factors critical to iontophoretic assessment of vascular …moor-inst.jp/products/factors critical...

Factors Critical to Iontophoretic Assessment of VascularReactivity: Implications for Clinical Studies of

Endothelial Dysfunction

†Jane Elizabeth Ramsay, *William Russell Ferrell, †Ian Andrew Greer, and‡Naveed Sattar

From the Departments of *Medicine, †Obstetrics and Gynaecology and ‡Pathological Chemistry, Royal Infirmary,Glasgow, Scotland, U.K.

Summary: Skin microvascular responses to iontophoresis of acetylcholine,an endothelium-dependent vasodilator, and sodium nitroprusside, an endothe-lium-independent vasodilator, were measured using a laser Doppler imagerwhose software controlled iontophoretic current delivery in an integratedfashion. A novel feature involved monitoring voltage across the iontophoresischambers during current application (total charge: 8 mC). Both drugs elicitedvasodilatation but with differing magnitudes and time courses, whereas cur-rent delivery with only vehicle (0.5% NaCl) present was ineffective. Duringdrug delivery a three- to fourfold difference in calculated skin resistance wasobserved between subjects, with higher resistance being associated with lowerdilator responses to both drugs. There was a significant (p < 0.0001) linearinverse correlation between perfusion.time and resistance.time integrals forboth acetylcholine (r� −0.86) and sodium nitroprusside (r � −0.96). Thiswas corrected in individual subjects by multiplying individual perfusion val-ues by the resistance.time integral, which reduced response variability. Cy-clooxygenase inhibition by aspirin apparently attenuated acetylcholine andsodium nitroprusside vasodilator responses but after correcting for skin resis-tance there was no longer any difference. Monitoring voltage across the ion-tophoretic circuit is critical, as effective drug delivery in individual subjects isinfluenced by the circuit resistance that can be corrected for. These findingshave implications for clinical studies that use the iontophoresis technique forassessing vascular function. Key Words: Iontophoresis—Vasodilatation—Endothelium—Resistance.

Increasing evidence relates impaired endothelial vaso-motor function to coronary heart disease (1). Endothelialdysfunction is not limited to the coronary circulation butis also detected in the peripheral circulation proportion-

ate to the degree of endothelial dysfunction occurring inthe coronary arteries (2). Furthermore, assessment of en-dothelial vasomotion in these peripheral arteries has beenshown to correlate with and relate directly to coronary

Received April 24, 2001; revision accepted July 4, 2001.Address correspondence and reprint requests to Dr. William R.

Ferrell at Department of Medicine, Royal Infirmary, Queen Elizabeth

Building, 10 Alexandra Parade, Glasgow G31 2ER, Scotland, UK.E-mail: [email protected].

Journal of Cardiovascular Pharmacology™39:9–17 © 2002 Lippincott Williams & Wilkins, Inc., Philadelphia

9

dysfunction (3). Thus there is considerable interest innoninvasive methods for assessing peripheral vascularfunction. Iontophoresis for transdermal delivery of thevasodilator agents acetylcholine and sodium nitroprus-side is increasingly used for this purpose. The techniqueis based on the fact that a charged molecule migratesacross the skin under the influence of an applied electri-cal field and ionized drug delivery is dependent on themagnitude of the applied current and its duration (current× time � charge, in Coulombs). In the past iontophoresishas been used in conjunction with laser Doppler flow-metry, a noninvasive method for assessing microvascularperfusion at a single point (4). More recently, iontopho-resis has been combined with laser Doppler imaging,which reduces measurement variability (5,6) because,unlike Doppler flowmetry, laser Doppler imaging mea-sures perfusion across many points (7) and thus an av-erage measure of perfusion can be computed for anychosen area. Iontophoresis of acetylcholine tests endo-thelial function since binding to muscarinic receptorswith subsequent generation of nitric oxide requires intactendothelial cells and is therefore said to be “endothe-lium-dependent.” Vasodilatation is ultimately mediatedby action of nitric oxide on vascular smooth muscle (viathe cyclic guanosine monophosphate pathway) and soiontophoresis of sodium nitroprusside, a nitric oxide do-nor, is used as an “endothelium-independent” control.This methodology has been widely used to investigatemicrovascular function in various disease states, mostcommonly diabetes mellitus, in which endothelial dys-function has been implied by decreased response to ace-tylcholine iontophoresis (8–10).

