Maximal Exercise Tolerance in Chronic Congestive Heart Failure

Maximal Exercise Tolerance in Chronic Congestive Heart Failure* Relationship to Resting Left Ventricular Function Edgar S. Carell , MD; Srinivas Murali, MD; Douglas S. Schulman , MD; Tulia Estrada-Quintero, MD; and Barry F . Uretsky, MD

The relationship between maximal exercise tolerance and resting radionuclide indexes of left ventricular systolic and diastolic function were evaluated in 20 ischemic and 44 idiopathic cardiomyopathy patients with New York Heart Association class 2-4 chronic congestive heart failure. Left ventricular ejection fraction, peak systolic ejection rate, peak diastolic filling rate, time to peak filling from end-systolic volume, and fractional filling in early diastole were measured from the radionuclide ventriculogram. All patients underwent symp tom-limited exercise testing with on-line measurement of oxygen consumption. In the ischemic group, all of the radionuclide indexes correlated poorly with maximal exercise oxygen consumption (Vo2max) except the peak systolic ejection rate which correlated modestly (r=0.58, p<0.05). Peak systolic ejection rate was significantly lower (p14 mL/kg/min). In the idiopathic group, none of the radionuclide indexes correlated well with Vo2max; and all indexes were similar in patients with and without marked exercise intolerance.

I mpaired exercise tolerance is the most frequent

symptom of chronic congestive heart failure (CHF).l Previous studies in CHF patients have shown a poor correlation between resting indices of left ventricular systolic function and indices of maximal exercise tolerance.v'' There is speculation that resting right ventricular ejection fraction may predict exercise tolerance in some patients, although this is controversial.f-S" Diastolic left ventricular dysfunction defined as an impairment of ventricular relaxation, abnormal chamber stiffness, or muscle stiffness often accompanies systolic dysfunction in the majority of chronic CHF patients and may be an important determinant of symptoms, functional capacity, and possibly surviva1. 6,8-10 A few small studies have suggested a direct correlation between exercise *From the Division of Cardiology, University of Pittsburgh School of Medicine, Pittsburgh. Manuscript received January 19, 1993; revision accepted February 14, 1994.

These data suggest that (1)resting left ventricular ejection fraction poorly predicts maximal exercise capacity in both ischemic and idiopathic cardiomyopathy and (2) resting peak systolic ejection rate, peak diastolic filling rate, and fractional filling in early diastole may predict exercise tolerance in ischemic but not idiopathic cardiomyopathy. (Chest 1994; 106:1746-52)

CHF=congestive heart failure; FF= fractional fillingin the first third of diastole; HRPFR=peak diastolic filling rate normalized to resting heart rate; L VEF=left ventricular ejection fraction; NYHA=New York Heart Association; PFR=peak diastolic filling rate; SER=normalized left ventricular peak systolic ejection rate; TPF=time to peak filling from end-systolic volume; Vco2max=maximal carbon dioxide production; VEmax= maximal minute ventilation; Vo2max=maximal oxygen consumption

Key words: congestive heart failure; exercise tolerance; left ventricular diastolic function ; left ventricular systolic function

tolerance and left ventricular diastolic filling in CHF patients. ll ,12 Heart failure patients with preserved exercise tolerance have been shown to have a higher peak diastolic filling rate during exercise compared with those with marked exercise intolerance.f The majority of patients in these studies, however, had CHF because of ischemic heart disease. Whether left ventricular diastolic filling predicts exercise tolerance in CHF patients who suffer from idiopathic cardiomyopathy has not been carefully studied. The objective of our study was to compare the relationship between maximal exercise tolerance and resting left ventricular systolic and diastolic function in chronic CHF patients with ischemic cardiomyopathy to those with idiopathic cardiomyopathy. M ETHODS

Study Patient s

15261

Sixty-four patients with left ventricular systolic dysfunction (left ventricular ejection fraction :545%) and New York Heart Association (NYHA) functional class 2-4 chronicCHF, whowere referred to our Cardiac Functional TestingLaboratory for func-

1746

Maximal Exercise Tolerance in Patients with Chronic CHF (Carell et al)

Reprint requests: Dr. Murali, University of Pittsburgh Medical Cent er, 538 Scaife Hall, 3550 Terrace Street , Pittsburgh, PA

Table I-Characteristics of Study Patients Ischem ic (n= 20) Mean age, yr

Idiopathi c (n = 44)

57± 10 49 ±II (range, 21-81) (range, 33-73)

