an individual level. The UM
may be assessed through different approaches, among which
(i) that based on combined data from internal quality con-
trol (IQC) and external quality assessment (EQA), and (ii)
that based on combined data from IQC and calibration
uncertainty [28,29]. Although the first one is the most con-
venient for medical laboratories, there is no available data
from EQA program allowing the estimation of mean biases
between methods for CRP concentrations near 1 mg/L.
Regarding the second approach, even though calibrator
uncertainties are available, they are unusable for determining
the UM in our study since they correspond to concentra-
tions much higher than 0.9 to 1 mg/L CRP. Indeed, the con-
centration and uncertainty of calibrator (C.f.a.s proteins, cat.
no. 11355279216; reference material: CRM470) is
81.1 ± 0.96 mg/L (CV 1.18%) for wr-CRP, and
89.4 ± 0.55 mg/L (CV 0.62%) for hs-CRP (certificates of
traceability and uncertainty from Roche Diagnostics; ver-
sions November 2017). Whatever it be, the imprecision
uncertainty is part of both approaches; in our study, it
ranged from nearly 4.6 to 11.1% for serum or plasma pools
of wr-CRP or hs-CRP at 0.8–1.1 mg/L, depending on the
analyzers (Supplementary Table S1). Given such impreci-
sions, lowering the wr-CRP cutoff from 1.0 to 0.9 mg/L is
5
debatable, from an analytical standpoint. Nevertheless, this
possible cutoff adaptation remains to be validated on the
basis of imprecisions determined with much more IQC val-
ues (i.e. daily assayed over several months that could lower
the imprecision), and with appropriate EQA or calibrator
concentrations (i.e. 0.9–1 mg/L).
From a cardiac standpoint, a hs-cTnT concentration
above nearly 7 ng/L and within the 99th percentile of the
general population may be predictive of all-cause mortality
and cardiovascular mortality [30–34]. In our study,
unadjusted hs-cTnT increased significantly with hs-CRP, wr-
CRP and Adj-wr-CRP, which is consistent with other stud-
ies, displaying a proportional relationship between CRP
above 1 mg/L and hs-cTnT above 7 ng/L [31–35].
However, considering the within-LGI-CVR subrange, the hs-
cTnT levels did not differ in our study, regardless of the
type of CRP used for risk classification, reinforcing the
assumption that replacing the hs-CRP by the wr-CRP assay
would not change the cardiovascular risk interpretation.
Supporting this assumption, strong correlations between wr-
CRP and hs-CRP at low concentrations have been shown in
patients with stable/unstable angina, myocardial infarction
[27], or with acute ischemic stroke/transient ischemic
attack [36].
Conclusion
Overall, our results support those of previous studies, show-
ing a strong correlation between wr-CRP and hs-CRP at
concentrations
5 mg/L. Our study also assumes that the
LGI-CVR under-estimation using initial wr-CRP values
could be corrected after a linear regression-based adjust-
ment, and by lowering the cutoff of the moderate risk range
from 1.0 to 0.9 mg/L. However, further studies are needed to
determine accurately the uncertainty measurement at such
low concentrations. If it turns out to be less than ten
percent, laboratories equipped with Roche analyzers could
reasonably replace the hs-CRP assay by the less expensive
wr-CRP for routine evaluation of low-grade inflammation-
associated cardiovascular risk.
Acknowledgements
The authors are grateful to Vincent Fitzpatrick for the
English rereading.
Disclosure statement
No potential conflict of interest was reported by the authors.
References
[1] Ross R. Atherosclerosis-an inflammatory disease. N Engl J Med.
1999;340:115–126.
[2] Taube A, Schlich R, Sell H, et al. Inflammation and metabolic
dysfunction: links to cardiovascular diseases. Am J Physiol
Heart Circ Physiol. 2012;302:H2148–H2165.
