sign in sign up
<<
Home , Methionine

LETTERS

blood glucose concentrations were mea- lispro at 40 Ϯ 3 min and aspart at 49 Ϯ 3 OBSERVATIONS sured, as indicated in Fig. 1. If blood glu- min after injection, respectively (P ϭ cose was 3.5 mmol/l or lower, 20 ml 0.01). The maximum concentra- glucose 30% was injected. One patient tion was 316 Ϯ 31 pmol/l on insulin lis- Direct Comparison of was excluded from the analysis because, pro and 295 Ϯ 27 pmol/l on insulin by mistake, he took a large extra dose of aspart (NS) (Fig. 1). The increase from 0 and insulin the night before the study. to 15 min after injection was 109 Ϯ 17 Aspart Shows Small Free insulin was measured after poly- pmol/l after injection of insulin lispro and Differences in glycol precipitation by Mercodia 53 Ϯ 11 pmol/l after injection of insulin Iso-Insulin (ELISA; Mercodia AB, Uppsala, Plasma (P ϭ 0.02). Fasting insulin lispro Sweden), a two-site immunoassay levels reached 50% of peak concentration Profiles After containing two monoclonal antibodies at 20 Ϯ 1 min and aspart at 30 Ϯ 3 min against insulin. Identical results were ob- (P ϭ 0.02) (Fig. 2). The decrease of free Subcutaneous tained when equimolar concentrations of Injection in Type 1 insulin concentration from peak concen- human insulin, insulin lispro, and insulin tration to 50% of the maximum concen- aspart were tested, indicating 100% Diabetes tration was found at 113 Ϯ 10 min during cross-reactivity between lispro, aspart, insulin lispro and 154 Ϯ 14 min during and human insulin in this assay. insulin aspart (P ϭ 0.02) (Fig. 2). The Blood glucose was analyzed with the fasting blood glucose concentration was wo rapid-acting insulin analogs, lis- Hemocue method (Hemocue, Mission Vie- 11.2 Ϯ 1.0 mmol/l before injection of in- pro and aspart, are now available for jo, CA). Serum IGF binding –1 was sulin aspart and 13.7 Ϯ 1.4 mmol/l before clinical use (1). determined by an immuno-enzymometric T injection of insulin lispro (NS). The The aim of our study was to compare assay with a kit from Medix Biokemica the plasma-insulin profiles of these ana- (Kauniainen, Finland). Differences among course of the blood glucose profiles was logs after subcutaneous injection in pa- groups were tested with Wilcoxon’s similar, with peak concentrations after 40 tients with type 1 diabetes. signed-rank test. Areas under the curve min, and there was no difference between Fourteen patients with type 1 diabe- (AUC) were calculated with the trapezoi- total AUC. tes (six men and eight women, [mean Ϯ dal method. The concentration of fasting IGFBP-1, Ϯ SEM] 35.4 Ϯ 3.3 years of age [range 22– The fasting-free insulin concentration a liver-derived protein (2), was 12.3 3.4 Ϯ Ϯ ␮g/l on insulin aspart and 16.8 Ϯ 3.9 ␮g/l 59], HbA1c 7.3 0.3% [reference range was 56 16 pmol/l before administration 3.2–5.4], BMI 24.7 Ϯ 1.1 kg/m2, and di- of insulin aspart and 30 Ϯ 15 pmol/l be- on insulin lispro. The concentrations fell to abetes duration 22.9 Ϯ 2.6 years) were fore administration of insulin lispro (NS) 8.3 Ϯ 1.4 and 13.5 Ϯ 4.5 ␮g/l 2.5 h after recruited for the study. Only two patients (Fig. 1). Both insulin analogs gave marked injection and to 7.4 Ϯ 1.6 and 11.5 Ϯ 4.1 had measurable C- levels (0.04 peaks of free insulin concentrations— ␮g/l after 5.5 h, respectively (all differ- and 0.11 nmol/l). All patients were on multiple-injection therapy with a break- fast insulin dose of 11.1 Ϯ 0.7 U (range 6–14) and no intermediate-acting insulin in the morning. The study was designed as a single blind randomized crossover study. On the first day, seven patients were random- ized to insulin lispro (Humalog, U-100; Eli Lilly, Indianapolis, IN), and the other seven were randomized to insulin aspart (NovoRapid, U-100; Novo-Nordisk, Bagsveard, Denmark). On the second study day (5–21 days later), the patients received the alternative insulin analog. The patients continued their usual insulin treatment between the study days. All patients arrived to the clinic fast- ing. After an initial blood sampling, they were given 10 U s.c. of one of the in the abdominal wall at 7:30 A.M. A stan- dardized breakfast, which had an energy content of 418 kcal and a nutri- ent content of 21 g protein, 11 g fat, and Figure 1—Plasma concentrations of free insulin in 13 patients with type 1 diabetes after a 10 U 59 g carbohydrates, was served immedi- single subcutaneous injection of insulin lispro (⅜) and insulin aspart (ⅷ) at 7:30 A.M. immediately ately thereafter. Plasma free insulin and before breakfast. The values are means Ϯ SEM.

