Mechanism and significance of a decrease in ejection fraction during exercise in patients with coronary artery disease and left ventricular dysfunction at rest

88

JACC Vol 3. No 1 January 1984 88-97

Mechanism and Significance of a Decrease in Ejection Fraction During Exercise in Patients With Coronary Artery Disease and Left Ventricular Dysfunction at Rest MICHAEL B. HIGGINBOTHAM, MB, R. EDWARD COLEMAN, MD, ROBERT H. JONES, MD, FREDERICK R. COBB, MD Durham . North Carolina

The purpose of this study was to determine whether an exercise-induced decrease in ejection fraction in patients with coronary artery disease and left ventricular dysfunction at rest represents ischemia or the nonspecific response of a compromised left ventricle to exercise stress. Accordingly, radionuclide ejection fraction responses of 246 patients with coronary artery disease and an ejection fraction at rest of less than 0.50 were compared with those of a "nonischemic" control group of 48 patients with idiopathic dilated cardiomyopathy and a similar degree of ventricular dysfunction. The significance of the ejection fraction response in the group with coronary artery disease was further examined by relating it to the angiographic extent of coronary artery disease, severity of angina, incidence of chest pain and electrocardiographic ST segment depression during exercise and longterm prognosis. The ejection fraction decreased by ~ 0.01 and ~ 0.05 during exercise in 48 and 28%, respectively, of the patients with coronary artery disease compared with only 8 and 2 %, respectively, of the patients with cardiomyopathy. When exercise was limited by fatigue at a

The demonstration of reversible myocardial ischenua In patients With coronary artery disease IS Important In the selecnon of appropnate therapy When left ventncular funcnon at rest is normal, Ischemia often can be detected dunng exercise stress by the demonstration of electrocardiographic

From the Cardiovascular Divrsionof the Department of Medicine. and the Department of Radiology, Durham Veterans Adrmrnstranon Medical Center and Duke Umversny Medical Center, Durham. North Carolina This study was supported m part by Research Grant HLl7670 from the National Heart. Lung. and Blood Institute. Bethesda. Maryland Dr Hrggmbotham received support from the National Heart Foundation of Austraha, Woden, AustralianCapital Terntory, Australia Manuscnptreceived March 22, 1983, revised manuscnpt received August 10, 1983, accepted August 16, 1983 Address for repnnts Fredenck R Cobb, MD, Divisionof Cardiology (lIlA), Durham Veterans Adrrnmstranon Medical Center, 508 Fulton Street, Durham, North Carolma 27705 © 1984 by the Amencan College of Cardiology

submaximal heart rate, the ejection fraction decreased in 25% of the patients with coronary artery disease but in none of the patients with cardiomyopathy. Patients with coronary artery disease whose ejection fraction decreased during exercise had a significantly higher incidence of three vesseldisease, exercise-induced chest pain or ST depression and late mortality than did patients whoseejection fraction did not decrease. These relations were confirmed equally in subgroups of patients with moderate (ejection fraction 0.30 to 0.49) and severe (ejection fraction < 0.30) left ventricular dysfunction. Thus, in patients with coronary artery disease and left ventricular dysfunction at rest, a decrease in ejection fraction during exercise is more likely to indicate ischemia than a nonspecific left ventricular response to exercise stress. In the individual patient, a decrease of 0.05 or greater, or a decrease during submaximal exercise, appears to be highly specificfor ischemia. A decrease in ejection fraction identifies a subgroup of patients with a high prevalence of multivessel coronary artery disease and a high risk of death during long-term follow-up on medical therapy.

changes (1), changes In radionuchde Indexes of global and regional left ventncular function (2-5), and thallium perfUSIOn defects (6,7) Specific diagnostic cntena have been established by comparing the responses of patients With coronary artery disease With those of normal control subjects. In patients who have depressed left ventncular funcnon at rest , the diagnosis of Ischemia may be more difficult for the following reasons First, because previous myocardial mfarcuon usually results In electrocardiographic changes and abnormalities of wall motion and myocardial perfusion at rest, It IS difficult to Interpret additional changes that may occur during exercise (8-11) Second, because the ejection fracnon at rest IS an Important determinant of the ejection fraction response to exercise In patients With coronary artery disease (10), the diagnostic cntena used In patients With normal left ventncular function at rest may not 0735-1097/84/$3 00

JACC Vol , No I January 1984 88-97

89

HIGGINBOTHAM ET AL EXERCISE RfWON5E WITH CARDIAC DYSFUNCTION I\T REST

be appropnate m patients with abnormal function at rest Third. myocardial mfarction reduces the amount of myocardium available for mamtammg global left ventncular function dunng exercise, a ventncle that I~ abnormal at rest may be unable to respond to the added demands ot exercise. and the ejection fraction may decrease even m the absence of rscherma The concept that exercise stres-, may reveal myocardial dysfunction not apparent at rest has been discussed by other mvestigators concerned with valvular ( 12) and myocardial (13) disorders The present study was performed to clanfy the mechamsm and significance of the ejection fraction re~pon~e to exercise m patients with coronary artery disease and left ventncular dysfunction at rest It was reasoned that the followmg observations would support the hypothesr- that a decrease m ejection fraction represents ischerma rather than a nonspecific response to exercise stress I) Patients who have left ventncular dysfunction but a very low probabihty of developmg ischemia do not expenence a decrease m ejection fraction dunng exercise 2) A decrease m ejection fraction dunng exercise I~ related to chrucal and electrocardiographic mdicators of ischemia. angiographrc extent of coronary artery disease and subsequent mortality To examme these hypotheses, the results of rest and exercise first pass radionuchde angiography performed in 246 consecunve patients with angrographically proven coronary artery disease and reduced left ventncular function were compared with those of 48 patients with idiopathic dilated cardiomyopathy In addition, the ejection fracnon responses m patients with coronary artery disease were correlated with cluneal and electrocardiographic indexes of ischenua. extent of anatomic coronary artery disease and subsequentmortality

