JACC Vol. 9, No.5May 1987:1091-8

1091

Enhanced Myocardial Washout and Retrograde Blood Delivery WithSynchronized Retroperfusion During Acute Myocardial Ischemia

BING-LO CHANG, MD, J, KEVIN DRURY, MD, FACC, SAMUEL MEERBAUM, PHD, FACC,

MICHAEL C. FISHBEIN, MD, FACC, JAMES S. WHITING, PHD, ELIOT CORDAY, MD, FACC

Los Angeles, California

The effects of synchronized coronary venous retroper­fusion of arterial blood on myocardial washout werestudied with digital subtraction angiography in 10closedchest dogs during balloon occlusion of the proximal leftanterior descending coronary artery. The center lumenof the intracoronary balloon catheter was used for se­quential injections of 1 ml (meglumine diatrizoate) Ren­ografin-76, and contrast washout rate was determinedby videodensitometry in myocardial regions subservedby the left anterior descending coronary artery. Beforecoronary artery occlusion, washout rate was 22.4 ± 2.7mm" (mean ± SEM). Five minutes after occlusion, andimmediately before synchronized retroperfusion, wash­out rate dropped sharply to 2.0 ± 0.7 min -I. Twenty­five minutes after occlusion, with 50 milmin synchro­nized retroperfusion treatment applied for 5 minutes,washout rate was 5.0 ± 1.5 min -I. Thus, synchronizedretroperfusion significantly (p < 0.05) accelerated con­trast disappearance over that during presynchronizedretroperfusion ischemia.

To determine the effects of synchronized retroper-

Experimental synchronized coronary venous retroperfusionof arterial blood during acute myocardial ischemia has beenreported (1-5) to enhance regional myocardial perfusion,improve cardiac function and reduce infarct size. However,the mechanisms by which this retrograde treatment benefits

From the Division of Cardiology, Department of Medicine and theDepartment of Pathology, Cedars-Sinai Medical Center, University of Cal­ifornia, Los Angeles, School of Medicine, Los Angeles, California. Thisstudy was supported in part by Grants HL J7561-11 and HL 14644-11from the National Heart, Lung, and Blood Institute, National Institutes ofHealth, Bethesda, Maryland and the Medallion Group of Cedars-SinaiMedical Center, the W.M. Keck Foundation, the Ahmanson Foundation,Mr. and Mrs. Harry Roman, Mr. J.C. Dunas, Mr. and Mrs. Hal Wallis.Mrs. Dorothy Forman, Mr. and Mrs. Nicolai Joffe, Mr. and Mrs. NormanTyre and the Emanuel Borinstein Family Foundation, Los Angeles, Cal­ifornia.

Manuscript received February 18, 1986; revised manuscript receivedAugust 27, 1986, accepted October 22, 1986.

Address for reprints: Bing-Io Chang, MD, Cedars-Sinai Medical Cen­ter, Halper Building #325, 8700 Beverly Boulevard, Los Angeles, Cal­ifornia 90048.

© 1987 by the American College of Cardiology

fusion on retrograde delivery to the ischemic myocar­dium, monastral blue dye was retroinfused through thesystem into the great cardiac vein before the dog waskilled. Transverse heart slices were then studied by lightmicroscopy, and regional intravascular dye content wasscored from 0 to 3 (0 = no dye, 3 = maximal dye).After great cardiac vein synchronized retroperfusion,blue dye content in capillaries of ischemic anterior andnonischemic posterior aspects of the left ventricle was2.3 ± 0.5 versus 0.7 ± 0.3, respectively (p < 0.05).Conversely, anterograde left ventricular dye injectionperformed in four control dogs with equivalent left an­terior descending coronary artery occlusion but withoutsynchronized retroperfusion resulted in greater blue dyedelivery to the nonjeopardized posterior myocardial re­gions. Thus, synchronized retroperfusion of the greatcardiac vein provides selective delivery of substrate tothe acutely ischemic myocardium and appears to sig­nificantly enhance washout from these regions.

