|
 |
|
|||||||
|
|||||||||
© 2004 by the American Diabetes Association, Inc.
Letters: Observations |
The Korle-Bu Hb Variant in Caucasian Women With Type 1 Diabetes
A pitfall in the assessment of diabetes control
1 Department of Diabetology, Fundación Sardà Farriol, Barcelona, Spain
2 Department of Biochemistry, Quality Medical Service Laboratory Medical Center, Barcelona, Spain
Address correspondence to Dr. Ana Chico, Fundación Sardà Farriol, Department of Diabetology, Paseo Bonanova 69, Barcelona 08020, Spain. E-mail: 28299acb{at}comb.es
Measuring HbA1c concentrations in diabetic patients is an established procedure for evaluating the long-term control of diabetes. The Diabetes Control and Complications Trial confirmed the direct relationship between diabetes complications and HbA1c levels in type 1 diabetic patients. As a result, both the American Diabetes Association and the European Group for the Study of Diabetes have drawn up guidelines for assessing glycemic control by measuring HbA1c levels. However, in spite of advances in standardizing methods for measuring HbA1c concentrations, an increasing number of Hb variants produce false HbA1c determinations.
We report the first case of the Korle-Bu Hb variant in a Caucasian woman, which is also the first case described in a diabetic subject. We also describe the interference of this variant in some of the methods used to determine HbA1c concentrations. In our patient, HbA1c levels were underestimated for 20 years and, as a result of this misleadingly good metabolic control, the patient has developed microangiopathic diabetes complications.
A 29-year-old Caucasian type 1 diabetic woman was referred to our center in order to optimize her glycemic control because she was planning to become pregnant. Diabetes had been diagnosed 20 years earlier, and she had since been treated with NPH insulin twice a day. Glycemic control had been relatively acceptable (laboratory HbA1c 
7–7.5%; reference values 4–6%) from the start. Her therapeutic treatment was changed to multiple insulin injections (regular insulin before breakfast, lunch, and dinner and NPH insulin at bedtime), and screening for diabetes complications was started. Severe retinopathy requiring laser therapy was found, and symptomatic peripheral neuropathy was confirmed after an electromyographic study. Fortunately, the urinary albumin excretion rate was normal, and no macrovascular complications were found. After 2 months of the new therapeutic approach, HbA1c levels were measured by her reference laboratory and a value of 3.9% was obtained. However, capillary glucose measurements did not agree with that result due to the absence of frequent hypoglycemias. We decided to measure HbA1c again at our center using the DCA 2000 system (reference values <6%) and obtained a value of 6.9%. Three months later, HbA1c levels were measured by her reference laboratory and also at our center using the DCA 2000 system. The discrepancy was confirmed again (3.7 vs. 6.7%, respectively), demonstrating that the laboratory had underestimated the HbA1c value by 50%. It was suspected that a Hb variant might be present that could not be properly detected by the laboratory’s system, so electrophoretic analysis of the Hb was performed. An abnormal band was detected and, after sequenciation, corresponded to a Korle-Bu Hb variant present in heterozygosis of 40% (Fig. 1). A study of the family showed the same mutation to be present in her mother. Her brother chose not to be tested, and no other relatives were available for study.
|
More than 700 characterized Hb variants have been reported (1). The Korle-Bu variant of Hb, also known as Hb G-Accra, is one of the variants affecting the molecule’s ß-chains. It is the consequence of a mutation in codon 73 of the Hb gene (GAT
AAT), causing a single amino acid substitution in the ß-chain (AspAsn). This mutation is very uncommon and has only been described in a few cases affecting black families from Ghana, Ivory Coast, Jamaica, Mexico, and Costa Rica (2–5). To our knowledge, this is the first reported case to date in a Caucasian subject. The patient said her family was from Spain and denied having any ancestors from other countries or of a different ethnic origin. In heterozygosis, no phenotypic abnormalities have been detected in the subjects carrying the mutation. Hematology is normal, and no modifications in oxygen affinity or Hb stability have been found. No homozygote cases have been reported. All of the few cases of subjects with the Korle-Bu Hb variant reported to date corresponded to subjects screened to determine the prevalence of Hb variants in different populations. No cases involving diabetic subjects had been described before, and the effect of the variant on HbA1c determinations was unknown.