Methods Study patients. A consecutive series of 246 patients with coronary artery disease and left ventricular dysfunction as defined by a radionuchde ejection fraction of levs than o 50 was studied Two hundred twenty of these patients were men and 26 were women. ages ranged from 25 to 72 years (mean 51) Coronary artery disease was confirmed in all patients by cardiac cathetenzation, It showed 75% or greater dIameter narrowmg of three vessels In 119 patlenb. of two vessels m 71 patlent~ and of one ve~"el In 56 patients There wa~ a hIstory of chest pam m 236 patlent<, and ot myocardIal mfarctlOn m 187 patlent~ The electrocardIogram at rest showed Q waves of 0 04 second or greater In 177 patIents Of the 246 patIent". 31 (I )I'k) had neIther a documented hIstory of myocardIal mfarctlOn nor Q waves of 0 04 second or greater on the electrocardIOgram Control group To compare patIent~ WIth coronary artery disease With patIents who had comparable left ventncular dy~functlon but a very low probabIlIty of developmg I~ch­ emIa dunng exerCise, a group of 48 patIents WIth IdlO-

pathic dilated cardiomyopathy was abo studied In 16 of these patients cardiomyopathy had been confirmed by cardiac cathetenzation and in 32 patients the diagnosis was based on chmcal and electrocardrographrc cntena alone Of the 16 patients who underwent cathetenzation, the procedure was performed to exclude valvular heart disease In 4, In the remammg 12 patients. coronary angiography and left ventriculography were performed a" avsociated procedures dunng catheterization for myocardial biopsy The decrsion to perform cardiac cathetenzanon wa-, made by the patient'< physician and was not mfluenced by the results of the radionuclide angiogram All 32 patients diagnosed chrncally presented for radionuchde angiography WIth a provisional dragnovi-, of congestive cardiomyopathy All had a history of congestive heart tailure Without chest pam, and none had electrocardiographic Q waves of 0 04 second or greater Because the purpose of this part of the study was to obtam a "model" of left ventncular dysfunction Without ischerrua to act a" a control group for the patients WIth coronary artery disease. the group With cardiomyopathy was further "elected on the baSI'> of their chmcal and electrocardiographic re
Table I. Baselme Charactcnvnc- of 48 Pauent-, With Nomschermc Cardiomyopathy Cardiac Catheten7 allon 16)

(n =

AgelH) Se\ Abnormal ECG STT

LYH LBBB RBBB REF

No Cardiac Cathetenzauon 32)

(n =

46 ± II 10 M 6 F 15

50 ± 14

~

17

3 3

5

I 029 ± 0 II

25 M 7 F 27 3 2 029 ± 0 II

ECG = e!ectrocardlllgram, F = female LYH = lett ventnLUlar hypertrophy LBBB and RBBB = lett and nght bundle branch bloc\., re'pectlvely REF = re,t eJecllon tracllon M = male, ST/T = ST-T wave ahnormaltlle,

90

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

study of 36 patients who had left ventncular function determmed by both techniques, the mcan difference (± standard deviation) between ejection fraction measurements WID>, respectively, 0 05 ± 0 02 at rest/correlation coefficient Ir] = 0.94) and 0 06 ± 004 dunng exercise (r = 091), the change 10 ejection fraction from rest to exercise vaned by o 04 ± 0 03 Despite minor differences 10 the exercise protocol used for the two techniques, similar exercise end pomts were used for all studies Beta-adrenergic block109 drugs were discontmued for at Icast 24 hours before the exercise test 10 207 of the 246 patients, the rernammg 39 took propranolol on the day of the study at the request of their personal physician Exercise protocol After the acquismon of the radionuclide angiogram at rest, each subject performed upright bicycle exercise sittmg on an isokmcnc bicycle ergometer For first pass studies, exercise work load commencedat 200 kpm/rrun and was mcreased by 100 kpm/mm each minute For equihbnum studies, exercise work load was mcreased by 150 kpm/mm every 3 mmutes to allow adequate time for data acquisruon A standard 12 lead electrocardiogram was recorded before the exercise study Precordial leads Y5 and Y6 were momtored contmuously, and the standard limb leads and leads Y-+ to Y6 were recorded dunng each mmute of exercise Blood pressure was measured by cuff manometry and recorded at rest and every 2 rnmutes dunng exercise Exercise was continued until one of the followmg end pomts was reached moderate chest pam, honzontal or downslopmg ST segment depression of I mY or more (2 mY With ST changes at rest), a heart rate 85% or greater than the age-predicted maximum (target heart rate) or severe fatigue Radtonucltde acquisition and data analysts First pass radronuchdeangiography was performed 10 the antenor projection usmg methods previously descnbed (14,15) For each radionuchde acquistion, 10 to IS mCI of technetium99m pertechnetate was dissolved 10 less than I cc normal saline solution and was flushed 10 as a bolus With 10 to 20 cc normal saline solution Using a multicrystal gammacamera equipped With a I inch (2 54 ern) parallel hole colhrnater, counts were acquired at 25 ms mtervals for I mmute Data were processed USIng the computer and software of the Baird-Atomic System 77 Data from three to SIX individual beats produced an average or representative cycle Ejection fraction was calculated from background-corrected end-diastolic and end-systolic counts Ejection fraction = End-dIastolIc counts - End-systolIc counts End-diastolic counts Wall motion was assessed from a stanc display of the supenmposed end-diastolic and end-systolic penmeters denved from the representative cycle The static Image was

