Ten-year standardization of lipids and high-sensitivity C-reactive protein in a randomized controlled trial to assess the effects of statins on secondary stroke prevention: Japan Statin Treatment Against Recurrent Stroke

Abstract

Background

The Japan Statin Treatment Against Recurrent Stroke (J-STARS) is a clinical trial that administered pravastatin for secondary stroke prevention. Lipid concentrations are a well-established risk factor for cerebrovascular diseases. Elevated high-sensitivity C-reactive protein (hs-CRP) indicates a high risk of inflammatory reactions. In clinical trials, internationally approved standardization is essential for obtaining study results that are comparable with those from overseas. Therefore, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG) and hs-CRP were standardized throughout a 10-year study period.

Methods

J-STARS specified a single clinical laboratory for blood analyses. Four lipids were evaluated by calculating the total error: accuracy (absolute mean %bias vs. reference value) + precision (1.96 among-run coefficient of variation [CV, %]). Accuracy for hs-CRP was ensured using a calibrator traceable to the international plasma protein reference material and precision was evaluated by CV.

Results

Average total errors (standard deviation, %) throughout the study period were as follows: TC 1.35% (0.290%), HDL-C 2.45% (1.087%), LDL-C 2.65% (0.956%) and TG 3.70% (0.559%). Four lipids met the performance criteria of the US Centers for Disease Control and Prevention (CDC). The precision of hs-CRP was 3.28% (0.627%), which met the performance criterion established by the American Heart Association/CDC.

Conclusions

Based on standardization, the results of J-STARS appear to be comparable with those of similar intervention-based clinical studies on statins overseas. These study results will contribute to the establishment of preventive measures against recurrent stroke in Japanese patients. J-STARS is registered in ClinicalTrials.gov under NCT00221104.

Keywords

J-STARS standardization lipids hs-CRP CDC CRMLN

Introduction

A large number of intervention studies have been performed on the initial and secondary prevention of arteriosclerotic cardiovascular events by statins (3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors).^1–5 The Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL)⁶ study was the first clinical trial to investigate the secondary preventive effects of statins on cerebrovascular diseases, such as stroke. In SPARCL, atorvastatin was administered at a high dose (80 mg/day) to patients with a medical history of stroke. Atorvastatin is a potent lipid-lowering statin that is used to prevent recurrent stroke. The effects of atorvastatin were investigated by comparisons between statin and placebo groups in a double-blind study. In the SPARCL study, the statin dose was markedly higher than that approved in Japan; therefore, the findings obtained in the SPARCL study may not be applicable to Japanese patients.

The Japan Statin Treatment Against Recurrent Stroke (J-STARS)⁷ (this study was registered in ClinicalTrials.gov under NCT00221104) is an investigator-initiated 10-year clinical trial that involved disease subtype at registration and the administration of pravastatin at a low dose (10 mg/day) to patients aged between 45 and 80 years old with a medical history of non-cardioembolic ischaemic stroke such as lacunar and atherothrombotic infarctions. In this study, the usefulness of statins for the secondary prevention of non-cardioembolic ischaemic stroke in Japanese patients was investigated, and the significance of statins for each disease subtype was also examined.

Serum lipid concentrations have been identified as a risk factor for ischaemic cardiovascular events^1–5 and elevated high-sensitivity C-reactive protein (hs-CRP) indicates a high risk of inflammatory reactions.^8–10 In clinical trials or epidemiological studies, internationally approved standardization is essential for obtaining high-quality study results that are comparable with those from overseas. Therefore, the standardization of four lipids including total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and triglycerides (TG) was performed and hs-CRP concentrations were investigated throughout the 10-year study period. The clarification of measurement performance for these items through the standardization preceded the analytical process of J-STARS.^7,11

Materials and methods

A total of 123 regional core hospitals throughout Japan collected blood samples from patients. J-STARS specified a single clinical laboratory in Tokyo, Japan, named SRL, for blood analyses throughout the study period. Eight blood chemistry markers including TC, HDL-C, LDL-C, TG and hs-CRP were analysed. SRL used enzymatic methods to measure TC and TG concentrations and homogeneous methods to assay HDL-C and LDL-C concentrations. The method of hs-CRP is described in the section of hs-CRP standardization. As an internal quality control, in-house pooled serum prepared at SRL was used in lipid and hs-CRP assays for precision control.

