JACC Vol. 18, No. I July 1991:271-82

271

Synchronized Coronary Venous Retroperfusion for Support And Salvage of Ischemic Myocardium During Elective and Failed Angioplasty SHEILA KAR, MD, J. KEVIN DRURY, MD, FACC, ISTVAN HAJDUCZKI, MD, NEAL EIGLER, MD, Y ASUSHI W AKIDA, MD, FRANK LITV ACK, MD, F ACC, NEIL BUCHBINDER, MD, FACC, HAROLD MARCUS, MD, FACC, ROLF NORDLANDER, MD, ELIOT CORDAY, MD, FACC

Los Angeles, California

To determine the safety and efficacy of synchronized coronary venous retroperfusion during brief periods of ischemia, 30 pa­tients undergoing angioplasty of the left anterior descending coronary artery were studied. Each patient underwent a mini­mum of two angioplasty balloon inflations. Alternate dilations were supported with retroperfusionj the unsupported inflations served as the control inflations. Synchronized retroperfusion was performed by pumping autologous femoral artery blood by means of an electrocardiogram-triggered retroperfusion pump into the great cardiac vein through a triple lumen 8.5F balloon-tipped retroperfusion catheter inserted percutaneously from the right internal jugular vein. Clinical symptoms, hemodynamics and two-dimensional echocardiographic wall motion abnormalities were analyzed.

Retroperfusion was associated with a lower angina severity score (0.8 ± 1 vs. 1.2 ± 1) and delay in onset of angina (53 ± 31 vs.37 ± 14 Sj P < 0.05) compared with the control inflations. The magnitude of ST segment change was 0.11 ± 0.14 mV with retroperfusion and 0.16 ± 0.17 mV without treatment (p < 0.05). The severity of left ventricular wall motion abnormality was also

Percutaneous transluminal coronary angioplasty is an effec­tive treatment for coronary artery stenosis, but results in transient interruption of regional blood supply, which can cause myocardial ischemia with consequent hemodynamic, electrophysiologic, metabolic and functional derangements (1-3). These ischemic changes are of particular concern in patients with a large area of jeopardized myocardium or significant preexisting left ventricular dysfunction. Further-

From the Division of Cardiology, Department of Medicine, Cedars-Sinai Medical Center, and the University of California, Los Angeles School of Medicine, Los Angeles, California. This study was supported in part by Grants HL 14644-15 and HL 17561-15 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland and Retroperfusion Systems Inc., Costa Mesa, California.

Manuscript received July 23, 1990; revised manuscript received January 21, 1991, accepted January 29, 1991.

Address for reprints: Eliot Corday, MD, Cedars-Sinai Medical Center, Halper Research Building, Room 325, 8700 Beverly Boulevard, Los Angeles, California 90048.

©1991 by the American College of Cardiology

significantly (p < 0.01) reduced with retroperfusion compared with control (0.7 ± 1.4 [hypokinesia] vs. -0.3 ± 1.6 [dyskinesia]). There were no significant changes in hemodynamics, except in mean coronary venous pressure, which increased from 8 ± 3 mm Hg at baseline to 13 ± 6 mm Hg with retroperfusion. Four patients required prolonged retroperfusion for treatment of an­gioplasty-induced complications. The mean retroperfusion dura­tion in these patients was 4 ± 2 h (range 2 to 7). In the three patients who underwent emergency bypass surgery, the coronary sinus was directly visualized during surgery and found to be without significant injury.

There were no major complications. Minor adverse effects were transient atrial fibrillation (n = 2), jugular venous catheter insertion site hematomas (n = 4) and atrial wall staining (n = 1), all of which subsided spontaneously. Thus, retroperfusion signif­icantly reduced and delayed the onset of coronary angioplasty­induced myocardial ischemia and provided effective supportive therapy for failed and complicated angioplasty.

(J Am Coli CardioI1991;18:271-82)

more, acute closure of the dilated coronary artery occurs in approximately 5% of all angioplasty procedures (4). This complication is usually managed by coronary bypass surgery but may sustain a mortality rate as high as 7% (5,6).

Among techniques employed in an attempt to mitigate these ischemic derangements are the administration of phar­macologic agents and infusion of oxygenated fluorocarbon solutions or arterial blood through the angioplasty catheter (7-11). These techniques allow extension ofthe angioplasty balloon inflation time, but have certain limitations. The requirement of some of the latter techniques for a larger catheter system limits their ability to cross a severely stenotic lesion or pass distally within a tortuous vessel. Perfusion may also be ineffective when there are additional sequential stenoses or when the perfusion catheter obstructs a significant side branch. Percutaneous femorofemoral car­diopulmonary bypass, hemopump left ventricular assist and intraaortic balloon pumping devices may provide systemic

0735-1097/91/$3.50

272 KAR ET AL. SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

JACC Vol. 18. No. I July 1991 :271-82

circulatory support and reduce myocardial oxygen demand; however, blood supply to the jeopardized myocardium is not enhanced by these techniques and therefore myocardial ischemia may persist (12-14).

Synchronized coronary venous retroperfusion is a tech­nique that provides perfusion of the patient's arterial blood to the jeopardized myocardium retrogradely through the coronary veins. Because perfusion is retrograde rather than anterograde, its efficacy is not influenced by the number, location or degree of stenoses, tortuosity or branching of the coronary artery. This method has been extensively investi­gated in animals (15-22) and has been shown to provide significant myocardial support in preliminary clinical studies (23-30).

Coronary angioplasty offers a unique opportunity to examine the physiologic response to recurrent transient ischemic episodes (3). Therefore, using angioplasty as a short and reproducible ischemic model, the efficacy of various support systems has been studied (7-13). The pur­pose of the present study was to determine the safety and efficacy of a new synchronized coronary venous retroperfu­sion system (Retroperfusion Systems, Inc.) in this ischemic model.

Methods Study design. The study included 43 patients with ~70%

stenosis of the left anterior descending artery scheduled for angioplasty of the artery at Cedars-Sinai Medical Center. The patients were selected if the investigator believed that balloon inflation would result in significant ischemia by obstruction of blood flow to the first septal perforator or a large first diagonal branch of the left anterior descending coronary artery. Informed consent was obtained from all patients with use of an institutional and Food and Drug Administration-approved human investigation form. In all patients, a 9F sheath was introduced percutaneously into the right internal jugular vein for insertion of the retroperfusion catheter, followed by placement of 8F arterial sheaths in both femoral arteries.