1120 DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 Letters

tively rapid rise of free insulin levels after 3. Home PD, Barriocanal L, Lindholm A: injection of insulin lispro might be an ad- Comparative pharmacokinetics and phar- vantage. The slightly slower decrease of macodynamics of the novel rapid-acting free insulin concentrations after the insu- insulin analogue, insulin aspart, in healthy lin peak of insulin aspart might influence volunteers. Eur J Clin Pharmacol 55:199– 203, 1999 the need for daytime basal insulin and the 4. Heinemann L, Heise T, Jorgensen LN, need for a snack between the main meals Starke AA: Action profile of the rapid of certain patients. acting insulin analogue: human insulin We previously found no difference in B28Asp. Diabet Med 10:535–539, 1993 IGFBP-1 concentrations when using hu- 5. Kang S, Creagh F, Peters JR, Brange J, Vo¨l- man regular insulin or insulin lispro (7), und A, Owens DR: Comparison of sub- and in this study there was no difference cutaneous soluble human insulin and between insulin aspart and lispro. insulin analogues (AspB9, GluB27; AspB10; B28 The main finding of our study is that Asp ) on meal-related plasma glucose the free insulin profiles of aspart and lis- excursions in type 1 diabetic subjects. Di- pro resemble each other, but insulin lis- abetes Care 14:571–577, 1991 Figure 2— Time (minutes) from subcutaneous 6. Lefebvre PJ, Scheen AJ: The postprandial injection of 10 U of insulin lispro (Ⅺ) and in- pro shows a more rapid uptake, reaches state and risk of cardiovascular disease. sulin aspart (f) from fasting levels to 50% of the maximum peak concentration earlier, Diabet Med 15 (Suppl. 4):S63–S68, 1998 the peak free insulin concentration, peak con- and shows a more rapid decline than in- 7. Hedman CA, Lindstrom T, Arnqvist HJ: centration, and 50% decrease from peak con- sulin aspart. We believe this finding may Treatment with insulin lispro changes the centration in 13 patients with type 1 diabetes. be of clinical importance. insulin profile but does not affect the The values are means Ϯ SEM. plasma concentrations of insulin-like 1 CHRISTINA A. HEDMAN MD growth factor-1 and insulin-like growth- 2 TORBJO¨RNLINDSTRO¨MMD, PHD factor binding protein-1 in type 1 diabe- 1,2 ences between the insulin analogs were HANS J. ARNQVIST MD tes. Clin Endocrinol. In Press nonsignificant). The time from injection to half From the Division of Internal Medicine, the 1Depart- ment of Medicine and Care and the 2Division of the maximum value and to peak insulin Cellbiology, Department of Biomedicine and Sur- Erythropoietin concentration was significantly shorter gery, Faculty of Health Sciences, Linko¨ping Univer- Treatment of during insulin lispro, indicating a faster sity, Linko¨ping, Sweden. absorption of this insulin. Address correspondence to Christina Hedman, Postural The impression from previous studies MD, Division of Internal Medicine, Department of Medicine and Care, The University Hospital, SE-581 Hypotension in is that insulin aspart gives a somewhat 85, Linko¨ping, Sweden. E-mail: christina.hedman@ Anemic Type 1 broader peak than lispro (3,4). We found lio.se. a significantly more rapid lowering of Diabetic Patients the free insulin concentration to 50% of With Autonomic the peak concentration with lispro. The Acknowledgments— Financial support was plasma profile of free insulin obtained is obtained from the Swedish Medical Research Neuropathy dependent on the rate of absorption of Council (04952), the Swedish Diabetes Asso- ciation, Barndiabetesfonden, and the County insulin from subcutaneous tissue and the of O¨ stergo¨tland, Sweden. No support was ob- A case study of four patients elimination of insulin from the circula- tained from any pharmaceutical company for tion. Because the metabolic clearance of this study. insulin analogs is predominately receptor We thank research nurses Ewa Bergstedt, ostural hypotension (PH) is a feature mediated (5) and because lispro and as- Christina Dahlgren, and Kerstin Krogager and of autonomic failure that may result part have about the same affinity for the dieticians Karin Fridell, Ann-Charlotte, and P from various causes and can be very (1), it seems probable Orre Pettersson for their excellent help. difficult to treat. Recently, we and others that they will have about the same clear- (1–3) have shown that some patients with ance rate and that the lowering of free in- ●●●●●●●●●●●●●●●●●●●●●●● severe symptomatic autonomic neuropa- sulin after the peak is mainly due to the References thy from type 1 diabetes (type 1 DAN) have rate of absorption. 