3018

Cyclosporine Impairs Release ofEndothelium-Derived Relaxing Factors in

Epicardial and Resistance Coronary Arteries

Krishnankutty Sudhir, MD, PhD; John S. MacGregor, MD, PhD; Teresa DeMarco, MD;Christianne J.M. De Groot, MD; Robert N. Taylor, MD, PhD; Tony M. Chou, MD;

Paul G. Yock, MD; Kanu Chatterjee, MB, FRCP

Background Cyclosporin A is reported to impair endothe-lium-mediated vasorelaxation and induce endothelin releasein some noncoronary vascular beds. We wished to determinewhether acute cyclosporine administration induces endothelialdysfunction in coronary conductance or resistance arteries.Methods and Results We examined the effect of intracoro-

nary acetylcholine, N&-nitro-L-arginine methyl ester (L-NAME), L-arginine, nitroglycerin, and adenosine before andafter acute cyclosporine administration (3 mg/kg IV over 30minutes) in anesthetized dogs. Flow velocity was measuredwith a 0.014-in Doppler wire to assess resistance vessel re-sponses, and epicardial coronary lumen area was simultane-ously measured with a 4.3F, 30-MHz imaging catheter insertedover the Doppler wire. In 6 dogs, acetylcholine-induced in-crease in flow velocity was attenuated by cyclosporine invehicle (137% to 55% at 10-5 mol/L, P<.001), as was acetyl-choline-induced epicardial vasodilation (14.1% to 6.7% at 10-5mol/L, P<.001). Vasodilation in response to intracoronarynitroglycerin (200 ,ug) and adenosine (6 mg) were unchanged

E ndothelial dysfunction observed in the coronaryarteries of orthotopic cardiac transplant recip-ients1,2 is said to be due to the interaction of

several factors such as hypercholesterolemia,34 hyper-tension, and immune-mediated vascular disease.5 Therole of the immunosuppressive agent cyclosporin A,widely used after cardiac transplantation, has not beenexamined as a potential cause of coronary endothelialdysfunction. In noncoronary beds such as the rat renalartery,6 the isolated rat aorta,7 and human subcutane-ous resistance vessels,8 cyclosporine has been shown toimpair endothelium-mediated vasorelaxation. In addi-tion, cyclosporine is reported to induce release ofendothelin,9,10 which may mediate its renal vasoconstric-tor effect."1 Further, a direct toxic effect of cyclosporinehas been shown in cultured rat microvascular endothe-lial cells.12,13

In the present study, we hypothesized that cyclospor-ine might induce endothelial dysfunction in the coro-nary circulation. We examined the effect of acutecyclosporine administration in vivo in the canine coro-

Received April 7, 1994; revision accepted July 11, 1994.From the Cardiovascular Research Institute and Division of

Cardiology, University of California at San Francisco.Correspondence to Krishnankutty Sudhir, MD, PhD, Alfred

Baker Medical Unit, Baker Medical Research Institute, Commer-cial Rd, Prahran, VIC 3181, Australia.C 1994 American Heart Association, Inc.

by cyclosporine. Epicardial vasoconstriction with L-NAME(10-4 mol/L) was reduced by cyclosporine (Pre, 7.4+0.9%;Post, 2.6+1.2%; P=.04), but L-arginine (10-4 mol/L) had noeffect after cyclosporine. In another 5 dogs, pure cyclosporineimpaired acetylcholine-induced vasodilatation to the samedegree as cyclosporine in vehicle (Cremophor); vehicle infu-sion did not impair endothelial function. In 5 more dogs,cyclosporine did not increase either arterial or coronary sinusconcentrations of endothelin-1.

Conclusions The present study shows that cyclosporineacutely impairs release of endothelium-derived relaxing factorin canine conductance and resistance coronary arteries andprovides evidence for decreased epicardial nitric oxide releaseafter cyclosporine. The potential contribution of acute cyclo-sporine-induced coronary endothelial dysfunction to post-transplant vasculopathy needs further study. (Circulation.1994;90:3018-3023.)Key Words * acetylcholine * N'O-nitro-L-arginine methyl

ester * endothelin * transplantation

nary epicardial and microcirculation using simultaneoustwo-dimensional (2D) and Doppler ultrasound.'4 Spe-cifically, we studied vasorelaxation responses to endo-thelium-dependent and endothelium-independent va-sodilators to determine the influence of cyclosporine onendothelial function. We also examined the effect ofacute cyclosporine administration on endothelin releasein the peripheral circulation and across the coronaryvascular bed.

