The Echocardiographic Assessment of Pulmonary Artery Pressure and Pulmonary Vascular Resistance

Transcription

1 The Echocardiographic Assessment of Pulmonary Artery Pressure and Pulmonary Vascular Resistance By STEPHEN HIRSCHFELD, M. D., RICHARD MEYER, M. D., DAVID C. SCHWARTZ, M.D. JOAN KORFHAGEN, U.T., AND SAMUEL KAPLAN, M.D. SUMMARY Serial assessment of the status of the pulmonary vascular bed requires repeat cardiac catheterization. We have demonstrated that right ventricular systolic time intervals (RVSTI) may be measured from the pulmonary valve echo. The right ventricular ejection time (RVET) and right pre-ejection period (RPEP) were measured in normal patients. The RVET and RPEP decreased with increasing heart rate but increased with age. The RPEP/RVET, however, was uninfluenced by either age or heart rate. The RPEP/RVET was, therefore, determined from the pulmonary valve echo in 64 patients with congenital heart disease who underwent cardiac catheterization. Increased pulmonary artery diastolic pressure (PADP), pulmonary vascular resistance (PVR) and mean pulmonary artery pressure (MPAP) resulted in an increased RPEP/RVET. The use of the RPEP/RVET permitted the serial echographic evaluation of the pulmonary vascular bed in selected patients; marked elevation of the ratio indicated the presence of pulmonary hypertension. Downloaded from by on October 5, patients WITH LARGE left-to-right shunts, such as ventricular septal defect (VSD), patent ductus arteriosus (PDA), atrioventricular canal (AVC), and atrial septal defect (ASD), as well as patients with transposition of the great vessels (TGV), with and without a VSD, may develop hypertensive pulmonary vascular disease. -4 It has been shown that patients with severe pulmonary vascular obstructive disease complicating their congenital heart lesion are high risk surgical candidates and survive longer if corrective surgery is not undertaken.5' 6 It is, therefore, important to select patients for surgery prior to the development of advanced pulmonary vascular changes. Patients with congenital heart disease, who are at risk to develop pulmonary vascular disease, may require repeat cardiac catheterization to permit serial evaluation of pulmonary vascular resistance (PVR) and pulmonary artery diastolic pressure (PADP) in order to determine the timing of corrective surgery. The advantage of a noninvasive technique that would permit the serial assessment of the pulmonary vascular bed and indicate the presence of pulmonary hypertension is easily appreciated. To date, the noninvasive assessment of the pulmonary vascular bed has been technically difficult. We From the Department of Pediatrics, College of Medicine, University of Cincinnati and Children's Hospital, Cincinnati, Ohio. Supported by NIH Grant 5 TOI HL052 and the Southwestern Ohio Chapter of the American Heart Association. Address for reprints: Richard Meyer, M.D., Division of Cardiology, Children's Hospital, Cincinnati, Ohio 229. Received January. 95; revision accepted for publication May have described the measurement of right ventricular systolic time intervals (RVSTI) from the pulmonary valve echo.9 In addition, the RVSTI of patients with TGV demonstrated marked prolongation of the right ventricular pre-ejection period (RPEP) and shortening of the right ventricular ejection time (RVET) in response to the increased afterload of systemic blood pressure.9 Leighton et al.'0 measured RVSTI at cardiac catheterization and noted that the duration of RPEP correlates with changes in PADP. The purpose of this study was to observe the relationship of PADP, PVR, and mean pulmonary artery pressure (MPAP) to the RVSTI derived from the pulmonary valve echo, and to determine if alteration of the ratio RPEP/RVET permits the noninvasive assessment of PADP, PVR, and MPAP. Methods The echocardiograms were obtained with a Hoffrel lolb ultrasonoscope. Strip chart recordings of the echocardiograms were obtained with a Cambridge Multichannel Physiological Recorder, Amplifier Type 252. A 5.0 MHz, /4 inch diameter, unfocused transducer was used to record the pulmonary valve echo in infants