- Email: [email protected]
Effects of mitral stenosis on pulmonary venous flow as measured by Doppler transesophageal echocardiography
Effects of Mitral Sfenosis on Pulmonary Venous Flow as Measured by Doppler Transesophageal Echocardiography Allan L. Klein, MD, Alexander S. Bailey, MS, Gerald I. Cohen, MD, William J. Stewart, MD, Kathleen Husbands, RN, Gregory L. Pearce, MS, and Ernest0 E. Salcedo, MD Pulmonary venous flow as assessed by transophageal echocardiography (7EE) is influenced by changes in left atrial pressure and function. In mitral stenosis (MS), normal left atrial hemodynamia are altered because there is a prolonged decay in diastolic pressure from the left atrium to the left ventricle and atrial function may be altered because of atrial fibrillation. To assess the effect of the prolonged atrial diastolic pressure decay caused by MS on pulmonary venous flow, we studied 27 patients with MS (mitral valve range 0.7 to 2.4 cm*) by pulse&wave Doppler TEE of the left or right upper pulmonary vein, and corn pared results with those of 13 normal subjects. of the 27 subjects with MS, 61% showed a blunted systolic flow pattern and 39% showed a normal flow pattern with greater systolic to diastolic flow ratio. Patients with atrial fibrillation had a predominantly blunted pattern, whereas patients with normal sinus rhythm exhibited both blunted and normal flow patterns. Patients with MS had a lower pulmonary venous peak systolic flow velocity and a longer diastolic pressure halftime than cow trol subjects. Pulmonary venous peak systolic flow velocity was significantly decreased in the presence of atrial fibrillation (p = 0.004). The mitral valve pressure halftime significantly correlated with pulmonary venous diastolic pressure halftime (r = 0.54; p = 0.004) mitral valve area (r = -0.46; p = 0.02). In conclusion, it was found that MS alters pulmonary venous flow patterns, showing a decreased pulmonary venous systolic flow and a prolonged diastolic flow, whi& may be useful in assessing the hemodynamics of MS. (Am J Cardiol1993;72:66-72)
From the Department of Cardiology, The Cleveland Clinic Foundation, Cleveland, Ohio. Manuscript received September 10, 1992; revised manuscript received and accepted January 20, 1993. Address for reprints: Allan L. Klein, MD, Desk F15, The Cleveland Clinic Foundation, Department of Cardiology, 9500 Euclid Avenue, Cleveland, Ohio 44195-5064.
66
THE AMERICANJOURNALOF CARDIOLOGY VOLUME72
P
ulmonary venous flow, as assessedby transesophageal echocardiography (TEE), is intluenced by changesin left atrial pressureand function in various diseases.1,2 It is normally composedof systolic and diastolic forward flow, as well as a small reversed flow during atrial contraction. The systolic wave is related to atrial relaxation after atrial contraction, and the descent of the base of the heart. The diastolic flow is related to the opening of the mitral valve, the rapid inflow of blood into the left ventricle, and a decay in left atrial to left ventricular pressuregradient. The reversedflow is associated with retrograde flow after atrial contraction.1,2In mitral stenosis(MS), there is an alteration of normal left atrial hemodynamicsbecausethe narrowed mitral valve prolongs the decay in diastolic pressure from the left atrium to the left ventricle, and atrial function may be impaired secondaryto atrial fibrillation.3,4 The effect of this prolonged left atrial pressuredecay in diastole and altered atrial function (atrial fibrillation) on pulmonary venous flow has not yet been characterizedby TEE. This study evaluates with TEE the effect of MS on pulmonary venous flow patterns in a large patient population, and determineswhether these patternsmay be useful in assessingthe hemodynamicsof MS. METHODS Patient group: The study was performed between
July 1989 and May 1991 on 27 patients (23 women and 4 men, mean age 54 + 14 years [range 30 to 771) with MS, who were referred to the cardiac function laboratory for TEE evaluation. Because of the known effect of mitral regurgitation on pulmonary venous -Bow,~,~patients with significant mitral regurgitation (>2+), as determined by Doppler color flow mapping, were excluded from the study. Normal control wbjec& Thirteen subjects(8 women and 5 men, mean age 42 k 13 years) without organic diseaseas evidenced by TEE with adequatepulmonary venous flow recordings served as a normal control group for comparison. Wiogr@ii examination: All patientsunderwent a complete outpatient TEE examination according to standard methods7with a 5 MHz monoplane transducer connected to a commercially available Doppler echocardiograph (Acuson model 128, Mt. View, California, or Hewlett-Packard model 7202OA, Andover, Maryland). All patients also underwent a complete transthoracic study using a 2.5 MHz transducer before the transesophagealstudy.
JULY 1,1993
Pulmonary venous flow measurements: Pulsedwave Doppler TEE of the left or right upper pulmonary vein was performed from the transverseshort-axis view of the pulmonary veins and left atrium.2,5With use of the orange color flow of the pulmonary veins emptying into the left atrium, the sample volume was placed 1 to 2 cm distal to the opening of the pulmonary vein in the left atrium.2,5The left upper pulmonary vein tracing was used in 23 patients (SS%), and the right upper pulmonary vein was used in 4 patients (15%). Pulmonary venous flow was recorded on a hard copy at a speedof 50 or 100 mm/s. Echocardiographic measurements: M-mode measurementsfrom the parastemalshort axis of the left ventricle were used to derive the left ventricular end-diastolic and end-systolic dimensions, fractional shortening and left atrial size. The ejection fraction was computed by the method of Quinones et al8 Continuous-wave Doppler transthoracic echocardiography acrossthe mitral valve was performed to measure the mean pressure gradient and pressure halftime. The mitral valve area was calculated as 22O/pressure halftime by the method of Hatle et al? Pulmonary venous Row measurements All measurements were calculated with an off-line digitizing computer (Dextra Medical, Irvine, California). Mean values were obtained by averaging 3 to 6 consecutive systolic, diastolic and atria1 reversal peak flow velocities (Figure 1). Peak velocities were measured from their baselines to their maximal velocities. Pulmonary venous systolic flow was describedas normal or blunted according to acceptedmethods.6A normal pattern was defined as a systolic to diastolic flow ratio of 21. A blunted pattern was defined as a systolic to diastolic flow ratio between 0 and 1. The duration and time velocity integrals were obtained by digitizing the darkestportion of the pulmonary venous flow recordings. The pressure halftime of the pulmonary venous diastolic wave was calculated by multiplying the deceleration time of the diastolic wave from the peak to the baseline by the constant,0.29, similar to the method of Teague et al.‘O Cardiac catheterization: Eighteen of the 27 patients (67%) had a cardiac catheterizationat a mean time of 19 days before the TEE study.The mean pressuregradient and pulmonary capillary wedge pressurewere recorded, and the mitral valve area was calculated according to the Gorlin formula.‘l Statistical analysis: Comparisons between the experimental and control groups were performed with linear contrasts. Relationships between pulmonary vein, mitral valve and hemodynamic variables were examined with Pearson correlation coefficients. Differences between atria1 fibrillation and normal sinus rhythm subgroups and the control group were examined with analysis of variance techniques using Bonferroni multiple comparison adjustment methods. Nonparametric methods (Kruskal-Wallis and Wilcoxon rank-sum tests) were used where distribution assumptionswere not met. A p value co.05 was considered statistically signilicant. Data are presentedas mean If: SD.
