Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y VO L . 7 1 , N O . 1 5 , 2 0 1 8

ª 2 0 1 8 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N

P U B L I S H E D B Y E L S E V I E R

REVIEW TOPIC OF THE WEEK

Myocardial Interstitial Fibrosis inHeart FailureBiological and Translational Perspectives

Arantxa González, PHD,a,b Erik B. Schelbert, MD, MS,c Javier Díez, MD, PHD,a,b,d Javed Butler, MD, MPH, MBAe

ABSTRACT

ISS

FrobC

Sp

syl

Me

co

Ma

Myocardial interstitial fibrosis contributes to left ventricular dysfunction leading to the development of heart failure.

Basic research has provided abundant evidence for the cellular and molecular mechanisms behind this lesion and the

pathways by which it imparts a detrimental impact on cardiac function. Translation of this knowledge, however, to

improved diagnostics and therapeutics for patients with heart failure has not been as robust. This is partly related to the

paucity of biomarkers to accurately identify myocardial interstitial fibrosis and to the lack of personalized antifibrotic

strategies to treat it in an effective manner. This paper summarizes current knowledge of the mechanisms and detri-

mental consequences of myocardial interstitial fibrosis, discusses the potential of circulating and imaging biomarkers

available to recognize different phenotypes of this lesion and track their clinical evolution, and reviews the currently

available and potential future therapies that allow its individualized management in heart failure patients.

(J Am Coll Cardiol 2018;71:1696–706) © 2018 by the American College of Cardiology Foundation.

B eyond the cardiomyocyte-centric view ofheart failure (HF), it is now accepted that al-terations in the interstitial extracellular ma-

trix (ECM) and the coronary microcirculation alsoplay a major role in the development of pathologicalstructural myocardial remodeling that determinesthe evolution of HF. Myocardial interstitial fibrosis(MIF) is defined by the diffuse, disproportionateaccumulation of collagen in the myocardial intersti-tium. MIF contributes to left ventricular (LV)dysfunction in many disorders and predisposes pa-tients to develop HF with either preserved ejectionfraction (HFpEF) or reduced ejection fraction(HFrEF). Although ample research evidence explainsthe mechanisms of MIF, translation of this knowledgeto improved diagnostics and therapeutics for HF hasnot been fully realized. This review summarizes the

N 0735-1097/$36.00

m the aProgram of Cardiovascular Diseases, Centre for Applied Medical

IBERCV (Centro de Investigación Biomédica en Red Enfermedades Card

ain; cDepartment of Medicine, Heart and Vascular Institute, University o

vania; dDepartment of Cardiology and Cardiac Surgery, University of Nava

dicine, University of Mississippi, Jackson Mississippi. All authors have rep

ntents of this paper to disclose.

nuscript received November 11, 2017; revised manuscript received Febru

knowledge regarding mechanisms involved in thepathogenesis and consequences of MIF in HF, dis-cusses the potential of circulating and imaging bio-markers available to recognize different MIFphenotypes and track their evolution, and reviewsthe available and future therapies that may allowindividualized HF management targeting MIF.

HISTOCHEMICAL ASPECTS OF MYOCARDIAL

INTERSTITIAL FIBROSIS. Histologically MIF isdefined by the diffuse deposition of excess fibroustissue (i.e., collagen types I and III fibers) relative tothe mass of cardiomyocyte within the myocardialinterstitium. There are 2 principal types of MIF (1). Inthe reparative or replacement fibrosis, MIF replacessmall foci of dead cardiomyocytes, forming micro-scars (2,3) (Figure 1). In the reactive fibrosis,

https://doi.org/10.1016/j.jacc.2018.02.021

Research, University of Navarra, Pamplona, Spain;

iovasculares), Carlos III Institute of Health, Madrid,

f Pittsburgh School of Medicine, Pittsburgh, Penn-

rra Clinic, Pamplona, Spain; and the eDepartment of

orted that they have no relationships relevant to the

ary 12, 2018, accepted February 15, 2018.

ABBR EV I A T I ON S

AND ACRONYMS

CITP = collagen Type I

telopeptide

CMR = cardiac magnetic

resonance

CVF = collagen volume

fraction

ECM = extracellular matrix

ECV = extracellular volume

HF = heart failure

MIF = myocardial interstitial

J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8 González et al.A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6 Myocardial Interstitial Fibrosis in HF

1697

accumulation of fibrous tissue in the perivascularspace around intramural coronary arteries and theperimysium and endomysium changes the normallythin fibrous tissue layer around the cardiac musclebundles and individual cardiomyocytes into thickersheaths (2,3) (Figure 1). It is unclear whether these 2types represent truly different entities as they coexistin most patients.