1. Kurtzhals P, Scha¨ffer L, So¨rensen A, Kris- a normocytic associated with eryth- From the clinical point of view, the tensen C, Jonassen I, Schmid C, Tru¨b T: ropoietin (EPO) deficiency. The treatment most important observation is that insulin Correlations of receptor binding and met- of these patients with EPO rapidly corrects analogs are absorbed much faster than abolic and mitogenic potencies of insulin their anemia (1) and improves their over- human insulin after subcutaneous in- analogs designed for clinical use. Diabetes all well-being. Preliminary observations by jection, with higher insulin peaks and 49:999–1005, 2000 Hoeldtke et al. (3) of the effect of a maxi- 2. Conover CA, Butler PC, Wang M, Rizza shorter duration of action (3,4). Because RA, Lee PDK: Lack of mum of 9 weeks of EPO treatment on high postprandial glucose and lipid levels effect on insulin-associated suppression blood pressure (BP) demonstrated a pos- have been emphasized as risk factors, es- of insulinlike growth factor binding pro- itive response in four cases of PH in type 1 pecially in type 2 diabetes (6), this may be tein 1 in humans. Diabetes 39:1251–1256, DAN patients, but evaluation of BP was of importance. In this respect, the rela- 1990 limited to only two to three readings be-

DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 1121 Letters

Figure 1—Monthly average systolic BP on standing during the control month, the third month on EPO, and third month off EPO at the four different times of the day. E, patient 1; ᭛, patient 2; Ⅺ, patient 3; ‚, patient 4. For statistical analysis, see text. fore and after EPO treatment. These re- months on treatment, and 3 months off nificance of the BP response to treatment sults stimulated the present study, in which treatment). The study was stopped in pa- with EPO was determined by fitting a we investigated the effect of EPO treatment tient 1 after 3 months on EPO because of nonlinear function of time to each pa- for 3 months on supine and standing BP in serious problems from foot sepsis. Three tient’s series of mean BP data (4). The ap- four anemic EPO-deficient type 1 DAN patients were already on fludrocortisone, plied function modelled a constant BP patients. the dose of which was kept constant dur- during the control phase, followed by a The patients were treated with recom- ing the study. gradual increase after EPO administration binant human EPO (25 IU/kg s.c. thrice During the 7- month study, the pa- and a gradual decrease after withdrawal of weekly). All four patients were women tients performed regular BP readings EPO. This model of assumed BP response (48, 41, 30, and 30 years of age with a themselves using an automatic sphygmo- was developed on the basis of observa- duration of diabetes of 13, 24, 19, and 11 manometer (Omron 2000) after undergo- tions made in patients with chronic renal years, respectively) and each had normo- ing several training sessions according to failure during EPO treatment; it is specific cytic anemia with inappropriately low se- a standardized protocol. The BP record- in that it only detects BP responses attrib- rum EPO levels. The known causes of ings were taken before meals and insulin utable to the EPO stimulus and therefore anemia were excluded. All of the patients injections at four different times (morn- excludes measurement errors. included in the study had pronounced ing, midday, afternoon, and bedtime) Pronounced symptomatic PH was symptomatic PH and at least one other thrice weekly. The supine BP reading was present in the four patients during the symptom of autonomic neuropathy. The compared with the lowest value within 5 control month (monthly average systolic autonomic function tests were severely min of standing. Each patient took be- postural fall: 44.9 Ϯ 3.5, 41.4 Ϯ 6.4, abnormal. The patients were followed up tween 672 and 1,584 BP readings during 52.3 Ϯ 5.1, and 21.7 Ϯ 4.9 mmHg, re- for 7 months (1-month control phase, 3 the course of the study. The statistical sig- spectively). EPO treatment for 3 months

1122 DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 Letters resulted in an increase in standing systolic changes, although EPO-increasing red 495–499, 2001 and diastolic BP in all four patients at all cell mass does not normally increase 3. Hoeldtke RD, Streeten DHP: Treatment of four times throughout the day (Fig. 1). blood volume (8). Despite the above- orthostatic hypotension with erythropoi- The BP decreased again after EPO with- mentioned indirect effects of EPO, some etin. N Engl J Med 329:611–615, 1993 drawal (Fig. 1). After 3 months of EPO authors postulate that EPO exhibits direct 4. Seber GAF, Wild CJ: Nonlinear Regression. New York, Wiley, 1998 treatment, standing systolic BP in the four pressor effects