Relation between left ventricular shape and exercise capacity in patients with left ventricular dysfunction

JACC Vol . 22, No . 3 September 1993 :750-7

751

FART FAILU

Relation B ween Lef' Ventricul7i Patients With Left V - . lcula ,

e

MARC D . TISCHLER, MD, FACC, JOELY51 NUGGEL, RN, DAVID T . BOROWSKI, MD, MARTIN M . LEWINTER, MD, FACC Vererlueei

Objectives. The aim of this study was to identify dynamic predictors of exercise duration in patients with systolic fell ventricular dysfunction and to test the hypothesis that left ventricular shape is an independent determinant of exercise duration in these patients . Background. Measurements of left ventricular volumes and ejecflon fraction at rest do not predict exercise capacity in patients with systolic left ventricular dysfunction . Left ventricular shape at rest has been reported to be an indc g-eadent determinant of exercise duraticai in these patients. The signot sauce of alterations ins left ventricular shape that occur during dynamic exercise has suit been investigated . Methods . Twenty-one paitients with a documented ejection fraction <40% performed sylaptora-has ked graded upright bicycle exercise with simultaneous quantitative two-dimenslonal echorailiography. End-diastolic Wattle, end-systolic volume, stroke volume, ejection fraction and spkaericity index were measured at rest and peak exercise .

Results. Eleven patients exercised beyond stage 11 (6 min, SO W), averaging 0 .9 ± 1 .9 min ; 10 patients were unable go complete stage 11, averaging 4 .9 ± 0 .9 min . No patient developed clinscal evidence of ischemia during the exercise period . Of the echocardiographic variables considered, only end-systolic and end-diastolic sphericity indexes at peak exercise (r = 0 .809 and 0 .711, respectively) and the change in end-systolic sphericity index during exercise (r = 0 .697) were strongly correlated with exercisei duration . Conclusions . Conventional descriptors of left ventricular function are poor predictors of exercise capacity . Dynamic changes in heart shape correlate strongly with exercise duration and may be important determinants of exercise capacity in patients with systolic left ventricular dysfunction . (J Am Coll Cordial 1993,22:751-7)

Measurements of left ventricular volumes and ejection fraction at rest do not predict maximal exercise duration in patients with congestive heart failure due to systolic left ventricular dysfunction (1-5) . Several investigators have described abnormalities in left ventricular mechanics resulting from alterations in left ventricular shape . In dilated cardiomyopathy, for example, left ventricular chamber shape is more spheric than normal and end-systolic wall stress is correspondingly greater than normal (6) . However, whether alterations in left ventricular shape, especially those occurring during exercise, are determinants of exercise capacity has not been investigated . Lamas et A . (7) performed biplane ventriculography in 40 patients with congestive heart failure on the basis of anterior myocardial infarction . Of the angiographic variables examined, the calculated end-diastolic sphericity index was the only independent predictor of exercise duration . This interesting observation has not been reproduced and is based on angiographic imaging performed at rest . The use of exercise echocardiog-

raphy provides the opportunity to examine dynamic changes in left ventricular shape and volumes at rest and during exercise . Accordingly, the present study was designed to identify dynamic predictors of exercise duration in patients with systolic left ventricular dysfunction and to test the hypothesis that left ventricular shape is an important independent determinant of exercise duration in these patients .

From the Cardiology Unit, Medical Center Hospital of Vermont and University of Vermont, Burlington, Vermont . Manuscript received July 29, 1992 ; r,.-vised manuscript received January 14, 1993, accepted March 2, 1993 . Address for correspondence: Marc D. Tischler, MD, Medical Center Hospital of Vermont, McClure 1, Burlington, Vermont 05401 . 01993 by the American College of Cardiology

Methods Study patients. Between September 1991 and August 1992, 28 patients were prospectively identified with an ejection fraction <40% and no significant valvular heart disease or active symptoms of ischemic heart disease . Ejection fract ,47n was determined by radionuclide ventriculography (n = 18), two-dimensional echocardiography (n = 6) or contrast ventriculography (n = 4) . Seven of the 28 patients were excluded from the study because of inadequate echocardiographic windows (n = 1), inconsistent supine and upright images (n = 5) or development of chest pain associated with electrocardiographic (BCC') and echocardio1) . The graphic evidence of ischemia during exercise (n remaining 21 patients (13 men, 8 women) formed the study group . Their age ange was 25 to 76 years (mean ± SD 58 ± 13) . Seven patients were in New York Heart Association 0735-1097193156.OD



