JACC Vol. 16, NO.4October 1990:795-803

CLINICAL STUDIES

Changes in Left Ventricular Diastolic Performance After AorticBalloon Valvuloplasty: Acute and Late Effects

795

KHALID H. SHEIKH, MD, CHARLES J. DAVIDSON, MD, MICHAEL B. HONAN, MD,

THOMAS N. SKELTON, MD, KATHERINE B. KISSLO, ROMS, THOMAS M. BASHORE, MD, FACC,

WITH THE TECHNICAL ASSISTANCE OF ROBERT BURGESS, BS, RT, LAURENCE SPERO, BSE

Durham, North Carolina

To evaluate acute and follow-up changes in left ventriculardiastolic performance, simultaneous digital left ventriculog­raphy and micromanometry were performed in 49 patientsundergoing aortic balloon valvuloplasty. All patients im­proved symptomatically after valvuloplasty, and 26 re­turned 6.3 ± 1.5 months later for follow-up catheterization.

Immediately after valvuloplasty, aortic valve area in­creased (before 0.5 ± 0.2 versus after 0.8 ± 0.2 cm2

, p <0.01), cardiac output (before 4.3 ± 1.2 versus after 4.4 ±1.3 liters/min) and ejection fraction (before 51 ± 18%versus after 52 ± 17%)did not change and diastolic indexesworsened, signified by a decrease in peak filling rate (before247 ± 80 versus after 226 ± 78 mils, p < 0.01) and increasein the time constant of isovolumetric relaxation (T) (before78 ± 29 versus after 96 ± 40 ms, p < 0.01) and the modulusof chamber stiffness (before 0.107 ± 0.071 versus after0.141 ± 0.083, p < 0.01).

At follow-up catheterization, 16 patients continued tohave symptomatic improvement (group 1) and 10 had

Percutaneous aortic balloon valvuloplasty is an alternativetreatment for patients with valvular aortic stenosis whoeither refuse or present an unacceptable risk for surgicalvalve replacement. Symptomatic improvement occurs inmost patients immediately after valvuloplasty (1-10) and ismaintained in 30% to 50% of patients followed up for 6months (2-10), Although data exist regarding acute and latechanges in aortic valve area and left ventricular systolicperformance after valvuloplasty (1-10), few data are avail­able regarding acute and late changes in diastolic perfor-

From the Department of Medicine, Division of Cardiology. Duke Univer­sity Medical Center. Durham, North Carolina. This study was supported inpart by Grants HL-07101, HL-I27IS and CA-37S86 from the U.S. PublicHealth Service. Washington. D.C.

Manuscript received December 13. 1989: revised manuscript receivedApril 11. 1990. accepted April 19. 1990.

Address for reprints: Khalid H. Sheikh. MD. Box 3260. Duke UniversityMedical Center. Durham. North Carolina 27710.

'91990 by the American College of Cardiology

recurrence of symptoms (group 2). Aortic valve area,cardiac output and ejection fraction at follow-up catheter­ization in both groups were similar and unchanged fromvalues before valvuloplasty. However, diastolic indexes andleft ventricular mass were improved compared with valuesbefore valvuloplasty in group 1 and worse in group 2 asfollows: peak filling rate: group 1 +8% versus group 2-20%, P = 0.001; time constant of isovolumetric relax­ation: group 1 -11 % versus group 2 +35%, P =0.008;chamber stiffness: group 1 -27% versus group 2 +26%,P =0.001; mass: group 1 -9% versus group 2 +6%, P =0.004.

These data indicate that 1) diastolic performance wors­ens immediately after valvuloplasty, consistent with thedevelopment of ischemia during balloon inflation; and 2)changes in left ventricular diastolic indexes and mass,rather than aortic valve area or systolic indexes, correlatebest with symptomatic status 6 months after valvuloplasty.

(J Am Coil CardioI1990;16:795-803)

mance (II). Given that diastolic performance indexes aresensitive to the development of myocardial ischemia (12-15),acute changes in these indexes after valvuloplasty mayprovide important insights into the development of the globalischemia that is postulated to occur during valvuloplasty.Furthermore, given that abnormalities of diastolic perfor­mance may be a primary cause of symptoms in patients withaortic stenosis (16-21), improved diastolic performance mayaccount for improved symptoms late after valvuloplasty. Toexamine these issues, invasively measured hemodynamicindexes of diastolic performance were examined in a con­secutive series of patients before. immediately after and 6months after successful aortic valvuloplasty.

MethodsStudy patients. Between April 1987 and October 1988, 69

patients underwent successful and uncomplicated aortic

0735-1097/90/$3.50

796 SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

JACC Vol. If>, No.4OClober 1990:795-803

valvuloplasty. From this group. 20 patients were excludedbecause of either hemodynamic instability prompting urgentvalvuloplasty (n = 3) or technically inadequate hemody­namic (n = 5) or echocardiographic data. or both (n = 12).The remaining 49 subjects had compensated New YorkHeart Association functional class III or IV congestive heartfailure or angina at the time of valvuloplasty. By clinicalnecessity, all medications were continued. Patients gaveinformed consent to a protocol approved by the institutionalreview board for all phases of this study.

Echocardiography. Two-dimensional. M-mode and Dop­pler echocardiograms were obtained immediately before andafter valvuloplasty with use of a Hewlett-Packard 77020A or77020CF imaging system fitted with either a 2.5 or 3.5 MHztransducer. Echocardiograms were analyzed off-line withuse of a commercially available computer and software(Nova Microsonics). Chamber size and wall thickness weremeasured from the parasternal M-mode tracing at the mitralchordal level. with values averaged for three consecutivebeats (22). Left ventricular mass was measured by anautomated biplane Simpson's rule method, using the averagevalue of manually traced end-diastolic epicardial and en­docardial outlines from three consecutive cardiac cycles inapical two and four chamber views (23.24). The mean valuesfor inter- and intraobserver variability for mass measure­ments, each performed at least 30 days apart. were 4.2% and3.59(. respectively. Wall stress was determined by themethod of Grossman et al. (25), using chamber size and wallthickness echocardiographic measurements and microman­ometer left ventricular pressures obtained within 5 to 10 minof each other (26.27).

