logy 128 (2008) 364–373www.elsevier.com/locate/ijcard

International Journal of Cardio

Cardiovascular magnetic resonance signs of ischemiain hypertrophic cardiomyopathy

Paola Melacini a,⁎, Francesco Corbetti b, Chiara Calore a, Valentina Pescatore a,Gessica Smaniotto c, Andrea Pavei a, Fabiana Bobbo a, Luisa Cacciavillani a, Sabino Iliceto a

a Department of Cardiac, Thoracic and Vascular Sciences, University of Padua Medical School, Padua, Italyb Department of Radiology, University of Padua Medical School, Padua, Italy

c Department of Biology, University of Padua, Padua, Italy

Received 22 January 2007; received in revised form 12 June 2007; accepted 30 June 2007Available online 23 July 2007

Abstract

Background: Recurrent myocardial ischemia has been recognized as playing an important role in the pathophysiology of hypertrophiccardiomyopathy (HCM) and cardiovascular magnetic resonance (CMR), with or without gadolinium, is a promising method of evaluatingfibrosis, edema and hypoperfusion. The aim of this study is to evaluate the interrelationship between late enhancement (LE) and other signsof ischemia, such as edema and perfusion defects, and to relate them to clinical data in order to describe the stage of the disease.Methods: Forty-four patients were evaluated by CMR cine images, T2-weighted sequences for edema and LE sequences. First-pass perfusionstudywas obtained in 37 patients. Acute-subacute ischemic events were clinically defined as the presence of chest pain or new onset of ST-segmentdepression, end-stage phase by left ventricular ejection fraction b50% and maximal left ventricular wall thickness b25 mm.Results: Intramural patchy LE was found in 35/44 (80%) patients; extensive LE in 4/44 (9%). Edema was present in 24/44 (54%) patients andperfusion defects in 17/37 (46%). Simultaneous presence of patchy LE, edema and hypoperfusion in corresponding segments, wassignificantly associated to acute–subacute ischemic-phase parameters (p=0.02; RR 1.99, 95% C.I. 0.77–5.02). Extensive LE and perfusiondefects in the absence of edema were significantly related to end-stage HCM (pb0.001; RR 13.7, 95% C.I. 1.83–102.05).Conclusions: Using CMR in patients with HCM, we found focal tissue abnormalities consistent with regional ischemia at various stages.CMR provides important, clinically relevant information on the acuity, extent and functional relevance of ischemic injuries in HCM.© 2007 Elsevier Ireland Ltd. All rights reserved.

Keywords: Hypertrophic cardiomyopathy; Myocardial ischemia; Magnetic resonance imaging

1. Introduction

Hypertrophic cardiomyopathy (HCM) is a genetic myo-cardium disease caused by mutations in the genes encodingsarcomeric proteins and is characterized by typical asym-metric left ventricular hypertrophy, myocardial disarray, in-

⁎ Corresponding author. Department of Cardiac, Thoracic and VascularSciences, University of Padua, Policlinico, Via Giustiniani 2-35128 Padua,Italy. Tel.: +39 049 8211776; fax: +39 049 8761764.

E-mail address: paola.melacini@unipd.it (P. Melacini).

0167-5273/$ - see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ijcard.2007.06.023

terstitial fibrosis and small vessels disease [1,2]. The outcomeand prognosis of HCM depends on many different patho-physiological mechanisms. Nowadays, ischemia is recog-nized as playing an important role in the natural history ofHCM patients. Acute ischemic events may act as a trigger forelectrical instability and consequent malignant arrhythmias,as well as for recurrent myocardial ischemia which may leadto the replacement fibrosis responsible for disease progres-sion [3], severe heart failure and death [4,5].

Cardiovascular magnetic resonance (CMR), with orwithout contrast media (gadolinium), is now considered tobe a sensitive method for identifying the presence, location

365P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

and extent of myocardial infarction in both acute and chronicsettings [6–8]. The mechanism of late enhancement (LE) iscorrelated to the interstitial kinetics of gadolinium contrastmedia, “inert” extracellular agents that are excluded from theintracellular space by intact sarcolemmal membranes [9].Myocardial regions with increased gadolinium concentra-tions are considered to correspond to the absence of viablemyocytes, including post-ischemic necrosis, edema, or areaswith augmented extracellular space, such as fibrotic tissue[10–13]. Late-enhancement is a common feature in patientswith HCM and its extension has been described as beingrelated to the severity of hypertrophy, electrocardiographicalterations [14], functional parameters, progression of thedisease and risk of sudden death [15–18]. Until now LE inHCM has been related, as in chronic myocardial infarction, tomyocardial fibrosis and increased collagen content [19,20].However, as Knaapen et al. recently demonstrated in a studycomparing CMR and PETwith oxygen-15-labeled water andcarbon monoxide [21], LE in HCM might correspond notonly to fibrosis [19,20,22,23], but also to edema that wasdocumented, first as intracellular, and subsequently as inter-stitial fluid extravasation, in the early stage of ischemic events[22].

The aim of this study was to use CMR to identify signs ofischemia, such as edema and hypoperfusion, and to correlatethem with LE and with clinical–functional data in order toidentify the stage of the disease.

2. Methods

2.1. Patient population

The patient cohort consisted of 44 consecutive patientsreferred to Padua University HCM outpatient clinic. Initialevaluation was performed from October 2003 to November2004. HCM was diagnosed using conventional echocardio-graphic criteria: maximal left ventricular wall thickness≥15 mm in adult patients in the absence of other diseases(aortic stenosis, systemic hypertension, amyloidosis) thatcould account for left ventricular hypertrophy [1,2,24].

