bekam, DM.pdf

Blood Letting in High-Ferritin Type 2DiabetesEffects on vascular reactivity

JOSE MANUEL FERNANDEZ-REAL, MD, PHD1

GEORGINA PENARROJA, MD2

ANTONI CASTRO, MD, PHD2

FERNANDO GARCIA-BRAGADO, MD, PHD2

ABEL LOPEZ-BERMEJO, MD, PHD1

WIFREDO RICART, MD1

OBJECTIVE In a recent study, iron chelation with deferoxamine led to improvement ofendothelial dysfunction in patients with coronary artery disease. We tested the hypothesis thatdecreasing circulating iron stores might improve vascular dysfunction in patients with type 2diabetes and increased serum ferritin concentration.

RESEARCH DESIGN AND METHODS A total of 28 type 2 diabetic male patientswith serum ferritin levels200 ng/ml (18% of consecutive type 2 diabetic men attending ouroutpatient clinic) were randomized to iron depletion (three extractions of 500 ml blood at2-week intervals; group 1A) or to observation (group 1B). C282Y mutation was absent in allpatients. Vascular reactivity (high-resolution external ultrasound) was evaluated at baseline andat 4 and 12 months thereafter. The two groups of patients were matched for age, BMI, pharma-cological treatment, and chronic diabetic complications.

RESULTS Endothelium-dependent vasodilation remained essentially unchanged in bothgroups of patients. In contrast, the vasodilation induced by glyceryl trinitrate (GTN) improvedsignificantly after iron depletion (P 0.006). These changes occurred in parallel to decreases intransferrin saturation index and HbA1c levels (0.6%, P 0.05) only in group 1A patients. Thebest predictor of the modifications in endothelium-independent vasodilation was the change inHbA1c levels. Changes in endothelium-independent vasodilation also correlated with the changein serum ferritin (r 0.45, P 0.04). At 12 months, transferrin saturation index andGTN-induced vasodilation returned to values similar to those at baseline in both groups ofsubjects.

CONCLUSIONS Iron depletion improves vascular dysfunction in type 2 diabetic patientswith high ferritin concentrations. The mechanisms by which these changes occur should befurther investigated.

Diabetes Care 25:22492255, 2002

F erritin gene expression increases inthe course of atherosclerotic plaqueformation (1). Iron is a transitionmetal that can easily become oxidized andthus act as an oxidant. A possible link be-tween iron and atherogenesis has been

suggested by the finding that iron chela-tion blocks oxidation of LDL, whereasiron released from heme and ferritin fa-vors oxidation of LDL (2). In fact, the gen-eral effect of catalytic iron is to convertpoorly reactive free radicals such as H2O2

into highly reactive ones, such as the hy-droxyl radical. In vitro addition of oxy-gen-derived free radical scavengers orantioxidants can reverse defective endo-thelium-dependent relaxation in experi-mental diabetes (3 6) or in diabeticpatients (7). Impaired vascular responsesto vasodilators, such as acetylcholine andglyceryl trinitrate (GTN), are usuallyfound in patients with type 2 diabetes(8 10).

In experimental models, iron has anadverse effect on endothelium (11) andaccelerates the development of athero-sclerosis (12,13). In patients with ironoverload, midsize arteries are character-ized by an eccentric hypertrophy and de-creased distensibility (14). These findingsseem to be linked to iron-induced fibro-genesis, determining an increased totalcollagen content in arteries from thesepatients (15). There is also some evidenceof iron-dependent growth of arterial walltissue (16). Iron chelation by deferox-amine inhibits vascular smooth musclecell proliferation (17).

Diabetes-induced endothelial dys-function is prevented by long-termtreatment with the modified iron chela-tor hydroxyethyl starchconjugated de-feroxamine (18). The coronary arteryresponses to cold pressor test and papav-erine are improved by acute administra-tion of deferoxamine in type 2 diabeticpatients (19). Deferoxamine also im-proved nitric oxide (NO)-mediated vaso-dilation in patients with coronary arterydisease (20). Iron depletion, commonlyused to treat patients with hemochroma-tosis, has been demonstrated to be safe indiabetic subjects (21). Blood letting hasbeen shown to increase the oxidation re-sistance of serum VLDLs/LDLs in menwho regularly smoke (22). Oxidized lowLDL impairs cyclic GMPmediated dila-tions (23), and, thus, endothelial functionis expected to improve after iron deple-tion. We found no previous controlledclinical trials of the effect of iron depletionon vascular function. For that reason, wecarried out a clinical trial to test the effects

From the 1Unit of Diabetes, Endocrinology and Nutrition, University Hospital of Girona Dr Josep Trueta,Girona, Spain; and the 2Department of Internal Medicine, University Hospital of Girona Dr Josep Trueta,Girona, Spain.

Address correspondence and reprint requests to J.M. Fernandez-Real, MD, PhD, Unit of Diabetes, Endo-crinology and Nutrition, Hospital de Girona Dr Josep Trueta, Ctra. Franca s/n, 17007 Girona, Spain.E-mail: [email protected].

