A 52-year-old woman with a medical history of hepatitis B, hyperlipidemia, hypertension, anemia, and depression presented to the internal medicine clinic for a routine visit. Laboratory tests 3 months previously had revealed an impaired fasting glucose concentration of 5.9 mmol/L (106 mg/dL) [reference interval, 3.9–5.6 mmol/L (70–100 mg/dL)]. Therefore, a hemoglobin (Hb)2 A1c analysis was performed. The initial Hb A1c evaluation by cation-exchange HPLC (CE-HPLC) (Hb A1c Program on the VARIANT II TURBO Link System; Bio-Rad Laboratories) showed an Hb A1c value of 115.8% (reference interval, 4.0%–6.0%) (Fig. 1).
In an effort to determine if the unusual Hb A1c result was due to potential hemoglobinopathies, we performed an Hb variant analysis with the Bio-Rad VARIANT CE-HPLC β-Thalassemia Short Program. The analysis revealed the absence of Hb A and the presence of sickle cell Hb (Hb S) (37.4%), along with normal Hb A2 (3.2%) and Hb F (<1.0%) (Fig. 2). Also evident was another large peak (53.0%) that eluted earlier than Hb A, which we called P2. This study suggested the presence of an Hb variant with a chromatographic retention time virtually identical to that of Hb A1c, in addition to Hb S (Figs. 1 and 2). A subsequent Hb electrophoretic analysis at pH 6.0 (QuickGel Acid; Helena Laboratories) identified Hb S and another abnormal band with a mobility similar to Hb F (not shown).
To identify the Hb variants, we investigated DNA sequences corresponding to the patient's β-globin genes. This analysis identified a substitution at codon 6 [GAG to GTG (Glu to Val)] on one allele, corresponding to Hb S, and a substitution at codon 1 [GTG to GCG (Val to Ala)] on the other allele, corresponding to Hb Raleigh.
QUESTIONS TO CONSIDER
What are the various types of methods used for measuring Hb A1c?
How do Hb variants interfere with each of these Hb A1c methods?
What actions should be taken when a spurious Hb A1c result is present?
The presence of these hemoglobinopathies suggested that the spurious Hb A1c result obtained with the CE-HPLC method was due to the elution of Hb Raleigh, which has a retention time similar to that of Hb A1c. We evaluated the Hb A1c result with a turbidimetric inhibition immunoassay (Dimension® Clinical Chemistry System; Siemens) and obtained an Hb A1c value of 4.1%, which was not consistent with the impaired fasting glucose concentration of 5.9 mmol/L (106 mg/dL).
Hb A1c is produced by nonenzymatic addition of a glucose molecule to the N-terminal valine residue on the β chain of Hb A. Glycation of the N-terminal residue changes its structure and decreases the positive charge of Hb A. The American Diabetes Association has recommended Hb A1c as an indicator of long-term glycemic control in patients with diabetes mellitus and for the screening and diagnosis of diabetes mellitus; an Hb A1c cutoff value of 6.5% has been suggested (1).
Methods of Hb A1c analysis can be divided into 2 categories: methods based on molecular charge and those based on structure. The former category includes CE-HPLC and electrophoresis, and the latter includes immunoassays, boronate affinity chromatography, and mass spectrometry (2). In CE-HPLC and electrophoresis assays, Hb A1c can be separated from Hb A because glycation of the N-terminal valine decreases the positive charge. Therefore, charge-based methods may be affected by posttranslational modifications (e.g., carbamylation and acetylation) (3) or by Hb mutations (2) that alter the charge. Common Hb traits such as Hb AS and Hb AC, however, do not interfere with the CE-HPLC Hb A1c assay used in this study (Bio-Rad VARIANT II TURBO) (4). Immunoassays use antibodies that target N-terminal glycated amino acids on the β chain to quantify Hb A1c, and the Hb A1c percentage is calculated from the Hb A1c and Hb concentrations (2). Thus, any factor that prevents glycation or any mutation in the epitope of the N-terminal amino acids that affects antibody recognition will produce erroneous results. Additionally, patients with increased Hb F (>10%) will have a falsely low Hb A1c value by immunoassay because the γ chain shares only 4 of the first 10 amino acids with the β chain of Hb A and has little to no immunoreactivity with most antibodies used in Hb A1c assays (2). In the boronate affinity chromatographic assay, boronic acid reacts with the cis diol groups created by glycation, thereby allowing glycohemoglobins such as Hb A1c to be separated from Hb A (2). On the other hand, Hb variants with excessive glycation, such as Hb Himeji, can interfere with boronate affinity chromatography (5). Mass spectrometric assay, an IFCC reference method, specifically measures the glycated N-terminal valine of the Hb A β chain (6), but the prohibitive cost of a mass spectrometer and the complicated nature of its installation and operation are likely to preclude its use in most clinical laboratories in the near future (2).