MethodsSixteen dogs (mean body weight, 27±0.9 kg) were anesthe-

tized with Innovar (0.04 mg/kg SC) and sodium pentobarbital(15 mg/kg IV), with additional doses as needed to maintainthe level of anesthesia. The dogs were mechanically ventilatedwith room air, blood pressure was continuously monitoredfrom a cannula placed in the left internal carotid artery, andheart rate was monitored from the ECG. All studies con-formed to the "Position of the American Heart Association onResearch Animal Use" (November 11, 1984).

Catheterization ProceduresIn the first 11 dogs, the left main coronary artery was

cannulated via the transfemoral approach under fluoroscopicguidance by use of an 8F canine guiding catheter (AdvancedCardiovascular Systems). A 0.014-in Doppler wire (FloWire,Cardiometrics Inc) was first introduced through the 8F guidingcatheter, after which the 2D ultrasound imaging catheter(Cardiovascular Imaging Systems, CVIS) was introduced di-

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

Sudhir et al Cyclosporine and Coronary Endothelial Function 3019

rectly over the Doppler wire into the mid segment of thecircumflex coronary artery. As previously described, theDoppler transducer was positioned approximately 2 cm distalto the tip of the imaging catheter.14 In the next 5 dogs, thecoronary sinus was cannulated via the right internal jugularvein with an 8F multipurpose guiding catheter. This catheterwas used to sample coronary venous blood for endothelinlevels.

2D Ultrasound System Description andImage Analysis2D ultrasound images were obtained with a commercially

available imaging system (CVIS). The ultrasound catheter(4.3F) has a fixed 30-MHz transducer and a rotating mirrorassembly. Images were displayed on a video monitor; axialresolution was -150 ,um and lateral resolution -250 ,um.Gain, contrast, and reject settings were adjusted by the oper-ator to yield a well-balanced gray-scale appearance on thevideo display. Real-time images were stored on high-qualitysuper VHS videotape for subsequent off-line analysis. Aspreviously described,'5 selected portions of the videotape weredigitized (12 bits, Rasterops 324) in real time (30 frames persecond) and transferred to a computer disk for off-line deter-mination of luminal area.

Doppler Ultrasound System DescriptionCoronary blood flow velocities were measured by use of a

steerable Doppler guide wire. This guide wire system has aminiature ultrasound crystal that transmits signals at a carrierfrequency of 12 MHz and receives pulsed-wave Dopplerultrasound signals. The ultrasound beam has an angular widthof approximately 250, and the sample volume is located at adistance of 5 mm from the guide wire tip. The Doppler signalsare analyzed by a FloMap instrument (Cardiometrics Inc) thatuses dedicated digital signal processing chips to perform thefast Fourier transformation required for the spectral display.The signals were then transformed into a gray scale, and theresultant spectrum was displayed on a monitor screen. TheECG, the arterial pressure wave tracing, and quantitativemeasurements of average peak velocity (APV) were alsodisplayed. The monitor display was continuously recorded on aVHS videotape.

Vasoactive AgentsDimensions and flow in the coronary arteries are influenced

by perfusion pressure, which varies with systemic arterialblood pressure.'6 To minimize effects on the systemic vascula-ture that would alter blood pressure and thus affect interpre-tation of results, all drugs (with the exception of cyclosporineand Cremophor) were infused directly into the coronarycirculation through the guiding catheter in the left maincoronary artery. Previous studies with saline infusions ofsimilar volume have shown that the "bolus" effect had littleinfluence on flow beyond 20 seconds.17 Measurements ofcoronary artery cross-sectional area (CSA) were thereforemade at 30 seconds, and every 30 seconds thereafter with eachadministration. Measurements of flow velocity were madecontinuously on-line. Observations reported indicate peakeffects of the administered drugs. Sufficient "washout" periodsof 15 minutes were allowed between administration of differ-ent drugs, with documentation that coronary artery dimen-sions and flow velocity returned to the baseline state. Cyclo-sporine and Cremophor were obtained from SandozPharmaceutical Corp and administered in the doses describedbelow after dilution in 20 mL of autologous blood. Adenosinewas obtained from Fujisawa Pharmaceuticals. All other drugswere obtained from Sigma Chemical Co.