Iontophoresis has a number of advantages. It providesa direct assessment of microvascular function, is simpleto use, and, most importantly for clinical application, isnoninvasive. However, some important factors that mayinfluence iontophoretic drug responses and interpretationof iontophoresis have not been addressed. One of these isthe assumption that drug delivery is solely influenced bythe magnitude of current applied and its duration(charge). This ignores the fact that the electrical proper-ties of skin differ between subjects and could impact oneffective drug delivery.

By monitoring voltage across iontophoresis chambers,we investigated for the first time whether variation of theelectrical properties of skin between normal subjectsinfluences the magnitude of the responses to acetylcho-line and sodium nitroprusside. The results of these in-vestigations have critical implications for future clinicalstudies of endothelial function that use iontophoresis ofvasoactive drugs.

METHODS

Experiments were performed in 20 healthy subjectsages 22–50 years of both sexes with no history of pe-ripheral vascular abnormalities such as Raynaud’ssyndrome, dermatologic diseases, or systemic diseaseprocesses such as diabetes mellitus, and all were non-smokers. All subjects fasted overnight and were asked torefrain from drinking any fluids except water beforemeasurements, which were undertaken in a temperature-controlled room (23 ± 1°C) and all subjects were allowedto acclimatize for 30 min before measurement. The studywas performed according to the Declaration of Helsinki,the institutional ethics committee approved procedures,and informed consent was obtained.

IontophoresisDrug delivery was achieved using a battery-powered

constant-current iontophoresis controller (MIC-1e; MoorInstruments Ltd., Axminster, U.K). The chambers usedfor iontophoresis (ION 6; Moor Instruments Ltd.) wereconstructed of Perspex (internal diameter, 22 mm; area,3.8 cm2) with an internal platinum wire electrode. Twochambers were attached to the skin of the volar aspect ofthe forearm by means of doubled-sided adhesive disks,avoiding hair, broken skin, and superficial veins. Thechambers were connected to the anode and cathode con-nections on the iontophoresis controller. A digital mul-timeter was connected in parallel to monitor voltageacross the chambers (Fig. 1A). Because a constant cur-rent source was used, resistance values were calculatedfrom the recorded voltages using Ohm’s Law.

Control of current delivery was programmed into thesoftware for the laser Doppler imager such that currentwas switched on at the beginning of a scan and remainedon throughout the scan until the start of the followingscan. The current was then either left on for the next scanor was switched off once the total charge had been de-livered. Current duration was determined by the timetaken to complete each scan (50 s) multiplied by the totalnumber of scans programmed. To limit the iontophoresisdose, resulting from relatively long scan times, low cur-rents were used: the protocol involved incremental cur-rent delivery with four scans at 5 �A, four at 10 �A, fourat 15 �A, and two at 20 �A, giving a total charge of 8mC (Fig. 2A). Each frame is associated with the currentdelivery during that scan, although the resulting vascularresponse could be delayed owing to the time required forchemical factors to initiate it. A 2.5-ml dose of 1% ace-tylcholine chloride (Sigma Chemical Co., St. Louis, MO,U.S.A.) was introduced into the anodal chamber and 2.5ml of 1% sodium nitroprusside (Sigma) was placed in the

J. E. RAMSAY ET AL.10

J Cardiovasc Pharmacol™, Vol. 39, No. 1, 2002

cathodal chamber. Thus both agents were deliveredsimultaneously during each period of current administra-tion. The vehicle for these drugs was 0.5% NaCl. Fluidwas prevented from escaping by placing circular 32-mmcoverslips over the chambers.

Perfusion measurementsNoninvasive measurement of skin perfusion was

performed by means of a laser Doppler imager (MoorInstruments Ltd.) equipped with a red laser (wavelength,633 nm; power, 1 mW; beam diameter, 1 mm). The

FIG. 2. A. Scans taken with sodium nitroprusside (SNP) in the upper chamber and acetylcholine (ACh) in the lower chamber. Differ-ences in time courses of the drug-induced vasodilator responses are clear. The numbers in each frame refer to the applied current (inµA). B. Repetitive scans in a single subject with 0.5% NaCl vehicle. Each scan duration was 50 s, with no delay between scans.