Coronary angioplasty. Coronary angioplasty was per­formed in the standard manner. The effects of a minimum of two angioplasty balloon inflations were studied in each patient. Retroperfusion was applied during alternate balloon inflations, the order of which was randomized. In some patients, the first angioplasty balloon inflation was of brief duration (30 s) and performed to ensure proper positioning of the balloon within the stenosed segment of the coronary artery. This inflation was not included in the analysis.

All patients received heparin to maintain the activated clotting time between 360 and 480 s. Hemodynamic variables (heart rate, aortic blood pressure and coronary venous pressure) and a single precordial electrocardiographic (ECG) lead (V 3 to V J was monitored and continuously recorded at 25 mmls on an Electronics for Medicine VR-12 physiologic recorder.

Figure 1. Triple lumen 8.5F balloon-tipped gas inflatable retroper­fusion catheter. (Reproduced with permission from Retroperfusion Systems, Inc., Costa Mesa, California.)

The time to onset of angina was determined during all unsupported and retroperfusion-supported-balloon infla­tions. The severity of the pain was assessed on a semiquan­titative scale of 0 to 4 (0 = no pain, 1 = mild, 2 = moderate, 3 = severe, 4 = intolerable). ST segment change was measured 0.06 s after the J point and expressed as absolute change from the baseline measurement.

Quantitative two-dimensional echocardiography. Two­dimensional echocardiograms were recorded before and during each balloon inflation in apical two- or four-chamber views using an Advanced Technology Laboratories Ultra Mark VIII echocardiograph. The echocardiographic win­dow, contrast settings and depth were optimized during baseline imaging and then kept constant throughout the study. The images were digitized from videotape into a cine loop format using a commercial computerized analysis sys­tem (Microsonics, Inc.). Because the duration ofthe treated and control balloon inflations was not identical, the un­treated being of shorter duration, echocardiographic images during treated inflations were analyzed at similar durations as the matched untreated inflations. A modified centerline analysis method as described by Jaffe et al. (9) was used to quantify extent and severity of regional left ventricular wall motion abnormality. Extent was expressed as the total number of chords of the entire left ventricular perimeter that were affected. Severity was the degree of percent chordal shortening in the 10 most affected contiguous chords. Global left ventricular ejection fraction was calculated with the modified Simpson's rule.

The retroperfusion system. The retroperfusion device consists of an electronic pumping console, a specially de­signed retroperfusion catheter (Fig. 1) and an arterial cath­eter. Arterial blood is withdrawn from the femoral artery through a 25 cm long 8F catheter, featuring end and mUltiple side holes. This catheter is introduced through a sheath

KAR ET AL. 273 JACC Vol. 18, No. I July 1991:271-82 SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

PTCA CATHETER

Figure 2. Schematic representation of the method by which syn­chronized coronary venous retroperfusion is accomplished. Arterial blood is obtained from the femoral artery and shunted through the pump console into the great cardiac vein by means of the retroper­fusion catheter. AI = anterior interventricular; EM = electromag­netic; LAD = left anterior descending coronary artery; PICA = percutaneous transluminal coronary angioplasty; SRP = synchro­nized retroperfusion.

placed in the femoral artery not used for angioplasty. The arterial catheter is connected to the inlet of the disposable retroperfusion pump cassette, which is housed within the system console. Arterial blood is pumped during diastole by means of a piston whose motion is triggered by the R wave of the monitored ECG. The outlet of the pumping cassette is connected to the infusion lumen of the coronary sinus catheter (Fig. 2), which is a radiopaque 8.5F triple lumen catheter, featuring an inflatable balloon located 1 cm from the distal end of the catheter. Carbon dioxide gas is delivered through the catheter's second lumen to inflate the balloon during diastole (maximal diameter lO mm at full inflation). The third retroperfusion catheter lumen is used to monitor coronary venous pressure. During systole, the retroperfu­sion flow is interrupted and the balloon deflated to facilitate coronary venous drainage.

Coronary sinus catheterization. An attempt was made in each case to place the catheter deep into the coronary sinus, preferably into the great cardiac vein with the tip of the catheter distal to the posterior vein of the left ventricle. Contrast material was hand-injected through the infusion lumen of the catheter into the coronary veins with the catheter balloon inflated to confirm both adequate catheter positioning and proper occlusion of the coronary vein by the balloon (Fig. 3). In three patients, a guide wire was used to facilitate catheter placement.

Modifications of the retroperfusion system. Changes were made in the catheter configuration and pump system during

the course of the study. The objective was to optimize the retroperfusion system to permit use of larger diameter bal­loons on the retroperfusion catheter and obtain higher flow levels, and consequently higher coronary venous pressures, and thus be able to perform retroperfusion from proximal positions in the coronary sinus when distal catheter place­ment was not possible. The first eight patients were studied with use of an autoinflatable balloon retroperfusion catheter with two lumens and a maximal flow rate of 96 mllmin. The next eight were studied with use of a triple lumen gas­activated balloon catheter with dual-valve pumping cas­settes, providing maximal flow rates of 124 mllmin. The last 14 patients were studied with use of the same triple lumen gas-activated retroperfusion catheter, but with a bladder pump cassette delivering maximal flow rates of 250 mllmin.

Statistics. All data are presented as mean values ± 1 SD. The change from preangioplasty balloon inflation to peak angioplasty balloon inflation in hemodynamic variables, magnitude of ST segment change and echocardiographic data was analyzed by the two-tail paired t test. Comparisons between the untreated and retroperfusion balloon inflation groups were made by the unpaired t test with these excep­tions: the time of onset of chest pain and severity score of angina were analyzed by the Mann-Whitney rank-sum test and the incidence of ST segment change and angina was compared with chi-square analysis.

Results Feasibility. Catheterization of the coronary sinus was

successful in 36 (84%) of 43 patients, with a mean insertion time of2.6 min (range 0.5 to 10) after the catheter was placed in the right atrium. Of the seven unsuccessful cases, the jugular vein could not be catheterized in three, the catheter could not be placed in the coronary sinus in three and fluoroscopy demonstrated staining of the right atrial wall after contrast injection through the catheter in one case. In this patient no further attempts at coronary sinus catheter­ization were made, the staining disappeared after 10 min and the patient's echocardiogram was normal. Six of the 36 patients undergoing catheterization did not complete the study. In three of the six, coronary angioplasty could not be performed for technical reasons; in one patient, ventricular fibrillation developed during the first control angioplasty balloon inflation and the patient was sent directly for coro­nary bypass surgery and in two patients, a stable coronary venous catheter position could not be maintained with pumping.