JACC Vol 3, No I January 1984 88-97

divided mto three segments corresponding to the antenor, apical and mfenor walls of the left ventncle Wall motion was graded for each segment from 0 to 6 where 0 = normal, I = less than Y2 wall hypokmetrc, 2 = greater than Y2 wall hypokmetic, 3 = less than Y2 wall akmetic, 4 = greater than Y2 wall akinetic, 5 = less than Y2 wall dyskmetic and 6 = greater than Y2 wall dyskmetic To distmguish between global and regional left ventncular dysfunction, asynergy of wall motion was defined as a difference of at least two grades between two segments Asynergy that was present only dunng exercise and resultmg from a decrease 10 regional wall motion was termed exercise-induced asynergy Equiltbrtum radionuchde angiography was performed after 10 VIVO labelmg of red blood cells WIth 30 mCI of technetlUm-99m A SIngle-crystal gamma-camera, equipped WIth a high sensinvity collimator, was interfaced WIth a Medical Data Systems A2 computer for data collection and subsequent analysis End-diastohc and end-systolic regions of interest were Identified us109 a sermautomated edge detection algorithm, ejection fraction was calculated from background corrected end-diastolic and end-systolic count measurements Wall motion was assessed 10 the left antenor obhque projection, septal, apical and posterolateral regions were graded as hypokmetic. akmetic or dyskmetic 10 a manner Similar to that used for the first pass studies Follow-up, Of the 246 patients With coronary artery dISease, 96 underwentcoronary artery bypass surgery All were operated on wrthm 4 months (56% withm 4 weeks. 89% within 8 weeks) and thus did not contribute significantly to long-term follow-up The remammg ISO patients were treated medically and constituted the medical follow-up group Information concernmg the mortahty of all patients had been obtained within 3 months of the study, either by letter or telephone call to the patient or by contactmg the patient's personal physician Follow-up duration was less than 10 months 10 49 patients, 10 to 19 months 10 21 patients, 20 to 29 months 10 29 patients and 30 months or more 10 51 patients Statistical analysis, Ejection fraction changes from rest to exercise were analyzed by paired t tests The relation between ejection fraction changes dunng exercise and at rest 10 the group WIth coronary artery disease was exarmned by lmear regression analysis Unpaired t tests were used for all mtergroup compansons The overall distnbution of angma class and vessel disease 10 the patient groups was compared by chi-square analysis A log rank test was used to compare survival curves for patients With different ejecnon fraction responses PatientsWith coronary arterydisease and an ejection fraction at rest of 0 30 or greater were termed subgroup I and those With an ejection fraction at rest of less than 0 30 were termed subgroup II All analyses were apphed to the total group of patients WIth coronary artery disease and both subgroups All group values are expressed as mean ± standard deviation

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

JACC Vol 3, No I January 1984 88-97

fracuon response to exercise /1l patients with coronary artery disease IS shown 111 Figure 2 Lmear regression analysis

Results Exercise response. Maximal work load achieved was similar for patients with coronary artery disease (638 ± 198 kpm/mm) and patients With cardiomyopathy (560 ± 188 kpm/rmn) Both groups exercised to sirrular maximal heart rates (135 ± 27 and 137 ± 21 beats/nun, respecnvely) Patients with cardiomyopathy proven by cardiac cathetenzation achieved a SImilar maximal work load (604 ± 243 kpm/rmn) and heart rate (\36 ± 18 beats/mm) to that of patients who did not have cardiac cathetenzanon (548 ± 159 kpm/rnm and 137 ± 22 beats/mm, respectively) Of the 246 patients with coronary artery disease. 162 achieved 85% of the predicted maximal heart rate compared with 25 of 48 patients with cardiomyopathy Seventyrune patients with coronary artery disease expenenced chest pam and 65 had diagnostic ST segment depression dunng exercise, one patient with cardiomyopathy confirmed by cardiac cathetenzation complamed of chest pam, but none had new or additional ST segment changes Exercise was lumted by fatigue m 52 patients with coronary artery disease and m all patients with cardiomyopathy Ejection fraction response to exercise. lndrvidualejection fracnon responses to exercise are shown m FIgure I Ejection fraction did not change from rest (0 34 ± 0 19) to exercise (0 35 ± 0 12) in patients with coronary artery disease, but mcreased from 0 29 ± 0 II to 0 36 ± 0 15 (probability [p] < 0 000 I) m patients with cardiomyopathy The ejection fracnon decreased dunng exercise m 117(48%) of the patients with coronary artery disease compared with only 4 (8%) of the patients with cardiomyopathy A decrease in ejection fraction of 0 05 or greater was seen m 69 panents (48%) with coronary artery disease compared with only I patient (2%) with cardiomyopathy Although the 39 patients tested while receivmg propranolol achieved lower heart rates than did patients not takmg propranolol (119 ± 28 compared with 138 ± 34 beats/nun), the ejection fraction decreased m a Similar proportion in the two groups (16 [41 %] of 39, and 101 [49%] of 207, respectively)