Quantitatively, MIF is characterized by the in-crease in the percentage of total myocardial tissueoccupied by collagen fibers, denoted as collagenvolume fraction (CVF) and determined in myocardial

FIGURE 1 Myocardial Interstitial Fibrosis Shown on

Endomyocardial Biopsy

(Upper) Endomyocardial tissue from a patient with severe

aortic stenosis and heart failure showing myocardial interstitial

fibrosis. Sections were stained with picrosirius red, and collagen

fibers were identified in red. Collagen deposits were identified

as thin bands surrounding individual cardiomyocytes or

groups of cardiomyocytes (a), micro-scars (b), and large

strands diffusely localized across the interstitium (c) and around

intramyocardial vessels (d). (Magnification �40). (Lower)

Endomyocardial tissue from the same patient. Sections were

stained with specific monoclonal antibodies against collagen

types I and III, and fibers were identified in green and

yellow, respectively, shown on confocal microscopy.

(Magnification �60). Reprinted with permission from

Echegaray et al. (7).

fibrosis

MMP = matrix

metalloproteinase

PICP = carboxy-terminal

propeptide of procollagen

Type I

PIIINP = amino-terminal

propeptide of procollagen

Type III

samples by collagen-specific stains (4).Although MIF is patchy, the area of fibrosisincreases from the outer to the inner third ofthe ventricular free wall, probably due totransmural pressure gradient, wall stress,and coronary microcirculation alterationsthat are present in ischemic and nonischemiccardiac diseases (3,5).

There are also qualitative changes in thecollagen composition. In HF with hyperten-sive heart disease (6) or aortic stenosis (7), thecollagen type I VF (CIVF)-to-collagen type IIIVF (CIIIVF) ratio is abnormally increased dueto an excess of collagen type I fibers. How-ever, in ischemic cardiomyopathy, the CIVF-to-CIIIVF ratio is decreased due to an excess ofcollagen type III fibers (8), suggesting thatcollagen dysregulation may depend on theunderlying clinical scenario. The physico-chemical properties of collagen fibers alsovary. The insolubility and stiffness of fibers

depend on the degree of intermolecular covalentlinkage, or cross-linking, among their constitutive fi-brils (9). Increased cross-linking is reported in pa-tients with HF and increased LV stiffness (10–12).These observations suggest that both the extent offibrosis and its composition are relevant. Addition-ally, it is likely that these alterations are dynamic,which are reflected in variable MIF phenotypes overtime.

MECHANISMS OF MYOCARDIAL INTERSTITIAL

FIBROSIS. MIF represents a final common lesionfollowing a variety of injuries caused by an intrinsiccardiac disease or by systemic factors activated in thecontext of extracardiac comorbidities such as arterialhypertension, diabetes mellitus, and chronic kidneydisease (13). The process of MIF develops in severalphases (Central Illustration).

Tr igger ing st imul i . Cardiomyocyte death is oftenthe triggering event responsible for the initiation offibrosis in reparative MIF. In reactive MIF, variedstimuli (e.g., pressure overload, ischemia, or meta-bolic injury) may trigger the fibrotic response in theabsence of cell death (2,3,13). Several cell types areimplicated in the fibrotic response, either directly byproducing fibrous tissue (myofibroblasts) or indirectlyby secreting profibrotic mediators (macrophages,mast cells, lymphocytes, cardiomyocytes, andvascular cells) (2,3,13).Generat ion of myofibroblasts and profibrot i cact ivat ion . Beyond resident fibroblasts, circulatingand resident fibroblast progenitor cells, includingfibrocytes, epicardial epithelial cells undergoing

CENTRAL ILLUSTRATION Developmental Phases in Myocardial Interstitial Fibrosis and Their Impact on the Heartand Related Clinical Consequences

Cardiomyocyte Deathand/or

Mechanical, Ischemic, or Metabolic Injury

Biomarkers to identifyindividual phenotypes

Direct activation

involvement of othercell types

Profibrotic factors

EC

FB

Proliferation and differentiationto a secretory phenotype

MyoFB

Fibrogenicsecretome

Targets to treatindividually

and effectively

Diffuse deposition of highlycross-linked collagen fibers

Myocardial Interstitial Fibrosis

Alterations of myocardialphysical properties

Alterations of cellular interactionsand tissue architecture

Alterations of ECMreservoir function

Left ventriculardysfunction

Propensity toarrhythmias

Reductionof perfusion

ADVERSE CLINICAL OUTCOMES

González, A. et al. J Am Coll Cardiol. 2018;71(15):1696–706.