FIG. 1. A. Diagram of the experimental arrangement. Acetylcholine (ACh) was placed in the anodal chamber and sodium nitroprusside(SNP) in the cathodal chamber. A digital voltmeter (V) was connected between the two chambers. B. Backscattered light intensityprovides the photo image and the Doppler-shifted component provides the flux image. This scan was taken at the conclusion of currentadministration. Flux is color-coded with lowest perfusion in dark blue (0 PU) and highest in dark red (500 PU).

IONTOPHORETIC ASSESSMENT OF VASCULAR FUNCTION 11


technique is based on the Doppler shift imparted by mov-ing blood cells in the underlying tissue to the backscat-tered light. The laser is scanned in a raster fashion overboth chambers and through the coverslips. The backscat-tered light is collected by photodetectors and convertedinto a signal proportional to perfusion in arbitrary per-fusion (flux) units (PU) that is displayed as a color-codedimage on a monitor (Fig. 1B). Perfusion measurementswere obtained using the imager manufacturer’s imageanalysis software by outlining a region of interest (ROI)around the internal circumference of the chamber. Sta-tistical analysis of the ROI was subsequently performedoffline to yield the median flux value across approxi-mately 700 measurement points. Twenty repetitive scanswere taken, the first being a control (before current ad-ministration), followed by the incremental current pro-tocol described above (14 scans), and followed by fivefurther scans with no current administration. The bio-logic zero was measured by taking a single scan afteroccluding the arterial blood supply with a sphygmoma-nometer cuff. This was found to be consistent betweenand within subjects with a mean (±SEM) of 23.9 ± 0.72 PU.

To determine whether the combination of fluid in theiontophoresis chambers and the coverslip affects the fluxsignal, scans were obtained comparing basal perfusion inthe anodal chamber (containing acetylcholine) and in thecathodal chamber (containing sodium nitroprusside) tothat occurring in a surrounding area of skin. The lattershowed significantly (p � 0.011, one-way ANOVA; n� 14) higher basal perfusion (71.6 ± 6.4 PU; mean ±SEM) compared with the values at the acetylcholine-containing chamber (54.4 ± 2.7 PU) and the sodiumnitroprusside-containing chamber (53.9 ± 3.7 PU), indi-cating that the chambers produce some attenuation ofthe laser Doppler imaging signal, but there was no dif-ference between chambers. Although some signal attenu-ation occurs, perfusion changes in response to interven-tion are unaffected. After occlusion of the blood supplyto the arm, the ratio of the increase in perfusion frombiologic zero showed no significant differences betweenthe anodal and cathodal chambers and adjacent skin(1.93 ± 0.09, 2.23 ± 0.15 and 2.18 ± 0.16, respectively;p � 0.26).

DrugsIn separate experiments, seven subjects were admin-

istered 600 mg aspirin orally 30 min before iontophoresisto inhibit prostaglandin synthesis. This has been shownto be an adequate period to produce maximal inhibitionof endothelium-derived prostacyclin (11). Aspirin wasdissolved in orange juice to disguise its taste, which

allowed use of plain orange juice as a placebo. Measure-ments with placebo or aspirin were performed on oneoccasion each, separated by a minimum of 14 days.

Statistical analysesMeasurement of responses was performed using raw

values, but in some cases an assessment of the overallresponse to drugs was obtained by taking the area underthe perfusion.time curve. For resistance data, resistance-.time integrals were computed. Comparisons were byANOVA or Student’s t tests, paired or unpaired as ap-propriate. All tests were two-tailed and data are ex-pressed as means ± SEM or ± SD. Because of the markeddifferences in variances between basal perfusion valuesand the maximal responses to the drugs, but the similarcoefficients of variation, log10 transformation of the datawas performed to equalize the variances and thereby per-mit parametric data analysis. The variance ratio, com-paring maximal drug responses to the control (beforecurrent application) scan, was found to vary between9–60 for the raw perfusion data, but after log transfor-mation the same variance ratio range was reduced to0.7–1.2.

Based on the raw perfusion.time integrals, the mean(± SD) between-day coefficient of variation for the ace-tylcholine response, measured in four subjects on 2 sepa-rate days, was 6.4 ± 3.3% while the within-day, between-site coefficient of variation, measured in both forearmson the same morning in four subjects, was 8.9 ± 5.3%.The same coefficients of variation for acetylcholine log-transformed perfusion data were 0.81 ± 0.39% and 1.10± 0.64%, respectively.