revealed a significant negative correlation between ejection fraction at rest and the change from rest to exercise (r = - 0 26), the ejection fraction decreased durmg exercise m 100 (57%) of 179 patients whose ejecnon fraction at rest was greater than 0 30 (subgroup I) , compared With 17 (24%) of 67 patients whose ejection fraction at rest was less than o 30 (subgroup II) The relation between exercise end point and ejection fraction response IS Illustrated /1l Table 2 Exercise was

stopped by fatigue before 85% of the predicted maximal heart rate was achieved ("madequate" end pomt) in 52 (20%) of the 246 patients With coronary artery disease and in 23 (50%) of the 48 patients With cardiomyopathy For all patients, mcludmg both coronary artery disease subgroups, the ejection fraction decreased less often when the exercise end pomt was inadequate than when It was adequate However, differences m exercise end pomt clearly did not account for the difference m ejection fraction response between coronary artery disease and cardiomyopathy groups when the end pomt was adequate the ejection fracnon decreased in 104 (53%) of 194 patients With coronary artery disease compared WIth 4 (16%) of 25 patients WIth cardiomyopathy (p < 0 00 I) When exercise was madequate the ejection fraction decreased in 13 (25%) of 52 patients With coronary disease but m none of the patients With cardiomyopathy (p < 0001) Wall motion. Of the 246 patients With coronary artery disease, abnormal wall motion was seen m 239 patients at rest and in 240 patients dunng exercise Asynergy, that IS, a segmental wall motion abnormality. was present in 79 patients at rest and m 106 patients dunng exercise Exerciseinduced asynergy occurred in 30 patients and was accornparned by a decrease in ejection fracnon m 20 Abnormal wall motion was noted at rest III all 48 patients WIth cardiomvopathv and during exercise In 45 Left ven-

tncular a~ynergy was seen m 12patients at rest and 9 patients dunng exercise Exercise-induced asynergy was not seen in the group With cardiomyopathy

The effect of the ejection fraction at rest on the ejection

z

o

CARDIOMYOPATHY

CO RO NARY ARTE RYDIS EASE

(0' 48 )

(0 ' 2461

v
SU BGROUP II:SUBG ROU PI I

'"

u-

I I

-=l -..

I

o

o o

~

~ 40

/

I

I

I-

I

t

,'

, .,'1',

r / •

";' e

I

.

••

:, .... it': .•

:;.

'J;".;i1I

w

<.f)

G

o- cc rd.cc cotheten zc non

~ 20

>< w

20

40

60

91

80

RESTI NG EJECTION

20

40

FRACTION

60

80

/ Figure 1. Ejection fraction at rest and dunng exercise In 48 patients WIth nomschermc cardiomyopathy (left panel) and 246 patients With coro nary artery disea se and ejection fraction at rest less than 0 50 (right panel ) POIntsappeanng below the hne of idennty Indicate a decrease In ejection fracnon from rest to exerci se, pomts on or above the lme represent no change or an Increase In ejection fraction Two coronary artery disease subgroups are separated for descnptive purposes accordmg to ejection fraction at rest above or below 0 30

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

92

r'- 26, pcO 001

I

tI

SUBGROUP][

+ 20

I-

U

I I

I

+10

« Z

. . .:... ...l;: -; :':

0

0

"

• " ' .. I·~ ';'

uUJ

,

'"

•

... ": :.1 l:::··.::e:.: . . • ",: ,;.,... I;' .:: : :" .:' ..... if

. ':'

;::::

."

.1

:

--,


I.-

:

,

u, '"

UJ

•

I

z

0

SUBGROUP I

- 10

1

,',,",,'

"" -:

i: ',,'. ..

I

- 20

c:er:

••••

I

'I'

"

:

I I

"

I

10

20

30

40

50

RESTING EJECTION FRACTION

Figure 2, Relation between ejection fraction atrest and the change

m ejection fraction from rest to exercise (Il ejection fraction) m patients with coronary artery disease and ejection fraction at rest less than 0 50 An ejection fraction at rest of 0 30, used for separanon of the two subgroups, IS intermediate between the highest and lowest values seen

Relation of ejection fraction change to severity of angina in patients with coronary artery disease. The dIStnbunon of subjects according to the seventy of angina was Similar for the total group and each subgroup (Table 3) In the total group, the proportion of patients m each functional class was the same for patients whose ejection fraction decreased as for patients whose ejection fraction increased or did not change (Fig 3) Separate analysis of subgroups I and II revealed a lower incidence of class II angina m panents of subgroup II whose ejection fraction decreased durmg exercise (p < 001) No other significant differences were seen Relation of ejection fraction change to exercise response in patients with coronary artery disease, Chest pam was seen m 32% and ST depression m 26% of the