CM ¼ cardiomyocyte; EC ¼ endothelial cell; ECM ¼ extracellular matrix; FB ¼ fibroblast; IC ¼ inflammatory cell; MyoFB ¼ myofibroblast.

González et al. J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8

Myocardial Interstitial Fibrosis in HF A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6

1698

TABLE 1 Major Extracellular Alterations of Collagen Types I and

III Turnover Involved in Myocardial Interstitial Fibrosis

Increased availability of myofibroblast-secreted procollagen types Iand III precursors and procollagen and collagen processingenzymes

Increased bone morphogenetic protein-1 or procollagen C-terminalproteinase and procollagen N-terminal proteinase-mediatedconversion of procollagen precursors into microfibril-formingmature collagen

Increased spontaneous microfibril assembly to form fibrils

Increased lysyl oxidase- and non-lysyl oxidase-mediated cross-linkingof fibrils to form stiff collagen type I and III fibers that are highlyresistant to degradation

Unchanged or decreased collagen fiber degradation by matrixmetalloproteinases and cathepsins

J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8 González et al.A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6 Myocardial Interstitial Fibrosis in HF

1699

epithelial-to-mesenchymal transition, and endothe-lial cells undergoing endothelial-to-mesenchymaltrans-differentiation, are also sources of myofibro-blasts (14). A number of physical and chemical fac-tors present in the injured myocardium stimulatemyofibroblast generation (14). Whereas inflammatorycytokines, chemokines, and reactive oxygen speciesmay be more important in reparative fibrosis, me-chanical stress, the renin-angiotensin-aldosteronesystem (RAAS), and fibrogenic growth factors suchas transforming growth factor-b and connective tis-sue growth factor appear to be involved in bothreparative and reactive fibrosis (14). When activatedby these factors, the fibroblast proliferates and dif-ferentiates into a nonproliferative secretory pheno-type, the myofibroblast, which combinesultrastructural and phenotypic characteristics ofsmooth muscle cells acquired through formation ofcontractile stress fibers and the expression of a-smooth muscle actin, with an extensive endoplasmicreticulum, a feature of synthetically active fibro-blasts (14). The myofibroblast fibrogenic responseinvolves the action of sequence-specific DNA bindingtranscription factors, as well as epigenetic changesand alterations in regulatory noncoding RNAs.Particularly in the heart, small noncoding microRNAshave been critically involved in the control of myo-fibroblast processes such as ECM synthesis andcytokine secretion, as well as in the regulation offibroblast proliferation (15).

The myofibroblast’s secretome consists of pro-collagen types I and III, molecules requisite toregulate extracellular fibrillary collagen types I andIII turnover, and autocrine and paracrine factorsthat simulate their metabolic activity, perpetuatingfibrogenesis (16). The myofibroblast also secretes inexcess other nonstructural ECM macromoleculessuch as matricellular proteins that, by coordinatingand integrating cell–cell and cell–matrix in-teractions play an important role in the regulationof fibrosis (16).

Extrace l lu lar fibrogenes is . The excess of collagentype I and III fibers, characteristic of MIF, results fromthe combination of several extracellular alterationsleading to the predominance of fiber formation anddeposition over its degradation and removal (Table 1)(17). Oxidation of specific collagen lysine by enzymesof the lysyl oxidase family is critical to ensure thatformation of fibers exceeds degradation (18). In fact,through oxidation, lysine acquires reactive aldehydesthat condense with other lysines and build covalentbonds between polypeptide chains of adjacentcollagen fibrils (i.e., cross-links that determine both

the formation and deposition of the final fiber and itsresistance to degradation by matrix metal-loproteinase [MMPs]) (19).