RESULTS Clinical, B and Doppler miid find ings: Patients with MS were predominantly women and
were older than the control group (Table I). All twentyseven patients had native MS with a mean mitral valve area of 1.2 cm2. Eighteen patients (67%) were in normal sinus rhythm, whereas 9 (33%) had atrial fibrillation. Patientswith atrial fibrillation had a faster heart rate
FWJRE 1. A t rancesophageal echocardiographii recardi~ of ri&t pulmonary (PULM) venous flow and mitral inflow in a normal subject showing a greater systolic (S) to diastolic (D) flow. The pulmonary venous diastolic pure halftime (DPM) was calculated as the deceleration time (DT) (time from peak diastolic flow to its baseline), multiplied by 0.29.10 Mitral inflow shows greater early (E) than late (A) filliN velocity. The pressure halftime (PRT) of mitral inRow was calculated as the deceleration time x0.29. DUR = dura tion.
PULMONARYVEIN BY TEE IN MITRALSTENOSIS 67
and larger left atria than the group with normal sinus rhytbrn and the control group (p ~0.05). Pulmonary venous flow: Of the 27 subjects with MS, 17 (63%) had a blunted systolic flow pattern and 10 (37%) had a normal flow pattern with a greater sys-
tolic to diastolic flow. Patients with atrial fibrillation had predominantly a blunted pattern (8 of 9), whereas patients with normal sinus rhythm exhibited both blunted (9 of 18) and normal (9 of 18) flow patterns. Pulmonary venous flow velocities are listed in Table
I\BLE I Clinical, Two-Dimensional and Doppler Echocardiographic atients with Mitral Stenosis and of 13 Control Subjects
Characteristics of 27
Experimental Group (n = 27)
Men/women Heart rate (beats/min) New York Heart Association (III/IV) Age (years) Left ventricular diastolic cavity size (mm) Left ventricular systolic cavity size (mm) Left atrium (mm) Mitral pressure halftime (ms) Mean mitral gradient
Total
Sinus Rhythm (n = 18)
Atrial Fibrillation (n = 9)
Control Subjects (n = 13)
4123 82 r 34 18*
3115 82 f 14$ 135
l/8 100 + 36 5t
518 90 ir 18 0
54 f 13* 49 -t 6
47 k ll$ 48 f 6
67 t 8t 50 r 6
42 r 13 47 2 6
31 -+6
30 f 4
33 f 7
30 f 6
53 f 10* 185 f 54
50 f 6§$ 180 ? 52
60 f 127 197 f 61
35 f 60 -
12 + 4
12 -+ 5
1.2 f 0.4 15 f 3
1.2 + 0.3 14 f 2
1.2 f 0.5 15 f 6
1.2 +- 0.6
1.1 f 0.5
1.5kO.9
24 k 8
25 f 8
22 f 7
11 rt:3
(mm Hg) (Doppler) Doppler valve area (cm21 Mean mitral gradient (mm Hg) (cath) Valve area (cm*) (cath) Pulmonary capillary wedge pressure (mm Hg) *p < 0.05 between tp < 0.05 b&an $p -C0.05 between §p < 0.05 between Data are expressed
the experimental group and controls. the atrial fibrillation subgroup and controls. the normal sinus rhythm and the atrial fibrillation the normal sinus rhythm subgroup and controls. as mean k SD.
-
subgroups.
I
I
TABLE II Transesophageal Echocardiographic Measurements of Pulmonary Venous Flow Variables in 27 Patients with Mitral Stenosis and in 13 Control Subjects Experimental Group (n = 27) NSR (n = 18)
Total
AFIB (n = 9)
Control Group (n = 13)
Peak Flow Velocities (cm/s) Systolic Diastolic Systolic/diastolic Atrial reversal
ratio
38 43 1.0 24
T + f f
19* 21 0.6* 9*
47 44 1.3 24
+ 2 f f
16$ 23 0.7$ 95
23 f 15t 40 r?: 19 0.6 z!c0.3t
55 39 1.5 17
f f + 2
17 11 0.4 5
5 f 4t 10 k 8 0.6 f 0.4t
14 f 5 7k2 2.2 f 0.7 2&l
-1
Flow Velocity Integrals (cm) Systolic Diastolic Systolic/diastolic Atrial reversal
ratio
7 -c 4* 10 + 6* 0.9 2 0.5* 321
9 f 4§$ 10 * 5 1.1 + 0.551: 3tl
Time Integrals (ms)
tp ~0.05 tp ~0.05 §p ~0.05
68
between the atrial fibrillation subgroup and controls. between the normal sinus rhythm and the atrial fibrillation between the normal sinus rhythm subgroup and controls.
THE AMERICAN JOURNAL OF CARDIOLOGY
VOLUME 72
subgroups.
JULY 1,1993
II. The overall peak systolic velocity and systolic to diastolic flow ratio were lower in the MS than in the control group; however, peak diastolic flow velocities did not differ between the groups. In contrast, the peak atrial reversed flow velocity was higher in the MS patients in normal sinus rhythm than in the control subjects. The diastolic flow velocity integral, duration of diastolic flow, and the pressurehalftime of the diastolic flow were greater in the study group than in the control subjects (Figures 2 to 5).
systolic flow and nonsimultaneous pulmonary capillary wedge pressure.