JACC Vol 3 No I January 1984 88-97

patients with coronary artery disease (Table 4) The mCIdence of chest pam and ST segment depression tended to be lower m subgroup II than m subgroup I, but this difference was not significant There was a strong association between exercise indexes of Ischemia and the directron of ejection fraction change dunng exercise (Fig 4) Both chest pam and ST depression occurred m a larger proportion of patients whose ejection fraction decreased than m those whose ejection fracnon did not decrease, these differences were seen m the total group and in each subgroup Among the 117 patients whose ejection fraction decreased dunng exercise, the decrease was accornpamed by either chest pam or ST segment depression in 79 patients (68%) The data m Table 4 and Figure 4 also Illustratethe higher incidence of a decrease m ejection fraction dunng exercise compared with chest pam and ST depression The ejection fraction decreased m 117 patients and chest pam occurred m 79 patients and ST depression m 65 patients Whereas 63 (54%) of the 117 patients whose ejection fraction decreased had no ST depressron, only 12 (18%) of the 65 patients with ST depression had no decrease in ejection fraction Relation of ejection fraction change to anatomic extent of coronary artery disease, Table 5 lists the number of diseased vessels in the total group and in each subgroup The distnbution was comparable, approximately 50% of each group having three vessel coronary artery disease Figure 5 shows the directional change m ejection fraction dunng exercise as a function of the number of diseased vessels Patients whose ejection fraction decreased had a higher prevalence of three vessel disease and a lower prevalence of one vessel disease than did patients whose ejection fracnon increased or remained unchanged The prevalence of two vessel disease tended to be lower m patients whose ejection fraction decreased None of the 14 patients who had single vessel coronary artery disease and ejection fracnon at rest of less than 0 30 (subgroup II) showed a further reduction m ejection fraction dunng exercise

Table 2. Proportion of Patients With a Decrease m Ejection Fraction During Exercise, Related to Quality of End Pomt

Group

Cardiomyopathy

Coronary artery disease Subgroup I (REF 0 30 to 0 49) (n = 178) Subgroup II (REF < 0 30) (n = 68)

Adequate" End Pomt

1P < 0001

4/25

1041194~ 89/148 15/46

Inadequate* End Pomt 0123

1P < 0001

13152~

Adequate vs Inadequate (p value) 0035

< 0001

11/31

0030

2121

0022

*Exerclse was defined as adequate when limited by chest pam or ST segment depression, or when 85% of the predicted maximal heart rate was achieved Exercise was madequate when limited by fatigue before achievement of the target heart rate REF = ejecnon fraction at rest

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

JACC Vol 3, No I January 198488-97

Table 3. Distnbution of Patients With Coronary Artery Disease and Ejection Fracnon at Rest Less Than 0 50, Accordmg to Angina Class NYHA

Angina

cause of the small number of patients In this subgroup (p = 0 10)

Class

II

III

IV

Total patients (n = 246)

9% (21)

23% (57)

22% (53)

46% (115)

Subgroup I

8% (14)

25% (44)

19% (33)

48% (87)

10% (7)

24% (16)

24% (16)

42% (29)

(REF 0 30 (n =

to 0 49) 178)

Subgroup II <

(REF

(n

0 30) = 68)

Numbers mparentheses represent numbers of patients NYHA York Heart Associanon, other abbreviations as before

93

=

New

Relation of ejection fraction change to mortality in patients with coronary artery disease. Twenty-six (17%) of the 150 patients who were treated medically died dunng the follow-up penod. Figures 6 and 7 relate survival to ejection fraction at rest and dunng exercise The degree of left ventncular dysfunction at rest was highly predictive of mortality; 10% of the patients m subgroup I died compared with 33% of those in subgroup II (p < 0 001) The ejection fraction response to exercise was related to survival In the total group (p = 0002) Fourteen deaths

(23%) occurred in the 61 patients whose ejection fraction decreased dunng exercise compared with 12 (13%) of 89 patients whose ejection fraction increased or remained the same In subgroup I, mortality was high among patients whose ejection fraction decreased (9 l20%] of 50 patients), by companson, only 1 (2%) of 51 patients whose ejection fraction mcreased or was unchanged died dunng the followup penod (p = 0 008) Although ejection fraction at rest was the major detenmnant of the high mortality In subgroup II, 5 (45%) of 11 patients whose ejection fraction decreased died compared with 11 (31%) of 38 patients whose ejection fraction Increased, this difference was not significant be-

Mortality differences between these subgroups cannot be explained by differences m age, mean age was between 50 and 52 years

Discussion The uutial question addressed m this study was whether the presence of ventncular dysfunction at rest predisposes patients to a further decrease m function dunng exercise stress, with a consequent reduction in ejection fraction ThIS question IS potentially Important not only for patients with coronary artery disease, but also for the assessment and follow-up of patients with valvular heart disease For example, Borer et al (12) and Peter and Jones (16) showed that the ejection fraction may decrease dunng exercise in many patients with chronic aortic regurgitation and no eVIdence of left ventncular dysfunction at rest, It has been suggested that this may represent mild left ventncular dysfunction which IS unmasked by the added stress of exercise However, these Investigators have not excluded the pOSSIbihty that a decrease in ejection fraction in patients with valvular regurgitation may reflect myocardial Ischemia or changes m ventncular loading rather than bemg a marker of left ventncular dysfunction Role of ischemia in the ejection fraction decrease during exercise. If a decrease in ejection fraction dunng exercise IS related to myocardial dysfunction Itself, one might expect It to be independent of commonly accepted indexes of myocardial Ischemia, and generally proportional to the degree of left ventncular dysfunction The findings of the present study suggest that the opposite IS true In patients with coronary artery disease, a decrease in ejection fraction became less frequent as left ventncular dysfunction became more severe, and It was related to cluneal and electrocardiographic Indexes of left ventncular Ischemia and multivessel coronary artery disease Furthermore, It occurred Infrequently In pattents WIth cardiomyopathy, who had left

g~ } (NYHA)

Angina Class

80

~m ~nz:

TOTAL GROUP

~

tj 60

z

UJ

~

L!.J

40

"" Cl.