PATHOPHYSIOLOGICAL AND CLINICAL ASPECTS OF

MYOCARDIAL INTERSTITIAL FIBROSIS. The accu-mulation of fibrotic tissue is maladaptive and leads toleft ventricular (LV) dysfunction, arrhythmia,impaired myocardial oxygen availability, and pooroutcomes (Central Illustration).LV dysfunct ion . Classically, MIF has been relatedto diastolic dysfunction. However, endomyocardialbiopsies demonstrate CVFs that are similar inHFpEF than in nonischemic HFrEF (10,20). Thus,the possibility exists that the effect of excesscollagen on LV function is modulated by changes incollagen quality and spatial organization. In thisregard, increased collagen cross-linking is associatedwith LV stiffness and diastolic dysfunction in HF(10–12). In addition, a greater association existsbetween collagen type I fibers than type III fibersand increased stiffness in HFpEF (7). The realign-ment of collagen and cardiomyocytes, which mayimpair transmission of cardiomyocyte force andmyocardial contractility, is associated with systolicdysfunction in HF (21).Other a l terat ions . MIF may promote atrial andventricular arrhythmia by creating a vulnerable sub-strate for re-entrant activity (22). Additionally, myo-fibroblasts can modulate electrical activity ofcardiomyocytes through direct physical interactionswith these cells or through secreted paracrine factors,contributing to the ventricular arrhythmogenic ac-tivity of MIF (22). Ventricular arrhythmias are relatedto the extent of MIF independently of other con-founders, including LV dysfunction (23).

Deposition of fibrotic tissue increases oxygendiffusion distance leading to the impairment of oxy-gen supply to the cardiomyocytes (24). In this regard,perivascular fibrosis has been shown to be inverselycorrelated with coronary flow reserve in HF (25).

FIGURE 2 Association Between Circulating Biomarkers and

Myocardial Interstitial Fibrosis

6C

5

4

3

2

1

00 1 2 3 4

r = –0.460P = 0.005

5Serum CITP:MMP-1 Ratio

Myo

card

ial C

olla

gen

Cros

s-Li

nkin

g

B40

30

20

10

00 4 8 12

r = 0.784P < 0.001

20Serum PIIINP (g//L)

Myo

card

ial C

olla

gen

Type

III (

%)

A20

15

10

5

00 10050 150 200

r = 0.774P < 0.001

250Serum PICP (µg/L)

Myo

card

ial C

olla

gen

Type

I (%

)

Continued on the next column

González et al. J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8

Myocardial Interstitial Fibrosis in HF A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6

1700

Cl in i ca l outcomes. MIF also adversely influencesHF outcome. The extent of fibrotic deposits is asso-ciated with the degree of LV functional improvementand mortality after aortic valve replacement (26) andwith death and cardiac events in HFrEF (20). Theextent of fibrosis also predicts effectiveness of long-term HF therapy (e.g., improvement with beta-blockers is likely to occur in patients with lessfibrosis) (27). Besides the quantity of collagen depo-sition, its quality also influences outcome. Increasedcollagen type I cross-linking is associated with hos-pitalization risk in HF (28), and the coincidence ofincreased cross-linking with high collagen type Ideposition is associated with both hospitalization forHF and mortality (29).

NONINVASIVE DIAGNOSIS OF MYOCARDIAL INTER-

STITIAL FIBROSIS. Because MIF influences the clin-ical evolution of HF patients, integrating itsevaluation into the assessment and management ofthese patients may be warranted. Although endo-myocardial biopsy is the gold standard method fordiagnosing MIF, due to its several limitations, alter-native noninvasive methods (e.g., biomarkers) areneeded for routine practice.Ci r cu la t ing b iomarkers . Among the many pro-posed circulating biomarkers of MIF, only 3 demon-strate an association with histologically provenMIF (30).

Serum carboxy-terminal propeptide of procollagen

type I. Serum carboxy-terminal propeptide ofprocollagen type I (PICP) is generated during theextracellular conversion of procollagen type I intocollagen type I by the enzyme bone morphogeneticprotein-1 or procollagen carboxy-terminal proteinase(31). A net release from the heart into the circulationhas been reported in HF (32), suggesting a cardiacorigin for systemic PICP in this syndrome. SerumPICP concentrations correlate with CVF and CIVF(11,32,33) in HF (Figure 2). Serum PICP levels and

FIGURE 2 Continued

(A) Association of serum PICP with the myocardial deposition of

collagen type I fibers in HF patients. Adapted with permission

from López et al. (11). (B) Association of serum PIIINP with the

myocardial deposition of collagen type III fibers in HF patients.

