Sage Journals: Discover world-class research

Abstract

Background

Asymptomatic variant haemoglobin is increasingly being found in the measurement of glycated haemoglobin (HbA_1c) for the management of diabetes mellitus. We compared the HbA_1c concentrations measured by high-performance liquid chromatography (HPLC) and immunoassay and glycated albumin (GA) concentrations and calculated the respective ratios in order to classify the variant haemoglobin.

Methods

Twenty different haemoglobin variants from 43 subjects were identified by mass spectrometry and DNA analysis. Since GA accurately reflects glycaemic control in patients with variant haemoglobin, we calculated respective ratios of HbA_1c and GA. Haemoglobin variants causing a low ratio of HbA_1c measured by HPLC (HPLC-HbA_1c) to GA with a normal ratio of HbA_1c measured by immunoassay (IA-HbA_1c) to GA were classified as C1. A further classification of α and β was used with abnormalities of the α chain or β chain in the haemoglobin gene. Other haemoglobin variants were classified as non-C1. Eight diabetic patients with stable glycaemic control were used as controls.

Results

Twenty forms of variant haemoglobins were classified as C1α (2 variants; I-Interlaken and Hb J-Meerut), C1β (15 variants) and non-C1 (3 variants; Hb Himeji, Hb Woolwich, Hb Peterborough). Positive correlations between GA and HPLC-HbA_1c or IA-HbA_1c were seen in the C1β patients with diabetes mellitus. The regression line between GA and HPLC-HbA_1c, but not that between GA and IA-HbA_1c, showed a downward shift in comparison with the data obtained from the diabetic controls.

Conclusions

Variant haemoglobin could be classified by calculating the ratios of HPLC-HbA_1c, IA-HbA_1c and GA.

Introduction

Variant haemoglobin results from mutations of amino acids in haemoglobin. A total of 1128 variant forms have been reported worldwide as of December 2011.¹ About 20% of the variants produce clinical phenotypes such as haemolytic anaemia, polycythaemia and metohaemoglobinaemia, while the remaining 80% demonstrate no abnormal phenotype. Although variant forms of haemoglobin have traditionally been found in patients with haematological diseases, asymptomatic forms are often found in the measurement of glycated haemoglobin (HbA_1c) for the management of diabetes mellitus.^2–4 Currently, high-performance liquid chromatography (HPLC) is most used to measure HbA_1c, since the technique is more accurate and reliable than other methods. In patients with variant haemoglobin, their HbA_1c concentrations measured by HPLC (HPLC-HbA_1c) are higher or lower than the actual levels corresponding to glycaemia, leading to misjudgement of the glycaemic control status. In contrast, HbA_1c measured by immunoassay (IA-HbA_1c) usually is not influenced by variant forms of haemoglobin, and this method is recommended for distinguishing the majority of variant haemoglobin and making decisions about their glycaemic control state.⁵ However, variant forms of haemoglobin in which the lifespan of red blood cells is shortened or glycation is increased or decreased have been reported.⁶ With these variant haemoglobins, even IA-HbA_1c sometimes yields abnormal results.

Glycated albumin (GA), another glycaemic control marker, is considered to accurately reflect glycaemic control in patients with variant haemoglobin.⁷ Through calculation of the ratio of HPLC-HbA_1c or IA-HbA_1c to GA (HPLC-HbA_1c/GA or IA-HbA_1c/GA), it is possible to judge the discrepancy between HbA_1c levels and glycaemic control state because the HbA_1c/GA ratio is not influenced by glycaemic control state.^8,9 Furthermore, no classification schemes have been set for variant haemoglobin based on the reasons by which HbA_1c is incorrectly measured. We therefore measured HPLC-HbA_1c, IA-HbA_1c and GA values in patients with variant haemoglobin and aimed to classify variant forms of haemoglobin using the HbA_1c/GA results.

Subjects and methods

Subjects and controls

We investigated 20 different variant forms of haemoglobin from 43 subjects (Table 1). As previously reported, identification of variant haemoglobin was performed by electrospray ionization/mass spectrometry, tandem mass spectrometry and DNA analysis.⁸ Eight patients with type 2 diabetes mellitus with stable glycaemic control were used as diabetic controls. All steps of this study, including DNA analysis, were approved by the Ethics Committees at Osaka Medical College and the study complied with the ethical guidelines of the Helsinki Declaration as revised in 2000. All subjects provided their written informed consent.