Received for publication 24 January 2002 and accepted in revised form 20 August 2002.Abbreviations: CV, coefficient of variation; GTN, glyceryl trinitrate; MDA, malondialdehyde; NO, nitric

oxide.A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion

factors for many substances.

E m e r g i n g T r e a t m e n t a n d T e c h n o l o g yO R I G I N A L A R T I C L E

DIABETES CARE, VOLUME 25, NUMBER 12, DECEMBER 2002 2249

of reducing stored and circulating iron onendothelial function in type 2 diabeticsubjects with increased serum ferritin.

RESEARCH DESIGN ANDMETHODS

Inclusion and exclusion criteriaA total of 28 type 2 diabetic men wereprospectively recruited from diabetesoutpatient clinics on the basis of the fol-lowing criteria: 1) serum ferritin level200 ng/ml on two separate determina-tions, with at least a 1-month interval, and2) stable metabolic control in the previous6 months. Approximately 18% of consec-utive type 2 diabetic men attending ouroutpatient clinic fulfilled these criteria.Exclusion criteria included the following:1) clinically significant hepatic, neurolog-ical, endocrinologic, or other major sys-temic disease, including malignancy; 2)history or current clinical evidence ofhemochromatosis, or presence of theCys282Tyr mutation; 3) history of drugor alcohol abuse, defined as consumingover 80 g/day, or serum transaminase ac-tivity over twice the upper limit of nor-mal; 4) an elevated serum creatinineconcentration; 5) an acute major cardio-vascular event in the previous 6 months;6) acute illnesses and current evidence ofacute or chronic inflammatory or infec-tive diseases; 7) transfusion history, oriron or vitamin therapies in the previousyear; 8) history of disturbances in ironbalance (e.g., hemosiderosis from anycause, atransferrinemia, paroxysmal noc-turnal hemoglobinuria, or iron defi-ciency); and 9) mental illness renderingthe subject unable to understand the na-ture, scope, and possible consequences ofthe study. Informed written consent wasobtained after the purpose, nature, andpotential risks were explained to the sub-jects. The experimental protocol was ap-proved by the Hospital Ethics Committee.

Study protocolAll patients underwent a full medical his-tory including age, duration of diabetes,BMI, eating habits, smoking habits, bloodpressure, total cholesterol, and a full ex-amination to screen for diabetic compli-cations. The clinical diagnosis of diabeticretinopathy was based on the examina-tion of the ocular fundus after dilatationof the pupils by experienced ophthalmol-ogists. Simplex retinopathy was definedas one or more microaneurysms or hem-

orrhages. Diabetic macroangiopathycomplications were diagnosed accordingto clinical findings, Doppler sonography,and angiopathy. Persistent microalbu-minuria was defined as albumin excretionrate of 30300 mg/day. Patients consid-ered eligible to participate in the studymet with the doctor 4 weeks before, every2 months during the first 4 months, andevery 4 months thereafter. The patientswere instructed to record any episode ofsymptomatic hypoglycemia daily.

Study designDiabetic patients with elevated serum fer-ritin concentrations were randomized toeither an iron depletion group (group 1A;n 13) or an observation group (group1B; n 15) according to a randomizationtable that included age, BMI, and HbA1c.The iron depletion intervention consistedof three extractions at 2-week intervals atweeks 0, 2, and 4. Each time, 450 g (500ml) of blood was drawn. After blood do-nation, blood volume was restored to nor-mal within 2448 h by hemodilution.This hemodilutional effect was usuallymanifest 1 week after phlebotomy, withsignificant reductions in the hematocritlevels, which returned to baseline level af-ter 4 weeks (24). Thus, the patients werestudied at baseline, at 4 months, and at 12months after iron depletion. All subjectswere instructed to keep their usual treat-ment with insulin or hypoglycemicagents, diet, and exercise during the studyperiod.

MeasurementsEach subject was studied in the researchlaboratory in the postabsorptive state.The room was quiet, lights were dimmed,and the temperature was controlled at23C. BMI was calculated as weight (inkilograms) divided by height (in meters)squared. The subjects waist was mea-sured with a soft tape midway betweenthe lowest rib and the iliac crest. The hipcircumference was measured at the wid-est part of the gluteal region. The waist-to-hip ratio was then calculated. Bloodpressure was measured in the supine po-sition on the right arm after a 10-min rest;a standard sphygmomanometer of appro-priate cuff size was used, and the first andfifth phases were recorded. Values used inthe analysis were the average of threereadings taken at 5-min intervals. Alco-hol, caffeine, and all medications, includ-ing sulfonylurea, metformin, and insulin,

were withheld within 12 h of the differenttests.

