Diastole is Not a Single Entity Four Components of Diastolic Dysfunction

Transcription

1 Physiology of Diastolic Function Made Easy James D. Thomas, MD, FACC, FASE Director, Center for Heart Valve Disease Bluhm Cardiovascular Institute Professor of Medicine, Feinberg School of Medicine, Northwestern University Chicago, Illinois Conflicts of interest: GE, Abbott, Edwards (honoraria) Spouse employment: Bay Labs Diastole is Not a Single Entity Four Components of Diastolic Dysfunction 1. Fill on a stiff pressure-volume loop 2. Delay LV relaxation 3. Lose diastolic suction 4. Suffer atrial systolic failure These rarely occur in isolation but considering them separately helps to understand diastole Page 1

2 How to Get Diastolic Dysfunction 1) Fill on Stiff P-V Loop Systolic Failure Diastolic Failure Normal Stiffer portion of same LV P-V curve Decreased diastolic chamber distensibility LV Pressure Left Ventricular Volume Left Ventricular Volume Lorell BH. Ann Rev Med 1991;42: Sources of Passive Elasticity Component Contribution Collagen ++++ Titin ++++ Actin ++ Intermediate + Microtubules - Passive Tension in Cardiac Muscle: Contribution of Collagen, Titin, Microtubules, and Intermediate Filaments Granzier HL, Irving TC Biophysical Journal 1995; 68: Page 2

3 Diastology 218 Available Echocardiographic Methods Transmitral velocity profiles Pulmonary venous profiles Doppler tissue imaging Color Doppler M-mode Future directions Mitral Inflow: E/A Velocity SV at MV Leaflet Tips Page 3

4 Left Ventricular Pressure Key parameters of the mitral inflow pattern: E velocity Decel time A velocity A Duration E/A ratio MV Deceleration Physical Determinants Decreased Diastolic Chamber Distensibility Left Ventricular Volume The stiffer the ventricle, the more rapid the pressure rise and reversal of p and the more rapid the deceleration dv/dt MVA/(LA-LV Compliance) Page 4

5 DT (msec) Physical Determinants of Deceleration Stiffer Ventricle = Shorter Decel Time Compliant p Compliant S A A S V v Stiff S A p Stiff t A S V DT 1/LV stiffness 3 25 y = e x r =.87 p < KLV (mmhg/ml) DT~14 msec stiffness ~.3 mmhg/ml Little et al. Circulation 1995;92:1933 Garcia et al. Am J Physiol 21;28:H554 Page 5

6 Survival [%] Transmitral Flow and Prognosis Restrictive Cardiomyopathy DT > 15 ms p<.1 DT < 15 ms Klein et al., Circ 1991; 83: Months Some Ventricles LOOK Like Amyloid Are there new diagnostic methods for less obvious ones? Page 6

7 HCM AS Phelan, Collier et al. Heart 212; 98: Amyloid: Apex > 2x Base & Mid Phelan, Collier et al. Heart 212; 98: Page 7

8 E/A ratio 1 How to Get Diastolic Dysfunction 2) Delay relaxation p(t) = p e -t/ mmhg LAP p 1 mmhg 1 2 Delayed Pseudonormal Normal relaxation Rate of rise is proportional to growth of transmitral pressure gradient, p, and d p/dt = LAP/ Preload vs Relaxation Confounding Effects best Garcia et al, JACC 1998;32:865 Diastolic function worst Page 8

9 Effect of Relaxation on LV Inflow = 28 LA = 8 = 31 LA = 8 = 43 LA = 8 With delayed relaxation, acceleration is slowed and E peak is lower. Choong, et al, Circ 1987 Effect of LAP on LV Inflow = 29 LA = 5 = 31 LA = 9 = 31 LA = 13 With rising LA pressure, acceleration is faster and E peak is higher. Choong, et al, Circ 1987 Page 9

10 Exercise Can Unmask Delayed Relaxation Relaxation Time Constant ( ) Diastolic dysfunction LVEDP Diastolic dysfunction Normal Normal Heart Rate (bpm) Heart Rate (bpm) Diastology 218 Available Echocardiographic Methods Transmitral velocity profiles Pulmonary venous profiles Doppler tissue imaging Color Doppler M-mode Future directions Page 1

11 Pulmonary Vein Doppler Acquisition AR S1 S2 D Pulmonary Vein Doppler Determinants of Waves Atrial contractility LA pressure LV stiffness MV area PVs1 PVs2 PVd LV relaxation MV area MR Cardiac output Atrial contractility LA pressure LV stiffness MV area PVa LV contractility Descent of annulus MR Page 11