[3] Biasucci LM. CDC; AHA. CDC/AHA workshop on markers of
inflammation and cardiovascular disease: application to clinical
and public health practice: clinical use of inflammatory markers
6 D. MONNERET ET AL.
in patients with cardiovascular diseases: a background paper.
Circulation. 2004;110:e560–e567.
[4] Ridker PM, Cushman M, Stampfer MJ, et al. Inflammation,
aspirin, and the risk of cardiovascular disease in apparently
healthy men. N Engl J Med. 1997;336:973–979.
[5] Rifai N, Ridker PM. High-sensitivity C-reactive protein: a novel
and promising marker of coronary heart disease. Clin Chem.
2001;47:403–411.
[6] Emerging Risk Factors Collaboration, Kaptoge S, Di
Angelantonio E, Lowe G, et al. C-reactive protein concentration
and risk of coronary heart disease, stroke, and mortality: an
individual participant meta-analysis. Lancet. 2010;375:132–140.
[7] Harutyunyan MJ, Mathiasen AB, Winkel P, et al. High-sensitiv-
ity C-reactive protein and N-terminal pro-B-type natriuretic
peptide in patients with stable coronary artery disease: a prog-
nostic study within the CLARICOR trial. Scand J Clin Lab
Invest. 2011;71:52–62.
[8] Zacho J, Tybjaerg-Hansen A, Nordestgaard BG. C-reactive pro-
tein, genetically elevated levels and risk of ischemic heart and
cerebrovascular disease. Scand J Clin Lab Invest. 2009;69:
442–446.
[9] Ridker PM. A test in context: high-sensitivity C-reactive pro-
tein. J Am Coll Cardiol. 2016;67:712–723.
[10] Ridker PM, Rifai N, Pfeffer MA, et al. Inflammation, pravasta-
tin, and the risk of coronary events after myocardial infarction
in patients with average cholesterol levels. Cholesterol and
Recurrent Events (CARE) Investigators. Circulation.
1998;98:839–844.
[11] Ridker PM, Rifai N, Clearfield M, et al. Measurement of C-
reactive protein for the targeting of statin therapy in the pri-
mary prevention of acute coronary events. N Engl J Med.
2001;344:1959–1965.
[12] Cushman M, Arnold AM, Psaty BM, et al. C-reactive protein
and the 10-year incidence of coronary heart disease in older
men and women: the cardiovascular health study. Circulation.
2005;112:25–31.
[13] Laaksonen DE, Niskanen L, Nyyss€ onen K, et al. C-reactive pro-
tein in the prediction of cardiovascular and overall mortality in
middle-aged men: a population-based cohort study. Eur Heart
J. 2005;26:1783–1789.
[14] Rogowski O, Vered Y, Shapira I, et al. Introducing the wide
range C-reactive protein (wr-CRP) into clinical use for the
detection of microinflammation. Clin Chim Acta. 2005;358:
151–158.
[15] Maharshak N, Arbel Y, Gal-Oz A, et al. Comparative analysis
of Bayer wide-range C-reactive protein (wr-CRP) and the
Dade-Behring high sensitivity C-reactive protein (hs-CRP) in
patients with inflammatory bowel disease. J Dig Dis.
2008;9:140–143.
[16] Grootendorst DC, de Jager DJ, Brandenburg VM, et al.
Excellent agreement between C-reactive protein measurement
methods in end-stage renal disease patients–no additional
power for mortality prediction with high-sensitivity CRP.
Nephrol Dial Transplant. 2007;22:3277–3284.
[17] Kanaan N, Goffin E, Maisin D, et al. CRP measurement: does
the assay matter in hemodialysis patients? Clin Nephrol.
2008;70:503–507.
[18] Ziv-Baran T, Shenhar-Tsarfaty S, Etz-Hadar I, et al. The ability
of the wide range CRP assay to classify individuals with low
grade inflammation into cardiovascular risk groups. Clin Chim
Acta. 2017;471:185–190.
[19] Kendall MG. (1976). Rank correlation methods. 4th edn.