Hemodynamics. Right and left heart catheterization wasperformed from the femoral approach. Baseline hemody­namic measurements were obtained at least 15 min after thepatient had undergone coronary arteriography and aortogra­phy using a low osmolarity. nonionic contrast agent. Coro­nary artery disease was defined as ::::.75% luminal diameternarrowing. Expired gas analysis was performed with aSensormedics MMC Horizon system. Right heart and pul­monary artery pressures were obtained with a 7F Critikoncatheter. Blood oxygen content and saturation were mea­sured on a calibrated Instruments Laboratories 282 CO­oximeter. Cardiac output was by the direct Fick method.

Simultaneous left l'entriClllar and aortic pressures weremeasured with an 8F dual sensor, high fidelity. Millarmicromanometer catheter. Micromanometer pressure cath­eters were balanced externally and then to a fluid-filledlumen pressure once inserted into the ventricle. Data weretransferred to a Hewlett-Packard 8086A computerized pres­sure recorder and displayed on a Gould 1000-C electronicstrip chart. All pressure data. the maximal rate of rise of leftventricular pressure (dP/dt) and the electrocardiogram wererecorded at 5 ms intervals. The analog data were digitally

converted and simultaneously stored with 30 frames/sdigitalventriculography data using an ADAC 4IOO-C system.

A diRitall'entriculoRram was performed at held inspira­tion. utilizing a 256 x 256 x 8 bit deep matrix in the 30° rightanterior oblique projection. Thirty-five to 40 ml of nonioniccontrast medium was power injected at 12 to 15 mils directlyinto the ventricle. Ventricular ectopic and immediate pre­and postectopic beats were excluded. The first five to sixbeats after contrast injection were used for analysis. Digitaldata were mask-mode subtracted using R wave-gatedsubtraction methods during postprocessing analysis. Ven­tricular volumes were derived using semiautomated com­mercially available software. Calibration factors were deter­mined by an imaging grid. Volumetric data obtained in thismanner had previously been validated in this laboratoryusing 20 human heart casts of known volumes (unpublishedobservations). from which a regression equation relating thetrue volume to the volume obtained by digital subtractionventriculography had been derived (28). The resulting vol­ume curve and time-matched pressure data were used toconstruct a pressure-volume loop for each study.

Aortic m/l'e area was calculated by the Gorlin equation(29). Aortic and mitral insufficiency were graded at catheter­ization using previously established criteria (0). Left ven­tricular end-diastolic pressure was obtained at the peak ofthe R wave on the ECG. The time constant of isovolumetricrelaxation (7) was determined by the method of Weiss et al.ot) from the linear relation of left ventricular pressure anddP/dt in early diastole. determined between peak -dP/dt to5 mm Hg above the end-diastolic pressure. The meancorrelation coefficient. using 7 to 20 data points per study.was -0.959 (range -0.912 to -0.997).

Instantaneous diastolicjillinR rates were calculated every5 ms after mitral valve opening. based on volumes derivedfrom digital ventriculography. The raw volume-time datawere smoothed by digital filtering using a fifth-grade movingaverage (32). The instantaneous slope (dV/dt) of thesmoothed volume-time curve was computed and plotted,with peak filling rate defined as the maximal increase involume over a 5 ms interval during the first half of diastole.

Chamber stiffness was determined over the range oflowest diastolic pressure to end-diastolic pressure by themethod of Gaasch et al. (33) according to the equation: P =be Kpv. When differentiated. this equation transforms todP/dV = Kp + b. where P = pressure. V '" volume, Kp = themodulus of chamber stiffness and b is a constant.

Valvuloplasty technique. Valvuloplasty was performedby a retrograde arterial technique as previously described(1). All patients had at least two 10 to 40 s dilations usingeither single 20 or 23 mm or double 10 and 20 mm aorticvalvuloplasty balloons. After valvuloplasty, digital left ven·triculography. micromanometry and echocardiography wererepeated within 10 min of the last dilation.

JACC Vol. 16, No.4October 1990:795-803

SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

797

Figure 2, Symptomatic grouping of 49 patients before aortic balloonvalvuloplasty (PBAV) and 26 patients at follow-up catheterization.Functional class = highest New York Heart Association class forangina or heart failure. Numbers in boxes = numbers of patients.

Figure 1. Patient selection and outcome. Numbers in parentheses =numbers of patients. AVR = aortic valve replacement; Cath =

catheterization; PBAV = percutaneous aortic balloon valvulo­plasty.

Follow-up data. All 49 patients undergoing aortic balloonvalvuloplasty were discharged from the hospital with symp­tomatic improvement. Follow-up symptomatic status wasassessed by periodic outpatient clinic visits and telephoneinterviews. The clinical course of all initially eligible patientsis displayed in Figure I. Patients who were alive and had nothad aortic valve surgery were eligible for follow-up cathe­terization. Of 36 eligible candidates, 26 consented to repeatcatheterization, which was performed between 5 and 7months after valvuloplasty in 23 patients and between 4 andII months in all 26 patients. No patient had a cardiac eventduring the time from valvuloplasty to follow-up catheteriza­tion. All patients were clinically stable at the time of repeatcatheterization.

Patients were divided into two R'oups at the time offollow-up catheterization according to symptom status (Fig.2). Group I consisted of 16 patients who were symptomati­cally improved at follow-up catheterization, as defined byNew York Heart Association functional class I or II for both

congestive heart failure and angina. Group 2 consisted of 10patients who had recurrent symptoms, with class III or IVcongestive heart failure or angina. Medications in use at thetime of repeat catheterization were continued. At follow-upcatheterization, all subjects underwent echocardiography.immediately followed by simultaneous digital left ventricu­lography and micromanometry. Repeat valvuloplasty wasperformed in seven patients at this time. All data wereanalyzed without knowledge of symptom status.