20 EF. N' 117

SUBGROUP I: RESTING EF 30 - 49

SUBGROUP II: RESTING EF < 30

Figure 3. Relation between New York Heart AsSOCiatIOn (N Y H A ) functional class and the ejecnon fraction response to exercise m patients with coronary artery disease and ejection fraction at rest less than 0 50 The prevalence of each angina class was compared between patients whose ejection fraction decreased dunng exercise (EF t) and patients whose ejection fraction increased or did not change (EF i -) Compansons were made for the total group. and for each subgroup separately Sigrnficant differences are indicated *p < 0 02 versus EF t N = number of patients

94

HIGGIN BOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

Table 4. Incidence of Exertional Chest Pam and ST Depression 10 Patients With Coronary Artery Disease and Ejection Fracuon at Rest Less Than 0 50

JACC Vol 3. No I

January 1984 88-97

Tab le 5. Distnbunon of Patients With Coronary Artery DIsease and Ejection Fraction at Rest Less Than 0 50. According to the Number of Diseased Vessels

Exercise Response

Number of Diseased Vessels

Chest Pam

ST Depression

Total patients (n = 246 )

32 % (79)

26 % (65 )

Total patient s 246 )

Subgroup 1 (REF 0 30 to 0 49 )

34 % (6 1)

29 % (52 )

Subgroup 1 (REF 0 30 to 0 49)

(n

=

178 )

=

25 % (17)

19% (13)

68)

Figure 4. Relation between the ejection fracnon response to exercise and exercise-Induced chest pam (open columns) or ST depression (hatc hed columns), or both. 10 patients With coronary artery disease and ejection fraction at rest less than 0 50 Companson groups and abbreviatrons are Similar to Ihose 10 FIgure 3 Symbols denote sigmficant differences compared with EF ! (* p < 0 02, **p < 00001. t p < 001 , ttp < 0 05)

o Chest Poon r:lI ST Depressron

80 TOTAL GROU P Ql

c

60

Ql

"'C

vc

40

~

20

=

3

23% (56)

29 % (7 1)

48 % (119 )

24'* (42)

31 % (56 )

45 % (80)

21 % (14)

23 % (16)

56 % (38)

(7 8)

68)

Abbrevianons as before

ventncular dysfunction but a very low probabihty of developing ischemia The Increase In ejection fraction In most patients WIth ca rdiomyopathy suggests that abnormal but noruscherruc myocardium usually responds to the surnuh accompanying exercise With an Increase In contracnhty Therefore, the decrease In ejection fraction dunng exercise that was seen In 117 of 246 patients with coronary artery disease was more likely to represent myocardial ischemia than a nonspecific decompensation dunng stress Criteria for di a gn osing isch emia d u r ing exe rcise. The ability to detect a subgroup of panents With a hig h hkehhood of having reversible myocardial ischerrua may be Important In selectmg appropnate populations for the study of medical or surgical Interventions However, the cluneal utihty of a test depends on ItS rehabihty In mdividual patients The interpretanon of an indrvidual response IS always more dif-

100

=

Subgroup II (REF < 0 30) (n

Abbrevrations as before

u

=

(n

Subgroup II (REF < 0 30) (n

(n

2

SUBGROU PI RE STINGEf 30-4 9

SUBGRO UP IT : RESTI NG Ef < 30

ficult than the separation of subgroups, partly because of the techmcal hrmtanons Inherent 10 a Single measurement and partl y because of the vanabrhty of biologic responses However, two relanvely specific cntena for diagnosing ischemia can be Inferred from the responses In our gro up WIth cardiomyopathy The first cntenon IS a decrease In ejection fract ion dunng subrnaximal level s of exercise which was seen In 25 % of the patients WIth coronary artery disease but In no patient WIth cardiomyopathy, the second IS a decrease In ejection fraction by 0 05 or greater WhICh was seen In 28 % of the group WIth coronary disease but In only 2%, or 1 of the 48 patients WIth cardiomyopathy Although the aSSOCIatIOn between ejection fracnon change , chest pain and ST segment depression dunng exercise and the number of diseased vessels strongly supported the Idea of an ischermc rather than noms chemic mechamsm for a decrease In ejecnon fraction In the total group With coronary artery disease, the weak correlations did not Improve the mterpretation of mdividual responses Coronary a rtery d isease versus cardi omyopathy . A selected group of patients With cardiomyopathy was used In the present stud y specifically as a model of left ventncular dy sfunction In the absence of ischerrua We thought that this would be the best available control group for determining the relation between rscherrna and the ejection fraction response However, the companson of patients With cardiomyopathy and patients With coronary artery disease may not be entirely valid , SInce It assumes that a ventncle With segmental dysfunction should, In the absence of Ischemia , respond the same as a ventncle WIth diffuse dysfunction Although It IS possible that differences In ventncular disease may result In different responses to exercise despite the same overall degree of dysfunction, this appears

unlikely EF. N '1 17

Because the patients WIth cardiomyopathy who did not undergo cardiac cathetenzation were selected to exclude either coronary artery disease or exercise-ind uced ischemia,

JACC Vol 3, No I January 198488-97

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

01} ~~

100 !!.

