Comprehensive Validation of Cardiovascular …circimaging.ahajournals.org/content/circcvim/early/2013/04/03/CIRC...Comprehensive Validation of Cardiovascular Magnetic Resonance Techniques

Comprehensive Validation of Cardiovascular Magnetic Resonance

Techniques for the Assessment of Myocardial Extracellular Volume

Miller et al: Validation of CMR Assessment of ECV

Christopher A. Miller, BSc, MBChB1,2,3; Josephine Naish, PhD2; Paul Bishop, MBChB4;

Glyn Coutts, PhD5; David Clark, BSc;6 Sha Zhao, PhD2; Simon G. Ray, MD1,3; Nizar Yonan, MD1;

Simon G. Williams, MD1; Andrew S. Flett, MD7,8; James C. Moon, MD7,8; Andreas Greiser, PhD;9

Geoffrey J.M. Parker, PhD2; Matthias Schmitt, MD, PhD1,2

1North West Heart Centre and The Transplant Centre, University Hospital of South Manchester,

Wythenshawe Hospital, Manchester, UK. 2Centre for Imaging Sciences & Biomedical Imaging

Institute, University of Manchester, UK. 3Cardiovascular Research Group, University of Manchester,

UK. 4Department of Pathology, University Hospital of South Manchester, Wythenshawe, Manchester,

UK. 5Christie Medical Physics and Engineering, The Christie Hospital, Manchester, UK. 6Alliance

Medical Cardiac MRI Unit, Wythenshawe Hospital, Manchester, UK. 7The Heart Hospital, London,

UK. 8Institute of Cardiovascular Science, University College London, London, UK. 9Healthcare

Sector, Siemens AG, Erlangen, Germany.

Correspondence to Dr Christopher Miller North West Regional Heart Centre University Hospital of South Manchester Wythenshawe, Manchester, UK Tel: +44 161 291 4940 Fax: +44 161 291 4940 Email: [email protected] DOI: 10.1161/CIRCIMAGING.112.000192

Journal Subject Codes: [30] CT and MRI; [124] Cardiovascular imaging agents/Techniques

pital of South MMMMananaaaaa

es & &&&&&& BiBiBiBiBiBiBiomomededdddddiiiicicii alalalalalalal IIIIIIIm

y f

o e

ical Physics and Engineering, The Christie Hospital, Manchester, UK

M t

y yyyy ofofofofof MMMMMannnnnchchchchheseesesester, UK. 3Cardiovascuulaaar Researccccch hhhh GrGrrrroouoo p, University of

off fff PPaPPP thologgggyy, Unininiversrsrsrsitiiii y HoHoHHH spppitalall off SSSououuouuththththth MMaanncheesttterr,rrr Wyytytyy hehehhh nnshaawwwe

ical Physics aaaaandndndndnd EEEEEngngngngngininnnneeeeeeeeeeririririringngngngng, , ,, , ThThThThThe eee ChChhChhriririrristststss ieee HHHHHososososospipipipip tatatatatal,ll,l MMMMManananananchester, UK

MRI Unittttt,, WyWyWyWyWythththththenenenenenshshhshhawawawawee e HoHoHoHoHospspspspspititititi alalalalal, , MaMaMaMaManncnnnchehehehehestststststererererer, , UKUKUKUKUK. 77777ThThThThTheeeee HeHeHeHeHearaararart Hospppit

by guest on July 8, 2018http://circim

aging.ahajournals.org/D

ownloaded from

by guest on July 8, 2018

http://circimaging.ahajournals.org/

Dow

nloaded from



ownloaded from



Dow

nloaded from



ownloaded from



Dow

nloaded from



ownloaded from








2

Abstract

Background—Extracellular matrix expansion is a key element of ventricular remodeling and a

potential therapeutic target. Cardiovascular magnetic resonance (CMR) T1-mapping techniques are

increasingly used to evaluate myocardial extracellular volume (ECV), however the most widely

applied methods are without histological validation. Our aim was to perform comprehensive validation

of; A, dynamic-equilibrium CMR (DynEq-CMR), where ECV is quantified using hematocrit-adjusted

myocardial and blood T1 values measured before and after gadolinium bolus; and B, isolated

measurement of myocardial T1, used as an ECV surrogate.

Methods and Results—Whole-heart histological validation was performed using 96 tissue samples,

analyzed for picrosirius red collagen volume fraction (CVF), obtained from each of 16-segments of

the explanted hearts of 6 patients undergoing heart transplantation who had prospectively undergone

CMR prior to transplantation (median interval between CMR and transplantation 29 days). DynEq-

CMR-derived ECV was calculated from T1 measurements made using a modified look locker

inversion recovery sequence before and 10- and 15-minutes post-contrast. In addition, ECV was

measured 2-20 minutes post-contrast in 30 healthy volunteers. There was a strong linear relationship

between DynEq-CMR-derived ECV and histological CVF (p<0.001; within-subject r=0.745, p<0.001;

r2=0.555; between-subject r=0.945, p<0.01, r2=0.893; for ECV calculated using 15-minute post-

contrast T1). Correlation was maintained throughout the entire heart. Isolated post-contrast T1

measurement showed significant within-subject correlation with histological CVF (r=-0.741, p<0.001;

r2=0.550 for 15 minute post-contrast T1), but between-subject correlations were not significant.

DynEq-CMR-derived ECV varied significantly according to contrast dose, myocardial region and

gender.

Conclusions—DynEq-CMR-derived ECV shows a good correlation with histological CVF throughout

the whole heart. Isolated post-contrast T1 measurement is insufficient for ECV assessment.

Key Words: collagen, histopathology, magnetic resonance imaging, myocardial fibrosis

o o ooooo hahahahahahahad d ddddd prprprprprprprososososososospepepepepepepectctctctctctctivivivivivivivelelelelelelelyyyyyy

splantntntntntntntatatatatatatatioioioioioioionnnnnnn 29292929292929 dddddddaayaaap ( p y

V o

E

n

p ( p y

V V VVV wwaww s calculululululatataata eddddd fffffrororororom m mm m TTTTT1111 mememmemeaaasurrrrreeemenenntsssss mmmmmadadadadade usussinininnnggggg a momomommodididididifififififiededdeded llllloooooooook kkkk lo

seqeqeqqqueueueueuencncncncncee ee bebebeefofofoff rerererre aaaaanddddd 11110-0-0-0-0- andndndndnd 151515155-m-m-m-m-minininininututututu esesesese pppppososososost-t-t-tt-cococococontntntntntrararastss . IIIIIn nnnn adadadaa dididdid titititit onononno , EEE

nutes popp st-contttrast inii 33330 0000 hhehh alllllththhththyyy volulll nttteers. ThThThThere eee wwwaww s a aaaa stronggg linear



ownloaded from


3

Expansion of the myocardial interstitial space is a feature common to a range of cardiac pathologies,

and appears fundamental to the process of adverse left ventricular (LV) remodeling.1-4 Interstitial

expansion, in part due to accumulation of collagen, is associated with changes in mechanical and

electrical properties of the myocardium and as such may represent a sentinel phenotype, transitional

between ‘healthy’ myocardium and ‘diseased’ myocardium associated with increased mortality risk.5-9

Furthermore, interstitial expansion is reversible and a potential therapeutic target.10-12 Quantification of

the myocardial interstitial space, or extracellular volume (ECV), may therefore represent an important

diagnostic and prognostic biomarker.

Endomyocardial biopsy represents the current gold standard for assessment of the myocardial

interstitium, however its invasive nature, and lack of whole heart coverage, restrict its clinical use for

this purpose. Cardiovascular magnetic resonance (CMR) techniques that potentially allow non-

invasive evaluation of the interstitial space have generated considerable recent interest.

Gadolinium chelates are standard extracellular contrast agents that potently shorten T1 relaxation time.

Gadolinium concentration is directly related to change in the relaxation rate R1 (where R1 is the

reciprocal of T1). Therefore T1 measurement (‘mapping’) before and after contrast administration can

be used to calculate gadolinium concentration in myocardium and blood. At contrast agent

equilibrium, gadolinium concentration is equal in blood and myocardium, and since the volume of

distribution of gadolinium (or ECV) in blood is known from the hematocrit, so myocardial ECV can

be calculated.

This approach requires a steady state to exist between blood and myocardial contrast agent in order for

the effect of contrast agent kinetics to be removed. Using a primed gadolinium infusion to achieve

contrast agent equilibrium (Equilibrium Contrast CMR, EQ-CMR), Flett et al13 demonstrated a strong

correlation between CMR measurement of myocardial ECV and histological collagen volume fraction.

smsmsmsmsmsmmenenenenenenent t ttttt ofofofofofofof ttttttthehehehehehehe mmmmmmmyoyoyoyoyoyoyocaccacacaaca

erageeee rerererererereststststststs riririririririctctctctctctc iiiiiiitststststststs cccccl, g ,

iovascular magnetic resonance (CMR) techniques that potentially al o

n

, g ,

iooooovvavv scular mmmmmagagagnnennn tititititicccc c rererereresososososonananananancncncn e (C(C(C(C( MRRR) tetetetetechchchchchninnnn quququesese ttttthahahahh t popopopopoteeteteentntntntntiaaaalllllllllly yyyy alalalalalloll

n ofofofoff ttttthehehehehe iiiiinnnnnteeersrsrsr tititttitititiiialalalll spaaaaacecececece hhhhhavvvvvee eee gegegegegenenenenenerarararr tetetetet d dddd cococococonsnsnsnsnsidididididerererererababababablelele recececececenenenenent tttt innnnnteteteteterererererestststs . nn



ownloaded from


4

However, due to the potentially cumbersome nature of this technique, two alternative methods of

assessing ECV have become more widely adopted.

The first, which involves a contrast agent bolus (i.e. without a subsequent infusion), assumes contrast

agent kinetic effects to be negligible due to a dynamic equilibrium between blood and myocardium

(dynamic-equilibrium CMR; DynEq-CMR).9, 14-19 In the second method, an isolated myocardial T1

measurement is made at a fixed time-point following a bolus of contrast agent (i.e. without pre-

contrast, or blood T1 measurements) and is used as a surrogate measure of ECV (“isolated-T1”).20-23

However, despite these techniques being applied to a rapidly expanding number of pathologies, both

have potential shortcomings; the assumed dynamic equilibrium proposed in the first method may not

always hold true in reality, and the second technique may be confounded by factors such as body fat

percentage, renal function and hematocrit. Systematic in vivo histological validation of these

techniques in humans is lacking.24-26

The aims of the current study were to first, provide comprehensive, whole-heart, histological

validation of the DynEq-CMR technique for measurement of myocardial ECV. This, at the same time,

allowed assessment of the validity of isolated-T1 technique. Second, we aimed to assess the effect of

contrast agent dose, post-contrast acquisition time, myocardial regionality, cardiac cycle and gender

on DynEq-CMR ECV measurement.