(J Am Coli CardioI1987;9:1091-8)

the jeopardized myocardium are still inadequately under­stood. Experimental studies employing radioactive micro­spheres (6-9) have demonstrated various improvements ofacutely ischemic myocardial perfusion, ranging from 10 to40% of normal levels. However, this magnitude of myo­cardial perfusion enhancement does not appear to fully ex­plain the observed dramatic improvement in regional myo­cardial function during synchronized retroperfusion and thereported infarct size reduction (1-3,5). This apparent dis­crepancy could be due to inadequacies in measurement tech­niques, particularly the coronary venous administration ofradionuclide microspheres for evaluation of retrograde en­hancement of myocardial perfusion. Another possibility isthat synchronized retroperfusion effectiveness is due notonly to selective arterial blood delivery to jeopardized isch­emic myocardium, but also to an enhanced washout of met­abolic byproducts from the acutely ischemic zone, whichimproves cardiac muscle function and myocardial viability.

0735-1097/87/$3.50

1092 CHANG ET AL.SYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCARDIUM

lACC Vol. 9. No.5May 1987:1091-8

Mohl et al. (10) demonstrated an enhancement of myo­cardial washout during pressure-controlled intermittentcoronarysinusocclusion, a technique that producespressurechanges within the coronary venous system without arterialblood delivery. Accordingly, we applied coronary digitalsubtraction venography and angiography to assess both ret­rograde contrastdistributionand washoutfrom acutely isch­emic myocardial regions, withand withoutthe synchronizedretroperfusion intervention. Microscopic studies of myo­cardial tissue were also performed to evaluate the retroper­fusion-induced retrograde penetration of a vascular markerinto the myocardial microcirculation.

MethodsAnimal preparation. Ten closed chest dogs weighing

23 to 37 kg were premedicated with morphine sulfate (1.5mg/kg body weight) and anesthetized with intravenous so­dium thiopental (30 mg/kg). After endotracheal intubation,ventilation was maintained with a respirator (Harvard Ap­paratus)and ethrane wasadministered and titratedto providea steady state of light anesthesia. Heparin (10,000 IV) wasgiven intravenously before instrumentation. To prevent ar­rhythmias, lidocaine (40 mg) was administered intrave­nouslyby bolus injectionfollowedby a continuous injectionof 2 mg/min. Arterial blood pressure and an electrocardio­gram (ECG) were monitored continuously, using an Elec­tronics for Medicine physiologic recorder. An autoinflatableballoon retroperfusion catheter was inserted under fluoro­scopic control through the left internal jugular vein into thegreat cardiac vein. A 4F double lumen balloon-tipped cath­eter was inserted through the carotid artery and positionedimmediately proximal to the firstdiagonal branchof the leftanterior descending coronary artery.

Synchronized retroperfusion system. A synchronizedretroperfusion system (Fig. I) was used to deliver arterialblood into the great cardiac vein, as previously described(1-3). Briefly, arterial blood was shunted from the brachialartery into the synchronized retroperfusion pump, whichwas triggered by the ECG. The blood was delivered ret­rogradelyduring diastole throughthe retroperfusion catheterinto the great cardiac vein and toward the myocardial zonesubservedby the occluded left anteriordescending coronaryartery. Diastolic autoinflation of the retroperfusion catheterballoon produces a brief and phased obstruction of the greatcardiac vein, which helps propel the arterial blood into theanterior interventricular veins and their smaller branches.During systole, retroperfusion flow is interrupted by thesynchronized pump, causing collapse of the catheter bal­loon, which facilitates drainage of coronary venous bloodthrough the coronary sinus into the right atrium. The pump­ing system was regulated to operate at several levels ofretroperfusion flow from 25 to 100 mllmin.

AORTIC ARCH

Figure 1. Schematic of the synchronized retroperfusion (SRP)system. Arterial blood is shunted from a brachial artery to thepump chamber. The upward movement of a piston into thepumpchamber during diastole propels the arterial blood into the greatcardiac vein. Adouble lumen balloon-tipped catheter is positionedproximal to the left anterior descending coronary artery (LADART) for coronary occlusion and contrast agent injections. AI =anterior interventricular; EM = electromagnetic.

Image processing system. Fluoroscopy was performedin a left lateral projection at a fixed X-ray tube potentialand current, typically 75 kV and 5 to 10 mAo Images wererecorded on videotape beginning 5 seconds before contrastinjection to establish the baseline video intensity, and for30 secondsafter the injectionto observe washoutof contrastmedium from the myocardium. The videotape images werethen processedby a GouldlDeAnza IP5500 image processorby wayof a videotime basecorrector (HarrisVideo System,model 516), to acquire the time-intensity curves from spe­cificcontrast-opacified myocardial regionsof interest. Myo­cardial washout rate (k, min-I) was derived from a mon­oexponential fit to the initial descending portion of the time­intensity curve immediately after peak intensity.