on vascular smooth muscle 5. Jandeleit K, Heintz B, Gross-Heitfeld E, patients increased from 5 to 30, 6 to 15, 5 cells (9) or enhances DNA synthesis in Kindler J, Sieberth HG, Kirsten R, Nelson to 13, and 9 to 17 mmHg in the morning, smooth muscle cells, thereby contribut- K: Increased activitiy of the autonomic midday, afternoon, and at bedtime, re- ing to vascular hypertrophy (10). EPO nervous system and increased sensitivity spectively. The standing mean BP in the might also have a direct trophic effect on to angiotensin II infusion after therapy morning increased significantly in all pa- the nervous system, as shown recently in with recombinant human erythropoietin tients (P ϭ 0.001–0.036). There was also neuroblastoma cells (11). These interest- (Letter). Nephron 56:220–221, 1990 a significant increase in standing mean BP ing observations need to be developed, 6. Martin W, Smith JA, White DG: The in patients 2–4 (P ϭ 0.001), patients 3 and their relevance to effects of EPO on BP mechanisms by which hemoglobin inhib- and4(P ϭ 0.001 and 0.003), and pa- in our patients remains speculative. its the relaxation of rabbit aorta induced ϭ by nitrovasodilators, nitric oxide, or bo- tients 1, 2, and 4 (P 0.001–0.025) at In summary, we have demonstrated vine retractor penis inhibitory factor. Br J midday, in the afternoon, and at bedtime, that treatment with EPO increases the Pharmacol 89:563–571, 1996 respectively. There was usually a small standing BP in diabetic patients with se- 7. Raine ARG: Hypertension, blood viscos- and highly variable increase in supine BP, vere autonomic neuropathy. This effect, ity, and cardiovascular morbidity in re- so that the effect of EPO treatment on the combined with the correction of the ane- nal failure: implication of erythropoietin actual BP drop was also highly variable. mia, can lead to considerable improve- therapy. Lancet i:97–100, 1988 EPO treatment corrected the anemia in all ment in the well-being of these severely 8. Cotes PM, Pippard MJ, Reid CDL, Winearls four patients, and hemoglobin increased disadvantaged patients and may provide CG, Oliver DO, Royston JP: Characteriza- from 116, 99, 100, and 115 g/l to 140, 131, an additional treatment for seriously af- tion of the anemia of chronic renal failure 124, and 132 g/l, respectively. Hemoglo- fected patients when other measures have and the mode of its correction by a prep- aration of human erythropoietin (r-Hu bin decreased to the baseline level by the failed. EPO). Quart J Med 70:113–137, 1989 end of the third month off treatment. 9. Heidenreich S, Rahn KH, Zidek W: Direct 1 PH in diabetic autonomic neuropathy ANDREA S. WINKLER, PHD vasopressor effect of recombinant human 2 is extremely variable and difficult to as- SABINE LANDAU, PHD erythropoietin on renal resistance vessels. 3 sess. The study of treatment is thus ren- PETER J. WATKINS, MD, FRCP Kidney Int 39:259–265, 1989 dered exceptionally difficult. We believe 10. Gogusev J, Zhu DL, Herembert T, Am- that by using numerous objective mea- From the 1Diabetes Centre and the 2Movement Dis- marguellat F, Marche P, Drueke T: Effect orders and Autonomic Unit, King’s College Hospi- of erythropoietin on DNA synthesis, surements of BP (up to 1,584 readings), tal; and the 3Department of Biostatistics and we have reduced potential bias and proto-oncogene expression and phospho- Computing, Institute of Psychiatry, London, U.K. lipase C activity in rat vascular smooth present valid results. The greatest benefit Address correspondence to Dr. P.J. Watkins, Di- muscle cells. Biochem Biophys Res Commun abetes Centre, King’s College Hospital, Denmark to our patients was the increase in stand- 199:977–983, 1994 ing BP, which was significant at almost all Hill, London SE5 9RS, U.K. E-mail: peter.watkins1@ virgin.net 11. Campana WM, Misasi R, O’Brian JS: Iden- times throughout the day. This was ac- tification of a neurotrophic sequence in companied by a decrease in dizziness on erythropoietin. Int J Mol Med 1:235–241, standing, though we appreciate that this Acknowledgments— This study was sup- 1989 was an open study and therefore prone to ported by a grant from the British Heart Foun- bias. Of course, some of the improvement dation (RLN TAM, to Dr. K. Ray Chaudhuri in overall well-being might be attributed and Dr. Watkins) and by Janssen-Cilag Ltd. to the improvement of the anemia in all We are grateful to Dr. Chaudhuri, Department COMMENTS AND the patients. of Neurology; Dr. Henry Hambley, Depart- The effect of EPO in raising standing ment of Hematology; Dr. Iain Macdougall, De- RESPONSES partment of Nephrology; and Prof. Tim Peters, BP might be mediated through a number Department of Clinical at