The method initially used to analyze the patient's Hb A1c was a CE-HPLC assay. A CE-HPLC Hb variants analysis (Fig. 2), acid gel Hb electrophoresis, and DNA sequencing of β-globin genes demonstrated that the falsely increased Hb A1c value was due to Hb Raleigh, which eluted in the Hb A1c window (Fig. 1). Hb Raleigh is unique in that a mutation (T to C) at the second base of the codon encoding the first amino acid of the β chain changes the N-terminal valine to an alanine residue. This substitution would not necessarily induce any change in Hb A charge except that N-terminal alanines are immediately acetylated to acetylalanines soon after translation (7, 8). This acetylation decreases the positive charge to one similar to Hb A1c. Therefore, the retention times of Hb A1c and Hb Raleigh are virtually identical; the elution peaks of these 2 Hbs fall in the same window in the chromatogram (Fig. 1). Thus, the presence of Hb Raleigh produces a falsely increased Hb A1c value when the latter is assessed by CE-HPLC. Chen et al. reported a case of falsely increased Hb A1c due to Hb Raleigh in the Bio-Rad VARIANT CE-HPLC Hb A1c assay (9). In their case, the patient was heterozygous for Hb A and Hb Raleigh, and the falsely increased Hb A1c value of 46% included a small fraction of real Hb A1c. In our case, the patient was heterozygous for Hb S, in which the N-terminal valine of the β chain was glycated as Hb S1c, and Hb Raleigh, in which the N-terminal acetylalanine of the β chain could not be glycated. Therefore, Hb A1c did not exist in this patient. The spurious Hb A1c value (115.8%) primarily represented Hb Raleigh. Other Hb variants that produce similar interferences include Hb Graz, Hb Sherwood Forest, Hb South Florida, Hb Niigata, and carbamylated Hb A (2).
The turbidimetric inhibition immunoassay produced an Hb A1c value of 4.1%, which represents the Hb S1c percentage, an equivalent of the Hb A1c percentage. Hb S1c was underestimated because of the heterozygosity with Hb Raleigh. Hb S is characterized by a substitution at codon 6 (GAG to GTG) that leads to the replacement of a glutamic acid residue in the β chain by a valine residue. Although this substitution is at the sixth amino acid residue, the antibody used in the immunoassay still recognizes Hb S1c. The acetylated alanine at the N terminus of the Hb Raleigh β chain, however, cannot be glycated and therefore prevents reaction with the antibody in the immunoassay. When the Hb A1c percentage is calculated, the numerator includes Hb S1c only, but the denominator consists of both Hb S and Hb Raleigh, as well as small amounts of Hb A2 and Hb F. Therefore, the Hb A1c percentage for this patient (as measured by immunoassay) was underestimated by approximately 50%. This issue with the Hb A1c immunoassay has been discussed by Chen et al. (9) and Jain et al. (10). Similarly, boronate affinity chromatography assays also underestimate Hb A1c for patients with Hb Raleigh, because the N-terminal acetylalanine in the Hb Raleigh β chain cannot be glycated. Hb Raleigh has a decreased affinity for boronate in the column (9), although the column can still interact with other glycated residues. For these patients, Chen et al. (9) and Jain et al. (10) recommended the use of fructosamine (9), multiple measurements of capillary glucose throughout the day, or continuous glucose monitoring of the glycemia (10). We did not perform these tests for our patient because of insufficient sample. For patients with the Hb Raleigh trait, the IFCC tandem mass spectrometric assay may be the best method, because it measures Hb A and Hb A1c specifically. The IFCC reference method, however, would be useless for our patient unless a mass spectrometric assay were available to measure Hb S and Hb S1c, because she had no Hb A or Hb A1c.
POINTS TO REMEMBER
Hb A1c assays can be divided into methods that use molecular charge (CE-HPLC and electrophoresis) and methods that use molecular structure (immunoassays, boronate affinity chromatography, and mass spectrometry).
Hb variants (or their glycated forms) may interfere with Hb A1c assays based on CE-HPLC and electrophoresis by coeluting/comigrating with Hb A and/or Hb A1c. If an Hb variant's amino acid substitution occurs within the N-terminal β-chain epitope recognized by an anti–Hb A1c antibody or if a patient has a greatly increased Hb F percentage, Hb A1c immunoassays will be affected. Hb variants with decreased or increased glycation will interfere with boronate affinity chromatography. Mass spectrometry is the IFCC reference method and generally appears to be unaffected by the presence of genetic or chemical modifications to the Hb A molecule.
When a spurious Hb A1c result is obtained, the possibility of interference by Hb variants should be considered, and the interpretation of Hb A1c results should be based on the patient's medical history and other laboratory results. In addition, efforts should be made to identify the Hb variant, and alternative Hb A1c methods that are free of the interference should be used. If there is no appropriate method for a particular Hb variant, fructosamine, daily multiple testing of capillary glucose, or continuous glucose monitoring may be used to monitor glycemic control. These alternative tests may also be used for patients who have an altered erythrocyte life span and changes in the degree of glycation. Hb A1c testing cannot be used for these individuals.
The present case is an example of how Hb variants can interfere with Hb A1c assays to produce spurious results. When an aberrant Hb A1c value is generated and/or the value does not match the clinical impression, the possibility of interference by Hb variants should be considered, and the interpretation of Hb A1c values should be based on the patient's medical history and other laboratory results. Efforts should be made to identify the Hb variant, and alternative Hb A1c methods free of the interference should be selected to monitor the patient's glycemic control.
↵2 Nonstandard abbreviations:
- cation-exchange high-performance liquid chromatography;
- Hb S,
- sickle cell Hb.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 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.
Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
- Received for publication August 26, 2010.
- Accepted for publication October 28, 2010.
- © 2010 The American Association for Clinical Chemistry