Experimental procedure. Before insertion of the leftanterior descending coronary artery catheter, a digital sub­traction venogram was obtained by injecting 2.5 ml of (me­glumine diatrizoate) Renografin-76 into the great cardiacvein through the retroperfusion catheter, to determine thespecific myocardial venousanatomyin each dog. After ven­ography, I ml of Renografin-76 was injected into the leftanterior descending coronary artery, and myocardial per­fusion images were obtained. The artery was then occludedby inflating the intracoronary balloon, and after 5 minutesof ischemia, I ml of Renografin-76 was injected throughthe central lumen of the catheter to opacify the myocardiumdistal to the occlusion. The time course of contrast mediumdisappearance was determined by obtaining X-ray imagesfor 30 seconds after the injection. Synchronized retroper­fusionwasstarted 10minutesafterocclusionand myocardialperfusion images during treatment were obtained by in-

lACC Vol. 9. No. 5May 1987:109 1- 8

CHANG ET AL.SYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCA RDIUM

1093

jecting I ml of Renografin-76 through the central lumen ofthe balloon catheter duringfourdifferent retroperfusion flowrates (25, 50, 75 and 100mil min) administrated in randomsequence. Each retroperfusion flow was maintained for 5minutes , with a 5 minute untreated interval between suc­cessive flow levels. After completing the synchronized re­troperfusion study, images were againobtainedduring coro­nary occlusion, but without retroperfusion treatment. Beforethe dog was killed, 10% monastral blue dye (I mllkg) wasretrogradely infused through the synchronized retroperfu­sion pump into the great cardiac vein to determine the ef­fectiveness of blood delivery to the myocardial microcir­culation. Arterial bloodpressure and an ECG wererecordedin the control state, after occlusion and beforesynchronizedretroperfusion and after occlusion during synchronizedretroperfusion.

Postmortem study. The dogs were killed after the mon­astral blue dye retroinfusion. All hearts were sliced from

apex to base into four transverse slabs that were photo­graphed and then fi xed in formalin. Using light microscopy,the distribution of blue dye was determined within capil­laries, venules and arterioles in tissue samples from both

Figure 2. A, Injection of Renografin-76 into the left anterior de­scending coronary artery before occlusion resulted in completeopacification of the anterior left ventricular myocardium (2), fol­lowed by a rapid return to preinjection intensity levels within 4seconds (4). The washout rate (right) was 16 min- I. B, Myo­cardial opacification was obtained after injection of I ml Reno­grafin-76 into the left anterior descending artery distal to the bal­loonocclusion before synchronizedretroperfusion. Contrast injectionresulted in myocardial opacification of the ischemic zone, withpersistence of contrast 10 seconds after injection. The washoutrate was 1.2 min I . C, Application of synchronizedretroperfusion(50 mllmin) demonstrates a significant increase in ischemic zonewashout rate at 10seconds despite maintained coronary occlusion.The washout rate (k) was 3.8 min- I.

100% r---"c-------,

2 3 4 5 6 7 8 9 10TIME (sec)

100%r---=------,4

>­f­e;;Zwo

~~wa:

A

./

B

~(/)zwow>~...JWa:

o 1 23 <: 5 6 7 89 101112131415TIME (sed

cPRE

INJECTION5 10

POST INJECTION (sec)

..

~(/)

sow>~oJwa:

2 3 4 5 6 7 8 9 101112TIME (sed

1094 CHANG ET AL.SYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCARDIUM

lACC Vol. 9, No.5May 1987:1091-8

o

the ischemic anterior and nonischemic posterior aspects ofthe left ventricle. This assessment was carried out by twoobservers employing a semiquantitative scoring system (fromo = no dye to 3 = maximal dye content).

Statistical analysis. Results of contrast washout rate (k)measured in the same dog under the six conditions describedwere evaluated by a nonparametric repeated measures anal­ysis of variance (Friedman's analysis of variance) and amultiple comparison procedure (Fisher's least significantdifference test). All data are expressed as mean ± SEM.