Brachial artery vascular reactivityA high-resolution external ultrasound(128XP/10 mainframe with a 7.5-MHzlinear array transducer) (Toshiba SSH-140A) was used to measure changes inbrachial artery diameter in response to re-active hyperemia (leading to flow-mediatedendothelium-dependent dilation) and inresponse to 400 g sublingual GTN, anendothelium-independent direct smoothmuscle dilator, as described by Celerma-jer et al. (25). The lumen diameter of theartery was defined as the distance be-tween the leading edge of the echo of thenear walllumen interface to the leadingedge of the far walllumen interface echo(26). All scans were taken electrocardio-gram-triggered coincident with the Rwaveend diastolic. All images were re-corded with an super-VHS videotape(Panasonic MD-830AG). Endothelial-dependent vasodilation was elicited withhyperemia induced by inflation of a pneu-matic tourniquet placed around the fore-arm, distal to the scanned part of theartery, up to a pressure of 300 mmHg for5 min, followed by sudden deflation. Thismaneuver is recognized to increase shearstress on the endothelial cells, which inturn releases NO-producing vasodilation,allowing endothelial function to be tested(27). Endothelial-dependent vasodilationis expressed as the percentage of changein the arterial diameter 1 min after hy-peremia. Endothelial-independent vaso-dilation is induced after sublingualadministration of a 400-g metered doseof GTN, an exogenous NO donor (Solini-trina spray; Almirall Prodesfarma, Barce-lona, Spain) and expressed as thepercentage of change in the arterial diam-eter 3 min later. Reactive hyperemia is cal-culated as the percentage change betweenthe maximum flow recorded in the first15 s after cuff deflation and the flow dur-ing the resting scan.

A first scan was recorded after 10 minof resting in a quiet room in the supineposition. Then the tourniquet was in-flated for 5 min. A second scan was re-corded during 90 s beginning 10 s beforecuff deflation. After at least 10 more min-utes of rest, a new control scan was re-corded. A last scan was recorded from 2min after GTN administration during70 s. All images registered on a super-VHS tape were analyzed afterward by two

Blood letting and vascular reactivity

2250 DIABETES CARE, VOLUME 25, NUMBER 12, DECEMBER 2002

independent observers blinded to therandomization of the subject and thestage of the experiment. Each observeranalyzed the arterial diameter of four car-diac cycles for each condition, and thesemeasurements were averaged. Before theinitiation of the study in diabetic subjects,validation of this technique was per-formed through the evaluation of inter-and intraobserver repeatability in 22healthy subjects (12 men and 10 women,mean age 30.1 years [95% CI 27.133.2],BMI 22.6 kg/m2 [21.323.8]). Measure-ments were performed by two observers(A and B). Intraclass coefficient of corre-lation of fixed effects between observers Aand B was 0.90. Coefficient of variation(CV) between means obtained by observ-ers A and B was 9%. The CV obtained bya same observer was 3%. The repeatability(95% CI) was 0.27 mm (observer A). TheCV was 4% for observer B, with a repeat-ability (95% CI) of 0.39 mm. Within-subject variability for 5 consecutive days(five subjects) showed a CV of 6% (ob-server A) and 2% (observer B). The GTN-induced vasodilation correlated withbasal artery diameter (r 0.67; P 0.025) and flux-mediated vasodilation(r 0.68; P 0.021).

Analytical determinationsThe serum glucose concentrations weremeasured in duplicate by the glucose ox-idase method with the use of a BeckmanGlucose Analyzer II (Beckman Instru-ments, Brea, CA). HbA1c was measuredby high-pressure liquid chromatographywith the use of a fully automated glycosy-lated hemoglobin analyzer system (Hita-chi L-9100) . Serum ferr i t in wasdetermined by microparticle enzyme im-munoassay (AXSYMTM; Abbot Laborato-ries, Abbott Park, IL), with a intra- andinterassay CV of6%. Serum transferrin,transferrin saturation index, C-reactiveprotein (Beckman, Fullerton, CA), iron(Hitachi 917), and whole-blood hemo-

globin level and hematocrit levels (EDTAsample; Coulter Electronics, Hialeah, FL)were determined by routinary laboratorytests. Serum C-peptide concentrationswere measured using a fluorimetric im-munoassay (EG & G Wallac, Wallac Oy,Turku, Finland) with intra- and inter-assay CVs of 6%. Total serum choles-terol was measured through the reactionof cholesterol esterase/cholesterol oxi-dase/peroxidase, using a Hitachi 747.HDL cholesterol was quantified after pre-cipitation with polyethylene glycol atroom temperature. LDL cholesterol wascalculated using the Friedewald formula,when applicable. Total serum triglycer-ides were measured through the reactionof glycerol/phosphate/oxidase and perox-idase.

Plasma NO2/NO3

levels were mea-sured with the Nitric Oxide ColorimetricAssay Kit (Calbiochem-Novabiochem,San Diego, CA) based on a Griess reac-tion. Amounts of nitrite in the plasmawere estimated by a standard curve ob-tained from enzymatic conversion ofKNO3 to nitrite.

Lipid peroxidation was evaluatedthrough a modification of the thiobarbi-turic acid method, in which the concen-tration of malondialdehyde (MDA), astable end by-product of the metabolismof lipid peroxides, is specifically mea-sured by high-pressure liquid chromatog-raphy, according to Esterbauer andCheeseman (28).