752

TISCHLER ET AL. LEFT VENTRICULAR SHAPE AND EXERCISE CAPACIT

JACC Vol . 22, No . 3 Septem er 1993 :751 . .7

Figure 1. Supine (felt) nd upright icycle (right) end-di stolic (R w ve pe k) im ges

demonstr ting preserv tion of long- xis nd short- xis dimensions in p tient with severe left ventricul r dil tion,

function l cl ss 1, six in cl ss II nd eight in cl ss III . Medic tions t the time of testing included digoxin (n = 11), diuretic drugs (n = 12), nitr tes (n = 6), v sodil tors (n = 17), et - drenergic locking gents (n = 4) nd c lcium ch nnel nt gonists (n = 3). C uses of left ventricul r dysfunction included myoc rdi l inf rction (n = 12), idiop thic dil ted c rdiomyop thy (n = 8) nd perip rtum c rdiomyop thy (n = 1) . Eight p tients h d ngiogr phic lly norm l coron ry rteries, seven h d no signific nt perfusion norm lities during stress r dionuclide perfusion im ging nd three h d no ECG evidence of ischemi during st nd rd tre dmill exercise testing performed efore p rticip tion in the present investig tion . Three p tients h d no ngin or other evidence of coron ry rtery dise se ut did not undergo speci lized testing efore p rticip tion in this study . Exercise echoc rdiogr phy. B seline two-dimension l echoc rdiogr phic ex min tions were performed with the p tient supine in the left l ter l decu itus position nd g in while sitting st tion ry on n upright icycle with ph sed rr y ultr sonoscope device (Acuson XP-5) using 2 .5-MHz tr nsducer . Im ges were o t ined in sequenti l f shion from the p r stern l long- xis, p r stern l short- xis, pic l four-ch m er nd pic l two-ch m er views. To minimize foreshortening of the left ventricul r c vity, the tr nsducer w s held slightly l ter l to the pic l impulse to m ximize the tomogr phic pl ne of the left ventricle . The tr nsducer w s then ngul ted nteriorly nd posteriorly in sm ll degrees to record the gre test long- xis length . Those p tients with inconsistent supine nd upright pic l nd short- xis dimensions were excluded from the study (Fig. 1) . Blood pressure w s me sured with cuff sphygmo-

m nometer t the time of the seline echoc rdi-Gr phic ex min tion . P tients then performed symptom-limited gr ded upright icycle ergometry, eginning t work lo d of 25 with 25- increments in work lo d every 3 min . Convention l 12-le d ECG monitoring w s utilized . Blood pressure w s me sured every 2 min . Continuous twodimension l echoc rdiogr phic ex min tions were performed during exercise . D t n lysis . All echoc rdiogr ms were ev lu ted y single experienced echoc rdiogr pher (M . D . T .) without knowledge of the clinic l nd ECG response to exercise nd to ll other clinic l inform tion . Using Microsonics Im geVue orkst tion (Nov Microsonics Inc .) three to five c rdi c cycles were digitized t end-systole (time of sm llest c vity re , Fig. 2) nd end-di stole (R w ve pe k, Fig . 3) . End-systolic, end-di stolic nd stroke volumes nd ejection fr ction were c lcul ted s previously descri ed (8,9) . Left ventricul r volumes were c lcul ted y the short- xis re x length method: LVESV = 5/6 x A, x L, nd LVEDV = 5/6 x A d x Lj,

where LVESV nd LVEDV re left ventricul r end-systolic nd end-di stolic volumes, A, nd Avd re end--iystolic nd end-di stolic c vity re s in which the p pill ry muscles re reg rded s p rt of the left ventricul r c vity nd 1 S nd Ld re left ventricul r end-systolic nd end-di stolic lengths, respectively . Sphericity index w s defined s the left ventricul r long- xis length/left ventricul r di meter r tio (6, 10,11) nd w s determined t oth end-systole nd enddi stole . To determine intr o server v ri ility, linded post hoc n lysis w s performed in 10 r ndomly selected