30

.....c::'E~ 20~

•

DURING OCCLUSION

Figure 3. Washout rate (k) plottedas a Tukeyboxgraphfor sevenmeasurements in six dogs. The five horizontal lines on each boxgraph portray five percentiles with p values, from bottom to top,of 10, 25, 50, 75 and 90%. All values in the data set above the90th percentile and below the 10th percentile are graphed indi­vidually. The solid squares represent the mean value of washoutrate for each measurement. Contrast washout rates during syn­chronized retroperfusion at 50, 75 and 100 ml/min were signifi­cantly greater (p > 0.05) than the washout rate during occlusionwithout synchronized retroperfusion (pre SRP), but significantlyless thanthe washout rateswithout coronary occlusion (PreOccl.).

trast medium from the ischemic zone (5.0 ± 1.5, 5.2 ±1.5 and 6.5 ± 2.1 min - I, respectively) compared with thenonretroperfused ischemic states (p < 0.05). However, thecontrast washout rate during synchronized retroperfusion at25 ml/min (2.7 ± 1.1 min - I) was not significantly differentfrom the nonretroperfusion washout rate.

Great cardiac vein angiographic findings during coro­nary occlusion (Fig. 4). A high intensity opacification wasnoted as early as the third diastole after retrograde infusionof Renografin-76, demonstrating that the ischemic myo­cardial region distal to the left anterior descending coronaryartery occlusion was promptly retrogradely perfused by thesynchronized retroperfusion. Contrast medium was seen topenetrate into the small coronary veins, and a "myocardialblush" was visualized in the digital subtraction venogramof 7 of the 10 dogs. The contrast intensity gradually de­creased after completion of the Renografin injection. In onecase (Fig. 5) there was direct shunting of the contrast agentfrom the anterior interventricular vein into the right atrium.A right atrial coronary venous orifice separate from thecoronary sinus was observed postmortem after the injectionof microfil into the coronary sinus.

0

0 18QPre Pre 25 50 75 100 Post

Occl, SRP SRPI I

10

ResultsCoronary angiographic findings. Seven digital sub­

traction angiograms were performed in each of 10 dogs, fora total of 70 injections. Five of these 70 injections couldnot be subjected to washout analysis because of respiration­induced baseline variations, and 3 injections could not beanalyzed because of technical difficulties with the retroper­fusion pump. Complete data consisting of seven successiveanalyzable angiograms per dog were obtained in 6 of the10 dogs.

Compared with the untreated postocclusion measure­ments, synchronized retroperfusion did not cause any sig­nificant changes in heart rate or mean aortic pressure (Fig.2). Injection of Renografin-76 into the left anterior descend­ing coronary artery before occlusion (Fig. 2A) resulted inpronounced opacification of the anterior left ventricularmyocardium, followed by a rapid return to preinjection myo­cardial intensity levels within 4 seconds. The contrast wash­out rate (k) in this dog was 16.0 min - I.

After 1 ml of Renografin-76 was injected into the leftanterior descending coronary artery during its balloon oc­clusion (Fig. 2B), myocardial opacification of the anteriorwall was again noted but persistence of contrast mediumwithin the ischemic zone exceeded 20 seconds after theinjection. Contrast washout rate during balloon occlusionwas 1.2 min - 1 from the ischemic myocardial region ofinterest. During coronary occlusion treated with synchro­nized retroperfusion at 50 mllmin (Fig. 2C), a substantialenhancement of contrast washout from the ischemic zonewas noted at the 10 second postinjection measurement. By20 seconds the myocardial intensity had returned to prein­jection levels. The derived washout rate (k) in this case was3.8 min-I.

Myocardial washout rates (Fig. 3). During coronaryocclusion before synchronized retroperfusion, the washoutrate of myocardial contrast medium from the acutely isch­emic zone (2.0 ± 0.7 min -I) was significantly slower com­pared with that in the preocclusion state (22.4 ± 2.7 min -I)

(p < 0.001). The washout rates during occlusion before andafter the retroperfusion treatment were similar (2.0 ± 0.7versus 1.4 ± 0.2 min - I, P = NS). Synchronized retro­perfusion at flow rates of 50, 75 or 100 mllmin resulted insignificantly accelerated disappearance of myocardial con-

JACC Vol. 9. No.5May 1987:1091-8

CHANGET ALSYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCARDIUM

1095

SRP DIAS-2SRP DIAS-1PRE SRP

Figure 4. Six end-diastolic (OlAS) car­diac imagesobtainedby synchronized ret­rograde infusionof 2.5 ml of Renografin­76 through the synchronized retroperfu­sion (SRP) catheter into the great cardiacvein during left anteriordescendingcoro­nary artery occlusion. Myocardial opaci­fication notedas early as the third diastole(SRP OlAS 3) suggests an early cumu­lative effect of the retroinfusate with ret­roperfusion. The contrast intensity grad­ually decreased after completion of theRenografin injection.

