Comparison of Monoplane and Biplane Assessment of Regional left Ventricular wall Mot&n After

Thrombolytic Therapy for Acute Myocardial Infarction

DAVID R. HOLMES, Jr., MD, ALFRED A. BOVE, DALE G. GEHRING, BS, JAMES H. CHESEBRO,

and HUGH C. SMITH,

Regional left ventricular (LV) function was assessed using serial biplane orthogonal LV angiograms re- corded before and after repetfusion therapy for acute myocardial infarction (AMI) in 24 patients. Improvement in regional LV function was seen in only 5 patients after reperfu&n therapy when only the right anterior oblique view was analyzed; im- provement in regional wall motion was seen in 14 when biplane views were analyzed. Biplane analy- sis was particularly important in the 12 patients with right coronary artery occlusion, amoq whom the

MD, RICK A. NISHIMURA, MD, MD, ROBERT M. OWEN, BA, MD

ri9ht anterior oblique view showed improvement in only 1 patient but the left anterior oblique view showed improverr?ent In 6 patients (p <0.05). Bi- plane analysis is more sensitive than monoplane right ante&r oblique analysis alone for detecting improvement in LV function after repetfusion thera- py for AM. However, both views are complementa- ry, adding information about regional function not re- vealed by either view alone.

(Am J Cardiol 1967;59:793-797)

A lthough improved mortality and morbidity rates have been documented with reperfusion therapy dur- ing acute myocardial infarction (AMI), quantification of myocardial salvage is difficult.1-8 Because the hall- marks of coronary artery disease are regional ischemia and localized contraction abnormalities, reliance on global left ventricular (LV) function may be insuffi- cient. Also, reliance on only 1 angle of view for left ventriculography may underestimate the effects of re- perfusion on LV function. We assessed LV function before and after reperfusion for AM1 using serial bi- plane orthogonal LV angiograms, and evaluated the independent contributions of right anterior oblique and left anterior oblique views for assessing changes in regional LV function.

From the Division of Cardiovascular Diseases and Internal Medicine, the Department of Physiology and Biophysics, and the Section of Information Processing and Systems, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Manuscript re- ceived August 4, 1986; revised manuscript received November 24,1986, accepted November 25.1986.

Address for reprints: David R. Holmes, Jr., MD, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.

Methods Patients: The population consisted of 24 patients (20

men, 4 women) sequentially enrolled in the Multi- center Thrombolysis in Myocardial Infarction Trial, which evaluated the relative efficacy of tissue plas- minogen activator and streptokinase during AMI.g In all patients baseline LV and coronary angiograms were recorded immediately before administration of the thrombolytic agent and 10 days after attempted reperfusion. The specific projection angles for coro- nary and LV angiograms recorded at the initial study were reproduced at the lo-day study.

Left ventricular angiography and function: Bi- plane %-mm tine left ventriculograms, 30” right ante- rior oblique and 60’ left anterior oblique views, were recorded at a rate of 30 frames/s; biplane video re- cordings using progressive scanning were obtained si- multaneously.1° Hemiaxial views were not used. An- giovisP (45 ml) was injected (13 to 15 ml/s) with a multiple sidehole catheter placed retrogradely into the left ventricle. The second or third beat after contrast injection was analyzed; no post-ventricular premature beats were used.

Representative end-systolic and end-diastolic vid- eo frames at baseline and follow-up studies were de-

794 ASSESSMENT OF REGIONAL WALL MOTION

fined in both the 30° right anterior oblique and 60’ left anterior oblique views as the smallest and largest LV volumes. These video frames were displayed for im- age analysis in a computer image-processing system (General Electric DF.3009 and IPDC).