DISCUSSION Pulmonary venous flow measuredby Doppler TEE is markedly altered in the presenceof MS and the associated prolonged left atrial diastolic pressuredecay.12In contrast to the normal systolic and diastolic flows and the small reversed flow with atrial contraction, patients with MS (mean valve area 1.2 cm*) had a decreased Effect of atrial fibrillation on pulmonary venous peak systolic flow velocity, a prolonged duration and flow: Patients with atrial fibrillation had a markedly de- pressurehalftime of diastolic flow, and an increasedrecreasedpeak systolic velocity and systolic to diastolic versed flow during atrial contraction when compared flow ratio, when comparedwith control subjects,where- with normal control subjects. as there was no difference between them and patients Mechanism of pulmonary venous flow: The syswith normal sinus rhythm. Atrial fibrillation had no sig- tolic, diastolic and atrial reversalphasesof pulmonary venificant effect on peak diastolic flow. nous flow are intluenced by both atria1 and ventricular Univariate correlates of pulmonary venous flow: factors.*J3The atrial factors include left atrial contractilAmong patients with MS, the diastolic pressurehalftime ity and relaxation, left atrial pressureand compliance,decorrelated well with the mitral valve pressure halftime cay of left atria1pressure,and rhythm abnormalities.The (r = 0.54, p = 0.004), Doppler valve area (r =-0.46, ventricular factors include mitral annular displacement, p = 0.02), catheterization-derived valve area (r = -0.62, left ventricular compliance, and relaxation. During the p = 0.005) and mitral gradient (r = -0.47, p = 0.05) (Figures 6, top and bottom). Diastolic pressure halftime in patients with normal sinus rhythm correlated better with mitral pressurehalftime (r = 0.61; p = 0.007) than it did in patients with atrial fibrillation (r = 0.43; p = 0.25). There was no correlation betweenthe pulmonary venous
RGURE 2. A transesophageal recording of left pulmonary venous flow and transthoracic recording of mitral inflow in a 69year4d woman with mild mitral stenosis (MS). 7he sy+ tolic (S) to diastolic (D) flow ratio is more than normal, and atrial reversal (AR) is prominent. Note the normal diastolic pressure halftime (DPHT) of 70 ms and the small mean gr+ dient (MG) of 6 mm across the mitral valve. These recordings are not simultaneous.
FIGURE 3. A transesophageal recording of right pulmonary venous flow and transthoracic recording of mitral inflow in a 79year-old patient with severe mitral stenosis (MS). Note the markedly prolonged pulmonary venous diastolic (D) pressure halftime (DPHC) of 146 ms, as well as the prxb longed mean pressure gradient (MO) of 9 mm Rg. In addition, the pressure halftime (PHT) of the mitral inflow was markedly prolonged at 266 ms. Dther abbreviations as in Figures 1 and 2.
PULMONARYVEIN BY TEE IN MITRALSTENOSIS 69
systolic phase,the pulmonary veins are filled from right ventricular contraction with subsequentfilling of the left atriurn2J3 The diastolic phaseis affectedby the gradient from the pulmonary artery to the left atrium.2J3The atrial reversalflow results from atrial contraction causingretrograde flow in the pulmonary veins2 MS alters physiology and tiuences pulmonary venous flow in several ways. The first is that in patients with normal sinus rhythm, resistanceis increasedat the mitral valve, which may cause elevated left atria1 diastolic pressure,accounting for a decreaseddiastolic gradient into the left atrium from the pulmonary veins and thus, a normal flow pattern with greater systolic to diastolic flow. In addition, there may be increasedatrial afterload (at the mitral valve level) during atrial contraction, which may account for the increasedretrograde flow in the pulmonary veins.4The secondis that the mitral valve is thickened and may interfere with normal mitral annular displacementduring systole, resulting in increasedleft atrial pressure.This would account for the decreasedpulmonary venous systolic flow (blunted sys-
FlGURE 4. A transesophageal recording of left pulmonary ve nous flow and transthoracic recoftling of mitral inflow in a SO-yeadd patient with moderate mitral stenosis (MS) and atrial fibrillation (A FlB). The systolic (S) to diastolic (D) flow is decreased and the diastolic pressure hatftime (DPRT) is mildly prolonged at 85 ms in tha pulmonary vein retarding Note the mildly increased mean gradient (MG) of 10 mm Hg and a pressure halftime (PRT) of 143 ms in the mitral iaflow recordin&
70
THE AMERICANJOURNALOF CARDIOLOGY VOLUME72
tolic flow) observed in our study.14The third is the effect of the prolonged decay of left atria1to left ventricular diastolic pressure,which affectspulmonary venous flow diastolic filling, thus prolonging the duration and pressurehalftime of diastolic flow into the left atrium.4J2 The fourth reason is the effect of MS on atrial function.15 Patients with MS may have altered atrial myocardium, which may create the milieu for atrial fibrillation.16In patients with atrial fibrillation, timing of atrial contraction and relaxation is lost, resulting in absent early systolic filling; thus, the pulmonary venous systolic flow dependsentirely on descentof the baseof the heart during systole, which may already be affected by the anatomic derangement of MSi4 Without atrial relaxation, left atrial pressuremay be higher in systole, and thus the pulmonary veins may have a decreasedsystolic gradient into the left atrium, resulting in decreasedsystolic flow~>4’7>‘8 Univariate cotrelates of pulp venous flow in mitral stenosis: We found a good correlationbetweenin-
creasing diastolic pressure halftime (i.e., the pressure halftime of the diastolic flow into the left atrium) and increasing mitral valve pressurehalftime and valve area. This correlation reflects the prolonged left atrial pressure decay in increasing severity of MS, which prolongs the
FIGURE 5. A transesophageal recording of left pulmonary venous aad transthoacic recording of mitral inflow in an 80 year-old patient with severe mitral stenosis (MS) and atrial fibrillation (A FIB). lhe systolic (S) to diastolic (D) flow is markedly decreased and the diastolic pressure halftime (DPHT) is prolonged at 131 ms. Mitral inflow shows a markedly increased pressure gradient (MG) of 13 mm Rg and a prolonged pressure halftime (P)(T) of 314 ms.