less than six months old. In older children a 2. MHz, /4 inch diameter transducer, focused at 5 centimeters, was employed. The pulmonary valve echo was recorded from the second or third intercostal space with the sonar beam aimed laterally and superiorly." In infants the pulmonary artery was located by rotating from the aorta to the pulmonary artery without moving the transducer. RVSTI were measured as previously described by employing the pulmonary valve echo and electrocardiogram (ECG).9 RVET was measured from pulmonary valve opening to closure. The timing of the onset and termination of right ventricular ejection was facilitated when the anterior and posterior pulmonic cusps could be recorded

2 Downloaded from by on October 5, ECHO/PULMONARY VASCULAR BED simultaneously (fig. ). However, when only the posterior pulmonic cusp could be recorded, the onset of ejection was chosen as the point of rapid posterior leaflet motion at which the closed valve echo changed from a thick line to a very fine one (fig. 2A). The termination of ejection was determined by the junction of the fine leaflet echo with the thick closed valve echo following rapid anterior motion. RPEP (LPEP in TGV) was measured from the initial ventricular depolarization of the ECG, usually the Q wave, to the onset of pulmonic valve opening (figs. and 2). Validation for this method of measuring RVSTI has been accomplished by demonstrating that RVET measured from a simultaneous, direct pulmonary arterial pressure tracing and pulmonary valve echo differed by less than 5 msec in ten patients.9 RVSTI derived from our echocardiographic tracings were measured at a paper speed of 5 mm/sec and time lines of msec intervals. Measurements were made during the expiratory phase of respiration in older children. In infants with a rapid respiratory rate the shortest STI were selected to minimize the influence of respiratory variation. Five separate complexes were averaged to obtain the final RVSTI. The RPEP/RVET measured from echo (figs. and 2) was correlated with the PADP, PVR, and MPAP recorded at cardiac catheterization. The pressure recordings were obtained with either a 5 F, cm Swan-Ganz flow-directed infant angiography catheter,* a standard 5 F, cm Swan-Ganz flow-directed catheter,* or a 6 or F, cm NIH thin walled, woven dacron catheter.f The catheter was connected to a P2Db Statham strain gauge, and the pressure *Edwards Laboratories, Santa Ana, California tusci, Billerica, Massachusetts was recorded on an Electronics for Medicine optical recorder. The PVR was calculated in units of mm Hg/L/min/m2 by employing assumed oxygen consumptions as reported by Cayler et al. for infants and LaFarge and Miettinen' for children three years of age or older. Premedication for the catheterization was droperidol, 0. mg/kg (maximum 5 mg), and atropine, 0.0 mg/kg for all children older than months. Younger children were not premedicated. Anesthesia, when necessary, was maintained with intravenous ketamine hydrochloride, mg/kg. Material Two groups of patients were evaluated. Group I Forty-five normal patients, 24 females and 2 males, age two months to years, were studied echocardiographically to measure RPEP and RVET (table ). These patients did not undergo cardiac catheterization and gave their consent to be studied by echocardiography as part of the evaluation of an innocent murmur. Group II 64 Sixty-four patients, 26 females and males, age one month to 2 years, with a variety of congenital heart diseases, underwent elective cardiac catheterization and echographic examination (table 2). Echocardiograms were performed on the day of catheterization in 5 patients and within one month in four patients. Three patients (cases 59,, 64) with Eisenmenger's syndrome documented by cardiac catheterization, who had remained clinically stable, were studied -5 years after catheterization. Figure Right ventricular STIs are measured from the anterior and posterior pulmonic cusp echoes. The RPEP/RVET calculated from the pulmonary artery echo is 0 in a patient with a PADP of mm Hg, PVR of.5 units, and MPAP of mm Hg. PA = pulmonary artery. Time lines are msec.