Protocol 1In 6 dogs, by use of an approach similar to that previously

described18 and assuming a coronary blood flow of 80 mL/

min,'7 increasing concentrations of intracoronary acetylcho-line were administered over 1-minute periods to achieveconcentrations from 10-7 to 10-5 mol/L. Intracoronary N0-nitro-L-arginine methyl ester (L-NAME) was administeredsimilarly to achieve increasing concentrations from 10-6 to 10-4mol/L. Intracoronary L-arginine (L-Arg) was administeredimmediately after L-NAME (with no "washout" period be-tween them) over a 1-minute period to obtain a final concen-tration of 10-` mol/L. Intracoronary adenosine (6 mg) andnitroglycerin (200 ,ug) were given as boluses as previouslydescribed.16 Next, cyclosporine in Cremophor (3 mg/kg bodywt IV) was infused over a period of 30 minutes. Intracoronaryacetylcholine, L-NAME, L-Arg, adenosine, and nitroglycerininfusions were then repeated to assess the influence of cyclo-sporine on endothelium-dependent and endothelium-indepen-dent coronary vasomotion. A peripheral venous blood samplewas drawn 15 minutes after the termination of the cyclosporineinfusion for measurement of cyclosporine levels.

Protocol 2In the next 5 dogs, acetylcholine-induced vasodilation was

measured before and after a 30-minute infusion of the vehicle,Cremophor, 2 mL. Next, cyclosporine (3 mg/kg body wt) wasadministered as pure powder dissolved in 20 mL of autologousblood over a period of 30 minutes, and the effect of acetylcho-line-induced vasodilation was measured once again.

Protocol 3In a separate series of 5 dogs, arterial and coronary sinus

blood was collected in 5 mmol/L EDTA containing 1 ,g/mLaprotinin for determination of endothelin levels before andafter a 30-minute infusion of pure cyclosporine powder (3mg/kg body wt) dissolved in autologous blood.

Endothelin AssaysEndothelin-1 was measured in extracted plasma by a specific

radioimmunoassay that recognizes the mature 1-21 peptide(RPA 555) according to the following protocol: C18 Amprepcolumns (Amersham) were equilibrated by washing with 3 mLmethanol, 2 mL water, and 2 mL 10% acetic acid, and 0.5 to1.0 mL plasma with an equal volume of 20% acetic acid wasapplied. The columns were washed with 2 mL of 10% aceticacid, followed by 3 mL of ethyl acetate, and eluted with amixture of 80% methanol and 20% 50 mmol/L ammoniumbicarbonate. The eluates were evaporated under nitrogen gas,reconstituted in assay buffer, and assayed by a double-antibodyradioimmunoassay according to the manufacturer's recom-mendations. Cross-reactivities of the radioimmunoassay were52% for endothelin-3 and 0.4% for big endothelin. The assaywas linear for endothelin-1 in dog plasma over a range ofplasma volumes of 0.25 to 2 mL and over a concentration rangeof 2 to 400 pg/mL. The assay sensitivity was 1.2 pg/mL, withinterassay and intra-assay coefficients of variation of 15% and7%, respectively.

Calculations and Statistical AnalysisLuminal CSA at baseline and after administration of drugs

was determined by computer-assisted planimetry. Measure-ments were gated to end-systolic and end-diastolic dimensions,respectively. As previously described,'4 a mean CSA value wasobtained by correcting for the fractional diastolic time interval(fDTI, the duration of diastole as a fraction of the cardiaccycle length) and the fractional systolic time interval (fSTI, theduration of systole as a fraction of the cardiac cycle length) asfollows: mean CSA= (fDTI x end-diastolic CSA)+ (fSTIx end-systolic CSA). Volumetric coronary blood flow (CBF) wascalculated from the relation CBF=mean CSAxAPVx0.5, aspreviously described and validated in a dog model.'4'9 Con-centration-response curves for CSA and Doppler-derivedAPV were analyzed by one-way ANOVA for repeated mea-sures followed by a post hoc Student-Newman-Keuls test. The

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

3020 Circulation Vol 90, No 6 December 1994

Changes in Mean Arterial Pressure and Heart RateInduced by the Various Pharmacological Agents InfusedBefore and After Cyclosporine

Before AfterCyclosporine Cyclosporine

MAP, HR, MAP, HR,Drug mm Hg bpm mm Hg bpmBaseline 90+5 140+5 95±4 137+7

AcetylCholine 10-7 90±6 136±6 90+6 142+5

Acetylcholine 10-6 92±4 142+7 88+5 139+4

Acetylcholine 10-5 94+5 135+6 87±6 134±6

L-NAME 10-6 94±3 144-+-6 90+5 141+6

L-NAME 10-5 97+7 141±8 96±6 136±6

L-NAME 10-4 99+4 133±7 96+9 138±6

L-Arg 10-4 92±5 145±9 91+5 139±6

Adenosine 83±11 147±8 84+9 144±8

Nitroglycerin 200 ,ug 84±10 150±14 85±8 150±8

MAP indicates mean arterial pressure; HR, heart rate; bpm,beats per minute; L-NAME, Nw-nitro-L-arginine methyl ester; andL-Arg, L-arginine.

dose-response relations to acetylcholine and L-NAME beforeand after cyclosporine were compared by two-way repeated-measures ANOVA. The effects of adenosine and nitroglycerinbefore and after cyclosporine were compared by Student'spaired t test, as were endothelin levels. Values are expressed asmean ± SEM.