The retroperfusion technique was thus feasible in 34 (79%) of 43 patients, but because of the reasons just enu­merated, only 30 patients completed the protocol. These patients underwent 128 angioplasty balloon inflations: 61 inflations were treated with synchronized retroperfusion and 67 were untreated. The mean duration of the retroperfused and untreated inflations was 81 ± 21 and 73 ± 18 s,

274 KAR ET AL. SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

JACC Vol. 18. No. I July 1991:271-82

respectively (p < 0.01). The clinical characteristics of the 30 patients are listed in Table 1.

Hemodynamics. Heart rate and diastolic blood pressure did not change significantly during retroperfusion, but sys­tolic pressure decreased significantly during untreated balloon inflations (from 125 ± 22 to 108 ± 24 mm Hg) and during retroperfusion (from 120 ± 22 to 111 ± 21 mm Hg). The mean coronary venous pressure was 13 ± 6 mm Hg (range 4 to 26) with retroperfusion compared with 8 ± 3 mm Hg (range 1 to 14) without retroperfusion (p < 0.01). The peak coronary venous pressure was 29 ± 12 mm Hg (range 6 to 59) with retroperfu­sion and 12 ± 5 mm Hg (range 4 to 26) without retroperfusion (p < 0.01) (Fig. 4), Coronary venous pressure did not change during angioplasty balloon inflations.

Angina. Angina occurred in 40% of the angioplasty bal­loon inflations supported by retroperfusion and in 53% of the unsupported inflations. The time to onset of angina was 53 ±

Figure 3. Coronary venograms of two patients during left anterior descending artery angio­plasty, showing proximal filling of all the tribu­taries of the coronary veins and regurgitation past the nonocclusive synchronized retroperfu­sion catheter balloon (A) and distal penetration of the retroperfusate as a result of adequate retroperfusion balloon catheter position (B). The venograms were obtained by injecting con­trast material into the retroperfusion pump sys­tem (right anterior oblique view). AIV = ante­rior interventricular vein; B = retroperfusion catheter balloon; GCV = great cardiac vein; MCV = middle cardiac vein; PICA = angio­plasty balloon; PV = posterior vein of the left ventricle; SRP Cath = coronary sinus retroper­fusion catheter; W = angioplasty guide wire.

31 s in the retroperfusion-supported inflations compared with 37 ± 14 s without retroperfusion (p < 0.05). The angina severity score was 0.8 ± 1 with retroperfusion compared with 1.2 ± 1 without retroperfusion (p < 0.05).

ST segment changes. ST segment change (;:::0.1 mY) occurred in 46% of the balloon inflations with retroperfusion compared with 52% of the control inflations. The magnitude of the ST segment change was 0.11 ± 0.14 mY with retroperfusion and 0.16 ± 0.17 mY without retroperfusion (p < 0.05), despite longer duration of the inflations supported with retroperfusion.

Two-dimensional echocardiography (Table 2). Echocar­diograms were recorded in 23 of the 30 patients. Quantitation of the echocardiograms was not possible in eight patients because of poor image quality. Moreover, two patients did not develop any ischemia during the angioplasty balloon inflations performed without retroperfusion and were not

KAR ET AL. 275 JACC Vol. 18, No.1 July 1991:271-82 SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

Table 1. Demographics of the 30 Patients Who Completed the Study Protocol

Age Pt (yr)1

Past History Symptoms Medication Assoc

Diseases Cardiac

Catheterization Data

No.ofPTCA Inflations

No. Gender Weight

(kg) MI PTCA CABG CCS NTG CB BB ASA DM HTN RCA LCx LAD LMCA SRP­

EF(%) Untreated Supp

2 54/M 841M 67/M

4* 64/M 5 52/F 6 601M 7 53/M 8 53/M 9' 731M

10 50/F 11 12 13 14 15

46/M 49/M 501M 48/M 62/M

16 561M 17 63/M 18 391M 19 69/M 20 49/M 21 47/M

77 80 Ant 82 91 36 85 Ant 85 Inf 75 67 86 91 77 82

78 Ant

80 Ant, inf 73 64 59 84

+

+

+ + +

+ +

+

4 4

4

4 4

4

+ + +

+ + +

+ +

+

+ + +

+ +

+

+ + +

+

+ +

+ +

+

+

+

+

+ +

+ + + +

+ +

+ + +

+

+

+

+ + + Ocd

+ + +

+ + + + + + + + + + + + + + + + + +

>55 >55 >55 >55 >55

35-55 35-55

>55 >55 >55 >55

35-55 >55 >55 >55 >55 <35 <35

+ >55 + >55 + <55

2 2

2 2 2 2 2

2 2 2 2 2

2 2 2 2 2

1 2

2 2 2 2 4 2 2

22 711F 70 + +

4 2 4 4 4 4 4 4 2 4

+ + + + + Narrowing <35

2 2

3 4 5

23 73/F 66 Inf + + + + + +

+ + +

+ Narrowing <35 24 62/M 25 68/M 26 40/F 27* 72/F

28 29 30

421M 56/M 751F

73 75 Ant 89 66 Ant

89 91 63 + +

4

2 4 4

+

+ + + + +

+ + + + +

+ +

+ +

+

+

+ +

+ + +

+ 35-55 + +

35-55 35-55

+ Stenosis, 35-55

+ +

thrombus

+ Narrowing

35-55 35-55

<35

2 4

2 2 2

2 2 4

2

2

'Prolonged retroperfusion; Ant = anterior; ASA = aspirin; Assoc = associated; BB = beta-blocker; CABG = coronary artery bypass grafting; CB = calcium channel blocker; CCS = Canadian Cardiovascular Society class; DM = diabetes mellitus; EF = left ventricular ejection fraction; F = female; HTN = hypertension; Inf = inferior; LAD = left anterior descending coronary artery; LCx = left circumflex coronary artery; LMCA = left main coronary artery; M = male; MI = myocardial infarction; NTG = nitroglycerin; Occl = total occlusion; PTCA = coronary angioplasty; RCA = right coronary artery; SRP-Supp = retroperfusion-supported; patient numbers in italics indicate patients with analyzable echocardiograms; - = missing data. Note: All coronary artery stenoses are marked positive only if >70% narrowing except of the left main lesion.

included in the analysis. Therefore, echocardiograms were analyzed in 13 patients.

Global function (Table 2). Left ventricular ejection frac­tion decreased from 51 ± 13% to 43 ± 8% (p < 0.05) with angioplasty balloon inflations supported by retroperfusion compared with 52 ± 9% to 32 ± 10% (p < 0.01) in the untreated inflations. The difference in ejection fraction change was statistically significant (p < 0.01) between groups (Fig. 5A).

