Presentation Number: SAT-761
Date of Presentation: June 15th, 2013

Tejal Uday Shah*1 and Shailendra B Patel2
1Medical Coll of Wisconsin, Milwaukee, WI, 2Medical College of Wisconsin, Milwaukee, WI


Introduction: Measurement of glycosylated hemoglobin or HbA1c is widely accepted as a means to manage diabetes mellitus and can also be used to diagnose diabetes. Health care providers should, therefore, be aware of pitfalls that influence HbA1c and its measurement.

Clinical Case: The patient is an 81-year-old African American veteran with a history of type 2 diabetes mellitus, diagnosed three years previously with two fasting blood glucose values of 127 mg/dL and 139 mg/dL (n 70-99 mg/dl), who was referred to the endocrine clinic for management. He was on Metformin 500mg three times daily and reported compliance and tolerance of the medication. His fingerstick average pre-breakfast and pre-dinner were 128 mg/dl and 150 mg/dl (n 70-99 mg/dl), respectively. Patient’s HbA1c was checked and found to be 3.5% (n 4.0-5.7%) with his HbA1c range since diagnosis 3.0-4.1% (n 4.0-5.7%), proportionally low compared to his fingersticks. A hematological disorder was suspected but hematological parameters, including hemoglobin, hematocit, MCV, iron, TIBC, ferritin, folate, and vitamin B12 levels, were normal. Patient’s creatinine level when checked was 1.03 mg/dL (n 0.67-1.17 mg/dl) and total bilirubin was 0.3 mg/dL (n 0.2-1.0 mg/dl). Reticulocyte count and absolute reticulocyte count were respectively 1.06% (n 0.5-2.5 %) and 50.0 K/µL (n 25-100 K/ul). A hemoglobinopathy was suspected but a hemoglobin electrophoresis test revealed HbA1 97% (n 97-99%), HbA2 2.5% (n 0-3%), and HbF 0.5% (n 0-2%). A check of his fructosamine was 293 µmol/L (n 190-270 umol/L) with albumin of 3.4 g/dL (n 3.4-5.0 g/dl), consistent with his fingersticks and confirming his hyperglycemia.

We then measured the HbA1c using two different methods with the same blood sample revealing HbA1c values of 3.5% using high performance liquid chromatography and 7.2% using an immunoassay, suggesting that the discrepant HbA1c was due to some effect upon the method used. This again raised the possibility of a variant Hb. When the patient’s Hb was re-run using another instrument, a variant and split peak was detected. The chromatogram is similar to a rare Hb variant identified by sequencing to be Hb Silver Spring.

Clinical Lessons: There are many factors that influence HbA1c and its measurement, including increased or decreased red cell turnover, factors that increase or decrease glycation, conditions of erythrocyte destruction, and assay interferences. In this case, a rare Hb variant, which was not detected originally on standard electrophoresis, was responsible for the abnormally low HbA1c. In understanding that there are interferences, including Hb variants, that may affect HbA1c and its measurement, one should exert caution when using HbA1c alone as a means to manage or screen for diabetes, even if this is an Aunt Minnie.


Nothing to Disclose: TUS, SBP