SRP DIAS-3 SRP DIAS-4 SRP DIAS-8

Microscopic findings (Fig. 6). In five of six dogs, in­jection of monastral blue dye into the coronary veins duringsynchronized retroperfusion at a flow rate of 50 ml/minresulted in dye delivery to capillaries, venules and arteriolesof the ischemic myocardium (Fig. 6a and b). The content

Figure 5. An example in one dog of contrast agent shuntingafterretrograde injection into the great cardiac vein through a directanterior interventricular vein (AIV) to right atrium (RA) anasto­mosis (arrow). An extracoronary venousorificein the right atriumwas found postmortem.

of blue dye in the ischemic anterior wall was significantlymore pronounced than that in the nonischemic posterior wall(score 2.3 ± 0.5 versus 0.7 ± 0.3, P < 0.05) (Table I).In two dogs no dye was noted in the nonischemic posteriorwall microcirculation (Fig. 6c), and in one there was nodye in either the anterior or the posterior wall of the leftventricle. In the latter, necropsy revealed a shunt from theanterior interventricular vein to the right atrium, which wasdemonstrated premortem with digital venography (Fig. 5).In four other dogs, monastral blue dye was injected into theleft ventricle to determine the distribution of anterogradeperfusion in the presence of a left anterior descending coro­nary artery occlusiun. Anterograde infusiun resulted in sig­nificantly greater dye concentration in the nonjeopardizedposterior region than in the jeopardized anterior myocardialregions (2.5 ± 0.3 versus 1.3 ± 0.3, p < 0.05) (TableI). Table 2 provides a comparison of washout rate andmicroscopic findings in the six dogs in which both mea­surements were obtained. The dogs that exhibited a sharpincrease in washout rate as a result of synchronized retro­perfusion also showed the most significant delivery of dyeto the ischemic myocardium, presumably reflecting syn­chronized retroperfusion enhancement of the risk zone cir­culation.

DiscussionDigital subtraction angiography. The rate of myo­

cardial disappearance of various diagnostic agents (such asradioactive sodium and xenon) has been shown to be pro­portional to the degree of myocardial perfusion, and cantherefore be used to characterize ischemia in experimental

1096 CHANG ET AL.SYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCARDIUM

lACC Vol. 9. No.5May 1987:1091-8

Figure 6. a, Ischemic anterior wall of left ventricle showing nu­merous capillaries and an artery (arrow) containing blue dye. b,Higher magnification showing dye in venules as well (arrow). c,Nonischemic posterior wall with no dye in any vessel. Hematox­ylin-eosin stain x 40 (a and c) and x 69(b), all reduced by 20%.

and clinical settings (11-13). Recently, advances in digitalsubtraction radiography have enabled the visualization ofvascularstructures containing relatively low concentrationsof contrast medium (14). Enhancement of selective arterio­grams provides an excellent method for imaging both thecoronary anatomy and the regional myocardial flow distri-

bution, and permits quantification of myocardial perfusionand washout (14,15) . Computerized analysis of digital an­giographic myocardial contrast appearance and disappear­ance rates has been used to evaluate myocardial blood flow ,and a linearcorrelation has been found between myocardialcontrast washout rate and coronary artery blood flow asmeasured by an electromagnetic flowmeter (15). Thus, itappears that contrast dilution measurements obtained withdigital coronary angiography can characterize the regionaldistribution of myocardial perfusion and provide on-linemonitoring of changes in perfusion during therapeutic in­terventions.