.tACC Vol . 22, No . 3 Septem er 1993 :751-7

TISCHLER ET AL . LEFT VENTRICULAR SHAPE AND EXERCISE CAPACLr

75 -5

Figure 2 . Short- xis end-systolic im ges recorded t rest (left) nd during pe k exercise (right) in p tient who exercised for 9 min 39 s . Note the reduction in end-systolic circumference nd short- xis dimension t pe k exercise .

studies . The correl tion coefficients for left ventricul r enddi stolic nd end-systolic volumes were r = 0 .982 nd 0 .964, respectively . For the sphericity index t end-systole, r 0 .992 . St tistic l n lysis. All d t re presented s me n v lue t SD . The initi l d t n lysis w s designed to identify simple line r correl tions etween, echoc rdiogr phic v ri-

Figure 3. Apic l four-ch m er enddi stolic im ges recorded t rest (left) nd during pe k exercise (right) in p tient who stopped exercise fter 3 min 12 s ec use of dy pne nd f tigue . Note the sm ll reduction in long- xis length nd incre se in midch m er di meter.

les nd dur tion of exercise . In ddition, differences mong the 10 p tients un le to complete st ge II of the gr ded exercise protocol (Grcup I) nd the I I who successfully completed this st ge (Group II) were ssessed using the Student t test ( ll test st tistics two-t iled) . Differences were considered signific nt if the null hypothesis could e rejected t p < 0 .05 .



754

JACC Vol. 22, No . 3 Septem er 1993 :751-7

TISCHLER ET AL. LEFT VENTRICULAR SHAPE AND EXERCISE CAPACIT

T le 1 . Simple Correl tion Among Test V ri les nd Dur tion of Exercise Correl tion Coefficient (r)

V ri le

0 .271 0 .708 0 .220 0 .374 0 .468 0 .557 0 .066 0.003 0.077 0 .063 0 .187 0 .151 0 .451 0 .809 0 .268 0.711 0.697 0 .571

HR, rest HR, pe k SBP, rest SBP, pe k DBP, rest DBP, pe k ESV, rest ESV, pe k EDV, rest EDV, pe k EF, rest EF, pe k SI„ rest SI„ pe k Sid , rest Sid , pe k Ch nge in Sl, Ch nge in Sid

DBP = di stolic lood prcssurc ; EDV left ventricul r end-di stolic volume ; EF = left ventricul r ejection fr ction ; ESV = left ventricul r end-systolic volume ; HR = he rt r te ; SBP = systolic lood pressure ; Si d = sphericity index t end-di stole; SI, = sphericity index t end-systole.

Results CI ch r cteristics . P tients exercised for me n of 7 ± 2 .5 min (r nge 3 .1 to 11 .4), stopping ec use of f tigue (n = 8), dyspne (n = 6), com in tion of oth symptoms (n = 6) or light-he dedness (n = 1) . No p tient in the study group compl ined of chest discomfort or developed ECG evidence of ischemi . Ten p tients (Group 1) were un le to complete st ge II of the gr ded test (6 min, 50 ), ver ging 4 .9 ± 0.9 min of exercise . The rem ining I 1 p tients (Group II) ll completed st ge II, ver ging 8 .9 ± 1 .9 min . The two groups were comp r le in ge (59 ± 8 vs . 56 ± 16 ye rs, p = 0.56). Group I p tients stopped ec use of f tigue (n = 3), dyspne (n = 4), light-he dedness (n = 1) or the com in tion of dyspne nd f tigue (n = 2) . Group II p tients stopped ec use of f tigue (n = 5) or dyspne (n = 2), or oth (n = 4). P tients with prior myoc rdi l inf rction were evenly distri uted etween the two groups, with five such p tients in Group I nd sevem in Gr3 p II . Of the clinic l hemodyn mic v ri les ex mined (he rt r te nd systolic nd di stolic lood pressure), only he rt r te t pe k exercise h d signific nt correl tion with exercise dur tion, with r > 0.60 (r = 0 .708) (T le 1) . Simple correl tions (T le 1). Of the echoc rdiogr phic v ri les considered, only the end-di stolic nd end-systolic sphericity indexes t pe k exercise nd the net ch nge in end-systolic sphericity index during the exercise period exhi ited strong correl tions with dur tion of exercise . No other v ri le h d correl tion coefficient ?0 .60 . Su group T le 2). In comp ring Groups I nd II, there were highly signific nt differences in he rt r te t pe k exercise, oth systolic nd di stolic sphericity in-

