A 52-year-old African American male with previously diagnosed type 2 diabetes mellitus and hemoglobin SC disease underwent transfusion therapy for treatment of an acute episode of sickle cell crisis. Post-transfusion hemoglobin electrophoresis was ordered to assess the percent of his total hemoglobin that was hemoglobin A. The results of the post-transfusion electrophoresis were hemoglobin A: 63%, hemoglobin S: 20% and hemoglobin C: 17%. Hemoglobin fractions were also determined by HPLC to confirm the identity of hemoglobins S and C. Indeed, HPLC revealed hemoglobin A0: 55.7%, hemoglobin S: 13.2% and hemoglobin C: 11.9%. HPLC also revealed hemoglobin A2: 3.5% (reference interval: =<3.5%) and hemoglobin A1c: 9.9% (reference interval: =<6.0%).
Question: Can the hemoglobin A1c be interpreted to represent the patient's degree of glycemic (diabetic) control?
The authors do not believe that the A1c measured in this patient can be used to assess the patient's glycemic (diabetic) control.
This is a complicated case because the glycated hemoglobin A0 (constituting hemoglobin A1c measured by HPLC) is from the transfused blood as the patient produces no hemoglobin A0 of their own. Transfused blood should be glycated similarly to endogenously-produced blood containing hemoglobin A0 although transfused RBCs can have a shortened life span which could falsely lower the measured A1c . Because of the short duration of time between the patient's transfusion and the timing of the hemoglobin electrophoresis (e.g., <1 day), the A1c most likely represents the A1c in the transfused blood.
In a recent letter to Clinical Chemistry , Spenser and colleagues demonstrated that A1c decreased following transfusion in 69% of all diabetic patients studied (n = 45) and in 100% of diabetic patients whose pre-transfusion A1c was =>7.0% (n = 21). For patients whose A1c remained unchanged or increased post transfusion (31% of the subjects; n = 14), 89% of these subjects had A1c values <6.5% (e.g., in the nondiabetic range). On average in the entire study group, the absolute A1c level declined by ~0.8%. For patients with A1c values above the target range (i.e., =>7.0%; desirable: <7.0%), post transfusion the absolute A1c level declined by ~2.0%. The authors did not report whether any of the subjects in their study had a known hemoglobinopathy.
Many immunoassays can detect glycated hemoglobin S and glycated hemoglobin C; however such measurements in patients with hemoglobin SC disease should not be interpreted as a reflection of glycemic control because the shortened life span of such erythrocytes in hemoglobin SC disease will falsely lower the measured A1c.
One last point of interest is the question of A1c concentrations in banked blood. Spenser reported the A1c levels as 5.4% and 5.7% in 2 units of banked blood . These values increased only minimally with further blood storage.
The effect of blood transfusion on A1c levels might be summarized as follows: in patients with pre-transfusion A1c levels of =>7.0%, A1c tends to fall after transfusion. When the A1c is <7.0%, A1c rises minimally or not at all. Our case discussion points out that the A1c present in a patient may not relate to their glycemic control when the A1c is from transfused blood and the sample for A1c testing is drawn only a short time after transfusion.