The apex and the midportion of the aortic valve were identified in each view at end systole and end diastole, thus defining the long axis of the left ventri- cle? The endocardial surface was then scanned with a computer-generated trackball video cursor to create an outline of the LV cavity. This method allows visual confirmation that the traced outline coincides with the ventricular silhouette. The generated coordinates are transferred in real time to the computer and stored on a magnetic disk. After tracing the long axis and endo- cardial outline, the X-Y coordinate pairs are used to determine 50 diameters perpendicular to the long axis in the right and left anterior oblique views (Fig. 1). Simpson’s rule was used to calculate ventricular vol- ume and biplane ejection fraction.11

For regional analysis, the biplane images of the left ventricle were separated into 9 regions, as defined by the Coronary Artery Surgery Study.12 Segments 1 through 5 are represented in the right anterior oblique view and segments 6 through 9 in the left anterior oblique view [Fig. 1). The hemisegment percent short- ening was defined as [(end-diastolic length - end- systolic length)/end-diastolic length]. Mean shorten- ing for a region was calculated as the average shortening of all hemisegments in that region.

Fourteen normal patients evaluated for chest pain of uncertain origin who had normal coronary arteries served as control subjects. Average hemisegment shortening for each segment was defined as average hemisegment shortening in the control subjects minus hemisegment shortening in the patients divided by the normal standard deviation for that specific segment.”

8

RAO LAO

FlGURE 1. Division of left ventricle Into 5 right anterior oblique (RAO) segments (1, anterobasak 2, anterolateral; 3, apical; 4, dtaphragmatlc; 5, posterolateral) and 4 left anterior oblique (LAO) segments (8, basal septal; 7, apical septal; 8, posterolateral; 9, lateral). End-systolic and end-dlastollc segments were computer- drawn for each Coronary Artery Surgery Study segment. The heml- segments were divided so that each of regions 1 through 5 In the right anterior oblique view encompassed 20 hemlsegments and each of regions 8 through 9 in Ihe left anterior oblique view encom- passed 25 hemisegments. These segments (end-systole, ///; end- diastole, \\\) are drawn for segments 2 and 8.

Improvement was defined as ati increase in avetage hemisegment shortening of 0.5 standard deviation or more from normal in the regional LV segment before and 10 days after reperfusion..

Previous studies evaluated the relation of the arte- rial segment, electrocardiographic documentation of AM1 location and ‘regional wall motion abnormali- ties.13-16 We related arterial blood supply and the re- gional wall segment to correlate changes with the oc- cluded artery. With the scheme selected, the right coronary artery is represented in the right anterior oblique view by segments 4 and 5 and in the left anteri- or oblique view by segments 8 and 9. The left anterior descending or diagonal coronary artery is represented in the right anteri.or oblique view by segments 2 and 3 and in the left anterior oblique view by segments 6 and 7. The left circumflex coronary artery is represented in the right anterior oblique view by segments 4 and 5 and the left anterior oblique view by segments 8 and 9 (Fig. 1).

Percent improvement in LV function before and after thrombolytic therapy was compared by chi- square analysis. To compare the proportion of patients who showed improvement in the left anterior oblique views with that in patients who showed improvement in the right anterior oblique views, a l-tailed sign test was used.

Results AM1 was anterior in 11 and inferior in 13 patients.

The infarct-related artery was the right coronary ar- tery in 12 patients, left anterior descending or a large diagonal coronary artery in 11 and left circumflex cor- onary artery in 1 patient. Reperfukion was successful in 17 of the 24 patients (71%). After successful reperfu- sion, the mean visual estimated luminal diameter re- sidual stenosis was 85%. In 7 patientsl29%], the artery remained occluded.

Of the 17 patients with successful reperftiion, .I1 (64%) showed improvement in regional LV function [Fig. 2). In contrast, of the 7 patients with unsuccessful reperfusion, 3 (43%) showed improvement in regional LV function (p = 0.39).

There was a trend toward more improved LV re- gional function with early successful reperfusion, but it did not reach statistical significance. Of the 5 patients with successful reperfusion within 4 hours of pain on-

improved 11 3 Regional

left ventricular 64% 43%

function No improvement 6 4

n=17 n=7

Yes No

Reperfusion FIGURE 2. Ftelatlon between Improvement In reglonal left ventrlcu- lar function and outcome of reperfuslon (computer analysis).