T le 2. Su group Comp rison of Individu l Test V ri les Dur tion of Exercise V ri le HR, rest HR, pe k SBP, rest SBP, pe k DBP, rest DBP, pe k ESV, rest ESV, pe k EDV, rest EDV, pe k EF, rest EF, pe k SI„ rest St., pe k Sid, rest Sid, pe k Ch nge in SI, Ch nge in Sid Dt

<6 min 87 116 125 138 67 71 110 109 150 158 29 31 1 .75 1 .58 1 .71 1 .58 -0.17 -0.14

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

15 18 20 28 8 10 35 30 39 42 9 7 0 .24) 0.19 0.17 0.12 0.11 0.13

'6 min

p V lue

97 f 12 136 ± 24 131 ± 21 152 ± 16 75 ± 14 79 ± 13 111 ± 37 112 ± 43 162 ± 57 166 ± 58 31 ± 6 33 ± 7 1 .91 ± 0.28 1 .99 ± 0.26 1 .74 ± 0.21 1 .82 ± 0.22 0 .08 ± 0.18 0 .07 ± 0.18

0.12 0.04 0.49 0.17 0.13 0.13 0.97 0.86 0.61 0 .72 0 .61 0 .48 0 .14 0 .001 0 .78 0 .008 0.001 0.005

re expressed s me n v lue ± SD . A

revi tions s in T

le 1 .

dexes t pe k exercise (Fig . 4 nd 5) nd the ch nge in these two sphericity indexes during the exercise period (Fig . 6 nd 7). There were no other signific nt differences in ny of the other clinic l or echoc rdiogr phic v ri les considered .

Discussion Convention l descriptors of left ventricul r geometry nd function t rest re poor predictors of exercise c p city (1-5) . The present qu ntit tive echoc rdiogr phic study w s designed to ex mine correl tions etween dyn mic descriptors of left ventricul r function nd exercise dur tion nd to test the specific hypothesis th t p tients wi,.h systolic left ventricul r dysfunction who h ve more clnheric left ven-

Figure 4. End-systolic sphericity index t pe k exercise in p tients un le to exercise for 6 min (Group I) nd those le to exercise for >6 min (Group II) .

4 V m C nr' .fi

p

2.50

0 .001 0

2 .00

1 .50

T 0 1 .

•

1 .00 C) 5 U

0 .50 0 .00

w Group I

Group II



JACC Vol . 22, No . 3 Septem er 1993 -751-7

TISCHLER ET AL .

LEFT VENT R11CULAR SHAFT, ANT EXERCISE CAPACIT

0 .50 , p - 0 .006

9-0 .005 030 ! CL

1,50 8

OZO V IN te

Group I

Group 11

11houre 5 . End-di stolic Sphericit index t pe k exercise in the two p tient groups (defined in Fig . 4) .

tricul r sh pe h ve diminished exeteise c p city comp red with p tients with more ellipsoid ch m er geometry . Previous investig tions . Could et l. (12) recognized the import nt rel tion etween ch m er geomeiry nd w il stress in p tients surv,,ving myoc rdi l inf rction . Su sequent studies (13) suggested th t ch nges left ventricul r sh pe re rel ted to underlying left ventricul r dysfunction nd m y precede detect le hemodyn mic norm lities . In p tients with dil ted c rdiomyop thy, for ex mple, more spheric ch m er is ssoci ted with more severely depressed perform nce t rest nd gre ter end-systolic w ll stress (14) . The import nce of ch m er geometry in rel tion to exercise c p city, however, h s een incompletely descri ed . L m s et l . (7) used ipl ne ventriculogr phy to study 40 p tients fter zute nterior myoc rdi l inf rction . P tients were r ndomized to receive pl ce o or c ptopril nd then performed qu rterly symptom-hinited exercise stress tests . The ngiogr phic lly derived end-di stolic sphericity index w s the only independent predictor of exercise dur tion in the pl ce o group . Pl ce o-tre ted p tients with