The lipid and hs-CRP performance criteria that are internationally accepted and used in large-scale clinical trials and epidemiological studies are shown in Table 1 (TG criteria are tentative). These criteria were established in order to conform to those of the US National Cholesterol Education Program and the American Heart Association (AHA). Accuracy was defined in terms of mean %bias vs. reference value and precision in terms of the among-run coefficient of variation (CV, %), and total error is calculated as follows: accuracy (absolute mean %bias vs. reference value) + precision (1.96 among-run CV [%]). The among-run CV of a single quality control material was calculated using values measured from 20 separate runs at SRL.

Table 1.

Criteria for the measurement performance of TC, HDL-C, LDL-C, TG and hs-CRP.

	CDC/CRMLN criteria				AHA/CDC criterion
Performance	TC	HDL-C	LDL-C	TG	hs-CRP
Accuracy (mean %bias vs. reference value)	≤3	≤5	≤4	≤5 (Tentative)
Precision (among-run CV, %)	≤3	≤4	≤4	≤5 (Tentative)	<10
Total error (%)	≤8.9	≤13	≤12	≤15 (Tentative)

CDC: Centers for Disease Control and Prevention; CRMLN: Cholesterol Reference Method Laboratory Network; AHA: American Heart Association; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; TG: triglycerides; hs-CRP: high-sensitivity C-reactive protein; CV: coefficient of variation (%).

Note: TG criteria are tentative.

Total error = accuracy (absolute mean %bias vs. reference value) + precision (1.96 among-run CV, %).

TC standardization

Although SRL is a clinical laboratory, they have participated in the TC standardization programme by the US Centers for Disease Control and Prevention (CDC) for reagent manufacturers in order to obtain high-quality reliability in their assayed values. In this programme (Total Cholesterol Certification Protocol for Manufacturers – Revised, October 2004, http://www.cdc.gov/labstandards/pdf/crmln/RevisedTCprotocolOct04.pdf), at least 40 fresh individual human serum samples were collected at reagent manufacturers participating in the programme. As specified by the standardization programme, 20% of individual human serum specimens are expected to be within 3.103 to 4.655 mmol/L, 30% within 4.681 to 5.689 mmol/L, 30% within 5.715 to 6.724 mmol/L and 20% within 6.749 to 10.344 mmol/L for TC concentrations. They measured these split samples in duplicate in 20 separate runs over a one-month period using a routine enzymatic method at SRL. In the lipid reference laboratory at the National Cerebral and Cardiovascular Center (NCVC) at Osaka, Japan, which is one of the members of the Cholesterol Reference Method Laboratory Network (CRMLN),¹² each sample was measured in duplicate using the Abell-Kendall reference measurement procedure (RMP) for cholesterol.^13,14 The TC reference value serving as an accuracy index was measured using the RMP. At SRL, the among-run CV of a single quality control serum value measured from 20 separate runs as an internal precision control was assessed and used as a precision index. The reference value and the assayed value were both input into comparative evaluation software developed by the CDC and statistically analysed. The results obtained were sent to the CDC and evaluated based on the performance criteria (Table 1). The performance of SRL (accuracy, precision and total error) for TC was then obtained and shown year-wise in Table 2.

Table 2.

Measurement performance of TC, HDL-C, LDL-C, TG and hs-CRP assayed at SRL between 2004 and 2014.