Sex Male Female NYHA class

2 3 4 Exercise param eter s Exercise HR , beat s/min Exercise BP, mm Hg Systolic Diastolic VE, L/ min Vc02max, L/min V02max, mL /kg/min Anaerobic threshold V02

18 2

36

7 7 6

27 15 2

141 ±34 140±29 78± II 57±20 1.5±0.9 14.0±5.6 11.9±4.5

158±26* 155±28 80±9 59±22 1.9 ± 1.0 17.4±5.6° 13.6 ± 4.3

8

*Probability value less than 0.05. tiona I capacity assessmen t were retrospectively studied . The CH F was due to eithe r ischemic (n = 20) or idiopathic cardiomyopathy (n =44). Ischemic patients had angiographically documented multivessel coronary artery dise ase and idiopathic ca rd iom yopathy patien ts had CHF in th e absen ce of signifi cant coronary disease , valvu lar disease , hypertensive or hypertrophic heart disea se or heart disease due to systemic illn ess. Patient characteristics are listed in Table 1. This idiopathic patient group was significantly younger than the ischem ic group.

Exercise Tolerance Assessment Maximal exercise testing was performed in all patients who wer e in the postprandial state using either an increm en tal mul tist ag e tread m ill protocol (Modified Naughton protocol ) with 2-m in stages l 3 or a n electro nic bicycle ergom eter protocol (W asse rma n Ramp protocol) starting at a wor kload of 10 Wand increasing it by 10 w every minute.!" All pat ients exercised until th e development of severe dyspnea or leg fatigue. Patients who discontinued exercise du e to angina wer e excluded from the stud y. All patients were on a stable medical regimen comprising digoxin, d iuretics, angiotensin-converting enzyme inhibitors, nitrates, hydra lazine, and antiarrhythmic drugs in var ious combinations at the time of the exercise test. None of the medications were withheld on the day of the test. No patient was receiving either t1-adren ergic blockers or calcium-channel blockers. During each test, contin uous on-line breath-by-breath measurement of m inute ventilation, oxyg en consum p tion (voz), and carbon d ioxid e production was performed using a Sensormedics Metabolic Cart. Lactate thresho ld or onset of anaerobic metabolism (anaerobic threshold) was noninvasively derived from th e aforemention ed m etabolic parameters, using pr eviously described criteria. 1S. 17 The exercise da ta were comparable in the two groups exce pt that the idiopathic patients achieved a significantly h igher maximal exercise heart rate and V02 (V02max) compared with ischemic patients (Table 1).

Radionuclide Angio graphy Each patient also underwent multiple gated blood pool cardiac scintigraphy at rest in the supine position. Red blood cells wer e labeled using standard in vivo techniques with 25 mCi of technetium 99m. Im aging wa s don e with a General Electric Starcam mobile camera equipped with a medium sensit ivity

parallel hole collimator. The stud y was formatted at 32 frames per ca rd iac cycle . The be st left ante rior oblique position that separated the left ventricle from the right ventricle and le ft atrium was chosen. Using a commercial computer program (SAG E), semi automati c regions of int er est wer e gen erated over the left ventricle for eac h fram e in th e cardiac cycl e using a com bine d second derivative and count threshold algorithm. A ba ckground region was automatically generated la teral or inf erior to th e left ventricle in the end- systolic frame , thus avoiding overlap with high-count background area in th e spleen or descending aorta. From the regions of interest , a high temporal resolution ba ck ground correct ed left ventricle tim e activity curve was obtained. Left ventricular ejec tion fra ction (L VE F ) was cal culated as left ventricular end-d iastolic counts minus left ventricular endsystolic counts divided by left ventricular end- d iastolic counts. Th e tim e activity curve dat a was fitted using a fourth Fourier transform seri es. Th e maximum first derivative during dia stole was used to esta blish peak diastolic fillin g rate (P F R) wh ich was then normalized to end- d iastolic counts. Since PFR is affe ct ed by heart rate, it also was normalized to resting heart rate (H RPF R).18 The tim e to peak filling from end-systolic volume (T PF) and fractional filling in th e first third of diastole (FF) were derived .19 Peak systolic ejection rate (SER) also was determined from th e maximum first derivative during systole of th e time activity curve.