3 Downloaded from by on October 5, * ***l:5*e m-;t Figure 2 : * HIRSCHFELD ET AL. :... A) RVSTI are measured from the posterior pulmonic cusp echo in a patient with pulmonary hypertension. The RPEP/RVET is 0.9 in a patient with a PADP of 50 mm Hg and MPAP of 0 mm Hg. B) LVSTI are measured from the transposed pulmonic cusp echo in a patient with TGV and VSD. The LPEP/LVET is 0. with a PADP of mm Hg and MPAP of mm Hg. Time lines msec. Twenty-two patients had a VSD, six an AVC, an ASD, and six a PDA. Six patients had pulmonic stenosis (PS), and three had a VSD and PS, but their dominant lesion was right ventricular obstruction. Six patients had simple TGV, and patient 4 had previously undergone a Mustard procedure. Four patients had TGV and a large VSD. In patients with TGV, left, rather than right, STI were measured since the LV reflected hemodynamic changes in the pulmonary vascular bed (fig. 2B). No patient had complete right bundle branch block, as judged by electrocardiogram and vectorcardiogram, or incompetence of the tricuspid or pulmonic valve. Although patients (cases 2,, 6-4,, 2,, 4, 4, 46) were receiving digoxin and diuretics for congestive heart failure at the time of the study, they were adequately compensated. Sixteen patients (cases, 2, 6-9,,, 4,, 2, 2, 4, 55, 5, 6) had two to five echo studies over a period of three months to one year. Six of the repeat studies (cases 6, 9,,, 55, 6) were performed while breathing room air and repeated during oxygen inhalation. Patients 2 and 2 were studied prior to and after banding of the pulmonary artery. Results Group - Normal Patients (table ) Heart rate and age influenced RPEP and RVET in such a manner that RPEP and RVET shortened with increasing heart rate and decreasing age. The mean RPEP/RVET was 0 (range 0.-0.) and was not significantly altered by age or heart rate (figs. and 4). Group 2 - Patients with Congenital Heart Disease (table 2) The ratio RPEP/RVET was compared to the PADP, 0t 50 F F.. " I * t *. \ * *.o k - 0. => t * " ^ * { * * * * t * * \ ^ k 'In 90IIU 0 a HEART RATE Figure 0 0 The RPEP/RVET, RPEP and RVET are plotted against heart rate in normal patients. The RPEP and RVET decrease with increasing heart rate, while RPEP/RVET is uninfluenced. g

4 Downloaded from by on October 5, ECHO/PULMONARY VASCULAR BED Table Echographic RVSTI in Normal Patients Age QP RPEP RVET RPEP Patient (y rs) HR (msec) (msec) (msec) RVET / 2/ 2/ 2 / 2/ 2/ / 5/ 9/ ], 2/ 6/ / /2 4./2 4 /2 6 4/ 6/ 9 9 6/ ; ; , ( , Abbreviations: QP = Iight venitricular electromechanical systole; RPEP = right pre-ejection period; IVET = right ventricular ejection time; Hit = heart rate. PVR, and MPAP by linear regression analysis. The PADP, PVR, and MPAP also were divided into four groups in order to analyze the statistical likelihood of abnormal pulmonary hemodynamic parameters associated with a specific range of RPEP/RVET (table ). Pulmonary Artery Diastolic Pressure (PADP) When PADP was compared to RPEP/RVET in the 64 patients by linear regression analysis, the r value was 0.2 (fig. 5). The magnitude of peak right ventricular systolic pressure did not influence the RPEP/RVET as was evident in patients with severe 'S 0r 50 ES 0 - ot 0 Q, tq.z 0 lor 0. 0 s0 ).4- ).A-. c )2.2 '" * * * a. *~~~ ** AGE (yeors) 4 Figure 4 The RPEP/RVET, RPEP, and RVET are plotted against age in normal patients. The RPEP and RVET increase with advancing age, while RPEP/RVET is uninfluenced. PS (24, 5, 5) who had marked right ventricular hypertension but low PADP and a low RPEP/RVET. If the RPEP/RVET was below 0., the likelihood PULMONARY ARTERY DIASTOLIC PRESSURE (mmhg) Figure 5 64 Pulmonary artery diastolic pressure (mm Hg) was plotted against RPEP/RVET in 64 patients. The correlation coefficient was r = 0.2.

