Influence of in vivo hemoglobin carbamylation on HbA1c measurements by various methods

Increased carbamylated hemoglobin formed in erythrocytes during uremia may interfere with HbA(1c) assays, but few studies compared directly both parameters. We measured carbamylated hemoglobin by HPLC in 45 non-diabetic uremic patients (16 with acute and two with chronic renal failure, 27 with transplant recipients) as 57.8 +/- 22.3 mu g carbamylvaline /g Hb (mean +/- standard deviation) vs. 31.6 +/- 5.1 in 15 controls (+83%, p < 0.001). In these samples, HbA(1c) was evaluated by three ion-exchange HPLC methods, 1: Diamat (BioRad), 2: A1c2.2 (Tosoh) and 3: HA8140 (Menarini), and one immunoassay method (Tinaquant II Roche). Whichever the method, mean HbA(1c) values obtained increased in patients with high (> 60 mu g carbamylvaline/g Hb) vs. low (< 45) carbamylated hemoglobin values (+0.08 to 0.25% of total Hb), but differences were not significant. Minor peaks on the chromatograms were however increased in parallel to carbamylated hemoglobin. HbA(1c) values over 6% were found in 4, 1, 2 and 0 samples, with HPLC 1,2, 3 and immunoassay, respectively. Fructosamine values were not significantly altered. Our results show that Hb adducts, whether due to carbamylation or to other chemical reactions, interfere to a variable extent with different HbA(1c) assay methods, and confirm that HbA(1c) values should be interpreted with caution in uremic patients.