King’s of different mechanisms. Vasoconstriction College Hospital, London, for their consider- might occur after increased norepineph- able help. Hyperhomocys- rine levels after injection of EPO (5) or as teinemia and a result of increased binding of nitric ox- ●●●●●●●●●●●●●●●●●●●●●●● ide by the greater concentration of hemo- Macroangiopathy in globin, thus reducing its capacity to References Type 2 Diabetes vasodilate (6). Furthermore, elevated lev- 1. Winkler AS, Marsden J, Hambley H, els of hemoglobin may increase blood vis- Chaudhuri KR, Watkins PJ: Erythropoie- tin depletion and anemia in diabetes. Dia- cosity and thereby augment peripheral bet Med 16:813–819, 1999 levated is widely re- vascular resistance (7). EPO may also in- 2. Bosman DR, Winkler AS, Marsden JT, garded as an independent risk factor crease vascular sensitivity to angiotensin II, McDougall IC, Watkins PJ: Anemia with E for macrovascular disease, particu- thus further promoting vasoconstriction erythropoietin deficiency occurs early in larly in patients with type 2 diabetes (1). (5). BP is very sensitive to blood-volume diabetic nephropathy. Diabetes Care 24: The increased prevalence of macroangi-

DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 1123 Letters opathy in type 2 diabetes is further dem- (12). While the debate goes on regarding therapy for homocysteine. Lancet 354: onstrated by Buysschaert et al. (2). The the potential cardiovascular benefit of 1208–1209, 1999 authors found a significant increase in the HRT in general use, its selective use in 8. Van Baal WM, Smolders RGV, Van der prevalence of macroangiopathy in type 2 subjects with high homocysteine concen- Mooren MJ, Teerlink T, Kenemans P: diabetic patients with hyperhomocys- trations may be justified. It would there- Hormone replacement and plasma homo- teinemia. Their data, however, should not fore be important for Buysschaert et al. to levels. Obstet Gynecol 94:485– 491, 1999 be overinterpreted because the etiological clarify the distribution of homocysteine 9. Walsh BW, Paul S, Wild RA, Dean RA, role of homocysteine in in and concentrations in relation to Tracy RP, Cox DA, Anderson PW: The the presence of macroangiopathy is far the prevalence of HRT usage in their effects of hormone replacement therapy from clear (3). study. and raloxifene on C-reactive protein and First, the of homocys- In summary, there is little doubt that homocysteine in healthy postmenpausal teine is altered for a period of several days homocysteine is implicated in the patho- women: a randomized, controlled trial. following acute ischemic events, such as genesis of diabetic macroangiopathy. The J Clin Endocrinol Metab 85:214–218, 2000 myocardial infarction (4) and stroke (5), correlation of its level would be most 10. Blom HJ, Boers GHJ, Elzen JPAM, van as a result of an increased release from valuable in primary prevention studies Roessel JJM, Trijbels JMF, Tangerman A: damaged tissues. This is caused by methy- because its concentration in patients with Difference between premenopausal wom- lation of DNA, RNA, and various , macroangiopathy is difficult to interpret. en and young men in the transamination leading to an increase in S-adenosylho- pathway of catabolism and mocysteine and subsequently to in- the protection against vascular disease. creased homocysteine production (3). N. NORMAN CHAN, MRCP DCH Euro J Clin Invest 18:633–638, 1988 11. Finkelstein J: Methionine metabolism in Furthermore, nephropathy may reduce TRICIA M.M. TAN, MRCP STEVEN J. HUREL, PHD MRCP mammals: effect of age, , and hor- homocysteine clearance in type 2 diabetic mones on three of the pathway patients. To our knowledge, there are no in rat tissues. Arch Biochem Biophys 122: published studies examining the time From the Department of Diabetes, Endocrinology, 583–590, 1972 course of homocysteine concentrations and Metabolism, University College London Hospi- tals, London, U.K. 12. Olszewski AJ, McCully KS: Homocysteine following acute ischemia, which may be Address correspondence to Dr. N. Norman Chan, content of lipoproteins in hypercholester- silent. Hence, assessing homocysteine Department of Diabetes, Endocrinology and Metab- olemia. Atherosclerosis 88:61–68, 1991 levels in patients with preexistent mac- olism, South House, The Middlesex Hospital, Mor- roangiopathy may produce misleading re- timer Street, London,W1N 8AA, U.K. E-mail: sults. This may in part account for the [email protected]. conflicting results regarding the associa- Response to Chan tion of homocysteine levels with insulin ●●●●●●●●●●●●●●●●●●●●●●● et al.: hyperhomo- resistance that have been reported in var- ious studies (6). References cysteinemia and