5 Table 2 Summary of Echographic and Cardiac Catheterization Data in 64 Patients Downloaded from by on October 5, Age RPEP PVR Patient (yrs) HR Diagnosis RPEP RVET RVET PADP (units) MPAP / 2/ 2/ 2/ 2/ / / / 4/ 5/ 5 / 5/ 5/ 5/ 6/ 6/ / 9/ 9/ / 6/ 6/ 6/ / / 2 6/ 2 6/ 2 6/ 4 6/ TGV* 0 VSD 0 PDA TGV* TGV* 0 VSD AVC 0 TGV, VSD* 0 VSD PDA ASD VSD 0 AVC VSD PDA 0 TGV, VSD* 0 VSD 0 AVC 0 VSD VSD 0 VSD VSD AVC PS 0 TGV* 5 TGV* ASD 0 TGV, VSD* AVC 0 ASD 90 VSD VSD, PS 0 PS 0 VSD 0 VSD 0 PS VSD 0 PDA 0 ASD 0 VSD, PS 0 PS 0 ASD 0 TGV* 0 ASD 0 ASD 90 VSD ASD ASD 95 VSD 0 VSD, PS 0 PS ASD AVC ]00 VSD 0 VSD PS 0 ASD 0 VD TGV, VSD* 0 VSD PDA 90 VSD PDA VSD a o o ) *Refers to left ventricular STI. Abbreviations: ASD = atrial septal defect; AVC = atrioventricular canal; PA = pulmonary artery; PDA = patent ductus arteriosus; PS = pulmonic stenosis; PYR = pulmonary vascular resistance; RPRP = right pre-ejection period; RVET = right ventricular ejection time; TGV = transposition of the great arteries; VSD = ventricular septal defect; HR = heart rate.

6 ECHO/PULMONARY VASCULAR BED 64 Table Likelihood of Abnormal Pulmonary Hemodynamic Parameters Associated with RPEP/IRVET PADP (mm Hg) PVR (units) MIPAP (mm Hg) RPEP/RVET N= N-. N=5 N= N= N=2 N= N=29 N-i N=24 N= N=2 N= ) N= N-=2 N= N =5 N=2 N= i N= N=2 N=2 N= N = N= N=2 N= N=4 N = N= N= Abbreviations: MPAP = mean pulmonary artery pressure; N = number of patients in each category; PAI)P = pulmonary artery diastolic pressure; PVR = pulmonary vascular resistance; RPEP = right pre-ejection period; RVET = right ventricular ejection time. Downloaded from by on October 5, of a PADP < mm Hg was 0., and all patients with this low ratio had a PADP of mm Hg or less (table ). If the ratio was 0. or greater, the likelihood was 0.9 that the PADP was greater than mm Hg. Between there was a 0.66 chance of a PADP of greater than mm Hg, but predictability improved to 0.5 if the RPEP/RVET was 0.4 or greater. Pulmonary Vascular Resistance (PVR) The correlation coefficient of RPEP/RVET versus PVR was r= 0.69 in 50 patients (fig. 6). Fourteen patients, nine with TGV, three with large PDAs, and two with severe PS were excluded because of difficulty in calculating pulmonary blood flow and PVR. If the RPEP/RVET was below 0., the likelihood was 0.9 that the PVR was less than units and.0 that the PVR was less than 5 (table ). If the ratio was greater than 0., the PVR was always greater than 5 units. Mean Pulmonary Artery Pressure (MPAP) The correlation coefficient of MPAP versus 0 r= ~~~~ * ~~~~~~~ E~~~ASD o PDA 0O 2_46 _0_5 PS or VSu PS o TGVy I0 PULMONARY VASCULAR RESISTANCE (units) Figure 6 Pulmonary vascular resistance (units) was plotted against RPEP/RVET in 50 patients. The correlation coefficient was r = RPEP/RVET was r = 0.66, but the scatter of data was more pronounced than for PADP or PVR (fig. ). If the RPEP/RVET was below 0., the likelihood of a MPAP below 5 was 0. (table ). If the ratio was greater than 0., the likelihood of a MPAP greater than was 0., and this improved to 0.9 if the ratio was greater than 0.. Sequential Evaluation and Intervention Thirteen of the patients in group 2 were studied serially by echocardiogram and demonstrated marked stability of RPEP/RVET. The RPEP/RVET varied by less than ±0.0 on different echographic examinations. The ratio of the other three patients changed markedly but reflected alterations in the pulmonary vascular bed. The RPEP/RVET of patient 2 was 0.5 (0/0) when the PADP, PVR and MPAP were mm Hg, 5. units, and mm Hg, respectively. Six months later, after what appeared to be a clinically inadequate pulmonary artery banding, the RPEP/RVET was 0 (62/2), and the reduction in ~~~~~~~~~~~~~~~~~~ 0 + D, 02 0~~~~~~~~~~~~~~~~~~~ 0 2 ~~ * "t * o * ASD o + * * o ~~~~~~~~~~~~VSD + *co PDA 0 ++ PS orw. ~~~~~~~~~o TGV n. I MEAN PULMONARY ARTERY PRESSURE mmhg Figure VSD. PS 90 0 Mean pulmonary artery pressure was plotted against RPEP/RVET. The correlation coefficient was r = 0.66.

