Ventricular pump performance in patients with obstructed right ventricular-pulmonary artery conduits

Ventricular obstructed conduits

pump performance in patients with right ventricular-pulmonary artery

Postoperative data were obtained from 18 patients with partially obstructed right ventricular (RV) to pulmonary artery (PA) conduits, who were studied 1 to 9 years following a Rastelli operation. Age at operation was 1 to 8 months in seven patients (group I: infant group) and 2 to 9 years in the remaining 11 patients (group II: childhood group). The diagnosis was pulmonary atresia in eight patients, truncus arteriosus in seven, and transposition of the great arteries with ventricular septal defect and pulmonary stenosis in three. Porcine-valved conduits were inserted in 17 patients and an aortic homograft in one. All but seven patients were free of symptoms at the time of postoperative study. Neither peak RV pressures nor RV to PA gradients were different between groups. RV ejection fraction (EF) was decreased in group II (0.43 -t 0.11) but was normal (0.80 ? 0.10) in group I. In addition, there was a significant inverse relationship between RVEF and age at repair (r = 0.714; p < 0.005). RV end-diastolic volume (EDV) was normal or increased in all patients and did not differ between the two groups. Left ventricular (LV) ejection fraction was also decreased in the older group (0.58 ? 0.10 vs 0.88 + 0.08; p < 0.05), and there was a decrease in RVEF and/or LVEF from pre- to postoperative studies in one of six group I patients compared with four of five group II patients. Radionuclide ventriculography with exercise was performed in eight patients (seven in group II) and RVEF responses were abnormal in all of them, whereas LVEF responses were abnormal in seven of eight. A decrease in RVEF of > 5 EF units with exercise was associated with a resting gradient from the right ventricle to the pulmonary artery of >50 mm Hg. These data indicate that RVEF is usually abnormal in patients with obstructed RV to PA conduits, who are operated upon after infancy. The progressive decrease in RVEF with age indicates the inadequacy of hypertrophy to normalize RV pump function with time. Further data after relief of RV outflow obstruction are needed to discern risk factors for permanent RV dysfunction in this clinical setting. (AM HEART J 1988;112: 1271.)

Imre Palik,

M.D., Thomas P. Graham, Jr., M.D., and Judith Nashuille, Tenn., and Budapest, Hungary

Patients with right ventricular (RV) to pulmonary artery (PA) conduits usually develop partially obstructed prostheses with resultant RV pressure overload which frequently is progressive. The effect of the increased RV pressure on ventricular function and outcome is unclear. Of particular interest in regard to planning optimal timing for reoperation is the time course of irreversible ventricular dysfunction. Therefore, the specific objectives of this investigation were to determine right and left ventricular size and pump function by means of biplane angiographic techniques and to attempt to correlate these From the Division of Pediatric Cardiology, Vanderbilt Center, and the Hungarian Institute of Cardiology. Received accepted

for publication May 2, 1086.

Reprint requests: Professor gy, Vanderbilt Ilniversity,

Jan.

13, 1986;

T. P. Graham, Medical Center

revision Division North,

University

received

March

Medical 24. 1986;

of Pediatric CardioloNashville, TN 37232.

Burger, R.N.

data with hemodynamic measurements and historic data in 18 patients studied following successful operative repair of RV-PA discontinuity. METHODS

There were 18patients studied at Vanderbilt University Hospital from October, 1973,to December, 1984,who had previous conduit repair for RV-PA discontinuity. All data were obtained during routine diagnostic cardiac catheterization performed becauseof clinical evidence of conduit obstruction. Such evidence included grade 4-616systolic murmur at the upper left sternal border and/or signs of systemic venous congestion or effort limitation. A Hancock porcine-valved conduit had been used in 14 patients, a Carpentier conduit in three, and an aortic homograft in one. Conduit size ranged from 10 to 22 mm in diameter. Infants lessthan 2 years of agewere premeditated with meperidine, 2 mglkg, and hydroxyzine. 1 mg/kg, intramuscularly, given at least 90 minutes prior to cineangiocardiography. Occasionally, additional intravenous meperidine 1271