Exclusion criteria were the presence of an implantable de-vice, mechanical valve prosthesis, claustrophobia, atrial fibril-lation or ventricular arrhythmias at the time of examination.

For each patient we collected: clinical history (age, sex,family history [24] and symptoms); electrocardiographic data(presence of abnormal Q waves, assessed by Minnesota code[25]; ST depression; giant negative T waves; arrhythmias at24 h Holter monitoring) and echocardiographic morpholog-ical-functional parameters (left atrial diameter, maximal leftventricular wall thickness, left ventricular end diastolic dia-meter, presence and entity of left ventricular outflow tractobstruction, E-wave deceleration time).

Nine patients underwent coronary angiography for chestpain.

All patients underwent CMR: the median time that elapsedbetween baseline evaluation and a CMR study was 6 months.

A mean follow-up of 25 months was obtained in 41patients, which included clinical, echocardiographic, 24 hourHolter monitoring or implantable cardioverter defibrillator(ICD) recording evaluation.

All patients gave their informed consent to the protocolstudy.

2.2. CMR

Patients were examined in a supine position on a clinical1.0-T scanner (Harmony, Siemens, Erlangen, Germany)using cardiologic software by Siemens with system MReaseSYNGO 2002B. All images were acquired using a 4 channelphased-array receiver coil during repeated breath-holds ofvarying duration, which depended on heart rate and patientcompliance (12 to 15 s).

Functional evaluation was obtained through acquisition ofserial and contiguous 8 mm thick short axis views from themitral valve plane to the apex, carried out using a steady statefree precession sequence, and by deriving left ventricularend-diastolic and end-systolic volume, ejection fraction, leftventricular mass and maximal left ventricular wall thicknessusing cardiologic software (Argus, Siemens). Three addi-tional long axis views (2, 3, 4 chambers) were obtained fordynamic evaluation purposes [26].

To detect any abnormal signal consistent with edema wealso acquired both 3 short axis (at the level of the mitralvalve, papillary muscles and apex) and 3 long axis views (atthe same level of cine MR sequences) black blood TSE FatSuppressed T2 weighted sequences. This was done byadapting the acquisition parameters to heart rate and breath-hold capability (TR 2–3 RR; echo train 19–23: TE 92 ms;BW 235 Hz/pixel; mean voxel size 1.6×1.4×7 mm). Whenfindings were equivocal, as previously reported [27], addi-tional heavily T2 weighted black blood TSE inversion reco-very images were obtained, with a time inversion (TI) that waseffective in suppressing the signal from normal myocardiumon our CMR unit and able to highlight areas of “focal” edema(TR: 2 R–R; TE 54 ms; echo train 11–17, TI 220 ms).

In 10 patients with hyperintense areas on T2 images,additional phase reconstructed TSE inversion recovery imageswere also acquired both to confirm the presence of tissueedema and to rule out potential causes of the hyperintensesignal [28]. The aforementioned sequences were not per-formed in all cases to avoid an over long examination and therisk of image artifacts due to the patient moving.

On T2 weighted images, areas of hyperintense signalswere visually identified using a 17-segment model [29] andtheir signal intensity was evaluated on the basis of a region ofinterest of adequate size. For evaluation purposes, a referenceregion of interest of the same size was drawn and positionedin the visually normal adjacent myocardium, in order to avoidany mismatch in signal intensity measurements due todifferent distance from the receiver coil [30]. Hyperintenseareas with signal intensity ≥2 SD higher than the referenceregion of interest were considered significant for edema.

Table 1Patient characteristics

Characteristic Value

Age (years) 47±14Men, n (%) 29 (66)HCM family history, n (%) 28 (64)Sudden death family history, n (%) 10 (23)Chest pain, n (%) 11 (25)Q wave, n (%) 14 (32)ST depression, n (%) 14 (32)Giant negative T wave, n (%) 17 (39)Paroxysmal atrial fibrillation on Holter monitoring, n (%) 10 (23)SVT or NSVT on Holter monitoring or ICD querying, n (%) 20 (45)LVOTO ≥30 mm Hg, n (%) 13 (29)

HCM=hypertrophic cardiomyopathy; ICD=implantable cardioverter defi-brillator; LVOTO=Left ventricular outflow tract obstruction; NSVT=non-sustained ventricular tachycardia; SVT=sustained ventricular tachycardia.

366 P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

Following cine and T2 image acquisition, all patientswere injected with 0.2 mmol/kg gadolinium contrast medium(gadobenate dimeglumine, Multihance, Bracco Italy). In 37cases a bolus injection (3 ml/s) was performed and a restingfirst-pass study was acquired using a Saturation RecoveryTurbo Flash sequence (TR 330 ms; TE 1.79 ms; TI 169 ms;α 8°; voxel size 2.7×1.4×8 mm). In all these 37 patients, thefirst-pass study was performed in long axis by single sliceacquisition. In patients with evidence of “focal” areas ofedema on T2 weighted images, the first-pass study wasperformed by carefully positioning the slice over the areas ofhigh signal intensity. A standard 3 chamber long axis,encompassing the junction of the anterior septum with themyocardial wall, was used in patients with no evidence ofhyperintense areas on the T2 weighted study.

Perfusion defects at first-pass images were defined,visually, as areas of persistent hypoenhancement followingcontrast medium bolus injection or delayed wash in ≥2 s[31].