Adapted with permission from Klappacher et al. (39). (C)

Association of the serum CITP-to-MMP-1 ratio with myocardial

collagen cross-linking in HF patients. Adapted with permission

from López et al. (28). CITP ¼ carboxy-terminal telopeptide of

collagen type I; HF ¼ heart failure; MMP ¼ matrix metal-

loproteinase; PICP ¼ carboxy-terminal propeptide of pro-

collagen type I; PIIINP ¼ amino-terminal propeptide of

procollagen type III.

J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8 González et al.A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6 Myocardial Interstitial Fibrosis in HF

1701

CVF change in parallel in response to torsemide (33)and spironolactone (34) in HF. Serum PICP isassociated with the severity of HFrEF (35), andwith mortality in HFpEF (36) and HFrEF (37).Of interest, the serum PICP-to-serum PIIINP ratio isrelated to malignant ventricular arrhythmogenesisin HF (38).Serum amino-terminal propeptide of procollagen

type III. Most serum (PIIINP) is generated duringthe extracellular conversion of procollagen type IIIto collagen type III by the enzyme procollagen amino-terminal proteinase (31). Serum PIIINP concentrationcorrelates with CIIIVF in HF (Figure 2) (39). Inaddition, the reduction in the extent of CVF in HFpatients treated with spironolactone is accompaniedby reductions in serum PIIINP (34). Serum PIIINP isassociated with severity (40) and outcomes in HF ofdifferent causes regardless of EF (36,40).Serum collagen type I telopeptide-to-serum matrix

metalloproteinase-1 ratio. As collagen cross-linkingdetermines the resistance of the collagen fiber toMMP degradation, the higher the cross-linking ofcollagen type I fibers, the lower the cleavage of thepeptide collagen type I telopeptide (CITP) by theenzyme MMP-1. Thus, the serum CITP-to-serumMMP-1 ratio is inversely correlated with myocardialcollagen cross-linking (28) (Figure 2). The CITP-to-MMP-1 ratio is independently associated with therisk of HF hospitalization (28). The combination oflow CITP-to-MMP-1 ratio and high PICP identifies HFpatients with highest risk (29).

Testing for circulating biomarkers of MIF presentsseveral limitations (30). They are not cardiac-specific,and changes in their concentrations may representintegrated abnormalities of the cardiovascularcollagen and/or influence of comorbidities affectingcollagen metabolism.Imag ing biomarkers . Recent studies have shownprogress in noninvasive assessment of MIF by usingimaging modalities, namely cardiovascular magneticresonance (CMR).Late gadolinium enhancement. Late gadoliniumenhancement (LGE) assessed by CMR detects focalmyocardial fibrosis, which often is prognosticallyrelevant, especially if it involves the midwall of theinterventricular septum (41). Abundant studies affirmits adverse consequences regarding mortality, ven-tricular arrhythmia, and hospitalization for HF.Currently, the CMR GUIDEd (Cardiac Magnetic Reso-nance GUIDEd Management of Mild-moderate LeftVentricular Systolic Dysfunction; NCT01918215) studyrandomized patients with focal fibrosis on LGEand mild to moderate systolic dysfunction toreceive implantable cardioverter-defibrillator or loop

recorder implantation, examining incident ventricu-lar arrhythmia. LGE requires spatial heterogeneity todetect focal fibrosis (42). LGE cannot quantify ordetect diffuse fibrosis, which limits its ability to trackantifibrotic response to intervention.Extracellular volume. CMR-assessed extracellularvolume (ECV) quantifies the interstitial space, and ifthere is no myocardial edema or amyloidosis, thenECV is an excellent measurement of diffuse MIF (42).In addition, whereas MIF refers to elevated collagenconcentration in myocardium rather than its totalquantity, ECV can be multiplied by LV mass tocompute the total myocardial fibrosis burden(Figure 3) (43,44).