Regional function (Table 2). The ischemic zone percent chordal shortening decreased from 1.5 ± 1.7% at baseline to 0.7 ± 1.4% (mild hypokinesia) during the balloon inflations with retroperfusion, whereas in the control inflations, the ischemic zone percent chordal shortening decreased from 1.6 ± 1.6% to -OJ ± 1.6% (p < 0.01), the negative value

indicating dyskinesia of the regional wall (Fig. 5B). The difference in percent chordal shortening during angioplasty without and with retroperfusion reached statistical signifi­cance (p < 0.01). The extent of left ventricular wall motion abnormality during the angioplasty balloon inflation was significantly less with retroperfusion (21 ± 9%) compared with that during inflation without retroperfusion (36 ± 12%) (p < 0.01) (Fig. 5C).

Hematology. The mean red blood cell counts, hemato­crit, hemoglobin and platelet counts were 4.5 million/mm3, 40.4%, 13.7 g/dl and 212,OOO/mm3, respectively, before ret­roperfusion and 4.4 million/mm3, 39.8%, 13.3 g/dl and 205,OOO/mm3 after completion of the protocol. The plasma­free hemoglobin increased from 5.6 mg/dl before retroperfu­sion to 21.7 mg/dl after retroperfusion (p = 0.25). These

276 KARET AL. SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

JACC Vol. 18, No.1 July 1991:271-82

60

C. :r: ~ 50 .JJ * 0:: ::J ~ 40 w 0:: Q.

~ 30 w > >-0:: ~ 20 0 gj u

10 t f i i f i

SRP SRP SRP SRP OFF ON OFF ON

MEAN MAXIMUM

Figure 4. Mean and maximal coronary venous pressure with and without synchronized retroperfusion (SRP) at three different flow rates in all the patients. Triangles = retroperfusion flow rate of 83 ± 12 mVmin; closed circles = flow rate of 115 ± 10 mVmin; open circles = retroperfusion flow rate of 203 ± 44 mVmin.

values include the four patients who underwent prolonged retroperfusion pumping.

Complications. There were no major complications due to retroperfusion. Minor adverse effects included transient atrial fibrillation (n = 2), hematoma at the internal jugular vein insertion site (n = 4) and atrial wall contrast agent staining (n = O. All these complications resolved completely without specific therapy.

RetroperJusion During Complicated/Failed Angioplasty,' Case Reports

Four patients required longer periods of synchronized retroperfusion because of postangioplasty abrupt occlusion in three patients and intolerable angina during the untreated angioplasty balloon inflations in the fourth patient.

Table 2. Echocardiographic Results in 13 Patients

Baseline

Regional function

Untreated Balloon Inflations

Case 1 (Fig. 6). A 64 year old man presented with progressive exertional angina due to a 95% proximal left anterior descending coronary artery stenosis involving the first septal perforator (Fig. 6A). Five minutes after comple­tion of the angioplasty-retroperfusion protocol, he suddenly developed severe chest pain, hypotension (blood pressure <60 mm Hg) and complete heart block; global hypokinesia was seen on the echocardiogram (Fig. 6C). Intracoronary nitroglycerin, right ventricular pacing and dopamine infusion did not reverse the cardiogenic shock, which was deter­mined by angiography to be due to postangioplasty total occlusion of the proximal left anterior descending coronary artery (Fig. 6B). The retroperfusion catheter was reposi­tioned and retrograde pumping begun in the hope of main­taining viability of the jeopardized myocardium. Within minutes the chest pain abated, blood pressure stabilized at 90 to 100 mm Hg and echocardiography showed marked improvement in global wall motion of the left ventricle (Fig. 6C). The patient was supported by retroperfusion for 2.5 h while several angioplasty balloon dilations were attempted but failed and the patient was sent for emergency coronary bypass surgery. The patient underwent successful double­vessel bypass grafting. The coronary sinus appeared normal on direct visualization during surgery. The patient had an uneventful hospital course and left the hospital with a left ventricular ejection fraction of 61% on radionuclide scans.

Case 2. A 73 year old man with diabetes, hypertension and a past history of myocardial infarction presented with severe chest pain. The ECG revealed complete right bundle branch block and the coronary angiogram showed 30% stenosis of the left main coronary artery, two severe serial stenoses in both the mid left anterior descending and prox­imalleft circumflex coronary arteries and a severe proximal stenosis in the right coronary artery. During the angioplasty procedure for the left coronary arteries, both vessels kept intermittently occluding with development of significant ischemia as evidenced by severe angina. Echocardiograms were not recorded in this patient. Retroperfusion was per­formed throughout the 2 h angioplasty procedure with relief of the angina. Eventually both vessels were dilated with residual stenoses of 40%, but the patient died 7 days later.

Balloon Inflation Baseline

Retroperfusion During Inflations

Balloon Inflation

Severity WMA (%CS) 1.6 ± 1.6 -0.3 ± 1.6' 1.5±1.7 0.7 ± l.4t Extent WMA (%L V) 36 ± 12 21 ± 9t

Global function Ejection fraction (%) 52 ± 9 32 ± 10' 51 ± 13 43 ± 8H

*p < 0.01 versus baseline, tp < 0.01 versus untreated; tp < 0.05. %CS == percent chordal shortening; %LV == percent of the left ventricle; WMA == wall motion abnormality.

KAR ET AL. 277 JACC Vol. 18, No.1 July 1991 :271-82 SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

UNTREATED RETROPERFUSION

z 70 0 f= u 60

1 « If 50 z 0 40 f= u 30 w ~ w

!z 20

w 10 u 0:: W 0 0..

A PRE PTCA INF. INF.

PRE PTCA INF. INF.

UNTREATED RETROPERFUSION

7

6 ~ (!) z 5 Z W I- 4 0:: 0 a; 3

~ Cl 2 0:: 0 :I: U I-Z 0 W u ffi -\ >< 0..

-2

-3

B PRE PTCA PRE PTCA INF. INF. INF. INF.

70

60 • •

~ 50 • I.L.