In all patients, contrast LE was evaluated 10 min aftercontrast medium injection and the same views obtained incine study were acquired by a segmented turbo flash inver-sion recovery sequence, modulated in relation to heart fre-quency and breath-hold capability (TR 1 or 2 R–R, K space/TR 17–21 lines, TE 6 ms, flip angle 30°, BW 150 Hz/pixel,resolution 134×256, typical voxel size of approximately1.8×1.3×8 mm, TI optimized for each patient, range 190–220 ms); late enhanced areas were considered, by definition[32], as those characterized by a signal intensity at least400% higher than the signal from normal myocardium. Thepresence of LE was visually assessed on all short axes and on3 long axis slices. Spatial extension of LE was assessedreferring to a 17-segment model.

CMR images were analyzed independently by 2 obser-vers blinded to patient name, clinical, electrocardiographic,echocardiographic and procedural characteristics.

2.3. Definitions

Acute-subacute ischemic events were defined as eitherpresence of chest pain or new onset of ST-segment depression.End-stage phase was classified by left ventricular ejectionfraction b50% and maximal left ventricular wall thicknessb25 mm.

2.4. Genotype testing

Mutation screening of several sequences of genes MYH7,MYBPC3, TNNI3 and TNNT2, considered mutation-proneon the basis of reported data, was performed in 24/44patients. In particular, exon 2–40 of β-myosin heavy chain,MYH7, exon 1–35 of myosin binding protein C, MYBPC3,exon 1–8 of troponin I, TNNI3 and exon 8–16 of troponin T,TNNT2, were screened. Each DNA segment was PCRamplified using primers designed for to the purpose. PCRproducts were analyzed by denaturing high performance

liquid chromatography (DHPLC), followed by automaticDNA sequencing of segments showing extra peaks.

All novel mutations were checked for polymorphism, bytesting a large cohort of healthy unrelated DNA samples (150individuals for missense variations, deletions and insertions;250 for splice-site variations).

2.5. Statistical analysis

Results are expressed as mean value±SD for continuousvariables and as frequency with percentage for categoricalresults. Differences between means were tested by unpairedStudent t test. Frequencies were compared by chi square testor Fisher's exact text, when appropriate. Linear correlationswere used to study relationships between continuous vari-ables. All probability values reported are 2-sided and aprobability value b0.05 was considered to be statisticallysignificant. All calculations were performed by SPSS/WIN12 and STATA 7.0.

3. Results

3.1. Patient characteristics

The clinical, electrocardiographic and echocardiographicdata at time of enrollment are summarized in Table 1.The mean age was 47±14 years: 4 (9%) patients wereb30 years old, 10 (23%) were 30 to 39 years old, 12 (27%)were 40 to 49, 7 (16%) 50 to 59 and 11 (25%) more than60 years old.

Of the 44 patients, 28 (64%) had a family history of HCMincluding 10 families in which at least one relative had diedsuddenly of HCM.

Causative sarcomere protein mutations were found in 8/24 patients: 5 in MYH7, 2 in MYBPC3 and 1 in TNNI3(Table 2).

Eleven patients were symptomatic for chest pain: 9 ofthem underwent coronary angiography that excluded thepresence of significant coronary artery stenosis (among these

Table 2Genetic, clinical, echocardiographic and CMR data of the eight genotyped patients

Id Gender Age(years)

FamilyHCMhistory

FamilySCDhistory

Genemutations

NYHAclass

Max LVT(mm)

Outflowgradient(mm Hg)

LA(mm)

DT(ms)

PAF NSVT EF(%)

Mass(g)

LE(numberofsegments)

Edema FP SCD/HFDor TX

1 Male 24 + O MYH7 A 20 O 35 280 O O 54 60 1 O O SCDEX13Gly407Cysnovel

2 Female 33 O O MYH7 A 28 O 40 214 O O 68 63 4 O O OEX19Gly716Argknown

3 Male 34 + O MYH7 III 36 20 51 113 O + 61 111 4 + + OEX22Met877Ilenovel

4 Female 35 + O MYH7 III 18 O 51 88 + O 57 60 13 O − TXEX19Arg719Glnknown

5 Female 67 + + MYH7 III 22 O 57 560 + O 50 67 5 O O OEX36Thr1760Metnovel

6 Male 36 + O MYBPC3 A 28 O 45 − O O 70 109 1 O O OINTR4-5538-2A→Cnovel

7 Male 40 + O MYBPC3 II 32 100 56 147 O O 60 137 2 + − OEX25Ala848Glunovel

8 Female 57 + O TNNI3 II 19 O 47 115 + O 45 86 1 + O OEX8Val188Alanovel

−, Data not available; +, present; O, absent; A, asymptomatic; CMR, cardiovascular magnetic resonance; DT, deceleration time; EF, ejection fraction; EX, exon;FP, first pass; HCM, hypertrophic cardiomyopathy; HFD, heart failure death; INTR, intron; LA, left atrium; LE, late enhancement; LVT, left ventricular thickness;MYBPC3, myosin binding protein C; MYH7, β-myosin heavy chain; NYHA, New York Heart Association; NSVT, non-sustained ventricular tachycardia; PAF,paroxysmal atrial fibrillation; SCD, sudden cardiac death; TNNI3, troponin I; TX, cardiac transplantation.