Statistical analysis. Baseline clinical characteristics be­tween groups were compared by unpaired Student's t testsfor continuous variables and chi-square analysis for dichot­omous variables. Intra- and intergroup comparisons of allechocardiographic and hemodynamic variables were byWilcoxon two-sample tests. Comparisons of differences be­tween groups for interval changes in variables were per­formed using the Wilcoxon signed-rank test. A p value<0.05 was considered statistically significant.

ResultsBaseline characteristics. Baseline characteristics of the 49

patients undergoing valvuloplasty as well as of the twopatient groups at follow-up catheterization are shown inTable I. There were no differences among the 49 patients atbaseline evaluation and between either group at follow-upcatheterization in terms of age, gender. incidence of previ­ous myocardial infarction noted by either electrocardio­graphic mCG) criteria or history, or prevalence or severityof coronary artery disease. By virtue of fewer symptoms,patients in group I were taking fewer medications and atlower dosages than were those in group 2. Valvuloplasty inboth groups was accomplished with similar balloon sizes andnumber and duration of balloon inflations. Both groupsunderwent follow-up catheterization at similar intervals aftervalvuloplasty.

Acute hemodynamic changes. Hemodynamic variablesbefore and after valvuloplasty for all 49 patients are shown inTable 2. Heart rate, cardiac output and ejection fractionwere unchanged, but mean aortic gradient decreased andmean aortic valve area increased. Left ventricular end­diastolic pressure was unchanged, whereas end-diastolicvolume decreased, probably as a result of procedure-relatedblood loss. All diastolic indexes worsened, with an increasein peak -dP/dt. T and chamber stiffness and a decrease inpeak filling rate. Both peak and end-systolic wall stress werereduced as a result of a decrease in left ventricular systolicpressure.

ChanRes in both aortic and mitral valve regurgitationwere minimal after valvuloplasty. Mitral insufficiency wasnoted in 20 patients before valvuloplasty (3+ in 3 patients,2+ in 4 and I+ in 13); it was unchanged in 18 patients anddecreased by one grade in 2 after valvuloplasty. New mild(I +) mitral insufficiency was detected in two patients. Aortic

Age 75 • 7 yrs, Time to follow-up6.5 • 2.1 months

Age 73 • 6 yrs, Time to follow-up6.4 • 1.0 months

Group ISymptomatically improved(n=t6)

Group 2Recurrent symptoms(n=10)

(3) Hemodynamic Instability

(17) Technically Inadequate data

(10) Deaths

(3)AVR

Follow-upPre-PBAV

All PBAV(1<PBAV Study(1<Follow-up

(36)

l--)~ (10) Refused Cath

Repeat 6 month Cath(26)

4

a

IIIIIIoso~ 2.Q

j3

798 SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

Table 1. Baseline and Follow-Up Patient Characteristics

lACC Vol. 16. No.4October 1990:795-803

Follow-Up Catheterization

Age (yr)Men/womenSymptoms

CHF class III or IVCHF class I or IIAngina class III or IVAngina class I or IIPrevious MI

Coronary artery diseaseI vessel2 vessel3 vessel

MedicationsDiuretic drugsCalcium channel antagonistsNitrates

Valvuloplasty procedureBalloon size (mm)

202310 + 20

No. of inflationsMax duration (s)

Follow-upTime (mo.)Range (mo.)

InitialValvuloplastyAll Subjects

(n = 49)

74 ± 822/27

45 (92%)316%)

21 (43%)7 (14%1

II (22%)26 (53'7r.l]()

97

41 (84%)26153%)22 (45%)

40182%15 (10%)4 (8%)

3.9 ± 2.225 ± 10

Improved Symptoms(group I. n = 16)

73 ± 68/8

14188\7,)

8 (50\7,)

46 (38\7,)

3

II (70\7,16 138\7, 17144%1

14 (88%)I 16\7,)116%)

3.7 ± 2.431 ± 14

6.4 ± 1.0(5.1-9.2)

Recurrent Symptoms(group 2. n = 10)

75 ± 74/6

6160%)4140%)8 (80%)

o2 (20%)

5150%1I3I

101100\7,)6160%)8 180\7,)

8 (80%)2120\7,)

o3.2 ± 2.526 ± II

6.5±2.1(4.0-11.4)

Values for age. number and duration of inflations and follow-up time are mean values ± I SO. All other valuesrepresent number of patients. Balloon size = largest balloon used: CHF = congestive heart failure: class = NewYork Heart Association functional class: Max duration = duration of longest balloon inflation for the largest balloonused; MI = myocardial infarction.

insufficiency was noted in 42 patients before valvuloplasty(3+ in 2 patients, 2+ in 12 and I+ in 28); it increased by onegrade in 2 patients, decreased by one grade in 4 and wasunchanged in 36. New mild (1 +) aortic insufficiency wasnoted in two patients.

Hemodynamics at follow-up. Follow-up echocardio­graphic and hemodynamic assessment was performed in 26patients (16 in group 1and 10 in group 2). Of the 10 patientswho refused repeat catheterization, symptom status wassimilarly distributed as in the group consenting to catheter­ization, with 6 patients having symptomatic improvementand 4 having recurrent symptoms. Hemodynamic and echo­cardiographic findings immediately before or after valvulo­plasty were not different between those patients returningfor follow-up catheterization and those who did not. Thir­teen patients were not eligible for follow-up catheterizationbecause of either aortic valve replacement (n = 3) or death(n = 10). All 10 patients who died had a cardiac death, with8 dying because of progressive heart failure and 2 dyingsuddenly. Clinical and hemodynamic characteristics before

and after valvuloplasty were similar in all respects betweenthe group of patients who died and those who did not, exceptfor a lower baseline ejection fraction in the former (37 ± 13%versus 54 ± 17%. p < 0.05).