Due to partial volume effects, ECV measurementsare limited mostly to the mid myocardium becausepixels straddling the subendocardial or subepicardialborder may be contaminated by blood pool orepicardial fat, given the limited spatial resolution ofthe images. This is important because there may be asubendocardial predilection for MIF in ischemiccardiomyopathy (3) and aortic stenosis (5). AlthoughMIF varies spatially in severity and homogeneity andmatching biopsy samples to the same area on ECVmeasures is difficult, ECV has been validatedrobustly relative to myocardial CVF (45). Of interest,one study leveraging whole-heart ECV measure-ments prior to heart transplantation using whole-heart histologic quantification after organ extrac-tion yielded the best correlations (46). Despite thesechallenges, ECV is reproducible (47–50), is associatedwith outcomes (20,43,44,51) (Figure 4), and sur-passes LVEF in risk stratification (44). Of note, ECVappears to associate with outcomes more so thanLGE (44).

MIF can be also be estimated by ECV measured bycardiovascular computed tomography (CCT) scan. Infact, CCT-measured ECV correlated as well as CMR-measured ECV with histological CVF (52).Native T1. Native pre-contrast CMR T1 measurementscan also assess MIF, and although the validation dataare not as robust as for ECV, it does exhibitassociation with outcomes (53).

SPECIFIC THERAPY FOR MYOCARDIAL INTERSTITIAL

FIBROSIS. MIF represents a likely therapeutic targetfor cardiac protection and improved outcomes in HF.

Available and emerging antimyocardial interstitialfibrosis treatments. Histomorphological data fromsome clinical studies with drugs used as standardtherapy in cardiac diseases support the notionthat MIF can be targeted. For instance, treatmentof hypertensive patients with the angiotensin-converting enzyme inhibitor lisinopril was able to

FIGURE 3 Extracellular Volume Assessment With Cardiac Magnetic Resonance Imaging

D25

20

15

10

5

018 20 22 24 26 28 30 32 34 36

No Nonischemic LGE (n = 936) Nonischemic LGE (n = 236)

38 40 42 44 46 48Myocardial ECV (%)

Freq

uenc

y (%

)

Extracellular volume (ECV) maps generated from T1 maps can display normal myocardium (A) as well as severe diffuse myocardial fibrosis (B, black arrows) that is not

detectablewith LGE imaging (C), and there was overlap in the distributions of ECV in thosewith andwithout evident LGE (D). Semiautomated quantitative LGE thresholding

techniques (C) using 2 common methods failed to identify the severe diffuse myocardial fibrosis present in null myocardium (row B). In contrast, focal myocardial fibrosis

manifests on late gadolinium enhancement (LGE) imaging (red arrows) and ECVmapping (white arrows). The upward shift of ECV distributions for thosewith focal LGEwas

small comparedwith the spectrum of ECV (D). Midmyocardial ECVwasmeasured to avoid contamination from partial volume effects from limited spatial resolution and/or

misregistration errors, depicted by the green-colored pixels along the blood pool and myocardium interface. Adapted with permission from Schelbert et al. (44).

González et al. J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8

Myocardial Interstitial Fibrosis in HF A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6

1702

FIGURE 4 Association Between Extracellular Volume and Prognosis

A0.6

0.4

0.2

0.00 1 2 3

141188 80

Logrank = 57p < 0.0001

42111 12440529 205 112329 38241277 113 77182 275161 23 1132 6

197Numbers at Risk

5713107717

ECV < 25%25% ≤ ECV < 30%30% ≤ ECV < 35%35% ≤ ECV < 40%40% ≤ ECV 810 7 58 2

Years After CMR

Prob

abili

ty o

f HHF

B0.6

0.4

0.2

0.00 1 2 3

142188 80

Logrank = 60p < 0.0001

43111 13444537 207 112336 38252288 116 79190 275866 27 1338 8

197Numbers at Risk

5713107717

ECV < 25%25% ≤ ECV < 30%30% ≤ ECV < 35%35% ≤ ECV < 40%40% ≤ ECV 911 7 59 2

Years After CMR

Prob

abili

ty o

f Dea

th

C0.6

0.4

0.2

0.00 1 2 3

141188 80

Logrank = 92p < 0.0001

42111 12440529 205 112329 38241277 113 77182 275161 23 1132 6

197Numbers at Risk

5713107717

ECV < 25%25% ≤ ECV < 30%30% ≤ ECV < 35%35% ≤ ECV < 40%40% ≤ ECV 810 7 58 2

Years After CMR

Prob

abili

ty o

fHH

F or

Dea

th

40% ≤ ECV 35% ≤ ECV < 40% 30% ≤ ECV < 35% 25% ≤ ECV < 30% ECV < 25%

Association between increasing degrees of ECV in noninfarcted myocardium with increased risks of adverse events. (A) First HHF; (B) all-cause

death; and (C) HHF or death in HF patients. Adapted with permission from Schelbert et al. (44). CMR ¼ cardiovascular magnetic resonance;

ECV ¼ extracellular volume; HHF ¼ hospitalization of heart failure.