Table 3Cardiovascular magnetic resonance data

Variable Value

Maximal LV WT (mm) 25±6LV mass (g/m2) 94±26LV ejection fraction (%) 56±13Edema, n (%) 24 (54)First-pass ⁎ positive, n (%) 17 (46)LE, n (%) 39 (89)

Patchy (involving 1 to 8 LV segments) 35 (80)Extensive (involving 9 or more LV segments) 4 (9)

LE=Late-enhancement; LV=left ventricular; WT=wall thickness.⁎ Performed in 37 patients.

367P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

patients 5 also had elevated troponin I). The remaining2 patients with chest pain were a 47-year-old woman and a40-year-old man, both without any risk factor for coronaryartery disease. At baseline or during follow-up, 20 patientshad sustained or non-sustained ventricular tachycardia onHolter monitoring or ICD querying, and 10 patients hadparoxysmal atrial fibrillation.

3.2. CMR

3.2.1. Functional studyCMR findings are detailed in Table 3. The maximal left

ventricular wall thickness and left ventricular mass,derived from cine short axis acquisitions, ranged from15 to 33 mm (mean 25±6 mm) and from 53 to 156 g/m2

(average 94±26 g/m2), respectively; the left ventricularejection fraction ranged from 30 to 72% (mean 56±13%).

3.2.2. Late enhancement and T2 studiesIn a small number of cases (5/44, 11%) both precontrast

T2 and LE studies were negative. Late enhancement was

368 P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

found in 39 of the 44 patients (89%) who showed a variablespatial extension: in 35 patients (80%) LE involved from 1 to8 left ventricular segments (mean number of segmentsinvolved=2.7±1.5), it was patchy, intramural and mostoften localized within hypertrophied regions, particularly atthe junction of the interventricular septum and the rightventricular free wall. Conversely, in 4 patients (9%), LEinvolved diffusely 9 or more adjacent segments (meannumber of segments involved=13.3±1.3) and was alsofrequently transmural.

On T2 weighted images, areas of significantly increasedsignal intensity were found in 24/44 (54%) patients andshowed marked hypointensity when evaluated by phasereconstructed TSE inversion recovery sequences, thusconfirming long T1 and T2 relaxation times and, therefore,the presence of edema (Fig. 1A). In all cases, edema waslocated within the most hypertrophied regions and corre-sponded, for size and location, to patchy areas of LE. Noneof the 4 patients with extensive LE showed edema.

3.2.3. Perfusion studyFirst pass study was performed in 37 patients and

disclosed perfusion defects in 17 cases (46%). In 14/37patients (38%) perfusion defects were detected within areasof edema and patchy LE on T2 and delayed images; 3additional perfusion defects were observed in patientswithout edema and all of these showed extensive LE ondelayed images. In the remaining 20 cases, perfusion studywas negative, irrespective of the presence of edema and/orLE.

Fig. 1. Cardiovascular magnetic resonance images showing spatial relation betweenfor edema (upper row) and late-enhancement (LE) images (lower row) in 4 represeTSE inversion recovery image (A′) is indicative of long T1 relaxation time, thus

Representative examples of the relation between areas ofedema and LE are given in Fig. 1, the correlation betweenareas of edema and hypoperfusion in patients with patchy LEin Fig. 2, and an example of hypoperfusion in a patient withextensive LE but without edema in Fig. 3.

3.3. Relation between CMR findings and clinical-ECG-functional features

Simultaneous detection of increased T2 signal intensityconsistent with edema, the presence of patchy LE (involvingfrom 1 to 8 left ventricular segments) and perfusion defects atfirst-pass study were associated with chest pain or new onsetof ST segment depression (RR 1.99, 95% C.I. 0.77–5.02;p=0.02). The presence of extensive LE (involving ≥9 seg-ments) and hypoperfusion at first-pass, in the absence ofedema, were significantly associated with impaired systolicfunction (ejection fraction b50%) and left ventricular wallthickness b25 mm ( pb0.001; RR 13.7, 95% C.I. 1.83–102.05).

A weak but statistically significant negative correlationwas found between the number of segments showing LE andboth left ventricular ejection fraction and deceleration time(r=−0.45; p=0.003, and r=−0.40; p=0.01).

3.4. Relation between CMR findings and outcome

After only 4 months of follow-up, a 24-year-old patientcarrying a novel mutation in β-myosin heavy chain (MYH7Gly407Cys) and without any risk factor for sudden death,

tiny areas of increased signal intensity (arrowheads) on T2 weighted imagesntative patients (A, B, C, D). In case A, hypointensity in phase reconstructedconfirming the presence of tissue edema.

Fig. 2. T2 weighted (upper row), first-pass frame (FP) (middle row), and late-enhancement (LE) (lower row) images (long axis) in 3 patients with focalhypoperfusion. The correspondence among focal edema, hypoperfusion, and LE is well demonstrated in cases A, B, C (arrowheads); note the questionable darkzone (arrow) in LE image of case B, not surely explained by “microvascular obstruction”, because of the normal perfusion on the FP image.

369P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

died suddenly. CMR had demonstrated the presence of anintramural septal gross LE patch (Table 2, #1). Two patientshad aborted sudden deaths caused by ventricular tachycardiaand/or fibrillation and received ICDs for secondary preven-tion. Both of them presented end-stage HCM forms with

Fig. 3. T2 weighted (on the left), first-pass (FP) frame (in the middle), and late-enhend-stage phase hypertrophic cardiomyopathy (HCM) (ejection fraction = 30%), shis a 37-year-old female with HCM diagnosed when she was 5 years old. Twehypertrophy and serial examinations demonstrated progressive thinning and dilatin

wide LE extension. Seven more patients received ICDs forprimary prevention.