Tahle 3 ShOll'S hemodynamic and echocardiographic var­iahles hefore and after valvuloplasty and at follow-up cath­eterization in groups I and 2. Both groups were comparablein all respects before and after valvuloplasty and at follow-upcatheterization. except for lower T and chamber stiffnessvalues and higher peak filling rates in group I than in group2. The median percent changes from before valvuloplasty tofollow-up catheterization for aortic valve area and indexes ofleft ventricular mass and systolic and diastolic performanceare shown in Figure 3. No difference between the groups wasdetected in terms of the change in aortic valve area, cardiacoutput or ejection fraction. However, all diastolic indexeshad worsened in group 2, in contrast to the improvementnoted in group 1. This was also accompanied by a reductionin the left ventricular mass index in group I compared withgroup 2, without an accompanying change in wall thickness

lACC Vol. 16. No.4October 1990:795-803

SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

799

Table 2. Immediate Hemodynamic Changes After Valvuloplasty

'p < 0.05 compared with before valvuloplasty. -dP/dt = maximal rate ofdecrease of left ventricular pressure: EDWS = end-diastolic wall stress:ESWS = end-systolic wall stress: LV = left ventricular: LVlDd = leftventricular end-diastolic internal dimension: Kr = modulus of chamberstiffness: PSWS = peak systolic wall stress: 1 = time constant of isovolumet­ric relaxation.

in either group. Although not indicated in Figure 3, heartrate. mean aortic gradient. left ventricular systolic pressure.frequency or severity of valvular regurgitation. wall stressand chamber dimensions were equal to prevalvuloplastyvalues and were not different between the two groups.

DiscussionACllte Hemodynamic Cha/1/{es

Relation to previous studies. Changes in left ventriculardiastolic performance occurring during and after aortic val­vuloplasty have received little attention. Stoddard et al. (II)noted prolongation in 7' and a reduction in peak -dP/dtimmediately after valvuloplasty. Nishimura et al. (34) ob­served alterations in the Doppler diastolic flow pattern.suggesting restriction of mitral flow during balloon occlusionthat persisted after balloon deflation. The findings of ourstudy extend these observations by demonstrating worsen­ing of indexes reflecting all aspects of diastolic performance.including isovolumetric relaxation. rapid passive filling andchamber compliance shortly after valvuloplasty.

Potential mechanisms. Worsening of diastolic perfor­mance is likely a result of transient myocardial ischemia thatoccurs during balloon occlusion of the aortic outflow tract.This may occur by I) a decrease in aortic systolic and

diastolic pressures. resulting in a lowered coronary perfu­sion pressure and hence diminished myocardial oxygendelivery; and 2) an acute elevation in afterload that results inincreased myocardial oxygen demand (35-37). Decreases incoronary blood flow and an increase in myocardial oxygenuptake and lactate production have been reported to accom­pany valvuloplasty (36). The resultant myocardial "stun­ning" may take several minutes to days to resolve (38).Although in most cases transient balloon occlusion is clini­cally well tolerated. hypotension. ventricular arrhythmias.angina and myocardial infarction may all occur during andimmediately after valvuloplasty (1.6).

No significant change in ejection fraction was noted inthis study. Previous reports have noted both improvement(1-5) and worsening 05.36) of systolic indexes after valvu­loplasty. Because diastolic indexes may be more sensitive tothe effects of myocardial ischemia than systolic indexes(12-15), it is not surprising that diastolic indexes mayworsen. without a change in systolic indexes.

Potential limitations. Confounding factors to consider inevaluating changes in diastolic indexes include the effects ofcontrast injections. changes in heart rate. valvular insuffi­ciency. loading conditions and baseline left ventricular fillingpattern. Small volumes of low osmolarity contrast agentshave only minimal effects on both systolic arid diastolichemodynamics. particularly when assessed within six beatsof contrast injection (39-41). Heart rate al1d the frequencyand severity of valvular regurgitation were similar bothbefore and after valvuloplasty.

Loadil/R cOl/ditiol/s \\'ere, 110\\'el'er, altered by I'all'u/o­p/astv. Preload. as assessed by end-diastolic volume. andafterload. as assessed by end-systolic wall stress. were bothreduced after valvuloplasty. Although peak -dP/dt. peakfilling rate and chamber stiffness may be affected by changesin load. 7' appears to be less affected (42-45). Furthermore.Diver et al. (46) recently demonstrated that an isolatedreduction in preload and afterload. resulting from nitroprus­side infusion in patients with aortic stenosis. results in nochange in T or peak filling rate. despite lowering of bothend-diastolic pressure and volume.

A limitation of peak fillil/R rate as an index of diastolicpe,:!iml/al/ce in aortic stenosis is that although mild diastolicdysfunction may cause a decrease in early diastolic filling.early diastolic filling may be the greatest contributor to leftventricular filling when the most profound diastolic perfor­mance abnormalities exist (19,47,481. It is noteworthy thatalthough most patients exhibited a decline in peak filling rateafter valvuloplasty. all six patients who had an increase inpeak filling rate had elevated (>20 mm Hg) baseline leftventricular end-diastolic pressure (mean ± SD 30 ± 8 mmHg). Thus. an increase in peak filling rate in these patientsmay also signify worsening diastolic performance.