7 Downloaded from by on October 5, 64 pressures was confirmed at cardiac catheterization. The PADP, PVR, and MPAP were mm Hg, 0.5 units, and 2 mm Hg. A similar situation was experienced with patient 2. Following pulmonary artery banding, the ratio fell from 0. (2/90) to 0 (50/25). The PADP had fallen from mm Hg to mm Hg and the MPAP from mm Hg to 5 mm Hg. When patient 4 was echoed at age 4, the ratio was 0.2, and the PADP and MPAP were mm Hg and 24 mm Hg. One year later the ratio increased to 0.4, and the PADP and MPAP were 5 mm Hg and 50 mm Hg. Six patients (6, 9,,, 55, 6) inhaled oxygen at the time of catheterization to evaluate its effects on the pulmonary vascular bed. Intracardiac pressure recordings, oximetry, and echocardiograms were recorded prior to and after administration of 0% oxygen. The RPEP/RVET of patients 9,, 55, 6 did not change in response to oxygen and reflected the unresponsiveness of the hypertensive pulmonary vascular bed to oxygen. In patient the RPEP/RVET was 0. in room air with a PADP of mm Hg and diminished to 0.29 in oxygen with a PADP of mm Hg. The ratio of patient 6 was 0 when the PADP was mm Hg. RPEP/RVET decreased to 0 in oxygen when the PADP was mm Hg and pulmonary blood flow had doubled. Discussion A systolic time interval is the phase of electromechanical systole the duration of which is governed by four basic factors: ) preload, indicated by diastolic volume or ventricular end-diastolic pressure; 2) afterload, reflected by arterial diastolic pressure; ) contractile state of the myocardium; and 4) the rate and sequence of intraventricular electrical conduction. 4, The pre-ejection period includes all electrical and mechanical events from the onset of ventricular depolarization to opening of the semilunar valves.'5 The phase of isovolumic contraction is a major segment of the pre-ejection period. In acute animal studies of canine left ventricular function, isovolumic contraction lengthened in response to increased afterload.' The ejection phase of systole was shortened by increased afterload, but the change was less significant than that of the pre-ejection period.` Recent work in adults has also indicated that the RPEP lengthened in response to acute and chronic elevation of pulmonary artery pressure.' ` Of primary importance in this study was the effect of long term elevation of afterload in the pulmonic circuit on RVSTI (LVSTI in patients with TGV). The ratio LPEP/LVET has become a useful single expression of left ventricular performance.'5 This ratio has the advantage of accentuating deviations in both HIRSCHFELD ET AL. intervals and thus may reflect abnormality when neither measure is clearly abnormal. In addition, the ratio is uninfluenced by a wide variation in cardiac rate. ` RPEP/RVET was chosen as an expression of altered right ventricular performance resulting from pulmonary hypertension in order to compare RVSTI in patients of different age and heart rate. Although Leighton et al.'0 reported that RVET shortened with increasing heart rate, while RPEP remained unchanged, our data did not support this statement. Leighton primarily studied adult patients whose slow cardiac rates differ greatly from those of a pediatric age group. Eighty percent of our patients were five years of age or less, and only ten patients had a heart rate less than 0. This marked difference in patient selection and range of cardiac rates may have been responsible for the variation between the studies. Because our study indicated that both RPEP and RVET are similarly influenced by age and heart rate, the RPEP/RVET, which was uninfluenced by age