Statistical methodsDescriptive results of continuous vari-ables are expressed as means SD. Beforestatistical analysis, normal distributionand homogeneity of the variances wereevaluated using Levenes test, and thenvariables were given a log-transformationif necessary. We used the 2 test for com-parisons of proportions, and we used one-way ANOVA with Bonferroni correction,

unpaired, or paired t tests for compari-sons of quantitative variables.

RESULTS The two groups of pa-tients were comparable in age (group 1A,54.4 8.2 vs. group 1B, 55.7 8 years),BMI (28.7 2.3 vs. 30.5 3.2 kg/m2),waist-to-hip ratio, systolic blood pressure(135.1 23.9 vs. 137.3 17.5 mmHg),diastolic blood pressure (82 15.9 vs.82.1 7.5 mmHg), proportion of smok-ers (n 3 vs. n 2), serum ferritin(460 109 vs. 566 369 ng/ml), phar-macological treatment (insulin, bigua-nides, sulfonylureas, acarbose, statins,fibrates, ACE inhibitors, -blockers, aspi-rin, and allopurinol), and chronic dia-betic complications (macroangiopathy,diabetic retinopathy, and microalbumin-uria). Of note is the near-identical initialblood HbA1c levels (6.27 0.95 vs.6.39 1.2%) because the patients werestratified according to this parameter.Baseline endothelium-dependent and-independent vasodilations were not sig-nificantly different in the two groups ofpatients (Table 1). After the study inter-vention, body weight, blood pressure,and blood hematocrit levels comprisedthe majority of the biochemical parame-ters that were measured, and drug treat-ment did not significantly change ineither group (4-month follow-up: BMI:group 1A: 29.1 2.7 vs. group 1B:30.2 3.7; systolic blood pressure:133.3 22.6 vs. 140 15.3; diastolicblood pressure: 71.2 12 vs. 85 10.6;hematocrit: 42.6 3.7 vs. 44.3 2.7;fasting glucose: 161 31 vs. 165 45;fasting C-peptide: 2.76 1.4 vs. 2.65 1.4; cholesterol: 182 30 vs. 195 40;LDL cholesterol: 109 24 vs. 117 32;and triglycerides: 194 70 vs. 156 87). As expected, serum ferritin, trans-ferrin saturation index, and blood hemo-globin decreased significantly at 4 monthsonly in patients after iron depletion (4-

Table 1Baseline and follow-up of brachial artery vascular reactivity

Group 1A (n 9) Group 1B (n 8)

Baseline 4 Months 12 Months P Baseline 4 Months 12 Months P

Baseline vessel size (mm) 4.28 0.41 4.14 0.53* 4.16 0.55 0.03 4.28 0.55 4.44 0.6 4.29 0.50 NSBaseline flow (ml/min) 0.50 0.07 0.48 0.123 0.52 0.15 NS 0.59 0.16 0.62 0.23 0.55 0.21 NSEndothelium-dependent vasodilation (%) 2.7 3.7 4.3 4 2.9 2.5 NS 4.3 3.9 3.8 4.7 4.4 4.8 NSEndothelium-independent vasodilation (%) 10.6 3.1 18.4 7* 12 7 0.006 13.2 8.7 12.4 6.3 12.8 5.4 NS

Data are means SD. *Significantly different from baseline.

Fernandez-Real and Associates


month follow-up: ferritin: group 1A:232 110 vs. group 1B: 507 438;transferring saturation index: 19.89 6.1 vs. 31 12.5; and hemoglobin:14.12 1.2 vs. 15.1 0.8). In parallel tothese changes, blood HbA1c decreasedsignificantly only in these patients (meandifferences 0.61, 95% CI 0.17 to1.05; P 0.01). They maintained thesimilar agents for the treatment of diabe-tes and showed a nonsignificant tendencytoward an increased number of hypogly-cemic events. None of them required hos-pitalization. C-reactive protein excludedsignificant acute inflammation in all sub-jects during the study period and re-mained essentially unchanged (4-monthfollow-up: group 1A: 0.57 0.5; group1B: 0.7 1.3). At baseline, serum MDA (asurrogate marker of free radical produc-tion) was not associated with any param-eter of iron metabolism (r 0.27, NS).Serum MDA and NO2

/NO3 did not

significantly change during the 4-monthperiod in either group (4-month follow-up: MDA: group 1A: 15.6 7.4 vs. group1B: 19.2 8.5; NO2

/NO3: group 1A:

299 157 vs. 268 76).Endothelium-dependent vasodila-

tion did not change significantly in eithergroup of subjects (Table 1). In contrast,GTN-induced vasodilation improved sig-nificantly after iron depletion in group 1Apatients in parallel to changes in HbA1cand the transferrin saturation index (Fig.1). Significant improvements in GTN-induced vasodilation were observed innearly 80% of patients, whereas it re-mained essentially unchanged in the restof patients. No significant differences inclinical or biochemical characteristicswere found between the subjects who re-sponded to phlebotomy versus the sub-jects who did not respond. No significantchanges were observed in group 1B pa-tients. The best predictor of the observedchanges in endothelium-independent va-sodilation was the change in HbA1c levelswhen all subjects were considered as awhole (r0.48, P 0.02). Changes inendothelium-independent vasodilationalso correlated with the change in serumferritin (r0.45, P 0.04), and a ten-dency was observed with the change inblood total hemoglobin in group 1A sub-jects (r 0.48, P 0.09, n 9). Al-though no significant changes inendothelium-dependent vasodilation wasobserved in either group, post hoc analy-sis indicated that it was associated with

changes in endothelium-independent va-sodilation (r 0.55, P 0.01) and alsowith the decrease in HbA1c (r 0.43,P 0.02) when all the subjects wereconsidered as a whole. At 12 months offollow-up, all clinical and biochemicalparameters remained unchanged in thecontrol group. In the phlebotomy-treatedgroup, hemoglobin, transferrin saturation,

and HbA1c returned to baseline levels,whereas ferritin concentrations remainedsignificantly lower compared with base-line. Similarly, all vascular reactivity pa-rameters remained unchanged at 12months of follow-up in the control group,whereas vessel size and endothelium-independent dilation returned to baselinelevels in the treated group.

Figure 1Changes in HbA1c, transferrin saturation index, and endothelium-independent vaso-dilation after blood letting (group 1A subjects) or sham blood letting (group 1B). *Significantlydifferent from baseline; significantly different from the same parameter at 4 months.



CONCLUSIONS Transition met-als and active oxygen species act as vascu-lar smooth muscle cell growth factors(16). Iron overload causes an early alter-ation of arterial wall structure and func-tion in humans, characterized byeccentric hypertrophy (14). A thicker vas-cular wall implies that the poorly vascu-larized intimal layer is less easily reachedby oxygen and other nutrients. This hy-pertrophy is reversible after iron deple-tion (14), whereas iron chelation bydeferoxamine inhibits vascular smoothmuscle cell proliferation (17). In thepresent study, we observed increasedGTN-induced vasodilation of forearmconduit vessels in type 2 diabetic patientsafter iron depletion. Vasodilation in re-sponse to GTN is mediated by vascularsmooth muscle cells. Thus, iron depletionresults in increased arterial distensibility(14) and raised vasodilatory response assurrogates of improved smooth musclecell function. These findings occurred inparallel to decreased HbA1c concentra-tions (Fig. 1). The parallel modificationsobserved in these variables suggest thatthey are inter-related events. Transitionmetalcatalyzed reactions play a majorrole in the development of vascular dys-function in experimental diabetes (29).Although our study cannot elucidate theprecise mechanisms involved, iron isknown to generate reactive oxygen spe-cies, especially hydroxyl radical, viaFenton chemistry. The importance of hy-droxyl radicals for impaired endothelialfunction in diabetes has been stressedelsewhere (18,29).

Superoxide reacts with NO to pro-duce peroxynitrite, which impairs vasodi-lation and can nitrosylate proteins to altertheir function (30). Iron chelation withdeferoxamine improves vascular dysfunc-tion in patients with coronary artery dis-ease (20) and experimental models (18).Our observation of improved vascular re-sponses to GTN after decreasing ironstores evokes a scenario in which bio-chemical rather than structural defects arepresent and may be reversible. Changes intransferrin saturation index, blood hemo-globin, and HbA1c, but not in serum fer-ritin (an indicator of tissue iron stores),mirrored the wax and wane of vascularresponses. These findings suggest thatblood itself is an important source of tran-sition metals that impair vascular func-tion. Exposure of normal blood vessels toHbA1c has been shown to inhibit vascular

relaxation (31). Endothelial cells are ableto incorporate cell-free hemoglobin, cre-ating a new way for circulating hemoglo-bin to be in close contact with NO (32).

Arterial smooth muscle responses toincreasing doses of GTN are signifi-cantly lower in diabetic subjects (3335).In the larger study sample to date, diabe-tes, larger vessel size, and reduced endo-thelium-dependent dilation were allindependently associated with impairedGTN-related vasodilation on multivariateanalysis (35). These findings are in linewith the parallel changes observed in en-dothelium-dependent and -independentvasodilation (r 0.55, P 0.01) and thereduction in arterial diameter after the im-proved responses to GTN found in thisstudy.