12 LVEDP [mmhg] Predicting LVEDP From Mitral and PV A-wave Duration Mitral valve 6 E A t 45 3 r =.68 6 Pulmonary vein -4 S D AR t (PV - MV A-duration) [msec] Rossvoll & Hatle, et al., JACC 1993; 21: The Problem with All Flow-Based Indices of Diastolic Function Preload Sensitivity Really good E/A ratio Good Bad Really bad Diastolic Function Really, really bad Page 12

13 Needed: measures of LV systolic and diastolic function that are less dependent on preload Diastology 218 Available Echocardiographic Methods Transmitral velocity profiles Pulmonary venous profiles Doppler tissue imaging Color Doppler M-mode Future directions Page 13

14 Tissue Doppler Imaging Blood: High velocity, low amplitude Tissue: Low velocity, high amplitude S E A Myocardial Wall Velocities Do not pseudonormalize Mitral inflow Mitral annulus Normal Delayed relaxation Pseudonormal Restrictive Sohn et al., JACC 1997; 3: Page 14

15 Transmitral velocity [cm/s] Annular velocity [cm/s] Constriction vs Restriction M-mode Doppler MV Flow Normal Restriction Constriction E 5cm/s 5cm/s E E 2cm/s TDE Ea 5cm/s Ea 5cm/s Ea Garcia et al., JACC 1996; 27: Constriction vs Restriction Doppler Differentiation Mitral Pulsed Doppler 12 Annular DTI p= p= Garcia et al., JACC 1996; 27: C R C R Page 15

16 Assessment of LV Relaxation DTE E-wave Inversely Related to PW-E w Anterior MI Controls E/A >1 E/A <1 If E LAP/ and Ea 1/, then LAP E/Ea Tau Oki et al, Am J Cardiol 1997;79:928 Estimation of P LA The Magic of E/e Nagueh et al. JACC 1997;3: Page 16

17 E/E'sep Estimation of P LA The Magic of E/e Does this always work? Not if the heart is normal Estimation of Left Atrial Pressure Subjects without Heart Disease y =.19x r =.1 SEE: 3.4 mmhg Pulmonary Capillary Wedge Pressure (mmhg) Firstenberg et al. J Am Coll Cardiol 2; 36: Page 17

18 Estimation of P LA The Magic of E/e Does this always work? Not if the heart is normal And not if the heart is really, really sick Circ 29; 119: 62-7 EF = 24% E = 135 cm/sec Lateral e = 2 E/Ea = 67 PCW = 14 mmhg EF = 31% E = 89 cm/sec Lateral e = 6.9 E/Ea = 14 PCW = 33 mmhg Page 18

19 Correlation Mitral E/E - PCWP Mullens et al. Circulation 29; 119: 62-7 Correlation Mitral E/E Direct LAP HCM Patients Geske et al. Circulation 27; 116: Page 19

20 Don t forget left atrial area!! The HbA1c of end-diastolic pressure How to Get Diastolic Dysfunction 3) Lose Diastolic Suction P Normal Ischemic Courtois et al. Circulation 199;81: Page 2

21 IVPG are Critical During Exercise Diastole Disproportionately Shortened IVPG IVPG Diastolic Filling Time Full Cardiac Cycle Cheng, Circ Res 1992;7:9-19 Invasive Measurement of Intraventricular Pressures AO LA RV Apex Mid Base Completely impractical in clinical practice Page 21

22 Color M-Mode Methodology Color M-Mode Methodology Page 22

23 Measurement of Propagation Velocity 1 sec LV V p = 3 cm/sec 1 cm LA Color M-mode Doppler Does Not Pseudonormalize Normal Delayed relaxation Restrictive Page 23

24 p IV (Doppler) IVPG (mmhg) CMM Calculation of IVPG E A Time (s) P LV apex LV base v v v s t s Greenberg et al. Am J Physiol 21;28:H Euler equation Intraventricular Pressure Gradient 8 6 y =.87x +.22 r =.96 SEE =.35 mmhg p < p IV (catheter) Greenberg et al: Am J Physiol 21; 28; H257-H2515 Page 24