April 1, 1987 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 59 795

set, 4 (SO%] had regional wall improvement in a LV segment supplied by the infarct-related artery. Of the 7 patients with unsuccessful reperfusion, 3 (43%) had improvement in regional LV function (p = 0.6). Of the 12 with successful reperfusion more than 4 hours after symptom onset, 7 (58%) had improved regional LV function.

The changes in LV regional wall function were compared in the 30° right anterior oblique view, 60’ left anterior oblique view and both views (Fig. 3). Of 11 patients with left anterior descending artery occlusion, improvement in regional LV function was seen in 4 in the right anterior oblique view alone, in 5 in the left anterior oblique view alone and in 7 when both views were assessed.

In the 12 patients with right coronary occlusions, the left anterior oblique view documented improve- ment in regional LV function in 6 and the right anterior oblique view showed improvement in only 1 patient (p <O.O5) (Fig. 3). When both views were analyzed, 6 of these patients showed improved regional LV function.

In the patient with occlusion of the left circumflex artery, only the left anterior oblique view showed re- gional wall improvement [Fig. 3).

The importance of biplane assessment of regional and global LV function in the 24 patients is shown in Figure 3. When the right anterior oblique view was used alone, only 5 patients showed improvement, but when biplane wall motion analysis was included, 14 showed regional LV improvement. Two patients showed improvement in a right anterior oblique seg- ment but not in a left anterior oblique segment: 9 pa- tients showed improvement in left anterior oblique segments but not right anterior oblique segments (1 circumflex, 3 left anterior descending or diagonal and 5 right coronary arteries]. Three patients showed im- provement in at least 1 segment in both right and left anterior oblique views.

Ejection fraction increased by more than 5% in 9 of 17 patients with reperfusion: 4 of 8 with anterior AMI and left anterior descending reperfusion, 4 of 8 with inferior AM1 and right coronary reperfusion and the patient with inferior AM1 and left circumflex artery reperfusion (Table I).

An example of regional LV function immediately before and 10 days after reperfusion for inferior AM1 is shown in Figure 4 and Table II. Before treatment, the anterior wall moved well. The inferior wall showed significant hypokinesia in right anterior oblique segments 4 and 5 in the distribution of the infarct-related artery. The right anterior oblique view showed relatively little change from baseline to the lo- day study; significant hypokinesia of segments 4 and 5 remained. In contrast, the left anterior oblique view showed obvious improvement in systolic function; se- vere hypokinesia was noted before reperfusion, but function was relatively normal at 10 days.

Discussion Coronary disease produces heterogeneous effects

on LV function so that reliance on global measures of ventricular function alone may be inadequate.17J8

TABLE I Relation Between Ejection Fraction and Reperfusion in 24 Patients

Location of Occlusion

Successful reperfusion

increased EF Increased (>5%)

EF, successful reperfusion

LAD Right

(n = 11) (n = 12)

8 0

5 4 418 41.5

LC

(n = 1)

1

1 111

EF = ejection fraction; LAD = left anterior descending artery; LC = left circumflex artery: n = number of patients; Right = right coronary artery.

AM1 exaggerates the problem of assessment of LV function because some non-infarct-related segments may be hyperkinetic and partially cancel the effects of the hypokinetic infarcted areas.lgJO Early studies documented the importance of analyzing biplane an- giograms in patients with chronic coronary artery dis- ease.18J1 The left anterior oblique view offers informa- tion about the septal and posterolateral walls, which are not seen when only the right anterior oblique view is used. This view is particularly important with cir- cumflex coronary artery occlusions. Despite these is- sues, there is debate as to the need for analysis of biplane regional LV function for gauging the effect of thrombolytic therapy during AMLz2

Assessment of regional LV function is complex. Al- though several mathematical and projection models can be used for both tine and radionuclide LV angiog- raphy, 7~18~*g~23~24 none is ideal. We use a centerline method and measure percent hemisegment shortening from the centerline. The centerline and hemisegments are recorded in both views separately for end diastole and end systole. With this method, systolic movement of the whole heart is accounted for by allowing the long axis to act as a moving reference system to adjust

0 I

Patients, no.