Data Analysi s Based on the Vozmax, ische m ic and idiopathic groups were arbitrarily divided into those with relatively preserved maximal exercise tolerance (Vo zmax > 14 mL /kg/min) and those with impaired maximal exer cise tolerance (V02m ax :514 mL /kg/ min). This arbitrary classification was based upon previously descr ibed sur vival diff erences in th ese groups.P' Student's t test was utilized for comparison of dat a between ischemic and id iopathic cardiomyopathy gro ups and between patients with relatively preserved exercise toler ance and those with impaired exercise capacity. Analysis of variance was used to compare data among patients strati fied by NYHA functional class. Correlations between radionuclide indexes and e xercise variables were per formed using th e Pearson product moment correlation. Kendall rank correlation was used to correlate radionuclide and exercise parameter s with N YH A fun ctional class. Statistical signifi cance was d efined as a probability value of less than 0.05 . All data a re expressed in mean ± SD. RESULTS

Badionuclide Data All resting radionuclide parameters were comparable in the ischemic and idiopathic patient groups (Tab le 2). The LVEF, SER, and PFR were considerably reduced in both groups (normal LVEF, 69 ± 7%; SER, 3.6 ± 0.7 end-diastolic coun t [EDC]/ s; PF R, 3.2 ±0.7 EDC/s) ; TPF and FF also were lower (norma l TPF, 138±28 ms; FF, 40± 16%). Resting heart Tab le 2-Resting Radionu clide Data in the Study Patien ts Param eters

Ischemic (n = 20)

Idiopathic (n = 44)

Probability

LVEF , % SER, EDC /s PFR , EDC /s TPF, ms FF , %

25±10 1.3±0.7 1.3±0.6 117 ±41 28±10

27±9 1.4±0.6 1.3±0.5 128 ±62 23±12

NS NS NS NS NS

CHEST / 106 / 6 / DECEMBER , 1994

1747

50 40

s ~

•

•

20 10 0

'0

., " <,

~ . ...

30

r..

3.0

•

••

..II

u

•

e

I':l l>:

r..

r= .3 8

0-

2.0

1.0 0 .5

10

15

20

25

30

V02 max (mt/min/kg)

v

3 .5 r -.---,.-.------r---, 3.0

~ 2 .5 til

G§. l>: c.l Cf.]

2.0 1.5 1.0 0 .5

• I

•

•

E ci'"

II ., g

<,

e

e:

r= .58

5

10

r..

E:: ...J

15

20

25

~

0 .04.--.---,.0 .03

;;-----: •

50 ~

I::

40 30 20

I

10

o

5

•

10

r= .32 15

20

25

30

V02 max (mt/min /kg)

•••

0 .02 0.01

'0

200

~ 150

.§.

r.. 100

i=

50

• • I


••

G§. c.l

2.0 1.5 1.0

~...

en 0 .5

2.0


o

1.5

I':l

e

1.0

l>:

r..

0 .0

r = .2 1

'---...L.---'--'---"----'---'

5

10 15 20 25 30 35

E

I

~

I::

r =. -10

•

-'---'----'

20

25

30

V02 max (ml/min/kg)

• • ~4

40

.·...:..

30 20 10

. .... -

o '----'----'------'--'-----l.----' 5

10 15 20 25 30 35

V0 2 max (ml /min /kg)

r= .25

10 15 20 25 30 35

0 .0 4 r -. ---.--.---.--,---,

....-....

0..

A

ci-

" <,

0 .0.3


U

Q

e

e:

.~

_ • • •

0 .02

..••.•• -.

..

_ ~.

.~

0 .0 1

• • ",

0-

l>:

r = .18

:I: 0.00 '----'----'-------'------'-----'----'

5 10 15 20 25 .30 .35 V02 m ax (m t/m in / kg)

70 r--.----r----,---r--,-----,

50

5

••

V0 2 m ax (mt/min /kg )

V02 max ( m l/min/kg)

~

. "",... ~. .·... -. · ..

0 .5

0-

.... ... ,-----------.

250 r -_._, ----,----.----,----,

60

•• • • •• • •

0 '-----'-----l._ 5 10 15

.. . . ~ - .. ...

~ 2 .5

~

r= .45

"

<,

V0 2 max (mt/min /kg)

3 .0

••

2.5

'0

0 .0

3.5 r-.--.-......--.--,.--,

•

3 .0

10 15 20 25 30 35

5

•

~

-.

r = .21

,.----.--,

. ~.. ...

•

.~.

20

0

30

250 . - . - - , . -,------.----,

~

30

10

0 .00 '-----'-----l._-'---'----' 5 10 15 20 25 30 V02 m a x (ml/min/kg)

V0 2 m ax (mt/m in/kg)

60

• ._ .t •

V02 m ax (rot/min/kg)

0..

0 .0 '-----'-----l._ -'---'-----' 5 10 15 20 25 30

•

f.-

40

g

r= .34

... ~ ·.....• ...

50

•

~ •• • •

1.5

0 .0 5

.

