Diabetes Research and Clinical Practice 64 (2004) 181–183
Glibenclamide-induced acute haemolytic anaemia revealing a G6PD-deficiency Stéphane Vinzio a,∗ , Emmanuel Andrès b , Anne-Elisabeth Perrin a , Jean-Louis Schlienger a , Bernard Goichot a b
a Service de Médecine Interne et de Nutrition, Hˆ opital de Hautepierre, 67098 Strasbourg Cedex, France Service de Médecine Interne, Diabète et Maladies Métaboliques, Hˆopital Civil, 67091 Strasbourg Cedex, France
Received in revised form 5 November 2003; accepted 12 November 2003
Abstract A 58-year-old woman was admitted at diagnosis of type 2 diabetes without keto-acidosis. Blood glucose was normalized initially with insulin. Whilst taking glibenclamide, she developed acute haemolysis. She was homozygous for glucose-6-phosphate dehydrogenase (G6PD) deficiency and had no other factors predisposing haemolysis. We reviewed the literature and discuss the relationship between glibenclamide and haemolytic crisis and between G6PD-deficiency and diabetes. © 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Diabetes mellitus; Glucose-6-phosphate dehydrogenase deficiency; Glibenclamide; Acute haemolytic anaemia; Oxidative stress
1. Introduction
2. Case report
Various drugs and infections may trigger acute haemolytic anaemia in glucose-6-phosphate dehydrogenase (G6PD) deficiency [1]. Oral hypoglycaemic drugs are not classically involved but particular circumstances could precipitate haemolysis among diabetic patients: hypoglycaemia, keto-acidosis and corrections of hyperglycaemia. We report a case of glibenclamide-induced acute haemolytic anaemia in a G6PD-deficient woman admitted at diagnosis of type 2 diabetes. We review the literature and discuss the relationship between G6PD-deficiency and diabetes.
A 58-year-old female of Burkina Faso origin, was admitted with polyuria, polydipsia and weight loss of 4 kg of 2 months duration. She was not on any medications and had no major illness in the past. She had no family history of diabetes mellitus or G6PD-deficiency. Physical examination was normal without fever, pallor, jaundice or pain. Fasting plasma glucose was 34.5 mmol/l, HbA1C 15.9% without ketonuria. There was iron overload: sideremia 27 mol/l (range 9–30), siderophiline saturation rate 78% (normal < 35) and ferritin 516 ng/ml (range 9–120) without signs of haemolysis (Hb 12.2 g/dl, MCV 91 fl, 106,010 reticulocytes/mm3 , total bilirubin 15.1 mol/l) or haemoglobin electrophoresis abnormality (Hb A 96.6%, A2 2.4%, F 1%). Screening
∗ Corresponding author. Tel.: +33-3-88-12-75-90; fax: +33-3-88-12-75-96. E-mail address:
[email protected] (S. Vinzio).
0168-8227/$ – see front matter © 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2003.11.006
182
S. Vinzio et al. / Diabetes Research and Clinical Practice 64 (2004) 181–183
for HFE mutation (primary hemochromatosis) was negative. Intravenous insulin therapy and diabetic diet were started. After blood glucose was controlled, insulin was withdrawn and metformin (850 mg three times a day) and glibenclamide (5 mg twice daily) were introduced. Five days later nausea, vomiting associated with abdominal and back pains occurred without fever or clinical evidence of infection or gastrointestinal bleeding. Haemoglobin level fallen to 9.4 g/dl with 204,600 reticulocytes/mm3 (normal < 100,000) or 6.6% (normal < 1.8) without schizocytes or Heinz bodies on peripheral smear. White cell count, C-reactive protein and bilirubin were in the normal range. Haptoglobin was normal (0.67 g/l, range 0.34–2.00) and Coombs test gave a negative result. Antinuclear antibody and cryoglobulinemia were absent. Cobalamin, folate levels and fasting blood glucose were in normal range (respectively, 1.27 g/l, 5.1 g/l, 4.8 mmol/l). Pyruvate-kinase and phosphohexose-isomerase levels were also normal but G6PD was very low at 0.4 UI/g Hb (range 4.0–6.8) corresponding to a homozygous deficiency. Molecular study of the G6PD gene (by PCR and reverse dot blot) showed homozygosity for the mutations 376 A → G and 202 G → A (G6PD A-variant). Abdominal ultrasound and gastroscopy gave normal results. Glibenclamide was withdrawn and blood tests returned to normal. One month after discharge, haemoglobin and haptoglobin rose, respectively, to 11.5 g/dl and 0.96 g/l whilst taking metformin (850 mg three times a day) associated with a stringent diabetic diet.