80

Number of Diseased Vessels

SUBGROUP II: RESTI NG EF < 30

SUBGROUP I: RESTING EF 30 - 49

TOTAL GROUP

w

u

z w

t

w

C£:

o,

t

40

I"'

**

*1""

r

.

I"'

'"

20

I""

Figure 5. Relation between anatomic extent of coronary artery disease and the ejection fraction response to exercise 10 patients with coronary artery disease and ejection fraction at rest less than 0 50 Companson groups and abbreviations are similar to those 10 Figures 3 and 4 Significant differences versus EF t are indicated (*p < 0 02, **p <

oom, tp

< 001)

'---

EF. N' 117

EF'"

EF.

129

101

EF.

EF'" 77

16

our data do not necessanly charactenze the ejection fraction responses In a representative population with cardiomyopathy It IS entirely possible that some patients excluded from the study because of chest pam or ST segment depression dunng exercise In fact had normal coronary artenes, and may have demonstrated a decrease In ejection fraction Clearly, however, the response In such patIents could not have been Interpreted as nomschermc Studies aimed at descnbmg the response In cardiomyopathy would require a prospectIve design In which a consecutIve group of patients underwent cardiac cathetenzatIon and radionuchde

angiography Clinical and electrocardiographic features ofischemia. These features are seen Infrequently In patIents with coronary artery disease and left ventncular dysfunction dunng

Figure 7. Survival curves for 150 medically treated patients with coronary artery disease and ejection fraction at rest less than 0 50, related to the direction of ejection fraction change from rest to exercise A poorer survival IS seen for patients whose ejection dunng exercise than for those 10 whom fraction decreased (EF It mcreased or did not decrease (EF i ~) This IS confirmed for each coronary artery disease subgroup

t)

CORONARY ARTERY DISEASE TOTAL GROUP

100

89

EFf-.

-~1 ,~-~~~~~-"'-... ------~2...--1._"\. 3-L

80

'--li----'-----,_____ EF. 24

--

T Figure 6. Cumulative survival

150 medically treated patients

In

with coronary artery disease and ejection fraction at rest less than related to the ejecuon fraction response to exercise Data are represented as In Figure I

o 50,

CORONARY ARTERY DISEASE FOLLOW-UP Z

f=

SUBGROUP .IT: SUBGROUP I

.

u


Ct:

LL

./

60

I I.

Z

0 f= uw w

V)

U

Ct:

w >< w

/ ~"'o':'1{

20

::.4' ••• J

0'1 • )( .0,0

/

0/

/

/

/

/ /

/

/

SUBGROUP II 100

38

/

y ..,'

.( ·I:J°· . ~

/0 /'.

/

/

.;to..

I ' /-.:

...,w 40

/

.,Y

JI":~.

/

/

I

.' ALIVE 0 ' DEAD

BO

.'

P

60

i

T

1

10 20 30 40 50 60

70 80 90

RESTING EJECTION FRACTION

!

10

20

_

-----14------

60

0 80

95

,

30

FOLLOW-UP (MONTHS)

;0

0.002

96

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

exercise stress testing (8,9) In the present study, chest pam occurred m only 32%, and ST segment depression m only 26% of patients dunng exercise, the mcidence was slightly lower m the subgroup of patients with an ejection fraction at rest of less than 0 30 The finding that neither chest pam nor ST segment depression was observed m 38 of 117 patients whose ejection fraction decreased dunng exercise appears to support earlier observations (17, 18) that changes m left ventncular function are more sensitive to Ischemia than are either chmcal symptoms or the electrocardiogram The development of new or additional wall monon abnormalities dunng exercise also was uncommon m patients with coronary artery disease and left ventncular dysfunction (30 of 246 patients) This low incidence of exercise-induced asynergy may have resulted from the assessment of wall motion In a single (antenor) projection, and from the difficulty of resolving a change in wall motion In the presence of extensive preexistmg wall motion abnormalities The responses In our patients wrth coronary artery disease show clearly that Ischemia cannot be predicted by coronary anatomy alone The cluneal , electrocardiographic and radionuclide responses were heterogeneous whether patients had one, two or three vessel disease, some patients with one vessel disease appeared to develop Ischemia and some with three vessel disease did not Prognostic impUcations. Left ventncular function at rest IS a strong and independent predictor of mortality in patients with coronary artery disease (19-21) In a study mvolvmg 1,214 medically treated patients with coronary artery disease, Hams et al (21) demonstrated a 2 year mortality rate of 5% In patients with normal left ventncular funcnon, 15% in patients with moderate and 40% in patients with severe left ventncular dysfunction In the present study, subgroups I and II correspond to moderate and severe left ventncular dysfunction, mortality rate at 2 years in these subgroups was approximately to and 33%, respectIvely The ejection fraction response to exercise provided sigmficant prognostic mformation In the present study, especially m patients with an ejection fraction at rest of 0 30 or greater (subgroup I), only 1 (2%) of 51 patients m this subgroup whose ejection fraction increased or remained unchanged dunng exercise died dunng the follow-up penod compared with 9 (18 %) of 50 patients whose ejection fraction decreased As ejection fraction at rest was a strong predictor of the high mortality in subgroup II, the Influence of the ejection fraction response to exercise was of less prognostic Importance, a decrease In ejection fraction dunng exercise Identified patients in subgroup II with a mortality of 45% compared with 31 % In those patients whose ejection fraction increased or did not change Although the Identification of a group of patients with reversible myocardial Ischemia and a poor prognosis while undergoing medical therapy has potentially Important therapeutic imphcations, further studies will be needed to define whether the ejection fraction re-