RESULTS

Responses to acetylcholine and sodium nitroprussideIontophoresis of acetylcholine and sodium nitro-

prusside resulted in progressive increase in perfusion,although with differing time courses (Fig. 2A). Acetyl-choline showed a more rapid onset than sodium nitro-prusside but also a rapid decline once current was ter-minated. This rapid fall in the acetylcholine responsemay be due to the presence of acetylcholinesterase inhuman blood vessels (12). Current administration withonly vehicle present failed to elicit vascular responses(Fig. 2B).

The time course of the responses to both drugs forseven subjects is shown in Figure 3A. The drug re-sponses differed significantly (p < 0.001; two-wayANOVA), but it is clear that the current protocol useddid not elicit hyperemic responses with only the vehicle



present. With responses expressed relative to cumulativecharge, both acetylcholine and sodium nitroprussidereached a plateau by the time about half the charge wasdelivered (Fig. 3B).

A striking observation is that drug delivery is influ-enced not only by the applied charge but also by theelectrical resistance of the skin, which varied in differentsubjects (Fig. 4). There is an inverse linear relationshipbetween the perfusion integral and the resistance integralfor both acetylcholine (Fig. 5A; r � −0.86) and sodiumnitroprusside (Fig. 5B; r � −0.96) (i.e., higher resistanceis associated with smaller vasodilator responses for bothdrugs and vice versa). These relationships were signifi-cant, p < 0.0001 in both cases. Correcting for this vari-able, by dividing by the integral of conductance (thereciprocal of resistance) over time, or more simply bymultiplying individual perfusion values by the integral ofresistance over time, normalizes responses. Correlatingthese corrected perfusion integrals to their respective re-sistance integrals now yields nonsignificant r2 values of0.011 for acetylcholine and 0.025 for sodium nitroprus-side (Fig. 5C, D), indicating the effectiveness of thiscorrection and lowering apparent intersubject variability.The coefficient of variation for the log perfusion integralof acetylcholine was 3.36% (n � 7) before correction forresistance and 1.86% thereafter. The respective coeffi-cients for sodium nitroprusside were 2.18% and 1.56%,respectively. These corrections reduce variability be-tween subjects by about one half and one third, respec-tively. The between-day coefficient of variation of the

acetylcholine response for the same four subjects de-scribed in methods after correction for resistance was3.04 ± 0.67% (mean ± SD), which was about half of theuncorrected value.

Influence of prostaglandins on cutaneousvascular responses

The vascular responses to both acetylcholine andsodium nitroprusside appeared to be significantlydepressed after administration of aspirin (p < 0.0001;two-way ANOVA; Fig. 6 A, B). However, it was found

FIG. 3. A. Time course of response to iontophoretic administration of acetylcholine (filled squares) and sodium nitroprusside (opensquares) dissolved in 0.5% NaCl as well as the response to this vehicle at the anode (filled circles) and cathode (open circles).Dose-dependent increase in perfusion is significant (p < 0.0001, one-way ANOVA) for drugs but not for vehicle. Mean ± SEM; n = 7subjects. B. Same perfusion data as in A, but plotted against cumulative charge.

FIG. 4. Relationship between perfusion.time integrals and theresistance.time integrals for the response to acetylcholine (ACh)in a series of subjects. The inverse nature of the relationship isclear.



that resistance during acetylcholine and sodium nitro-prusside administration was significantly greater (p �0.03; Student paired t test, n � 7; means ± SD) afteraspirin (6.25 ± 2.08 M� min) than with placebo admin-istration (4.09 ± 1.04 M� min). After correcting forresistance, there was no longer any significant differencebetween the responses for acetylcholine or sodium nitro-prusside (p � 0.64 and p � 0.75, respectively; two-wayANOVA; Fig. 6 C, D). Measurement of drug responsesin four subjects on two separate days but without aspirinadministration showed no significant differences in resis-tance between days (4.32 ± 0.94 and 4.74 ± 0.6 M� min).