Methods

Study Design

All research was carried out at University Hospital of South Manchester NHS Trust, UK. An ethics

committee of the UK National Research Ethics Service approved the study and written informed

consent was obtained from all participants. The work was conducted according to the Helsinki

Declaration.

dededededededed d ddddd bybybybybybyby fffffffacacacacacacactototototototorsrsrsrsrsrsrs sssucucucucucucuchhhhhhh

ical vavavavavavavalilililililil dadadadadadadatititititititiononononononon ooooooofffff ty g

a

rrent study were to first, provide comprehensive, whole-heart, histolog

y ggg

aaaaansnsnssn is lackkkkinnnnngg.24-4-44 262626262

rrent study yy were to fifififif rst,t,t, pppro iviiiidddedd comprpp ehen isiive, ,, whwhwhwwhollle-e-e-hhhehh art,,, histologg



ownloaded from


5

The study comprises 3 parts: (1) Phantom studies; performed in order to validate the accuracy of the

T1 mapping sequence used and to calculate a T1 heart-rate correction algorithm (see Online Data

Supplement); (2) Histological validation. CMR was performed prospectively in patients awaiting heart

transplantation. When these patients subsequently underwent transplantation, the explanted hearts

were used to provide whole-heart histological validation of the DynEq-CMR and isolated-T1

techniques; and (3) The effect of contrast dose, post-contrast acquisition time, myocardial regionality,

cardiac cycle and gender on DynEq-CMR were assessed in healthy volunteers.

CMR imaging

All CMR imaging was performed with the same 1.5 tesla scanner (Avanto; Siemens Medical Imaging,

Germany) equipped with a 32-element phased-array coil.

T1 measurements were made using an electrocardiogram (ECG)-gated single-shot modified Look

Locker inversion recovery (MOLLI) sequence as described by Messroghli et al.27 Typical parameters

were FOV 340x255mm, matrix 192x138, 8mm slice thickness, flip angle 35°, iPAT factor (GRAPPA)

2 with 24 reference lines, 6/8 partial Fourier k-space sampling, acquisition time 201ms for a single

image, initial effective inversion time (TIeff) 100ms with a TIeff increment of 80ms. To sample T1

recovery, serial single-shot diastolic images were acquired every heart beat after 3 non-selective

adiabatic inversion pulses (i.e., 3, 3, and 5 images after each inversion pulse, totaling 11 images), with

3 dummy heart beats before the second and third inversion pulses to allow recovery (17 heart beat

total acquisition duration).

The gadolinium contrast agent used throughout the study was gadopentetate dimeglumine (Gd-DTPA;

Magnevist; Bayer Healthcare), administered as a single bolus of 2mL/s followed by a 30mL saline

chaser bolus delivered at the same flow rate using a power injector.

w f

e a

5 t

wwwwwerrrre made uuuuusisisiss nggggg aaaaan nnnn eleleleleleceeee trtrtrtrtrococooo ararardidididiiogogo raammm (EE(EE(ECGCGCGCGCG)-gagaateteteeed dd dd ssingngngngnglelellele-s-s-s-s-shohohohohot ttt t momomomomodidddd f

ecovovovovovererererery yyyy (M(M(M(M(MOLOLOLOLOLLILILILL ))))) sequququququenenenenence aaaaass sss dededededescscscscscririririribebebebebedd ddd bybybybyby MMMMMesesesesessrsrsrsrsrogogogoghlhlllhliiiii etetetee aaaaalll.ll 27777 TTTTTypypyppy icicicica

5mm, ,, matrix 111119292929292x11138383833 , ,, 88m888 m slllllice hthththiciiii kkknk ess,,, ffflililiiip pp anananannglglle 3535353535°,, iPAT fact



ownloaded from


6

Histological validation

All patients on the heart transplant waiting list at University Hospital of South Manchester NHS Trust,

UK (1 of 6 UK adult heart transplant centers), between January 1st 2011 and July 1st 2012 were

screened for study eligibility. Of the 54 patients on the waiting list during this period, 41 had an

intracardiac device that prohibited CMR, and 2 of the 13 patients without devices were deemed too

unwell for CMR by the supervising medical team. Eleven patients were therefore invited for CMR, of

which 2 refused consent. The 9 remaining patients underwent CMR, of which 6 underwent heart

transplantation. As pre-specified, no patients with acute myocardial inflammation (known or suspected

myocarditis, acute myocardial infarction) or cardiac amyloidosis were included.

The CMR protocol included LV short-axis MOLLI sequence acquisition at basal, mid and apical

ventricular levels before, and 10- and 15-minutes after a bolus of 0.20mmol/kg Gd-DTPA contrast

agent (Figure 1). Standard long- and short-axis steady-state free precession (SSFP) cine imaging was

performed in order to assess LV mass and volumetric parameters. Standard late gadolinium

enhancement (LGE) imaging was performed at least 10 minutes following contrast agent

administration using spoiled gradient echo segmented inversion recovery, and phase sensitive

inversion recovery (PSIR) segmented gradient echo, sequences.28 Blood samples were taken at the

time of CMR in order to measure hematocrit.

At the time of transplantation, the explanted hearts were immediately fixed in 10% buffered formalin.

The hearts were cut at basal, mid and apical LV levels (using the MOLLI slice positions as guidance,

Figure 1) and 16 tissue blocks were taken from the LV of each heart (96 samples in total) according to

the American Heart Association/American College of Cardiology 16-segment model29 before being

embedded in paraffin and stained with picrosirius red. High-power magnification (x200) digital

images, excluding perivascular areas, underwent automated image analysis (macro written in

ImageJ30). As described previously,13 a combination of standard deviation from mean signal and

isodata automatic thresholding derived the collagen area, expressed as a percentage of total

myocardial area, excluding fixation artifact. Twelve high-power fields were assessed per segment

ooooooon n nnnnn atatatatatatat bbbbbbbasasasasasasasalalalalalalal, ,,,, , , mimimimimimimid d ddd d d ananananananan

mmololololololol/k/k/k/k/k/k/kggggggg GdGdGdGdGdGdGd DDDDDDDTTTTTPTPT, g

r i

E imaging was performed at least 10 minutes following contrast agen

, g

tttttanananndard lonnnnng-g-g-gg aaaannnnd d d dd shshshshshororororort-tttt axaxaxaxaxisiisiss steteteteeadaddady--stattttte e e ee frfrfrfrfreeeeee prprp ececccceesesee siononononon ((SSSSSSSSSSFPFPFPFPFP) ) ))) cicicicicinnennn

r to oo asasasasassesesess ssssssssss LLLV VVVV mamamamam ssssssss andndndndnd vvvvvoluuuuumememememetrtrtrtrtricicicicic ppppparararara amamamamametetetetetererererers.s SSSSStatatatatandndndddarrrrrdd ddd lalalaaatetetetete gggggadadadaddolololoo inininnni

E)) imaggging gg was pepp ffrffformed dd attt lllllea tst 1110 000 iimiinutes fofff lllllllllowowowowowinii g g g gg cocccc ntrast agegg n



ownloaded from


7

(1152 high-power fields were assessed in total) and mean histological collagen volume fraction (CVF,

%) and CVF heterogeneity (coefficient of variation between fields) were obtained. In order to

determine whether CVF measurements were affected by the degree of magnification used to acquire

the histological images, CVF measurements were repeated at 2 additional magnifications (x50 and

x100) in 2 randomly selected tissue blocks from each patient (12 samples (13%) in total, i.e. an

additional 288 histological images).

Effect of contrast agent dose, acquisition time, myocardial regionality, cardiac cycle and gender

Thirty completely asymptomatic healthy volunteers, with no known risk factors or history of cardiac

disease, normal physical examination and normal ECG, underwent CMR. (Subjects were not patients

who had been referred for CMR that was subsequently found to be normal). CMR (cine and LGE

imaging) was normal in all cases.

Volunteers were prospectively split into 3 age- and sex-matched groups, and received Gd-DTPA

contrast agent at the following doses: Group A 0.10mmol/kg, Group B 0.15mmol/kg and Group C

0.20mmol/kg.

The CMR protocol included a MOLLI sequence acquisition in short-axis at mid-ventricular level

before and at 2, 4, 6, 8, 10, 12, 14, 15, 16, 18 and 20 minutes post-contrast administration. In addition,

MOLLI imaging was also performed in early systole (150ms after the R wave) pre-contrast and at 10

and 15 minutes post-contrast administration. Standard long- and short-axis SSFP cine imaging was

performed in order to assess LV mass and volumetric parameters. LGE imaging of the entire heart in

short-axis was performed using single-shot PSIR imaging.31 Blood samples were taken at the time of

CMR in order to measure the hematocrit.

CMR image analysis

For T1 relaxation time measurements, endocardial and epicardial contours were drawn on the MOLLI

images using Osirix Imaging Software (Pixmeo; Switzerland; v4.0). An additional region of interest

rrrrrrrmamamamamamamal)l)l)l)l)l)l)...... . CMCMCMCMCMCMCMR R RRRRR (c(c(c(c(c(c(cininininininine e eeeee

r

he following doses: Group A 0.10mmol/ , Group B 0.15mmol/kg and

rospepepepepectctctctctivivvivvelee y yyy spspsplililiiittt iniinini toooo 33333 aaaaaggegg - ananannand dddd seseses x-x-x-xx mamamamamatctctctcchehehehehed dddd grgrgrgrg ouououououpspsps, ananananand dddd rererererecececeeeivivvivvedededded GGGGG

he following gg dodddd ses: GGGGGrouppp AAAAA 00000 1.1000m00 mol/l/l/l kgkkgk , ,, GrGGG oup p p p BBBBB 0.000 1515151515mmol/kg gg and



ownloaded from


8

(ROI) was drawn in the blood pool, avoiding papillary muscles and trabeculae, and the anterior right

ventricular septal insertion point was marked. ROIs were manually translated on each TIeff time

image to allow motion compensation. The same ROIs were used on corresponding pre- and post-

contrast images. To obtain voxel-wise T1 relaxation times, a 3 parameter fit to the signal intensity, S as

a function of TIeff was performed according to S(TIeff) = A - Be(-TIeff/T1*) and T1 was calculated as

T1=T1*((B/A)-1). Fitting was carried out using MatLab (The MathWorks; USA; vR2009a). Using the

anterior septal insertion point as reference, T1 maps were segmented according to the American Heart

Association/American College of Cardiology model.29 Mean voxel T1 relaxation time in each segment

before and after contrast was then used to calculate segmental myocardial ECV according the

following formula:

Extracellular volume fraction (ECV) = x (1- hematocrit).

Where the partition coefficient, = R1(myocardium) / R1(blood). R1 is proportional to contrast

agent concentration. R1 = R1(post-contrast) – R1(pre-contrast).

Separate ECV calculations were performed using the 10- and 15-minute post-contrast T1 values

acquired in patients awaiting transplantation, and using the 2-20 minute post-contrast T1 values

acquired in healthy subjects.