2.5

'0

•

200

.. .. • ·.....

~ 150

~ -" -

.§.

r.. 10 0

0f-

•

50

•

••

r= . 15

o '---...L.---'--'---"-- - '---' 5

10 15 20 25 .30 3 5

V02 m a x (ml/m in / kg )

FIGURE 1. Corre lation betw een restin g rad ionu clide measures of left ventricul ar function and maximal exercise tolerance in ischem ic cardiom yopathy patients . Corre lation with Vo2m ax was poor (p =NS) for LVEF, PFR , TPF, and FF but mod est and significant (p < O.05) for SER and HRPFR.

FIGURE 2. Cor re lation betwee n resting radio nuclide measures of left ventricul ar function and maximal exercise toler anc e in idiopathi c ca rdiomy opathy pati ents. All measures of resting left ventric ular systolic and diastolic fun cti on corre lated poorl y (p = NS) with Vo -rnax.

rate was comparable in both groups (82 ± 17 vs 90 ± 16 beats per minute).

cantly lower in ischemic patients with poor exercise tolerance (1.05 ± 0.5 vs 2.1 ±0.8 EDC/s; p
Relatio nship Between Resting Radionuclide Data and Maxi mal Exercise Capacity In the ischem ic gro up, resting LVEF, PF R, TPF , and FF correlated poorly (p =NS) with Vo2max (r=0.38, 0.34, - 0.10, and 0.32, respectively); however, there was a modest significant correlation with SER (r = 0.58; p<0.05) and HRPFR (r = 0.45; p<0.05 [Fig 1]). On the other hand , in the idiopathic group, V02max correlated poorly (p = NS) with resting LVEF (r = 0.21), PFR (r = 0.25), SER (r = 0.21), TPF (r = 0.15), FF (r = - 0. 14), and HRPFR (r = 0.18 [Fig 2]). In both groups, the resting LVEF was similar in patients with preserved (Vo 2max > 14 mL /kg/min) and impaired (Vo 2max < 14 mL /kg/min) exercise tolerance (F ig 3). Th e PFR was significantly less in ischemic patients with impaired exercise tolerance (1.09 ± 0.53 vs 1.63±0.50 EDC /s, p <0.05) as were HRPFR (0.013 ± 0.007 vs 0.022 ± 0.009 EDC /s/beats per minute; p<0.05) and FF (31.2 ± 9,4 vs 49.6 ± 11.0%, p < O.OI). The SER also was signifi1748

Relationship of Radionuclide and Exercise Data to New York Heart Association Functional Class There were no significant differences in the radionuclide indexes among NYHA class 2, 3, and 4 patients (Table 3). Resting LVEF, SER, PFR , TPF, and FF correlated poorly with NYHA functional class (r=-0.19, - 0.29, - 0.24, - 0.17, and - 0.21, respectively) . There was a significant correlation between V02max and NYHA functiona l class (r=-0.80, Maximal Exerc ise Tolerance in Patients wit h Chron ic CHF (Carell et all

1

3D

,...'.01

,....---,

]

3

~ ,

20

..

;":

'0

I3CHBW.C

IDI OPA"nII C

to

I D IOPA TH IC

lSCHBWJC

ID IO P A THIC

lB H RN JC

. . o.OJ

i

~

3

j, .".. , IDIOPATlUC

IDIOPATHIC

I8CHZNIC

J8HZIolIC

ID IOPA.THI C

IllH ZNIC

F IGURE 3. Restin g rad ionu cl id e m easur es of left ven tricula r fun ction in ischem ic and id iopat hic ca rdiomyopath y pati ent s by degr ee of exe rcis e impairment. T he SE R, PFR , HR PFR , and FF we re significan tly (p < O. 05) lower in idiopathic ca rdiom yopathy patients with marked exe rcise int olerance (V0 2 :5 14 mL/ kg /min , open bars) com pared with th ose with pr eser ved exe rcise ca pacity (Vojmax > 14 m L / kg / mi n , hatched bars).

p
Parameters

NYHA Class 2 (n = 34)

NYHA Class 3 (n = 22)

LVEF, % SER, EDC /s PFR, EDC /s TPF, ms FF, %

28±8 1.6± 0.6 1.4±0.5 134 ± 50 29±1l

25 ± 1l 1.2 ± 0.7 1.2 ± 0.5 123±41 25±9

NYHA Class 4 (n = 8) 24 ±9 1.1 ±O.5

1.4 ± O.5 115 ± 47 21 ±12

Tab le 4- Relationship of Exercise Parame ters to Ne w York Hea rt Association Class