40 !~ 0 l-

I Z 30 w -f

u 0:: 20 -w a... • f

10 • . 0 C UNTREATED RETROPERFUSIDN

Figure 5. Echocardiographic findings in 13 patients. A, Ejection fraction by the centerline method. The individual and mean ejection fractions obtained by quantitation of the two-dimensional echocar­diogram in both the retroperfusion-supported and unsupported (untreated) inflations before (PRE INF.) and during (PICA INF.) angioplasty balloon inflation. B, Severity of wall motion abnormality (expressed as percent chordal shortening) obtained by quantitation of the apical four chamber echocardiogram by the centerline method, showing individual and mean values for the groups with and without retroperfusion support before (PRE INF.) and during (PICA INF.) angioplasty balloon inflation. C, Individual and mean values for the extent of the left ventricle (LV) with severe wall motion abnormality during angioplasty balloon inflation (expressed as a change from the baseline) in the inflations with and without retroperfusion support. See text for p values.

At postmortem examination, the cause of death was found to be a fresh infarction in the left circumflex coronary artery territory. The coronary sinus and veins were normal.

Case 3 (Fig. 7). A 72 year old woman with diabetes, hypertension and congestive heart failure presented with unstable angina and anterolateral ischemic ECG changes. She had 90% complex stenoses of both the proximal left anterior descending (Fig. 7 A) and left circumflex coronary arteries and severe sequential stenoses of the right coronary artery. The angioplasty procedure of the proximal left ante­rior descending coronary artery was complicated by rupture of the angioplasty balloon two times, followed by dissection of the left main coronary artery (Fig. 7B). The angioplasty procedure was therefore abandoned. The patient underwent prophylactic retroperfusion during the angioplasty proce­dure and while awaiting coronary bypass surgery; retroper­fusion continued for a total of 7 h from the start of angio­plasty to the institution of cardiopulmonary bypass in the operating room, during which time she was clinically stable. At operation, the coronary sinus was found to be normal and there was no evidence of acute myocardial infarction. At 1 year, the patient is asymptomatic and the ECG shows only a loss of the R wave in lead VI along with the preexisting anterolateral ischemic changes.

Case 4. A 62 year old man was admitted to the hospital with unstable angina. Angiography revealed two discrete 95% proximal stenoses of the left anterior descending coro­nary artery. During angioplasty, he began complaining of severe angina when the guidewire was inserted across the lesion even before the angioplasty balloon catheter was inserted. Synchronized retroperfusion was started and alle­viated the angina. Echocardiograms were not recorded in this patient. Retroperfusion was continued during 13 succes­sive angioplasty balloon inflations and thus the patient had to be excluded from the protocol because control inflations were not possible. Because angioplasty was not successful, the patient underwent coronary bypass surgery with saphe­nous vein grafts to the diagonal, circumflex and posterior descending arteries and an internal mammary graft to the left anterior descending coronary artery. The coronary sinus and veins were normal. The patient is asymptomatic at 2 years.

Discussion In this study in 36 patients coronary sinus retroperfusion

during balloon angioplasty was safe, feasible and effective in ameliorating and delaying the onset of ischemia. Further­more, retroperfusion provided myocardial support during complicated and failed angioplasty and reversed cardiogenic shock in one patient.

Safety of the retroperfusion technique. There are three safety issues: 1) safety of coronary sinus catheterization through the jugular vein; 2) the effects of retroperfusion on the coronary venous system; and 3) the effects of retroper­fusion on the formed blood elements. Coronary sinus cath-

278 KAR ET AL. SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

JACC Vol. 18, No. I July 1991 :271-82

eterization is a safe technique for measuring coronary blood flow (31-33). In our study, the small hematomas at the jugular venous insertion site could probably have been avoided by a left brachial artery approach. Nevertheless, the hematomas resolved without specific treatment and no pa­tient required blood transfusion. Atrial fibrillation occurred probably secondary to irritation of the atrial wall. It resolved spontaneously and was not associated with any chest pain or hemodynamic effects. In one patient, there was evidence of injury to the atrial wall as demonstrated by staining after contrast injection through the retroperfusion catheter. How­ever, because of the injury , retroperfusion was not employed to avoid the potential for extensive damage.

The effects of synchronized retroperfusion on the coro­nary sinus and coronary veins have been studied extensively in the experimental model (15-22) and humans (23-30). On direct visualization of the coronary sinus in three patients undergoing coronary bypass surgery and in one patient at postmortem examination, Gore et al. (28) noted minor bruis-

Figure 6. Case I . Left coronary artery angiogram taken before angioplasty, showing tight proximal stenosis in the left anterior descending coronary artery (arrow) but with some distal filling (A) and after angioplasty when the patient was in cardio­genic shock, showing total occlusion of the artery (arrow) and very poor distal filling (B) . C, End­systolic two-dimensional echocardiographic im­ages recorded during abrupt occlusion of the left anterior descending coronary artery as a result of dissection (left) and after 150 min of retroperfusion (right) . Note the severe global hypokinesia of the left ventricle during shock, which was reversed to normal contraction after retroperfusion as repre­sented by the significantly reduced end-systolic cavity.

ing of the coronary sinus and one instance of a microscopic nonocclusive thrombus. These findings were corroborated in our study, in which the coronary sinus was visualized in the three patients who were sent for coronary bypass surgery and one in whom a postmortem study was performed. The coronary sinus was normal in three and only superficially bruised in one.

Previous studies (16,17) indicate that severe damage to venules occurred when coronary venous pressures exceed 60 mm Hg peak or 40 mm Hg mean pressure. In our study, the average mean and peak coronary venous pressures were kept well below those levels at 13 ± 6 and 29 ± 12 mm Hg, respectively. The pump is equipped with a safety mechanism that terminates retroperfusion when coronary venous pres­sures exceed prescribed limits.

We also confirmed prior studies (16,28) that showed no significant changes in red cell, white cell and platelet counts or plasma-free hemoglobin levels. Thus, the safety of the retroperfusion technique was confirmed in the current study

KARET AL. 279 JACC Vol. 18, No.1 July 1991:271-82 SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

Figure 7. Case 3. Left coronary angiogram taken before angioplasty, showing tight proxi-mal stenosis in the left anterior descending A coronary artery (arrow) (A) and after angio-plasty, showing left main coronary artery dis­section (arrow) (B). This patient underwent retroperfusion for 7 h and was clinically stable until she underwent cardiopulmonary bypass in the operating room.

by the absence of any hemodynamic, hematologic, morpho­logic or clinically evident adverse effects.