In addition to decreased responsive-ness of vascular smooth muscle to NO,the mechanisms responsible for endo-thelial dysfunction and reduced endo-thelium-dependent vasodilation inpatients with diabetes are not completelyunderstood. Decreased synthesis or re-lease of NO by endothelial cells is onepossibility. We observed no significantchanges in serum NO2

/NO3 concen-

tration and no significant variations inendothelium-dependent vasodilationduring the study. Another possible mech-anism is increased inactivation of endo-thelium-derived NO by oxygen-derivedfree radicals, which damages endothelialmembrane receptors for vasodilator ago-nists (3,4). We observed that vasculardysfunction did not run in parallel withthe serum MDA concentration. However,we studied only diabetic patients with sta-ble metabolic control, and thus the lack ofsignificant modifications in endothelium-dependent responses could be related togood baseline risk factor control. The de-crease in HbA1c after iron depletion wassimilar to that obtained after deferox-amine therapy in type 2 diabetic patients(0.5 and0.6%, respectively [36,37]),in parallel to decreased transferrin sat-uration index and serum ferritin con-centration. Blood hematocrit did notsignificantly change, excluding hemodi-lution as a confounding factor (24).Bleeding produced a significant decreasein serum glucose in diabetic patients (21)and healthy subjects (38). Long-termtreatment of diabetic rats with hydroxy-ethyl starch conjugated deferoxaminedid not modify serum insulin or bloodglucose but caused a reduction in HbA1c

(39). This chelator also reduced glyca-tion of albumin in vitro through elevatedglucose concentrations. Transition metalsplay an important role in protein glyca-tion induced by hyperglycemia. In fact,both HbA1c and serum glucose arestrongly associated with serum ferritinlevels, even in healthy subjects (40,41).Increased iron stores predicted the devel-opment of diabetes in epidemiologicalstudies (42,43). Interestingly, a lowerprevalence of diabetes was observedamong blood donors in a recent study(44). The impact of iron depletion on vas-cular dysfunction and metabolic controlin diabetic patients needs to be confirmedin a large-scale study because of the im-portant public health implications.

Current research is examining strate-gies that might improve endothelial func-tion. The early implementation of suchstrategies in the disease process wouldprevent or delay atherogenesis. Up untilnow, the therapies that have demon-strated to improve endothelial functionhave been lipid lowering (45,46), oral ad-ministration of L-arginine (47), ACE inhi-bition (48), estrogen therapy (49),administration of antioxidants (vitaminC) (7,50) and insulin (10), and allopuri-nol treatments (51), among others, withthe latter two being the only treatmentstested in diabetic patients. Our findingscontribute to a greater understanding ofthe factors that modulate the vasorelaxantresponse to NO. Whether blood lettingshould be considered adjuvant therapy inhigh-ferritin type 2 diabetes requires fur-ther study.

Acknowledgments This work was par-tially supported by grant 98/0808 from theFondo de Investigaciones Sanitarias, NationalHealth Institute of Spain.

The authors thank Dr. Dolores Cabrero andNuria Aleixandre for technical assistance.

References1. Pang J, Jiang MJ, Chen YL, Wang FW,

Wang DL, Chu SH: Increased ferritin geneexpression in atherosclerotic lesions.J Clin Invest 97:22042212, 1996

2. Abdalla DSP, Campa A, Monteiro MP:Low density lipoprotein oxidation bystimulated neutrophils and ferritin. Ath-erosclerosis 97:149159, 1992

3. Hattori Y, Kawasaki H, Abe K, Kanno M:Superoxide dismutase recovers alteredendothelium-dependent relaxation in di-abetic rat aorta. Am J Physiol 261:H1086



H1094, 19914. Langenstroer P, Pieper GM: Regulation of

spontaneous EDRF release in diabetic rataorta by oxygen free radicals. Am J Physiol263:H257H265, 1992

5. Diederich D, Skopec J, Diederich A, DaiFX: Endothelial dysfunction in mesen-teric resistance arteries of diabetic rats:role of free radicals. Am J Physiol 266:H1153H1161, 1994

6. Rosen P, Ballhausen T, Bloch W, AddicksK: Endothelial relaxation is disturbed byoxidative stress in the diabetic rat heart:influence of tocopherol as antioxidant.Diabetologia 38:11571168, 1995

7. Ting HH, Timimi FK, Boles KS, CreagerSJ, Ganz P, Creager MA: Vitamin C im-proves endothelium-dependent vasodila-tion in patients with non-insulin-dependent diabetes mellitus. J Clin Invest97:2228, 1996

8. Goodfellow J, Ramsey MW, LuddingtonLA, Jones CJ, Coates PA, Dunstan F, LewisMJ, Owens DR, Henderson AH: Endothe-lium and inelastic arteries: an earlymarker of vascular dysfunction in non-insulin-dependent diabetes. BMJ 312:744745, 1996

9. Williams SB, Cusco JA, Roddy M-A, John-stone MT, Creager MA: Impaired nitricoxide-mediated vasodilation in patientswith non-insulin-dependent diabetesmellitus. J Am Coll Cardiol 27:567574,1996

10. Vehkavaara S, Makimattila S, SchlenzkaA, Vakkilainen J, Westerbacka J, Yki-Jarvinen H: Insulin therapy improves en-dothelial function in type 2 diabetesmellitus. Arterioscler Thromb Vasc Biol 20:545550, 2000

11. Lekakis J, Papamicheal C, Stamatelopou-los K, Cimponeriu A, Voutsas A, VemmosK, Mavrikakis M, Stamatelopoulos S:Hemochromatosis associated with endo-thelial dysfunction: evidence for the roleof iron stores in early atherogenesis. VascMed 4:147148, 1999