Is the myocardium being perfused by synchronizedretroperfusion? It has been reported (1,2,3,6,16) that syn­chronized coronary venous retroperfusion provides diastol­icallyaugmented retrograde perfusion of an ischemicregionand facilitates coronary venous drainage in systole. Ourstudy characterized the coronary venous circulation duringsynchronized retroperfusion using digital subtraction ven­ograms. Thus, Figure 4 illustrates the diastolic delivery ofarterial blood during synchronized retroperfusion into theregional coronary veins that subservedthe acutely ischemiczone. Despite the systoliccollapse of the synchronized ret­roperfusion balloon, the contrast agent retrogradely pro­pelled in each diastole is seen to penetrate the ischemicmyocardial microvasculature within as few as three cardiaccycles from the start of synchronized retroperfusion. Dia­stolic delivery was further corroborated in the same dogsby microscopic evidence of dye delivery to the microcir­culation (Fig. 6, Table 2). It has been previously reported(17) that great cardiac vein dye bolus injection results in apreferential distribution in left ventricular regions suppliedby the occluded coronary artery. The observed myocardialdistribution of the retrogradely infused monastral blue dyein this studyclearlydemonstrates the abilityof synchronizedretroperfusion to deliver substrate to the critically under­perfused tissue.

Does washout play an important role in synchronizedretroperfusion? During acute myocardial ischemia, sub­cellular ischemic derangements and loss of fluid controlresult in myocardial edema and leakage into the interstitiumof intracellular contents through dysfunctioning cell mem­branes (18). In the absence of adequate ischemic zone cir­culation, toxic metabolites tend to accumulate, as reflected

Table 1. Scoring of Regional Blue Dye Delivery in the Mid-Left Ventricular Transverse Slab

A

Retrograde (n = 6)

p A

Anterograde (n = 4)

p

Mid-LVslab

3.1 ± 0.6* 1.7 ± 0.6 2.0 ± 0.4* 3.5 ± 0.5

*p < 0.05 versus posterior wall values are mean ± SEM. A = anterior wall; n = no. of dogs; P =

posterior wall .

JACC Vol. 9, No.5May 1987:1091-8

CHANGET AL.SYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCARDIUM

Table 2. Comparison of Washout Rate Versus Blue Dye Content in Six Dogs

Washout Rate (min-I) During Content of Dye in IschemicCoronary Occlusion Myocardium

Dog No. Pre-SRP During SRP A C V

I 1.4 3.0 2 3 3

2 1.2 8.0 I 3 3

3 2.0 3.0 I 2 2

4 1.4 5.5 2 3 3

5 1.4 1.4 0 0 0

6 1.4 4.4 2 3 3

A = arterioles;C = capillaries; SRP = synchronized retroperfusion; V = venules.

1097

by a reduction of myocardial tissue pH, and this may pre­cipitate a vicious metabolic cycle leading to further dete­rioration, expansion and acceleration of the ischemic injury(19). Mohl et al. (10) demonstrated that myocardial fluidwashout is induced by intermittent coronary sinus occlusion,a technique employing automatically triggered intermittentballoon inflation and deflation without arterial blood deliv­ery. Our experiments using digital subtraction angiographyindicated that synchronized retroperfusion results in a sig­nificant acceleration of myocardial contrast disappearance,suggesting enhanced myocardial fluid washout from theischemic region.

How does coronary veno-venous shunting affect syn­chronized retroperfusion? Direct connections between thecoronary venous system and cardiac chamber were de­scribed by Thebesius (20) in 1716. Maurer et al. (21) usingretrograde echo contrast injections in vivo, and indocyaninegreen dye injections postmortem, demonstrated direct drain­age into the right and to a lesser degree the left side of theheart. Palkaska and Kolff (22) using postmortem radio­opaque microfil injections found a larger number of veno­venous anastomoses between the anterior cardiac veins andthe great cardiac vein in human hearts compared with thosein dogs. OUf study, using digital subtraction coronary ven­ography, demonstrated the presence of significant apicalcoronary veno-venous anastomoses (2 of 10 dogs) and sep­arate right atrial venous connections (I dog). In the presenceof coronary artery occlusion, retrograde injection into cor­responding coronary veins resulted in dye delivery to boththe ischemic myocardial microvasculature and the posteriorcoronary veins through apical anastomoses. With impendingapplication of the newer retrograde interventions employingthe coronary venous system as an alternative pathway toreach jeopardized myocardium (4), the anatomic variabilityand flow resistance of such coronary venous connectionsmight assume importance, because an unknown fraction ofthe retroperfusate could be shunted directly into the cardiacchambers or nonjeopardized zones rather than into jeopar­dized myocardium. A special case in point was Dog 5, inwhich direct venous shunting to the right atrium was re­flected in the absence of blue dye content in the ischemic

myocardium and in failure of enhancement of measuredmyocardial contrast washout (Table 2). In contrast, Dog 6exhibited an abundant retrograde blue dye delivery into theischemic microcirculatory bed, along with significant in­crease of the myocardial washout rate. It is anticipated thatcomputer-aided digital coronary venography will allow de­lineation of individual anatomic pathways and will permitassessment of anticipated effectiveness of coronary venousinterventions.