3 6 9 12 I I I I

RAO LAO

Combination

RAO LAO

Combination

Improvement

FIGURE 3. Change in regional left ventricular functlon In right anterl- or oblique (RAO) view, left anterior obllque (LAO) view and both views (computer analysis). When only the RAO view was used, 5 patients (21%) showed Improvement In reglonal LV function; when biplane views were used, 14 (55%) showed Improvement. Clrc = circumflex coronary artery; LAD = left anterlor descending coro- nary artery; RCA = right coronary artery.

796 ASSESSMENT OF REGIONAL WALL MOTION

TABLE II Calculated Data for Patient with lnferlor InfarctIon’

Segment

Time of Study 1 2 3 4 5 6 7 0 9

Average % Hemisegment Shortening

Baseline 34 28 25 26 22 14 6 24 0 IO-day 50 34 28 23 14 51 40 41 34

Deviation from Normalt

Baseline -0.62 -0.86 -1.72 -1.38 -0.53 -0.88 -1.82 -1.75 -2.35 1 O-day 0.15 -0.56 -1.35 -1.59 -1.05 1.25 0.48 -0.31 -0.15

l Angiograms of this patient are shown in Figure 4. TAverage hemisegment shortening for each segment was defined as the amount of deviation from normal ex-

pressed in standard deviations. The deviation for the specific segment was the average hemlsegment shortening obtained in normal controls minus the hemisegment shortening In the patient divfded by the normal standard devl- ation for that specific segment. As defined, improvement was an increase in average hemisegment shortening of 10.5 standard deviation from normal. In this patlent. improvement was noted In segments 6 through 9.

for motion of the heart and of the patient, which are independent of the actual shortening of the myocardi- al fibers. Also, to adjust for the effects of global ven- tricular function on regional function, we relate aver- age hemisegment shortening in patients to that in normal control subjects.

In addition to concerns about the mathematical for- mula selected for analysis, it may be difficult to relate the site of arterial occlusion to a specific LV segment. There is a correlation between the location of coronary arterial narrowing and regional wall motion abnor- malities in patients with AMI.13-16J8 The Coronary Ar- tery Surgery Study uses LV segments corresponding to frequently involved arterial segments: inferior or lat- eral segments in both views are related to the right or

the circumflex coronary artery, and anterior and api- cal septal segments are related to a left anterior de- scending or diagonal coronary artery. Variability and overlap, however, do exist, so sequential changes in regional function must be analyzed in the same seg- ments from baseline to follow-up.

In this study, recovery of LV function after attempt- ed reperfusion was variable. With earlier reperfusion, a larger percentage of patients had improvement in regional LV function. With reperfusion within 4 hours of pain onset, 4 of 5 patients showed improvement. A similar proportion with successful reperfusion of either the right or the left anterior descending coro- nary artery showed improvement in global ejection fraction.

FIGURE 4. Left ventricular anglograms of a patient with lnferlor acute myocardlal Infarction. A, right anterior obllque (RAO) vlews. Leff, dlastole (fop) and systole (hffom) lmmedlately before thrombolytlc therapy. Anterior wall motion ls well preserved, but lnferlor wall segments (segments 4 and 5) show marked hypoklnesla. R&h& dlastole (fop) and systole (boffom) 10 days after successful WdUSiOn.

Regional wall motlon showed llttle change from the Initial to the follow-up study. B, left anterlor oblique (LAO) views. reff, dlastole (fop) and systole (bottom) lmmedlately before thrombolytic therapy. R/g/M, dlastole (fop) and systole (boffom) 10 days after successful reperfuslon. Systolic function markedly Improved, severe hypoklnesla was noted before reperfuslon, but function was relatively normal at 10 days.