and heart rate, permitted serial examination of the same patient and allowed the evaluation of patients of different ages and heart rates. The number of patients is not sufficient to evaluate whether age or heart rate is the dominant factor influencing RVSTI. The influence of age may largely be due to elevated heart rates at younger ages. In our laboratory it has been possible to record the opening and closure of the pulmonary valve for the determination of RVET in 0-% of infants and 50-% of older children. A single complete pulmonic cusp echo was recorded in all patients with pulmonary hypertension and in all children with moderate left-to-right shunts without pulmonary hypertension. Although premature pulmonary valve closure occurred in several patients, this presented no difficulty in measuring RVET because the pulmonary valve reopened in late systole and inscribed a discrete point of valve closure. The correlation of RPEP/RVET with three commonly employed hemodynamic measures of pulmonary hypertension was good. The best correlation was with PADP, which reflects resistance in the pulmonary circuit. Above a PADP of mm Hg there was deviation from a linear correlation, and this may reflect the effects of other hemodynamic variables on RVSTI at this high level of PADP. In our study the correlation of RPET/RVET with PVR was viewed with caution since assumed values for oxygen consumption were employed for the calculation of pulmonary blood flow. The MPAP reflected systolic, as well as diastolic, pulmonary artery pressure and represented the poorest correlation with RPEP/RVET.

8 Downloaded from by on October 5, ECHO/PULMONARY VASCULAR BED The likelihood of a given range of RPEP/RVET predicting abnormal pulmonary hemodynamic parameters was also examined. If the RPEP/RVET was less than 0. the likelihood of a PADP < mm Hg, a PVR < units, and MPAP < 5 mm Hg was about 0.9. When the RPEP/RVET was 0. the likelihood was about 0.9 of a PADP > mm Hg, a PVR > 5 units, and MPAP > mm Hg. Patients with TGV are at particular risk for the early development of pulmonary vascular disease.4 The LVSTI of patients with TGV reflected alterations of pressure and resistance in the pulmonic circuit and were characterized by a low pre-ejection period to ejection time ratio.` The ratio of patients 2, 4, and 59 was markedly increased in response to elevated pulmonary artery pressure and demonstrated that this technique detected elevation of pressure and resistance even when the pulmonary artery originated from an anatomic left ventricle. An important extension of this technique of assessing pulmonary hypertension is the differentiation of fixed, organic changes in the pulmonary vasculature from reversible vasoconstrictive influences. One commonly employed method has been to measure PVR and PADP at cardiac catheterization when the patient is breathing room air and to repeat the measurements in high concentrations of oxygen. There is a predictable decrease of PVR and PADP and increase in pulmonary blood flow if hypoxic vasoconstriction had contributed to the pulmonary hypertension. In our six patients the RPEP/RVET was able to detect the presence or absence of a response to oxygen inhalation. It is also important to evaluate patients following pulmonary artery banding. Our echographic technique predicted the reduction of PADP, PVR and MPAP in two patients with pulmonary artery bands, despite the fact that the bands appeared clinically inadequate. Although elevation of systolic pressure is present proximal to an adequate band, the PADP distal to the band is low which is reflected by a low RPEP/RVET. If complete right bundle branch block (CRBBB) was present, this alone prolonged the RPEP. Postoperative patients with CRBBB could not be evaluated by this technique. In our