Parameters Exercise HR, beats/ min Exercise BP, mm Hg Systolic Diastolic VEmax Vco2max V02max Anaerobic threshold V02

NYHA Class 2 (n = 34)

NYHA Class 3 (n = 22)

NYHA Class 4 (n = 8)

169±24

140± 21*

117 ± 25 f1

155± 25 81±9 68 ±20 2.4 ± 0.9 20.5 ± 4.7 15.8± 3.9

151 ± 35 78 ±10 49 ± 17* 1.2 ± 0.5* 12.8 ± 1.3* 1O.7±2.0*

125 ± 18 t 75± 12 42± 19 t 0.8 ±0.4t 8.4± 1.1tl 7.8± 1.3t

*Probab ility value less than 0.01, NYHA class 3 vs class 2. t Probab ility value less than 0.01, NYHA class 4 vs class 2. [ Probability value of less than 0.05, NYHA class 4 vs class 3.

hea rt ra te and V0 2ma x wer e significan tly lower in class 4 pati ents compared with class 3 pa tie nts. D ISCUSSION

Exercise Tolerance in Congestive Heart Failur e Since chronic CH F pa tients ofte n are sym pto matic only d ur ing exertion, it is im portant to assess their fun ctional status in terms of their exercise toler ance .21,22 Clinical classifica tion of CHF seve rity such as N YHA function al class relies solely upon th e patient' s subjective assessment of his or her degree of ph ysical impairment and ma y therefore be inadeq ua te in grading CHF. 22,23 Objective assessme nt of exe rcise toleran ce by non inv asive measurement of V0 2max (aerobic capacity ) and ana erobic threshold during incremental uprigh t exe rcise therefore has been sugges te d as a more reliable measur e of fun ction al status in ch ron ic CHF patients.22 Ma ximal exe rcise V0 2 or V02max is d et ermined by cardi ac output response to exercise and ma xim al oxygen extraction .24 Oxygen extract ion by metabolizin g tissues is not genera lly impaired in chronic CH F pat ient s' P and may in fact be au gmented whe n com pare d with normal subjects.25-27 Th erefo re , V oernax primarily reflect s th e level of ca rdi ac output achi eved during maximal exe rcise , ie, ca rdi ac reser ve. Impaired exercise tolerance in chronic CHF, howe ver , is not on ly relate d to skeletal muscle underperfusion, but also to peripheral abnormalities such as impaired periphe ra l vasodilatory ca pacity , muscle atrophy, an d decr eased oxida tive metabolism .28-3o In this study, non e of the resting radi onuclide measures of left ventricular systo lic or diastoli c CHEST 1106 / 61 DECEMBER , 1994

1749

functi on correlated with NYHA functional class. In fact , ther e wer e no significant diff er en ces in any of th e radi onuclide ind exes amo ng th e diff er ent fun ctional classes. Correlati on bet ween V0 2m ax and NYHA class was good but not exce llen t in both ische mic and idi opathic ca rd iomyopa thy pati ents. T h us, it would seem th at although subjective NYHA functio na l class assessment is a somew hat reliabl e m easur e of functiona l capacity in CHF patients, it has limitati ons.

Left Ve ntricular Systolic Fun ction and Exercise Tolerance Th e result s of our study confirm th e pr evious observa tion that restin g left ventr icular ejectio n fra ction does not accurat ely pr edi ct maximal exercise toler ance in chronic CH F .2,5,31 The L VEF did not correla te well with V0 2max, wh ether CH F was due to ische mic or idiopathic card iomyopathy, and th er e was no diff er enc e in th e restin g L VEF between patients with relatively pr eser ved exe rcise tolerance and those with im pa ired exe rcise tolerance. Most previo us stud ies describing th e lack of correlation be tween restin g LVE F and exe rcise tolerance in C HF pat ients involved sma ll study populati ons composed predominantly or en tirely of pati ent s with ische mic heart disease .2-5 Baker et al" report ed a poor correla tion bet ween LVEF and V0 2max (r = 0.08), in 25 men with chronic CHF, 12 of wh om had ischemic heart disease. Franciosa et aJ2 also described a poor corre lation (r = 0.06) bet ween resting L VEF and exercise tim e in 21 chronic CHF pati ents, 16 of whom had ischemi c card iomyopa thy. Th e inability of resting LVEF to pred ict exe rcise tolerance in chronic C H F may relat e to seve ra l mech ani sms. Exer cise toler an ce ma y be pr eser ved in some patients with poor left ventricular systolic functi on becau se of th e ability to tolerate elevated pulmonary arte ry wedge pr essur es without developing dyspn ea, increased pulm onary lymphat ic flow th at limits pulmonar y veno us conges tion, pr eser vati on of a ppropriate chronotr opi c response, ad equate left ventricular dilatation during exercise, ability to further acti vat e th e alrea dy stim ulated neurohormonal mechanisms and chronic cha nges in left ventricular com pliance th at limit aug me nta tion of filling pr essures during exe rcise." Further, right ventric ular dysfunction may coex ist, particularl y in pati ents with id iopa thic ca rdiomyopa thy and exe rcise may precipita te ischem ia in some ische mic heart disease pati ents.V Th e relat ionship bet ween restin g SER and exe rcise toleran ce has not been syste ma tically eva luated pr eviously. Both restin g and exe rcise SER pr eviously have been shown by Heo et al5 to be similar in ischemic patients with normal exe rcise tolerance and severe exercise intolerance. Th ere are no data corre1750