1272

Table

Pal&

Graham,

and Burger

American

December 1966 Heart Journal

1. Preoperative patient data .- ____-

Patient No. Group 1

LVP

8 9 10

group):

2 wk 4mo 2 mo 6 mo 3 wk II

(childhood

group):

NA NA NA NA 30 25 NA

15 75 80 91 85 83 81

PAt, VSD Truncus I Truncus II Truncus I Truncus II Truncus II PAt, VSD

9515 85/5 85/19

85/l 2 loo/5 go/10

50 NA 30

80 72 83

T/F, BT shunt PAt, Waterston shunt TGA, VSD, PS, BT shunt PAt, VSD, BT shunt Truncus II, pulmonary artery band TGA, VSD, PS, BT shunt PAt, VSD, Potts shunt Truncus II, pulmonary artery band PAt, VSD, BT shunt TGA, VSD, PS, BT shunt PAt, VSD, BT shunt

after

infancy

3 wk 12 yr

loo/10 lOO/lO

lOO/lO 100/g

NA 50

61 89

13

6 yr

100/10

100/12

80

81

14 15

9 yr 5.5 yr

100/13 11016

120114 110/6

90 100

89 85

16 17

5.5 yr 3 mo

115/7 90/7

110/7 loo/9

NA 75

83 62

18

7 mo

NA

95/9

NA

74

LVP = left ventricular pressure; NA = not available; PAt = tetralogy of Fallot; TGA = transposition of the great arteries;

(0.25 mg/kg) was required during the procedure. In children older than 2 years, Innovar

(neuroleptic

combination

of droperidol and fentanyl) was used in a dose of 0.025 cc/kg (maximum 1.0 cc intramuscularly for premedication. Rarely, additional intravenous Innovar wasrequired during the procedure, in a doseone-fourth to one-half the original dose. Pressureswere obtained prior to cineangiocardiography with

the

use of side-hole

catheters

which

were

zero

referenced to midchest. Oxygen saturations were determined by meansof reflectance oximetry. RV volumeswere calculated from biplane cineangiocardiogramsfollowing injection into the right ventricle, right atrium, or superior vena cava. Left ventricular (LV) volumes were calculated following RV, PA, or LV injections. All data were calculated during sinus rhythm, and ectopic or postectopic beats were not used. In 15 of the 18 studies,

Diagnosis

in infancy

11 12

Abbreviations: BT = Blalock-Taussig; RVP = right ventricular pressure; T/F *Age at initial catheterization.

Or

10518 113/10 90/8 92/9 80/15 105/10 8014

Repair

1 yr 2 days 1 wk

Systemic

(75)

repair

(mm HAT)

llO/lO 115/15 95110 1 lO/lO 84/10 105/10 %I15

1.5mo

RV-PA

(mm Hg)

Primary

2 mo

2 3 4 5 6 7 Group

(mm Hg)

Age* I (infant

RVP

angiograms

were

suitable for analysis of end-diastolic and end-systolic volumes of both ventricles, while in three studies only ejection fraction (EF) data were available. Methodology for these calculations and normal values for comparison have been published previously.*.2 Patient ventricular volume data are expressedin milliliters/square meters or in terms of a percentage of predicted normal values. Values between 75% and 125% of predicted normal are within the normal range of variation for these variables.‘T2

= pulmonary atresia; PS = pulmonary VSD = ventricular septal defect.

stenosis;

Preoperative patient data and hemodynamics prior to conduit repair are shown in Table I. Primary diagnoses include severe tetralogy of Fallot or pulmonary atresia with ventricular septal defect in eight patients, truncus arteriosus in seven, and transposition of the great arteries with ventricular septal defect and pulmonary stenosis in three. These conditions were lumped together because they share the characteristics of systemic RV pressurepreoperatively and obstructed conduits postoperatively with the potential for permanent RV dysfunction. Patients with residual left-to-right ventricular shunts or LV outflow tract obstruction following conduit repair were excluded. Patients were divided into seven children with primary repair in infancy (group I) and 11patients with late repair after palliative procedures (group II). In Table II, the age at conduit repair, time from repair to recatheterization, ventricular pressures,resting gradients, and clinical condition of the patients at recatheterization are shown. Three patients in group I and two patients in group II had mild-to-moderate aortic regurgitation as estimated from aortic root angiograms(mildfilling only part of the left ventricle and clearing in one or two beats; moderate-filling the entire left ventricle and taking more than two beats to clear). All seven patients in group I have had reoperation and have done well. AI1 11 patients in group II have had