Figure 6 . Ch nge in end-systolic sphericity index etween rest nd pe k exercise in the two p tient groups (defined in Fig . 4), x

• V

0 .50 P -

•

owl

0 .301

C6 Co

0 .10

0 • 10

-0 .10

T 0

-0.30

2

I I

U1

2 •

-0.50

Group I

Group 11

Group I

Group J1

Figure 7 . Ch nge in end-di stolic sphericity index etween vest nd pe k exercise in the two p tient groups (defined in Fig, 4),

the most norm l sphericity index lso h d the most severe he rt f ilure nd highest specific ctivity sc le scores . These findings, however, h ve not een confirmed . furthermore, ll im ging w s performed t rest t time remote from the ctu l exercise period . In the present investig tion, exercise echoc rdiogr phy w s used to ex mine left ventricul r geometry nd volume during oth rest nd gr ded exercise . Consistem with most previous findings (1-5), none of the convention l rest or dyn mic descriptors of left ventricul r function correl ted signific ntly with exercise dur tion . In contr st to the results of L m s et l . (7), signific nt correl tion etween exercise dur tion nd left ventricul r sh pe t rest w s not found . However, oth the systolic nd di stolic sphericity indexes t pe k exercise nd the ch nge in sphericity during the exercise period were strongly correl ted with exercise dur tion . These results suggest th t dyn mic ch nges in geometry m y e import nt predictors of exercise c p city in p tients with systolic left ventricul r dysfunction . Potenti l mech nisms. Although our study design does not llow specific conclusions s to whether sphericity differences in our p tients re directly rel ted to exercise toler nce or represent second ry effect, there re re sons to elieve th t preserv tion of wore elliptic sh pe might le d to incre sed exercise c p city . In c nine ventricles, the gener tion of restoring forces (due to contr ction to elow equili rium volume) th t ssist filling is closely rel ted to eccentricity (15) . At sm ll end-systolic volumes ( elow the equili rium volume), s occur during exercise, the ventricle ecomes progressively more ellipsoid s it gener tes more restoring forces . In nim l models nd in hum n he rts, left ventricul r systolic "twist," or torsion l motion, h s een thought to e one mech nism for stor ge of potenti l energy for restoring forces (16,17) . This type of three-dimension l deform tion is intim tely rel ted to the sh pe of the ventricle . End-systolic sh pe norm lities m y therefore reflect ility to gener te restoring forces. Furthermore, reduced gener tion of restoring forces lso decre ses energy losses