With respect to chamber stiffness. it is possible that theincrease in the modulus of chamber stiffness after valvulo-

77 :!: 140.8 :!: n.2*35 :!: 12*

180:!: 32*4.4 :!: 1.352 :!: 1724 :!: 10

AfterValvuloplasty

(n = 49)

20 :!: 9123 :!: 43*

1.473 :!: 421 *96 :!: 40*

226 :!: 78*0.141 :!: n.08)·

142 :!: 58*84 :!: 46*19:!: 14

1.4 :!: OJ4.4 :!: 0.8

76:!: 140.5 :!: 0.259:!: 17

209 :!: 384J :!: 1.251 :!: 1826:!: 10

BeforeValvuloplasty

(n = 49)

22 :!: 9130 :!: 42

1.602 :!: 42078 :!: 29

247 :!: gOn.107 :!: 0.071

171 :!: 7198:!: 5222:!: 161.4 :!: OJ4.5 :!: 0.8180:!: 48

Heart rate (beals/min)Aortic valve area (em 2

)

Mean aortic gmdient (mm Hg)LV systolic pressure (mm Hg)Cardiac output (liters/min)LVejection fraction (91)

Mean pulmonary pressure(mm Hg)

LV end-diastolic pressure (mm Hg)LVend-diastolic volume ImilPeak -dP/dt Imm Hg/s)rims)Peak filling rate (ml/s)Kp

PSWS (X 103 dynes/cm 2)

ESWS (X 103 dynes/cm~)

EDWS (X 103 dynes/cm~)

LV posterior wall thickness (em)LVlDd (em)LV mass index (g/m2)

800 SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

lACC Vol. 16, No, 4October 1990:795-803

Table 3. Immediate and Follow-Up Hemodynamic Changes in Patient Groups After Valvuloplasty

Group Pre Post Follow-Up

Heart rate (beats/min) 78 ± 14 79 ± 16 74 ± 142 76 ± 11 80 ± 13 77 ± II

Aortic valve area (em") 0.5 ± 0,1 0.8 ± 0.2* 0.6 ± 0.20.5 ± 0.1 0.8 ± 0.2* 0.5 ± 0.2

Mean aortic gradient (mm Hg) 65 ± 18 38 ± 13* 58 ± 172 61 ± 23 33 ± 11* 55 ± 18

LV systolic pressure (mm Hg) I 221 ± 40 187 ± 38* 223 ± 332 213 ± 30 188 ± 22* 217 ± 40

Cardiac output Oiters/min) I 4.7 ± 1.2 4.8 ± 1.5 4.8 ± 1.54.5 ± 0.9 4.4 ± 1.2 4.2 ± 1.2

LV ejection fraction (%) 59 ± 16 62 ± 15 59 ± 1456 ± 18 57 ± 14 52 ± 15

Mean pulmonary pressure (mm Hg) 27 ± 12 25 ± 10 24 ± II26 ± 10 23 ± II 25 ± 10

LV end-diastolic pressure (mm Hg) I 21 ± 7 19 ± 9 18 ± 82 19 ± 10 17 ± 8 21 ± 8

LVend-diastolic volume (ml) 128 ± 48 129 ± 50 124 ± 402 121 ± 34 107 ± 35 127 ± 33

Peak -dP/dt (mm Hg/s) 1.543 ± 461 1.377 ± 411* 1,838 ± 4392 1.678 ± 274 1,468 ± 355* L742 ± 471

rIms) 80 ± 29 106 ± 42* 69 ± 1668 ± 21 95 ± 47* 89 ± 34t

Peak filling rate (mI/s) 259 ± 99 232 ± 90* 290 ± 106258 ± 60 246 ± 69* 213 ± 47*t

Kp 0.092 ± 0.057 0.137 ± 0.100* 0.061 ± 0.035*0.084 ± 0.030 O.l30 ± 0.053* 0.127 ± 0.093*t

PSWS (X JOl dynes/em") 149 ± 51 122 ± 43* 160 ± 532 165 ± 55 137 ± 43* 167 ± 56

ESWS (X 10J dynes/em") 72 ± 38 60 ± 32* 76 ± 362 87 ± 36 76 ± 34* 94 ± 42

EDWS (X JOl dynes/em") 16 ± 10 14 ± 8 15 ± 92 18 ± 13 14 ± 9 19 ± 14

LV posterior wall thickness (em) 1.4±OJ 1.5 ± OJ 1.4±OJ2 1.4±OJ 1.4 ± OJ 1.5 ± OJ

LVIDd (em) I 4.2 ± 0.6 4.1 ± 0.6 4.2 ± 0.62 4J ± 0.8 4.2 ± 0.8 4.5 ± 0.6

LV mass index (g/m") I 173 ± 45 159 ± 40163 ± 45 174 ± 48

*p < 0.05 compared with before valvuloplasty in the same group: tp < 0.05 compared with group 1for specificvariable. Group 1 = improved symptoms (n = 16): Group 2 = recurrent symptoms (n = 10): Post = aftervalvuloplasty: Pre = before valvuloplasty: other abbreviations as in Table 2.

plasty may have been due to a smaller end-diastolic volume.This in itself would result in a steeper diastolic pressure­volume curve unrelated to the development of myocardialischemia, causing an impairment in chamber compliance(33,49),

Hemodynamic Changes at Follow-Up

Relation to previous studies. Consistent with previousreports (3-10), only 22 (44%) of the 49 study patients hadcontinued symptomatic improvement at follow-up catheter­ization. Symptom status at follow-up catheterization was notrelated to baseline clinical characteristics, baseline indexes

of aortic stenosis severity or hemodynamics or changes inaortic valve area and hemodynamics that resulted fromvalvuloplasty, Furthermore, as also previously reported(3-11), no differences between group I (improved symp­toms) and 2 (recurrent symptoms) in either mean gradient oraortic valve area at follow-up catheterization were noted,nor were changes from the postvalvuloplasty aortic valvearea to follow-up aortic valve area (restenosis) able todistinguish patients with alleviated symptoms from thosewith recurrent symptoms.