3. Discussion In a G6PD-deficient subject, haemolysis may occur as a result of ingestion of various drugs, infection and more specifically among diabetic subjects, hypoglycaemia [2], blood glucose normalisation [3,4] or keto-acidosis [5], although the later is controversial [6]. All these conditions were absent in our case and blood glucose normalisation had been achieved for several days when acute haemolysis appeared. Moreover after glibenclamide was withdrawn, haemoglobin steadily rose to normal and haemolysis did not recurr. Glibenclamide is a second-generation sulphonylurea derived from sulphonamides, with hy-
poglycaemic action without antibacterial or diuretic properties. Haematological side effects of sulphonylureas include autoimmune haemolytic anaemias [7]. In our case there was no biological argument for an autoimmune mechanism. The delay between first exposure and haemolysis was brief and direct and indirect antiglobin tests were negative. Moreover a similar case has been described with a man of Sardinian origin who was known to have diabetes and G6PD-deficiency (Mediterranean type). He presented with acute haemolysis after glibenclamide was begun [8]. Therefore we consider that glibenclamide could initiate haemolysis in G6PD-deficient erythrocytes and should be use with caution. Concerning metformin, three cases of metformin-induced haemolysis have been reported in the literature, only one with significantly decreased G6PD activity [9]. Because haemoglobin level remained stable whilst our patient took only metformin regimen, a metformin-induced haemolysis seems unlikely. The initial iron overload probably reflected a chronic and low level haemolysis related to G6PD-deficiency. Since the first description of G6PD-deficiency [10] and a first anecdotal report of diabetes-G6PDdeficiency association in 1964 [11], many studies have shown a statistically significant relationship between G6PD-deficiency and the diabetic state. In 1969, Eppes et al. [12] reported higher blood glucose levels after intravenous glucose tolerance test in 10 G6PDdeficient non diabetic subjects compared with ten non deficient subjects. Monte Alegre et al. [13] demonstrated a lower first phase (first 30 min) insulin level after oral or intravenous glucose tolerance test, emphasising the role of G6PD in glucose-induced insulin release. Epidemiological studies have demonstrated a positive association between G6PD-deficiency and type 2 diabetes in the Chinese and Indian populations of Singapore [14]. Similarly an excessive prevalence of G6PD-deficiency has been reported among diabetic subjects of different non-Mediterranean populations: Saudi (12.4% versus 2.0%) [15], and Africans (19.4% versus 10.6%) [16]. Recent advances about oxidative stress and glutathione metabolism provide interesting hypothesis concerning diabetes and G6PD-deficiency association. The administration of oxidant drugs (alloxan, streptozocin) is used to induce diabetes in animals, illustrating the sensitivity of pancreatic beta-cells
S. Vinzio et al. / Diabetes Research and Clinical Practice 64 (2004) 181–183
to oxidants compared with other somatic cells. In red cells glutathione and glutathione peroxidase are the main route of elimination of hydrogen peroxide (H2 O2 ). The NADPH needed as reductant is provided by G6PD. Thus in G6PD-deficient subjects, H2 O2 is periodically raised resulting in deficient red cell haemolysis. As beta-pancreatic cells are highly sensitive to oxidative stress, West postulates that a regular excess of oxidant related to G6PD-deficiency may lead to diabetes, pre-existing oxidative stress becoming a part of diabetes development and not only a consequence [17]. Moreover, while catalase is a second route of disposal H2 O2 in red cells, Goth and Eaton [18] demonstrate that the prevalence of diabetes is increased in patients will catalase deficiencies. Many studies have demonstrated a higher prevalence of G6PD-deficiency among non-Mediterranean diabetic subjects. This relation still remains unexplained but episodic H2 O2 accumulation related to G6PD-deficiency may damage beta-cells that are more sensitive to oxidant stress than other somatic cells. Thus screening for G6PD-deficiency could be suggested in diabetic subjects from particular ethnic origins. Oral hypoglycaemic drugs are not classically involved in acute haemolytic anaemia in G6PD-deficient subjects. However our case report shows that they should be used with heightened awareness and will careful monitoring in the weeks following the onset of treatment.