JULY 1,1993
FlGURE 6. Scattelplots showing the relatiorr ship between pulmonary venous diastolic pressure halftime, mitral valve pressure halftime (top) and Doppler valve area (hottom) in 31 patients with atrial fibrillation and normal sinus tiythm with mitral stem sls.
0
50
100
150
200
250
300
350
400
Mitral Valve Pressure Half Time
zoo.
*.
*. -*.. -.
-.
‘.
.g 150l‘1= I” 2! SlW-------____
-em_
tz I? ,o 0 9” .n *-
*.
0 -? 0
50
100
150
.mJ
250
330
*.
*. 350
I 4w
Mitral Valve Area
pressure halftime across the mitral valve as well as the Wing of the pulmonary veins during diastole.2,4J2 Previous studies: In a transthoracic Doppler echocardiography study of the right upper pulmonary veins in 15 patients with MS, Keren et al4 described the effects of severe MS, as well as the effect of normal sinus rhythm and at&l fibrillation, on pulmonary venous flow. In patients with severe MS and normal sinus rhythm, they found the pulmonary capillary wedge pressure to RGURE 7. A pulse&wave Doppler tansesophageal recordia of the left superior pulmonary vein and a simultaneous left atrial (LA) pressure recording in a 37-yealcdd woman with severe mitral stenosis (MS) and atrial fibrillation (A FlB). Note ths prolonged diastolic pressure halfttme (DPRT) of 130 ms (arrow), coincident with the left atrial pressure decay, from the v wave to the troup of the y descent (arrow). D q diastolic; S q systolic. PULMONARY VEIN BY TEE IN MITRAL STENOSIS
71
be lower during systole than during diastole, accounting for increasedpulmonary venous systolic and decreased diastolic flow. Our study contirms the results of Keren et al in a larger patient population using the more acceptable and reliable transesophagealmeasurementsof both left and right upper pulmonary venous flo~.~ A recent investigation during mitral balloon valvuloplasty in MS showed that the blunted systolic flow pattern in the pulmonary veins before valvuloplasty immediately ‘ ‘normalizes” after the procedure, which increasesthe mitral valve area. Study limitations: The major limitation of this study is that we did not have simultaneous measurements of left atrial pressureand pulmonary venous flow and mitral valve flows. The correlation of pulmonary venous diastolic pressure halftime and mitral v@e pressure halftime would have undoubtedly been higher with simultaneous recordings. Recently, we simultaneously measuredleft atrial pressureand pulmonary venous flow in a patient with severe MS and found that prolonged left atrial pressuredecay coincided with prolonged pressure halftime and duration of diastolic filling (Figure 7). Pulmonary venous flow may be intluenced by many other confounding factors, including left atrial volume and atrial ejection fraction, left atrial compliance loading conditions, cardiac output, and left ventricular compfimce.Zl920
1. Keren G, Sherez J, Megidish R, Levitt B, Laniado S. Pahnonary venous flow pattern - its relationship to cardiac dynamics. A pulsed Doppler echocardiographic study. CircuUofi 1985;71:1105-1112. 2. Klein AL, Tajik AJ. Doppler assessment of pulmonary venous flow in healthy subjects and in patients with heart disease. J Am Sot Echo 1991;4:37%392. a. Bmmwald E. Valvular heart disease. In: Braunwald E, ed. A Textbook of Cardiovascular Medicine. Philadelphia: WB Saunders, 1992: 1007-1011. 4. Keren G, Pa&s A, Miller HI, Scherez J, La&lo S. Puhnomuy venous flow determined by Doppler echocardiography in mitral stenosis. Am J Cardiol 1990.65:
72
THE AMERICANJOURNALOF CARDIOLOGY VOLUME72
246-249. 5. Caste110R, Pearson AC, Lenzen P, Labovitz AJ. Effect of mitral regurgitation on pulmonary venous velocities derived from tramesophageal echocardiography color-guided pulsed-Doppler imaging. J Am CON Cardiol 1991;17:149%1506. 9. Klein AL, Obarski TP, Stewart WJ, &sale PN, Pearce GL, Husbands K, Cosgrove DM, Sakedo EE. Tramesophageal Doppler echocardiography of pulmonary venous flow: a new marker of mitral regurgitation severity. J Am Co11 Cardiol 1991;18:2518-526. 7. Seward JB, Khandheria BK, Oh JK, Abel MD, Hughes RW, Edwards WD, Nichols BA, Freeman WK, Tajii AJ. TrTsesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 1988;63:649-680, 8. Quinones MA, Waggoner AD, R&to LA, Nelson JG, Young JB, Winters WL, Rib&o LG, Miller RR. A new, simplified and accurate method for determining ejection fraction with hwdimensional echocardiography. Circulation 1981;64:
744753. 9. Hatle LK, Brubakk A, Tromsdal A, Angelsen B. Noninvasive assessment of pressure drop in mitral stenosis by Doppler ultrasound. Br Heart J 1978;40: 13 l-140.
10. Teague SM, Heinsimer JA, Anderson JL, Sublett I$ Olson EG, Voyles WF, Thadani U. Quantification of aortic regurgitation utilizing continuous wave Doppler nltrasonnd J Am Co11 Cardiol 1986;8:592-599. li. Cohen MV, Gorlin R. Modified orifice equation for the calculation of mitral valve area. Am Heart J 1972;84:839-840. 12. Klein AL, Cohen GI, Davison MB, Stewart WJ, Husbands 5, Pearce GL, Salcede EE. Effect of mitral stenosis on pulmonary venous flow by hansesophageal echocardiography (abstr). J Am Sot Echo 1991;4:297. 13. Yellm EL, Meisner JS, Nikolic SD, Karen G. ‘Ihe scientic basis for the relations between pulsed-Doppler trammitral velocity patterns and left heart chamber properties. Echocardiography 1992;9:313-338. 14. Natarajan D, Sharma VP, Chandra S, Dhar SK, Gaga M, Caroli B. Effects of percutanmus balloon mitral valvotomy on puhnonary venous flow in severe mitral stenosis. Am J Cardiol 1992;69:81&812. 15. Keren G, Et&n T, Sherez J, Z.&x AA, Megidish R, Miller HI, La&do S. Atrial fibrilltion and atrial enlargement in patients with mitral stenosis. Am Heart J 1987;114:1146-1155. 19. Thiedemann K-V, Ferrans VJ. Left atrial ultrastmchue in m&al valvular disease. Am J Path01 1977:89:575-604. 17. Keren G, Sonnenblick EH, LeJemtel TH. Mitral atmlus motion. Relation to pulmomuy venous and transmitral flows in normal subjects and in patients with dilated cardiomyopathy. Circtdation 1988;78:621-629. 19. Keren G, Bier A, Sherez J, Miura D, Keefe D, LeJemtel T. Atrial contraction is an important determinant of pulmonary venous flow. J Am Coil Cardiol 19867:
693-695. 19. Basnight MA, Gonzalez MS, Kershenovich SC, Appleton CP. Pulmonary venous flow velocity: relation to hemodynamics, mitral flow velocity and left atrial volume, and ejection fraction. J Am Sot Echo 1991;4:547-558. 20. Bartzckis T, Lee R, Yeoh TK, Grosin H, Schnittger I. Tramesophageal echoDoppler assessment of pulmonary venous flow patterns. J Am Sot Echo 1991;4:
457-464.