lating SER to exercise toler an ce in idi opathic ca rd iomyopa thy pat ients. In th e pr esent study, SER was reduced in both ischemic an d idi opathic pat ients. Th er e was a mod est correlation bet ween restin g SER and V0 2max in ische mic patients. F ur the r, pati ent s with ische mic d isease and im pair ed exe rcise toler ance had a significan tly lower SER than th ose with pr eser ved exe rcise tolerance (1.05 ± 0.5 vs 2.10 ± 0.80 EDC js ; p
Left Ve n tricular Diastolic Function and Exercise Tolerance Lef t ventric ular d iastolic filling abnorma lities are freque ntly seen in associa tion with decreased left ven tric ular systolic funct ion in chronic CHF second ary to both ische mic and idi opathic ca rdio myo pathy .8-10,34 Mec hanisms include myocardial ischemia , alte ra tion in the ph ysical pr operties of th e myoca rd ium du e to myocyte destructi on and fibrosis, increased sym pathetic ner vous syste m activity with ~- l - receptor down regul ati on ca using reduced m yoca rdial relaxation rat e and cha nge in load ing conditions." Restin g PFR in both groups in thi s study wer e sim ilar and considerably lower th an normal. Resting PFR correla ted poorl y with NYH A fun ctional class; and correlation between restin g PFR and V02max was poor in both grou ps. In contrast to idiopathic ca rd iomyopathy pati ents, however , ische mic patients with impair ed exe rcise tolerance had a significantly lower PFR com pared with th ose with pr eserved exe rcise toler an ce. Soufe r et all l pr eviously report ed in a sma ll study that resting PFR was a good pr ed ictor of maximal exe rcise V0 2 in CHF patients. However , in th at study only 9 pat ients we re evaluated; the mean L VEF was 40% sugges ting a some wha t diff erent patient populat ion com pa red with th e present stu dy; and majority of pat ients had left ventricul ar d ysfunction seconda ry to isch emic disease . We also noted a moder at e correlation bet ween HR PFR and V0 2max in ischemic pati ents. The HRPFR also was signifi cantly lower in ische m ic patients with impaired exe rcise toler an ce com pa red with tho se Maximal Exercise Tolerance in Patients with Chron ic CHF (Carell et al)

with pr eser ved exercise tolerance , and this differ en ce was not apparent in th e idiopathic group. Perhaps th e mechanism for exe rcise intol erance is diff er ent in patients with ischemic and idiopathic cardiomyopathy. Isch emic patients in our study who had lower PFRS ma y have had significant silent myocardial ischemia at rest and th er efore a poorer exe rcise tolerance . Further , silent myocardial ischemia induced by exercise may ha ve incr eased left ventricular stiffness, thus contributing toward a rapid rise in left ventric ular filling pressur es and limiting exerc ise tolerance .f This inter esting hyp oth esis also m ay explain th e lack of correlation between resting PFR and exercise tolerance in idiopathic ca rd iomy opathy patients. Furth er stud ies are needed to clarify if th is, in fact, is th e case. Other measures of di astolic functi on such as TPF and FF wer e also reduced in both groups and correlated poorl y with NYH A class. The TPF and FF correlated poorl y with exercise toleran ce in both ischem ic and idiopathic ca rd iomyopa thy patient s. However , like PFR , F F was significantly lower in ische mic pat ients with poor exe rcise toler ance com pare d with tho se with pr eser ved exe rcise toler an ce.