12. Araujo JA, Romano EL, Brito BE, Parthe V,Romano M, Bracho M, Montano RF,Cardier J: Iron overload augments the de-velopment of atherosclerotic lesions inrabbits. Arterioscler Thromb Vasc Biol 15:11721180, 1995

13. Jacob HS: Newly recognized causes ofatherosclerosis: the role of microorgan-isms and vascular iron overload. J Lab ClinMed 123:808816, 1994

14. Failla M, Giannattasio C, Piperno A, Ver-gani A, Grappiolo A, Gentile G, Meles E,Mancia G: Radial artery wall alterations ingenetic hemochromatosis before and afteriron depletion therapy. Hepatology32:569573, 2000

15. Cardoso LEM, Mourao PAS: Composi-tional and structural alterations of glyco-saminglycans associated with the

complications brought about by thalasse-mia major: a case report. Angiology47:175183, 1996

16. Rao GN, Berk BC: Active oxygen speciesstimulate vascular smooth muscle cellgrowth and protooncogene expression.Circ Res 70:593599, 1992

17. Porreca E, Ucchino S, Di Febbo C, Di Bar-tolomeo N, Angelucci D, Napolitano AM,Mezzetti A, Cuccurullo F: Antiprolifera-tive effect of desferrioxamine on vascularsmooth muscle cells in vitro and in vivo.Arterioscler Thromb Vasc Biol 14:299304, 1994

18. Pieper GM, Siebeneich W: Diabetes-in-duced endothelial dysfunction is pre-vented by long-term treatment with themodified iron chelator, hydroxyethylstarch conjugated-deferoxamine. J Car-diovasc Pharmacol 30:734738, 1997

19. Nitenberg A, Paycha F, Ledoux S, Sachs R,Attali J-R, Valensi P: Coronary artery re-sponses to physiological stimuli are im-proved by deferoxamine but not byL-arginine in non-insulin-dependent dia-betic patients with angiographically nor-mal coronary arteries and no other riskfactors. Circulation 97:736743, 1998

20. Duffy SJ, Biegelsen ES, Holbrook M, Rus-sell JD, Gokce N, Keaney JF Jr, Vita JA:Iron chelation improves endothelial func-tion in patients with coronary artery dis-ease. Circulation 103:27992804, 2001

21. Bofill C, Joven J, Bages C, Vilella E, Sans T,Cavalle P, Miralles R, Llobet J, Camps J:Response to repeated phlebotomies in pa-tients with non-insulin-dependent diabe-tes mellitus. Metabolism 43:614620,1994

22. Salonen JT, Korpela H, Nyyssonen K,Porkkala E, Tuomainen TP, Belcher JD,Jacobs DR Jr, Salonen R: Lowering ofbody iron stores by blood letting and ox-idation resistance of serum lipoproteins: arandomized cross-over trial in malesmokers. J Intern Med 237:161168, 1995

23. Galle J, Bauersachs J, Busse R, Bassenge E:Inhibition of cyclic AMP- and cyclic GMP-mediated dilations in isolated arteries byoxidized low density lipoproteins. Arte-rioscler Thromb 12:180186, 1992

24. Ebert RV, Stead EA, Gibson JG: Responseof normal subjects to acute blood loss.Arch Intern Med 68:578590, 1941

25. Celermajer DS, Sorensen KE, Gooch VM,Spiegelhalter DJ, Miller OI, Sullivan ID,Lloyd JK, Deanfield JE: Non-invasive de-tection of endothelial dysfunction in chil-dren and adults at risk of atherosclerosis.Lancet 340:11111115, 1992

26. Wendelhag I, Gustavsson T, Suurkula M,Berglund G, Wikstrand J: Ultrasoundmeasurements of wall thickness in the ca-rotid artery: fundamental principles anddescription of a computerized analysingsystem. Clin Physiol 11:565577, 1991

27. Buga GM, Gold ME, Fukuto JM, IgnarroLJ: Shear stress-induced release of nitricoxide from endothelial cells grown onbeads. Hypertension 17:187193, 1991

28. Esterbauer H, Cheeseman KH: Determi-nation of aldehydic lipid peroxidationproducts: malonaldehyde and 4-hy-droxynonenal. Methods Enzymol 186:407420, 1990

29. Cameron NE, Cotter MA: Effects of an ex-tracellular metal chelator on neurovascu-lar function in diabetic rats. Diabetologia44:621628, 2001

30. Beckman JS, Beckman TW, Chen J, Mar-shall PA, Freeman BA: Apparent hydroxylradical production by peroxynitrite: im-plications for endothelial injury from ni-tric oxide and superoxide. Proc Natl AcadSci U S A 87:16201624, 1990

31. Angulo J, Sanchez-Ferrer CF, Peiro C,Marn J, Rodriguez-Manas L: Impairmentof endothelium-dependent relaxation byincreasing percentages of glycosylatedhuman hemoglobin: possible mecha-nisms involved. Hypertension 28:583592, 1996