Pre-contrast and 15-minute post-contrast MOLLI image analysis was independently repeated in 50%

(3 patients, 48 segments) of patients who had CMR prior to undergoing transplantation, and 33% of

the healthy subjects (10 patients, 60 segments) by a second observer in order to assess interobserver

variability of ECV measurement. Patients were randomly selected.

LV mass, end-diastolic volume, end-systolic volume and ejection fraction were quantified from SSFP

images using CMRtools (Cardiovascular Imaging Solutions, UK). LGE images were reported visually

( ) ( )

n l

n

( ) ( )

n coooooefefefefeffifififf cicicicicienenent,tttt ===== R1(m(m(m(m(myyoyyy caaaaardrdrdrdrdiuiuiuiuium)m)m)m)m) /// RRRRR11111(b(b(b(b(blolololooododododod).)))) RRRRR1 iiiiissss prprprprpropopopoo ororororortitititit onononoo aaalaa

n. R11 = R11(p(posttt-co tntrast) )) – RRRRR11(p((p( re-contrast)t)t)t .



ownloaded from


9

by 2 experienced operators (C.A.M and M.S.). Segments were recorded as containing no LGE, infarct-

typical LGE or infarct atypical LGE.

Statistical analysis

All data was analyzed in a blinded fashion, with independent analysis of CMR (C.A.M. and M.S.) and

histology (P.B.) data. Statistical analysis was performed using SPSS (IBM, USA; v19). Continuous

variables are expressed as mean ± SD. For the histological validation, regression analysis using

generalized estimating equations (GEE) in order to adjust for repeated measurements within each

subject was used to assess the relationship between DynEq-CMR-derived ECV and histological CVF.

Within subject and between subject correlations were calculated using the methods described by Bland

et al.32, 33 The same analyses were used to assess the relationship between ‘isolated’ post-contrast T1

measurements and histological CVF. As pre-specified, analyses were repeated after excluding

segments containing infarct-typical LGE and after excluding segments containing any LGE (infarct-

typical or -atypical). Characteristics of healthy volunteers were compared across the three groups

using one-way analysis of variance, as were pre-contrast T1 measurements. The effect of contrast

agent dose on repeated post-contrast T1 measurements, and the effect of contrast agent dose,

myocardial region and sex on repeated ECV measurements, were assessed using a repeated measures

analysis of variance model. In order to assess the change in ECV measurement over time, ECV values

calculated at 2- and 20-minutes post-contrast were compared within each contrast agent dose group

using a paired t-test. Diastolic and systolic ECV measurements were compared within each contrast

dose group using a paired t-test. Interobserver agreement was evaluated using the repeatability

coefficient,34 which calculates the range within which measurements by two different observers are

expected to lie for 95% of subjects, and intraclass correlation coefficient (ICC).

Results

Histological validation

Characteristics of each of the six patients who received a heart transplant following CMR are

summarized in Table 1. Median time between CMR and transplantation was 29 days (in 5 patients the

eeeeeeeenenenenenenen ‘‘‘‘isisisisisisisolololololololatatatatatatatededededededed’’’’’’ popopopopopooststststststst

repeaaaaaaateteteteteteteddddddd afafafafafafafteteteteteteterrrrrr exexexexexexexcclccg p p , y p

ng infarct-typical LGE and after excluding segments containing any L

l e

f

g p p , y p

nggg g iinii farct-tytyyyypipipipp caaaaalllll LGLGLGLGLGE EE E E annnnnd dddd afaffafa teteter rrrr eeexcluuudinininining sesesesesegmgmgmenennnntstststst cononononontattatataininininininininining ggg anananananyyyyy L

l). CCCCChahahahaharararararactctctctcterererrisisistititicscscscc ooof hehehehehealalalalalthththtthy vovovovovolululululuntntnnnteeeeeeeeeersrsrsrsr wwwwwererererereee ee cococococompmpmpmpm ararararededdded aaaaacrcrcrcrcrosoooo s ss thththththe eee ththththt rerrrr

lysyy is of variance, ,, as were prpp e-co tntra tstt TTTT11 measurememememmentstsss. TThT e effect of



ownloaded from


10

interval was 40 days or less, but in 1 patient the interval was 276 days). Mean histological CVF was

21.6±12.4%; range 3.3% to 55.2% (Figure 2). Mean DynEq-CMR-derived ECV calculated using the

10-minute post-contrast T1 values was 43.8±7.0%; range 31.1% to 65.1% (median within subject

range 20.2%), and using the 15-minute post-contrast T1 values was 43.9±6.7%; range 30.9% to 68.4%

(median within-subject range 18.2%).

There was a significant linear relationship between ECV measured by DynEq-CMR, using both the

10- or 15-minute post-contrast T1 values, and histological CVF (p<0.001 for both using GEE; Figure

3). Whilst high for both, the correlation was marginally higher when ECV was calculated using the

15-minute post-contrast T1 values (within-subject r=0.745, p<0.001; r2=0.555; between-subject

r=0.945, p<0.01, r2=0.893) compared to when the 10-minute post contrast values were used (within-

subject r=0.728, p<0.001; r2=0.530; between-subject r=0.880, p<0.01, r2=0.774). The linear

regression equation for the relationship between DynEq-CMR-derived ECV using the 15-minute post-

contrast T1 values and histological CVF was: histological CVF = 1.45 x ECV – 42. The correlation

between mean ECV and CVF on a per individual basis was r=0.945, p=0.004; r2=0.893 (Figure 4).

There was a significant linear relationship between isolated post-contrast T1 measurements made at

10- and 15-minutes post-contrast and histological CVF (p<0.001 for both using GEE; Figure 3),

however this was largely driven by the within-subject correlations (isolated 10 minute post-contrast

T1: r=-0.690, p<0.001; r2=0.475; isolated 15 minute post-contrast T1: r=-0.741, p<0.001; r2=0.550).

The between-subject correlations were not significant (isolated 10 minute post-contrast T1: r=-0.028,

p=0.96; isolated 15 minute post-contrast T1: r=-0.207, p=0.69). There was no correlation between pre-

contrast T1 values and histological CVF (p=0.437 using GEE; within-subject r=0.138, p=0.192;

between-subject r=0.199, p=0.71).

In light of these findings, all subsequent analyses were performed using ECV calculated with DynEq-

CMR using the 15-minute post-contrast T1 values.

tttttttrararararararaststststststst vvvvvvvalalalalalalalueueueueueueues s sssss wewewewewewewererererererere uuuuuuu

r2=0000000 777777774747474747474))))))) TTTTTTThehehehehehehe lllllin; ; j , p , )

n for the relationship between DynEq-CMR-derived ECV using the 15

and histological CVF was: histological CVF = 1.45 x ECV – 42. The

V (

; ; j , p , )

n n nnn foofoor the relalalalalatititit onononnnshshshshshipipipipip bbbbbetetetetetwwewww enenene DDDDDyyynEqEqq-CCCCCMRMRMRMRMR-d-d-d-d-dereriviivededededed ECVCVCVCVCV uuuuusisisisisingngngng ttttthehehehehe 15

anddnddd hhhhhisisisisistooototololololl gigigig cacacal lll CVCVCVCC F FFFF wawawawawass:ss hhhhhisisisisistototototolololoologigigigigicacacaaal lll CVCVCVCVCVFF FFF ===== 111.11 4545454545 xxx EEEEECVCVCVCVCV – 444442.2222 TTTTTheheheee

V and CVF on a pepp r iinii didididii iviiidduddd lallll bbbbasiisii was r=0000.9994545454545, ,, p=p=p=pp 0.000 00000000004;; r2=0.893 ((



ownloaded from


11

There was a significant linear relationship between ECV and histological CVF in septal and non-septal

segments (p<0.001 using GEE for both), with minimal difference in within-subject and between

subject correlations (septum: within-subject r=0.750, p<0.001; r2=0.560; between-subject r=0.940,

p<0.01, r2=0.884; non-septum: within-subject r=0.722, p<0.001; r2=0.521; between-subject r=0.941,

p<0.01, r2=0.885). The significant linear relationship between ECV and histological CVF was also

maintained across ventricular levels (p<0.001 using GEE for both basal and mid ventricular segments,

and apical segments), with minimal difference in within-subject and between subject correlations

(basal and mid ventricle: within-subject r=0.727, p<0.001; r2=0.529; between-subject r=0.960,

p<0.01, r2=0.922; apical ventricle: within-subject r=0.748, p<0.001; r2=0.559; between-subject

r=0.892, p<0.01, r2=0.796).

Infarct-typical LGE was present in 32 segments and infarct-atypical LGE was present in a further 10

segments. Only 1 segment displayed LGE throughout its entirety. When segments containing infarct-

typical LGE were excluded (analysis performed on 64 segments), the linear relationship between ECV

and histological CVF was maintained (p<0.001 using GEE; within-subject r=0.682, p<0.001;

r2=0.465; between-subject r=0.912, p<0.01, r2=0.832; Figure 5). Likewise, the linear relationship

between ECV and histological CVF remained when segments containing any LGE (infarct-typical and

infarct-atypical patterns) were excluded (analysis performed on 54 segments) (p<0.001 using GEE;

within-subject r=0.652, p<0.001; r2=0.426; between-subject r=0.843, p<0.02, r2=0.711; Figure 5).

Effect of contrast agent dose, myocardial regionality, cardiac cycle and gender on ECV

There were no significant differences in subject characteristics between contrast agent-dose groups

(Table 2).

Pre-contrast myocardial (group A 1051±49ms; group B 1045±49ms; group C 1040±43ms; p=0.87)

and blood (A 1678±98ms; B 1645±118ms; C 1686±101ms; p=0.66) T1 relaxation times were not

significantly different between groups. Mean myocardial (A 542±65ms; B 465±69ms; C 407±55ms;

p<0.001) and blood (A 407±73ms; B 307±67ms; C 252±48ms; p<0.001) T1 relaxation times averaged

GE wwwwwwwasasasasasaas ppppppprererererereresesesesesesesentntntntntntnt iiiiinp g yp p

s a

e

V 0

2

p g yp p

segegegge ment ddddisisisssplplplplplayayyyyededededed LLLLLGEGEGEGEGE ttttthrhrhrhrh ouououo ghghghghhoout tiits enenenenentititititirererereretyy. WhWhWhWhWheeen sssssegegegeegmemememementnttntnts cococococontnntntn a

exccccclululululudededded d dddd (a(a(a(( nanananalylylyysisisisis s ssss peeeerfrfrfrfrfororororormemememeedd ddd ononononon 6666644444 seseseeegmgmgmgmgmenenenenentststststs), tttthehehehehe lllinininii eaaaaarrr rr rereererelllall titiitiionononononshshshhs ipipipip

VF was maint iiaiiined ddd (p(p((( 0<0<0<00 0.000010101011 u iisiinggg GGGGGEEEEEEEEE;;; wiiii hthththiniii -ssssubububububjejej ctctcttt rrrrr=0.682, ,, p<pp 0

222



ownloaded from


12

over all time points post-contrast shortened significantly as contrast dose increased (Figure 6).