DISCUSSION

There is mounting interest in the measurement ofvascular function in cardiovascular research, becauseendothelial dysfunction may be an early feature of theatherogenic process and is amenable to therapeutic in-tervention. It can be argued that endothelial functionintegrates the stress of other risk factors. Laser Dopplerimaging with iontophoresis is the most recent techniqueused for assessing endothelial function and is particularlyattractive because of its noninvasive nature. Althoughiontophoresis only assesses the cutaneous microcircula-

FIG. 5. A. Relationship between the log transformed perfusion.time integral and the resistance.time integral for acetylcholine (ACh).B. Relation between the log-transformed perfusion.time integral and the resistance.time integral for sodium nitroprusside (SNP) (n = 14and p < 0.0001 for both A and B). C. Correction of log-transformed perfusion.time integral (multiplying flux by the resistance.time integral)for ACh and plotting against the resistance.time integral no longer yields a significant correlation. D. Similarly, the corrected data for SNPno longer shows a correlation.



tion, this is effectively a robust surrogate marker of vas-cular function in other vascular beds. Reduced respon-siveness to iontophoretic administration of acetylcholinehas been observed in diabetes (8–10) and hypercholes-terolemia (13), and in both these conditions there is aparallel reduction of the acetylcholine response in theforearm circulation (predominantly a skeletal musclevascular bed) assessed by venous occlusion plethysmog-raphy (14,15). Moreover, attenuated response in the skinof heart transplant patients to acetylcholine iontophoresis(16) is paralleled by reduced responsiveness of coronaryblood vessels to acetylcholine in this group (17). In ad-dition, acetylcholine-induced vasodilatation is reduced inboth forearm musculature (18,19) and skin (19) in pa-tients with essential hypertension. Thus many conditionsaffecting the cardiovascular system appear to result inglobal endothelial dysfunction and assessment of the cu-taneous microcirculation yields valuable insights into pe-ripheral vascular function.

The current protocol used in the current study resultedin an overall lower charge (8 mC) than that used in

previous studies (5,6) and this, combined with a largesurface area for iontophoresis and the use of a weaksaline vehicle, prevented the development of hyperemicartefacts at the cathode. This obviated the need to use atopical anesthetic such as lidocaine-prilocaine cream(5,6), which produces cutaneous vasoconstriction (20)and which presents a problem because any vasodilatorresponses would then be superimposed on a basal vaso-constrictor tone. In preliminary experiments we have ob-served that iontophoresis of acetylcholine and sodiumnitroprusside to skin treated with lidocaine-prilocainecream elicited substantially reduced vasodilatorresponses (by 30%–50%, unpublished observations) us-ing the protocol used in the present investigation.

Past work aimed at establishing the role of prostaglan-dins in mediating the response to acetylcholine has pro-duced contradictory results. It was previously observedthat the vasodilator response to iontophoresis of acetyl-choline was reduced by administration of oral (21) orintravenous aspirin (22). However, Morris and Shore (5)found no difference in the acetylcholine response after

FIG. 6. Vascular responses to administration of acetylcholine (ACh; A) and sodium nitroprusside (SNP; B) in 0.5% NaCl vehicle before(open symbols) and after (closed symbols) administration of aspirin. For both ACh and SNP the responses before and after administrationof aspirin differ significantly (p < 0.0001, two-way ANOVA). Vascular response to administration of ACh (C) and SNP (D) before (opensymbols) and after (closed symbols) administration of aspirin after correcting flux values for resistance in each subject. (Mean ± SEM;n = 7 subjects).



oral aspirin. Our finding of an apparent difference in thevascular responses to both acetylcholine and sodium ni-troprusside after oral aspirin suggests a possible expla-nation for discrepancies between these earlier studies.After correction for resistance, there was no longer anydifference in the acetylcholine and sodium nitroprussideresponses and this may indicate that variations in resis-tance across the iontophoresis circuit during drug admin-istration could have been a confounding factor in previ-ous studies. Aspirin itself could have changed skinresistance, and the time control experiment supports this,as no change in resistance was found when measured onseparate days but without aspirin administration. Thechange in resistance could not be ascribed to variationsin room temperature, because all measurements were un-dertaken in a temperature-controlled room and subjectswere allowed to acclimatize. It is therefore possible thatin previous studies, variations in skin resistance mighthave influenced effective drug delivery, leading to dif-fering results. This emphasizes the importance of cor-recting for resistance to avoid potentially spurious re-sults, particularly if measurements are taken at differentsites or on different days.

The variation in calculated resistance between subjectssuggests that there may be variable numbers of resistancepathways in the skin of different subjects. As the equip-ment and the composition of the solutions used was thesame between subjects, the site of the resistance must berelated to the skin. It might be the case that lower resis-tance is associated with a greater availability of low-resistance pathways, such as sweat ducts or hair follicles,which are close to blood vessels. This is quite likelygiven that sweat glands and their associated ducts, aswell as hair follicles, are known to be richly vascularized(23). Higher resistance may be associated with high-resistance pathways for ion flow, mostly likely throughthe stratum corneum, and these are more remote fromblood vessels. Thus effective drug delivery could differ,even though the charge remains constant. In effect, al-though the same total amount of drug may be delivered,the vascular response could differ depending on the rela-tive number of low-resistance pathways available.