Table 1

Variant haemoglobin specimens in the present study

Type	N	DM*	Variant name	Substitution	References
C1β	2	2	Buzen	β138Ala → Thr
C1β	2	0	E	β26Glu → Lys
C1β	2	0	G-Coushatta	β22Glu → Ala
C1β	2	0	G-Szuhu	β80Asn → Lys
C1β	7	0	Hamadan	β56Gly → Arg
C1β	6	3	Hikari	β61Lys → Asn
C1β	1	0	Hirose	β37Trp → Ser
C1β	3	0	Hoshida	β43Glu → Gln
C1β	1	0	J-Lome	β59Lys → Asn
C1β	1	0	Masuda	β114Leu → Met
C1β	1	0	Moriguchi	β97His → Tyr
C1β	7	4	Riyadh	β120Lys → Asn
C1β	1	0	Shizuoka	β108Asn → His
C1β	1	0	Stanmore	β111Val → Ala
C1β	1	0	Yoshizuka	β108Asn → Asp
C1α	1	0	I-Interlaken	α15Gly → Asp	13,14
C1α	1	0	J-Meerut	α2Ala → Glu	15
Non-C1	1	0	Himeji	β140Ala → Asn	16,17
Non-C1	1	0	K-Woolwich	β132Lys → Gln	18
Non-C1	1	0	Peterborough	β111Val → Phe	19

*Number of patients with diabetes mellitus

Laboratory measurements

HbA_1c was measured by HPLC (HPLC-HbA_1c) using HA-8150 (Arkray Inc., Kyoto, Japan)³ and by immunoassay (IA-HbA_1c) using DCA-2000 (Bayer Diagnostics, Boston, MA, USA)¹⁰ in these subjects. All these methods for HbA_1c were corrected as a National Glycohemoglobin Standardization Program (NGSP) equivalent value (%) calculated using the formula HbA_1c (%) = HbA_1c (Japan Diabetes Society: JDS) (%) + 0.4%, considering the relational expression of HbA_1c (JDS) (%) measured using previous Japanese standard substances and measurement methods and HbA_1c (NGSP).¹¹ Serum GA was measured by HPLC.¹²

Since GA accurately reflects glycaemic control in patients with variant haemoglobin, we calculated the respective ratios of GA and HbA_1c and aimed to classify the variant forms of haemoglobin according to HbA_1c/GA. Variant forms of haemoglobin demonstrating low HPLC-HbA_1c/GA (<0.35) with normal IA-HbA_1c/GA (0.3–0.5) were classified as C1. A further classification of α and β was used in order to differentiate between variant haemoglobin with abnormalities of the α chain and β chain in the haemoglobin gene. Other haemoglobin variants were classified as non-C1.

Statistical analysis

All data are shown as mean ± SD. For statistical analysis, unpaired Student's t-test was used to compare two groups. P values of <0.05 were considered to be statistically significant.

Results

According to the levels of HPLC-HbA_1c, IA-HbA_1c and GA, 20 variant forms of haemoglobin identified from 43 subjects in this study were classified as C1α (2 variant forms; Hb I-Interlaken and Hb J-Meerut), C1β (15 variant forms, n = 38) and non-C1 (3 variant forms; Hb Himeji, Hb K-Woolwich and Hb Peterborough) (Table 1). The majority of variant forms of haemoglobin were classified as C1β type. Among the 38 C1β patients, 10 patients (3 species) had diabetes mellitus while the other 28 patients (14 species) did not have diabetes mellitus.

We identified a positive correlation between GA and HPLC-HbA_1c (R = 0.692, P = 0.057) or IA-HbA_1c (R = 0.770, P = 0.025) in the diabetic controls (Figure 1). A positive correlation between GA and HPLC-HbA_1c (R = 0.625, P = 0.053) or LA-HbA_1c (R = 0.660, P = 0.038) was also seen in the C1β patients with diabetes mellitus. The regression line between GA and HPLC-HbA_1c, but not GA and IA-HbA_1c, showed a downward shift in comparison with the data obtained from the diabetic controls. A significant positive correlation between HPLC-HbA_1c and IA-HbA_1c was also seen in the diabetic controls (R = 0.972, P < 0.001) and the C1β patients with diabetes mellitus (R = 0.741, P = 0.014). The regression line between HPLC-HbA_1c and IA-HbA_1c also showed a downward shift in comparison with the data obtained from the diabetic controls.