Volume Number

Table

112 6

Ventricular

II. Postoperative

Patient NO.

patient

Group 1 2 3 4 r

I: Infant

ii 7 Group 8 9 10 11 12 13 14 15 16 17 18

82mo ma 1 mo II: Childhood group 2.5 yr 2.5 yr 8 Yr 4 yr 14 yr k-5Yr 9 yr 6 Y= 5.5 yr 6 yr 10 yr

CHF *Age tTime LDied BDied

4 2 3 4

and

conditiun

Symptom

free

92 93

Symptom Symptom SympLonr Symptom Symptom Svmpl OIII

free free free free free free

93 95 95 94 98 88 85 93 92 90 96

sympm free Tired easily Severe VHF1 Tiretl easily5 I‘ired easilyi ‘Tire4 easily Sym))tom l’ree Symptom free Symptom free (‘HF. t.ired easily 7’irec: easily%

8 2.5 5.4 4

115115 lOO/lO 11019 120113

125110 lOO/lO 7518 105/8

100 80 45 15

91 90 92 95

26.5 6.5

90/12 9516 92112

9516 60/5 9019

44 70 35

8 8.5 1 3.4 2 6 3 8 9 4 4.2

8015 135/11 125134 120117

9014 9518 70115 2OOl20 11514 135/22 115118 240110 120/10 101117 92112

70 62 10 170 x2 110 45 200 60 -. 66

145/17 125118 145111 13015 11716 100/2

.-.. -I_-

f’linkai

Svstemic . (%i

RVP (mm Hg)

OI

1273

wnduits

--.-.

RV-PA (mm Hgj

LVP (mm Hd

group mo mo mo mo

obstrwted

data

Time? (Yr)

Age*

performance

.---ut

= congestive heart failure; for other abbreviations, see Table I. at conduit surgery. elapsed from repair to postoperative catheterization. after reoperation. at reoperation.

Table III. Statistical ______~

comparison

of postrepair

cardiac

catheterization

data for groups

I and I1 -~

Data

Group

I

Group

..~~

II __-._

Age at repair (yr) Age at postrepair

catheterization

0.3 5.4 5.1 0.60 0.68 116 145 93 8.0 103 10.3 64.2

(yr)

Time after repair (yr) RVEF

LVEF RVEDV (I’:. normal) LVEDV (S normal) RVP (mm Hg) RVEDP (mm Hg) LVP (mm Hg) I,VEDP RV-PA

(mm Hg) gradient (mm

Hg)

All values are mean t SD; NS (not significant)

=p >

0.05

by unpaired

It_ 0.2 * 2.3 If: 2.2 k 0.10 f 0.08 + 34 + 35 2 27 2 1.9 +- 12 * 2.4 rt 24.1

6.4 11.6 5.2 0.43 0.56 119 159 128 12.8 125 14.9 81.0

+ _t i i ++ + -+ + t t i-

3.6 3.6 3.5 0.11 0.10 29 82 53 6.5 20 8.6 66.2

- ..--.-... p Value

_

.._

-


-co.01

--~

---..

<0.05 NS NS NS NS co.025 NS NS “_ ___._ ~_ ._.

t test for all data.

reoperation; there have been four deaths, two at reoperation and two from 6 to 24 months following operation with signs of abnormal ventricular function. These four deaths all occurred in symptomatic patients. Eleven patients also had EF data obtained by means of radionuclide techniques described previously.3 Rest and exercise data were available for the left ventricle in eight patients and for the right ventricle in seven patients. The lower limit of normal for resting LVEF is 55% and for RVEF is 35%, with an increase in EF of >5% (e.g., from 55% to 60%) the normal exercise response for both ventricles.

