Analytical performances of a glycated albumin assay that is traceable to standard reference materials and reference range determination

Abstract Background Glycated albumin (GA) is an intermediate‐term marker for monitoring glycemic control (preceding 2–3 weeks) in patients with diabetes mellitus. We evaluated the performance of Lucica Glycated Albumin‐L, a new GA assay that is traceable to standard reference materials and determined the reference range in healthy subjects without diabetes. Methods The performance and reference range studies were conducted in accordance with Clinical and Laboratory Standards Institute (CLSI) Guidelines. The traceability was established using reference material recommended by the Japan Society of Clinical Chemistry (JSCC). Results The coefficient of variation (CV) of overall repeatability, within‐laboratory precision, and overall reproducibility values of GA values were not more than 2.6%, 3.3%, and 1.6%, respectively, among laboratories. The GA values showed good linearity from 173 to 979 mmol/mol (9.4%–54.9%) across the assay range. The GA reference range in 262 healthy subjects was between 183 and 259 mmol/mol (9.9%–14.2%) while that of subjects with diabetes was 217–585 mmol/mol (11.8–32.6%). The reagent was stable for 2 months on the bench at room temperature. The limits of blank, detection, and qualification were 6.9, 7.9, and 9.7 μmol/L for GA concentration, and 3.8, 7.0, and 21.8 μmol/L for albumin concentration, respectively. Hemoglobin slightly affected the assay, while other classical interfering substances had no significant impact. Conclusions The present GA assay shows comparable performance to current clinical assays and could be used for intermediate‐term monitoring of glycemic control in diabetes patients.


| INTRODUC TI ON
Diabetes is a complex, chronic illness requiring continuous medical care with multifactorial risk-reduction strategies beyond glycemic control. 1 Large-scale clinical studies such as the Diabetes Control and Complications Trial Research (DCCT), 2 the Kumamoto Study, 3 and the U.K. Prospective Diabetes Study (UKPDS) 4 showed that improving glycemic control inhibits the occurrence and progression of diabetic complications. Hemoglobin A1c (HbA1c) reflects the concentration of mean plasma glucose (MPG) over the last 2-3 months and has been the gold standard for monitoring glycemic control of diabetic patients in clinical practice. 1 However, HbA1c levels may be inaccurate for hemoglobin variants and abnormal hemoglobin metabolism including anemia, decreased renal function, and gestational diabetes. 5 Glycated Albumin (GA) is an intermediate-term glycemic indicator that reflects the glycemic control status for the previous 2-3 weeks due to the albumin half-life. GA is an important HbA1c substitute for assaying samples with hemoglobin variants and abnormal hemoglobin metabolism because it is unrelated to this pathway. More importantly, GA changes quicker than HbA1c and might be a useful marker for detecting short-term changes of glycemic control during treatment. 6 Interventions and Complications (DCCT/EDIC). 10 GA is associated with vascular outcomes and mortality from the Atherosclerosis Risk in Communities (ARIC) study, which followed nearly 11,104 patients over 20 years. 11 There are also similar reports from Chinese and Japanese groups that shows GA is closely related to the onset and progression of diabetic complications in Asian population. [12][13][14] An Italian group summarized recent updates and advantages of GA as a biomarker for predicting and stratifying the cardiovascular risk. 15 Historically, GA levels are determined using several methods including boronate affinity chromatography, 16,17 ion-exchange chromatography, 18,19 thiobarbituric acid (TBA) assay, 18 immunoassay, 20,21 enzyme-linked boronate immunoassay, 22 and high-performance liquid chromatography (HPLC). 23,24 In 2002, an enzymatic method for GA measurement was developed by Kohzuma et al, and the improved method in 2004 is traceable using the HPLC method and widely used in clinical practice. 25,26 However, the target molecules and reference GA ranges differ for each method due to the definition of GA levels. 27 The Committee on Diabetes Mellitus Indices and the Japan Society of Clinical Chemistry (JSCC) recommended a reference procedure based on isotope dilution mass spectrometry (IDMS) for measuring GA and distributed reference materials for GA determination to standardize GA measurements. 28 In this study, we evaluated the performance of an enzymatic GA assay that is traceable to the JSCC-recommended standard reference materials. We compared the standardized new GA assay with the old assay and developed the GA reference range in healthy subjects without diabetes in the United States.

| Observed value study
The observed value of GA in subjects with diabetes was confirmed following CLSI Guideline EP28-A3C. One hundred and fifty subjects with diabetes (≥18 years of age) from eight different sites were tested. Subjects with end-stage renal disease, chronic kidney disease of Stage 3 or greater, liver cirrhosis, uncontrolled or untreated thyroid disease, a history within the last 6 months of a blood transfusion, and any other acute or chronic conditions that may significantly influence albumin or glucose metabolism in the opinion of the investigator were excluded from participation. The

| Method comparison study
The methods used for comparison studies were the HPLC traceable assay 26 and the present assay. The present assay under evaluation uses a calibrator that is traceable to standard reference material.
The HPLC traceable assay and current assay were used to test 1813 patient samples collected from a multicenter cohort study.
The study subjects were identified from the results of a study comparing GA to other glycemic indices. 30 The correlation between the two assays was studied, and Bland-Altman plots were employed to determine the differences. The method comparison study was approved by The Tulane University Biomedical IRB 618). All patients provided informed consent for participation in the study.