756

JACC Vol. 22, No. 3 Septem er 1993 :751-7

TISCHLER ET AL . LEFT VENTRICULAR SHAPE AND EXERCISE CAPACIT

cross the mitr l v lve (15) . Thus, it is possi le th t preserv tion of more elliptic ch m er sh pe improves ventricul r perform nce nd exercise c p city y virtue of effects th t optimize filling . The dyn mic ch nges in left ventricul r geometry th t occur during exercise h ve complex determin nts such s v rious peripher l nd neurohumor l responses th t influence lo d nd contr ctility (18-23), myoc rdi l perfusion nd intrinsic myoc rdi l norm lities such s the extent nd severity of fi rosis . Ultim tely, however, the sh pe t end-systole is determined y interrel tionships etween the intrinsic contr ctile perform nce of the ventricle nd its lo d . Bec use of the strong correl tion etween he rt r te t pe k exercise nd exercise dur tion, we suggest th t possi le expl n tion for the sh pe differences t end-systole is th t they reflect ltered neurohumor l end-org n sensitivity or other peripher l neurohumor l effects th t result in dverse ventricul r lo ding during exercise, or oth . Thus, neurohumor l ch nges th t ffect he rt r te might e linked to other neurohumor l ch nges th t influence lo ding . The minim l or nonexistent differences in left ventricul r ejection fr ction t rest nd during exercise suggest th t the sh pe differences were not rel ted to differences in "glo l" left ventricul r contr ctile perform nce, s does the diversity of our p tient cohort, which included persons with ischemic nd nonischemic dise se . Fin lly, lthough it is not possi le to completely exclude ischemi during exercise, our study design m kes this n unlikely f ctor . Limit tions . Two-dimension l echoc rdiogr phy m y provide t ngenti l cuts r ther th n true me sures of the left ventricul r long xis (24) . By recording only high qu lity digit l im ges, m king repe ted me surements of ch m er dimensions nd excluding those p tients whose supine nd upright dimensions were not the s me, we minimized potenti l confounding y this phenomenon . As mentioned previously, the p tient cohort w s heterogeneous with respect to the origin of left ventricul r dysfunction. The r tion le for including the two groups of p tients (with nd without ischemi ) is th t they eh ve identic lly with respect to sphericity . The heterogeneity of the p tient up supports unifying hypothesis th t left ventricul r sh pe is n independent determin nt of exercise toler nce irrespective of the c use of left ventricul r dysfunction. Fin lly, while the cohort size nd overl p of d t with reg rd to seline sh pe m ke sphericity index t ny single point in v me poor predictor of individu l response, the group responses with reg rd to ch nges in end-systolic nd end-di stolic sh pe were highly uniform . C . Twenty-one p tients with documented left ventricul r ejection fr ction <40 performed symptomlimited gr ded upright icycle ergometry . Qu ntit tive ech io hic n lysis w s performed on digit lly cquired two-dimension l echoc rdiogr phic im ges o t ined t rest nd during exercise . None of the convention l descriptors of left ventricul r function (th t is, end-systolic volume nd ejection fr ction) correl ted with exercise dur -

tion . Left ventricul r sh pe during pe k exercise nd the ch nge in left ventricul r sh pe th t occurred during exercise correl ted strongly with exercise dur tion . The underlying mech nism of sh pe differences nd how they rel te to determin nts of exercise toler nce merit future investig tion .