Comparison between groups was by nonpaired t test with p < 0.05 considered significant with two-tailed analysis. Regression analysis was by simple linear regression equation techniques. RESULTS

Fig. 1 illustrates postrepair EF data for both patient groups. The left panel shows RVEF data with the infant repair group having a normal value which was significantly greater than the childhood repair group. For the infant group RVEF ranged

1274

Palik, Graham, and Burger

American

1.00

i

0.8

t

0.6

LV.E.F.

R.V.6.F.

December 1986 Heart Journal

LEGUlD

0.60 o.:o P-ml

0.43

“3 g

0.4

b k 2

0.2

Fig. 1. Mean + SD for postoperative right and left ventricular ejection fractions for patients with infant vs childhood conduit operation.

G.“.E.F.=69.6-2.317fAG6

1.00 1

PREOP OX

IN VRI

6.E.lr7.8%

0.2 1 I

0.

1

2

4

6

i

10

AGE OF POSTOPER*TIVE

12 CA7WMEMZA7IlM

14

1’6

16

20

IVG)

Fig. 2. Right ventricular ejection fraction as a function of age at postoperative catheterization.

between 0.47 and 0.70 and only one of seven had a value below our lower limit of normal of 0.49.2 For the older group RVEF ranged between 0.19 and 0.54, and 7 of 11 had individual values lower than normal. Data for LVEF are shown in the right panel with the infant group again showing a higher value. For the infant group LVEF ranged between 0.60 and 0.79, and all were above our lower limit of noraml of 0.55.’ For the older group, LVEJD ranged between 0.45 and 0.70 with 4 of 11 less than 0.55. A statistical comparison of postrepair cardiac catheterization data for the two patient groups is shown in Table III. The time from repair to postoperative catheterization was not different between the two groups, but age at repair and age at postrepair study were lower for the primary repair group. Although mean values for right and left heart pressures as well as RV to PA gradient were less for the primary repair group, only the peak LV pressure difference was statistically significant.

Fig. 2 shows RVEF as a function of age at postrepair catheterization. There is a significant inverse relationship between these two variables as shown. In addition, a similar inverse relationship was found between RVEF and age at repair (r = 0.714; p < 0.005; SEE +- 0.085). Linear regression analysis also was used in an attempt to find a relationship between RVEF and RV peak pressure, end-diastolic pressure, or RV-PA conduit gradient. None of these variables was related significantly to RVEF. Four patients in the infant group and five patients in the older children group had both pre- and postoperative RVEF data available (Fig. 3). None of the infants but three of five older patients had a fall in RVEF (change >0.05 EF unit) between studies. Six infants and four children had LVEF data with one of six infants and all four older children showing decreases in EF from pre- to postoperative study. Radionuclide RVEF data are shown in Fig. 4 for seven patients who had both rest and exercise

Volume Number

112 6

Ventricular

performance

.

TRIJNCUS

1.

PULWfW+Ry

LVLF.

R.V.E.F.

and obstructed

ATRESIA

c~)nduits

1275

I !

i 0.

GROUP!

-

GRoupI

OROUP II

Fig. 3. Right and left ventricular ejection fractions obtained prior to and following initial conduit

operation for infant (group I) and childhood (group II) repair groups.

RADlDWCuM

E.F. DATA

1.00

ATtnJNc4Js

-7

a

--.-I

E 0.8 2 u.

i, r’ 0.