J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8 González et al.A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6 Myocardial Interstitial Fibrosis in HF

1703

TABLE 2 Examples of Emerging Strategies for the Development

of Targeted Therapies for Myocardial Interstitial Fibrosis

At the level of the triggering stimuli

� Antagonism of relaxin 2 receptor (e.g., serelaxin)� Partial agonism of the adenosine A1-receptor (e.g.,

capadenoson)� Blockade of monocyte chemoattractant protein-1 CCR2

receptor (e.g., cenicriviroc)

At the level of myofibroblast generation and profibrotic activation

� Genetic modulation (e.g. , histone deacetylases inhibitors, anti-noncoding microRNA compounds, peroxisome proliferator-activated receptor agonists)

� Paracrine modulation (e.g., recombinant neuregulin-1b, stemcell therapy, antifibrotic growth factors [e.g., hepatocytegrowth factor, insulin-like growth factor-1, basic fibroblastgrowth factor], blockade of pro-fibrotic factors [e.g.,connective tissue growth factor, interleukin-10])

At the level of extracellular fibrogenesis

� Modulation of BMP-1-mediated fibrillary collagen formation(e.g., BMP-1 inhibitors)

� Modulation of lysyl oxidase-mediated cross-linking (e.g., lysyloxidase-like 2 inhibitors, anti-noncoding RNA compounds)

� Stimulation of cardiac lymphangiogenesis (e.g., vascularendothelial growth factor-C)

BMP ¼ bone morphogenetic protein.

TABLE 3 Proposed C

Myocardial Remodelin

Post-InterventionResults

Predominant MIF regre

Predominant cellmass regression

Mixed MIF and cellmass regression

*LV mass � ECV. †LV mass

ECV ¼ extracellular volu

González et al. J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8

Myocardial Interstitial Fibrosis in HF A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6

1704

reduce CVF, and this effect was associated withimproved LV diastolic dysfunction (54). Similarly,treatment of hypertensive patients with theangiotensin-II receptor blocker losartan diminishedCVF, and this was accompanied by a decrease inLV stiffness (55). In addition, the mineralocorticoidreceptor antagonist spironolactone ameliorated LVdiastolic dysfunction and reduced LV stiffness inassociation with reduction of MIF in HF patients (34).On the other hand, in patients with either HFpEFor HFrEF, administration of torsemide in addition tostandard HF therapy was associated with reductionsin myocardial expression of active lysyl oxidase(LOX), the degree of collagen cross-linking, CVF andCIVF, normalization of LV stiffness, and improvementof function in 80% of the patients, without LVenlargement (56).

ardiac Magnetic Resonance-Based Assessment of the Reversal of

g Following an Intervention

Change in TotalInterstitial

Compartment*Change in Total

Cell Compartment† Change in ECV

ssion Decrease Minimal change Decrease

Minimal change Decrease Increase

Decrease Decrease Decrease or increasedepending onpredominance

� [1 � ECV].

me; LV ¼ left ventricular; MIF ¼ myocardial interstitial fibrosis.

Most of these clinical studies are small or havenot been maintained for long enough periods. Inaddition, although some conventional therapies,such as RAAS inhibitors, reduce MIF in humans,MIF persists in patients with HF, even when theytreated with these agents, indicating a need todevelop novel and effective anti-MIF therapies.Although many emerging therapies targeting MIFare promising in preclinical models (57,58) (Table 2),many clinical trials of novel antifibrotic drugs havefailed. Therefore, it is important to rationalize drugdiscovery by using meaningful in vitro models todiscard irrelevant molecules in terms of efficacy andpharmacokinetic and toxicological profiles at anearly stage. In this regard, recently an algorithm forselection of new antifibrotic factors to be furthertested as potential therapeutic targets has beenproposed (59).