ResultsIn all 16 dogs, interpretable 2D images and Doppler

flow spectra were obtained. There were no instances ofarrhythmias, spasm, or thrombosis during catheter/Doppler wire placement.

Resting Coronary Dimensions and Flow andCyclosporine LevelsMean resting heart rate and mean arterial pressure at

baseline are shown in the Table. Mean resting coronaryCSA was 9.2+0.6 mm2, and the mean value for APVwas 27±6 cm/s. Mean volumetric coronary blood flow atbaseline was 75.5±15.2 mL/min. Peak cyclosporine lev-els achieved 15 minutes after the end of the infusionwere 791±98 ,g/L.

Effect of Cyclosporine on Acetylcholine-InducedIncreases in Coronary CSA and Flow Velocity

Acetylcholine caused a dose-dependent increase inmean coronary CSA and APV, as shown in Fig 1. Themean increase in CSA at 10-5 mol/L was 14.1+2%, andthe mean increase in APV was 137±45% at the 10-5mol/L concentration of acetylcholine. As a result, volu-metric coronary blood flow increased by 172±42% at the10-5 mol/L concentration of acetylcholine. The peakeffect on CSA was observed at 30 seconds, and flowvelocity reached its peak at 40±5 seconds. Mean arterialpressure remained unchanged. A transient slowing of theheart rate was observed at the highest concentration(10-5 mol/L), lasting about 5 to 10 beats, with fullrecovery at 30 seconds. No other significant changes inheart rate were observed with any dose (Table).

CSA(sq.mm)

10 P

Post 28-

Baseline -7 -6 .5Acetylcholine

logM

APV (cm/sec)60

50

40

30

20Post 1

10Baseline -7 -6 -5

AcetylcholinelogM

FIG 1. Graphs showing effect of increasing concentrations ofacetylcholine on epicardial coronary cross-sectional area (CSA)measured by intravascular ultrasound imaging and averagepeak velocity (APV) measured by intravascular Doppler. * indi-cates no cyclosporine; c, Cremophor (vehicle); *, cyclosporine+ vehicle; and o, pure cyclosporine powder.

After administration of cyclosporine, mean arterialpressure was unchanged, as was heart rate (Table). Therewas a 5±2% decrease in baseline CSA (P=.08) but nochange in APV, and volumetric blood flow decreased to61.9±6.5 mL/min (P=.09). There was a significant atten-uation of the increase in CSA to 6.7±1.3% at 10-5 mol/L(F=12.83, P=.016 for precyclosporine versus postcyclo-sporine dose-response curves from two-way ANOVA)and APV to 55 ±22.1% at 10-5 mol/L (F=37.44, P<.01 forprecyclosporine versus postcyclosporine dose-responsecurves from two-way ANOVA) in response to acetylcho-line. At the 10` mol/L concentration of acetylcholine, theincrease in volumetric flow after cyclosporine was reducedto 65±18% (P<.01). With respect to both CSA and APV,the concentration-response curve to acetylcholine afteradministration of pure cyclosporine powder was identicalto that after cyclosporine in Cremophor (Fig 1). Cremo-phor vehicle per se did not result in any shift in theacetylcholine concentration-response relation with re-spect to either CSA or APV (Fig 1). There was a smalldecrease in baseline CSA in response to Cremophor,which did not reach statistical significance (P=.15) (Fig 1).

Effect of Cyclosporine on L-NAME- andL-Arg-Induced Changes in Coronary CSA andFlow Velocity

Before cyclosporine administration, L-NAME causeda significant dose-dependent decrease in CSA, as shownin Fig 2, the peak effect occurring at 6±2 minutes.

* Pre CYACSA (sq.mm) T M Post CYA

9.51

9.0_

8.5

8.01Baseline -6 -5 -4 L-Arg

L-NAME (log M) (-4 logM)FIG 2. Graph showing effect of increasing concentrations ofN0-nitro-L-arginine methyl ester (L-NAME) and L-arginine (L-Arg)on epicardial coronary cross-sectional area (CSA) measured byintravascular ultrasound imaging. CYA indicates cyclosporin A.

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

Sudhir et al Cyclosporine and Coronary Endothelial Function 3021

Pre-CyA12

CSAsq.mmn

Pre-CyA Post-CyAFIG 3. Bar graph showing coronary flow reserve measured asthe ratio of peak to baseline flow velocity in response toadenosine, showing no change after administration of cyclospor-ine. CyA indicates cyclosporin A.