April I,1987 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 59 797

Use of regional wall analysis with biplane projec- tion views added clinically important information. Only 5 of the 24 patients showed regional LV improve- ment when the right anterior oblique view was used, but 14 showed regional LV improvement when bi- plane wall motion analysis was used. Of 12 patients with right coronary artery occlusion, only 1 patient showed improvement in the right anterior oblique view. In contrast, 6 of 12 patients showed improve- ment in 1 or 2 regional segments in the left anterior oblique view. For the 24 patients as a whole, more showed improvement in the left anterior oblique seg- ments. Both views, however, offered complementary information. Two patients showed improvement only in a right anterior oblique segment and 9 showed im- provement in the left anterior oblique segments alone.

In the 1 patient with circumflex coronary artery occlusion, improvement was identified in only the left anterior oblique view. The posterolateral wall seg- ment supplied by the circumflex artery may be missed with only monoplane right anterior oblique LV views20 Assessment of LV function with the mono- plane right anterior oblique view may therefore un- derestimate improvement in function after lytic therapy.

Limitations: Biplane projection analysis has limita- tions. Standard left anterior oblique views foreshorten septal segments. Angulated views minimize this prob- lem. However, with left anterior oblique cranial an- gulation, the lower apical septal and posterolateral segments may be superimposed on the diaphragm, making analysis difficult. In addition, normal values for wall motion are not available for such projection views. Finally, specific LV wall segments and the arte- rial branches involved may show variability. Serial studies of regional function in the same patient mini- mize this problem.

Irrespective of these limitations, biplane projection views for assessment of regional LV wall motion add significant information to that obtained by monoplane right anterior oblique assessment. The projections are complementary, adding information about regional function not obtained by either view alone.

References 1. Kennedy IW, Ritchie JL, Davis KB, Fritz Jr<. Western Washington Random- ized Trial of Intracoronary Streptokinase in Acute Myocordial Infarction. N Engl 1 Med 1983;309:1477-1482. 2. Simoons ML, Serruys PW, van den Brand M, Biir F, de Zwaan C, Res J. Verheugt FWA, Krauss XH, Remme WJ, Vermeer F, Lubsen 1. Improved survival after early thrombolysis in acute myocardial infarction: a rando- mised trial by the Interuniversity Cardiology Institute in The Netherlands. Lancet 1985;2:578-581.