Second, in their study, Buysschaert et 1. Hoogeveen EK, Kostense PJ, Beks PJ, Mac- Kaay AJC, Jakobs C, Bouter LM, Heine RJ, macroangiopathy al. (2) have rightly taken into account the Stehouwer CDA: in type 2 diabetes prevalence of potential confounders of is associated with an increased risk of plasma homocysteine concentrations, cardiovascular disease, especially in non- namely fibrate and metformin therapy insulin-dependent diabetes mellitus: e thank Chan et al. (1) for their (2,7). However, the authors failed to com- a population-based study. Arterioscler ment on the use of hormone replacement Thromb Vasc Biol 18:133–138, 1998 interest in our work (2) and their therapy (HRT) in their cohort of patients 2. Buysschaert M, Dramais A-S, Wallemacq W relevant comments. We agree (mean age 63 Ϯ 10 years), the majority of PE, Hermans MP: Hyperhomocysteine- that there is still much to be learned about whom were women (82 of 122). Com- mia in type 2 diabetes: relationship to the relationship between homocysteine bined - HRT has macroangiopathy, nephropathy, and in- and atherosclerosis, as was recently sulin resistance. Diabetes Care 23:1816– stressed by Christen et al. (3). We also been shown to lower homocysteine con- 1822, 2000 centrations in postmenopausal women agree that homocysteine metabolism is 3. Dudman NPB: An alternative view of ho- altered for a few days after acute isch- (8,9), with the greatest benefit being seen mocysteine. Lancet 354:2072–2074, 1999 in women with high levels of homocys- 4. Egerton W, Silberberg JS, Crooks R, Ray emic events, as reported by Egerton teine (8). may lower homocys- C, Xie L, Dudman NPB: Plasma levels of et al. (4) and Lindgren et al. (5). There- teine through several mechanisms that homocysteine during acute phase re- fore, patients with acute cardiovascular include changes in the transamination sponse. Am J Cardiol 77:759–761, 1996 events were excluded from our analysis pathway of methionine catabolism (10) 5. Lindgren A, Brattstrom LE, Norrving B, and could not interfere with the inter- and an increased activity of renal methio- Hultberg B, Anderson A, Johansson BB: pretation of our results. nine synthase (11), the enzyme responsi- Plasma homocysteine in the acute and Chan et al. (1) pertinently mention convalescent phase after stroke. Stroke 26: the role of fibrates and metformin as po- ble for the remethylation of homocysteine 795–800, 1995 to methionine. Additionally, estrogens 6. Chan NN: Homocysteine and insulin lev- tential confounders of homocysteine con- may increase LDL-receptor expression, els in type 2 diabetic patients (Letter). Di- centrations. Our data in type 2 diabetic which facilitates LDL binding to homo- abetes Care 23:1041, 2000 patients confirm that fibrates significantly cysteine and its subsequent clearance 7. Chan N, Chan JCN: Implication of fibrate increased homocysteine levels, whereas

1124 DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 Letters there was no influence from metformin cause of various confounding factors, in- among postmenopausal women. JAMA treatment. cluding the use of HRT. 281:1817–1821, 1999 Chan et al. (1) also comment aptly on postmenopausal use of estro-progester- M. BUYSSCHAERT, MD one hormone replacement therapy M.P. HERMANS, MD (HRT). Recent reports show that homo- The Role of Hepatitis cysteine levels increase after menopause From the Department of Endocrinology and Nutri- C Virus Genotypes in (6,7) and that HRT lowers homocysteine tion, St. Luc Universite´ Catholique de Louvain, Brussels, Belgium. Development of concentrations (8,9). Address correspondence to Prof. M. Buysschaert, In our cohort, 84% of women (n ϭ Service d’Endocrinologie et , Cliniques Autoimmune 68) had reached menopause, and reliable Universitaires St Luc Avenue Hippocrate 54, UCL Diseases information on both menopause and HRT 54.74B-1200Brussells,Belgium.E-mail:buysschaert @diab.ucl.ac.be. status was available for 46 of them. A total of 20 (43%) of the latter patients were n the study by Betterle et al. (1), it was ϩ ●●●●●●●●●●●●●●●●●●●●●●● treated with HRT (HRT group), whereas stated that there was no association be- Ϫ 26 had no hormonal substitution (HRT References I tween the occurence of organ-specific/ group). Age, diabetes duration, BMI, 1. Chan NN, Tan T, Hurel SJ: Hyperhoma- nonorgan-specific autoimmune diseases and chronic hepatitis C virus (HCV) in- smoking prevalence, HbA1c, and creati- cysteinemia and macroangiopathy in type nine clearance were comparable between 2diabetes(Letter).DiabetesCare24:1123– fection. It was evident in previous studies the two groups. In contrast, menopause 1124, 2001 (2,3) that there was a clear association duration was