2

4

6

8

?r 10

12

c

14

16

18

I 20

AGE Al STUDY WR)

Fig. 4. Radionuclide right ventricular ejection fractions at rest and at peak exercise.Arrows

connect rest

and exercise data that were obtained at the samestudy.

evaluation. Only one of these patients was in the infant repair group. Three of seven had a marked decrease in EF with exercise, while four of seven showed no change (change from rest to exercise <0.05 EF units). Thus all seven patients had an abnormal exercise response. Radionuclide EF data are plotted in Fig. 5 in a similar manner for the left ventricle for eight patients. Six patients showed no significant change in LVEF with exercise, while one patient increased from 0.62 to 0.70 and one decreased from 0.46 to 0.51. Fig. 6 shows a change in RVEF with exercise as a function of the RV-PA gradient obtained at catheterization prior to the exercise study for five patients in whom these data were available. There is a significant inverse relationship between these variables with higher gradients associated with a greater

decrease in RVEF with exercise. In addition, a decrease in EF of >5 EF units is associated with an RV-PA gradient of >50 mm Hg. DISCUSSION

Progressive RV hypertension following RV to PA conduit placement is a common problem in children following a Rastelli type of operation.4-10 The causes of RV pressure overload include porcine valve failure, intimal peel of the conduit, compression or abnormal conduit angulation, and rapid childhood growth coupled with a fixed conduit orifice. This known natural history of conduit surgery mandates reoperation which optimally should be performed before permanent RV dysfunction occurs. The clinician frequently is confronted with the problem of an asymptomatic child with a narrowed conduit and significant RV pressure overload. Both patient size

1276

Palik, Graham, and Burger

American

RADIOWCLIDE

December 1986 Heart Journal

EF. DATA

l.OCI-

“f 5

0.4

x 0.2

I P

0.

L

2

4

6

8

10

12

14

16

18

1 20

AQEATSTlKWNIi)

Fig. 5. Radionuclide left ventricular ejection fractions at rest and at peak exercise.

Fig. 6. Changein right ventricular ejection fraction with exerciseas a function of resting right ventricular

to pulmonary artery gradient (RVEF = 0.2599- 0.0787 [gradients]; SEE = 2.87; r = 0.825;p < 0.05).

and projected future growth are confounding factors in the decision regarding a delay in reoperation. To date, there are no data regarding the use of either an invasive or noninvasive test to determine optimal timing for reoperation. Our study was performed in an attempt to characterize the effect of obstructed conduits on right and left heart volumes and EFs. Patients with three separate congenital defects were studied together because they have the common characteristics of RV pressure overload secondary to conduit obstruction and operative closure of a ventricular septal defect. This approach appears reasonable since the data indicate similar findings for all patient groups and can be seen in Figs. 2 to 5 where preoperative diagnoses are indicated.

The results show a clear difference in RVEF between patients in whom repair was done in infancy and those in whom repair was done later in childhood. This difference could not be directly correlated statistically to differences in RV-PA gradient, RV peak pressure, or time elapsed between operation and restudy, since these variables were not significantly different between groups. Possible causes for the differences in RVEF include a longer period preoperatively for RV fibrosis to occur as a result of chronic hypoxemia and intermittent myocardial hypoxia during times of oxygen supplydemand imbalance, the chronic effects of prior shunting producing volume overload with subsequent dysfunction, and a difference in myocardial damage at the time of surgery resulting from inade-