32. Faivre-Fiorina B, Caron A, Fassot C, FriesI, Menu P, Labrude P, Vigneron C: Pres-ence of hemoglobin inside aortic cells af-ter cell-free hemoglobin administration inguinea-pigs. Am J Physiol 276:H766H770, 1999

33. McVeigh GE, Brennan GM, Johnston GD,McDermott BJ, McGrath LT, Henry WR,Andrews JW, Hayes JR: Impaired endo-thelium-dependent and independent va-sodilation in patients with type 2 (non-insulin-dependent) diabetes mellitus.Diabetologia 35:771776, 1992

34. Watts GF, OBrien SF, Silvester W, MillarJA: Impaired endothelium-dependentand independent dilatation of forearm re-sistance arteries in men with diet-treatednon-insulin-dependent diabetes: role ofdyslipidemia. Clin Sci 91:567573, 1996

35. Adams MR, Robinson J, McCredie R,Seale JP, Sorensen KE, Deanfield JE, Cel-ermajer DS: Smooth muscle dysfunctionoccurs independently of impaired endo-thelium-dependent dilation in adults atrisk of atherosclerosis. J Am Coll Cardiol32:123127, 1998

36. Cutler P: Deferoxamine therapy in high-ferritin diabetes. Diabetes 38:12071210,1989

37. Redmon JB, Pyzdrowski KL, RobertsonRP: No effect of deferoxamine therapy onglucose homeostasis and insulin secretionin individuals with NIDDM and elevatedserum ferritin. Diabetes 42:544549,1993

38. Facchini FS: Effect of phlebotomy onplasma glucose and insulin concentra-tions (Letter). Diabetes Care 21:2190,1998

39. Wolff SP, Jiany ZY, Hunt JV: Protein gly-



cation and oxidative stress in diabetesmellitus and aging. Free Radic Biol Med10:339352, 1991

40. Fernandez-Real JM, Ricart W, Arroyo E,Balanca R, Casamitjana R, Cabrero D,Fernandez-Castaner M, Soler J: Serumferritin as a component of the insulinresistance syndrome. Diabetes Care 21:6268, 1998

41. Tuomainen T-P, Nyysonen K, Salonen R,Tervahauta A, Korpela H, Lakka T,Kaplan GA, Salonen JT: Body iron storesare associated with serum insulin andblood glucose concentrations. DiabetesCare 20:426428, 1997, 1996

42. Salonen JT, Tuomainen T-P, NyyssonenK, Lakka H-M, Punnonen K: Relation be-tween iron stores and non-insulin-depen-dent diabetes in men: case-control study.BMJ 317:727730, 1998

43. Ford ES, Cogswell ME: Diabetes and se-rum ferritin concentration among U.S.adults. Diabetes Care 22:19781983, 1999

44. Ascherio A, Rimm EB, Giovannucci E,

Willett WC, Stampfer MJ: Blood dona-tions and risk of coronary heart disease inmen. Circulation 103:5257, 2001

45. Treasure CB, Klein JL, Weintraub WS,Talley JD, Stillabower ME, Kosinski AS,Zhang J, Boccuzzi SJ, Cedarholm JC, Al-exander RW: Beneficial effects of choles-terol-lowering therapy on the coronaryendothelium in patients with coronary ar-tery disease. N Engl J Med 332:481487,1995

46. Anderson TJ, Meredith IT, Yeung AC, FreiB, Selwyn AP, Ganz P: The effect of cho-lesterol lowering and antioxidant therapyon endothelium-dependent coronary va-sodilation. N Engl J Med 332:488493,1995

47. Clarkson P, Adams MR, Powe AJ, DonaldAE, McCredie R, Robinson J, McCarthySN, Keech A, Celermajer DS, DeanfieldJE: Oral L-arginine improves endotheli-um-dependent dilation in hypercholes-terolemic young adults. J Clin Invest 97:19891994, 1996

48. Mancini GB, Henry GC, Macaya C,ONeill BJ, Pucillo AL, Carere RG, War-govich TJ, Mudra H, Luscher TF, KlibanerMI, Haber HE, Uprichard AC, Pepine CJ,Pitt B: Angiotensin-converting enzyme:inhibition with quinapril improves endo-thelial vasomotor dysfunction in patientswith coronary artery disease. Circulation94:258265, 1996

49. Gilligan DM, Quyyumi AA, Cannon RO:Effects of physiological levels of estrogenon coronary vasomotor function in post-menopausal women. Circulation 89:25452551, 1994

50. Levine GN, Frei B, Kouloukis SN, Ger-hard MO, Keaney JFJ, Vita JA: Ascorbicacid reverses endothelial vasomotor dys-function in patients with coronary arterydisease. Circulation 93:11071113, 1996

51. Butler R, Morris AD, Belch JJF, Hill A,Struthers AD: Allopurinol normalizes en-dothelial dysfunction in type 2 diabeticswith mild hypertension. Hypertension 35:746751, 2000



Documents

bekam, DM.pdf