The inverse relationship between drug response andresistance, with calculated resistance integrals showingon average more than twofold variation between sub-jects, indicates that skin resistance may contribute to in-tersubject variability of drug responses. This has impor-tant implications for interpreting responses to drugadministration in clinical studies. This factor has notbeen taken into account in any previous study using ion-tophoresis and variations in observed responses betweensubject groups could have been influenced by a system-

atic variation in resistance rather than a true difference invascular reactivity. This said, it remains likely that dia-betic patients do have vascular dysfunction, as most stud-ies using venous occlusion plethysmography and intra-arterial drug administration show a reduced response toacetylcholine compared with control subjects (14,24,25).However, observed differences to iontophoresis of drugsat different sites could be explained by variations in skinresistance. Reduced vascular responses have been ob-served in the skin over the dorsum of the foot comparedwith the forearm for both normal (10,26) and diabeticsubjects (27). Similarly, marked differences in responseto iontophoretic administration of acetylcholine occur atvarious sites on the hand and forearm in normal subjects(28), assessed by using laser Doppler flowmetry. Evenwhen laser Doppler imaging is used, between-site va-riations in the magnitude of responses occur, althoughvariability can be reduced by ensuring that consecutivemeasurements are taken from the same site (6). Thisreinforces the need to monitor voltage during iontopho-resis so that resistance can be estimated. Only by doingthis is it possible to accurately assess the extent to whicheffective drug delivery is affected by skin resistance andto correct for this variable.

In conclusion, this study has shown for the first timethat resistance is an important but previously unrecog-nized variable that influences iontophoresis. The inverserelationship between skin resistance and blood flow re-sponses to both acetylcholine and sodium nitroprussideindicates that resistance influences effective drug deliv-ery. We anticipate that correction for resistance, coupledwith use of appropriate vehicles, chambers, and ionto-phoresis protocols, will lead to significant improvementof the iontophoresis technique and permit its additionaldevelopment, thereby further increasing its robustness asa noninvasive tool for assessing endothelial function inclinical studies.

Acknowledgment: The support of the Scottish HospitalsEndowment Research Trust and the North Glasgow Uni-versity NHS Trust is gratefully acknowledged.

REFERENCES

1. Sattar N, Petrie J, Jaap A. The atherogenic lipoproteinphenotype and vascular endothelial dysfunction. Athero-sclerosis 1998;138:229–35.

2. Anderson TJ, Uehata A, Gerhard MD, et al. Close rela-tionship of endothelial function in the human coronary andperipheral circulation. J Am Coll Cardiol 1995;26:1235–41.

3. Zeiher AM. Endothelial vasodilator dysfunction: pathoge-netic link to myocardial ischemia or epiphenomenon? Lan-cet 1996;348:S10–S12.



4. Nilsson GE, Tenland T, Oberg PA. A new instrument forcontinuous measurement of tissue blood flow by light beat-ing spectroscopy. IEEE Trans Biomed Eng 1980;27:597–604.

5. Morris SJ, Shore AC. Skin blood flow responses to theiontophoresis of acetylcholine and sodium nitroprusside inman: possible mechanisms. J Physiol 1996;496:531–42.

6. Kubli S, Waeber B, Dalle-Ave A, et al. Reproducibility oflaser Doppler imaging of skin blood flow as a tool to assessendothelial function. J Cardiovasc Pharmacol 2000;36:640–48.

7. Wårdell K, Jakobsson A, Nilsson GE Laser Doppler per-fusion imaging by dynamic light scattering. IEEE TransBiomed Eng 1993;40:309–16.

8. Morris SJ, Shore AC, Tooke JE. Responses of the skinmicrocirculation to acetylcholine and sodium nitroprussidein patients with NIDDM. Diabetologia 1995;38:1337–44.

9. Pitie D, Watkins PJ, Edmonds ME. NO-dependent smoothmuscle vasodilatation is reduced in NIDDM patients withperipheral sensory neuropathy. Diabetic Med 1997;14:284–90.

