Evaluation of newly-developed glycated hemoglobin clinical analytic reagents and chromatography column on Tosoh HLC-723 G8 Analyzer

Objective To evaluate the performance of newly developed glycated hemoglobin (HbA1c) clinical analytic reagents and HPLC columns, applied on Tosoh HLC-723 G8 Analyzer. Methods Newly developed reagents and columns were used on a Tosoh HLC-723 G8 Analyzer (standard mode) system to measure both of qulity contorls and the clinical blood samples to evaluate the performances of these newly developed prodcuts including precision, accuracy, linearity, carryover, bias evaluation, correlation with commercial reagents, and stability according to CLSI recommendations. Results The CV of intra-assay precision and inter-assay precision of quality control and clinical blood sample assays using Lirimax products were both less than 3.00%. And the REs of accuracy were less than 6.00%. Linearity: R2 = 0.9993 in the concentration range 4.77%–14.67%. Carryover: 0.05%. The Bland-Altman mean difference: −0.003583% HbA1c (CI: 0.07398: −0.08115); Passing-Bablok regression: y = 1.0022(0.9984:1.006)x-0.01097(-0.03776: 0.01582), R2 = 0.9996. Stability evaluation was also acceptable. Conclusion The performance of newly developed products was well evaluated for HbA1c measurement on a TOSOH G8 Analyzer which shows excellent suitability for clinical assay.


Introduction
Nowadays 500~600 million people suffered from diabetes all over the world and around 200 million people were unconscious prediabetes [1][2][3][4].About 10%-13% of population in China is diabetic or prediabetic because of unhealthy lifestyle and changes in diet [5,6].If prediabetic patients were diagnosed and treated in earlier stage, their blood glucose levels could be reverted back to normal status.In addition, maintainin normal range of blood glucose in diabetic patient leads to decreased morbidity and mortality caused by diabetic complication.Therefore, the early and precise diagnosis was essential for patients with diabetes and prediabetes.
As known, HbA1c is a specific subfraction of glycated hemoglobin, which indicates average level of blood glucose concentration over 8-12 weeks, to assess long-term glycemic control and therapeutic effect as well as a predictor for the risk of developing microvascular complications of diabetes [7][8][9].Many clinical studies and trials suggested that HbA1c levels may be associated with cardiovascular disease [10], diabetic stroke, diabetic nephropathy, chronic microvascular complications and fundus oculi lesion, etc.So far HbA1c has been recommended for screening and diagnosis of diabetes by American Diabetes Association [11][12][13].Similar recommendations were proposed by WHO as well as other countries.
Until now, various methods have been developed and updated such as enzymatic assay, capillary electrophoresis, boronate affinity HPLC, ion-exchange HPLC and immunoassay, ect.Among them, enzymatic assay, boronate affinity HPLC and ion-exchange HPLC were popular in clinical HbA1c analysis [14].Ion-exchange HPLC, considered as the golden standard of clinical HbA1c analysis, is the most popular method for HbA1c analysis [15,16].Its analytical system usually includes automatic glycohemoglobin analyzer machine, elution buffer, hemolysis reagent, chromatography column, calibtrator and qulity control.The performance of machine usually cannot be modified after leaving the factory [15,[17][18][19].Except analyzer, the precise analysis still need specific elution buffer, hemolysis reagent, chromatography column, calibtrator and qulity control.Elution buffer could gradiently eluted all hemoglobin from chromatography column.Hemolysis reagent is cell lysis buffer which disrupted erythrocytes to release hemoglobin and clean equipment's pipes.Chromatography column as a hemoglobin adsorption and dissociation carrier is the core of components.Calibrator adjusts the machine to get accuracy value.And quality control is used to evaluate system working satus and performance.
Recently studies focus on comparing the performance of different Automatic Glycohemoglobin Analyzer or Hemoglobin Testing System.However, few cared the performance of reagents and columns, which were also crucial in glycohemoglobin analysis.
In this study, we conducted a comprehensive evaluation of newly developed reagents and columns for use with the Tosoh HLC-723 G8 Analyzer.Our aim was to assess whether they would be compatible with the related analyzer system and produce reliable glycohemoglobin analysis results.

Reagents
The standard references of glycated hemoglobin (GBW09181a, GBW09182a, and GBW09183a) were obtained from Clinical Test Center of Beijing Hospital Ministry of Health in China.
The evaluated HbA1c diagnostic reagents and columns: Calibrators

Instrumentation
HLC-723 G8 Automatic Glycohemoglobin Analyzer standard mode (TOSOH Corporation, Japan) was utilized in the whole evaluation.The method for glycohemoglobin analysis was recommended by the NGSP (National Glycohemoglobin Standardization Program) and IFCC (International Federation of Clinical Chemistry and Laboratory Medicine).The analyzer was calibrated at the beginning of the measurement, according to the manufacture's instructions using manufacturer-provided calibrators.

Clinical blood specimens
The clinical samples in EDTA anti-coagulant tubes were collected from Department of Clinical Laboratory, the Air Force Characteristic Medical Center of PLA, China and stored at 2-8 • C for up to 72 h.
For intra-assay precision, the same lot of hemolysis reagent (only for blood samples), elution buffer and columns were utilized for 10 tests of quality control and clinical blood samples for 10 times on the calibrated Analyzer.
For inter-assay precision, the clinical blood samples and quality control were measured for 10 times with 3 lots of hemolysis reagent (only for blood samples), elution buffer and columns on the calibrated Analyzer.The mean, standard deviation and CV were calculated.