Measurements of mean ECV averaged over all time points were significantly higher in group A

compared to group B and group C (A 27.7±3.7%; B 25.8±3.4%; C 25.8±2.8%; p<0.001), but the

difference between groups B and C was not significant (Figure 6). Mean measured ECV increased

linearly over time in each group; between the 2 and the 20 minute post-contrast acquisitions measured

ECV increased from 27.2±2.7% to 28.8±3.4%, p=0.020 in group A; 25.3±2.8% to 26.5±3.2%,

p=0.004 in group B; and 25.2±1.7% to 26.2±2.1%, p=0.068 in group C.

ECV varied significantly between myocardial regions. In each group ECV was highest in the septum

and lowest in the lateral wall, as exemplified by group C (results for other groups were similar):

anterior 24.9±2.5%; septal 27.0±2.5%; inferior 25.2±2.1%; lateral 24.2±2.2%; p<0.001. ECV did not

differ significantly between diastole and systole in any group (A diastolic ECV 28.5±3.2%, systolic

ECV 28.1±3.4%, p=0.26; B diastolic ECV 26.0±3.4%, systolic ECV 25.9±3.3%, p=0.74; C diastolic

ECV 25.7±2.0%, systolic ECV 25.5±2.6%, p=0.56).

Mean ECV was significantly higher in females than in males in each group (A female 29.6±3.0%,

male 25.4±3.0%, p<0.001; B female 27.4±2.7%, male 23.6±2.9%, p<0.001; C female 26.1±2.8%,

male 24.7±2.5%, p=0.027).

Comparison of ECV in health and disease

In healthy subjects receiving 0.20mmol/kg contrast agent (i.e. group C), mean segmental ECV

calculated using 15-minute post-contrast T1 values was 25.5±2.6%. In patients prior to transplantation

(who received the same type and dose of contrast agent), mean ECV, also calculated using 15-minute

post-contrast T1 values, in segments without LGE was 41.4±5.0% (p<0.001 when compared with

healthy subjects), and in segments with LGE was 47.0±7.4% (p<0.001 when compared with healthy

subjects and when compared with segments without LGE; Figure 7). In patients prior to

transplantation, there was a significant difference in mean ECV between segments without LGE,

segments with infarct-atypical LGE (45.8±4.7%) and segments with infarct-typical LGE (47.4±8.1%;

2222222±2±2±2±2±2±2±2.2.2.2.2.22.2%;%;%;%;%;%;%; ppppppp<0<0<0<0<0<0<0.0.0.0.000001010101010101..

olic ECECECECECECECVVVVVVV 28282828282828 5555555±3±3±3±3±3±3±3 2y y g p (

p=0.26; B diastolic ECV 26.0±3.4%, systolic ECV 25.9±3.3%, p=0. 4

y y ggg p (

p=0=0=0=0.26; B ddddiaiaiaiaiassstolololololicicicicic EEEEECVCVCVCVCV 222226.66.6.0±0±0±3.333.3 4%444 ,, sssystststststolololololicicicicic EEEEECVCVCVCVV 22222555.9±9±9±9±9±3.3333 3%3%3%3%3%, p=p=p=p=p 0.0.0.0.0.747777

ystototototolililililiccccc ECECECECECV VVVV 25252522 .5555±2±±±± .66666%%%,%% p=0=0=00=0.555556)6)6)6)6).



ownloaded from


13

p<0.001). On post-hoc analysis the difference in ECV between segments with infarct-typical LGE and

segments without LGE was significant (p<0.001) but the difference between segments with infarct-

atypical LGE and segments without LGE was not (p=0.125; Figure 7).

Mean pre-contrast myocardial T1 relaxation time was significantly higher in patients prior to

transplantation than in healthy subjects (1187±163ms versus 1045±46ms, p<0.001).

Repeatability

Interobserver variability for DynEq-CMR-measurement of ECV in healthy subjects was 2.3% (ICC

0.932), ranging from 1.4% (ICC 0.969) for septal segments, to 2.7% (ICC 0.914) for non-septal

segments. Interobserver variability for DynEq-CMR-measurement of ECV in patients prior to

transplantation was 5.1% (ICC 0.883), ranging from 2.9% (ICC 0.944) for septal segments, to 5.9%

(ICC 0.850) for non-septal segments, and ranging from 4.5% (ICC 0.907) for basal and mid

ventricular segments, to 6.7% (ICC 0.798) for apical segments. For histological CVF assessment,

interobserver variability was 2.8% (ICC 0.994), however, within the same tissue sample, CVF was not

uniform between high-powered fields, with a mean standard deviation of 49% normalized to CVF%.

Mean CVF measurements did not vary significantly according to histological magnification (x200

14.0±11.2%; x50 14.4±12.5%; x100 14.0±11.7%; p=0.507).

Discussion

This is the first human study to provide comprehensive histological validation of the DynEq-CMR

technique for quantification of myocardial ECV, and the first to provide LV histological validation of

the isolated post-contrast T1 measurement technique. Indeed, this is the largest histological validation,

in terms of number of tissue samples, of any CMR ECV-quantification method, and the only to

provide whole heart tissue corroboration.

The DynEq-CMR technique relies on the assumption of a two-compartment model, whereby a steady

state is assumed to exist between the intravascular and interstitial compartments, with equal contrast

EEEEEEECVCVCVCVCVCVCV iiiiiiin n nnnnn papapapapapapatititititititienenenenenenentststststststs ppppppprrrrrrr

) forr seseseseseseseptptptptptptptalalalalalalal sssssssegegegegegeggmmmmemmem n( ), g g ( ) p g

n-septal segments, and ranging from 4.5% (ICC 0.907) for basal and m

nts, to 6.7% (ICC 0.798) for apical segments. For histological CVF ass

a

( ), g g ( ))) p g

nn-nnn ssess ptal segggggmemememm ntntntntnts,s,s,s,s aaaaandndndndnd rrrrrananananangigigig nnng fffrrromm 4.5%5%5%5%5% (((((ICICICICICC C C C 0.00 90909099 7))))) ffffforrrrr bbbbbasasasasalalalalal aaaaandndndndnd m

nts, tototototo 66666.77777% %%%% (I(I(I(I(ICCCCCCCCCC 00000.79898989898) )))) fofofofofor apapapapapicicicicicalalalalal sssssegegegegegmemememementntntntntsss.ss FFFFFororororor hhhhhisisisiistottottolololooogigigigg cacacacacal CVCVCVCVCVF FFFF asasasaa s

ability yy was 2.8%8%8%8%8% (((((ICICICCICCC CCC 00.00 9999999994)4)4)4)4), ,, hhhohh wever,,, witthihihih n thththththe eee sassss meee tttttiiisii sue samppple



ownloaded from


14

agent concentrations in each, due to rapid exchange of contrast agent between the compartments. The

small increase in ECV over time seen here (Figure 6), which is in keeping with the findings of Kawel

et al19 and Shelbert et al,14 suggests that the two-compartment model may be limited and an

incomplete dynamic equilibrium between blood and myocardium is achieved. The lack of equilibrium

may be due to penetration of gadolinium into other ‘compartments’ such as bone and synovial fluid,

and due to faster renal clearance than exchange rate between compartments. Indeed the latter reason in

particular may explain the significantly higher ECV values, and greatest increase in ECV over time,

seen in the healthy subjects receiving the lowest contrast dose.

As a result of the incomplete dynamic equilibrium, the correlation between DynEq-CMR-derived

ECV and histological CVF changed over time. Nevertheless, the relationship between DynEq-CMR-

derived ECV and histological CVF in the current study is comparable to that found in the study by

Flett et al13 using the EQ-CMR technique (r2=0.80), in which basal septal ECV was compared with

histological CVF of tissue obtained from the basal septum at surgical biopsy in patients undergoing

valve replacement for aortic stenosis (18 patients) or myectomy for hypertrophic cardiomyopathy (8

patients). The DynEq-CMR technique used here however offers advantages over the EQ-CMR

technique in terms of being substantially simpler to perform and less time consuming, and could easily

be incorporated into routine scanning protocols.

There was a strong linear relationship between CMR-derived ECV and histological CVF across the

whole spectrum of ECV and CVF, which is an important finding as it means that ECV is suitable for

stratifying patients based on CVF. Nevertheless, the intercept of the linear regression equation for the

relationship between ECV and CVF was not zero and the slope was different from one (unity),

findings which are in keeping with those of Flett et al13 and Messroghli et al.25 As discussed by Di

Carli et al,35 the y-intercept represents a surrogate for all of the multiple components of the myocardial

interstitium, which as well as collagen include non-collagenous proteins, fibroblasts, endothelial cells

and vessels. In addition, whilst we excluded patients with specific myocardial inflammatory

conditions per se, all subjects had end-stage heart failure and as such myocardial inflammation, and

iiiiionononononononshshshshshshshipipipipipipip bbbbbbbetetetetetetetweweweweweweweenenenenenenen DDDDDDD

to thhhhhhhatatatatatatat fffffffouououououououndndndndndndnd iiiiiiinnnnnnn thttttg y p

h m

of tissue obtained from the basal septum at surgical biopsy in patients u

m

g y p

heheheheh EEEQ-CMMMR RRR R tetetechchchchchnininininiquququququeee ee (r(r(r(r(r22222=0=0=0=0.8.. 0)00)0)0 ,, in wwwhihihihihichchchchch bbbbbassallll sseepepee ttallll ECECECECECV V V V V wawawawaasssss cococococomm

of titititiissssssssssueueuueu ooooobtbtbtbb aiaiaia nenened dddd frffrfrf ommmmm ttttthehehehehe bbbasasasasa alalalalal sssssepepepepeptututututum mmmm atatatatat sssssurururururgigigigigicacacacacallll bibibibb opopopppsysysysysy iiiiinn nnn papapapapatitititit enenennentststsss u

for aortic steno iisiiis (1(1(11188 8 papp titiiienttts))))) or myyyectomyyy ffffor hhhhhypypypyy ertrtrtrtrtropopopoo hic cardiom



ownloaded from


15

hence edematous expansion of the interstitial space without collagen deposition, may have been

present, which may be one explanation for the higher mean pre-contrast T1 seen in pre-transplant

patients compared to healthy volunteers. As such it is logical to assume that CMR-derived ECV does

not only reflect CVF, however it appears that excess ECV beyond the baseline is explained by

increases in CVF, provided other causes of ECV expansion are excluded. The y-intercept and slope of

the equation may also have been influenced by changes in the extracellular compartment relating to

the tissue processing itself, and by the CVF quantification technique.