Figure 1

Correlation between GA and HbA_1c in the C1β patients with variant haemoglobin with or without diabetes mellitus. The correlations between GA and HPLC-HbA_1c (a), GA and IA-HbA_1c (b), and IA-HbA_1c and HPLC-HbA_1c (c) in the C1β patients with variant haemoglobin with (closed circles) or without diabetes mellitus (open circles) and the diabetic controls (closed squares). The regression lines for the C1β patients with diabetes mellitus (straight line) and the diabetic controls (dotted line) are also shown. GA, glycated albumin; HbA_1c, glycated haemoglobin; HPLC, high-performance liquid chromatography; HPLC-HbA_1c, HbA_1c measured by HPLC; IA-HbA_1c, HbA_1c measured by immunoassay

Table 2 presents the results obtained in the present study. Although GA and IA-HbA_1c in the C1β patients with diabetes mellitus were not significantly different from the diabetic controls, their HPLC-HbA_1c, HPLC-HbA_1c/GA and HPLC-HbA_1c/IA-HbA_1c were significantly lower than the diabetic controls. GA, HPLC-HbA_1c and IA-HbA_1c concentrations in the C1β patients with diabetes mellitus were significantly higher than the C1β patients without diabetes mellitus, but the HPLC-HbA_1c/GA and the HPLC-HbA_1c/IA-HbA_1c were not significantly different between both groups. The IA-HbA_1c/GA in the C1β patients with or without diabetes mellitus was not significantly different from that in diabetic controls. The HPLC-HbA_1c/GA and the HPLC-HbA_1c/IA-HbA_1c in the C1α patients were higher than that in the C1β patients. In the non-C1 patients, the HPLC-HbA_1c/GA and the HPLC-HbA_1c/IA-HbA_1c were different from those in the controls. Furthermore, the IA-HbA_1c/GA in the patient with Hb Himeji and in the patient with Hb Peterborough was also different from that in the controls.

Table 2

GA values, HbA_1c values measured by HPLC and immunoassay, and ratios of the three results in types of variant haemoglobin

Type	Species	n	Diabetes mellitus	GA (%)	HPLC-HbA_1c (%)	IA-HbA_1c (%)	HPLC-HbA_1c/GA	IA-HbA_1c/GA	HPLC-HbA_1c/IA-HbA_1c
Control	–	8	+	21.8 ± 1.6	8.3 ± 0.6	8.1 ± 0.6	0.38 ± 0.02	0.37 ± 0.02	1.02 ± 0.02
C1β with diabetes mellitus	3	10	+	21.0 ± 4.5	4.6 ± 1.6*	7.4 ± 2.1	0.22 ± 0.06*	0.35 ± 0.07	0.63 ± 0.16*
C1β without diabetes mellitus	14	28	−	14.1 ± 1.1**	3.1 ± 0.8**	5.3 ± 0.4**	0.22 ± 0.06	0.38 ± 0.04	0.59 ± 0.01
C1α	2	2	−	14.5	4.2	5.3	0.29	0.36	0.79
Non-C1 (Hb Himeji)	1	1	−	14.5	4.2	8.7	0.29	0.60	0.48
Non-C1 (Hb K-Woolwich)	1	1	−	16.0	33.6	6.1	2.1	0.38	5.5
Non-C1 (Hb Peterborough)	1	1	−	14.2	2.7	4.2	0.19	0.30	0.64

GA, glycated albumin; HbA_1c, glycated haemoglobin; HPLC, high-performance liquid chromatography; HPLC-HbA_1c, HbA_1c measured by HPLC; IA-HbA_1c, HbA_1c measured by immunoassay

*P < 0.001 versus control

**P < 0.001 versus C1β with diabetes mellitus

Discussion

We aimed to classify variant haemoglobin on the basis of the reasons by which HbA_1c is measured as abnormal values. In patients with variant haemoglobin complicated with diabetes mellitus, because HbA_1c is also affected by the glycaemic control state, evaluation of the HbA_1c value is difficult. We therefore calculated the respective ratios of HPLC-HbA_1c or IA-HbA_1c and GA and were able to identify a C1 group and a non-C1 group of variant forms of haemoglobin by comparing these ratios.