Sustained or non-sustained ventricular tachycardia hadbeen documented in 20 patients (including the two above-mentioned patients with aborted sudden deaths).

ancement (LE) (on the right) images (long axis) concerning to a patient withowing hypoperfusion and extensive LE in the absence of edema. The patientnty years previously echocardiography showed severe asymmetric septalg left ventricle.

370 P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

Remarkably, in all of them, CMR revealed the presence of LEwith a wide extension range (from one patch to an extensivepattern) and edema was found in 14 patients. Arrhythmicevents were more frequent in patients with LE (20/39, 51% vs0/5, 0%, p=0.039), although 19 patients with LE did notshow ventricular arrhythmias (sensitivity of LE in detectingventricular arrhythmias: 100%, specificity: 21%, positivepredictive value: 51% and negative predictive value: 100%)as well as in patients with edema (14/24, 58% vs 6/20, 30%p=0.05).

After 42 months of follow-up a 35-year-old female patientcarrying a known mutation of β-myosin heavy chain (MYH7Arg719Gln) and a restrictive form of HCM, underwent car-diac transplantation for progressive diastolic heart failure.CMR showed an extensive “reticular-like” LE pattern but noedema (Table 2, #4).

4. Discussion

The clinical course of HCM is the result of a complexinteraction of pathophysiological mechanisms. Both leftventricular hypertrophy, with or without outflow tract obs-truction, and structural myocardial alterations, such as inter-stitial fibrosis, disarray and microvascular disease, are allresponsible for diastolic dysfunction, left ventricular remo-deling, ischemic events and arrhythmias. Because the in-fluence of each factor varies greatly in different patients, theclinical course and natural history of the disease are extre-mely heterogeneous [33]. Acute–subacute ischemic events,as well as chronic myocardial ischemia, are recognizedas playing an important role in the course of the disease.Recently, using positron emission tomography, Olivottoet al. have demonstrated that severe microvascular dysfunc-tion is a potent long-term predictor of adverse remodelingand systolic dysfunction [3].

Nowadays, cardiovascular magnetic resonance is not onlya sensitive technique for the study of HCM, because ofits utility – from both the morphological and the functionalpoint of view – as a gold standard for left ventricular massand ejection fraction calculation [34], but also, it is nowconsidered to be a promising non-invasive method for tissuecharacterization [18–35]. Since 2002, several studies havedemonstrated the presence of LE in HCM patients andhave revealed correlations between LE extent and leftventricular hypertrophy, left ventricular function determi-nants, and risk factors for sudden death [15–18]. A study byMoon et al. and a case report by Papavassiliu et al. in end-stage HCM patients who performed CMR before hearttransplantations, have proved that the histological basis ofLE was the presence of fibrotic tissue and, in particular, thatLE was detected when the percentage of collagen exceeded15% [19,20].

As reported in previously published studies [15–17] wefound LE in 39 (89%) out of 44 patients. On the basis of itsspatial extension, LE was either patchy and intramural orextensive and transmural. The patchy pattern was found in

35 patients and, as described by Choudhury et al. [15], waspredominantly located within hypertrophied regions, partic-ularly at the junction of the interventricular septum and theright ventricular free wall. Extensive pattern was limited to 4patients: 3 patients with end-stage phase and 1 with arestrictive HCM form characterized by normal ejectionfraction, mild hypertrophy, huge atrium and restrictive leftventricular rapid filling. This latter patient carried a mutationin β-myosin heavy chain (MYH7 Arg719Gln) previouslydescribed by some authors as benign [36] and, more recently,by some others as malignant [37,38]. Interestingly the patientdescribed by Huang et al. who harbored the same MYH7mutation showed a restrictive HCM form with a clinicalphenotype very similar to ours [37].

We found a weak, but statistically significant, negativecorrelation between the number of left ventricular segmentswith LE and the ejection fraction of the left ventricle. Thefindings in this study of end-stage phase patients agree withthose of the latest study by Harris et al., who evaluatedprevalence, clinical profile and significance of left ventricularremodeling in end-stage HCM patients and identified largeareas of LE, which were frequently transmural and weredistributed diffusely throughout the ventricular septum andleft ventricular free wall [18]. We also found a statisticallysignificant negative correlation between the number of leftventricular segments with LE and E-wave deceleration timeat echo PW Doppler, thus confirming diastolic dysfunctionconsequent to myocardial tissue alteration.

For the first time, in HCM patients, we also performed T2weighted sequences for the detection of an abnormal signalconsistent for edema [39–42] on all our study population.We found 24 (54%) patients showed increased signalintensity. Late-enhancement was detected with a focal pat-tern and located at the same level of increased T2 signalintensity in all these patients. Conversely, none of thepatients, either those without LE or those with extensive LE,showed increased T2 signal intensity. These findings couldbe explained by the hypothesis that, in the early phase ofacute ischemia, reduction of membrane ATPase induces aninflux of sodium and water within damaged myocardial cellscausing ischemic myocytes to swell followed by an inter-stitial fluid extravasation [43] that may be responsible for thedetection of both edema and LE at CMR. Similarly, inpatients with myocardial infarction, Abdel-Aty et al. havedemonstrated that an imaging approach combining LE andT2-weighted CMR is able to accurately differentiate acutefrom chronic injuries [44]. In addition, comparing theirfindings of CMR and the perfusable tissue index obtainedwith PET with oxygen-15-labeled water and carbon mono-xide in patients with HCM, Knaapen et al. have demon-strated that LE may represent not only fibrosis but also otherconditions, such as edema, inflammation, and necrosis [21].Moreover, in an experimental study in HCM hamsters, Asoet al. [45] found that gadolinium concentrations variedduring the different stages of the disease: higher accumula-tion in the early stages of inflammation and lymphocyte

371P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

infiltration and lower levels of accumulation in the laterstages, which were characterized by more densely packedcollagen fibers.