The inability to relate symptoms at follow-up catheteriza­tion to aortic valve area has led some (4) to attribute aplacebo effect to valvuloplasty. However, the discordance

JACC Vol. 16. No.4October 1990:795-1103

SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

801

provement has been accompanied by regression of hypertro­phy and improvement in both (ystolic and diastolic perfor­mance despite the presence of a significant residual stenosisacross most prosthetic valves (47,52-54). Left ventricularhypertrophy may regress as early as 3 to 5 weeks after valvereplacement (54). Not only may absolute wall thicknessdecrease. but ventricular remodeling, characterized bychanges in both wall thickness and chamber dimension, mayoccur (52-54). Diastolic performance, in particular, appearsto be closely related to ventricular mass and improves withregression of hypertrophy (17-19). Of note. a recent stuc'y bySafian et al. (8) indicated that some patients whose ejectionfraction improved 3 months after valvuloplasty also had adecrease in the ratio of wall thickness to chamber dimension.suggestive of early left ventricular remodeling.

Potential mechanisms. Valvuloplasty may improve dias­tolic performance by causing a decrease in afterload. result­ing in a decrease in wall stress. If maintained, a decrease inwall stress should promote regression of hypertrophy (25). Asecond beneficial effect that may contribute to improveddiastolic performance is a decrease in transvalvular aorticgradient. resulting in improved coronary blood flow andreserve as documented by Smucker et al. (35). A decrease inwall stress and improvement in myocardial ischemia maypromote left ventricular remodeling and further improve­ment in ventricular performance. In support of this mecha­nism. patients with improved symptoms had a decrease iventricular mass and chamber size. but not an appreciahledecrease in left ventricular wall thickness.

Precautions. Previous studies have emphasized that val­vuloplasty should be regarded as a palliative procedure forpatients who present a high operative risk or are unwilling toundergo aortic valve replacement. The results of this studyshould not be taken as an endorsement that valvuloplasty iscurative. Instead, these results indicate that there is ahemodynamic basis to explain why some patients mainlainsymptomatic improvement despite valve restenosis. Nota­bly. 6 months after aortic valvuloplasty. improvement insymptom status is only demonstrable in 44% of patients andno baseline or postvalvuloplasty variables predict whichpatients are likely to maintain symptomatic benefit. Further­more. although group I (improved symptoms) had a smalldecrease in left ventricular mass, this was largely due to areduction in cavity size and not a decrease in wall thickness.Also. because group I had a slight yet insignificantly higherbaseline left ventricular mass than group 2. this may haveresulted in a greater ability to detect a decrease in mass ingroup 1. Finally. at follow-up catheterization hath end­systolic and peak systolic wall stress had returned to base­line levels in group I. which may signify a continuedstimulus for further hypertrophy and loss of the hemody­namic and symptomatic benefit occurring 6 months aftervalvuloplasty. Further data are necessary to determine if

....AVA CO EF EDP EDV PFR TAU Kp PWT LVMI.80 .30 .30 .04 .04 .0001 .008 .001 .29 .004

o Grol4l1 (0-16)• GrOl4l 2 (n-l0)

o-10

-20

-30-40 l- -.J

+30

+60 r-----------------,+50

+40

pvalue

Change fromPre-PBAV 10 +20

Fonow-up (%) +10

Figure 3. Changes in hemodynamic and echocardiographic varia­bles from before valvuloplasty to follow-up catheterization. Barsrepresent median %change from before percutaneous aortic balloonvalvuloplasty (Pre-PBAV) to follow-up catheterization for groups Iand 2. Numbers shown below the variables represent p values thatcompare the significance of changes in group I versus group 2.AVA = aortic valve area; CO = cardiac output; EDP = end­diastolic pressure; EDV = end-diastolic volume: EF = ejectionfraction; Kp = modulus of chamber stiffness: LVMl = left ventric­ular mass index: PFR = peak filling rate: PWT = posterior wallthickness: TAU = time constant of isovolumetric relaxation.

between aortic valve area and symptom status may be due toseveral factors. One such factor may be that the Gorlinequation, by virtue of the use of a fixed constant andinaccuracy in the presence of low flow states and valveinsufficiency. is an imperfect measure of aortic stenosisseverity (50,51).

The relation of symptoms in aortic stenosis to left l'en­tricular filllction independent of aortic mIre area is \I"el/recof.:nized (/6-21). However. this study and others (3-11)indicate that changes in left ventricular systolic performancedo not correlate with symptom status after aortic valvulo­plasty. Although baseline clinical characteristics such as theextent and severity of coronary disease and left ventricularejection fraction were similar in groups I and 2 and nopatient had clinical. angiographic, hemodynamic or echocar­diographic evidence of a primary cardiomyopathy. left ven­tricular dysfunction due to a nonvalvular. nonischemic causecannot be excluded.

The primary finding of this study is that symptom status 6months after mlvuloplasty is closely related to c/wnRcs inleft ventricular diastolic pelfornwllce. This issue was alsoaddressed by Stoddard et al. (\ I). who found that diastolicperformance changes at follow-up study after aortic valvu­loplasty did not correlate with functional status. Their studywas, however, limited by the fact that diastolic performanceat follow-up evaluation was assessed by Doppler mitral flowindexes rather than hemodynamics. Data in our laboratory(48) indicate that Doppler mitral flow indexes do not relate tohemodynamic variables of left ventricular relaxation orchamber stiffness in patients with aortic stenosis.

Pertinent to the results of this study are those in patientsafter aortic valve replacement in whom symptomatic im-

802 SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

JACC Vol. 16, NO.4October 1990:795-803

hemodynamic and symptomatic improvement is maintainedbeyond this early follow-up period.

Conclusions. These results indicate that aortic valvulo­plasty is an effective palliative treatment for selected pa­tients with aortic stenosis. Immediately after valvuloplasty,diastolic performance worsens, probably because of tran­sient myocardial ischemia during balloon occlusion of theaortic outflow tract. Maintained symptomatic improvement6 months after valvuloplasty is closely related to improveddiastolic performance and not to systolic performance oraortic valve area. Adecrease in wall stress and improvementin myocardial ischemia, resulting in left ventricular remod­eling and regression of hypertrophy, are the likely mecha­nisms responsible for symptomatic and hemodynamic im­provement. Why some patients maintain symptomaticimprovement and others do not and the further course ofthose who are improved require further investigation.