Performance of lipids and hs-CRP	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	Average % (SD, %) between 2004 and 2014
1. TC accuracy: mean %bias vs. reference value	−0.1	−0.1	−0.9	−0.9	−0.3	−0.5	0.1	0.1	−0.6	−0.6	−0.6	−0.40 (0.346)
TC precision: among-run CV (%)	0.6	0.6	0.4	0.4	0.5	0.4	0.3	0.3	0.4	0.4	0.4	0.43 (0.096)
TC total error (%)	1.4	1.4	1.7	1.7	1.3	1.3	0.8	0.8	1.5	1.5	1.5	1.35 (0.290)
2. HDL-C accuracy: mean %bias vs. reference value	1.0	1.2	1.2	1.2	−1.0	0.0	0.0	1.2	0.3	0.3	0.3	0.52 (0.679)
HDL-C precision: among-run CV (%)	1.7	1.1	1.1	1.1	1.0	0.7	0.7	0.7	0.5	0.5	0.5	0.87 (0.352)
HDL-C total error (%)	4.4	3.4	3.4	3.4	3.0	1.4	1.4	2.6	1.3	1.3	1.3	2.45 (1.087)
3. LDL-C accuracy: mean %bias vs. reference value	−0.7	−0.7	0.3	0.3	1.7	−1.4	−1.4	3.0	2.5	2.5	2.5	0.78 (1.624)
LDL-C precision: among-run CV (%)	0.7	0.7	0.4	0.4	0.6	0.6	0.6	0.6	0.5	0.5	0.5	0.55 (0.099)
LDL-C total error (%)	2.1	2.1	1.1	1.1	2.9	2.6	2.6	4.2	3.5	3.5	3.5	2.65 (0.956)
4. TG accuracy: mean %bias vs. JMA consensus value	−0.1	−0.4	0.0	−1.0	−1.1	−1.9	−1.7	−1.3	0.7	−0.4	−0.5	−0.70 (0.743)
TG precision: among-run CV (%)	1.5	1.4	2.3	1.0	1.0	1.1	1.2	1.6	1.6	1.8	1.7	1.47 (0.374)
TG total error (%)	3.0	3.1	4.6	2.9	3.1	4.0	4.0	4.4	3.8	4.0	3.8	3.70 (0.559)
5. hs-CRP precision: among-run CV (%)	–	2.9	2.0	4.0	2.5	3.7	3.9	3.9	3.5	3.1	3.3	3.28 (0.627)

SD: Standard deviation; CV: coefficient of variation (%); JMA: Japan Medical Association; SRL: A clinical laboratory in Tokyo for blood analyses specified by J-STARS; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; TG: triglycerides; hs-CRP: high-sensitivity C-reactive protein.

Note: TG criteria are tentative. Total error = accuracy (absolute mean %bias vs. reference value) + precision (1.96 among-run CV, %).

HDL-C standardization

Although SRL is a clinical laboratory, they have participated in the CDC HDL-C standardization programme for reagent manufacturers. In this programme (HDL Cholesterol Certification Protocol for Manufacturers, November 2002, http://www.cdc.gov/labstandards/crmln_manufacturers.html#HDL), at least 40 fresh individual human serum samples were collected at reagent manufacturers. Samples were prepared to include at least five samples each within concentration range of 0.517–0.750, 0.776–1.009, 1.034–1.267, 1.293–1.526 and 1.552–1.784 mmol/L for HDL-C concentrations. They measured these samples in duplicate in five runs over a one-month period using a routine homogeneous method at SRL. The lipid reference laboratory at Osaka measured each sample in duplicate using the HDL-C reference method for CRMLN, the designated comparison method (DCM).^15,16 Specimens with a TG concentration of >2.258 mmol/L were excluded from measurements by the DCM¹⁵ because elevated TG concentrations interfere with the precipitation reactions. The reference value serving as an accuracy index was measured by HDL-C DCM. At SRL, the among-run CV of a single quality control serum value measured from 20 separate runs as an internal precision control was assessed and used as a precision index. The reference value and assayed value were both input into CDC comparative evaluation software and statistically analysed. The results obtained were sent to the CDC and evaluated based on the assessment criteria (Table 1). The performance of SRL for HDL-C was then obtained (Table 2).

LDL-C standardization

Although SRL is a clinical laboratory, they have participated in the CDC LDL-C standardization programme for reagent manufacturers. In this programme (LDL Cholesterol Certification Protocol for Manufacturers, June 2006, http://www.cdc.gov/labstandards/pdf/crmln/MFRLDLJune2006final.pdf), at least 40 fresh individual human serum samples were collected at reagent manufacturers. As specified by the standardization programme, 20% of individual human serum samples are expected to be <2.586 mmol/L, 30% within 2.586 to 3.362 mmol/L, 30% within 3.388 to 4.138 mmol/L and 20% within 4.163 to 10.344 mmol/L for LDL-C concentrations. They measured these materials in duplicate in five runs over a one-month period using a routine homogeneous method at SRL. The lipid reference laboratory measured each sample in duplicate using the LDL-C RMP, a beta-quantification method using ultracentrifugation.¹⁷ The reference value serving as an accuracy index was assessed by the LDL-C RMP. At SRL, the among-run CV of a single quality control serum value measured from 20 separate runs as an internal precision control was assessed and used as a precision index. The reference value and measured value were both input into CDC comparative evaluation software and statistically analysed. The results obtained were sent to the CDC and evaluated based on the performance criteria (Table 1). The performance of SRL for LDL-C was then obtained (Table 2).