Finally, we evaluated myocardial perfusion in 37 out of the44 patients using rest first-pass images after gadolinium bolusinjection. Seventeen patients showed perfusion defects and, inthe assessed regions, the perfusion abnormalities were spa-tially associated with regions of LE. As is well known, myo-cardial ischemia in HCM patients may be caused by smallvessel disease, interstitial fibrosis, myocardial bridging, impai-red diastolic relaxation and decreased capillary-to-myocardialfiber ratio (“mismatch”) [46]. Several different techniqueshave demonstratedmyocardial ischemia and reduced coronaryflow reserve in HCM [47–50] and recently CMR first-passperfusion studies have been playing an increasingly importantrole [51,52]. In our study, the simultaneous presence of patchyLE, edema and hypoperfusion in the corresponding segments,had a statistically significant association with acute–subacuteevents i.e. chest pain or new onset ST segment depression.Of note, extensive LE and perfusion defects in the absenceof edema were significantly related to the end-stage phase ofthe disease, characterized by left ventricular ejection fractionb50% and maximal wall thickness b25 mm. Patients withperfusion defects, but without signs of edema, were morelikely to be in a chronic phase of the disease, with the presenceof an expanded interstitial matrix, caused by fibrosis, that hadled to reduced perfusion and was responsible for LE detection,in the absence of acute ischemic injuries.

A 25 month follow-up revealed that 20 patients had hadventricular arrhythmic events ranging from asymptomatic nonsustained ventricular tachycardia to sustained ventriculartachycardia responsible for aborted sudden death. At CMRLE was detectable in all these patients and it was associated toedema in 14 patients. This interesting finding suggests thatacute–subacute ischemic phase tissue alterations may be trig-gers for the electrical instability responsible for life-threateningarrhythmic events and it is supported by a previous pathologicstudy in young HCM patients who died suddenly [22].

CMR data in genotyped patients can add new insight tophenotypic characterization but because of the small numberof patients and the well known clinical heterogeneity furtherstudies are needed before postulating a correlation betweenpathogenic gene mutations and CMR patterns.

5. Limitations

1. First-pass image study was performed in a single sliceacquisition either centered on areas of signal, or using astandard 3 chamber long axis encompassing the junction ofthe anterior septum with the myocardial wall (in patientswho did not show hyperintense areas in a T2 weightedstudy). Presence of perfusion defects outside the selectedslice cannot be excluded.

2. Since a quantitative analysis of LE extension lay outsidethe scope and purpose of this study, we performed only a17 segment-based semi-quantitative evaluation of LE. In

particular, we focused on the presence of edema, perfu-sion defects and LE in the same area.

6. Conclusions

CMR is now considered a promising non-invasive methodfor tissue characterization: the ability to evaluate the patternand extension of LE in association with the sequences foredema and with perfusion study at first-pass, allowed us todetect acute-subacute ischemic events in vivo.

This is the first report of a comprehensive approach tovisualizing the various components of tissue alterations inHCM patients in vivo and our findings support the reasonablepatho-physiological hypothesis which states that, in the acutephase, the mechanisms leading to LE are likely to be celldamage, “focal” edema, and cellular death whereas in thechronic phase, i.e. in patients with perfusion defects butwithout edema, LE is mainly due to fibrosis, a tissue with anaugmented interstitial matrix.

This study highlights the growing clinical utility of CMRboth in the early detection of non-homogeneous zones thatmay reveal the electrical instability substrates responsible forarrhythmic events and in the evaluation of extensive fibrosisleading to systolic and/or diastolic dysfunction and poorprognosis.

Acknowledgement

Dr Melacini has received research grants from ItalianMinistry for Scientific and Technological Research (MURST-COFIN 2002 prot. 2002065749 002, and 2004 prot.2004064807 002).We would like to thank Dr FrancescoMaddalena for his assistance with statistical analyses.

References

[1] Richardson P, McKenna WJ, Bristow M, et al. Report of the 1995World Health Organization/International Society and Federation ofCardiology Task Force on the Definition and Classification ofcardiomyopathies. Circulation 1996;93:841–2.

[2] Maron BJ, McKennaWJ, Danielson GK, et al. ACC/ESC clinical expertconsensus document on hypertrophic cardiomyopathy: a report of theAmerican College of Cardiology Task Force on Clinical Expert Con-sensus Documents and the European Society of Cardiology Committeefor Practice Guidelines (Committee to Develop an Expert ConsensusDocument on Hypertrophic Cardiomyopathy). J Am Coll Cardiol2003;42:1687–713.

[3] Olivotto I, Cecchi F, Gistri R, et al. Relevance of coronary microvascularflow impairment to long-term remodeling and systolic dysfunction inhypertrophic cardiomyopathy. J Am Coll Cardiol 2006;47:1043–8.

[4] Varnava AM, Elliott PM, Sharma S, McKenna WJ, Davies MJ.Hypertrophic cardiomyopathy: the interrelation of disarray, fibrosis,and small vessel disease. Heart 2000;84:476–82.

[5] Varnava AM, Elliott PM, Mahon N, Davies MJ, McKenna WJ.Relation between myocyte disarray and outcome in hypertrophiccardiomyopathy. Am J Cardiol 2001;88:275–9.

