Medical Device Link.

REGULATIONS & STANDARDS

Randie R. Little, PhD, is an associate professor in the department of Pathology and Anatomical Sciences and the Department of Child Health at the University of Missouri School of Medicine, Columbia, MO. She is codirector of the Diabetes Diagnostic Laboratory, the coordinator of the NGSP network and a member of the NGSP steering committee, and a member of the IFCC working group on HbA1c. She can be reached at littleR@ health.missouri.edu.

Glycated hemoglobin, which is reported as HbA1c, measures metabolic control in patients with diabetes mellitus. The way in which HbA1c test results should be reported in the future has recently gene-rated a lot of debate. This debate has emerged due to the development of a definitive reference method for HbA1c measurements. This article explores the controversy surrounding this analyte.

Setting the Stage

Curt L. Rohlfing is a technical writer and research analyst at the Diabetes Diagnostic Laboratory in the Department of Pathology and Anatomical Sciences at the University of Missouri School of Medicine, Columbia, MO. He can be reached at rohlfingC@ health.missouri.edu.

Glycated hemoglobin results from the postranslational nonenzymatic attachment of glucose to hemoglobin. HbA1c is a specific glycated hemoglobin formed at the N-terminus of the hemoglobin beta chains, which is the most predominant glycation site. However, glycation of hemoglobin also occurs at the amino terminus of the hemoglobin alpha chains and at epsilon amino groups on lysine residues. Ion-exchange chromatography and immunoassay methodologies measure HbA1c, while boronate affinity methods measure total glycated hemoglobin. The results of these methods are highly correlated, and virtually all methods are standardized to express results as HbA1c regardless of the actual hemoglobin species measured.

Long-term clinical studies of patients with diabetes, most notably the Diabetes Control and Complications Trial (DCCT) and the United Kingdom Prospective Diabetes Study (UKPDS), have established a direct link between HbA1c levels and risks for developing diabetes-related complications including retinopathy, nephropathy, and neuropathy.^1,2 After the DCCT ended in 1993, major diabetes organizations issued guidelines for using HbA1c tests in clinical care, including HbA1c target levels for minimizing risks of complications.

For example, since 1994, the American Diabetes Association has recommended a target level of less than 7% HbA1c for most people with diabetes.³ The test is used not only by physicians and other healthcare providers but also by patients in their efforts to attain adequate glycemic control and to prevent or delay diabetic complications.

However, the glycated hemoglobin measurements from different assay methods used by laboratories varied considerably. Assay methods measured and, more importantly reported, different glycated hemoglobin fractions. Early studies had demonstrated the feasibility of standardizing glycated hemoglobin test results from different assay methods to a common reference.⁴ However, data from a 1993 survey by the College of American Pathologists (CAP; Northfield, IL) showed that labs were still reporting results on three different number scales (i.e., HbA1, HbA1c, or total glycated Hb). Even within those categories, the results varied considerably.⁵

Because of this situation, in 1993 AACC formed a subcommittee to determine the best way to standardize glycated hemoglobin test results. The subcommittee recommended the use of Bio-Rex 70 ion-exchange high- performance liquid chromatography (HPLC) as the comparative method for standardization of HbA1c. This method had demonstrated long-term stability in the DCCT, and could be used in a standardization program until more-definitive reference methods were developed.

Figure 1. (click to enlarge) Glycated hemoglobin results from the College of American Pathologists in 1993 and 2006. Each point and bar represent the mean ±2 standard deviation of results for a single method. Diamonds represent results reported as HbA1, squares represent results reported as HbA1c, and circles represent results reported as total GHB. The dotted horizontal line represents the NGSP/DCCT target value.

After the AACC subcommittee was disbanded in 1996, the National Glycohemoglobin Standardization Program (NGSP) was formed to implement the subcommittee’s recommendations.⁶ The NGSP has worked with IVD manufacturers and laboratories to establish their traceability to the DCCT/UKPDS values. Such efforts have improved the comparability of glycated hemoglobin test results reported by labs. For example, based on CAP data from late 2006, more than 99% of laboratories reported results as HbA1c or equivalent, and used assay methods that NGSP certified as being traceable to the DCCT/UKPDS values. Such methods include immunoassay, ion-exchange HPLC, and boronate affinity HPLC. In addition, results for a given specimen showed less variability than before, and are now compared to NGSP-assigned values (see Figure 1).

In 1995, the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC; Milan, Italy) working group on HbA1c standardization was formed to develop a definitive reference method and reference materials for HbA1c measurement.⁷ In the IFCC reference method, hemoglobin is cleaved by endoproteinase Glu-C. The resulting glycated and nonglycated n-terminal hexapeptides of the b chain are separated by HPLC. This is followed by quantification using either electrospray ionization mass spectrometry or capillary electrophoresis. HbA1c is measured as the ratio of glycated to nonglycated b-N-terminal peptide, and is reported as a percentage. IFCC established a global network of laboratories that analyzes HbA1c using its reference method, and provides value-assigned secondary reference materials to IVD manufacturers.

Figure 2. (click to enlarge) The relationship between HbA1c measured by the NGSP and IFCC networks. The upper green line is the regression line. The solid line is y = x.

Since the European Union’s IVD Directive requires IVDs to be traceable to a higher-order reference method in order to be sold in Europe, most manufacturers of HbA1c tests now document their traceability to the IFCC reference. The IFCC and NGSP laboratory networks demonstrate a linear relationship and correlation (r2 > 0.99), although the absolute numbers are different (see Figure 2). A master equation has been developed that can convert results between the IFCC and DCCT/UKPDS systems.⁸ IFCC results are about 1.9 units lower than DCCT/UKPDS results in the normal range, and about 1.3 units lower at elevated HbA1c (see Figure 2).