References1. McKay RG. Safian RD. Lock JE, et al. Assessment of left ventricular and

aortic valve function after aortic balloon valvuloplasty in adult patientswith critical aortic stenosis. Circulation 1987:75: 192-203.

2. Cribier A, Savin T, Berland J. et al. Percutaneous transluminal balloonvalvuloplasty of adult aortic stenosis: report of 92 cases. J Am ColiCardiol 1987;9:381-6.

3. Block PC. Palacios IF. Clinical and hemodynamic follow-up after percu­taneous aortic valvuloplasty in the elderly. Am J Cardiol 1988:62:760-3.

4. Desnoyers MR. Isner JM. Pandian NG, et al. Clinical and noninvasivehemodynamic results after aortic balloon valvuloplasty for aortic steno­sis. Am J Cardiol 1988:62: 1078-84.

5. Litvack F, Jakubowski AT. Buchbinder NA. Eigler N. Lack of sustainedclinical improvement in an elderly population after percutaneous aorticvalvuloplasty. Am J Cardiol 1988;62:270-5.

6. Safian RD. Berman AD. Diver OJ. et al. Balloon aortic valvuloplasty in170 consecutive patients. N Engl J Med 1988;319:125-30.

7. Nishimura RA, Holmes DR Jr. Reeder GS, et al. Doppler evaluation ofresults of percutaneous aortic balloon valvuloplasty in calcific aorticstenosis. Circulation 1988:78:791-9.

8. Safian RD, Warren SE. Berman AD, et al. Improvement in symptoms andleft ventricular performance after balloon aortic valvuloplasty in patientswith aortic stenosis and depressed left ventricular ejection fraction.Circulation 1988:78: 1181-91.

9. Berland J. Cribier A, Savin T. Lefebre E. Koning R. Letac B. Percuta­neous balloon valvuloplasty in patients with severe aortic stenosis andlow ejection fraction. Circulation 1989:79: 1189-96.

10. Lewin R. Dorros G. King J, Mathiak L. Percutaneous transluminal aorticvalvuloplasty: acute outcome and follow-up of 125 patients. J Am ColiCardioI1989:l4:121O-7.

11. Stoddard M, Vandormael M, Pearson A, et al. Immediate and short-termeffects of aortic balloon valvuloplasty on left ventricular diastolic functionand filling in humans. J Am Coli CardioI1989;14:1218-28.

12. Aroesty JM, McKay RG. Heller GV. Royal HD. Als AV, Grossman w.Simultaneous assessment of left ventricular systolic and diastolic dys­function during pacing-induced ischemia. Circulation 1985:71:889-900.

13. Charlat ML, O'Neill PG, Hartley CJ. Roberts R. Bolli R. Prolongedabnormalities of left ventricular diastolic wall thinning in the "stunned"myocardium in conscious dogs: time course and relation to systolicfunction. J Am Coli Cardiol 1989:13:185-94.

14. Barry WHo Brooker JZ, Alderman EL. Harrison DC. Changes in diastolicstiffness and tone of the left ventricle during angina pectoris. Circulation1974:49:255-63.

15. Bertrand ME. Lablanche JM, Fourrier JL. Traisnel G. Mirsky I. Leftventricular systolic and diastolic function during acute coronary arteryballoon occlusion in humans. J Am Coli CardioI1988;12:341-7.

16. Stewart S, Mason DT. Braunwald E. Impaired rate of left ventricularfilling in idiopathic hypertrophic subaortic stenosis and valvular aorticstenosis. Circulation 1968;37:8-14.

17. Eichhorn P, Grimm J, Koch R, Hess 0, Carroll J. Krayenbuehl HP. Leftventricular relaxation in patients with left ventricular hypertrophy sec­ondary to aortic valve disease. Circulation 1982;7: 1395-404.

18. Fifer MA. Borow KM, Colan SO. Lorell BH. Early diastolic leftventricular function in children and adults with aortic stenosis. J Am ColiCardiol 1985:5: 1147-54.

19. Lavine SJ. Follansbee WP. Shreiner DP, Amidi M. Left ventriculardiastolic filling in valvular aortic stenosis. Am J CardioI1986:57: 1349-55.

20. Fuji J. Watanabe H. Koyama S. Kato K. Echocardiographic study ondiastolic posterior wall movement and left ventricular filling by diseasecategory. Am Heart J 1979;98:144-51.

21. Peterson KL. Tsuji J, Johnson A. DiDonna 1. LeWinter M. Diastolic leftventricular pressure-volume and stress-strain relations in patients withvalvular aortic stenosis and left ventricular hypertrophy. Circulation1978:58:77-89.

22. Sahn OJ, DeMaria A, Kisslo J, Weyman A. Recommendations regardingquantitation in M-mode echocardiography: results of a survey of echo­cardiographic measurements. Circulation 1978;58: 1072-83.

23. Weyman AE. Doty WD. Left ventricle. In: Weyman AE, ed. Cross­Sectional Echocardiography. Philadelphia: Lea & Febiger. 1982:293-300.

24. Folland ED. Parisi AF, Moynihan PF. Jones DR. Feldman CL, Tow DE.Assessment of left ventricular ejection fraction and volumes of real-time.two-dimensional echocardiography: a comparison of cineangiographicand radionuclide techniques. Circulation 1979:60:760-6.

25. Grossman W. Jones D. McLaurin LP. Wall stress and patterns ofhypertrophy in the human left ventricle. J Clin Invest 1975:56:56-64.

26. Quinones MA. Mokotoff DV. Nouri S. Winters WL Jr. Miller RR.Noninvasive quantification of left ventricular wall stress: validation ofmethod and application to assessment of chronic pressure overload. AmJ Cardiol 1980:45:782-90.