patients the presence of incomplete right bundle branch block was usually an index of right ventricular hypertrophy and did not influence the length of RPEP. Nanda et al. have recently reported their experience in the recognition of pulmonary hypertension by echocardiographic examination of the pulmonary valve. Adult patients with pulmonary hypertension showed straightening of the diastolic configuration of the pulmonary cusp image, rapid opening movements of the pulmonary valve and absence of the "a" dip or pulmonary cusp motion produced by atrial systole. In our pediatric patients we were unable to record these findings consistently. The present study demonstrated that the measurement of RPEP/RVET by ultrasound provided an accurate assessment of the hemodynamic status of the pulmonary vascular bed. The determination of RPEP/RVET permitted the serial echographic evaluation of pulmonary hemodynamic parameters in selected patients; marked elevation of the ratio indicated the presence of pulmonary hypertension. As further experience with the application of the technique is gained, serial echocardiographic evaluation of patients with increased pulmonary blood flow should reduce the necessity of repeated cardiac catheterization for the assessment of pulmonary artery pressure and resistance. References 649. HEATH D, EDWARDS JE: The pathology of hypertensive pulmonary vascular disease. Circulation : 5, HEATH D, HELMHOLZ HF JR, BURCHELL HB, DUSHANE JW, EDWARDS JE: Graded pulmonary vascular changes and hemodynamic findings in cases of atrial and ventricular septal defect and patent ductus arteriosus. Circulation : 5, 95. HEATH D, HELMHOLZ HF JR, BURCHELL HB, DUSHANE JW, EDWARDS JE: Relation between structural changes in the small pulmonary arteries and the immediate reversibility of pulmonary hypertension following closure of ventricular and atrial septal defects. Circulation :, FERENCZ C: Transposition of the great vessels: Pathophysiologic considerations based upon a study of the lungs. Circulation : 22, CARTNIILL TV, DuSHANE JW, McGOON DC, KIRKLIN JW: Results of repair of ventricular septal defect. J Thorac Cardiovasc Surg 52: 46, HALLIDIE-SMITH KA, HOLLOMAN A, CLELAND WP, BENTALL HH, GOODWINJF: Effects of surgical closure of ventricular septal defects upon pulmonary vascular disease. Br Heart J :: 246, 969. HOFFMANN JIE, RUDOLPH AM: The natural history of isolated ventricular septal defect with special reference to selection of patients for surgery. Adv Pediatr : 5, 90. NANDA NC, GRAMIAK R, ROBINSON TI, SHAH PM: Echocardiographic evaluation of pulmonary hypertension. Circulation 50: 55, HIBS HEELD SS, MEYER RA, KAPLAN S: Measurement of right and left ventricular systolic time intervals by echocardiography. Circulation 5: 4, 95. LEIGHTON RF, WEISSLER AM, WEINSTEIN PB, WOOLEY CF: Right and left ventricular systolic time intervals. Am J Cardiol 2: 66, 9. GRAAMIAK R, NANDA NC, SHAH PM: Echo detection of the pulmonary valve. Radiology 2:, 92. CAYLER CG, RUDOLPH AM, NADAS AS: Systemic blood flow in infants and children with and without heart disease. Pediatrics 2:, 96. LAFARCE CG, MIETTINEN OS: The estimation of oxygen consumption. Cardiovasc Res 4: 2, 90

9 0 4. WAYNE HH: Non-invasive Techniques in Cardiology. Chicago, Year Book Medical Publishers, 9, p 6. WEISSLER AM, LEWIs RP, LEIGHTON RF: The systolic time intervals as a measure of left ventricular performance in man. In Progress in Cardiology, vol, edited by Yu PN, GOODWIN JF. Philadelphia, Lea and Febiger, 92, p 5 HIRSCHFELD ET AL.. WALLACE AG, MITCHELL JH, SKINNER NS, SAR2NOFF SJ: Duration of the phases of left ventricular systole. Circ Res : 6, 96. RUDOLPH AM, YUAN S: Response of pulmonary vasculature to hypoxia and H+ ion concentration. J Clin Invest : 99, 966 Downloaded from by on October 5,