There was no significant change in APV (31±6, 30 ±4,and 28±4 cm/s at concentrations of 10-6, 10-, and 10`4mol/L, respectively). There was a small decrease involumetric coronary blood flow at the iO-` mol/L con-

centration of L-NAME (12±6%, P=.04). No changes insystemic arterial pressure or heart rate were observedwith L-NAME (Table).

After cyclosporine administration, there was again a

6±2% decrease in baseline CSA (P=.06) but not APVand a significantly attenuated vasoconstrictor response

to L-NAME in the epicardial coronary arteries (Precy-closporine, 7.4±0.9% decrease in CSA; postcyclospor-ine, 2.6+1.2% decrease in CSA at 10` mol/L; P=.04).The coronary flow velocity response to L-NAME aftercyclosporine was not different from that before cyclo-sporine, with no change in APV at any dose ofL-NAME (28±7, 24±6, and 27+8 cm/s at concentra-tions of 10-6, i0-5, and 10`4 mol/L, respectively). Aftercyclosporine, there was no significant change in theeffect of L-NAME on volumetric coronary blood flow(10±3% at the 10-' mol/L concentration of L-NAME).L-Arg (10-4 mol/L), infused immediately after

L-NAME, reversed the epicardial vasoconstriction in-duced by blockade of NO synthesis before cyclosporine(Fig 2), the reversal being completed in 4±1 minutes.After cyclosporine, L-Arg, infused after L-NAME,failed to increase CSA. There was, however, no increasein flow velocity in response to L-Arg either before(APV 29±7 cm/s) or after (APV 31±5 cm/s) cyclospor-ine. There was no change in blood pressure or heart ratewith L-Arg either before or after cyclosporine (Table).After L-NAME (10-a mol/L), L-Arg increased volumet-ric coronary blood flow before cyclosporine by 23±4%but did not change volumetric coronary blood flow aftercyclosporine (-3±2%).

Effect of Cyclosporine on Coronary Flow ReserveA transient decrease in heart rate (range, 18 to 40

beats per minute) was observed in response to adeno-sine, lasting a maximum of 15 seconds, with full recov-

ery by 30 seconds. Blood pressure was not significantlyreduced (Table). Adenosine did not change coronaryCSA (baseline, 9.1+±0.5 mm2; adenosine, 9.3+0.4 mm2'P=NS). The ratio of peak flow to baseline flow in

response to intracoronary adenosine was 3.5±0.4 be-fore cyclosporine and 3.3±0.1 after cyclosporine(P=NS) (Fig 3). The peak response to adenosine was