Limitation of the study design. The videodensitometricanalysis of digital subtraction angiographic images providesregional myocardial washout appearance-disappearance curvesand washout rate measurements that have previously beenshown to correlate with radionuclide microsphere perfusiondata as well as electromagnetic coronary artery flow mea­surements. However, these validations and correlations werecarried out without the retrograde coronary venous inter­ventions used in our study. To evaluate and corroborate theinfluence of these retrograde manipulations on myocardialwashout, it would be desirable to also employ alternativetechniques such as thallium-20l scintigraphy. However, thewashout of thallium after intracoronary injection occurs witha half-life >60 minutes, which makes multiple assessmentsdifficult.

The microscopic evidence of blue dye in ischemic myo­cardial capillaries, venules and arterioles during synchro­nized retroperfusion is convincingly illustrative; however,it is far from being quantitative. Similarly, although thepresence of a myocardial blush in the digital venographicimages is well discerned, there is no quantitative analysisof the degree and extent of retrograde coronary venous pen­etration into the ischemic myocardium. For quantitativeevaluation of myocardial perfusion, appropriately designedradionuclide microsphere protocols should be employed.However, most of the validation studies employing radio­nuclide microspheres have been performed in the setting ofanterograde blood delivery. Available reports (6-8) revealdiscrepancies and variability of the estimated retrograde de­livery with coronary sinus interventions, presumably be­cause of differences in experimental design as well as mea­surement procedures. Moreover, the dynamics of retrograde

1098 CHANG ET AL.SYNCHRONIZED RETROPERFUSION IN ISCHEMIC MYOCARDIUM

JACCVol. 9, No.5May 1987:1091-8

microsphere delivery is not fully understood, particularlywhen the retrograde intervention is competing with or al­ternating with a substantial residual anterograde perfusion,which may potentially interfere with capillary trapping ofthe microspheres. Therefore, radiographic contrast digitalsubtraction venography and monastral blue dye injection,not involving capillary "trapping," were employed in ourstudy.

Clinical implications. Coronary venous retroperfusionhas been demonstrated in the animal laboratory to be aneffective treatment of acute myocardial ischemia. Clinicaltrials of synchronized retroperfusion have been initiated andpreliminary reports from these studies suggest ameliorationof ischemia with retroperfusion treatment during coronaryangioplasty (16) and unstable angina (23). However, evenin these early trials the technique has not been successfulin all patients, and definite evidence of efficacy has not asyet been demonstrated, partly because assessment of myo­cardial perfusion and infarct size reduction is difficult inhumans.

Our study suggests that the efficacy of retroperfusionmight be assessed in humans by analysis of vascular andmyocardial images obtained with digital subtraction venog­raphy and angiography. Thus, demonstration of large ve­nous shunts or insufficient retroperfusion flow rates, or both,might explain the lack of efficacy of the retrograde treatmentin some patients. Conversely, significant enhancement ofwashout measured by digital contrast angiography wouldindicate that synchronized retroperfusion might effectivelyincrease circulation to and from the ischemic myocardium.

We gratefully acknowledge the technical assistance of Willis Curtis Peaand Myles Prevost. We also thank Kenneth 1. Resser, MS for assistancewith the statistics and Jeanne Bloom for editorial assistance.

ReferencesI. Drury lK, Fishbein M, Meerbaum S, Corday E. Synchronized diastolic

coronary venous retroperfusion treatment of acute myocardial is­chemia; results of a preclinical safety and efficacy study. 1 Am ColiCardiol 1985;6:328-35.

2. Farcot lC, Meerbaum S, Lang TW, Kaplan L, Corday E. Synchro­nized retroperfusion of coronary veins: circulatory support of jeop­ardized ischemic myocardium. Am 1 Cardiol 1978;41:1191-120 I.

3. Meerbaum S, Lang TW, Osher IV, et al. Diastolic retroperfusion ofacutely ischemic myocardium. Am 1 Cardiol 1976;37:588-98.