[6] Kim RJ, Fieno DS, Parrish TB, et al. Relationship of MRI delayedcontrast enhancement to irreversible injury, infarct age, and contractilefunction. Circulation 1999;100:1992–2002.

372 P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

[7] Kim RJ, Wu E, Rafael A, et al. The use of contrast-enhanced magneticresonance imaging to identify reversible myocardial dysfunction.N Engl J Med 2000;343:1445–53.

[8] Wu E, Judd RM, Vargas JD, Klocke FJ, Bonow RO, Kim RJ.Visualisation of presence, location, and transmural extent of healed Q-wave and non-Q-wave myocardial infarction. Lancet 2001;357:21–8.

[9] Weinmann HJ, Brasch RC, Press WR, Wesbey GE. Characteristics ofgadolinium-DTPA complex: a potential NMR contrast agent. Am JRoentgenol 1984;142:619–24.

[10] Kim RJ, Choi KM, Judd RM. Assessment of myocardial viability bycontrast enhancement. In: Higgins CB, de Roos A, editors. Cardiovas-cularMRI andMRA. Philadelphia, PA: LippincottWilliams andWilkins;2003. p. 209–37.

[11] Kim RJ, Chen EL, Lima JA, Judd RM. Myocardial Gd-DTPA kineticsdetermine MRI contrast enhancement and reflect the extent and severityof myocardial injury after acute reperfused infarction. Circulation1996;94:3318–26.

[12] Flacke SJ, Fischer SE, Lorenz CH. Measurement of the gadopentetatedimeglumine partition coefficient in human myocardium in vivo:normal distribution and elevation in acute and chronic infarction.Radiology 2001;218:703–10.

[13] Rehwald WG, Fieno DS, Chen EL, Kim RJ, Judd RM. Myocardialmagnetic resonance imaging contrast agent concentrations afterreversible and irreversible ischemic injury. Circulation 2002;105: 224–9.

[14] Dumont CA, Monserrat L, Soler R, et al. Interpretation of electro-cardiographic abnormalities in hypertrophic cardiomyopathy withcardiac magnetic resonance. Eur Heart J 2006;27:1725–31.

[15] Choudhury L, Mahrholdt H, Wagner A, et al. Myocardial scarring inasymptomatic or mildly symptomatic patients with hypertrophiccardiomyopathy. J Am Coll Cardiol 2002;40:2156–64.

[16] Moon JC, McKenna WJ, McCrohon JA, Elliott PM, Smith GC,Pennell DJ. Toward clinical risk assessment in hypertrophic cardio-myopathy with gadolinium cardiovascular magnetic resonance. J AmColl Cardiol 2003;41:1561–7.

[17] Moon JC, Mogensen J, Elliott PM, et al. Myocardial late gadoliniumenhancement cardiovascular magnetic resonance in hypertrophic cardio-myopathy caused by mutations in troponin I. Heart 2005;91:1036–40.

[18] Harris KM, Spirito P, Maron MS, et al. Prevalence, clinical profile, andsignificance of left ventricular remodeling in the end-stage phase ofhypertrophic cardiomyopathy. Circulation 2006;114:216–25.

[19] Moon JC, Reed E, Sheppard MN, et al. The histologic basis of lategadolinium enhancement cardiovascular magnetic resonance inhypertrophic cardiomyopathy. J Am Coll Cardiol 2004;43:2260–4.

[20] Papavassiliu T, Schnabel P, Schroder M, Borggrefe M. CMR scarringin a patient with hypertrophic cardiomyopathy correlates well withhistological findings of fibrosis. Eur Heart J 2005;26:2395.

[21] Knaapen P, van Dockum WG, Bondarenko O, et al. Delayed contrastenhancement and perfusable tissue index in hypertrophic cardiomyo-pathy: comparison between cardiacMRI and PET. J NuclMed 2005;46:923–9.

[22] Basso C, Thiene G, Corrado D, Buja G, Melacini P, Nava A.Hypertrophic cardiomyopathy and sudden death in the young: pathologicevidence of myocardial ischemia. Hum Pathol 2000;31:988–98.

[23] Cecchi F, Olivotto I, Gistri R, Lorenzoni R, Chiriatti G, Camici PG.Coronary microvascular dysfunction and prognosis in hypertrophiccardiomyopathy. N Engl J Med 2003;349:1027–35.

[24] Spirito P, Bellone P, Harris KM, Bernabo P, Bruzzi P, Maron BJ.Magnitude of left ventricular hypertrophy and risk of sudden death inhypertrophic cardiomyopathy. N Engl J Med 2000;342:1778–85.

[25] Blackburn H. Classification of the electrocardiogram for populationstudies: Minnesota Code. J Electrocardiol 1969;2:305–10.

[26] Tarantini G, Cacciavillani L, Corbetti F, et al. Duration of ischemia is amajor determinant of transmurality and severe microvascular obstruc-tion after primary angioplasty: a study performed with contrast-enhanced magnetic resonance. J Am Coll Cardiol 2005;46:1229–35.

[27] Abdel-Aty H, Boyè P, Zagrosek A, et al. Diagnostic performance ofcardiovascular magnetic resonance in patients with suspected acute

myocarditis, comparison of different approaches. J Am Coll Cardiol2005;45:1815–22.

[28] Finn JP, Simonetti OP. Pulse sequence design in MRI. In: Edelman RR,Zlatkin MB, Hesselink JR, editors. Clinical magnetic resonanceimaging. 2nd Edition. Saunders; 1996. p. 145–76.