TG standardization

The external quality assessment of the clinical laboratories programme survey by the Japan Medical Association (JMA) is performed every year in October in Japan. In the JMA survey, TG samples with three concentrations were distributed to participating laboratories. They measured net TG (total glycerides−free glycerol) using a routine enzymatic method with an automatic free glycerol cancellation reagent,¹⁸ which is commonly used in Japan. They were evaluated using a four-step (A, B, C and D) relative evaluation system uniquely established by the JMA representing the deviation from the adjusted mean net TG value obtained after removing statistical outliers from the measurement results of all participating laboratories. The four-step relative evaluation system is defined as follows: A ≤ adjusted mean net TG ± 1 standard deviation (SD) of all participating laboratories after removing statistical outliers; B ≤ adjusted mean net TG ± 2 SD; C ≤ adjusted mean net TG ± 3 SD and D > adjusted mean net TG ± 3 SD. Regarding TG, the adjusted mean from approximately 2700 clinical laboratories was regarded as a consensus value instead of a reference value as an accuracy index by the RMP. Precision as among-run CV was evaluated from SRL internal quality control serum values measured from 20 separate runs. Total error was calculated from accuracy (absolute mean %bias vs. JMA consensus value) + precision (1.96 among-run CV [%]). The performance of SRL for TG was then obtained (Table 2). Total glycerides and free glycerol by gas chromatography-isotope dilution mass spectrometry (GC-IDMS) at Osaka have both been standardized by GC-IDMS at the CDC since April 2011.^19,20 The accuracy of net TG by GC-IDMS at Osaka in October 2013 showed −0.73% (n = 8) vs. CDC on average, which was in good agreement with that of the CDC reference value by GC-IDMS.

hs-CRP standardization

Sixty-nine hospitals participated in hs-CRP standardization in the J-STARS substudy. hs-CRP concentrations were measured at SRL. hs-CRP measurement conditions were as follows: measurement device – the BN ProSpec System of Siemens; measurement reagents – N-Latex CRP II CardioPhase hs-CRP, calibrator for instrument control and N-Rheumatology Standard SL. The accuracy of the nominal value of the standard substance for calibrations was tested based on the international plasma protein reference material (IRMM ERM-DA470), which has the highest ranking in the hs-CRP traceability system. In-house pooled serum from SRL was used for internal quality control. Precision was presented as the among-run CV of values measured for 20 separate runs in October every year. The precision performance criterion is shown in Table 1. hs-CRP concentrations were examined in a substudy (this study was registered in ClinicalTrials.gov as number NCT00361699) of J-STARS.

Statistical analysis

All statistical analyses were conducted using the protocol EP9-A from the Clinical and Laboratory Standards Institute for bias estimation²¹ and STATA12 analysis program. P-values for statistical tests were two-tailed, and the values of p < 0.05 were regarded as significant.

Results

Table 2 lists the measurement performance by routine analytical methods at SRL of accuracy, precision and total error for TC, HDL-C, LDL-C, TG and hs-CRP by year between 2004 and 2014 as well as the summarized average % (SD, %) throughout the study period. The average % (SD, %) of accuracy, precision and total error for TC for 10 years were −0.40% (0.346%), 0.43% (0.096%) and 1.35% (0.290%), respectively. Similarly, the average % of accuracy, precision and total error for HDL-C were 0.52% (0.679%), 0.87% (0.352%) and 2.45% (1.087%), respectively. The average % of accuracy, precision and total error for LDL-C were 0.78% (1.624%), 0.55% (0.099%) and 2.65% (0.956%), respectively. The average % of accuracy, precision and total error for TG were −0.70% (0.743%), 1.47% (0.374%) and 3.70% (0.559%), respectively; however, TG criteria are tentative (Table 1). The average % of precision for hs-CRP was 3.28% (0.627%). The performance of four lipids and hs-CRP over a 10-year period was stable every year within acceptable performance criteria (Table 2).