10. Caballero AE, Arora S, Saouaf R, et al. Microvascular andmacrovascular reactivity is reduced in subjects at risk fortype 2 diabetes. Diabetes 1999;48:1856–62.

11. Heavey DJ, Barrow SE, Hickling NE, et al. Aspirin causesshort-lived inhibition of bradykinin-stimulated prostacy-clin production in man. Nature 1985;318:186–8.

12. Walch L, Norel X, Leconte B, et al. Cholinergic control ofhuman and animal pulmonary vascular tone. Therapie1999;54:99–102.

13. Khan F, Litchfield SJ, Stonebridge PA, et al. Lipid-lower-ing and skin vascular responses in patients with hypercho-lesterolaemia and peripheral arterial obstructive disease.Vasc Med 1999;4:233–38.

14. Williams SB, Cusco JA, Roddy M-A, et al. Impaired nitricoxide-mediated vasodilation in patients with non-insulin-dependent diabetes mellitus. J Am Coll Cardiol 1996;27:567–74.

15. Lewis TV, Dart AM, Chin-Dusting JPF. Endothelium-de-pendent relaxation by acetylcholine is impaired in hyper-triglyceridemic humans with normal levels of plasma LDLcholesterol. J Am Coll Cardiol 1999;33:805–12.

16. Andreassen AK, Kvernebo K, Jørgensen B, et al. Exercisecapacity in heart transplant recipients: relation to impairedendothelium-dependent vasodilatation of the peripheralmicrocirculation. Am Heart J 1998;136:320–328.

17. Treasure CB, Vita JA, Ganz P Loss of coronary microvas-

cular response to acetylcholine in cardiac transplant pa-tients. Circulation 1992;86:1156–64.

18. Panza JA, Quyyumi AA, Brush JE, et al. Abnormal endo-thelium-dependent vascular relaxation in patients with es-sential hypertension. N Engl J Med 1990;323:22–7.

19. Rossi M, Taddei S, Fabbri A, et al. Cutaneous vasodilata-tion to acetylcholine in patients with essential hyperten-sion. J Cardiovasc Pharmacol 1997;29:406–11.

20. Bjerring P, Andersen PH, Arendt-Nielsen L. Vascular re-sponse of human skin after analgesia with EMLA cream.Brit J Anaesth 1989;63:655–60.

21. Khan F, Davidson NC, Littelford RC, et al. Cutaneousvascular responses to acetylcholine are mediated by a pros-tacyclin-dependent mechanism in man. Vasc Med 1997;2:82–6.

22. Noon JP, Walker BR, Hand MF, et al. Studies with ionto-phoretic administration of drugs to human dermal vesselsin vivo: cholinergic vasodilatation is mediated by dilatorprostanoids rather than nitric oxide. Br J Pharmacol 1998;45:545–50.

23. Ellis RA. Vascular patterns of skin. In: Montagna W, EllisRA, eds. Blood vessels and circulation. New York: Per-gamon Press, 1961:20–37.

24. Watts GF, O’Brien SF, Silvester W, et al. Impaired endo-thelium-dependent and independent dilatation of forearmresistance arteries in men with diet-treated non-insulin-dependent diabetes: role of dyslipidaemia. Clin Sci 1996;91:567–73.

25. Lim SC, Caballero AE, Smakowski P, et al. Soluble inter-cellular adhesion molecule, vascular cell adhesion mol-ecule, and impaired microvascular reactivity are earlymarkers of vasculopathy in type 2 diabetic individualswithout microalbuminuria. Diabetes Care 1999;22:1865–70.

26. Hu J, Norman N, Wallensteen M, et al. Increased largearterial stiffness and impaired acetylcholine induced skinvasodilatation in women with previous gestational diabetesmellitus. Br J Obstet Gynaecol 1998;105:1279–87.

27. Arora S, Smakowski P, Frykberg RG, et al. Differences infoot and forearm skin microcirculation in diabetic patientswith and without neuropathy. Diabetes Care 1998;21:1339–44.

28. Gardner-Medwin JM, Taylor JY, Macdonald IA, et al. Aninvestigation into variability in microvascular skin bloodflow and the responses to transdermal delivery of acetyl-choline at different sites in the forearm and hand. Br J ClinPharmacol 1997;43:391–7.



factors critical to iontophoretic assessment of vascular …moor-inst.jp/products/factors critical...

Documents