Flett et al13 excluded patients who demonstrated any LGE in the region of the biopsy. Wong et al,9

who demonstrated an association between DynEq-CMR-derived ECV and short-term all-cause

mortality (but without histological corroboration), excluded regions of myocardium in the vicinity of

infarct-typical LGE but included myocardium displaying infarct-atypical LGE. Others have quantified

ECV in myocardium exhibiting infarct-typical LGE (also without histological corroboration).16, 36

Expansion of the myocardial interstitial space appears to occur as a continuous spectrum. As such we

included all tissue samples in our analysis, regardless of LGE ‘status’, and simply quantified ECV in

each. Nevertheless, our study demonstrated that the correlation between DynEq-CMR-derived ECV

and histological CVF remained strong when segments containing any LGE were excluded and when

segments containing infarct-typical LGE only were excluded, as well as when segments containing

LGE were included.

Indeed, this study serves to highlight the potential shortcomings of the LGE technique relating to its

relative signal intensity nature. Whilst the LGE technique is very well established for infarction

detection and quantification, the ability of LGE to detect and quantify myocardial fibrosis in non-

ischemic cardiomyopathies is much less well validated. Myocardial fibrosis exists as a spectrum from

diffuse to focal. Such homogeneity, or lack of heterogeneity, is problematic for the LGE technique

since it relies on “normal” regions of myocardium to serve as reference for nulling. As a result, LGE

quantification (and hence “quantification of fibrosis” via the LGE technique) in non-ischemic

cardiomyopathies is dependent on the signal intensity threshold used.37 As demonstrated in Figure 5,

f f fffff mymymymymymymyocococococococararararararardididididididiumumumumumumum iiiiiiinn nn nn n ththththththth

cal LGLGLGLGLGLGLGEEEEEEE OOOOOOOthththththththerererererererssssss hy p y g yp

um exhibiting infarct-typical LGE (also without histological corrobora

myocardial interstitial space appears to occur as a continuous spectr m

samples in our ana sis, regardless of LGE ‘status and sim y quan i

y p y g yp

ummmm exhibitiiingngngngng infnfnfnfnfarararararctctctctct-t-t-t-t-typypyypy icicicicicalalalall LLLLGEGEGEGEG (aalssso wwwwwititititthohohohohouut hhhisisissstotototot logigigigigicacacaaal lll l cococococ rrrrrrrrroboboboboborororoora

myooooocacacacacardrdrdrr iaiaiaiaiallll inininntetetett rsrsrssstititiiitititit alllll ssssspaapapaaccecc aaaaappppppppppeaeaeaeaearsrsrsrsr tttooo oo ocococococcucucucucurrrr r asasasasas aaaa ccconononoontitt nununununuouououououss sss spspspppecececece trtrtrrrumumumumm

samppples in our analyly isiis,,, regggarddldldd ess offfff LLLLGEGEGEG ‘‘‘stattussss’,’,’’’ andndnddnd sssssimplp y yy quqq anti



ownloaded from


16

in the current study there was considerable overlap in histological CVF between segments that

displayed LGE and those that did not (although it should be recognized that CVF and ECV values in

the current study represent means for entire segments, with only 1 segment displaying LGE

throughout). Furthermore the difference in mean ECV between segments without LGE and those with

atypical LGE was not significant, although overall mean differences in ECV may not be reflective of

differences in individual patients. Nevertheless, in keeping with other work this study suggests that

myocardial fibrosis in non-ischemic cardiomyopathies may be better assessed using ECV techniques

rather than LGE.38

The correlation between DynEq-CMR-derived ECV and histological CVF was maintained throughout

the LV, although there was greater ECV measurement variability in non-septal compared to septal

regions and in apical compared to basal and mid-ventricular myocardium, likely reflecting the

comparatively thinner ventricular walls in these regions. The degree of variability in the mid-

ventricular slice was comparable to that reported in other studies;14, 16 variability at other ventricular

levels has not previously been reported.

Mean DynEq-CMR-derived ECV in healthy subjects was in keeping with that found in other studies.9,

14-16, 18, 19, 39 Interestingly, ECV was significantly higher in the septum compared to other myocardial

regions in healthy subjects. This is in keeping with histological data from healthy myocardium, which

demonstrates higher collagen content in the septum compared to other regions,40 and with DynEq-

CMR data reported by Kawel et al.19 This finding, which may be secondary to the septum being

exposed to mechanical strain from both ventricles, has important implications both in terms of CMR

ECV measurement and histological assessment of CVF, as septal sampling appears not necessarily

representative of other myocardial regions.

In addition, ECV was seen to vary significantly between genders. This is in keeping with the findings

of Sado et al39 who, using EQ-CMR in healthy subjects with a similar age range to here, found gender

was an independent predictor of ECV in multivariate analysis (myocardial mass, hematocrit and

ooooooon-n-n-n-n-n-n-seseseseseseseptptptptptptptalalalalalala cccccccomomomomomomompapapapapapaparererererereredddddd

um lililililililikekekekekekekelylylylylylyly rrrrrrrefefefefefefeflelelelelelelectctctctctinp y , y

nner ventricular walls in these regions. The degree of variability in the

a r

i

p y , y

nnenenenn rr rr ventriiiicuuuuulalalall r wawawawaalllllllllls s s s ininininin tttttheheheheh sseses regege ioiii nns. ThThThThThe dedddd gggreeeee ooooof ff ff vaaaaariririririabababbbilililililitititity yyy ininininin ttttthehhhh

as cococococompmpmpmpmparaaraarababbabblelele ttto oooo thtththt attttt rrrrreppepeppooroo teteeeedd ddd ininininin oooooththththherererere ssssstututututudididididiesesesees;;;14,4,4,44 161111 vvvaraa iaiaiaiaiabibibibb lililililityttytyt aaaatt ttt ototottotheheheeer

iously yy been reppported.dddd



ownloaded from


17

patient height were not); indeed, the absolute difference in ECV between genders reported by Sado et

al is very similar to that found here. The reasons for the sex difference are not clear. However given

that the correlation between ECV and histological CVF was maintained throughout the heart, with low

observer variability particularly in the mid-ventricular slice, this finding appears genuine i.e. it does

not appear to be related to potentially greater partial voluming effects in theoretically thinner female

LV walls. As Sado et al39 discuss, it may be that this gender difference in ECV contributes to known

differences in cardiovascular disease expression between males and females. Further investigation of

this finding is required.

Mean DynEq-CMR-derived ECV in segments containing LGE was similar to that reported by

Ugander et al,16 who assessed ECV in patients with cardiac disease exhibiting infarct-typical or -

atypical LGE undergoing clinical CMR. However, mean ECV in myocardium remote from segments

containing LGE in the pre-transplant patients in the current study was substantially higher than ECV

in myocardium remote from segments containing LGE in the study by Ugander et al.16 This is likely to

reflect the end-stage nature of cardiac disease in the cohort studied here. As a result, whilst there was

considerable overlap between ECV in healthy subjects and ECV in myocardium in patients with

cardiac disease but remote from LGE in the study by Ugander et al (and in other studies),9, 39 there was

very little overlap in the current study.

Iles et al found a significant correlation (r=-0.7, p=0.03) between LV mid-ventricular isolated 15-

minute post-contrast T1 values and histological CVF in 9 heart transplant recipients, although tissue

was of right ventricular origin and of small volume (obtained via transvenous endomyocardial biopsy),

and this technique has been increasingly applied as a surrogate for ECV.24 However isolated post-

contrast T1 measurements are confounded by a number of factors such as renal function, hematocrit,

body fat and myocardial steatosis. In keeping with this, whilst the within-subject correlation between

isolated post-contrast T1 measurements and histological CVF was high in the current study, there was

no significant correlation between subjects.

hihihihihihihibibibibibibibitititititititingngngngngngng iiiiiiinnnnnnnfafafafafafafarcrcrcrcrcrcrct-t-t-t-t-t-t-tytytytytytytyppppppp

cardiiiiiiiumumumumumumum rrrrrrremememememememotototototototeeeeeee ffrfff og g , y

m T

g i

g g , y

ththththt eeee pre-trannannnsppplaalaalantntntntnt pppppatatatatatieieieieientntntntnts ssss ininiini thhhhhe cuurrrrennnnntt ttt ststststtuduuudu yy wawawaaasss ss sssubsbsbsbsbstattatatantntntntntiaiaiaiai lllllllllly y yyy hihihihihighgggg

motetetetete fffffrororororom mmmm seseseegmgmgmgmenenenenentstt ccconononoontatatatataininininininggggg LGLGLGLGLGE EEEE inininini ttttthehehehehe ssssstututututudydydydydy bbbbby yyyy UgUgUgUgUgananananandededededer eteteete aaaaal.ll 16161666 TTT

ggge nature of ca ddrdddiiiaii c didididd sease iiniii thththht e cohhohhh rt studididid eddddd hhhhhererererere. AAAAAs ssss a result,,, whi



ownloaded from


18

Limitations

Whilst the number of tissue samples included in this study is large, the total number of patients is

relatively small. However this reflects the difficulty in obtaining whole-heart histological data from

patients who have recently undergone CMR, and who do not have an acute inflammatory myocardial

condition. A greater number of patients could potentially have been recruited by performing the study

using post-mortem tissue obtained from patients who had undergone CMR for other reasons and

subsequently died. However immediate formalin fixation would not have been possible, which may

have introduced error in histological CVF quantification, as may cause of death. In the current study,

hearts remained in vivo until explantation when they were immediately fixed in formalin, minimizing

‘post-mortem’ changes. The interval between CMR and transplantation could have resulted in changes

in myocardial collagen content. However in 5 of the 6 patients, the interval was minimal (40 days or

less). There was a small improvement in correlation between DynEq-CMR-derived ECV and

histological CVF when ECV was calculated using the 15-minute post-contrast T1 values compared to

when the 10-minute post contrast values were used. As such, the optimal post-contrast time for T1

acquisition was not ascertained and it is possible that the correlation may have improved further with

later post-contrast T1 measurements. However, the clinical status of the patients involved, i.e. end-

stage heart failure, necessitated relatively short scan duration. Finally, an alternative CMR method of

quantifying ECV that has been less commonly applied than the methods assessed here, in which

contrast agent kinetics are deconvoluted using mathematical modeling, was not investigated.41

Conclusions

This study provides comprehensive validation of the DynEq-CMR method for measurement of

myocardial ECV. Isolated post-contrast measurement of myocardial T1 is insufficient for ECV

assessment.

Acknowledgements

The authors would like to thank Siemens for providing access to the MOLLI Sequence (WIP

CV_MOLLI_388).

eeeeeeervrvrvrvrvrvrvalalalalalalal wwwwwwwasasasasasasas mmmmmmminininininininimimimimimimimalalalalalalal

CMRRRRRRR dddddddererererererrivivivivivivivededededededed EEEEEEECCCCVCCCp y q

w

t i

t d

p y q

wwwwwhehehehen ECV VV wawawaww s cacacacaalclclclclculululululatatatatateed d dd d usussusinii g ththhthe 15-mmmmmininininnuutuuu e ee popop ststststst-c-c-connnnntrtrtrtrtrassssst t t tt TTTTT111 vavavavav lululululueseeee

te pppppososososost tt cococooontntntrarararaststst vvvvvalalalala ueuu s ssss wewewewewerer uuuuusssededededed. AsAsAsAsAs sssssucucucucuchh,hhh ttttthehehehehe oooooptptptptptimimimimmalaa ppppposososoo t-t-t-t-t-cocococontntntntntrararararastststs ttti

t ascertained anddd dd ititit iiiiis popp ssibibibbibllelll thahhh tt thehhh correllal titiiionnnn mmmmmayyy hhhhhavaaaa e imprpp ovedt



ownloaded from


19

Sources of Funding

Christopher Miller is supported by a Fellowship from the National Institute for Health Research, UK

(NIHR-DRF-2010-03-98). Christopher Miller, Simon Williams, Nizar Yonan, and Matthias Schmitt

have received research funding from New Start Transplant Charity.