Table 3 shows a comparison of the adjusted mean net TG value by the JMA survey, assayed net TG value by SRL and target net TG value by GC-IDMS. TG samples with three concentrations provided by the JMA survey in 2013 were analysed using the GC-IDMS RMP at Osaka standardized by the CDC. The difference between the adjusted mean net TG value by the JMA survey and assayed net TG value by SRL was not significant in any survey sample at P < 0.05. Both values for each sample were almost equivalent. In the %bias (accuracy) of the adjusted mean net TG value by the JMA survey vs. the target net TG value by GC-IDMS, survey samples of level 1, level 2 and level 3 were −4.42%, −2.07% and 0.37%, respectively. The average %bias of the three levels was −2.04%. The adjusted mean net TG value by JMA was slightly lower at level 1 and level 2, but was slightly higher at level 3 than those by GC-IDMS. They all met the performance criteria by CDC/CRMLN for TG accuracy (≤5% [tentative]) (Table 1). P-values for differences between the assayed net TG value by SRL and GC-IDMS were 0.095 at level 1, 0.173 at level 2 and 0.721 at level 3. The difference in each sample was not significant at P < 0.05. In the %bias of the assayed net TG value by SRL vs. GC-IDMS, survey samples at level 1, level 2 and level 3 were −4.33%, −2.85% and −0.21%, respectively. The average %bias of three levels was −2.46%. The assayed TG values by SRL showed slightly lower at all three levels than those by GC-IDMS. However, they all met the performance criteria by CDC/CRMLN for TG accuracy (≤5% [tentative]) (Table 1).

Table 3.

Comparison of adjusted mean net TG by JMA, assayed net TG by SRL and target net TG by GC-IDMS using survey samples in 2013.

Survey sample for TG	Adjusted mean net TG value (a) by the JMA survey	Assayed net TG value (b) by SRL	Target net TG value (c) by GC-IDMS RMP	%Bias of adjusted mean net TG value (a) by JMA vs. GC-IDMS RMP (c)	P for difference (b and c) at P < 0.05	%Bias of assayed net TG value (b) by SRL vs. GC-IDMS RMP (c)
Level 1	1.060 mmol/L (n = 2766)	1.061 mmol/L (n = 2)	1.109 mmol/L (n = 8)	−4.42	0.095	−4.33
Level 2	1.514 mmol/L (n = 2764)	1.502 mmol/L (n = 2)	1.546 mmol/L (n = 8)	−2.07	0.173	−2.85
Level 3	1.908 mmol/L (n = 2764)	1.897 mmol/L (n = 2)	1.901 mmol/L (n = 8)	0.37	0.721	−0.21
				Average: −2.04		Average: −2.46

TG: triglycerides; JMA: Japan Medical Association; GC-IDMS: gas chromatography-isotope dilution mass spectrometry; RMP: reference measurement procedure; SRL: A clinical laboratory in Tokyo for blood analyses specified by J-STARS.

Note: TG criteria are tentative. The adjusted mean net TG value by the JMA survey is the net TG value (total glycerides−free glycerol) by the routine enzymatic methods of participating laboratories using automatic free glycerol cancellation reagent.

Table 4 shows the baseline profiles of cardiovascular risk factors in the J-STARS study.⁷

Table 4.

Baseline profiles of cardiovascular risk factors in J-STARS (n = 1578).

Age (years)	66.2 ± 8.5
Gender (male, %)	68.8
Height (cm)	160.3 ± 8.7
Weight (kg)	61.1 ± 10.2
Hypertension (%)	75.9
Diabetes mellitus (%)	23.3
TC (mmol/L)	5.431 ± 0.639
HDL-C (mmol/L)	1.384 ± 0.409
LDL-C (mmol/L)	3.349 ± 0.634
Triglyceride (mmol/L)	1.605 ± 0.838
Systolic blood pressure (mmHg)	137.1 ± 17.8
Diastolic blood pressure (mmHg)	79.3 ± 11.3
Concentrations of blood pressure ≥150/90 mmHg (%)	39.0
Fasting blood glucose (mmol/L)	6.528 ± 2.276
Use of antiplatelet agents (%)	90.7
Coronary artery disease (%)	5.1

Note: This table is adapted from Nagai et al.⁷ in the baseline article of J-STARS.