The majorities of variant haemoglobin measured in the present study were classified as C1β type. Alpha chain variants cause less discrepant results from actual HbA_1c values than β chain variants, because there are two genes (α1 and α2) for haemoglobin α chains. Two variant haemoglobins of C1α type were identified as Hb I-Interlaken^13,14 and Hb J-Meerut.¹⁵

We could not subcategorize non-C1 type since there were only three patients identified as non-C1. However, the HbA_1c/GA ratios of these patients with non-C1 type were different from each other. The HPLC-HbA_1c in patients with Hb K-Woolwich was known to be extremely high because the variant haemoglobin overlaps with HbA_1c on HPLC.¹⁶ However, since HbA_1c measured by immunoassay reflects their state of glycaemic control, the LA-HbA_1c/GA was similar to that of the controls. Hb Himeji is reported to be increased glycation.^17,18 Therefore, IA-HbA_1c in Hb Himeji was high, although HPLC-HbA_1c was low. Hb Peterbrough is known to be unstable haemoglobin.¹⁹ Therefore, IA-HbA_1c in Hb Peterpough showed a low value. As a result, the non-C1 type of variant haemoglobin includes various forms and thus it is necessary to subclassify the non-C1 type in future after collecting data from many non-C1 patients.

Fructosamine has been previously used as a measure of glycaemic control in patients with variant haemoglobin.⁵ However, as fructosamine is expressed as the concentration of glycated protein, its measured values are affected by serum protein concentration. With distorted serum protein concentrations, fructosamine cannot accurately reflect glycaemic control.²⁰ GA is considered superior to fructosamine as a clinical marker of glycaemic control, as this value is expressed as a ratio of glycated albumin to total (glycated and non-glycated) albumin.⁷ The present study is the first to examine GA concentrations in patients with variant haemoglobin systematically.

Several diseases showing abnormal HbA_1c values are known besides those involving variant haemoglobin. Haemolytic anaemia, liver cirrhosis, renal anaemia and iron deficiency anaemia under treatment show values that appear to be low,²¹ while iron deficiency anaemia shows values that appear to be high.²² Differentiating these diseases from variant haemoglobin is sometimes difficult. However, both HPLC and immunoassay yield similar HbA_1c results in the former, whereas both assays yield divergent HbA_1c results in all forms of variant haemoglobin (Table 2). As a result, variant haemoglobin and the other diseases can be differentiated in patients showing discrepancies between glycaemic control and HbA_1c values by measuring HbA_1c with HPLC and immunoassay and comparing both results.

In conclusion, variant forms of haemoglobin could be classified by measuring HbA_1c by HPLC and immunoassay and GA, and comparing the respective ratios of the three results. Differential diagnosis can also be made between variant haemoglobin and other conditions yielding abnormal HbA_1c values, using this approach.

DECLARATIONS

Competing interests: None.

Funding: No funding was required in this research.

Ethical approval: The ethics committee of Osaka Medical College approved this study and written informed consent was obtained from each subject.

Guarantor: AM.

Contributorship: AM researched data, contributed to the discussion and edited the manuscript. TK researched data, contributed to the discussion and edited the manuscript. SK contributed to the discussion, and reviewed and edited the manuscript. MK wrote and edited the manuscript and contributed to the discussions. All authors approved its final version.

Acknowledgements: None.

References

Wajcman

. Human Hemoglobin Variants, the Electric Versions. See http://globin.cse.psu.edu (last checked 20 December 2011)

Bry

, Chen

, Sacks

. Effects of hemoglobin variants and chemically modified derivatives on assays for glycohemoglobin. Clin Chem 2001;47:153–63

Schnedl

, Krause

, Halwachs-Baumann

, Trinker

, Lipp

, Krejs

. Evaluation of HbA_1c determination methods in patients with hemoglobinopathies. Diabetes Care 2000;23:339–44

Thomas

, Agosti

, Man

, Mastorides

. Screening for hemoglobinopathies during routine hemoglobin A1c testing using the Tosoh G7 Glycohemoglobin Analyzer. Ann Clin Lab Sci 2007;37:251–5