Disclosures

None.

References

1. Beltrami CA, Finato N, Rocco M, Feruglio GA, Puricelli C, Cigola E, Quaini F, Sonnenblick EH, Olivetti G, Anversa P. Structural basis of end-stage failure in ischemic cardiomyopathy in humans. Circulation. 1994;89:151-163. 2. Beltrami CA, Finato N, Rocco M, Feruglio GA, Puricelli C, Cigola E, Sonnenblick EH, Olivetti G, Anversa P. The cellular basis of dilated cardiomyopathy in humans. J Mol Cell Cardiol. 1995;27:291-305. 3. Weber KT, Brilla CG. Pathological hypertrophy and cardiac interstitium. Fibrosis and renin-angiotensin-aldosterone system. Circulation. 1991;83:1849-1865. 4. Swynghedauw B. Molecular mechanisms of myocardial remodeling. Physiol Rev. 1999;79:215-262. 5. Querejeta R, Lopez B, Gonzalez A, Sanchez E, Larman M, Martinez Ubago JL, Diez J. Increased collagen type i synthesis in patients with heart failure of hypertensive origin: Relation to myocardial fibrosis. Circulation. 2004;110:1263-1268. 6. Kawara T, Derksen R, de Groot JR, Coronel R, Tasseron S, Linnenbank AC, Hauer RN, Kirkels H, Janse MJ, de Bakker JM. Activation delay after premature stimulation in chronically diseased human myocardium relates to the architecture of interstitial fibrosis. Circulation. 2001;104:3069-3075. 7. McLenachan JM, Dargie HJ. Ventricular arrhythmias in hypertensive left ventricular hypertrophy. Relationship to coronary artery disease, left ventricular dysfunction, and myocardial fibrosis. Am J Hypertens. 1990;3:735-740. 8. Moreo A, Ambrosio G, De Chiara B, Pu M, Tran T, Mauri F, Raman SV. Influence of myocardial fibrosis on left ventricular diastolic function: Noninvasive assessment by cardiac magnetic resonance and echo. Circ Cardiovasc Imaging. 2009;2:437-443. 9. Wong TC, Piehler K, Meier CG, Testa SM, Klock AM, Aneizi AA, Shakesprere J, Kellman P, Shroff SG, Schwartzman DS, Mulukutla SR, Simon MA, Schelbert EB. Association between extracellular matrix expansion quantified by cardiovascular magnetic resonance and short term mortality. Circulation. 2012;126:1206-1216.. 10. Brilla CG, Funck RC, Rupp H. Lisinopril-mediated regression of myocardial fibrosis in patients with hypertensive heart disease. Circulation. 2000;102:1388-1393. 11. Diez J, Querejeta R, Lopez B, Gonzalez A, Larman M, Martinez Ubago JL. Losartan-dependent regression of myocardial fibrosis is associated with reduction of left ventricular chamber stiffness in hypertensive patients. Circulation. 2002;105:2512-2517. 12. Izawa H, Murohara T, Nagata K, Isobe S, Asano H, Amano T, Ichihara S, Kato T, Ohshima S, Murase Y, Iino S, Obata K, Noda A, Okumura K, Yokota M. Mineralocorticoid receptor antagonism

CCCCCCCigigigigigigigololololololola a aaaaa E,E,E,E,E,E,E, QQQQQQQuauauauauauauaininininininini i ii i i i F,F,F,F,F,F,F,hemiiiiiiicccc ccc cacacacacacaardrdrdrdrdrdrdioioioioioioiomymymymymymymyopopopopopopopaa

;C la e

, Brilla CG. Pathological hypertrophy and cardiac interstitium. Fibrosea e

;CAAAAA, FiFiFiFFinaaaatototototo NNNNN, Rocco M, Feruglio GGGGAA,A,AA Puricelli CCCCC, CiCiCiCiCigola E, Sonnenblaa aaa PPP.P The celllelllulululululaaaaarr rr bababababasisisisisis s sss ofofofofof dddddilililillata ededededd cccardiiommmmmyooooopapapapp thththhy ininininin hhumumumumumannnnns.s.s.s.s. J J J JJ MoMoMoMoMol l l ll CeCC

, BBBBBriririririlllllllllla aaaa CGCGCGCGCG. PaPaPaPP thththhtholololoo oggggicicicicicalalalalal hhypypypypypererererertrtrtrtrtropopopopo hyhyhyhyhy aaaaandndndndnd cccccararararardididididiacacacacac iiiiintntnterere stststststititiiti iuiuiuiuium.mm FFFFFibibibbibrororoooserone system. CiCiCiCiCircrcrcrcrculululululatatatttioiooioionnnnn... 19191919199191919191;8;8;8;8;83:3:333 18188884949494949-1118686868665.5555 auw B. Moleculalll r me hchhhhanisms offf f myyyocardididii lall remmmmododododo elllinininning.g.g.gg Physyy iol Re



ownloaded from


20

ameliorates left ventricular diastolic dysfunction and myocardial fibrosis in mildly symptomatic patients with idiopathic dilated cardiomyopathy: A pilot study. Circulation. 2005;112:2940-2945. 13. Flett AS, Hayward MP, Ashworth MT, Hansen MS, Taylor AM, Elliott PM, McGregor C, Moon JC. Equilibrium contrast cardiovascular magnetic resonance for the measurement of diffuse myocardial fibrosis: Preliminary validation in humans. Circulation. 2010;122:138-144. 14. Schelbert EB, Testa SM, Meier CG, Ceyrolles WJ, Levenson JE, Blair AJ, Kellman P, Jones BL, Ludwig DR, Schwartzman D, Shroff SG, Wong TC. Myocardial extravascular extracellular volume fraction measurement by gadolinium cardiovascular magnetic resonance in humans: Slow infusion versus bolus. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance. 2011;13:16. 15. Broberg CS, Chugh SS, Conklin C, Sahn DJ, Jerosch-Herold M. Quantification of diffuse myocardial fibrosis and its association with myocardial dysfunction in congenital heart disease. Circulation: Cardiovascular Imaging. 2010;3:727-734. 16. Ugander M, Oki AJ, Hsu L-Y, Kellman P, Greiser A, Aletras AH, Sibley CT, Chen MY, Bandettini WP, Arai AE. Extracellular volume imaging by magnetic resonance imaging provides insights into overt and sub-clinical myocardial pathology. European Heart Journal. 2012;33:1268-1278. 17. Lee JJ, Liu S, Nacif MS, Ugander M, Han J, Kawel N, Sibley CT, Kellman P, Arai AE, Bluemke DA. Myocardial T1 and extracellular volume fraction mapping at 3 tesla. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance. 2011;13:75. 18. Kawel N, Nacif M, Zavodni A, Jones J, Liu S, Sibley CT, Bluemke DA. T1 mapping of the myocardium: Intra-individual assessment of post-contrast T1 time evolution and extracellular volume fraction at 3t for gd-dtpa and gd-bopta. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance. 2012;14:26. 19. Kawel N, Nacif M, Zavodni A, Jones J, Liu S, Sibley CT, Bluemke DA. T1 mapping of the myocardium: Intra-individual assessment of the effect of field strength, cardiac cycle and variation by myocardial region. Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance. 2012;14:27. 20. Messroghli DR, Walters K, Plein S, Sparrow P, Friedrich MG, Ridgway JP, Sivananthan MU. Myocardial T1 mapping: Application to patients with acute and chronic myocardial infarction. Magnetic Resonance in Medicine. 2007;58:34-40. 21. Jellis C, Wright J, Kennedy D, Sacre J, Jenkins C, Haluska B, Martin J, Fenwick J, Marwick TH. Association of imaging markers of myocardial fibrosis with metabolic and functional disturbances in early diabetic cardiomyopathy. Circulation: Cardiovascular Imaging. 2011;4:693-702. 22. Turkbey EB, Gai N, Lima JAC, Van der Geest RJ, Wagner KR, Tomaselli GF, Bluemke DA, Nazarian S. Assessment of cardiac involvement in myotonic muscular dystrophy by T1 mapping on magnetic resonance imaging. HRTHM. 2012;9:1691-1697. 23. Gai N, Turkbey EB, Nazarian S, van der Geest RJ, Liu C-Y, Lima JAC, Bluemke DA. T1 mapping of the gadolinium-enhanced myocardium: Adjustment for factors affecting interpatient comparison. Magnetic Resonance in Medicine. 2010;65:1407-1415. 24. Iles L, Pfluger H, Phrommintikul A, Cherayath J, Aksit P, Gupta SN, Kaye DM, Taylor AJ. Evaluation of diffuse myocardial fibrosis in heart failure with cardiac magnetic resonance contrast-enhanced T1 mapping. Journal of the American College of Cardiology. 2008;52:1574-1580. 25. Messroghli DR, Nordmeyer S, Dietrich T, Dirsch O, Kaschina E, Savvatis K, O h-Ici D, Klein C, Berger F, Kuehne T. Assessment of diffuse myocardial fibrosis in rats using small-animal look-locker inversion recovery T1 mapping. Circulation: Cardiovascular Imaging. 2011;4:636-640. 26. Kehr E, Sono M, Chugh SS, Jerosch-Herold M. Gadolinium-enhanced magnetic resonance imaging for detection and quantification of fibrosis in human myocardium in vitro. Int J Cardiovasc Imaging. 2008;24:61-68. 27. Messroghli DR, Greiser A, Fröhlich M, Dietz R, Schulz-Menger J. Optimization and validation of a fully-integrated pulse sequence for modified look-locker inversion-recovery (molli) T1 mapping of the heart. Journal of Magnetic Resonance Imaging. 2007;26:1081-1086.

