Left VentricularLeft Ventricular FunctionFunctionLeft Ventricular Left Ventricular FunctionFunctionHow to Measure and How

Accurate is EchoAccurate is Echo

Neil J. Weissman, MDNeil J. Weissman, MDMedStar Health Research InstituteMedStar Health Research Institute

&Professor of MedicineGeorgetown University

Washington, D.C.

DisclosuresDisclosures

Grant support (to institution) for Core Lab activities:pp ( )

• Abbott Vascular• Boston Scientific• Boston Scientific• Biotronic• Direct Flow

Ed d Lif S i• Edwards LifeScience• Medtronic• MitrAligng• St. Jude• Sorin/ Carbomedics

For a full list, visit www.EchoCoreLab.org

Thanks to our ASE Past President:

Michael H. PicardMassachusetts General Hospital

Harvard Medical School

LV function in clinical practice: What is the echocardiographer asked?

• Diagnosis – systolic (and diastolic) dysfunction– Etiology for symptomsgy y p

• Assessing response to treatment• Assessing risk and prognosis• Assessing risk and prognosis

– Need for interventions• Defibrillators valve surgery meds CRT• Defibrillators, valve surgery, meds, CRT

– Timing of interventions

Systolic function by echo:an important marker of riskan important marker of risk

Post - MI CHF EF < 35%

GISSI 2Volpi A, et al, Circulation 1993;88:416

SOLVD Registry Data

GISSI - 2

Quinones et al : JACC 2000;1237-1244

When is it appropriate to use echo to quantify ventricular function ?quantify ventricular function ?

Wh h i f d• When ever echo is performed• Why ?

– Echo measures of ventricular function are all validated and standard

– Requesting MDs expect it and will use it• Keeps echo competitive with other modalities

• If concern that image quality is insufficient to measure LV systolic function– Then use contrast– Consider methods that do not require border delineation

J AM Soc Echocardiogr 2005; 18:1440‐1463

Approximately 5500 citations5500 citations

iASE in iTUNE

Cardiac ChamberQuantification: What is New?

Database 

RT3DEDeformation Imaging

J Am Soc Echocardiogr 2015;28:1‐39

Partition Values for Severity of yAbnormalities

Cutoffs based on SD◦ Data readily exist◦ Echo parameters are not normally distributed◦ Asymmetric distribution 

Cutoffs based on percentile values (95th)Cutoffs based on percentile values (95th) Cutoffs based on outcomes or prognosisCutoffs experienced based consensusLV EF LA LA si e and LV massLV EF, LA,  LA size and LV mass

Normal Reference Values for 2DE

Seven data bases (Asklepios FlemenghoSeven data bases (Asklepios, Flemengho, Cardia5, Cardia 25, Padua 3D Echo Normal, NorreStudy)Study)

No contrast studiesAge, gender, ethnicity, height and weightNl BP no diabetes nl BMI creatinineNl BP, no diabetes, nl BMI, creatinine, glomerular filtration rate, cholesterol, LDL and trigliceridesand triglicerides

Left Ventricle and Left AtriumLeft Ventricle and Left Atrium

How do we Assess LV Function ?EyeEye ballball LV Function ?

SubjectiveSubjective Experience dependentExperience dependentQualitativeQualitative p pp p Lack of standardizationLack of standardization Large interLarge inter-- and intraand intra--

AssessmentAssessmentobserver variabilityobserver variability

Left Ventricular Linear Measurement

lNormal Mildly Moderately Severely

Male

LV DiastolicDiameter/BSA 2.2‐3.0 3.1‐3.3 3.4‐3.6 >3.6

LV SystolicDiameter/BSA 1.2‐2.1 2.2‐2.3 2.4‐2.5 >2.5

FemaleNormal Mildly Moderately Severely

LV Diastolic 2 3 3 1 3 1 3 3 3 4 3 6 3 6LV DiastolicDiameter/BSA 2.3‐3.1 3.1‐3.3 3.4‐3.6 >3.6

LV SystolicDiameter/BSA 1.3‐2.1 2.2‐2.3 2.4‐2.6 >2.6

Left Ventricular Volumetric MeasurementMeasurement

TEICHHOLZ FormulaTEICHHOLZ Formula

Left Ventricular Volumetric MeasurementBiplane Disk Summation1

Area Length Method

2

LV Volumes by 2D

Normal Mildly Moderately Severely

Male

Normal Mildly Moderately Severely

LV DiastoilcVolume/BSA 34‐74 75‐89 90‐100 >100

LV SystoilcVolumer/BSA 11‐31 32‐38 39‐45 >45

Normal Mildly Moderately Severely

Female

LV Diastolic Volume /BSA 29‐61 62‐70 71‐80 >80

LV SystoilcLV SystoilcVolume/BSA 8‐24 25‐32 33‐40 >40

3

3D echo for volume and EFtriplane imaging and manual tracing

Linear regression of LVEF in all patients, measured by 3D echocardiography by Simpson's method (3DS) vs radionuclide

angiography (RNA)g g y ( )

Copyright ©1996 American Heart Association

Nosir, Y. F.M. et al. Circulation 1996;94:460-466

Results of LVEF measurements plotted as differences between methods and analysis of agreement