so 50

APV 4o T APV 40

cm/sec cm/sec

30 30'

~~~~~~~~20

Baseline Nitro Baseline Nitro

FIG 4. Bar graphs showing increase in cross-sectional area

(CSA) and average peak velocity (APV) in response to nitroglyc-erin before (left) and after (right) cyclosporine, showing nochange in nitroglycerin (nitro)-induced vasodilation after cyclo-sporine. CyA indicates cyclosporin A.

observed between 45 seconds and 1 minute afteradministration.

Effect of Cyclosporine on the Coronary VasodilatorResponse to NitroglycerinThere was no significant change in either heart rate or

blood pressure in response to intracoronary nitroglyc-erin 200 jig (Table). Nitroglycerin caused an increase inboth CSA and APV before cyclosporine administration.After cyclosporine, nitroglycerin-induced increases inCSA and APV were not significantly different fromthose before cyclosporine (Fig 4). The magnitude of theincrease in volumetric blood flow induced by nitroglyc-erin was also similar before and after cyclosporine(before cyclosporine, 76.3+13 to 115.4+11 mL/min;after cyclosporine, 65±4 to 107.9±9 mL/min). The peakresponse to nitroglycerin was observed 2 minutes afteradministration.

Effect of Cyclosporine on Endothelin LevelsCyclosporine did not increase either arterial or cor-

onary sinus endothelin (arterial: before cyclosporine,26.6±5.1 pg/mL; after cyclosporine, 26.0±2.3 pg/mL;P=NS; coronary sinus: before cyclosporine, 29.8±1.5pg/mL; after cyclosporine, 28.0±6.1 pg/mL; P=NS).

DiscussionThe present study demonstrates that acute intrave-

nous administration of cyclosporin A impairs acetylcho-line-induced vasorelaxation in the coronary circulation.This impairment was seen in both conductance andresistance coronary arteries and was not restoredacutely by L-Arg. In the epicardial coronary arteries,NO release is reduced after cyclosporine administra-tion. Endothelium-independent relaxation in responseto nitroglycerin and coronary flow reserve as assessedby intracoronary adenosine appeared to be preservedafter cyclosporine administration. No significant in-crease in either peripheral or coronary venous endothe-lin levels was observed after cyclosporine.

In general, impaired endothelium-dependent vasorp-laxation may be mediated by several potential mecha-nisms, such as decreased release of NO or a decreasedsensitivity of vascular smooth muscle to NO. Our datashowing a decrease in baseline CSA after cyclosporineand a significantly decreased vasoconstrictor response

Coronary flow reserve Post-CyA

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

3022 Circulation Vol 90, No 6 December 1994

to L-NAME in epicardial coronary arteries suggestimpaired production of NO in the conductance vesselsafter exposure to cyclosporine. Previous studies fromour laboratory18 and others20 suggest that NO produc-tion may contribute less to acetylcholine-induced va-sorelaxation in coronary resistance vessels comparedwith the conductance vessels, an observation that hasnow been reported in humans as well.21 In the presentstudy, L-NAME did not have any significant influenceon coronary flow, and cyclosporine did not influencethis response in the resistance vessels. This wouldsuggest that cyclosporine-induced impairment of acetyl-choline-induced vasorelaxation in coronary resistancearteries may be mediated by endothelium-derived relax-ing factors other than NO. Rather than interferencewith a specific endothelial pathway, a generalized toxiceffect on the endothelium in both the large and smallcoronary vessels could account for the impairment inacetylcholine-induced increases in coronary CSA andblood flow as well as a decrease in basal epicardialarterial release of NO. Indeed, cyclosporine has beenshown to directly cause abnormalities in the endothelialcytoplasm and nucleus and to inhibit endothelial cellreplication.1213 Toxic endothelial injury would also ex-plain the lack of response to L-Arg; by contrast, com-petitive inhibition of NO synthase should have beenreversed by increasing the concentration of its substrate,L-Arg.

Cyclosporine clearly did not impair endothelium-independent relaxation in our study, as demonstrated bypreserved relaxation in response to nitroglycerin. Nitro-glycerin acts via conversion to NO in vascular smoothmuscle22: hence, the responsiveness of vascular smoothmuscle to NO is not influenced by cyclosporine. Previ-ous studies with acute cyclosporine administration haveshown similar preservation of endothelium-indepen-dent relaxation in noncoronary vessels6,7; however, withmore prolonged periods of cyclosporine administration,impaired vasorelaxation in response to sodium nitro-prusside has been reported.23 While the predominanteffect of nitroglycerin is on coronary conductance arter-ies, intracoronary adenosine mainly dilates coronaryresistance arteries.14 In the present study, the ability ofthe microcirculation to maximally vasodilate was alsopreserved after cyclosporine administration, as shownby a preserved response to a dose of intracoronaryadenosine in excess of that shown to cause maximalhyperemia.24 This finding lends further support to anendothelium-specific toxicity of cyclosporine.