There are different methods for measuring HbA1c concentrations (6–8). The most frequent are boronate affinity or affinity-binding chromatography, cation-exchange chromatography, and automated high-performance liquid chromatography (HPLC) methods and immunoassays. Other methods include electrospray mass spectrometry and electrophoresis. Our patient’s HbA1c levels were measured at a laboratory using a well-validated HPLC method (Menarini HI-AutoA1c HA-8121). This method separates Hb species based on charge differences. Hb species elute from the cation-exchange column at different times with the application of buffers of increasing ionic strength. A spectrophotometer measures the concentration of Hb in each collected fraction, which is quantified by calculating the area under each peak. The following equation makes it possible to determine the amount of HbA1c in a given sample: percentage HbA1c = 100 x ([HbA1c/HbA] + HbA1c). In our case, the Hb variant (called HbX1c) and HbA1c separate, whereas HbA and HbX coelute, producing a spurious decrease in the calculated value of the percentage of HbA1c. The system used at our center to measure HbA1c (immunoassay method; Bayer DCA 2000) is based on antibodies that recognize the NH2-terminal–glycated amino acid in the context of three amino acids of the Hb ß-chain. These antibodies do not recognize other glycated Hb species, including chemically modified derivatives. The system is not affected by the Korle-Bu variant, which does not fall within the susceptible region.
We can conclude that glycemic control in diabetes must be evaluated not only through periodical HbA1c determinations but also with the information obtained from capillary glucose measurements. Any discrepancy found between the HbA1c and fingerstick measurements should be analyzed carefully, and the possible presence of Hb variants should be taken into account.
References
- Alphabetical hemoglobin variant list (Editorial). Hemoglobin 20:313–335, 1996[Medline]
- Ahern E, Ahern V, Serjeant GR, Serjeant BE, Seakins M, Darbre P, Middleton A, Lehmann H: Beta-chain variants in Jamaican newborns. Hemoglobin 2:495–502, 1978[Medline]
- Honig GR, Seeler RA, Shamsuddin M, Vida LN, Mompoint M, Valcour N: Hemoglobin Korle-Bu in a Mexican family. Hemoglobin 7:185–189, 1983[Medline]
- Elizondo J, Saenz JF, Alvarado ML, Ramon M: Hemoglobin Korle-Bu (alpha2-beta2 73 asp-asm) finding in Costa Rica. Sangre 21:54–56, 1976[Medline]
- Boissel JP, Fabritius H, Richard P, Wajcman H, Cabannes R, Labie D: Polymorphism of hemoglobins D in Ivory Coast: Hb Korle-Bu, Hb Avicenna and Hb Cocody. Nouv Rev Fr Hematol 23:197–201, 1981
- Bry L, Chen PC, Sacks DB: Effects of hemoglobin variants and chemically modified derivatives on assays for glycohemoglobin. Clin Chem 47:153–163, 2001
[Abstract/Free Full Text]
- Nakanishi T, Miyazaki A, Iguchi K, Shimizu A: Effect of hemoglobin variants on routine glycohemoglobin measurements assessed by a mass spectrometric method. Clin Chem 46:1689–1692, 2000
[Free Full Text]
- Schnedl WJ, Krause R, Halwachs-Baumann G, Trinker M, Lipp R, Krejs GJ: Evaluation of HbA1c determination methods in patients with hemoglobinopathies. Diabetes Care 23:339–344, 2000[Abstract]
CiteULike 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Diabetes | Diabetes Care | Clinical Diabetes | Diabetes Spectrum |