A-073 Novel Rare Hemoglobin Variant Causes False Elevation of HbA1c by Cation Exchange Chromatography

Abstract Background Sixteen years ago, hemoglobin A1c (HbA1c) became a cornerstone for diagnosis of diabetes mellitus (DM) when the IFCC and ADA recommend its use. Methods of measurement include high performance liquid-chromatography (i.e., cation-exchange and boronate-affinity), chemistry methods (i.e., enzymatic and immunoassay), and most recently capillary electrophoresis. One common and high-throughput method for HbA1c by cation-exchange chromatography is the BioRad D-100, which we employ in our lab. Cation-exchange methods generally excel at resolving and quantifying common hemoglobin variants. However, in practice, partial variant overlap with the HbA1c peak using our method means that ∼0.29% of our results are not reported. Peak integration is automatically performed by BioRad*s algorithm. In those cases, we may decide to reflex the sample for measurement by an alternate method. We recently came across the case of a patient without DM who measured 7.0% (Figure 1) with the D-100 method. The patient became concerned and requested recollection and measurement by another method and the result was 5.4%. We proceeded to investigate. Methods We observed no D-100 instrument or QC issues that could explain the discrepancy so, with patient cooperation, we drew their blood for genetic analysis of hemoglobin variants. The patient also had their blood collected on three separate occasions 1-3 months apart over the course of five months to determine if the initial increased result was reproducible and rerun by an alternate Roche Cobas C503 enzymatic method to confirm the discrepancy. The first two measurements were performed with column-cartridges in the middle of their lifespan. Consultation with BioRad indicated that column age may potentially be a factor leading to “tailing” of the suspect HbA1c peak so we replaced the column with a new one for the third measurement to eliminate this unlikely variable. Results The patient was consistently 7.0% across three D-100 measurements over five months and was consistently 5.4% by the Roche enzymatic method. In Figure 1A the patient*s peak also shows subtle evidence of a doublet at the top. Hemoglobin genetic analysis was largely normal except for the observation of a rare, benign Athens-Georgia variant, which was not quantified. We searched literature for prior observation of this variant interfering with cation-exchange or other HbA1c methods and could find none. In fact, there was one report that showed Athens-Georgia could be picked up as a peak by capillary electrophoresis (ClinChemLabMed 2019; 57(6):e108–e110), but did not interfere with quantifying HbA1c. There has been observation of another rare variant, Hb Wayne, causing unexpected elevation in HbA1c by cation-exchange methods (ProcBaylUnivMedCent, 2022;35(1):84–85), but again none for Athens-Georgia. Conclusion Once we had eliminated column age as a factor, we were confident that Athens-Georgia was the cause of the false elevation. This interference may not occur with all cation-exchange methods as elution-time, buffers, and column-resin are all variables. However, with the commonly employed BioRad D-100 method the interference is consistently present and not previously observed in literature or by BioRad. Labs running this method should be aware of the potential for this rare interference yielding false elevated HbA1c results.