JULY 1,1993
From the Department of Cardiology, The Cleveland Clinic Foundation, Cleveland, Ohio. Manuscript received September 10, 1992; revised manuscript received and accepted January 20, 1993. Address for reprints: Allan L. Klein, MD, Desk F15, The Cleveland Clinic Foundation, Department of Cardiology, 9500 Euclid Avenue, Cleveland, Ohio 44195-5064.
66
THE AMERICANJOURNALOF CARDIOLOGY VOLUME72
P
ulmonary venous flow, as assessedby transesophageal echocardiography (TEE), is intluenced by changesin left atrial pressureand function in various diseases.1,2 It is normally composedof systolic and diastolic forward flow, as well as a small reversed flow during atrial contraction. The systolic wave is related to atrial relaxation after atrial contraction, and the descent of the base of the heart. The diastolic flow is related to the opening of the mitral valve, the rapid inflow of blood into the left ventricle, and a decay in left atrial to left ventricular pressuregradient. The reversedflow is associated with retrograde flow after atrial contraction.1,2In mitral stenosis(MS), there is an alteration of normal left atrial hemodynamicsbecausethe narrowed mitral valve prolongs the decay in diastolic pressure from the left atrium to the left ventricle, and atrial function may be impaired secondaryto atrial fibrillation.3,4 The effect of this prolonged left atrial pressuredecay in diastole and altered atrial function (atrial fibrillation) on pulmonary venous flow has not yet been characterizedby TEE. This study evaluates with TEE the effect of MS on pulmonary venous flow patterns in a large patient population, and determineswhether these patternsmay be useful in assessingthe hemodynamicsof MS. METHODS Patient group: The study was performed between
July 1989 and May 1991 on 27 patients (23 women and 4 men, mean age 54 + 14 years [range 30 to 771) with MS, who were referred to the cardiac function laboratory for TEE evaluation. Because of the known effect of mitral regurgitation on pulmonary venous -Bow,~,~patients with significant mitral regurgitation (>2+), as determined by Doppler color flow mapping, were excluded from the study. Normal control wbjec& Thirteen subjects(8 women and 5 men, mean age 42 k 13 years) without organic diseaseas evidenced by TEE with adequatepulmonary venous flow recordings served as a normal control group for comparison. Wiogr@ii examination: All patientsunderwent a complete outpatient TEE examination according to standard methods7with a 5 MHz monoplane transducer connected to a commercially available Doppler echocardiograph (Acuson model 128, Mt. View, California, or Hewlett-Packard model 7202OA, Andover, Maryland). All patients also underwent a complete transthoracic study using a 2.5 MHz transducer before the transesophagealstudy.
JULY 1,1993
Pulmonary venous flow measurements: Pulsedwave Doppler TEE of the left or right upper pulmonary vein was performed from the transverseshort-axis view of the pulmonary veins and left atrium.2,5With use of the orange color flow of the pulmonary veins emptying into the left atrium, the sample volume was placed 1 to 2 cm distal to the opening of the pulmonary vein in the left atrium.2,5The left upper pulmonary vein tracing was used in 23 patients (SS%), and the right upper pulmonary vein was used in 4 patients (15%). Pulmonary venous flow was recorded on a hard copy at a speedof 50 or 100 mm/s. Echocardiographic measurements: M-mode measurementsfrom the parastemalshort axis of the left ventricle were used to derive the left ventricular end-diastolic and end-systolic dimensions, fractional shortening and left atrial size. The ejection fraction was computed by the method of Quinones et al8 Continuous-wave Doppler transthoracic echocardiography acrossthe mitral valve was performed to measure the mean pressure gradient and pressure halftime. The mitral valve area was calculated as 22O/pressure halftime by the method of Hatle et al? Pulmonary venous Row measurements All measurements were calculated with an off-line digitizing computer (Dextra Medical, Irvine, California). Mean values were obtained by averaging 3 to 6 consecutive systolic, diastolic and atria1 reversal peak flow velocities (Figure 1). Peak velocities were measured from their baselines to their maximal velocities. Pulmonary venous systolic flow was describedas normal or blunted according to acceptedmethods.6A normal pattern was defined as a systolic to diastolic flow ratio of 21. A blunted pattern was defined as a systolic to diastolic flow ratio between 0 and 1. The duration and time velocity integrals were obtained by digitizing the darkestportion of the pulmonary venous flow recordings. The pressure halftime of the pulmonary venous diastolic wave was calculated by multiplying the deceleration time of the diastolic wave from the peak to the baseline by the constant,0.29, similar to the method of Teague et al.‘O Cardiac catheterization: Eighteen of the 27 patients (67%) had a cardiac catheterizationat a mean time of 19 days before the TEE study.The mean pressuregradient and pulmonary capillary wedge pressurewere recorded, and the mitral valve area was calculated according to the Gorlin formula.‘l Statistical analysis: Comparisons between the experimental and control groups were performed with linear contrasts. Relationships between pulmonary vein, mitral valve and hemodynamic variables were examined with Pearson correlation coefficients. Differences between atria1 fibrillation and normal sinus rhythm subgroups and the control group were examined with analysis of variance techniques using Bonferroni multiple comparison adjustment methods. Nonparametric methods (Kruskal-Wallis and Wilcoxon rank-sum tests) were used where distribution assumptionswere not met. A p value co.05 was considered statistically signilicant. Data are presentedas mean If: SD.