[29] Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardizedmyocardial segmentation and nomenclature for tomographic imagingof the heart. A statement for healthcare professionals from the CardiacImaging Committee of the Council on Clinical Cardiology of theAmerican Heart Association. Circulation 2002;105:539–42.

[30] Arai AE. Magnetic resonance first pass perfusion imaging. Top MagnReson Imaging 2000;11:383–98.

[31] Taylor AJ, Al-Saadi N, Abdel-Haty H, Schultz-Menger J, MessroghliDR, Friedrich MG. Detection of acutely impaired microvascularreperfusion after infarct angioplasty with magnetic resonance imaging.Circulation 2004;109:2080–5.

[32] Simonetti OP, Kim RJ, Fieno DS, et al. An improved MRI technique forthe visualization of myocardial infarction. Radiology 2001;218: 215–23.

[33] Spirito P, Bellone P. Natural history of hypertrophic cardiomyopathy.Br Heart J 1994;72:S10–2.

[34] Rickers C, Wilke NM, Jerosch-Herold M, et al. Utility of cardiacmagnetic resonance imaging in the diagnosis of hypertrophic cardiomy-opathy. Circulation 2005;112:855–61.

[35] Nagueh SF, Mahmarian JJ. Noninvasive cardiac imaging in patientswith hypertrophic cardiomyopathy. J Am Coll Cardiol 2006;48:2410–22.

[36] Consevage MW, Salada GC, Baylen BG, Ladda RL, Rogan PK. A newmissense mutation, Arg719Gln, in the beta-cardiac heavy chain myosingene of patients with familial hypertrophic cardiomyopathy. Hum MolGenet 1994;3:1025–6.

[37] Huang X, Song L, Ma AQ, et al. A malignant phenotype of hyper-trophic cardiomyopathy caused by Arg719Gln cardiac beta-myosinheavy-chain mutation in a Chinese family. Clin Chim Acta 2001;310:131–9.

[38] Van Driest SL, Ackerman MJ, Ommen SR, et al. Prevalence andseverity of “benign”mutations in the beta-myosin heavy chain, cardiactroponin T, and alpha-tropomyosin genes in hypertrophic cardiomy-opathy. Circulation 2002;106:3085–90.

[39] Garcia-Dorado D, Oliveras J. Myocardial oedema: a preventable causeof reperfusion injury? Cardiovasc Res 1993;27:1555–63.

[40] Garcia-Dorado D, Oliveras J, Gili J, et al. Analysis of myocardialoedema by magnetic resonance imaging early after coronary arteryocclusion with or without reperfusion. Cardiovasc Res 1993;27: 1462–9.

[41] Miller S, Schick F, Scheule AM, et al. Conventional high resolutionversus fast T2-weighted MR imaging of the heart: assessment ofreperfusion induced myocardial injury in an animal model. MagnReson Imaging 2000;18:1069–77.

[42] Nilsson JC, Nielsen G, Groenning BA, et al. Sustained postinfarctionmyocardial oedema in humans visualised by magnetic resonanceimaging. Heart 2001;85:639–42.

[43] Willerson JT, Scales F, Mukherjee A, et al. Abnormal myocardial fluidretention as an early manifestation of ischemic injury. Am J Pathol1977;87:159–88.

[44] Abdel-Aty H, Zagrosek A, Schltz-Menger J, et al. Delayedenhancement and T2-weighted cardiovascular magnetic resonanceimaging differentiate acute from chronic myocardial infarction.Circulation 2004;109:2411–6.

[45] Aso H, Takeda K, Ito T, Shiraishi T, Matsumura K, Nakagawa T.Assessment of myocardial fibrosis in cardiomyopathic hamsters withgadolinium-DTPA enhanced magnetic resonance imaging. Invest Radiol1998;33:22–32.

[46] Wigle ED, Rakowski H, Kimball BP, Williams WG. Hypertrophiccardiomyopathy. Clinical spectrum and treatment. Circulation1995;92:1680–92.

[47] Cannon III RO, Rosing DR, Maron BJ, et al. Myocardial ischemia inpatients with hypertrophic cardiomyopathy. Circulation 1985;71:234–43.

373P. Melacini et al. / International Journal of Cardiology 128 (2008) 364–373

[48] Krams R, Kofflard MJ, Duncker DJ, et al. Decreased coronary flowreserve in hypertrophic cardiomyopathy is related to remodeling of thecoronary microcirculation. Circulation 1998;97:230–3.

[49] Nienaber CA, Gambhir SS, Mody FV, et al. Regional myocardial bloodflow and glucose utilization in symptomatic patients with hypertrophiccardiomyopathy. Circulation 1993;87:1580–90.

[50] Camici P, Chiriatti G, Lorenzoni R, et al. Coronary vasodilation isimpaired in both hypertrophied and nonhypertrophied myocardium ofpatients with hypertrophic cardiomyopathy. J Am Coll Cardiol1991;17:879–86.

[51] Sipola P, Lauerma K, Husso-Saastamoinen M, et al. First-Pass MRimaging in the assessment of perfusion impairment in patients withhypertrophic cardiomyopathy and the Asp175Asn Mutation of theαTropomyosin gene. Radiology 2003;226:129–37.

[52] Petersen SE, Jerosch-Herold M, Hudsmith LE, et al. Evidence formicrovascular dysfunction in hypertrophic cardiomyopathy: newinsights from multiparametric magnetic resonance imaging. Circula-tion 2007;115:2418–25.