Optimization of the retroperfusion technique. It has been suggested by previous studies (20,27,34) that high flows and increased coronary venous pressures are required to over­come the rich alternating drainage that may shunt the retroperfusate away from the ischemic bed. Therefore, the system was modified to increase the maximal flow rate from 100 to 250 mllmin. An attempt was made to correlate flow rate, coronary venous pressure, catheter position and effi­cacy. There was significant correlation (p < 0.01) between retroperfusion flow rate and coronary venous pressure; however, differences in echocardiographic wall motion could not be appreciated because of the small number of patients (n = 4) in the groups with different flow rates. In the three patients in whom there was no efficacy, the catheter tip was very proximal in the coronary sinus and venography showed inadequate retrograde penetration of the contrast

material. Thus, efficacy is related to multiple factors, in­cluding flow rate, coronary venous anatomy, catheter posi­tion relative to the site of coronary artery obstruction and pressure gradient between the coronary venous and distal arterial beds, as shown by Meesman et al. (35) in a canine model.

Feasibility of the technique. Feasibility of retroperfusion involves three issues: ease and rapidity of apphcatlon, special training requirements and patient applicability. The retroperfusion catheter was successfully positioned in 84% of patients and the technique was feasible in 79%. Catheter placement took 2.5 min (range 0.5 to 10) after placement of the catheter in the right atrium; this timing included the time taken by both experienced and inexperienced physicians who attempted the technique. An additional 20 min was required to prime the pump and start synchronized retro­perfusion. Neither a perfusionist nor a surgical team was required. Moreover, retroperfusion has been used success-

280 KARETAL. SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

JACC Vol. 18, No.1 July 1991:271-82

fully in hospitals lacking a catheterization laboratory, in­traaortic balloon pump or cardiac surgery facilities (29,30).

In this study, patients between 18 and 85 years of age were included; therefore, patient age was not a contraindi­cation to retroperfusion. We limited the use of retroperfu­sion to left anterior descending coronary artery support. The efficacy of retroperfusion in the reduction of left circumflex coronary artery ischemia in evolving myocardial infarction was reported by Barnett and Touchon (29).

Efficacy of retroperfusion. Experimental studies (15-22) of retroperfusion during acute ischemia have demonstrated a significant reduction in infarct size, enhanced myocardial perfusion, restoration of glucose metabolism and significant improvement in left ventricular function. Clinical studies (23-29) have suggested efficacy in the treatment of unstable angina, acute myocardial ischemia and as an adjunct during balloon angioplasty in the left anterior descending coronary artery distribution.

In this study the effects of retroperfusion on brief left anterior descending coronary artery occlusion during angio­plasty were studied in patients who were predicted to develop severe ischemia. Moreover, in this study design, each patient was his or her own control. No patient had left bundle branch block or QS patterns on the ECG in the region of interest, which could also affect the results. The efficacy of retroperfusion was demonstrated by significant diminu­tion and delay in the onset of ischemia, as evidenced by left ventricular wall motion, angina and ST segment change. This efficacy suggests that retroperfusion may be used prophylactically as a device for "supported angioplasty" in high risk patients, as described by Vogel et al. (14) when defining the role of cardiopulmonary bypass in high risk angioplasty. Also, in cases where prolonged angioplasty balloon inflation times are required, the support of retroper­fusion may be beneficial. Retroperfusion was not effective in all patients, but in the majority of patients the ischemic response was delayed or decreased although not totally abolished. The role of this technique, therefore, is as a prophylactic support technique during high risk and complex angioplasty rather than for routine angioplasty.

Prolonged retroperfusion for complicated angioplasty. The complication rate of coronary angioplasty in this study was 13%. This high rate may be attributed to the fact that patients predicted to develop significant ischemic events were selected. Abrupt occlusion occurred in two patients (Cases 1 and 2), reSUlting in cardiogenic shock and severe angina, respectively. In both patients conventional therapy was unsuccessful, but retroperfusion promptly reversed the ischemia. Thus, retroperfusion can reverse cardiogenic shock due to abrupt coronary occlusion during angioplasty and is helpful in difficult and prolonged multivessel angio­plasty complicated by symptomatic vessel closure.

Left main coronary artery dissection developed during angioplasty of the left anterior descending coronary artery in another patient (Case 3). This is a rare complication that has been reported by Lotan et al. (36). This patient also had a

thrombus in the left anterior descending coronary artery and was thus at greater risk for abrupt occlusion (27). Even though she was asymptomatic and hemodynamically stable, retroperfusion was performed prophylactically because of her high risk for a sudden severe ischemic event. This case suggests that synchronized retroperfusion may be helpful if kept as standby during high risk angioplasty.

Case 4 illustrates the development of severe ischemia due to occlusion by a guide wire of a coronary artery supplying a large portion of viable myocardium. The patient's severe chest pain was refractory to intracoronary nitroglycerin. Retroperfusion decreased the chest pain and allowed the patient to undergo angioplasty, although the latter was unsuccessful and emergency coronary bypass surgery was required. This case suggests the efficacy of retroperfusion in difficult angioplasty, where the ischemia is so significant that occasionally the procedure has to be abandoned.

It has been reported (37,38) that the incidence of Q wave infarction is 55% to 77% and the mortality rate 5% to 12% in patients requiring emergency coronary bypass surgery after failed angioplasty. The risk is greater when the failed angio­plasty is complicated by hemodynamic instability or the presence of ischemic symptoms (39). All three patients in our study who required emergency bypass surgery were in the greater risk group and yet did not develop Q waves or die after surgery.

These cases suggest that retroperfusion is an effective support system in the event of angioplasty-induced compli­cations. However, immediate permanent revascularization should always be the goal of therapy, retroperfusion provid­ing a bridge until reperfusion is established.

Retroperfusion and other myocardial support techniques. Systemic support systems (namely, the intraaortic balloon pump, percutaneous cardiopulmonary bypass and hemo­pump left ventricular assist devices) maintain peripheral perfusion and systemic hemodynamics but have no direct myocardial protective effects. In contrast, retroperfusion provides nutrient blood flow to the myocardium and en­hances washout of toxic metabolites from underperfused areas but does not directly support the systemic circulation. Therefore, there may be some value in combining retroper­fusion with one of the other techniques to provide both myocardial and systemic support in some very high risk patients.

Arterial blood perfusion (passive) or infusion (active) through the angioplasty catheter effectively prevents the onset of ischemia, allowing prolonged angioplasty balloon inflation times (7). The limitations of this technique are low flow rates, clot formation, inadequate driving pressure in severely hypotensive patients and a potential risk of perfo­ration or dissection of the coronary artery with the catheter tip. Moreover, because it provides perfusion to the territory supplied by the vessel in which it is placed, global myocar­dial perfusion is not provided as required in diffuse multi­vessel disease. During abrupt vessel closure after removal of

KAR ET AL. 281 JACC Vol. 18, No. I July 1991 :271-82 SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

the angioplasty balloon catheter and wire, recrossing the lesion may be unsuccessful in 30% to 50% of cases (39).