significantly shorter in 2. Buysschaert M, Dramais A, Wallemacq among the antigenic subtypes of the HCV, ϩ Ϫ PE, Hermans MP: Hyperhomocysteine- HRT group than in the HRT group prognosis of disease, autoimmune disease Ϯ Ϯ ϭ mia in type 2 diabetes: relationship to development, and response to the inter- (16 4 vs. 22 6 years, P 0.002). macroangiopathy, nephropathy, and in- ␣ ␣ There was a nonsignificant trend for low- feron- (IFN- ) treatment. Also, these sulin resistance. Diabetes Care 23:1816– studies stress that some specific subtypes, er homocysteine levels in patients with 1822, 2000 HRT (13.2 Ϯ 5.3 vs. 17.4 Ϯ 9.9 ␮mol/l, 3. Christen W, Ajani U, Glynn R, Hennek- e.g., HCV-1b, may have worse prognosis and poor response to the treatment (4). P ϭ 0.07). Folic acid and vitamin B con- ens C: Blood levels of homocysteine and 12 This result may be due to the fact that the centrations were comparable (7.4 Ϯ 3.9 increased risk of cardiovascular disease. various genotypes of the HCV have vari- vs. 7.2 Ϯ 3.6 ng/ml and 398 Ϯ 167 vs. Arch Intern Med 160:422–434, 2000 ϩ 4. Egerton W, Silberberg JS, Crooks R, Ray ous immunological responses in the host. 521 Ϯ 298 pg/ml) in the HRT and the Ϫ C, Xie L, Dudman NPB: Plasma levels of So, it seems logical that the development HRT groups, respectively. Plasma cho- homocysteine during acute phase re- of the HCV subtype–specific immune re- lesterol (C), LDL-C, and triglycerides ϩ sponse. Am J Cardiol 77:759–761, 1996 sponse may stimulate the formation of an- were also similar in the HRT and the 5. Lindgren A, Brattstrom L, Norrving B, Ϫ tibodies directed to some organs or may HRT groups, whereas HDL-C and Hultberg B, Anderson A, Johansson BB: be generalized, which are named organ- Ϯ Ϯ Plasma homocysteine in the acute and C/HDL ratio were 58 17 vs. 49 13 and/or nonorgan-specific autoimmune mg/dl (P ϭ 0.058) and 4.0 Ϯ 0.9 vs. convalescent phase after stroke. Stroke 26: Ϯ ϭ 795–800, 1995 diseases. Therefore, various types of viral 5.0 2.5 (P 0.068) in patients with antigens may play an important role dur- and without HRT, respectively. It is of in- 6. Hak E, Polderman KH, Westendorp I, Ja- kobs C, Hofman A, Witteman J, Stehou- ing the development of the various types terest to mention that macroangiopathy ϩ wer C: Increased plasma homocysteine of autoimmune diseases via hosts’ im- was significantly less frequent in HRT after menopause. Atherosclerosis 149:163– mune responses. In the study (1), there ϭ than in HRT – patients (10 vs. 39%, P 168, 2000 were some autoantibody positivity results 0.04), as also observed in nondiabetic 7. Morris MS, Jacques PF, Selhub J, Rosen- before IFN-␣ treatment (one patient with subjects (10). berg I: Total homocysteine and estrogen low-titer classic islet cell antibody pat- In view of the trend in association be- status: indicators in the third National tern, one with high levels of anti-GAD an- Health and Nutrition Examination Sur- tween the use of HRT and lower homo- tibodies, two with anti– cell cysteine levels, as well as the reduced vey. Am J Epidemiol 152:140–148, 2000 ϩ 8. Van Baal WM, Smolders R, Van der antibody, four with microsomal antibod- prevalence of macroangiopathy in HRT Mooren M, Teerlink T, Kenemans P: Hor- ies, two with thyroglobulin autoantibod- women, we agree with Chan et al. (1) that mone replacement therapy and plasma ies, five with parietal cell antibodies, it could be of interest to measure homo- homocysteine levels. Obstet Gynecol 94: and—from nonorgan-specific antibody cysteine in postmenopausal type 2 dia- 485–491, 1999 group—two patients with anti-nuclear betic patients as another incentive for 9. Walsh B, Paul S, Wild R, Dean R, Tracy R, antibody [ANA] positivity); but, it is not selective use of HRT when high homocys- Cox D, Anderson P: The effects of hor- clear which HCV subtypes were infecting teine is found. mone replacement therapy and raloxifene antibody-positive patients. Like this situ- In summary, our findings demon- on C-reactive protein and homocysteine ation, the HCV subtype–specific host- in healthy postmenopausal women: a ran- strate the current importance of homo- domized, controlled trial. J Clin Endocrinol immune response may be responsible, as cysteine determination in routine initial Metab 85:214–218, 2000 with the autoimmune diseases that devel- evaluations of all patients at risk for mac- 10. Ridker PM, Manson JE, Buring JE, Shih oped during IFN-␣ treatment. Therefore, roangiopathy. Additional caution is J, Matias M, Hennekens Ch: Homocys- it would have been useful in this study to needed when interpreting such data be- teine and risk of cardiovascular disease assess whether a clinical linkage was

DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 1125 Letters present between the viral subtype and autoimmunity. HCV genotypes Patients Autoantibodies Thyroid antibodies PCA ICA ANA 1b 23 (32.8) 4 (17) 0 1 2 1 AYLA SANVER, MD 1a 6 (8.6) 2 (33.3) 0 1 1 0 ALPER GURLEK, MD 2c 8 (11.4) 3 (37.5) 0 2 0 1 HALIS SIMSEK, MD 3a 29 (41.4) 6 (20.7) 4 1 1 0 GONCA TATAR, MD 4 4 (5.7) 0 (0) 0 0 0 0 Data are n and n (%). ANA, anti-nuclear antibodies; ICA, islet-cell antibodies; PCA, parietal cell antibodies. From the Hacettepe University Faculty of Medicine, Department of Internal Medicine, Ankara, Turkey. Address correspondence to Ayla Sanver, MD, Response to Sanver immune phenomena has not been Hacettepe University Faculty of Medicine, Depart- significantly increased with respect to the ment of Internal Medicine, 06100 Sıhhıye, Ankara, et al. normal population. These data are in con- Turkey. E-mail: [email protected]. flict with the fact that HCV chronic infec- n this issue of Diabetes Care, the letter tion may act as a trigger factor in autoimmunity. So, our data are not suffi- ●●●●●●●●●●●●●●●●●●●●●●● by Sanver et al. (1) reported previous studies that demonstrated evidence of cient to confirm the hypothesis that a spe- References I cific genotype may be correlated with the 1. Betterle C, Fabris P, Zanchetta R, Pedini B, a correlation among the hepatitis C virus (HCV) antigenic subtypes, prognosis of autoimmune response. We think that it Tositti G, Bosi E, De Lalla F: Autoimmu- will be necessary to carry out further stud- nity against pancreatic islets and other tis- disease, autoimmune disease develop- sues before and after interferon-␣ therapy ment, response to the interferon (IFN)-␣ ies, before and after IFN therapy in hu- in patients with hepatitis C virus chronic treatment, and various immunological re- mans, to clarify the specific role of genetic infection. Diabetes Care 23:1177–1181, sponses in the host. In our 70 patients, we HCV subtypes and the genetic suscepti- 2000 carried out the type of HCV genotypes as bility in inducing autoimmunity. 2. Mason A, Lau JYN, Hoang N, Qian K, Al- published (2). The correlations between exander GSM, Xu L, Guo L, Jacob S, Re- HCV genotypes and autoimmunity are re- CORRADO BETTERLE, MD genstein FG, Zimmerman R, Everhart JE, ported in Table 1. RENATO ZANCHETTA, MD Wasserfall C, Maclaren NK, Perrillo RP: From our data, the autoimmunity Association of diabetes mellitus and PAOLO FABRIS, MD chronic hepatitis C infection. Hepatology varies from 37.5% in 2c to 0% in 4 HCV- 29:328–333, 1999 genotype groups. Interestingly, all of the From the Department of Medical and Surgical Sci- 3. Quaranta JF, Tran A, Regnier D, Letestu R, patients who tested positive for thyroid ences, University of Padua, Padua, Italy. Beusnel C, Fuzibet JG, Thiers V, Rampal autoantibodies are in the 3a HCV geno- Address correspondence to Prof. Corrado Bet- type group, but it is important to consider terle, Department of Medical and Surgical Sciences, P: High prevalance of antibodies to hepa- Via Ospedale 105, Padua, Italy. titis C virus (HCV) in patients with anti- that in our group, 29 (41.4%) patients thyroid autoantibodies. J Hepatol 18:136– show 3a genotype. All patients who tested 138, 1993 positive for autoantibodies, except one, ●●●●●●●●●●●●●●●●●●●●●●● 4. Ozyilkan E, Erbas T, Simsek H, Telatar F, showed the positivity before IFN-␣ ther- References Kayhan B, Telatar H: Increased prev- apy. In the majority of the cases, autoan- 1. Sanver A, Gurlek A, Simsek H, Tatar G: alance of hepatitis C virus antibodies in tibodies increased under IFN therapy and The role of hepatitis C virus genotypes in patients with diabetes mellitus. J Intern the clinical autoimmune disease rarely development of autoimmune diseases. Di- Med 235:283–284, 1994 appeared. It is accepted that the autoim- abetes Care 24: 1125–1126, 2001 5. Pol S, Nalpas B, Bourliere M, Couzigou P, mune diseases are correlated with a par- 2. Betterle C, Fabris P, Zanchetta R, Pedini B, Abergel A, Zarski JP, Berthelot P, Brechot Tositti G, Bosi E, De Lalla F: Autoimmu- C: Combination of ribavirin and interfer- ticular genetic pattern of susceptibility nity against pancreatic islets and other tis- on-␣ surpasses high doses of interferon-␣ and that the environmental factors may sues before and after interferon-␣ therapy alone in patients with genotype-Ib-related act as trigger factors, but direct evidence in patients with hepatitis C virus chronic chronic hepatitis C. Hepatology 31:1338– of this phenomenon remains circumstan- infection. Diabetes Care 23: 1177–1181, 1344, 2000 tial. In our patients, the presence of auto- 2000

1126 DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001