Limitations

N YHA functiona l class in CHF pati ents correlates well with objective mea sur es of maxim al exe rcise toler an ce but not measur es of restin g left ventricular systolic or d iastolic functi on . Restin g L VEF is a poor pr edictor of ma ximal exe rcise tolerance in patients with CHF du e to ischemic or idiopathic cardiom yopathy. Resting SER , PFR , and FF ma y be useful in predicting maxim al exe rc ise toler ance in ischemic but not idi opathic pati ents. Larger , pr ospective studies ar e need ed to confirm th ese observa tions. ACKNOWLEDGMENT: We than k Alfred Cecchetti for assistance with statistical ana lysis and Marga ret Altvater for preparation of the man uscript. R EFERENCES

2 3 4 5

Our findings have to be interpreted in the context of certain limitations. First, exercise studies and radionuclide ang iography were not performed on th e same day. The stud ies wer e done within 3 months of each othe r, and th er e may have been int ercurrent change in th e patient's left ventricular systolic or diastolic fun cti on bet ween th e two tests. Every effort was mad e to ensure that patients we re clinically sta ble bet ween th e tests, thus making a significant change less likel y. All th e exercise studies were conducted afte r the patient was familiarized with th e exerc ise laboratory and personnel with an initial test. Second , patients were on various med ical regim ens for th e tr eatment of CHF , and we ca nnot exclude possible influ ences of drug th erapy on th e study results. Th e differ ent relati onships observe d bet ween radionuclide paramet er s and exercise tolerance in the isch emic and idiopathic pati ent groups could be expla ined by th e diff er ent medical regimens in the two groups. Eve ry patient was maintained on a stabl e medical regim en between th e time of radionuclide stud y and th e exercise test , and non e of th e ischemic heart d isea se patients developed an gin a during exercise. Fi nally, limitations of th e radionuclide ventriculogram in accurat ely measuring diastolic fun ction sho uld also be born e in mind. 35 C ONCLUSION

Despite th ese limitations, we can draw seve ra l conclusions from thi s study. Subjectiv e assessme nt of

6 7

8

9 10 II

12 13 14 15

Engler R, Roy F, Higgins CB, et al. Clinical assessment and follow-up of fun ctional capacity in patients with chronic congestive cardiomyopathy. Am J Card iol1 982; 49:1832-37 Fr anciosa JA, Park M, Levin TB. Lack of correlatio n between exercise capacity and indexes of resting left ventricular performance in heart failure. Am J Cardiol1981; 47:33-9 Higginbotham MB, Morris KG, Conn EH, et al. Determ inants of variable exercise performance among patien ts with severe left ventricular dysfunction. Am J Ca rdiol 1983; 41:51-60 Baker BJ, Wilen MM, Boyd LM, et al. Relation of right ventricu lar ejection fraction to exercise capacity in chronic left ventricular failure. Am J Ca rdiol 1984; 54:594-99 Heo J, Iskandri an A, Hakki A. Relation betwee n left ventricular diastolic function and exercise toleran ce in patient s with left ventr icular dysfunction: catheterization and cardi ovascular diagnosis. 1986; 12:311-16 Litchfield RL, Ker ber RE, Benge W, et al. Norma l exercise capaci ty in patients with severe left ventricular d ysfunction: comp ensator y mechanisms. Circulation 1982; 62:129-34 Haines DE , Beller GA, Watson DD , et al. A prospec tive clinical scintigraphic, angiog rap hic and fu nctional evaluation of patients aft er inferior wall myocardial infarction with and without right ventricular dysfunction . J Am Coli Cardiol 1985; 6:995-1003 Feldm an M, Alderma n J, Aroesty J, et al. Depression of systolic and diastolic myocard ial reserve during atrial pacing tachycardia in patients with dilated cardiomyopathy. J Clin Invest 1988; 82:1661-69 Pouleur H, Han et C, Curne 0 , et al. Focus on diastolic function: a new ap proach to hea rt failure therapy. Br J Clin Pharmacol 1989; 28:415-525 Grossman W, McLauren L, Rolett E. Altera tion in left ventric ular relaxation and diastolic comp liance in congestive cardiomyopathy. Car diovasc Res 1979; 13:514-22 Soufer R, Lindo C, Kanisk J, et al. Th e relationship of baseline peak filling rate to exercise capacity in congestive heart failure. Circu lation 1986; 74:II-138 Vanoverschelde J, Raphael D, Robert A, et al. Left ventricular filling in dilated cardiomyopathy: relation to funct ion class and hem odynamics. J Am Coli Cardiol 1990; 15:1288-95 Patterson JA, Naughton J, Pietra s RJ, et al. Treadmill exercise in assessme nt of the fu nctional capacity of patients with cardiac disease. Am J Ca rdiol 1972; 30:757-62 Hansen JE. Exercise instruments , schemes, and protocols for evaluating the dyspneic patient. Am Rev Respir Dis 1984; 129: 5525-27 Webe r KT, Kinasewitz GT, Janicki JS, et al. Oxygen utilization and ventilation du ring exercise in patients with chronic cardiac failur e. Circulation 1982; 65:1213-23 CHEST 1106/ 61 DECEMBER , 1994