Data are the mean ± standard deviation (SD) of continuous variables.

TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol.

Discussion

Selection for the measurement of lipids and hs-CRP at a single clinical laboratory and the significance of standardization

A total of 123 regional hospitals throughout Japan provided blood samples from 1578 patients to a single clinical laboratory, SRL in Tokyo. Numerous epidemiological and interventional studies have been conducted with measurements of markers in a single laboratory center. This is also described in relevant studies and textbooks. If measured at multiple laboratories, the reproducibility and accuracy of assayed values may not be maintained or assured at constant quality control for a prolonged time, resulting in misjudgments in the interpretation of J-STARS results. Therefore, J-STARS adopted a measurement system in a single laboratory during the study period in order to minimize annual changes in measurement results. In clinical trials and epidemiological studies on cardiovascular diseases, lipids, such as TC, TG and HDL-C, are analysed as important risk factors. However, different measurement methods, reagents, primary standards, analytical instruments and quality control systems are selected in various laboratories, with the reproducibility and accuracy of measurements differing among them; therefore, compatibility-assured results are not always obtained. We considered ‘standardization’, in which the accuracy of measurements is standardized in order to assure compatibility in accordance with internationally accepted criteria when comparing values among studies or reviewing annual changes, to be a basic and important requirement for maintaining the homeostasis of research and performed it accordingly. Thus, the results of this study suggest the importance of internationally accepted standardization for assuring the compatibility of clinical trials at the international level.

Lipid standardization excluding TG

In large-scale clinical trials or epidemiological studies on the inhibitory effects of statins on cerebro- and cardiovascular events conducted in Western countries, risk factors are generally detected from primary and secondary preventive measures by assaying lipid concentrations and investigating the relationship between major events including cerebral haemorrhage, stroke and myocardial infarction.^1–5 Since the risk factors vary depending on ethnicity, the duration of the study, types and doses of statins, duration of statin administration and the effects of statins on cerebro- and cardiovascular events typically show various patterns. We concluded that a demonstration of the high-level reliability of study results through the standardization of blood chemistry findings was necessary for J-STARS in order to receive affirmative approval for the J-STARS study results.¹¹ The central clinical laboratory of J-STARS, SRL, participated in the same lipid standardization programmes applied to reagent manufacturers conducted by the CDC and its international network, the CRMLN. The NCVC is an international member of the CRMLN, and the RMP for lipids at the NCVC have been standardized by the CDC since July 1992. The NCVC standardized all lipids measured in SRL, except for TG.

Although the standardization of blood chemistry results is generally only briefly or not described at all in clinical study reports, the standardization period (10 years) in this study was 2–3-fold longer than that of normal clinical studies. Fewer clinical studies in Japan were standardized through the CDC. Evidence from Asia, in which the prevalence of dyslipidaemia is generally lower than that in Western populations, has been limited. In order to obtain J-STARS results that were comparable with those of similar clinical studies performed in other countries, the establishment of reference values using the RMP was performed to ensure that the measured blood chemistry values were accurate and reliable. More than 40 clinical and epidemiological studies have participated in the WHO-CDC Cooperative Cholesterol and Triglycerides Standardization Program²² and CDC-National Heart, Lung, and Blood Institute Lipid Standardization Program²³ conducted by the CDC and CRMLN. J-STARS study is also one of the participating clinical trials in the standardization conducted by the CDC and the CRMLN.

TG standardization

Since the RMP was not employed in the JMA survey, the accuracy of the adjusted means of samples reported in the JMA summary is unclear. The assayed values of each participating laboratory are evaluated based on the adjusted mean values obtained through statistical analyses of values from all participating laboratories. The adjusted mean is a consensus value instead of accuracy by the RMP. Concerns were expressed regarding whether the adjusted mean has high-level reliability for accuracy. In order to overcome this issue, TG samples of the JMA survey in 2013 were measured only once using GC-IDMS RMP at the end of the J-STARS study period and its values were compared to assess accuracy.^19,20 In the average %bias of the three samples, the adjusted mean net TG value by JMA vs. GC-IDMS indicated −2.04% (Table 3). As a result, the adjusted mean as a consensus value was investigated where available because it met the CDC accuracy performance criteria for TG (≤5% [Tentative]) (Table 1). After the end of the J-STARS study, the measurement of TG using the JMA survey samples has been continued by the GC-IDMS RMP every year in 2014, 2015 and 2016. We are observing TG accuracy in JMA survey samples with the GC-IDMS RMP every year. Meanwhile, in the average %bias of three samples, the assayed net TG value by SRL vs. GC-IDMS was −2.46% (Table 3). Although the assayed net TG values by SRL were slightly lower than the target net TG values by GC-IDMS, they all met the performance criteria by CDC/CRMLN for TG accuracy (≤5% [tentative]) (Table 1). As a result, the assayed TG values by SRL provide reliable data in the J-STARS study.