RESULTS Clinical, B and Doppler miid find ings: Patients with MS were predominantly women and
were older than the control group (Table I). All twentyseven patients had native MS with a mean mitral valve area of 1.2 cm2. Eighteen patients (67%) were in normal sinus rhythm, whereas 9 (33%) had atrial fibrillation. Patientswith atrial fibrillation had a faster heart rate
FWJRE 1. A t rancesophageal echocardiographii recardi~ of ri&t pulmonary (PULM) venous flow and mitral inflow in a normal subject showing a greater systolic (S) to diastolic (D) flow. The pulmonary venous diastolic pure halftime (DPM) was calculated as the deceleration time (DT) (time from peak diastolic flow to its baseline), multiplied by 0.29.10 Mitral inflow shows greater early (E) than late (A) filliN velocity. The pressure halftime (PRT) of mitral inRow was calculated as the deceleration time x0.29. DUR = dura tion.
PULMONARYVEIN BY TEE IN MITRALSTENOSIS 67
and larger left atria than the group with normal sinus rhytbrn and the control group (p ~0.05). Pulmonary venous flow: Of the 27 subjects with MS, 17 (63%) had a blunted systolic flow pattern and 10 (37%) had a normal flow pattern with a greater sys-
tolic to diastolic flow. Patients with atrial fibrillation had predominantly a blunted pattern (8 of 9), whereas patients with normal sinus rhythm exhibited both blunted (9 of 18) and normal (9 of 18) flow patterns. Pulmonary venous flow velocities are listed in Table
I\BLE I Clinical, Two-Dimensional and Doppler Echocardiographic atients with Mitral Stenosis and of 13 Control Subjects
Characteristics of 27
Experimental Group (n = 27)
Men/women Heart rate (beats/min) New York Heart Association (III/IV) Age (years) Left ventricular diastolic cavity size (mm) Left ventricular systolic cavity size (mm) Left atrium (mm) Mitral pressure halftime (ms) Mean mitral gradient
Total
Sinus Rhythm (n = 18)
Atrial Fibrillation (n = 9)
Control Subjects (n = 13)
4123 82 r 34 18*
3115 82 f 14$ 135
l/8 100 + 36 5t
518 90 ir 18 0
54 f 13* 49 -t 6
47 k ll$ 48 f 6
67 t 8t 50 r 6
42 r 13 47 2 6
31 -+6
30 f 4
33 f 7
30 f 6
53 f 10* 185 f 54
50 f 6§$ 180 ? 52
60 f 127 197 f 61
35 f 60 -
12 + 4
12 -+ 5
1.2 f 0.4 15 f 3
1.2 + 0.3 14 f 2
1.2 f 0.5 15 f 6
1.2 +- 0.6
1.1 f 0.5
1.5kO.9
24 k 8
25 f 8
22 f 7
11 rt:3
(mm Hg) (Doppler) Doppler valve area (cm21 Mean mitral gradient (mm Hg) (cath) Valve area (cm*) (cath) Pulmonary capillary wedge pressure (mm Hg) *p < 0.05 between tp < 0.05 b&an $p -C0.05 between §p < 0.05 between Data are expressed
the experimental group and controls. the atrial fibrillation subgroup and controls. the normal sinus rhythm and the atrial fibrillation the normal sinus rhythm subgroup and controls. as mean k SD.
-
subgroups.
I
I
TABLE II Transesophageal Echocardiographic Measurements of Pulmonary Venous Flow Variables in 27 Patients with Mitral Stenosis and in 13 Control Subjects Experimental Group (n = 27) NSR (n = 18)
Total
AFIB (n = 9)
Control Group (n = 13)
Peak Flow Velocities (cm/s) Systolic Diastolic Systolic/diastolic Atrial reversal
ratio
38 43 1.0 24
T + f f
19* 21 0.6* 9*
47 44 1.3 24
+ 2 f f
16$ 23 0.7$ 95
23 f 15t 40 r?: 19 0.6 z!c0.3t
55 39 1.5 17
f f + 2
17 11 0.4 5
5 f 4t 10 k 8 0.6 f 0.4t
14 f 5 7k2 2.2 f 0.7 2&l
-1
Flow Velocity Integrals (cm) Systolic Diastolic Systolic/diastolic Atrial reversal
ratio
7 -c 4* 10 + 6* 0.9 2 0.5* 321
9 f 4§$ 10 * 5 1.1 + 0.551: 3tl
Time Integrals (ms)
tp ~0.05 tp ~0.05 §p ~0.05
68
between the atrial fibrillation subgroup and controls. between the normal sinus rhythm and the atrial fibrillation between the normal sinus rhythm subgroup and controls.
THE AMERICAN JOURNAL OF CARDIOLOGY
VOLUME 72
subgroups.
JULY 1,1993
II. The overall peak systolic velocity and systolic to diastolic flow ratio were lower in the MS than in the control group; however, peak diastolic flow velocities did not differ between the groups. In contrast, the peak atrial reversed flow velocity was higher in the MS patients in normal sinus rhythm than in the control subjects. The diastolic flow velocity integral, duration of diastolic flow, and the pressurehalftime of the diastolic flow were greater in the study group than in the control subjects (Figures 2 to 5).
systolic flow and nonsimultaneous pulmonary capillary wedge pressure.