Intracoronary perfusion of oxygenated Fluosol-DA 20% (Alpha Therapeutic Corporation), a fluosol emulsion, does improve left ventricular function, but does not return it to normal in experimental and clinical studies (8,9). There are also several limitations. It takes 30 to 40 min to defrost and oxygenate, which may limit its use in emergency situations; also, it is not as effective as arterial blood (40). Moreover, because of limitations in the total volume that may be infused, fluosol may be restricted in its use as support or maintenance therapy in failed angioplasty.

Because all the myocardial support techniques require special catheters to be introduced through the target lesion of the coronary artery, when access is lost and cannot be regained rapidly, coronary venous retroperfusion may pro­vide the only means of rapidly perfusing the myocardium until surgery. In addition, cardioprotective drugs and car­dioplegia may be delivered effectively by this route in the presence of an arterial occlusion (27).

Limitations of the study. In this study, a single precordial lead ECG was monitored. A 12 lead ECG or perhaps an epicardial electrogram would have provided a more sensitive marker of ischemia (41). Wall motion abnormality measure­ment obtained by two-dimensional echocardiography was feasible in only 50% of patients. The supine position of the patient posed problems in attaining good echocardiographic windows. Cardiac hemodynamic variables such as cardiac index and stroke work index were not measured in this study. These variables were monitored in a study by Cos­tantini et al. (26), who reported that retroperfusion signifi­cantly improved cardiac index and stroke work index. Their results suggest that retroperfusion does preserve cardiac hemodynamics during ischemia. In this study, only short duration angioplasty balloon inflations were studied. The results in the complicated cases, however, indicate its po­tential during prolonged ischemia. It is also important to note that this study deals with retroperfusion of the left anterior descending coronary artery territory only. The role of ret­roperfusion in other vessel territories remains to be ade­quately defined.

Retroperfusion is invasive and requires two additional puncture sites. However, unlike intraaortic balloon pumping or cardiopulmonary bypass, the vascular complications have been negligible as a result of the smaller size sheaths. Significant vascular occlusive disease was not encountered in this study but may pose a problem. If retroperfusion is performed in conjunction with intraaortic balloon pumping or the cardiopulmonary pump system, a separate sheath and arterial catheter are not mandatory. Blood may be obtained from the side arm of the angioplasty sheath in extreme conditions. As mentioned earlier, to prevent jugular venous hematoma, left brachial insertion sites may be used. Lastly, it is important to note that this technique requires training in right heart and coronary sinus catheterization before it can be applied.

Conclusions. This new synchronized coronary venous retroperfusion system was safe and feasible during coronary angioplasty. It significantly reduced and delayed the onset of ischemia during angioplasty and preserved regional and global myocardial function. It also reversed severe ischemia and cardiogenic shock during angioplasty complicated by abrupt coronary artery occlusion and supported these pa­tients until definitive revascularization therapy was carried out. Retroperfusion thus appears to be an effective standby technique for high risk and complex coronary angioplasty and a support device in complicated and failed cases.

We thank James S. Forrester, MD, FACC, Director of Cardiology, Cedars­Sinai Medical Center and Kenneth G. Lehmann, MD, Division of Cardiology, Long Beach Veterans Administration Medical Center for their valuable suggestions. We also thank Ronald Arzaga for typing the manuscript, Piper Fogle and Stephanie Smith for performing the echocardiograms and Jeanne Bloom for editorial assistance. We express our appreciation to Retroperfusion Systems, Inc. for supplying the equipment and catheters used in the study and to the cardiac catheterization laboratory staff at Cedars-Sinai Medical Center for cooperation and support.

References 1. Serruys PW, Wigins W, Van den Brand M, et al. Left ventricular

performance, regional blood flow, wall motion, and lactate metabolism during transluminal angioplasty. Circulation 1984;70:25-36.

2. Hauser AM, Gangadharan V, Ramos RG, Gordon S, Timmis GC. Sequence of mechanical, electrocardiographic and clinical effects of repeat coronary artery occlusion in human beings: echocardiographic observations during coronary angioplasty. J Am Coli Cardiol 1985;5: 193-7.

3. Wohlgelernter D, Cleman M, Highman HA, et al. Regional myocardial dysfunction during coronary angioplasty: evaluation by two-dimensional echocardiography and I2 lead electrocardiography. J Am Coli Cardiol 1986;7:1245-54.

4. Cowley MJ, Dorros G, Van Raden M, Detre KM. Acute coronary events associated with percutaneous transluminal coronary angioplasty. Am J CardioI1984;53: 12C-6C.

5. Detre K, Holubkor R, Kelsey S, et al. Percutaneous transluminal coro­nary angioplasty in 1985-1986 and 1977-1981: the National Heart, Lung and Blood Registry. N Engl J Med 1988;318:265-70.

6. Dorros G, Cowley MJ, Janke L, Kelsey SF, Mullins SM, Van Raden M. In-hospital mortality rate in the National Heart, Lung and Blood Institute Percutaneous Transluminal Coronary Angioplasty Registry. Am J Cardiol 1984;53: 17C-2IC.

7. Lasala JM, Cleman MW. Myocardial protection during percutaneous transluminal coronary angioplasty. Cardiol Clin 1988;6:329-43.

8. Cleman M, Jaffe CC, Wohlgelernter D. Prevention of ischemia during percutaneous transluminal coronary angioplasty by transcatheter infusion of oxygenated F1uosol DA 20%. Circulation 1986;74:555-62.

9. Jaffe CC, Wohlgelernter D, Cabin HS et al. Preservation of left ventric­ular ejection fraction during percutaneous transluminal coronary angio­plasty by distal transcatheter coronary perfusion of oxygenated Fluosol DA 20%. Am Heart J 1988;115:1156-64.

10. Lehmann KG, Atwood JE, Synder EL, Ellison RL. Autologous blood perfusion for myocardial protection during coronary angioplasty: a feasi­bility study. Circulation 1987;76:312-23.

11. Stack RS, Quigley PJ, Collins G, Phillips HR III. Perfusion balloon catheter. Am J CardioI1988;61:776-806.

12. Merhige ME, Smalling RW, Cassidy D, et aI. Effect of hemopump left ventricular assist device on regional myocardial perfusion and function, reduction of ischemia during coronary occlusion. Circulation 1989; 80(suppl III):III-158-68.