Halwachs-Baumann

, Katzensteiner

, Schnedl

, Pürstner

, Pieber

, Wilders-Truschnig

. Comparative evaluation of three assay systems for automated determination of hemoglobin A1c. Clin Chem 1997;43:511–7

Shimizu

, Nakanishi

, Miyazaki

. Detection and characterization of variant and modified structures of proteins in blood and tissues by mass spectrometry. Mass Spectrom Rev 2006;25:686–712

Koga

, Kasayama

. Clinical impact of glycated albumin as another glycemic control marker. Endocr J 2010;57:751–62

Koga

, Otsuki

, Matsumoto

, Saito

, Mukai

, Kasayama

. Negative association of obesity and its related chronic inflammation with serum glycated albumin but not glycated hemoglobin levels. Clin Chim Acta 2007;378:48–52

Bando

, Kanehara

, Kasayama

, Koga

. Association of serum glycated albumin to glycated haemoglobin A1C ratio with hepatic function tests in patients with chronic liver disease. Ann Clin Biochem 2009;46:368–72

10.

Nakanishi

, Miyazaki

, Shimizu

, Yamaguchi

, Nishimura

. Assessment of the effect of hemoglobin variants on routine HbA_1c measurements by electrospray ionization mass spectrometry. Clin Chim Acta 2002;323:89–101

11.

The Committee of Japan Diabetes Society on the diagnostic criteria of diabetes mellitus. Report of the Committee on the classification and diagnostic criteria of diabetes mellitus. J Diabetes Invest 2010;1:212–28

12.

Shima

, Ito

, Abe

, Hirota

, High-performance liquid chromatographic assay of serum glycated albumin. Diabetologia 1988;31:627–31

13.

Harano

, Harano

, Shibata

, Hb J Oxford [alpha 15 (A13) Gly→Asp] in Japan. Hemoglobin 1984;8:197–8

14.

Schiliro

, Rizzari

, Testa

, Lo Faro

, Comisi

, Russo

. Association of Hb S [beta 6(A3)Glu—-Val] and Hb I-interlaken [alpha 15(A13)Gly—-Asp] in a Sicilian man; review of the occurrence of Hb I-interlaken in Sicily. Hemoglobin 1989;13:403–10

15.

Blackwell

, Wong

, Wang

, Weng

, Liu

. Hemoglobin J Meerut: alpha120 Ala leads to Glu. Biochim Biophys Acta 1974;351:7–12

16.

Ringelhann

, Konotey-Ahulu

, Talapatra

, Nkrumah

, Wiltshire

, Lehmann

, Haemoglobin

. Woolwich (alpha 2, beta 2 132 lysine leads to glutamine) in Ghana. Acta Haematol 1971;45:250–8

17.

Ohba

, Miyaji

, Murakami

, Hb Himeji or beta 140 (H18) Ala→Asp. A slightly unstable hemoglobin with increased beta N-terminal glycation. Hemoglobin 1986;10:109–25

18.

Nishihara

, Koga

, Kadowaki

, Method-dependent HbA_1c values in a family with hemoglobin Himeji. Clin Chim Acta 2011;412:1689–92

19.

Nakanishi

, Miyazaki

, Kishikawa

, Shimizu

, Aoki

, Kikuchi

. Hb Peterborough [beta 111(G13)Val→Phe] in Japan; rapid identification by ESI/MS using proteolytic digests of oxidized globin. Hemoglobin 1998;22:23–35

20.

Armbruster

. Fructosamine: structure, analysis, and clinical usefulness. Clin Chem 1987;33:2153–63

21.

Jeffcoate

. Diabetes control and complications: the role of glycated haemoglobin, 25 years on. Diabet Med 2004;21:657–65

22.

Coban

, Ozdogan

, Timuragaoglu

. Effect of iron deficiency anemia on the levels of hemoglobin A1c in nondiabetic patients. Acta Haematol 2004;112:126–8

Classification of variant forms of haemoglobin according to the ratio of glycated haemoglobin to glycated albumin

Abstract

Background

Methods

Results

Conclusions

Introduction

Subjects and methods

Subjects and controls

Laboratory measurements

Statistical analysis

Results

Discussion

DECLARATIONS

References