Brewing Controversy

Since developing its reference method, the IFCC has proposed completely changing the system for reporting HbA1c test results from DCCT/UKPDS units to IFCC units. IFCC argued that its reference numbers are accuracy based and that the change is needed to establish metrological traceability and compliance with the IVD Directive. However, members of the clinical community have resisted such proposals. They contend that changing the system from the established DCCT/UKPDS units could create much confusion and negatively affect patient care. Since HbA1c test results are reported as a percentage of HbA1c under the DCCT/UKPDS system, changing to the IFCC units, which are also expressed as a percentage, would be especially confusing since it may not be clear which result is being reported.

For example, one study showed a deterioration in metabolic control among patients when the system for reporting HbA1c was changed despite extensive efforts to educate patients.⁹ The IVD Directive does not require HbA1c test results to be reported in IFCC units, since the master equation provides statistically valid conversions (with negligible added uncertainty) between the IFCC and DCCT/UKPDS systems. The ISO standard on metrological traceability of values assigned to calibrators also indicates that a correction factor is acceptable if the added uncertainty is taken into account and the details of the correction are available to users.

Most IVD manufacturers are already documenting their traceability to the IFCC reference method and reporting DCCT/UKPDS results in the field. In addition, except for Sweden and Japan which have their own national standardization schemes, virtually all other countries report HbA1c results in DCCT/UKPDS units.

In 2004, representatives of the clinical diabetes community, the American Diabetes Association (ADA), the European Society for the Study of Diabetes (EASD), and the International Diabetes Federation (IDF) formed a working group for the HbA1c assay. They discussed the ramifications of the IFCC reference method and a possible change in the units for reporting HbA1c test results.^10,11 They raised three options: continuing to report results as DCCT/UKPDS equivalents, changing to reporting in IFCC units, and changing to other units such as mean blood glucose (MBG).

These groups decided to establish a study to define the relationship between HbA1c and MBG. The purpose of this ongoing study is to assess whether MBG and HbA1c are sufficiently correlated, and if the correlation is constant across different ethnicities, such that HbA1c can feasibly be reported as MBG.

The rationale is that patients will find MBG easier to understand since they are the same units used in their blood glucose monitors. This would make it simpler for patients to use their fingerstick blood glucose monitors to achieve HbA1c goals. The working group also recommended that laboratories continue reporting HbA1c test results in DCCT/UKPDS units until the study is completed.

Moving Forward

Members of the clinical laboratory community have already raised their objections to reporting HbA1c as MBG because test results would be reported as a name and in units that do not reflect the analyte being measured. They are also concerned that due to biological factors, MBG and HbA1c may not be sufficiently correlated to warrant expressing HbA1c results in terms of MBG.

Meanwhile, IFCC proposed approving a document written by the Joint IFCC Committee on Nomenclature, Properties, and Units, and the International Union of Pure and Applied Chemistry (IUPAC; Research Triangle Park, NC) Subcommittee on Nomenclature, Properties, and Units. This document would require HbA1c test results to be reported in IFCC units as mmol HbA1c/mol Hb. The reported results would be 10 times the IFCC percentage of HbA1c (e.g., an IFCC result of 4% would be reported as 40 mmol/mol).

The document also states that the proper name for the test should be “Haemoglobin beta chain (Blood) – N-(1-deoxyfructos-1-yl) haemoglobin beta chain; substance fraction milimole per mole.” In addition, the document states that “the current abbreviation ‘HbA1c’ should not be used for the new specifically measured property” and a suggested term for everyday speech is “DOF-Hemoglobin,” which is a shortened version of 1-deoxyfructos-1-yl.

Reporting IFCC units as mmol/mol would avoid potential confusion with the DCCT/UKPDS system since the number scale is higher. However, clinical groups are still concerned that such changes in units and nomenclature could lead to confusion among patients and their healthcare providers, and deterioration of metabolic control among patients with diabetes. Member countries have voted on the IFCC document; if it is approved, it would obviate the proposal for reporting HbA1c as MBG, regardless of the study’s outcome.

AACC voted against IFCC’s recommendations and urged it to delay adopting the standards until the international diabetes community completes its clinical trial evaluating the feasibility of reporting HbA1c as MBG. NIH agreed, stating that “a precipitous change in the clinically reported value of HbA1c could be detrimental to patients’ health and reverse a decade of effort and progress.”¹²

The IFCC’s president responded by expressing the view that there is no foundation for the clinical community’s fear of a change in the way HbA1c is reported. Hicks contended that “the adoption of the proposed HbA1c reference method and units of measurement represent an important step forward in the care of patients with diabetes.”¹³

Conclusion

Various concerned parties agree that the IFCC reference method can and should be the anchor for worldwide standardization of measuring glycated hemoglobin, and that utilizing the same units worldwide is a desirable goal.^11,12 There is also general agreement that changing to a different system other than the DCCT/UKPDS units would require extensive and expensive efforts to reeducate physicians and patients, in addition to the costs to IVD manufacturers to change their HbA1c reporting systems.

Questions remain: Should there be a change to a new system for reporting HbA1c test results, and if so, what should be reported? Who should make this decision? Who will pay for reeducating patients and physicians? Most importantly, will a change in HbA1c reporting improve patient care, or could it increase risk for diabetes complications?

---