| Assay description and equations
Lucica® Glycated Albumin-L (Asahi Kasei Pharma) is an enzymatic GA assay that is traceable to the JSCC-recommended standard reference materials. This assay was used to determine GA expressed in mmol/mol, the ratio of GA concentration to albumin (Alb) concentration using calibrator for Lucica® Glycated Albumin-L and control for Lucica® Glycated Albumin-L. The assay was adapted for the Roche Modular P Chemistry Analyzer (Roche diagnostics). Results were automatically reported in mmol/mol. The analyzer calculates the GA value (mmol/mol) using the following calculation formula.
All the studies were conducted following the Clinical and Laboratory Standards Institute (CLSI) guideline including reference range study, observed value study, and performance study.

| Precision/Reproducibility
The precision study was performed in accordance with CLSI Guideline EP5-A3. The single-site precision study tested five serum pools at 2 runs/day in duplicates for 20 days (N = 80). The multi-site precision study tested three serum pools at 5 replicates/run, 1 run/day, for five testing days (N = 25) at 3 different laboratories. The time interval between measurements was set at ≥2 h when the precision test measurement was conducted twice a day. The repeatability and reproducibility were calculated as SD and CV%.

| Linearity
The linearity study was performed in accordance with CLSI Guideline EP6-A with 11 concentration levels prepared by mixing low and high GA value serum pools. Recovery (%) of the measured samples was calculated, and the polynomial evaluation of linearity was performed. The allowable acceptance criterion was ±5%.

| Traceability
The traceability system was established in accordance with "The Committee on Diabetes Mellitus Indices of the Japan Society of Clinical Chemistry-recommended reference measurement procedure and reference materials for glycated albumin determination. 28

| Stability
The stability test was performed in accordance with CLSI Guideline EP25-A. The fluctuations in the measurements were determined, and the allowable drift acceptance criterion was ±10%.

| Statistical analysis
The statistical analyses of reference range and observed value study were performed using SAS® version 9.3 or higher (SAS Institute).
The method comparison study was studied using the Passing-Bablok method (Starflex). The statistical software Analyse-it v4 (Microsoft) was used for the analysis of performance studies.

| Reference range study
Two hundred and sixty-two healthy subjects without diabetes were included in the reference range study, and the GA reference range was 183-259 mmol/mol (9.9%-14.2%) for the study population of nondiabetic healthy subjects. The reference study included 172 Caucasian subjects, 43 African American subjects, and 41 Asian subjects, which reflect the diversity of the U.S. population. No significant difference was observed between race, sex, age, and ethnicity subgroups (Table 1). was observed for group 1 and group 2, respectively. No significant difference was found between race, sex, age, and ethnicity subgroups (Table 2).

| Traceability
The calibrator and control were traceable to the JSCC-recommended

| Stability
The shelf-life for the reagents was 12 months when refrigerated within a temperature range of 2 and 8°C. The reagent was stable for 2 months on the bench at room temperature.

| Limit of quantification, limit of blank, and limit of detection
The LoQ, LoB, and LoD for GA and Alb concentrations were 9.7 μmol/L and 21.8, 6.9, and 3.8 μmol/L, and 7.9 and 7.0 μmol/L, respectively.

| Analytical interference
Unconjugated bilirubin, conjugated bilirubin, glucose, ascorbic acid, triglycerides, and uric acid showed no significant interference to the present GA assay at the following concentration (accepted bias

| Calculation formula simplification
The

| DISCUSS ION
In this study, a standardized GA assay that is traceable to JSCC- between African American and Caucasian samples. 24 The GA difference between African American and Caucasian subjects was similar to differences in HbA1c. 31 The GA reference range was 10.7%-15.1% in a healthy population of 1799 individuals (mean age: 55 years old; 51% female, 15% African American). 32 Therefore, the reference range in this study is lower compared with previous studies. Previous studies were based on subjects without diabetes, and there is a high possibility that both healthy subjects and prediabetic subjects were included. In contrast, prediabetic subjects, including impaired fasting glucose and impaired glucose tolerance, were excluded from this study. This inclusion/exclusion criteria is the likely main cause of the lower reference range in this work, although the limited sample size might also affect the results.
Although African Americans showed a higher GA reference

AUTH O R CO NTR I B UTI O N S
X.T and T.K. conceived the reference range study and participated in analysis and interpretation of the data. K.R and Y.N conducted the performance study and participated in analysis and interpretation of the data. All authors participated in the drafting and critical revision of the manuscript. All authors had full access to the data in the study and had final responsibility for the decision to publish.

ACK N OWLED G EM ENTS
We would like to thank the staff of Medpace, Inc., and Pacific Biomarkers for helping us with the reference study.

CO N FLI C T O F I NTE R E S T
All authors are employees of Asahi Kasei Pharma.

DATA AVA I L A B I L I T Y S TAT E M E N T
The dataset generated during the current study is not publicly available but is available from the corresponding author on reasonable request.

PATI E NT CO N S E NT S TATE M E NT
All patients provided informed consent for participation in the study.