r Cardiovasculararrrr MM

uemkkkkkkkee eeeee DADADADADADADA. TTTTTTT11 mamamamamamamapppplutionnnnnnn aaaaaaandndndndndndnd eeeeeeextxtxtxtxtxtxtrararararararacececcc lp 1

deN p- d

hai

p 1d-dtdtdtdtdtpppapp andndndndnd ggggdd-ddd bopta. Journal of cardidididd ovascular rrr r magngngngng etic resonance :eeeeetytytyy for Carrdididididiovovovvasasasascucucucuculalalaalarrr rr MaMaMaMaMagngngng ettticicicii Reesonononnnananaaa cecececece. 202020202 12122122;1;1;1;; 4:2626262626. NNNNNacccccif M, Zaavovodnnniii A,,,,, JJJJJones JJ,,, Liiuuu S,, SSSibbbbbleleleleleyyy CTCT, Blueueemkmkkmkke DAAAAA. TT1 mammap-innnndididididivivivivividududuuualalal aaaassssssesesesee smsmsmsms entntntntnt ooooof ffff thhhhheee ee efefefefeffefefefefectctctctct ooooof fff fififififieleleleleld dddd stststststrerereeengngngngngththththth, cacacaaardrdrdrdrdiaiaiaaacc ccc cycycycyyclclclclc e eeee anananaand Journal of cccararaara dididididiovovovvvasasasaa cucucuuculalalalalar rrrr mamamamamagngngngngneteteteticici rrrrresesesessonononananananncecececece :::: ooooofffffffffficicciciciaiaiaiaial l ll l jojj urnal of thagggnetic Resonance. 2020220201212122;1;1;114:4444 2727272727.i DR, WWWWWalalalalaltetetetetersrsrssrs KKKKK, PlPlPlPlPleieieieieinnn S,SSSS SSSSSpapapapaparrrrrrrrrowowowowow P,PPPP FFFFFririedededededriririririchchcchc MMMMMG,GGGG RRRRRidididididgwgwgwgwgwayayayayay JP, Siv



ownloaded from


21

28. Kellman P, Arai AE, McVeigh ER, Aletras AH. Phase-sensitive inversion recovery for detecting myocardial infarction using gadolinium-delayed hyperenhancement. Magn Reson Med. 2002;47:372-383. 29. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: A statement for healthcare professionals from the cardiac imaging committee of the council on clinical cardiology of the american heart association. Circulation. 2002;105:539-542. 30. Rasband WS. Imagej. Us national institutes of health, bethesda, maryland, USA, http://imagej.Nih.Gov/ij/, 1997-2012. 31. Ledesma-Carbayo MJ, Kellman P, Hsu LY, Arai AE, McVeigh ER. Motion corrected free-breathing delayed-enhancement imaging of myocardial infarction using nonrigid registration. J Magn Reson Imaging. 2007;26:184-190. 32. Bland JM, Altman DG. Calculating correlation coefficients with repeated observations: Part 1--correlation within subjects. BMJ. 1995;310:446. 33. Bland JM, Altman DG. Calculating correlation coefficients with repeated observations: Part 2--correlation between subjects. BMJ. 1995;310:633. 34. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307-310. 35. Di Carli MF, Kwong RY, Jerosch-Herold M. Insights into left ventricular remodeling through noninvasive measures of myocardial matrix expansion with cardiovascular magnetic resonance. Circulation. 2012;126:1179-1181. 36. Bauner KU, Biffar A, Theisen D, Greiser A, Zech CJ, Nguyen ET, Reiser MF, Wintersperger BJ. Extracellular volume fractions in chronic myocardial infarction. Investigative radiology. 2012;47:538-545. 37. Flett AS, Hasleton J, Cook C, Hausenloy D, Quarta G, Ariti C, Muthurangu V, Moon JC. Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging. 2011;4:150-156. 38. Kellman P, Wilson JR, Xue H, Bandettini WP, Shanbhag SM, Druey KM, Ugander M, Arai AE. Extracellular volume fraction mapping in the myocardium, part 2: Initial clinical experience. JCardiovasc Magn Reson. 2012;14:64. 39. Sado DM, Flett AS, Banypersad SM, White SK, Maestrini V, Quarta G, Lachmann RH, Murphy E, Mehta A, Hughes DA, McKenna WJ, Taylor AM, Hausenloy DJ, Hawkins PN, Elliott PM, Moon JC. Cardiovascular magnetic resonance measurement of myocardial extracellular volume in health and disease. Heart. 2012;98:1436-1441. 40. Medugorac I. Collagen content in different areas of normal and hypertrophied rat myocardium. Cardiovasc Res. 1980;14:551-554. 41. Jerosch-Herold M, Sheridan DC, Kushner JD, Nauman D, Burgess D, Dutton D, Alharethi R, Li D, Hershberger RE. Cardiac magnetic resonance imaging of myocardial contrast uptake and blood flow in patients affected with idiopathic or familial dilated cardiomyopathy. AJP: Heart and Circulatory Physiology. 2008;295:H1234-H1242.

t ventricular remomooooodcucucucucucuculalalalalalalar r rrrrr mamamamamamamagngngngngngngneteteteteteteticicicicicicic rrrrrrreseseseseseses

n ET RRRRRRReieieieieieieiseseseseseseserrrrrrr MFMFMFMFMFMFMF W, , , , , g y , ,o o

Hasleton J, Cook C, Hausenloy D, Quarta G, Ariti C, Muthurangu V, Mn ue

nv x

, , , , , g y , ,olululululummmemm fffffrararararaccctc ioioioioions in chronic myocarddiddd aaal infarctiooooon.nnnn InInInInInvvev stigative radiolo

HaHaHaHaassslss eton J, , CoCoookkk CCC, HaHHHH usenee lolooy D,D QQuuuartrtrtrtrta aa aa G,GG,GG AAriiiti CC,,, MuMuuMuuthuuru anana ggu VV, Mniququququueseseseses ffforororoor ttthehehehhe qqquauauauauantnnn iffffficicicicicatatatatatioiioii n nnnn ofofofofof mmmmmyyyyyocococoo arararara didididid alalalalal ssssscaccacc r rrrr ofofofofof dddddifififii fefeeefeririririringngngngng etitiitiiololololologogoggogyy yy uuue. JACC Carrrrdididididiovovovovovasasasasascc ccc ImImImmImagagagagaginininininggggg. 2020202020111111;4;4444:1:1:1::1505555 -1-1-115656565656. ,, Wilson JR, ,, XuXXXX e H,HH,HH BBBBandeddd ttttttiinii ii WPWPWPWPWP, ,, ShShShSShanbhbhbhhaggg SSSSSMM,M,MM DDDDruuuuueyeyeyee KM,, Ugagg n

volume fffffrararararactctctctctioioioioonnn nn mamamamamappppppppppinnng ggg ininininin ttttthehehehehe mmmmmyoyoyoyoyocacacacacardrdr iuiuiuiuumm,mmm pppararararartt ttt 2:2:2:2:2: IIIIInininininitititititialalalalal ccccclililililininn cal ex



ownloaded from


Table 1. Characteristics of patients who underwent cardiovascular magnetic resonance and heart transplantation

BMI indicates body mass index; BSA body surface area; eGFR estimated glomerular filtration rate; EDV end diastolic volume; ESV end-systolic volume; EF

ejection fraction. ‘Indexed’ refers to indexed to body surface area.

Patient Sex Diagnosis Age at

transplantation

(years)

Time between

CMR and

transplantation

(days)

Height

(m2)

Weight

(kg)

BMI

(kg/m2)

BSA

(m2)

eGFR

(mL/min/

1.73m2)

Indexed

EDV

(mL/m2)

Indexed

ESV

(mL/m2)

EF

(%)

Indexed

mass

(g/m2)

1 M DCM 44 10 1.83 80.4 24.0 2.0 74 269 199 26 52

2 M DCM 25 31 1.75 82.2 26.8 2.0 76 251 209 17 62

3 M IHD 48 276 1.68 73.5 26.0 1.8 65 142 112 21 58

4 M IHD 55 26 1.53 60.0 25.6 1.6 95 216 187 14 90

5 M IHD 63 40 1.74 68.0 22.5 1.8 82 106 70 33 54

6 M DCM 46 12 1.72 86.8 29.3 2.0 87 188 130 31 58

0000000 2222222.0.0.0.0.0.0.0

31 1.75 82.2 26.8 2.0

276 1.68 73.5 26.0 1.8

26 1.53 60.0 25.6 1.6

3333311111 1.75 82.222 22222 262222 .88888 2.0

222227676767676 1.6.6.66.688888 737777 .5.555 222226.6666 0 0 0 0 1.1111 888

2626 11111.5.553333 6060606060.00.000 222225.55.55 6 666 1.1.66



ownloaded from


Table 2. Characteristics of healthy subjects

Values are mean ± standard deviation. Age range is also given in brackets. BP indicates blood

pressure, eGFR estimated glomerular filtration rate, EDV end diastolic volume; ESV end-systolic

volume; EF ejection fraction. ‘Indexed’ refers to indexed to body surface area.

Overall

(n=30)

Group A

(0.10

mmol/Kg)

(n=10)

Group B

(0.15

mmol/Kg)

(n=10)

Group C

(0.20

mmol/Kg)

(n=10)

p value

Male 15 5 5 5

Age

Male

Female

45±13 (22-65)

45±15 (22-65)

45±12 (23-64)

44±14 (23-65)

44±15 (27-65)

45±15 (23-62)

45±14 (22-64)

46±17 (22-64)

44±12 (28-59)

46±13 (28-64)

44±15 (30-60)

47±12 (28-64)

0.98

Weight (Kg) 73.1±13.8 74.8±12.9 70.3±16.0 73.1±13.8 0.75

Height (m) 1.69±0.10 1.71±0.09 1.67±0.10 1.70±0.11 0.67

BSA (m2) 1.83±0.20 1.87±0.19 1.78±2.4 1.85±0.18 0.65

HR (bpm) 68±9 67±9 70±11 66±8 0.67

Systolic BP (mmHg) 114±11 116±11 114±11 112±12 0.63

Diastolic BP (mmHg) 67±11 71±11 65±13 66±7 0.41

eGFR (mL/min/m2) 95±17 94±11 100±20 91±19 0.50

Indexed EDV (mL/m2) 79±8 78±8 79±7 80±8 0.83

Indexed ESV (mL/m2) 26±5 27±5 27±5 26±5 0.78

EF (%) 67±5 66±4 66±4 68±5 0.45

BSA corrected Mass (g/m2) 45±8 44±9 43±8 47±7 0.46

0000.3.333333±1±1±1±1±1±1±16.6.66666 0000000 73737373737373

67±0000000 1010101010100 111.69 0. 0 .7 0.09 .67 0. 0 .

68±9 67±9 70±11

114±11 116±11 114±11 1

.69 0. 0 .7 0.09 .67 0. 0 .

1.8383838383±0±0±0±0±0.2.2.2220 000 0 1.111 8887±0±00.19 9 9 9 9 1.1.111 7778±2±2±2±2±2 4444.4 1.

666668±8±8±±±9 6767676767±9±9±9±9±9 7770±00 1111111111

111141414 11±1111 1111116±66 111111 11111141414±1±1±11±1111 111



ownloaded from


24

Figure Legends

Figure 1. Short-axis voxel-wise T1 relaxation time maps at basal (A and B; colors correspond to T1

relaxation time according to color bar legends), mid (D and E) and apical (G and H) ventricular levels

before (A, D, G) and 15-minutes after contrast agent bolus (B, E, H), with corresponding macroscopic

explanted heart sections (C, basal; F, mid and I, apical ventricular levels).