DISCUSSION Pulmonary venous flow measuredby Doppler TEE is markedly altered in the presenceof MS and the associated prolonged left atrial diastolic pressuredecay.12In contrast to the normal systolic and diastolic flows and the small reversed flow with atrial contraction, patients with MS (mean valve area 1.2 cm*) had a decreased Effect of atrial fibrillation on pulmonary venous peak systolic flow velocity, a prolonged duration and flow: Patients with atrial fibrillation had a markedly de- pressurehalftime of diastolic flow, and an increasedrecreasedpeak systolic velocity and systolic to diastolic versed flow during atrial contraction when compared flow ratio, when comparedwith control subjects,where- with normal control subjects. as there was no difference between them and patients Mechanism of pulmonary venous flow: The syswith normal sinus rhythm. Atrial fibrillation had no sig- tolic, diastolic and atrial reversalphasesof pulmonary venificant effect on peak diastolic flow. nous flow are intluenced by both atria1 and ventricular Univariate correlates of pulmonary venous flow: factors.*J3The atrial factors include left atrial contractilAmong patients with MS, the diastolic pressurehalftime ity and relaxation, left atrial pressureand compliance,decorrelated well with the mitral valve pressure halftime cay of left atria1pressure,and rhythm abnormalities.The (r = 0.54, p = 0.004), Doppler valve area (r =-0.46, ventricular factors include mitral annular displacement, p = 0.02), catheterization-derived valve area (r = -0.62, left ventricular compliance, and relaxation. During the p = 0.005) and mitral gradient (r = -0.47, p = 0.05) (Figures 6, top and bottom). Diastolic pressure halftime in patients with normal sinus rhythm correlated better with mitral pressurehalftime (r = 0.61; p = 0.007) than it did in patients with atrial fibrillation (r = 0.43; p = 0.25). There was no correlation betweenthe pulmonary venous
RGURE 2. A transesophageal recording of left pulmonary venous flow and transthoracic recording of mitral inflow in a 69year4d woman with mild mitral stenosis (MS). 7he sy+ tolic (S) to diastolic (D) flow ratio is more than normal, and atrial reversal (AR) is prominent. Note the normal diastolic pressure halftime (DPHT) of 70 ms and the small mean gr+ dient (MG) of 6 mm across the mitral valve. These recordings are not simultaneous.
FIGURE 3. A transesophageal recording of right pulmonary venous flow and transthoracic recording of mitral inflow in a 79year-old patient with severe mitral stenosis (MS). Note the markedly prolonged pulmonary venous diastolic (D) pressure halftime (DPHC) of 146 ms, as well as the prxb longed mean pressure gradient (MO) of 9 mm Rg. In addition, the pressure halftime (PHT) of the mitral inflow was markedly prolonged at 266 ms. Dther abbreviations as in Figures 1 and 2.
PULMONARYVEIN BY TEE IN MITRALSTENOSIS 69
systolic phase,the pulmonary veins are filled from right ventricular contraction with subsequentfilling of the left atriurn2J3 The diastolic phaseis affectedby the gradient from the pulmonary artery to the left atrium.2J3The atrial reversalflow results from atrial contraction causingretrograde flow in the pulmonary veins2 MS alters physiology and tiuences pulmonary venous flow in several ways. The first is that in patients with normal sinus rhythm, resistanceis increasedat the mitral valve, which may cause elevated left atria1 diastolic pressure,accounting for a decreaseddiastolic gradient into the left atrium from the pulmonary veins and thus, a normal flow pattern with greater systolic to diastolic flow. In addition, there may be increasedatrial afterload (at the mitral valve level) during atrial contraction, which may account for the increasedretrograde flow in the pulmonary veins.4The secondis that the mitral valve is thickened and may interfere with normal mitral annular displacementduring systole, resulting in increasedleft atrial pressure.This would account for the decreasedpulmonary venous systolic flow (blunted sys-
FlGURE 4. A transesophageal recording of left pulmonary ve nous flow and transthoracic recoftling of mitral inflow in a SO-yeadd patient with moderate mitral stenosis (MS) and atrial fibrillation (A FlB). The systolic (S) to diastolic (D) flow is decreased and the diastolic pressure hatftime (DPRT) is mildly prolonged at 85 ms in tha pulmonary vein retarding Note the mildly increased mean gradient (MG) of 10 mm Hg and a pressure halftime (PRT) of 143 ms in the mitral iaflow recordin&
70
THE AMERICANJOURNALOF CARDIOLOGY VOLUME72
tolic flow) observed in our study.14The third is the effect of the prolonged decay of left atria1to left ventricular diastolic pressure,which affectspulmonary venous flow diastolic filling, thus prolonging the duration and pressurehalftime of diastolic flow into the left atrium.4J2 The fourth reason is the effect of MS on atrial function.15 Patients with MS may have altered atrial myocardium, which may create the milieu for atrial fibrillation.16In patients with atrial fibrillation, timing of atrial contraction and relaxation is lost, resulting in absent early systolic filling; thus, the pulmonary venous systolic flow dependsentirely on descentof the baseof the heart during systole, which may already be affected by the anatomic derangement of MSi4 Without atrial relaxation, left atrial pressuremay be higher in systole, and thus the pulmonary veins may have a decreasedsystolic gradient into the left atrium, resulting in decreasedsystolic flow~>4’7>‘8 Univariate cotrelates of pulp venous flow in mitral stenosis: We found a good correlationbetweenin-
creasing diastolic pressure halftime (i.e., the pressure halftime of the diastolic flow into the left atrium) and increasing mitral valve pressurehalftime and valve area. This correlation reflects the prolonged left atrial pressure decay in increasing severity of MS, which prolongs the
FIGURE 5. A transesophageal recording of left pulmonary venous aad transthoacic recording of mitral inflow in an 80 year-old patient with severe mitral stenosis (MS) and atrial fibrillation (A FIB). lhe systolic (S) to diastolic (D) flow is markedly decreased and the diastolic pressure halftime (DPHT) is prolonged at 131 ms. Mitral inflow shows a markedly increased pressure gradient (MG) of 13 mm Rg and a prolonged pressure halftime (P)(T) of 314 ms.
JULY 1,1993
FlGURE 6. Scattelplots showing the relatiorr ship between pulmonary venous diastolic pressure halftime, mitral valve pressure halftime (top) and Doppler valve area (hottom) in 31 patients with atrial fibrillation and normal sinus tiythm with mitral stem sls.
0
50
100
150
200
250
300
350
400
Mitral Valve Pressure Half Time
zoo.
*.
*. -*.. -.
-.
‘.
.g 150l‘1= I” 2! SlW-------____
-em_
tz I? ,o 0 9” .n *-
*.
0 -? 0
50
100
150
.mJ
250
330
*.