27. Reichek N. Wilson J, SI. John Sutton M, Plappert TA. Goldberg S,Hirshfeld JW. Noninvasive determination of left ventricular end-systolicstress: validation of the method and initial application. Circulation 1982:65:99-108.

28. Fifer MA. Grossman W. Measurement of ventricular volumes, ejectionfraction. mass and wall stress. In: Grossman W. ed. Cardiac Catheteriza­tion and Angiography. 3rd ed. Philadelphia: Lea & Febiger. 1986:282-300.

29. Gorlin R. Gorlin SG. Hydraulic formula for calculation of the area of thestenotic mitral valve. other cardiac valves and central circulatory shunts.Am Heart J 1951:41:1-13.

30. Grossman W. Profiles in valvular heart disease. In: Ref 28:359-81.

31. Weiss JL. Frederiksen JW. Weisfeldt ML. Hemodynamic determinants ofthe time-course of fall in canine left ventricular pressure. J Clin Invest1976:58:751-60.

32. Savitzky A. Golay MJE. Smoothing and differentiation of data bysimplified least squares procedures. Anal Chem 1964:36: 1627-38.

33. Gaasch WHo Battle WE, Oboler AA. Banas JS. Levine HJ. Left ventric­ular stress and compliance in man with special reference to normalizedventricular function curves. Circulation 1972:45:746-62.

34. Nishimura RA. Holmes DR Jr. Reeder GS. Tajik AJ. Hatle LK. Dopplerechocardiographic observations during percutaneous aortic balloon val­vuloplasty. J Am Coli CardioI1988:11:1219-26.

JACC Vol. 16. No.4October 1990:795-803

SHEIKH ET AL.DIASTOLIC FUNCTION AFTER AORTIC VALVULOPLASTY

803

35. Smucker ML. Tedesco CL. Manning SB. Owen RM. Feldman MD.Demonstration of an imbalance between coronary perfusion and exces­sive load as a mechanism of ischemia during stress in patients with aorticstenosis. Circulation 1988:78:573-82.

36. Rousseau MF. Wyns W. Hammer F. Caucheteux D. Hue L. Pouleur H.Changes in coronary blood flow and myocardial metabolism during aorticballoon valvuloplasty. Am J CardioI1988;61:1080-4.

37. De Lew JS. Pan M. Romero M. Sancho M. Carrasco L. Physiopathologyof transient ventricular occlusion during balloon valvuloplasty for pul­monic or aortic stenosis. Am 1 Cardiol 1988:61:436-40.

38. Braunwald E. Kloner RA. The stunned myocardium: prolonged. postis­chemic ventricular dysfunction. Circulation 1982;66: 1146-9.

39. Bettmann MA. Higgins CB. Comparison of an ionic with a nonioniccontrast agent for cardiac angiography: results of a multicenter trial.Invest Radiol 1985;20:S70-4.

40. Mancini GB1. Atwood lE. Bhargava V. Slutsky R. Grover M. HigginsCB. Comparative effects of ionic and nonionic contrast materials onindexes of isovolumic contraction and relaxation in humans. Am 1Cardiol1984:53:228-33.

41. Vine DL. Hegg JD. Dodge HT. Stewart OK. Frimer M. Immediate effectof contrast medium injection on left ventricular volumes and ejectionfraction. Circulation 1977:56:379-85.

42. Gaasch WHo Carroll JD. Blaustein AS. Bing OHL. Myocardial relax­ation: effects of preload on the time courSe of isovolumetric relaxation.Circulation 1986;73: 1037-41.

43. Starling MR. Montgomery DG. Mancini 1. Walsh RA. Load indepen­dence of the rate of isovolumic relaxation in man. Circulation 1987:76:1274-81.

44. Martin G. Gimeno JV. Cosin 1. Guillem MI. Time constant of isovolumicpressure fall: new numerical approaches and significance. Am J Physiol1984:247:H283-94.

45. Ishida Y. Meisner lS. Tsujioka K. et al. Left ventricular filling dynamics:influence of left ventricular relaxation and left atrial pressure. Circulation1986:74: 187-%.

46. Diver 01. Royal HD. Aroesty 1M. et al. Diastolic function in patients withaortic stenosis: influence of left ventricular load reduction. 1 Am ColiCardiol 1988:12:642-8.

47. Murakami T. Hess OM. Gage lE. Grimm J. Krayenbuehl HP. Diastolicfilling dynamics in patients with aortic stenosis. Circulation 1986:73: 1162­74.

48. Sheikh K. Bashore T. Kitzman O. et al. Doppler left ventricular diastolicfilling abnormalities in aortic stenosis and their relation to hemodynamicparameters. Am 1 Cardiol 1989:63: 1360-8.

49. Gaasch WHo Levine Hl. Quinones MA. Alexander lK. Left ventricularcompliance: mechanisms and clinical implications. Am 1CardioI1976:38:645-53.

50. Cannon SR. Richards KL. Crawford M. Hydraulic estimation of stenoticorifice area: a correction of the Gorlin formula. Circulation 1985;71:1170-8.

51. Segall. Lerner Dl. Miller DC. Mitchell RS. Alderman EA. Popp RL.When should Doppler-determined valve area be better than the Gorlinformula'? Variation in hydraulic constants in low flow states. J Am ColiCardiol 1987:9: 1294-305.

52. Monrad ES. Hess OM. Murakami T. Nonogi H. Corin Wl. KrayenbuehlHP. Time course of regression of left ventricular hypertrophy after aorticvalve replacement. Circulation 1988:77: 1345-55.

53. Schwarz F. Flameng W. Thormann 1. et al. Recovery from myocardialfailure after aortic valve replacement. 1 Thorac Cardiovasc Surg 1978;75:854-64.

54. Cody RJ. Stephens DD. Walker Hl. Boucher CA. Early post-operativechanges in left ventricular dimensions and mass following valve replace­ment in adults with aortic stenosis. J Cardiovasc Surg 1981;22: 19-27.