4. Meerbaum S, Haendchen RV, Corday E, et al. Hypothermic coronaryvenous phased retroperfusion: a closed chest treatment of acute re­gional myocardial ischemia. Circulation 1982;65:1435-45.

5. Yamazaki S, Drury lK, Meerbaum S, Corday E. Synchronized coro­nary venous retroperfusion: prompt improvement of left ventricular

function in experimental myocardial ischemia. 1 Am Coli Cardiol1985;5:655-63.

6. Berdeaux A, Farcot lC, Bourdarias IP, Barry M, Bardet 1, GiudicelliIF. Effects of diastolic synchronized retroperfusion on regional coro­nary blood flow in experimental myocardial ischemia. Am 1 Cardiol1981;47:1033-40.

7. Chiu Cl, Mulder DS. Selective arterialization of coronary veins fordiffuse coronary occlusion. An experimental evaluation. 1 ThoracCardiovasc Surg 1975;70:177-82.

8. Hochberg MS, Roberts We, Morrow AG, Austen WG. Selectivearterialization of the coronary venous system. Encouraging long termflow evaluation utilizing radioactive microspheres. 1 Thorac Cardio­vase Surg 1979;77:1-12.

9. Hochberg MS. Hemodynamic evaluation of selective arterializationof the coronary venous system: an experimental study of myocardialperfusion using radioactive microspheres. 1 Thorac Cardiovasc Surg1977;74:774-83.

10. Mohl W, Punzegruber C, Moser M, et al. Effects of pressure-con­trolled intermittent coronary sinus occlusion on regional ischemic myo­cardial function. 1 Am Coli Cardiol 1985;5:939-47.

II. Bassingthwaighte 1. Physiology and theory of tracer washout techniquefor the estimation of myocardial blood flow: flow estimation fromtracer washout. Prog Cardiovasc Dis 1977;20:165-80.

12. Cannon P, Dell R, Dwyer E. Measurement of regional myocardialperfusion in man with 133-xenon and a scintillation camera. 1 ClinInvest 1972;51:964-77.

13. Kety SS. Measurement of regional circulation by the local clearanceof radioactive sodium. Am Heart 1 1949;38:321-7.

14. Vogel R, LeFree M, Bates E, et al. Application of digital techniquesto selective coronary arteriography: use of myocardial contrast ap­pearance time to measure coronary flow reserve. Am Heart 11984;107:153-64.

15. Whiting IS, Drury lK, Pfaff 1M, et al. Digital angiographic mea­surement of radiographic contrast material kinetics for estimation ofmyocardial perfusion. Circulation 1986;73:789-98.

16. Farcot lC, Berland 1, Cribier A, Letac B, Bourdarias IP. Diastolicsynchronized retroperfusion in the coronary sinus during percutaneoustransluminal angioplasty: preliminary experience (abstr). Circulation1985;72 (suppl IlI):III-470.

17. Karagueuzian HS, Ohta M, Drury JK, et a!. Coronary venous retro­perfusion of procainamide: a new approach for the management ofspontaneous and inducible sustained ventricular tachycardia duringmyocardial infarction. J Am CoIl Cardiol 1986;7:551-63.

18. Schaper J. Ultrastructure of the myocardium in acute ischemia. In:Schaper W, ed. The Pathophysiology of Myocardial Perfusion. Am­sterdam, New York, Oxford: Elsevier North-Holland, 1979:571-3.

19. Jennings RB, Reimer KA. Lethal myocardial ischemic injury. Am 1Pathol 1981;102:241-55.

20. Thebesius AC. Dissertatio medica de circulo sanguinis in corde. Lug­duni Batavorum apud 10h Langerak Arnold 1716:31.

21. Maurer G, Punzengruber C, Haendchen RV, et al. Retrograde coro­nary venous contrast echocardiography: assessment of shunting anddelineation of regional myocardium in the normal and ischemic canineheart. J Am Coli Cardiol 1984;4:577-86.

22. Palkaska E, Kolff WI. Anatomical basis for retrograde coronary veinperfusion: venous anatomy and veno-venous anastomoses in the heartsof humans and some animals. Minn Med 1980;63:795-801.

23. Gore JM, Weiner BH, Benotti JR, et al. Preliminary experience withsynchronized coronary sinus retroperfusion in humans. Circulation1986;74:381-8.