Copyright ©1997 American Heart Association

Buck, T. et al. Circulation 1997;96:4286-4297

LV Ejection Fraction

PreloadPreloadPreloadPreload

Heart RateHeart RateLV EFLV EF

AfterloadAfterload

ContractilityContractilityContractilityContractility

Potential problems with LVEF

• Load dependency• Measurement issuesMeasurement issues

– Endocardial dropout• Overestimation of volumeOverestimation of volume

– Foreshortening of the ventricle• Underestimation of volumeUnderestimation of volume• less effect on EF

Potential pitfalls of EF measurement ( ti d)(continued)

G i i– Geometric assumptions• Influence EF measure when LV distorted

– Regional dysfunctionRegional dysfunction• Over or under-represented with some methods

– Paradoxical septal motion, other discoordinations of contraction

• Underestimation of EF

– Heart rate effectsHeart rate effects• tachycardia

– reproducibility

Left Ventric lar Ejection FractionLeft Ventricular Ejection Fraction

Normal Mild Moderate SevereNormal Mild Moderate Severe

2015 >52 51‐41 40‐30 <302015 >52 51 41 40 30 <302005 >55 54‐45 44‐30 <30

LV Ejection Fraction

Normal Mildly Moderately Severely

Male

LVEF 52‐72 41‐51 30‐40 <30Female

Normal Mildly Moderately Severely

Female

LVEF 54‐74 41‐53 30‐40 <30

LV Global Longitudinal StrainLV Global Longitudinal Strain

Peak GLS in thePeak GLS in the range of -20% can be expected in abe expected in a healthy person

Low Flow ASC di l Cardio-oncology

ValvularR it tiRegurgitation

1. Normal or Hyperkineticyp

2. Hypokinetic (reduced (thickening)

3. Akinetic(absent or negligible g gthickening

4. Dyskineticy(systolic thinning or gstretching)

Real-time 3D echo and automated border detection: assessment of LV volumes and EFassessment of LV volumes and EF

What if image quality inadequate ?What if image quality inadequate ?

Use contrast for LVO to assessUse contrast for LVO to assess LV global and regional function

S l ll iSegmental Wall Motion

Without contrast With contrast

Comparison with RNA of echo LV EF by Simpson’s method fundamental or harmonicSimpson s method, fundamental or harmonic,

contrast or non-contrast

F F+CF F+C

H H+CH H+C

Nahar et al, AJC 2000;86:1358

Interobserver Variabilty LoweredInterobserver Variabilty Lowered with Contrast Down to MRI Levels

Hoffman et al, EHJ 2005;26:607-616

Other measures of LV systolic function that do not rely on endocardial border delineationrely on endocardial border delineation

• Isovolumic indices– dP / dt

• Ejection phase indices• Ejection phase indices– Time interval (Doppler)

• Tei indexTei index– M mode

• Fractional shortening (FS)• Velocity of circumferential fiber shortening (Vcf)

Quantitation of global LV systolic functionQ g y

• Isovolumic indicesdP / dt

• Easy to measure – MR CW DopplerA d• Automated

• Mean dP/dt correlates well but underestimates dP/dt max (dP/dt max depends on time of peak systolic pressure)

• Instantaneous dP/dt accurate measure of dP/dt max

MR t b t• MR must be present• Maximum spectra must be recorded

– Can use contrast to enhance weak signal

• Not truly isovolumic

Measuring mean dP/dT

• CW Doppler of MR

• Measure time interval for velocity to increasefrom 1 m/s to 3 m/sfrom 1 m/s to 3 m/s

• dP/dt = 32/t

Kolias, et al. Kolias, et al. JACCJACC 2000;36:15942000;36:1594

Improved dP/dt after CRT

Fan et al, JASE 2004:17:553

dP/dt by echo for HF outcomes

Kolias, et al. JACC 2000;36:1594

Quantitation of global LV systolic function:Ejection phase indicesEjection phase indices

Doppler total ejection isovolume indexDoppler total ejection isovolume indexTei index

• Doppler measure• No geometric assumptions• Less dependent on load• Less dependent on load

• Requires accurate IVRT, ET, ICT• Pseudonormalization

Prognostic value of Tei Index in CHF

Cardiac amyloidosis

Tei et al, JACC 1996;28:658-64

Idiopathic dilated cardiomyopathy

Am J Cardiol 1998;82:1071-1076, ; ;

2D strain without need for border delineationtracking speckle with 2 ROIs

ROI 1

g p

D(0) D(t)

aROI 2

y = 0.882x + 0.1413R2 = 0.9918P<0.0001

20

25

30

EC

HO

(%)

y = 1.0537x + 0.9193R2 = 0.9646P <0.0001

10

12

14

16

HO

(deg

)

0

5

10

15

0 5 10 15 20 25 30 35

ST

RA

IN b

y E

STRAIN by SONOMICROMETERS (%) 0

2

4

6

8

10

0 2 4 6 8 10 12 14

RO

TATI

ON

by

ECH

Tournoux et al, J Am Soc Echocardiogr 2008;21:1168-1179

y ( ) 0 2 4 6 8 10 12 14

ROTATION by CT scan (deg)

Apical Rotational Mechanics

Courtesy of Manni Vannan, MD

Summary– Quantitation of LVEF

• 2D Biplane Simpson or 3D• 2D Biplane Simpson or 3D• Still the foundation of LV systolic function• limitations exist but it remains a trusted measure that has tat o s e st but t e a s a t usted easu e t at as

prognostic value

– Semi-quantitative assessment of regional LV function

• Qualitative function, quantitative location and size

– Use of contrast to improve LV function assessment– Specialized conditions may require novel measures

( i k l ki i )(strain, speckel tracking, torsion)

Questions?Questions?