Volume Number

112 6

quate myocardial protection, A difference in myocardial protection could well have been present, since six of seven infants had repair performed with deep hypothermia and total circulatory arrest, whereas all of the older patients had repair by means of moderate hypothermia and cardiopulmonary bypass. In addition, the majority of older children showed a decrease in both RVEF and LVEF when pre- and postoperative values were compared, while only one infant showed such a decrease (Fig. 3). These data suggest a detrimental effect on ventricular performance as a result of inadequate myocardial protection in the older group, although such an effect remains unproved because of the multiple potential factors involved. The use of EF alone to characterize RV function is limited by the known effects of preload, afterload, and myocardial hypertrophy on this variable. In the present study, an estimation of preload in terms of either RV end-diastolic volume or end-diastolic pressure revealed no differences between the two groups. Afterload is a more difficult variable to estimate for the right ventricle, but peak RV pressure was not different between the two groups. End-systolic stress has not been quantified for the right ventricle, but may well have been higher for the older group, since peak RV pressure and endsystolic volume were greater in the older children. Thus an increased afterload in the older patients may have been partly or wholly responsible for the lower EF. Nevertheless, the younger patients were able to adapt to the abnormal RV pressure loading and by appropriate hypertrophy achieved normal RVEF and cardiac output, while the older patients were unable to achieve normal RV pump function. The abnormal LV pump function in older children suggests an overall depression of myocardial performance related to chronic volume overload, chronic hypoxemia, or differential effects of intraoperative myocardial protection. The ventricular functional abnormalities also correlated well with clinical findings. All seven patients operated upon in infancy were well prior to reoperation and have done well with repeat surgery. In contrast, 7 of 11 older patients were symptomatic prior to reoperation; two died at reoperation with low cardiac output, and two died later following conduit replacement with signs of impaired ventricular function. Evaluating any individual patient for conduit replacement, however, remains difficult as illustrated by two of our older patients. The first boy (No. 15, Table I) had an initial conduit operation at 6 years and when he was restudied 8 years later he had a RV pressure of 240/10 mm Hg but a normal

Ventricular

performance

and obstructed

imduits

1277

RVEF of 0.54. He subsequently had uneventful conduit replacement and is now asymptomatic with normal exercise tolerance and normal supine bicycle exercise test results. The second illustrative patient (No. 11, Table I) had initial conduit surgery at age 4 years with restudy 3% years later at which time his RV pressure was 200/20 mm Hg with a slightly decreased RVEF of 0.44. He died at reoperation with low cardiac output and clinical signs of ventricular dysfunction. The first patient did not appear to have suffered RV functional abnormalities despite a higher RV peak pressure presumably over a longer period of time than the second patient described. Frequent follow-up with recatheterization studies when clinical evidence for RV pressure overload exists seems the most prudent management option for these patients at the present time. Radionuclide studies may prove useful as a partially noninvasive test to detect early evidence of RV dysfunction. Our data on changes in RVRF with exercise (Fig. 6) suggest that a decrease in this variable by >0.05 EF units is usually associated with a prominent RV pressure overload and thus warrants invasive reinvestigation. Attempts to estimate gradients noninvasively with Doppler studies have not been convincing to date because of difficult and/or uncertain angles of incidence of the ultrasound beam with the turbulent jet. In summary, we have demonstrated impaired RVEF and LVEF in patients with partially obstructed RV to PA conduits, who had repair in childhood after initial palliative operation, whereas patients with primary repair in infancy showed normal EFs. Multiple factors probably account for this difference, including differences in myocardial protection at the time of operation, duration and degree of cyanosis prior to surgery, and the ability of the young patient to hypertrophy the right ventricle sufficiently to compensate adequately for an abnormal pressure load. The study indicates a need for further follow-up studies in patients prior to and after reoperation for conduit obstruction to determine the incidence, severity, and risk factors for permanent myocardial performance abnormalities. We acknowledge the tireless efforts of Mrs. Ms. Joy Phillips in preparing this manuscri~~t.

Anne

Kruger

and

REFERENCES

1. Graham TP Jr, Jarmakani JM, Canent KV Jr, Morrow MN. Left heart volume estimation in infamy and childhood. Reevaluation of methodology and normal values. Circulation 1971;43:895-904. 2. Graham TP Jr, Jarmakani JM, Atwood GF, Canent RV Jr. Right ventricular volume determinations in children. Normal values and observations with volume or pressure overload. Circulation 1973;47:144-53.