JACC Vol 3. No I January 1984 88-97

sponse to exercise has independent predictive value in addinon to other cluneal, nonmvasive and cardiac catheterizanon data Conclusions. The results of this study indicate that m patients With coronary artery disease and abnormal left ventncular function, a further decrease m ejection fraction durmg exercise IS more likely to represent Ischemia than a nonspecific response of the ventncle to exercise stress, even In the presence of severe left ventncular dysfunction Such a decrease in ejection fraction IS related to multivessel coronary artery disease and mortality during long-term followup on medical therapy We acknowledge Jaruce WIlson and Sharon Kamash for their assistance m the collection of data, and Catlue Collins for her excellent secretanal work

References McNeer lF, Margohs lR , Lee KL, et al The role of the exercise test m the evaluation of patients for ischemic heart disease CIrculation 178,57 64-70 2 Borer lS, Bacharach SL, Green MY, Kent KM, Epstein SE, Johnston GS Real-time radionuchde cineangiography in the nomnvasive evaluanon of global and regional left ventncular function at rest and dunng exercise in patients WIth coronary artery disease N Engl 1 Med 1977,296839-42 3 Jones RH, McEwan P, Newman GE, et al Accuracy of diagnosis of coronary artery disease by radionuchde measurements of left ventncular function dunng rest and exercise CIrculation 1981,64 586-600 4 Berger HJ, Reduto LA, Johnstone DE, et al Global and regional left ventncular response to bicycle exercise m coronary artery disease assessment by quanntanve radionuchde angiocardiography Am 1 Med 1979,66 13-20 5 Freeman MR, Berman DS. Stamloss H, et al Companson of upnght and supine bicycle exercise m the detection and evaluation of extent of coronary artery disease by equihbnum radionuchde ventnculography Am Heart 1 1981.102 182-9 6 Rigo P, Bailey IK, Gnffith LSC, et al Value and hmuations of segmental analysis of stress thaihum myocardial imagmg for localizauon of coronary artery disease Ctrculanon 1980.61 973-80 7 Caldwell JH, Hamilton GW , Sorenson SG, Ritchie n., Williams DL. Kennedy lW The detection of coronary artery disease WIth radionuchde techniques a companson of rest-exercise thaihum imaging and ejection fraction response Circulation 1980,61 610-9 8 Castellanet Ml, Greenberg PS, Ellestad MH Companson of SoT segment changes on exercise testmg WIth angiographrc findings m patients WIth pnor myocardial mfarction Am 1 Cardrol 1978,4229-35 9 Kramer N, Susmano A, Shelelle RB The "false negative" treadmill exercise test and left ventncular dysfunction Circulation 1978,57 763-8 10 Port S. McEwan P. Cobb FR, Jones RH Influence of resting left ventncular function on the left ventncular response to exercise m patients WIth coronary artery disease Circulation 1981.63 856-63 II

Veraru MS, Jhmgran S. Attar M. Rizk A, Quinones MA, MIller RR Poststress redtstnbution of thalhum-201 in patients with coronary artery disease , WIth and WIthout pnor myocardial mfarcuon Am J Cardiol 1979,43 1114-22

12 Borer lS , Bacharach SL, Green MY. et al Exercise-induced left ventncular dysfunction in symptomatic and asymptomatic patients WIth aortic regurgrtanon assessment WIth radionuchde cmeangiography Am 1 Cardiol 1978,42351-7

JACC Vol 3. No 1 January 1984 88-97

13 Das SK. Brady TJ, Thrall JH, PItt B Cardiac function with pnor myocardins J Nucl Med 1980,21 689-98

HIGGINBOTHAM ET AL EXERCISE RESPONSE WITH CARDIAC DYSFUNCTION AT REST

In

patients

14 Rerych SK, Scholz PM, Newman GE, Sabiston DC Jr. Jones RH Cardiac function at rest and dunng exercise In normals and In patients with coronary heart disease evaluation by radionuchde angiocardiography Ann Surg 1978,187449-56 15 Scholz PM, Rerych SK, Moran JF, et al Quantitative radionuchde angrocardiography Cathet Cardiovasc Diagn 1980,6 265-9 16 Peter CA, Jones RH Cardiac response to exercise In patients with chronic aortic regurgitauon Am Heart J 1982,10485-91 17 Newman GE, Rerych SK, Upton MT. Sabiston DC Jr. Jones RH Comparison of electrocardiographic and left ventncular functional changes dunng exercise Circulation 1980,62 1204-11

97

18 Upton MT, Rerych SK, Newman GE, Port S, Cobb FR. Jones RH Detecting abnormalities In left ventncular funcnon dunng exercise before angina and ST-segment depression CIrculation 1980,62 341-9 19 Bruschke AVG, Proudfit WL, Sones FM Jr Progress study of 590 consecutive nonsurgical cases of coronary disease followed 5-9 years II Ventnculographrc and other correlations Circulation 1973,471154-63 20 Burggraf GW. Parker JO Prognosis In coronary artery disease Angtographic. hemodynarmc, and cluneal factors Circulation 1975,51 146-56 21

HdlT1~ P1. Harrell FE, Lee KL, Behar VS, Rosati RA Survival In medically treated coronary artery disease Circulation 1979,60 1259-69