London: Charles Griffin.
[20] Davison AC, Hinkley DV. (1997). Bootstrap methods and their
application. Cambridge: Cambridge University Press.
[21] Olivieri F, Galeazzi R, Giavarina D, et al. Aged-related increase
of high sensitive Troponin T and its implication in acute myo-
cardial infarction diagnosis of elderly patients. Mech Ageing
Dev. 2012;133:300–305.
[22] Kuster N, Monnier K, Baptista G, et al. Estimation of age- and
comorbidities-adjusted percentiles of high-sensitivity cardiac
troponin T levels in the elderly. Clin Chem Lab Med.
2015;53:691–698.
[23] deFilippi C, Seliger SL, Kelley W, et al. Interpreting cardiac
troponin results from high-sensitivity assays in chronic kidney
disease without acute coronary syndrome. Clin Chem.
2012;58:1342–1351.
[24] Chung JZ, Dallas Jones GR. Effect of renal function on serum
cardiac troponin T-population and individual effects. Clin
Biochem. 2015;48:807–810.
[25] Kidney Disease: Improving Global Outcomes (KDIGO) CKD
Work Group. KDIGO 2012 clinical practice guideline for the
evaluation and management of chronic kidney disease. Kidney
Inter Suppl. 2013;3:1–150.
[26] Levey AS, Becker C, Inker LA. Glomerular filtration rate and
albuminuria for detection and staging of acute and chronic kid-
ney disease in adults: a systematic review. JAMA. 2015;313:
837–846.
[27] Arbel Y, Eros Y, Rogowski O, et al. Comparison of values of
wide-range C-reactive protein to high-sensitivity C-reactive pro-
tein in patients undergoing coronary angiography. Am J
Cardiol. 2007;99:1504–1506.
[28] White GH, Farrance I, AACB Uncertainty of Measurement
Working Group. Uncertainty of measurement in quantitative
medical testing: a laboratory implementation guide. Clin
Biochem Rev. 2004;25:S1–S24.
[29] Matar G, Poggi B, Meley R, et al. Uncertainty in measurement
for 43 biochemistry, immunoassay, and hemostasis routine ana-
lytes evaluated by a method using only external quality assess-
ment data. Clin Chem Lab Med. 2015;53:1725–1736.
[30] van der Linden N, Klinkenberg LJ, Bekers O, et al. Prognostic
value of basal high-sensitive cardiac troponin levels on mortal-
ity in the general population: a meta-analysis. Medicine
(Baltimore). 2016;95:e5703.
[31] de Lemos JA, Drazner MH, Omland T, et al. Association of
troponin T detected with a highly sensitive assay and cardiac
structure and mortality risk in the general population. JAMA.
2010;304:2503–2512.
[32] Hussein AA, Gottdiener JS, Bartz TM, et al. Cardiomyocyte
injury assessed by a highly sensitive troponin assay and sudden
cardiac death in the community: the cardiovascular health
study. J Am Coll Cardiol. 2013;62:2112–2120.
[33] deFilippi CR, de Lemos JA, Christenson RH, et al. Association
of serial measures of cardiac troponin T using a sensitive assay
with incident heart failure and cardiovascular mortality in older
adults. JAMA. 2010;304:2494–2502.
[34] Eggers KM, Al-Shakarchi J, Berglund L, et al. High-sensitive
cardiac troponin T and its relations to cardiovascular risk fac-
tors, morbidity, and mortality in elderly men. Am Heart J.
2013;166:541–548.
[35] Oluleye OW, Folsom AR, Nambi V, ARIC Study Investigators,
et al. Troponin T, B-type natriuretic peptide, C-reactive protein,
and cause-specific mortality. Ann Epidemiol. 2013;23:66–73.
[36] Shenhar-Tsarfaty S, Ben Assayag E, Bova I, et al. Wide-range
C-reactive protein efficacy in acute ischemic stroke patients.
Acta Neurol Scand. 2006;114:29–32.