December

Palik, Graham, and Burger

American

3. Parrish MD, Graham TP Jr, Born ML, Jones J. Radionuclide evaluation of right and left ventricular function in children: Validation of methodology. Am J Cardiol 1982;49:1241-7. 4. Norwood WI. Freed MD. Rocchini AP. Bernhard WF. Castaneda AR. Experience with valved conduits for repair of congenital cardiac lesions. Ann Thorac Surg 1977;24:223-32. 5. Ciaravella JM Jr, McGoon DC, Danielson GK, Wallace RB, Mair DD. Experience with the extracardiac conduit. J Thorac Cardiovasc Surg 1979;78:920-30. 6. Geha AS, Laks H, Stansel HC Jr, et al. Late failure of porcine valve heterografts in children. J Thorac Cardiovasc Surg 1979;78:351-64. 7. Saravalli OA, Somerville J; Jefferson KE. Calcification of

1966

Heart Journal

aortic homografts used for reconstruction of the right ventricular outflow tract. J Thorac Cardiovasc Surg 1980;80:90920. 8. Fishbein MC, Levy RJ, Ferrans VJ, et al. Calcification of cardiac valve bioprotheses. Biochemical, histologic, and ultrastructural observations in a subcutaneous implantation model system. J Thorac Cardiovasc Surg 1982;83:602-9. 9. McGoon DC, Danielson GK, Puga FJ, Ritter DG, Mair DD, Ilstrup DM. Late results after extracardiac conduit repair for congenital cardiac defects. Am J Cardiol 1982;49:1741-49. 10. Vergesslich KA, Gersony WM, Steeg CN, et al. Postoperative assessment of porcine-valved right ventricular-pulmonary artery conduits. Am J Cardiol 1984;53:202-5.

Beneficial effect of taurine in rabbits chronic congestive heart failure

with

To examine the effect of daily treatment with taurine on improving the status of congestive heart failure (CHF), we used rabbits with artificially induced aortic regurgitation. Ten rabbits were treated daily with taurine (100 mg/kg by mouth) and eight with guanidlnoethyl suifonate (GES) (100 mg/kg by mouth) immediately after induction of aortic regurgitation. The cumulative mortality rate at 8 weeks in the taurine-treated CHF group was 10% (1 of 10) compared with 53% (18 of 30) in the nontreated CHF group and 75% (8 of 8) in the GES-treated CHF group (p < 0.05). Although cardiac function (max dP/dt) in CHF rabbits was significantly decreased (p < O.OOl), taurine-treated CHF rabbits maintained the same values as control rabbits. Taurlne content of the left ventricular tissue of the CHF rabbits was significantly increased (P < 0.01). Administration of taurine and GES to control rabbits for 8 weeks affected neither the hemodynamics nor the taurine content of the heart. It was concluded that taurine slowed the rapid progression of heart failure and consequently prolonged life expectancy. (AM HEART J 1988;112:1278.)

Keiko Takihara, M.D., Junichi Azuma, M.D., Nobuhisa Awata, M.D., Hideyuki Ohta, M.D., Tomoyuki Hamaguchi, M.D., Akihiko Sawamura, M.D., Yuya Tanaka, M.D., Susumu Kishimoto, M.D., and Nick Sperelakis, Ph.D. Osaka, Japan, and Cincinnati,

Ohio

Taurine (2aminoethanesulfonic acid) is the most abundant free amino acid in mammalian heart tissue, making up approximately 50% of the free amino acid pool. The heart contains the highest From the Third Department of Internal Medicine, Osaka University Medical School, and the Department of Physiology and Biophysics, University

of Cincinnati

Supported HL-31942

in part by the from the National

Received accepted

for publication May 5, 1986.

Reprint requests: Medicine, Osaka 553, Japan.

1278

Junichi University

Medical

Center.

Uehara Memorial Foundation Institutes of Health. Jan.

17, 1985;

revision

received

and

by grant

April

1, 1986;

Azuma, M.D., Third Department of Internal Medical School, l-l-50, Fukushima, Osaka

concentration of taurine within the body.’ Although articles have been published on the function of taurine in various organs, its physiologic function in the heart has not been defined. Taurine has been found to potentiate the inotropic action of digitalis glycosides,2 prevent cardiotoxicity caused by elevated concentrations of digoxin and epinephrine,3 and antagonize the negative inotropism of low calcium concentrations4 and the positive inotropism of high calcium concentrations.6 Other potential functions of taurine include osmoregulation of the heart, and prevention of the calcium paradox7z 8 and the severity of lesions in cardiomyopathic hamsters.g The taurine content of the heart varies in differ-