Endothelin has been proposed as a potential media-tor of cyclosporine-induced vascular toxicity.9"11 In en-dothelial cells in culture,10 in rat glomeruli,1" in the dogrenal circulation,9 and in an anecdotal report in a renaltransplant patient,25 cyclosporine has been reported tocause release of endothelin. In our study, however,cyclosporine did not result in any elevation of peripheralendothelin levels. Further, there was no increase incoronary sinus concentrations of endothelin, suggestingno net release of endothelin across the heart in ourmodel. Our data do not, however, rule out potentialelevation of endothelin levels with more prolongedadministration of cyclosporine. The decrease in baselineCSA after administration of cyclosporine would alsoappear to be mediated by mechanisms other than

lease of endothelium-derived NO (see above) couldaccount for this epicardial vasoconstriction. In humanarteries, endothelium-derived relaxing factor (EDRF)inhibits endothelin-1-induced contractions, while endo-thelin-1 specifically attenuates the vasodilator effect ofEDRF in veins,26 suggesting that vascular tone is main-tained as a result of balance between these factors.Therefore, it is conceivable that in the presence of adeficiency of EDRF release after cyclosporine, endothe-lin may exert a greater vasoconstrictor effect evenwithout an increase in its secretion.

Cyclosporine is markedly hydrophobic27 and is usu-ally clinically administered in an oil-based vehicle, Cre-mophor. Cremophor has been shown to have markedhemodynamic effects; in particular, it can decreaserenal blood flow.28 In the present study, however, apartfrom a small decrease in baseline CSA that did notreach statistical significance, Cremophor administrationdid not influence endothelium-dependent vasorelax-ation in response to acetylcholine in either coronaryconductance or resistance arteries. Further, administra-tion of pure cyclosporine without vehicle yielded asimilar attenuation of acetylcholine-induced coronaryvasorelaxation as cyclosporine in Cremophor, suggest-ing that this impaired endothelium-dependent vasore-laxation was a direct result of cyclosporine administra-tion and not its vehicle.

Limitations of This StudyWe have made the assumption that in the normal

canine coronary arteries studied, changes in vascularreactivity observed by intravascular ultrasound at onecross-sectional location are representative of the wholeartery. This is an inherent limitation of the currenttechnology; in the future, three-dimensional reconstruc-tions from a series of cross-sectional "slices"29 mayallow inferences to be made on vasomotion in longervascular segments.The present study investigated the effects of acute

cyclosporine administration on coronary endothelialfunction in a canine model. The extent to which thesefindings relate to the clinical situation of chronic expo-sure to cyclosporine after cardiac transplantation isunclear. Hypertension, hypercholesterolemia, and im-mune-mediated vascular disease probably contribute tocoronary endothelial damage in cardiac transplant re-

cipients.1-5 We would suggest that the potential role ofchronic cyclosporine therapy in the development ofendothelial dysfunction and subsequent posttransplantvasculopathy requires further consideration.

AcknowledgmentsDr Sudhir was funded as a C.J. Martin Overseas Fellow by

the National Health and Medical Research Council of Aus-tralia and as a Post-Doctoral Fellow by the American HeartAssociation, California Affiliate.

References1. Fish RD, Nabel EG, Selwyn AP, Ludmer PL, Mudge GH, Kirsh-

enbaum JM, Schoen FJ, Alexander RW, Ganz P. Responses ofcoronary arteries of cardiac transplant patients to acetylcholine.J Clin Invest. 1988;81:21-31.

2. Mugge A, Heublein B, Kuhn M, Nolte C, Haverich A, Warnecke J,Forssmann W-G, Lichten PR. Impaired coronary dilator responsesto substance P and impaired flow-dependent dilator responses inheart transplant patients with graft vasculopathy. J Am Coll

increased endothelin release. A decrease in basal re- Cardiol. 1993;21:163-170.

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

Sudhir et al Cyclosporine and Coronary Endothelial Function 3023

3. Johnson MR. Transplant coronary disease: nonimmunologic riskfactors. J Heart Lung Transplant. 1992;11:S124-S132.

4. Eich D, Thompson JA, Ko DJ, Hastillo A, Lower R, Katz S, KatzM, Hess ML. Hypercholesterolemia in long-term survivors of hearttransplantation: an early marker of accelerated coronary arterydisease. J Heart Lung Transplant. 1991;10:45-49.

5. Radovancevic B, Poindexter S, Birovljev S, Velebit V, McAllisterHA, Duncan JM, Vega D, Lonquist J, Burnett CM, Frazier OH.Risk factors for development of accelerated coronary arterydisease in cardiac transplant recipients. Eur J Cardiothorac Surg.1990;4:309-313.

6. Diederich D, Yang Z, Luscher TF. Chronic cyclosporine therapyimpairs endothelium-dependent relaxation in the renal artery ofthe rat. JAm Soc Nephrol. 1992;2:1291-1297.