25 VO2 max (ml/kg/min) Correlation Between Delta IVPG and VO 2 max VO2max = 7.82 * IVPG R =.79 P < Delta IVPG (mmhg) Heart Failure Normal Subjects Rovner et al. Am J Physiol 25; 289: H281-8 Left Ventricular Torsion The spiral architecture of the LV produces base-apex torsion This stores energy in systole that is released in diastole (suction) Though important in LV mechanics, torsion has been difficult to measure MRI courtesy of Ed Shapiro, Johns Hopkins University Page 25

26 Torsion During Exercise 8 Rest Apical twist 18 Exercise Apical twist Notomi et al. Circ 26; 113: Timing and Magnitude of LV Mechanics Impact of Exercise Rest Exercise Long- and Short-axis, cm/s M C A O EJ A C M O Contraction/ Torsion EF (E) Expansion Untwisting Lengthening Time, % Systolic duration Torsion/ Untwisting, rad/s Long- and Short-axis, cm/s M C A O EJ A C M O EF (E) Contraction/ Torsion Untwisting Lengthening Time, % Systolic duration Expansion Torsion/ Untwisting, rad/s Long axis motion Radial motion Torsion Notomi et al. Circ 26; 113: Page 26

27 Torsion-Volume Loop TV 1 AC Ejection 75 Torsion, %max 5 25 MO Pk-E AO MC En-E LV volume, %SV Notomi et al. Circ 26; 113: LV Untwisting Predicts Suction Gradient 7.5 y = -.44x +.87 r =.72 IVPG, mmhg Peak LV untwisting, rad/s Notomi et al. Circ 26; 113: Page 27

28 Putting It All Together During systole, a significant amount of elastic energy is stored in the myocyte and the interstitum as torsion The earliest mechanical manifestation of diastole is an abrupt untwisting that is largely completed before the mitral valve opens This untwisting helps to establish a base-to-apex intraventricular pressure gradient in early diastole that assists in the low pressure filling of the heart Modulation of this mechanism allows the heart to augment its function many-fold during exercise How to Get Diastolic Dysfunction 4) Atrial Dysfunction Mitral Valve A Pulmonary Vein Assessment complicated by having two outlets and no isovolumic period AR Page 28

29 Three Components of Atrial Function Pump Conduit Reservoir S E D A Diseases with Atrial Dysfunction Atrial fibrillation/flutter Heart transplant Amyloidosis Mitral stenosis/regurgitation Atrial fibrosis AR Guidelines Approach to Grading Diastolic Dysfunction, ca. 29 Septal e Lateral e LA volume Big Problem There are 8 combinations of these parameters, but only 3 fit the algorithm! Septal e 8 Lateral e 1 LA < 34 ml/m2 Septal e 8 Lateral e 1 LA 34 ml/m2 Septal e < 8 Lateral e < 1 LA 34 ml/m2 E/A <.8 DT > 2 ms Av. E/e 8 Ar-A < ms Val E/A <.5 E/A DT 16-2ms Av. E/e 9-12 Ar-A 3 ms Val E/A.5 E/A 2 DT < 16 ms Av. E/e 13 Ar-A 3 ms Val E/A.5 Normal. function Normal function, Athlete s heart, or constriction Grade I Grade II Grade III Nagueh et al. JASE 29; 22: Page 29

30 A Room with Eight Ways Out But 5 of them are locked! How Well Do These Work in Practice? 41 consecutive patients, age 59±16 years (6%M) Using only the 3 primary classifiers (LAVi, septal and lateral e ), diastolic function could be assigned in only 34% of cases For the 5 secondary indices (E/A ratio, E deceleration time, E/E, PV AR reversal duration, and E/A with Valsalva), concordance (3+/5 indices in agreement) occurred in only 64% of cases. Let s take another swing at the guidelines! Page 3

31 Nagueh et al. JASE 216; 29: Grading Diastolic Dysfunction With LVEF<5% or Other LV Disease Page 31

32 How applicable are the new guidelines??? Stay tuned for Jae Oh s analysis Keys to Handling Discrepant Indices Discount technically limited indices Don t overinterpret garbage Look at the atrium Normal atrial size virtually precludes severe DD Large atrium must be explained but DD isn t only cause (consider AF, MR, and MS) Atrial systolic failure must be recognized Despite our linear grading scheme, diastole is far more complex Consider early and late diastole separately Page 32

33 Diastole is Not a Single Entity Four Components of Diastolic Dysfunction 1. Fill on a stiff pressure-volume loop 2. Delay LV relaxation 3. Lose diastolic suction 4. Suffer atrial systolic failure These rarely occur in isolation but considering them separately helps to understand diastole Page 33