References
[4]
[5]
[6]
[7] [8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[1] B. Beutler, G6PD deficiency, Blood 84 (1994) 3613–3636. [2] O. Shalev, R. Eliakim, G.Z. Lugassy, J. Menczel, Hypoglycemia-induced haemolysis in glucose-6-phosphate dehydrogenase, Acta Haematol. 74 (1985) 227–229. [3] M. Vanelli, L. Lucentini, P. Picco, et al., Blood glucose normalization-induced haemolysis in three adolescents with
[17] [18]
183
type 1 diabetes mellitus at onset and unknown G-6-PD deficiency, J. Pediatr. Endocrinol. Metab. 9 (1996) 193–196. A. Galli-Tsinopoulou, S. Nousia-Arvanitakis, Glucose-6phophate dehydrogenase deficiency-induced haemolysis in newly diagnosed diabetic monozygotic twins, J. Pediatr. Endocrinol. Metab. 13 (2000) 669–672. A.M. Gellady, R.D. Greenwood, G-6-PD haemolytic anaemia complicating diabetic ketoacidosis, J. Pediatr. 80 (1972) 1037–1038. O. Shalev, A. Wollner, J. Menczel, Diabetic ketoacidosis does not precipitate haemolysis in patient with the Mediterranean variant of glucose-6-phosphate dehydrogenase, Br. Med. J. (Clin. Res. Ed.) 288 (1984) 179–180. S.L. Abbate, B.J. Hoogwerf, Haemolytic anaemia associated with sulfonylurea use, Diabetes Care 13 (1990) 904–905. G. Meloni, T. Meloni, Glyburide-induced acute haemolysis in a G6PD-deficient patient with NIDDM, Br. J. Haematol. 92 (1996) 159–160. A. Meir, Y. Kleinman, D. Rund, N. Da’as, Metformin-induced hemolytic anemia in patient with gluco-6-phosphate dehydrogenase deficiency, Diabetes Care 26 (2003) 956–957. P.E. Carson, C.L. Flanagan, C.E. Ickes, A.S. Alving, Enzymatic deficiency in primaquine-sensitive erythrocytes, Science 124 (1956) 484–485. D. Chanmugan, A.M. Frumin, Abnormal oral glucose tolerance response in erythrocyte G6PD deficiency, N. Engl. J. Med. 271 (1964) 1202–1204. R.B. Eppes, A.M. Lawrence, J.V. McNamara, M.D. Powell, P.E. Carson, Intravenous glucose tolerance in Negro men deficient in glucose-6-phosphate dehydrogenase, N. Engl. J. Med. 281 (1969) 60–63. S. Monte Alegre, S.T. Saad, E. Delatre, M.J. Saad, Insulin secretion in patients deficient in glucose-6-phosphate dehydrogenase, Horm. Metab. Res. 23 (1991) 171–173. N. Saha, Association of glucose-6-phosphate dehydrogenase deficiency with diabetes mellitus in ethnic groups of Singapore, J. Med. Genet. 16 (1979) 431–434. G.A. Niazi, Glucose-6-phosphate dehydrogenase deficiency and diabetes mellitus, Int. J. Haematol. 54 (1991) 295–298. T.K. Saeed, H.A. Hamamy, A. A Alwan, Association of glucose-6-phosphate dehydrogenase deficiency with diabetes mellitus, Diabetes Med. 2 (1985) 110–112. I.C. West, Glucose-6-phosphate dehydrogenase: a candidate gene for diabetes, Diabetes Med. 19 (2000) 172–173. L. Goth, J.W. Eaton, Hereditary catalase deficiencies and increased risk of diabetes, Lancet 356 (2000) 1820–1821.