13. Alcan KE, Stertzer SH, Wallsh JE, DePasquale NP, Bruno MS. The rate of intra-aortic balloon counterpulsation in patients undergoing percutane­ous transluminal coronary angioplasty. Am Heart J 1983;105:527-30.

282 KARETAL. SYNCHRONIZED RETROPERFUSION DURING ANGIOPLASTY

JACC Vol. 18, No. I July 1991 :271-82

14. Vogel RA, Shawl F, Tommasco C, et aI. Initial report of the National Registry of elective cardiopulmonary bypass supported coronary angio­plasty. J Am Coll Cardioll990;15:23-9.

15. Meerbaum S, Lang TW, O'Shea JV, et aI. Diastolic retroperfusion of acutely ischemic myocardium. Am J CardioI1976;37:558-98.

16. Drury JK, Yamazaki S, Fishbein M, Meerbaum S, Corday E. Synchro­nized diastolic coronary venous retroperfusion: results of a preclinical safety and efficacy data. J Am Coll CardioI1985;6:328-35.

17. Yamazaki S, Drury JK, Meerbaum S, Corday E. Synchronized coronary venous retroperfusion: prompt improvement of left ventricular function in experimental myocardial ischemia. JAm Coll CardioI1985;5:655-63.

18. Geary GG, Smith GT, Suehiro GT, Seman C, Sire B, McNamara JJ. Quantitative assessment of infarct size reduction by coronary venous retroperfusion in baboons. Am J CardioI1981;48:1064-70.

19. Zalewski A, Goldberg S, Slysh S, Maroko PRo Myocardial protection via coronary sinus interventions: superior effects of arterialization compared with intermittent occlusion. Circulation 1985;71:1215-23.

20. Chang B-L, Drury JK, Meerbaum S, Fishbein MC, Whiting JS, Corday E. Myocardial washout and retrograde blood delivery with synchronized retroperfusion during acute myocardial ischemia. J Am Coll Cardiol 1987;9:1091-9.

21. Berdeaux A, Farcot JC, Bourdarias JP, Barry M, Bardet J, Giudicelli JF. Effects of diastolic synchronized retroperfusion on regional blood How in experimental myocardial ischemia. Am J CardioI1981;47:1033-40.

22. 0' Byrne GT, Nienaber CA, Miyazaki A, et aI. Positron emission tomog­raphy demonstrates that coronary sinus retroperfusion can restore re­gional myocardial perfusion and preserve metabolism. J Am Coll Cardiol 1991;18:257-70.

23. Berland J, Farcot JC, Barrier A, Dellac A, Gamra H, Letac B. Coronary venous synchronized retroperfusion during percutaneous transluminal angioplasty of left anterior descending coronary artery. Circulation 1990; 8 I (suppl IV):IV-35-42.

24. Hajduczki I, Kar S, Areeda J, et aI. Reversal of chronic regional dysfunction (hibernating myocardium) by synchronized diastolic coro­nary venous retroperfusion during coronary angioplasty. J Am Coll Cardioll990;15:238-42.

25. Nishida K, Kodama K, Mishima M, et al. Retrograde coronary venous myocardial contrast echocardiography in humans: can coronary venous retroinfusion deliver pharmacologic agents to the myocardium? (abstr). JAm Coll Cardioll990;15:195A.

26. Costantini C, Sampaolesi A, Pieroni M, et al. Coronary venous retroper­fusion support during high-risk coronary angioplasty in patients with acute ischemic syndromes (abstr). J Am Coll CardioI1989;13:625A.

27. Jacobs AK, Faxon DP. Retroperfusion and PTCA. In: Topol E, ed. Textbook of Interventional Cardiology. Philadelphia: WB Saunders, 1990:477-95.

28. Gore JM, Weiner BH, Benotti JR, et al. Preliminary experience with synchronized coronary sinus retroperfusion in humans. Circulation 1986; 74:381-8.

29. Barnett CR, Touchon CT. Use of coronary venous retroperfusion in acute myocardial ischemia. Cardiovasc Rev Rep 1990;10:64-5.

30. Barnett CR, Warren SG, Namay DN, Vaghan AN, Touchon RC. Coro­nary sinus retroperfusion in unstable angina and myocardial infarction (abstr). Proc Adv Med Instr 1990;1:12.

31. Ganz W, Tamuruk K, Marcus HJ, Donoso R, Yashida S, Swan HJC. Measurement of coronary sinus blood How by continuous thermodilution in man. Circulation 1971;44:181-94.

32. Bing RT, Hammond MN, Handelsman JC, et aI. Measurement of coro­nary blood How, oxygen consumption and efficiency of the left ventricle in man. Am Heart J 1949;38:1-24.

33. Faxon DP, Jacobs AK, Kellet MA, McSweeney SM, Coats WD, Ryan TJ. Coronary sinus occlusion pressure and its relation to intracardiac pressure. Am J CardioI1985;56:457-60.

34. Beat! KJ, Serruys PW, deFeyter P, Van den Brand M, Verdouw PD, Hugenholtz PG. Hemodynamic observations during percutaneous trans­luminal coronary angioplasty in the presence of synchronised diastolic coronary sinus retroperfusion. Br Heart J 1988;59: 159-67.

35. Meesman M, Karageuezian HS, Ino T, et aI. Selective perfusion of ischemic myocardium during coronary venous retroinjection: a study of causative role of venoarterial and venoventricular pressure gradients. JAm Coll CardioI1987;10:887-97.

36. Lotan C, Milgalter E, Gotsman M. Dissection of the left main coronary artery: a complication of PTCA to the anterior descending artery. Clin CardioI1988;11:120-1.

37. Vlietstra RE. Management of acute occlusion after percutaneous trans­luminal coronary angioplasty. Eur Heart J 1989;IO(suppl H):H-101-3.

38. Connor AR, Vlietstra RE, Schaff HV, Piehler JM, Pluth JR. Effect of clinical status after failed angioplasty on surgical morbidity and mortality (abstr). J Am Coll Cardiol 1987;9:88A.

39. Detre KM, Holmes DR, Holubkov R, and Registry Investigators. Con­sequences of occlusion during PTCA: the 1985-1986 NHLBI PTCA Registry (abstr). J Am Coll CardioI1989;13:230A.

40. Tokioka H, Miyazaki A, Fung P, et aI. Effect of intracoronary infusion of arterial blood or Fluosol-DA 20% on regional myocardial metabolism and function during brief coronary occlusions. Circulation 1987;75:473-81.

41. Taggart P, Sutton P, John R, Hayward R, Swanton H. The epicardial electrogram: a monophonic action potential recordings. Br Heart J 1989;62:342-52.