hs-CRP standardization

No criteria are currently available for assessing the precision and accuracy of hs-CRP values. Furthermore, no reliable standard material for calibrations or RMP has been established.^24,25 However, hs-CRP reference material, IRMM ERM-DA470, exists whereby values measured by several laboratories with high-level analytical techniques are evaluated as a highly reliable subject. This international reference subject is used for practical purposes as the primary reference material for the standardization of hs-CRP. Accuracy is ensured by the calibration of analytical instruments with standard serum originating from IRMM ERM-DA470. The AHA/CDC Scientific Statement⁸ proposed a CV of <10% within a concentration range of 300–10,000 μg/L for the precision performance criteria (Table 1). Therefore, the precision target was set at this criterion in J-STARS. The precision of hs-CRP by SRL throughout the study period was within the acceptable criteria established by the AHA and CDC (Table 2). hs-CRP is a marker of low-grade inflammation and its presence indicates an increased risk of vascular disorders such as cardiovascular disease, arteriosclerosis, diabetes mellitus, obesity and hypertension.^8–10 Serum lipid concentrations are also reported to be risk factors for these diseases. Therefore, lipid concentrations and hs-CRP were measured in J-STARS. The details of this substudy for hs-CRP will be reported elsewhere.

Conclusions

The results of the 10-year standardization for lipids and hs-CRP in J-STARS involving the administration of pravastatin suggest that the results of this clinical study are comparable with those of similar clinical trials performed in other countries. This also suggests the importance of the international standardization of measurements for a long period in a single laboratory in a clinical trial and markers as risk factors. Reliable results will contribute to the establishment of preventive measures against recurrent stroke in Japanese patients.

Footnotes

Acknowledgments

The authors would like to thank Hiroyasu Iso, M.D., Ph.D. at the Osaka University Graduate School of Medicine and Hubert W. Vesper, Ph.D. at the CDC for their reviews and comments. The authors also wish to thank Ms. Yukari Ichikawa for her excellent assistance in providing references and assisting with the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This study was initially supported by a grant from the Ministry of Health, Labour and Welfare, Japan. After governmental support expired, it was conducted in collaboration with Hiroshima University and the Foundation for Biomedical Research and Innovation.

Ethical approval

The ethics committee of the Hiroshima University Graduate School of Biomedical and Health Sciences approved the study. Ethical approval: 30002 for the main study (J-STARS) and 30003 for the substudy (hs-CRP).

Guarantor

MN.

Contributorship

All authors confirmed they have contributed to the intellectual content of this manuscript and have met the following three requirements: (a) significant contributions to the conception and design, acquisition of data or analysis and interpretation of data, (b) drafting or revising the article for intellectual content, and (c) final approval of the published article.

References

Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994; 344: 1383–1389.

Shepherd

Cobbe

Ford

et al.

Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group. N Engl J Med 1995; 333: 1301–1307.

Sacks

Pfeffer

Moye

et al.

The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events Trial investigators. N Engl J Med 1996; 335: 1001–1009.

The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. N Eng J Med 1998; 339: 1349–1357.

Nakamura

Arakawa

Itakura

et al.

for the MEGA Study Group. Primary prevention of cardiovascular disease with pravastatin in Japan (MEGA Study): a prospective randomised controlled trial. Lancet 2006; 368: 1155–1163.

Amarenco

Goldstein

Szarek

et al.

on behalf of the SPARCL Investigators. Effects of intense low-density lipoprotein cholesterol reduction in patients with stroke or transient ischemic attack: the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) Trial. Stroke 2007; 38: 3198–3204.