1751

16 Wasserma n K, Mcll roy MB. Detecting the threshold of anae robic me tabolism in cardiac pat ients during exercise. Am ] Ca rdiol 1964; 14:844-52 17 Wasserma n K, Whipp B], Koyal SN, et al. Anaero bic thr eshold and respiratory gas exchange during exercise. ] Appl Ph ysiol 1973; 35:236-43 18 Bonow RO, Bacharach SL, Green MV, et al. Impaired left ventricular diastolic filling in patients with coronary artery d isease: assessment with radionuclide angiography. Circu lation 1981; 64:315-19 19 Bashore TM, Shaffer P. Diastolic function. In: Gerson ML, ed. Card iac nuclear medicin e. New York: McGraw-Hill, 1987;187 20 Manci ni OM, Eisen H , Kussmaul W, et al. Value of peak exercise oxygen consumption for optima l timin g of cardia c transplant ation in ambulatory patient s with heart failur e. Cir culation 1991; 83:778-86 21 Willens H], Blevins RD, Wrisley 0 , et al. Th e prognostic value of functional capacity in pati ents with mild to moderat e heart failure. Am Heart] 1987; 114:377-82 22 Franciosa ]A, Ziesche S, Wilen M. Fun ctional capacity of patients with chronic left ventricular failur e. Am ] Med 1979; 67:460-66 23 Weber KT, Wilson ]R , Janicki ]S, et al. Exercise testing in the evaluation of the pat ient with chronic cardiac failur e. Am Rev Respir Dis 1984; 129:560-62 24 Weber KT, Janicki ]S. Cardiopulmonary exercise testing for evaluation of chronic cardiac failur e. Am ] CardioI 1985; 55: 22A-31A 25 Sullivan M], Knight ]0, Higgenbotham MB, et al. Relation between centra l and periph eral hemodynamics during exercise in patien ts with chronic hea rt failure: muscle blood flow is reduced with maintenance of arterial perfusion pressure. Circulation 1989; 80:769-81

1752

26 Wilson ]R, Marti n ] L, Fe rraro N. Impaired skeletal mu scle nutritive flow during exercise in patients with congestive heart failur e: role of cardiac pump dysfunction as determ ined by the effect of dobutam ine. Am ] Cardiol 1984; 53:1308-15 27 Zelis R, Longhur st], Capone R], et al. A comparison of regional blood flow and oxygen utilization du ring dy namic forearm exercise in normal subjects and patien ts with congestive heart failure. Circulation 1974; 50:137-43 28 Zelis R, Nellis SH, Longhurst ], et al. Abnormalities in the regionalcircu lations accompanyi ng congestive heart failur e. Prog Car diovasc Dis 1975; 18:181-99 29 Lipkin DP, Jones DA, Round ] M, et al. Abnormalities of skeletal muscle in pati ents with chro nic heart failure. Int] Cardiol 1988; 18:187-95 30 Wilson ]R, Fink L, Maris], et al. Evaluati on of energy metabolism in skeletal muscle of patients with heart failur e with gated phosphorus-31 nuclear magnetic resonance. Circulation 1985; 71:57-62 31 Meiler SE, Ashton]J, Moeschberger ML, et al. An analysis of the determinants of exercise perf ormance in congestive heart failur e. Am Heart] 1987; 113:1207-17 32 Fr agossa G, Bent i R, Sciammarella M, et al. Symptom-limited exercise testing causes sustained diastolic dysfunction in patient s with coronar y disease and low effort tolerance. ] Am Coli Cardiol 1991; 17:1251-55 33 Kappl er], Ziesche S, Nelson ], et al. The reproducibility of hemodynam ic and gas exchange data during exercise in patients with stable congestive heart failur e. Heart failur e 1986;2:157-63 34 Grossman W. Diastolic dysfunction and congestive heart failure. Circulation 1990; 81:1111-17 35 Plotnick GO. Changes in diastolic function-difficult to measure, harde r to inter pret. Am Heart] 1989; 118:637-41

Maximal Exerc ise Tolerance in Patients with Chronic CHF (Carellet all