References 1 . Leicr CV, Binkley PF, St rling RC, Huss-R ndolph P. Disp rity etween improvement in left ventricul r function nd ch nges in clinic l st tus nd exercise c p city during chronic enoximine ther py . Am He rt J 1989; 117:1092-8 . 2 . Meiler SE, Ashton JJ, Moesch erger ML, Unverfcrth DV, Leier CV . An n lysis of the determin nts of exercise c p city in congestive he rt f ilure. Am He rt J. 1987 ;113 :1207-17 . FR . 3 . Iliggen oth m MB, Morris KG, Cohn Eli, Colem n RE, Co Detrr:vin nts of v ri le exercise perform nce mong p tients with severe left ventricul r dysfunction. Am J C rdio11983 ;51 :52-60 . 4 . Fruncios JA, P rk M, Levine TB . L ck of correl tion etween exercise c p city nd indexes of re-Ang left ventricul r perform nce in he rt f ilure . Am J C rdiol 1981 ;47 :33-9. 5 . Szl chic J, M ssie BM, Kr mer BL, Topic N, Tu u J . Correl tes nd prognostic implic tion of exercise c p city in chronic congestive he rt f ilure . Ant J C rdiol 1985 ;55 :1037-42. 6. L skey !V, St. John Sutton M, Zeevi G, Hirshfeld J, Reichek N . Left ventricul r mech nics in dil ted c rdiomyop thy . Am J C rdiol 1984;54 : 620-5. 7. L m s GA, V ughn DE, P risi AF, Pfeffer MA . Effects of left ventricul r sh pe nd c ptopril ther py on exercise c p city fter cute nterior myoc rdi l inf rction. Am J C rdiol 1989 ;63 :1167-73 . 8. St . John Sutton M, Pl ppert T, Spiegel A, et l . E rly postoper tive ch nges in left ventricul r ch m er size, rchitecture, nd function in ortic stenosis nd ortic regurgit tion nd their rel tion to intr oper tive ch nges in fterlo d: prospective two-dimension l echoc rdiogr phic study. Circul tion 1 7 ;76:77-89. 9 . Schiller NB, Acquetell 11, Ports TA, et l . Left ventricul r volume from p ired ipl ne two-dimension l echoc rdiogr phy . Circul tion 1979;60: 547-55 . 10. Dougl s P, Morrow R, foli A ., Reichek N . Left ventricul r sh pe, fterlo d nd surviv l in idiop thic dil ted c rdiomyop thy . J Am Coll C rdiol 1989;13 :311--5 . 11 . D'Cruz LA, A oul tt H, Kill m H, Br dley A, H nd R . Qu ntit tive two-dimension l echoc rdiogr phic ssessment of left ventricul r sh pe in ischemic he rt dise se. J Chn Ultr sound 1989 ;17 :569-72. 12. Gould KL, Lipscom K, H milton H , Kennedy J . Rel tion of left ventricul r sh pe, function, nd w ll stress in m n. Am J C rdiol 1974 ;34 :627-34. 13. Tomlinson C . Left ventricul r geometry nd function in experiment l models of he rt f ilure. C n J C rdiol 1987 ;3 :305-10. 14. Borow K, L ng RM, Neum nn A, C rroll JD, R jfer SL Physiologic mech nisms governing hemodyn mic responses to positive inotropic ther py in dil ted c rdiomyop thy. Circul tion 1 ;77:625-37 . 15. Nikolic S. ellin E, D m M, Pgj ro 0, Fr y er R. . Rel tionship etween di stolic sh pe (eccentricity) nd p ssive el tic properties in c nine left ventricle. Am J Physiol 1590 ;259:H457-63 . 16. Ingels N, H nsen D, D ughters G, Stinson E, Alderm n E, Mill DC . Rel tion etween longitudin l, circumferenti l, nd o lique shortening nd torsion l deform tion in the left ventricle of the tr nspl nted hum n he rt. Circ Res 1989 ;64:915-27. 17. H sen D, D ughters G, Alderm n E, Ingels N, Miller DC . Torsion l deform tion of the left ventricul r midw ll in hum n he rts with intr myoc rdi l m rkers : region l heterogeneity nd sensitivity to the inotropic effects of rupt r te ch nges. Circ Res 1 8 ;62 :941-52 . 18. Freem n GL, Colston JT . Role of ventriculov scul r coupling in c rdi c response to incre sed contr ctility in closed-chest dogs . J Clin Invest 1990 ;86:1278-84. 19. Drexler H, Hiroi M, Riede U, B nh rdt U, Meinertz T, Just H . Skelet l muscle lood flow, met olism nd morphology in chronic congestive



JAUC Vol . 22, No . 3 Septem er 1993 :751-7

he rt f ilure nd effects of short- nd long-term ngiotensin-convening enzyme inhi ition . Am J C rdiol 1988 ;62 :82E-5E . 20 . Cohn K Levine TB, Oliv ri MT, et l . Pl sm norepinephrinc s guide to prognosis in p tients with chronic congestive he rt f ilure . N Engl J Med 1984 ;31 :819-23 . 21 . Thom s JA, M rks B11 . Pl sm norepinephrine in congestive he rt f ilure . Am J C rdiol 1978;41 :233-43, 22 . Buigoiv MR, Cins urg R, Minohe , et l . Decre sed c icch6 mine

iSCHLER ETAL . LEFT VENTRICULAR SHAPE AND EXERCISE CAPACIT

7 .57

sensifivit nd density i3 f iling r, m-t h he rs . N E gi I Mcd 1982307 :205-11 . 23 . Fowler MB, L ser J, Hopkins GL, Minu e . Bristow M . Assessment of the et - d,onergic receptor p thw y in the int ct f iling hurr y he rt . Circul tion 1996;74:1290--302 . 24 . Ey el R, S .hwcizcr P, L m eriz H, et l . F-choventniculogr physimult neuls n lysis of two-dinne ,sion l echoc rdiogr phy nd cineverotriculogr phy . Circul tion 1983 ;67 :205-15 .