7. Bossaller C, Forstermann U, Hertel R, Olbricht C, Reschke V,Fleck E. Ciclosporin A inhibits endothelium-dependent vasodi-lation and vascular prostacyclin production. EurJPharmacol. 1989;165:165-169.

8. Richards NT, Poston L, Hilton PJ. Cyclosporine A inhibitsrelaxation but does not induce vasoconstriction in human subcu-taneous resistance vessels. J Hypertens. 1989;7:1-3.

9. Carrier M, Tronc F, Stewart D, Pelletier LC. Dose-dependenteffect of cyclosporine on renal arterial resistance in dogs. Am JPhysiol. 1991;261:H1791-H1796.

10. Bunchman TE, Brookshire CA. Cyclosporine stimulated synthesisof endothelin by human endothelial cells in tissue culture. KidneyInt. 1990;37:365A. Abstract.

11. Kon V, Sugiura M, Inagami T, Harvie BR, Ichikawa I, Hoover RL.Role of endothelin in cyclosporine-induced glomerular dys-function. Kidney Int. 1990;37:1487-1491.

12. Lau DCW, Wong K-L, Hwang WS. Cyclosporine toxicity oncultured rat microvascular endothelial cells. Kidney Int. 1989;35:604-613.

13. Zoja C, Furci L, Ghilardi F, Zilio P, Benigni A, Remuzzi G.Cyclosporine-induced endothelial cell injury. Lab Invest. 1986;55:455-462.

14. Sudhir K, MacGregor JS, Barbant SD, Foster E, Fitzgerald PJ,Chatterjee K, Yock PG. Assessment of coronary conductance andresistance vessel reactivity in response to nitroglycerin, ergonovineand adenosine: in vivo studies with simultaneous intravasculartwo-dimensional and Doppler ultrasound. JAm Coll Cardiol. 1993;21:1261-1268.

15. Sudhir K, Fitzgerald PJ, MacGregor JS, DeMarco T, Ports TA,Chatterjee K, Yock PG. Transvenous coronary ultrasoundimaging: a novel approach to visualization of the coronary arteries.Circulation. 1991;84:1957-1961.

16. Klocke FJ, Ellis AK. Physiology of the coronary circulation. In:Chatterjee K, Karliner J, Rapaport E, Cheitlin MD, Parmley WW,

Scheinman M, eds. Cardiology. Philadelphia, Pa: JB Lippincott;1991:1.101-1.114.

17. Sudhir K, MacGregor JS, Gupta M, Barbant SD, Redberg R, YockPG, Chatterjee K. Effect of selective angiotensin II receptorantagonism and angiotensin converting enzyme inhibition on thecoronary vasculature in vivo: intravascular two-dimensional andDoppler ultrasound studies. Circulation. 1993;87:931-938.

18. Sudhir K, MacGregor JS, Barbant S, Gupta M, Yock PG, Chat-terjee K. Differential contribution of nitric oxide to the main-tenance of vascular tone in coronary conductance and resistancearteries. Am Heart J. 1994;127:858-865.

19. Chou TC, Sudhir K, Iwanaga S, Chatterjee K, Yock PG. Mea-surement of volumetric coronary blood flow using simultaneousintravascular two-dimensional and Doppler ultrasound: validationin an animal model. Am Heart J. 1994;128:237-243.

20. Chu A, Chambers DE, Lin C, Kuehl WD, Cobb FR. Nitric oxidemodulates epicardial coronary basal vasomotor tone in awakedogs. Am J Physiol. 1990;258:H1250-H1254.

21. Lefroy DC, Crake T, Uren NG, Davies GJ, Maseri A. Effect ofinhibition of nitric oxide synthesis on epicardial coronary arterycaliber and coronary blood flow in humans. Circulation. 1993;88:43-54.

22. Chung SJ, Fung HL. Identification of the subcellular site for nitro-glycerin metabolism to nitric oxide in bovine coronary smoothmuscle cells. J Pharmacol Exp Ther. 1990;253:614-619.

23. Huang HC, Rego A, Vargas R, Foegh ML, Ramwell PW.Nitroprusside-induced vascular relaxation is attenuated in organ-transplanted animals treated with cyclosporine. Transplant Proc.1987;19(suppl 5):126-130.

24. Laxson DD, Dai X-Z, Homans DC, Bache RJ. Coronary vaso-dilator reserve in ischemic myocardium of the exercising dog.Circulation. 1992;85:313-322.

25. Fogo A, Hakim RC, Sugiura M, Inagami T, Kon V. Severe endo-thelial injury in a renal transplant patient receiving cyclosporine.Transplantation. 1990;49:1190-1192.

26. Luscher TF, Yang Z, Tschudi M, von Segesser L, Stulz P, Bou-langer C, Siebenmann R, Turina M, Buhler FR. Interactionbetween endothelin-1 and endothelium-derived relaxing factor inhuman arteries and veins. Circ Res. 1990;66:1088-1094.

27. Kahan B. Individualization of cyclosporine therapy using pharma-cokinetic and pharmacodynamic parameters. Transplantation.1985;40:457-476.

28. Besarab A, Jarrell BE, Hirsch S, Carabasi RA, Cressman MD,Green P. Use of the isolated perfused kidney model to assess theacute pharmacologic effects of cyclosporine and its vehicle, Cre-mophor EL. Transplantation. 1987;44:195-201.

29. Kitney RI, Moura L, Straughan K. 3-D visualization of arterialstructures using ultrasound and Voxel modelling. Int J CardImaging. 1989;4:135-143.

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

ChatterjeeK Sudhir, J S MacGregor, T DeMarco, C J De Groot, R N Taylor, T M Chou, P G Yock and K

resistance coronary arteries.Cyclosporine impairs release of endothelium-derived relaxing factors in epicardial and

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1994 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.90.6.3018

1994;90:3018-3023Circulation. 

http://circ.ahajournals.org/content/90/6/3018World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on February 14, 2018http://circ.ahajournals.org/

Dow

nloaded from