Nagai

Kohriyama

Origasa

et al.

for the J-STARS Investigators. Rational, design, and baseline features of a randomized controlled trial to assess the effects of statin for the secondary prevention of stroke: the Japan Statin Treatment Against Recurrent Stroke (J-STARS). Int J Stroke 2014; 9: 232–239.

Pearson

Mensah

Alexander

et al.

Centers for Disease Control and Prevention; American Heart Association. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation 2003; 107: 499–511.

Ridker

Rifai

Pfeffer

et al.

Long-term effects of pravastatin on plasma concentration of C-reactive protein. The Cholesterol and Recurrent Events (CARE) Investigators. Circulation 1999; 100: 230–235.

10.

Albert

Danielson

Rifai

et al.

for the PRINCE Investigators. Effect of statin therapy on C-reactive protein levels: the pravastatin inflammation/CRP evaluation (PRINCE): a randomized trial and cohort study. JAMA 2001; 286: 64–70.

11.

Hosomi

Nagai

Kohriyama

et al.

The Japan Statin Treatment Against Recurrent Stroke (J-STARS): a multicenter, randomized, open-label, parallel-group study. EBioMedicine 2015; 2: 1071–1078.

12.

Myers

Kimberly

Waymack

et al.

A reference method laboratory network for cholesterol: a model for standardization and improvement of clinical laboratory measurements. Clin Chem 2000; 46: 1762–1772.

13.

Abell

Levy

Brodie

et al.

A simplified method for the estimation of total cholesterol in serum and demonstration of its specificity. J Biol Chem 1952; 195: 357–366.

14.

Nakamura

Iso

Kitamura

et al.

Total cholesterol performance of Abell-Levy-Brodie-Kendall reference measurement procedure: certification of Japanese in-vitro diagnostic assay manufacturers through CDC’s Cholesterol Reference Method Laboratory Network. Clin Chim Acta 2015; 445: 127–132.

15.

Kimberly

Leary

Cole

et al.

for the Cholesterol Reference Method Laboratory Network. Selection, validation, standardization, and performance of a designated comparison method for HDL-cholesterol for use in the cholesterol reference method laboratory network. Clin Chem 1999; 45: 1803–1812.

16.

Nakamura

Yokoyama

Kayamori

et al.

HDL cholesterol performance using an ultracentrifugation reference measurement procedure and the designated comparison method. Clin Chim Acta 2015; 439: 185–190.

17.

Nakamura

Kayamori

Iso

et al.

LDL cholesterol performance of beta quantification reference measurement procedure. Clin Chim Acta 2014; 431: 288–293.

18.

Nakamura

Iso

Kitamura

et al.

Comparison between the triglycerides standardization of routine methods used in Japan and the chromotropic acid reference measurement procedure used by the CDC Lipid Standardization Programme. Ann Clin Biochem 2016; 53: 632–639.

19.

Edwards

Stribling

Pyatt

et al.

Reference measurement procedure for total glycerides by isotope dilution GC-MS. Clin Chem 2012; 58: 768–776.

20.

Koyama I, Iso H, Kiyama M, et al. Establishment of practical procedure for measurement of total glycerides by Isotope Dilution/Gas Chromatography/Mass Spectrometry at the Osaka Medical Center for Health Science and Promotion (CRMLN lipid reference laboratory). Clin Chem 2011; 57 (Supplement): A129 [Abstract].

21.

Method comparison and bias estimation using patient samples; approved guideline. NCCLS EP9-A, Vol.15 No.17, December 1995.

22.

Myers

Cooper

Winn

et al.

The Centers for Disease Control-National Heart, Lung, and Blood Institute Lipid Standardization Program. An approach to accurate and precise lipid measurements. Clin Lab Med 1989; 9: 105–135.

23.

Cooper

Myers

Henderson

. Establishment of reference methods for lipids, lipoproteins and apolipoproteins. Eur J Clin Chem Clin Biochem 1991; 29: 269–275.

24.

Kimberly

Vesper

Caudill

et al.

Standardization of immunoassays for measurement of high-sensitivity C-reactive protein. Phase I: evaluation of secondary reference materials. Clin Chem 2003; 49: 611–616.

25.

Nakamura

Sato

Shimamoto

. Establishment of external quality control program for hs-CRP and three-year follow-up of the performance for precision and accuracy. J Atheroscler Thromb 2007; 14: 287–293.