*. 350
I 4w
Mitral Valve Area
pressure halftime across the mitral valve as well as the Wing of the pulmonary veins during diastole.2,4J2 Previous studies: In a transthoracic Doppler echocardiography study of the right upper pulmonary veins in 15 patients with MS, Keren et al4 described the effects of severe MS, as well as the effect of normal sinus rhythm and at&l fibrillation, on pulmonary venous flow. In patients with severe MS and normal sinus rhythm, they found the pulmonary capillary wedge pressure to RGURE 7. A pulse&wave Doppler tansesophageal recordia of the left superior pulmonary vein and a simultaneous left atrial (LA) pressure recording in a 37-yealcdd woman with severe mitral stenosis (MS) and atrial fibrillation (A FlB). Note ths prolonged diastolic pressure halfttme (DPRT) of 130 ms (arrow), coincident with the left atrial pressure decay, from the v wave to the troup of the y descent (arrow). D q diastolic; S q systolic. PULMONARY VEIN BY TEE IN MITRAL STENOSIS
71
be lower during systole than during diastole, accounting for increasedpulmonary venous systolic and decreased diastolic flow. Our study contirms the results of Keren et al in a larger patient population using the more acceptable and reliable transesophagealmeasurementsof both left and right upper pulmonary venous flo~.~ A recent investigation during mitral balloon valvuloplasty in MS showed that the blunted systolic flow pattern in the pulmonary veins before valvuloplasty immediately ‘ ‘normalizes” after the procedure, which increasesthe mitral valve area. Study limitations: The major limitation of this study is that we did not have simultaneous measurements of left atrial pressureand pulmonary venous flow and mitral valve flows. The correlation of pulmonary venous diastolic pressure halftime and mitral v@e pressure halftime would have undoubtedly been higher with simultaneous recordings. Recently, we simultaneously measuredleft atrial pressureand pulmonary venous flow in a patient with severe MS and found that prolonged left atrial pressuredecay coincided with prolonged pressure halftime and duration of diastolic filling (Figure 7). Pulmonary venous flow may be intluenced by many other confounding factors, including left atrial volume and atrial ejection fraction, left atrial compliance loading conditions, cardiac output, and left ventricular compfimce.Zl920
1. Keren G, Sherez J, Megidish R, Levitt B, Laniado S. Pahnonary venous flow pattern - its relationship to cardiac dynamics. A pulsed Doppler echocardiographic study. CircuUofi 1985;71:1105-1112. 2. Klein AL, Tajik AJ. Doppler assessment of pulmonary venous flow in healthy subjects and in patients with heart disease. J Am Sot Echo 1991;4:37%392. a. Bmmwald E. Valvular heart disease. In: Braunwald E, ed. A Textbook of Cardiovascular Medicine. Philadelphia: WB Saunders, 1992: 1007-1011. 4. Keren G, Pa&s A, Miller HI, Scherez J, La&lo S. Puhnomuy venous flow determined by Doppler echocardiography in mitral stenosis. Am J Cardiol 1990.65:
72
THE AMERICANJOURNALOF CARDIOLOGY VOLUME72
246-249. 5. Caste110R, Pearson AC, Lenzen P, Labovitz AJ. Effect of mitral regurgitation on pulmonary venous velocities derived from tramesophageal echocardiography color-guided pulsed-Doppler imaging. J Am CON Cardiol 1991;17:149%1506. 9. Klein AL, Obarski TP, Stewart WJ, &sale PN, Pearce GL, Husbands K, Cosgrove DM, Sakedo EE. Tramesophageal Doppler echocardiography of pulmonary venous flow: a new marker of mitral regurgitation severity. J Am Co11 Cardiol 1991;18:2518-526. 7. Seward JB, Khandheria BK, Oh JK, Abel MD, Hughes RW, Edwards WD, Nichols BA, Freeman WK, Tajii AJ. TrTsesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 1988;63:649-680, 8. Quinones MA, Waggoner AD, R&to LA, Nelson JG, Young JB, Winters WL, Rib&o LG, Miller RR. A new, simplified and accurate method for determining ejection fraction with hwdimensional echocardiography. Circulation 1981;64:
744753. 9. Hatle LK, Brubakk A, Tromsdal A, Angelsen B. Noninvasive assessment of pressure drop in mitral stenosis by Doppler ultrasound. Br Heart J 1978;40: 13 l-140.
10. Teague SM, Heinsimer JA, Anderson JL, Sublett I$ Olson EG, Voyles WF, Thadani U. Quantification of aortic regurgitation utilizing continuous wave Doppler nltrasonnd J Am Co11 Cardiol 1986;8:592-599. li. Cohen MV, Gorlin R. Modified orifice equation for the calculation of mitral valve area. Am Heart J 1972;84:839-840. 12. Klein AL, Cohen GI, Davison MB, Stewart WJ, Husbands 5, Pearce GL, Salcede EE. Effect of mitral stenosis on pulmonary venous flow by hansesophageal echocardiography (abstr). J Am Sot Echo 1991;4:297. 13. Yellm EL, Meisner JS, Nikolic SD, Karen G. ‘Ihe scientic basis for the relations between pulsed-Doppler trammitral velocity patterns and left heart chamber properties. Echocardiography 1992;9:313-338. 14. Natarajan D, Sharma VP, Chandra S, Dhar SK, Gaga M, Caroli B. Effects of percutanmus balloon mitral valvotomy on puhnonary venous flow in severe mitral stenosis. Am J Cardiol 1992;69:81&812. 15. Keren G, Et&n T, Sherez J, Z.&x AA, Megidish R, Miller HI, La&do S. Atrial fibrilltion and atrial enlargement in patients with mitral stenosis. Am Heart J 1987;114:1146-1155. 19. Thiedemann K-V, Ferrans VJ. Left atrial ultrastmchue in m&al valvular disease. Am J Path01 1977:89:575-604. 17. Keren G, Sonnenblick EH, LeJemtel TH. Mitral atmlus motion. Relation to pulmomuy venous and transmitral flows in normal subjects and in patients with dilated cardiomyopathy. Circtdation 1988;78:621-629. 19. Keren G, Bier A, Sherez J, Miura D, Keefe D, LeJemtel T. Atrial contraction is an important determinant of pulmonary venous flow. J Am Coil Cardiol 19867:
693-695. 19. Basnight MA, Gonzalez MS, Kershenovich SC, Appleton CP. Pulmonary venous flow velocity: relation to hemodynamics, mitral flow velocity and left atrial volume, and ejection fraction. J Am Sot Echo 1991;4:547-558. 20. Bartzckis T, Lee R, Yeoh TK, Grosin H, Schnittger I. Tramesophageal echoDoppler assessment of pulmonary venous flow patterns. J Am Sot Echo 1991;4:
457-464.
JULY 1,1993