Figure 2. Histological analysis. Tissue blocks taken from each segment29 were embedded in paraffin

and stained with picrosirius red in order to determine histological collagen volume fraction (CVF).

Examples shown demonstrate (A) 8%, (B) 16%, (C) 36% CVF.

Figure 3. Dynamic-equilibrium CMR-measured myocardial extracellular volume, calculated using 10-

minute (A) and 15-minute (B) post-contrast T1 values, plotted against histological collagen volume

fraction. Isolated post-contrast T1 measurements made at 10-minutes (C) and 15-minutes (D) post-

contrast plotted against histological collagen volume fraction. Symbols correspond to different

patients, as set out in the legends.

Figure 4. Mean individual dynamic-equilibrium CMR-measured myocardial extracellular volume,

calculated using 15-minute post-contrast T1 values, plotted against mean individual histological

collagen volume fraction.

Figure 5. Dynamic-equilibrium CMR-measured myocardial extracellular volume fraction plotted

against histological collagen volume fraction, split according to the presence (red symbols) or absence

(black symbols) of infarct-typical late gadolinium enhancement (LGE) (A), and split according to the

presence (red symbols) or absence (black symbols) of any LGE (infarct-typical or infarct-atypical

patterns) (B). Symbols correspond to different patients, as set out in the legends.

ular vovovovovovov lululululululumememememememe, , cacacaaaacalclclclclclclcuuuuuuu

- g

s

ainst histological collagen volume fraction. Symbols correspond to dif

-mimimimiminnununn tttte (B(B(B(B(B) pppppost-contrast T1 values, ppplotted agaiaiaiaiainst tttt hhihhh stological collag

ooosoo ttt-tt contrast TTT1 mem asssuuuru emennee tsss mmaaade aaat 111110-0-0-0-0-mimim nuutttes (C(CC) aaand 11111555-55 mim nuutttes

ainst histologigigigigicacacacacalll ll cococococ lllllll agagagagageneneenen vvvvvolololololumumumumume eee frffracacacacactitiititiononono . SySySySySymbmbmbmbmbololololo sssss cococococorrrrrrrrrrespond to dif

in the leeeeegegegegegendndndndnds.sss



ownloaded from


25

Figure 6. (A) Myocardial (Myo, dashed lines) and blood (continuous lines) T1 relaxation times plotted

against time after contrast (Gd-DTPA) administration. Group A (red) received 0.10mmol/kg contrast

agent, Group B (blue) received 0.15mmol/kg and Group C (green) received 0.20mmol/kg. Pre-contrast

(Time 0) myocardial and blood T1 relaxation times were not significantly different between groups.

Post-contrast T1 relaxation times shortened significantly as contrast dose increased. (B) Myocardial

and blood R1 (1/T1) values plotted against time after contrast administration. Labeling as per A. (C)

Dynamic-equilibrium CMR-measured myocardial extracellular volume fraction (ECV) plotted against

time following contrast administration. ECV was significantly higher in group A compared to group B

and group C, but there was no significant difference between groups B and C. ECV increased linearly

over time in each group. Error bars represent ±1 SD.

Figure 7. Dynamic-equilibrium CMR-measured segmental myocardial extracellular volume fraction

(ECV) in; healthy subjects (Healthy), segments without late gadolinium enhancement in patients prior

to transplantation (No LGE), segments with infarct-atypical LGE in patients prior to transplantation

and segments with infarct-typical LGE in patients prior to transplantation. ECV was calculated using

15-minute post-contrast (0.20mmol/Kg Gd-DTPA) T1 values throughout. Horizontal lines represent

mean ± 1SD. ECV was significantly higher in segments without LGE in patients awaiting

transplantation than in healthy subjects. ECV was significantly different between segments without

LGE, segments with infarct-atypical LGE and segments with infarct-typical LGE, but on post-hoc

analysis only the difference between segments with infarct-typical LGE and segments without LGE

was significant (p<0.001).

al extrtrtrtrtrtrtracacacacacacacelelelelelelellulululululululalalalalalalarrrrrrr vovovovovvov luq g y

s n

( n

c

q g y

suuuuubbjbb ects (HeHeHeHeHealaa thhhhhy)y)y)y)y), sesesesesegmgmgmgmgmenenenenentststtt witititii hohhh utut latatatatateee ee gagagagagadooddod liiiinniuiuuiuummm enenenenenhahahahahancncncncncememememmenenenenent tt t t inii

(Nooooo LLLLLGEGEGEGEGE),)))) ssssegegegmemememm ntntntnn s wiwiwiwiwiththththth infnfnfffarararararctctctctc -----atatataa ypypypypypicicicccalalalalal LLLLLGEGEGEGEGE iiiin nnnn papapapp tittittiennnnntststststs ppppprrrirr ororororor ttttto ooo trtrtranaaaa

infarct-tyyypip calllll LGLGLGLLGE EEEE iini pppatienttts prpp iiioii r ttto transpllpllantatatataatitititit on. EEECEE V was calc



ownloaded from




ownloaded from




ownloaded from




ownloaded from




ownloaded from




ownloaded from




ownloaded from




ownloaded from


J.M. Parker and Matthias SchmittRay, Nizar Yonan, Simon G. Williams, Andrew S. Flett, James C. Moon, Andreas Greiser, Geoffrey Christopher A. Miller, Josephine Naish, Paul Bishop, Glyn Coutts, David Clark, Sha Zhao, Simon G.

Assessment of Myocardial Extracellular VolumeComprehensive Validation of Cardiovascular Magnetic Resonance Techniques for the

Print ISSN: 1941-9651. Online ISSN: 1942-0080 Copyright © 2013 American Heart Association, Inc. All rights reserved.

TX 75231is published by the American Heart Association, 7272 Greenville Avenue, Dallas,Circulation: Cardiovascular Imaging

published online April 3, 2013;Circ Cardiovasc Imaging.

http://circimaging.ahajournals.org/content/early/2013/04/03/CIRCIMAGING.112.000192World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circimaging.ahajournals.org/content/suppl/2013/04/03/CIRCIMAGING.112.000192.DC1Data Supplement (unedited) at:

http://circimaging.ahajournals.org//subscriptions/

is online at: Circulation: Cardiovascular Imaging Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further information aboutnot the Editorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center,Circulation: Cardiovascular Imaging Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:



ownloaded from

http://circimaging.ahajournals.org/content/early/2013/04/03/CIRCIMAGING.112.000192

http://circimaging.ahajournals.org/content/suppl/2013/04/03/CIRCIMAGING.112.000192.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circimaging.ahajournals.org//subscriptions/


1

SUPPLEMENTAL MATERIAL

Phantom studies

Supplemental methods

Fifteen agarose and CuSO4 phantoms with T1 values of 175 – 1775ms (i.e. a range encompassing

physiological pre-‐ and post-‐contrast myocardial and blood T1 values), and a constant physiological T2

value (55±8ms), were scanned using the MOLLI sequence with simulated heart rates (50 to 110 beats

per minute, 10 beats increments). Reference T1 relaxation times were measured using an inversion

recovery spin echo sequence with 10 inversion times ranging from 25 to 5000ms (TR 10s, TE 8.1ms);

reference T2 relaxation times were measured using a spin echo sequence with 10 echo times ranging

from 10 to 200ms (TR 10s). Phantom imaging was repeated in order to assess inter-‐scan

reproducibility of MOLLI-‐derived T1 measurements.

Supplemental results

As has been demonstrated previously,1 the accuracy of the MOLLI sequence for measurement of T1

relaxation time was found to be dependent on T1 and heart-‐rate, with progressive underestimation

of T1 relaxation time, as compared with the inversion recovery spin echo-‐derived T1 measurements,

as T1 and heart rate increased (Supplemental Figure 1). Inter-‐scan reproducibility of phantom T1

measurements made using the MOLLI sequence was 2.1ms, (intraclass correlation coefficient 1.0).

Heart rate correction algorithm

Using the above phantom data, MOLLI-‐derived T1 relaxation times at different heart rates were fitted

to the inversion recovery spin echo-‐derived T1 relaxation times using second order polynomial fitting,

in order to derive the following correction algorithm,2 which was subsequently applied to all in-‐vivo

MOLLI-‐derived T1 measurements:

2

T1-‐IR = A*(T1-‐MOLLI)2 + B*(T1-‐MOLLI) + C

Where T1-‐IR = inversion recovery spin echo-‐derived T1 relaxation time; T1-‐MOLLI = MOLLI-‐derived T1

relaxation time; and A, B and C are as displayed in Supplemental Table 1. For the in vivo data, linear

interpolation of the coefficients between adjacent simulated heart rates was used.

Impact of heart rate correction algorithm

Application of the heart rate correction algorithm significantly increased mean pre-‐contrast

myocardial T1 relaxation times (for example in patients prior to transplantation, uncorrected T1

1086±129ms; corrected T1 1187±163ms; p<0.001). However the heart rate correction algorithm did

not lead to a significant difference in mean DynEq-‐CMR-‐derived ECV (ECV calculated using

uncorrected T1 values 43.8±6.8% versus ECV calculated using corrected T1 values 43.9±6.7%,

p=0.903). Furthermore, the linear regression equation for the relationship between DynEq-‐CMR-‐

derived ECV and histological CVF using uncorrected T1 values (histological CVF = 1.445 x ECV –

41.588) was almost identical to that when corrected T1 values were used (histological CVF = 1.451 x

ECV – 42.023). As such, for future work using DynEq-‐CMR-‐derived ECV calculated using the MOLLI

sequence applied here, a heart rate correction algorithm (which is cumbersome to apply) may not be

necessary. Nevertheless, for this validatory paper the heart rate correction algorithm was applied

throughout.

3

Supplemental Table 1

Heart rate A B C

50 0.0000671 1.013 -‐13.64

60 0.0000871 0.991 -‐9.38

70 0.0001215 0.951 -‐0.532

80 0.0001565 0.91 8.19

90 0.0001762 0.893 10.38

100 0.0002193 0.85 18.07

110 0.0002501 0.823 21.88

4

Supplemental Figure 1.

Black line represents the identity line.

5

Supplemental references

1. Messroghli DR, Greiser A, Fröhlich M, Dietz R, Schulz-‐Menger J. Optimization and validation of a

fully-‐integrated pulse sequence for modified look-‐locker inversion-‐recovery (MOLLI) T1 mapping of

the heart. J Magn Reson Imaging. 2007;26:1081-‐1086.

2. Lee JJ, Liu S, Nacif MS, Ugander M, Han J, Kawel N, Sibley, CT, Kellman P, Arai AE, Bluemke DA.

Myocardial T1 and Extracellular Volume Fraction Mapping at 3 Tesla. J Cardiovasc Magn Reson.

2011;13:75.

Documents

Comprehensive Validation of Cardiovascular …circimaging.ahajournals.org/content/circcvim/early/2013/04/03/CIRC...Comprehensive Validation of Cardiovascular Magnetic Resonance Techniques