3. Rentrop P. Smith H, Painter L. Holt J. Changes in left ventricular ejection fraction after intracoronary thrombolytic therapy: results of the Registry of the European Society of Cardiology. Circulation 1983;68:suppl I:I-55-I-59. 4. Leiboff RH. Katz. Rj, Wasserman AG. Bren GB, Schwartz H, Varghese Pj, Ross AM. A randomized, angiographically controlled trial of intracoronary streptokinase in acute myocardial infarction. Am 1 Cardiol 1984;53:404-407. 5. Anderson IL. Marshall HW, Bray BE, Lutz JR, Frederick PR. Yanowitz FG, Datz FL, Klausner SC, Hagan AD. A randomized trial of introcoronary strep- tokinase in the treatment of acute myocardial infarction. N Engl r Med 1983;308:1312-1318, 6. Khaia F, Walton IA Ir, Brvmer FT. Lo E, Osterberger L, O’Neill WW, Colfer HT, W&s R, Lee T..K&ian ?, Goldberg AD, Pitt B, Goldstein S. Intracoronary fibrinolytic therapy in acute myocardial infarction. Report of a prospective randomized trial. N Engl 1 Med 1983;308:1305-1311, 7. Sheehan FH, Mathey DG, Schafer 1, Dodge HT. Balson EL. Factors that determine recovery of left ventricular function after thrombolysis in patients with acute myocardial infarction. Circulation 1985;71:1121-1128. 6. Koren G. Weiss AT, Hasin Y, Appelbaum D, Welber S, Rozenman Y, Lotan C, Mosseri M, Sapoznikov D, Luria MH. Gotsman MS. Prevention of myocar- dial domage in acute myocardial ischemia by early treatment with intrave- nous streptokinase. N Engl [ Med 1985;313:1384-1389. 9. TIM1 Study Group. The Thrombolysis in Myocardial Infarction (TIMI) trial: phase I findings. N Engl 1 Med 1985:312:932-936. 10. Holmes DR, Bove AA, Wondrow MA, Gray JE. Video x-ray progressive scanning: new technique for decreasing x-ray exposure without decreasing image quality during cardiac catheterization. Mayo Clin Proc 1986:61:321- 326. 11. Bove AA, Kreulen TH, Spann JF. Computer analysis of left ventricular dynamic geometry in man. Am 1 Cardiol 1978;41:1239-1248. 12. Principal Investigators of CASS and Their Associates. The National Heart, Lung, and Blood Institute Coronary Artery Surgery Study (CA%]. Circulation 1981;63:suppl 1:&1-I-81. 13. Williams RA, Cohn PF, Vokonas PS, Young E, Herman MV, Gorlin R. Electrocardiographic, arteriographic and ventriculographic correlations in transmural myocardial infarction. Am f Cardiol 1973;31:595-599. 14. Herman MV. Heinle RA, Klein MD, Gorlin R. Localized disorders in myocardial contraction: asynergy and its role in congestive heart failure. N Engl r Med 1967;277:222-232. 15. Savage RM, Wagner GS, Ideker RE, Podolsky SA, Hackel DB. Correlation of postmortem anatomic findings with electrocardiographic changes in pa- tients with myocardial infarction: retrospective study of patients with typical anterior and posterior infarcts. Circulation 1977;55:279-285. 16. Bodenheimer MM, Banka VS, Helfant RH. Q waves and ventricular asynergy: predictive value and hemodynamic significance of anatomic local- ization. Am 1 Cardiol 1975;35:615-618. 17. Cohn PF, Herman MV, Gorlin R. Ventricular dysfunction in coronary artery disease. Am \ Cardiol 1974;33:307-310. lg. Herman MV, Gorlin R. In vivo angiagraphic pathoanatomy of the acute syndromes of coronary heart disease. Trans Assoc Am Physicians 1972;85: 231-246. 19. Sheehan FH, Mathey DG, Schafer 1, Krebber HJ, Dodge HT. Effect of interventions in salvaging left ventricular function in acute infarction: a study of introcoronary streptokinase. Am r Cardiol 1983;52:431-438. 20. Stack RS, Phillips HR III, Grierson DS, Behar VS. Kong Y. Peter RH, Swain JL. Greenfield JC Jr. Functional improvement of jeopardized myocar- dium following introcoronary streptokinase infusion in acute myocardial in- farction. J Clin invest 1983;72:84-95. 21. Cohn PF, Gorlin R, Adams DF, Chahine RA, Vokonas PS, Herman MV. Comparison of biplane and single plane left ventriculograms in patients with coronary artery disease. Am 1 Cardiol 1974;33:1-6. 22. Sheehan FH, Schafer J, Dodge HT. Wygant J, Mitten S, Bolson EL. RAO vs LAO regional wall motion changes in post-thrombolysis patients (abstr]. Cir- culation i985;72:suppl rrr:rrr-22. 23. Ohsuzu F, Boucher CA, Newell JB. Yasuda T, Gold HK, Leinbach RC. McKusick KA, Okada RD. Rosenthal S, Pohost GM, Strauss HW. Relation of segmental wall motion to global left ventricular function in acute myocardial infarction. Am J Cardiol 1983;51:1275-1281. 24. Neumann P. Schicha H, Tebbe U, Kreuzer H, Emrich D. Analysis of left ventricular regional motility: a comparison of different methods. Eur J Nucl Med 1984;9:205-208.