Accuracy
The standard references of glycated hemoglobins (No.GBW09181a, GBW09182a and GBW09183a) were measured for 5 times on the calibrated Analyzer, respectively.The mean, standard deviation and relative error were calculated.

Linearity
One low value blood sample (L, Hb1Ac<5.00%) and one high value blood sample (H, Hb1Ac>14.00%)were measured on the calibrated Analyzer.Then the samples were diluted 150-fold and mixed as indicated in Table 1.Then seven mixed samples were obtained and measured for 3 times on the calibrated Analyzer.

Bias and correlation
120 clinical blood samples were measured using the reagents and columns from both Lirimax and TOSOH, respectively.

Stability
For long term stability, the performance of reagents and chromatography column were tested on the 3rd,6th, and 9th month, respectively, as above-mentioned in methods.

Statistical analysis
Statistitics was performed using GraphPad Prism software, version 5.0 (GraphPad Software Inc, La Jolla, CA).The data were presented with means, SDs, coefficients of variation (CV), correlation coefficients (r).

Precision
Both intra-assay and inter-assay precision are summarized in Table 2 and Table 3.We tested them using quality control product as well as clinical blood samples.All CV were below 3.00% which meets CLSI (Clinical and Laboratory Standards Institute) criteria in HbA1c analysis.

Accuracy
The accuracy of noval reagents and columns was evaluated by measuring 3 levels of Standard References.All of the relative errors were less than 6.00%, a common criteria of National Health Commission of China in Table 4.

Linearity and carryover
HbA1c levels were ranged 4.77%-14.67%.The linear regression showed measured value well matched target value (R 2 = 0.9993) as shown in Fig. 1.The carry-over between high-and low-levels of HbA1c measured using the reagent and columns from Lirimax was

Bias and correlation
The results of measurements using reagents and columns from Lirimax and TOSOH G8 showed a high concordance in 120 clinical samples.The Bland-Altman plot showed a mean difference of − 0.003583% in HbA1c (CI: 0.07398: − 0.08115) in Fig. 2. Passing-Bablok regression between Lirimax and TOSOH G8 showed a slope of 1.0022 with a 95% CI 0.9984-1.006,and an intercept of − 0.01097 with CI (− 0.03776: 0.01582).The Pearson Correlation Coefficient was 0.9996 as shown in Fig. 3.

Stability
The product stability was also assessed.The accuracy, precision and linearity of the products from Lirimax on the 3rd, 6th, and 9th  month were evaluated.All of them were acceptable according to the common criteria.All data was shown in Table 5, Table 6 and Fig. 4.

Discussion
Since HbA1c was discovered in the 1960s, it has become one of the most widely used clinical biomarker of long-term blood glucose level control in prediabetic and diabetic patients.As mentioned, there are several popular methods to measure HbA1c.The guidelines for laboratory analysis of diabetes diagnosis gave a recommendation on the CV of intra-laboratory≤2.00%and the CV of inter-lab-oratory≤3.50% [20][21][22].HPLC assay was recommended by various international organizations incuding IFCC [23,24].
Subquently, different instruments based on HPLC appeared and updated in the last two decades.As known, the components of HbA1c analysis system (HPLC) included the reagents such as calibrators, quality control, hemolysis reagent, elution buffer and chromatography column besides analyzer.So far there are a number of studies on evaluation of these analyzers or on the comparison of different analyzers.However, few researcher considered the evaluation of performance of reagents and column, or whether the diagnostic kit was suitable for certain analyzer.Although Analyzer plays an important role in measuring HbA1c, no doubt the performance of analytical reagents and columns is also crucial for the reproducible and unbiased measurement of HbA1c in clinical diagnosis of prediabetes and diabetes [25,26].Therefore, we focused on the evaluation of accuracy, precision and stability of newly developped reagents and columns.
In the case of the precision, all CV of both intra-assay and inter-assay measuring quality control as well as clinical blood samples were below 3.00%.It indicated that these products possesed a good reproducibility under proper experimental conditions.However, precision was not enough to validate an excellent set of analytic products.It also need good accuracy to assess the difference between measured value and truth value [27].All REs were less than 6.00%, a suggested criteria of National Health Commission of China, lower than 8.00% as claimed by TOSOH.Well-performed precision and accuracy are fundamentally important for the process of diabetes moitoring.Linearity showed that the products give reliable results within a wide range of HbA1c levels.Moreover, considering the continuously measuring by automated machine, the ture value of samples might be contaminated by adjacent samples [28].So low    B. Yuan et al.

B.
Yuan et al.

Fig. 3 .
Fig. 3. Correlation of HbA1c results measured with reagents and columns from both TOSOH G8 and Lirimax.

Table 1
Preparation of linear samples mixture.

Table 2
Precision study of quality control.

Table 3
Precision study of clinical blood samples.

Table 4
Accuracy study of standard references.

Table 6
Stability study of precision.