Target ing myocard ia l interst i t ia l fibros is inc l in i ca l t r ia ls . The effects of MIF on cardiacfunctions and outcomes render its prevention orits regression possible therapeutic targets andendpoints for treatment efficacy in HF trials. ECVand circulating biomarkers appear suitable for MIFassessment in trials. Indeed, investigators havealready leveraged ECV for this purpose in a recenttrial (60). However, because reversal of myocardialremodeling following an intervention may includevariable regression of cell mass, ECM mass, orboth (61), it is important to measure both theinterstitial compartment (LV mass � ECV) and cellcompartment (LV mass � [1 � ECV]) and thendefine ECV-assessed regression of MIF as proposedin Table 3. Clearly, this can be a matter for futureresearch.

On the other hand, whereas data from large ran-domized clinical studies support a beneficial role formineralocorticoid receptor antagonists on collagenturnover, as assessed by reductions in circulatingPCIP and PIIINP, with a positive impact in post-infarct LV geometric remodeling and clinicaloutcome (62), recent data from another large ran-domized clinical trial failed to demonstrate any as-sociation between amelioration of the severity ofthe disease and changes in circulating PIIINP inpatients with HFpEF treated with either valsartan orLCZ696 (63).

Toward a persona l i zed therapy of myocard ia linterst i t i a l fibros is . The use of panels combiningcirculating and/or imaging biomarkers of MIF intranslational research studies may integratedifferent levels of information, overcome methodo-logical limitations, and contribute to a better

J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8 González et al.A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6 Myocardial Interstitial Fibrosis in HF

1705

phenotyping of patients, with a view to personal-izing HF therapy. From this perspective, and as anexample, it can be hypothesized that therapies ableto regress MIF as a result of reductions in both cross-linking and deposition of collagen type I fibers (e.g.,torsemide) (33,56) may be specifically beneficial forHF patients presenting with the combination ofincreased (LV mass � ECV) product, low-serumCITP-to-MMP-1 ratio and high-serum PICP. This hy-pothesis remains to be tested in an adequatelydesigned trial.

CONCLUSIONS AND PERSPECTIVES

MIF is critical to the evolution of pathological struc-tural myocardial remodeling and the development ofHF in patients with cardiac diseases, underscoring theneed to integrate it into diagnosis and management ofpatients (Central Illustration). Translational researchincorporating quantitative and qualitative measure-ment of MIF is warranted to yield new approachesaimed to prevent its detrimental clinical impact. Dueto the nature of the studies in which the biomarkers

of MIF have been investigated, the available infor-mation provides a snapshot of the evolutionary stateof the cardiac disease rather than a comprehensiveview on the progression of MIF. Therefore, theinvestigation of biomarkers to track MIF needs to be amajor focus in the coming years. Many distinctrelated cellular and molecular factors activatingthrough diverse pathways contribute to the initiationand progression of MIF. Therefore, a multiprongedapproach may be needed to understand and inter-vene upon with an integrated antifibrotic strategythat simultaneously targets important mediators andeffectors of the fibrogenic process to treat MIF.

ADDRESS FOR CORRESPONDENCE: Dr. Javier Díez,Program of Cardiovascular Diseases, Centre forApplied Medical Research, University of Navarra, Av.Pío XII 55, 31008 Pamplona, Spain. E-mail: jadimar@unav.es. OR Dr. Javed Butler, University of Mis-sissippi Medical Center, Department of Medicine(L650), 2500 N. State Street, Jackson, Mississippi39216. E-mail: jbutler4@umc.edu.

RE F E RENCE S

1. Weber KT, Pick R, Jalil JE, Janicki JS, Carroll EP.Patterns of myocardial fibrosis. J Mol Cell Cardiol1989;21 Suppl 5:121–31.

2. Weber KT, Sun Y, Guarda E. Structural remod-eling in hypertensive heart disease and the role ofhormones. Hypertension 1994;23;6 Pt 2:869–77.

3. Beltrami CA, Finato N, Rocco M, et al. Structuralbasis of end-stage failure in ischemic cardiomy-opathy in humans. Circulation 1994;89:151–63.

4. Hoyt RH, Ericksen E, Collins SM, Skorton DJ.Computer-assisted quantitation of myocardialfibrosis in histologic sections. Arch Pathol Lab Med1984;108:280–3.

5. Treibel TA, López B, González A, et al. Reap-praising myocardial fibrosis in severe aortic ste-nosis: an invasive and non-invasive study in 133patients. Eur Heart J 2018;39:699–709.

6. López B, González A, Querejeta R, Larman M,Rábago G, Díez J. Association of cardiotrophin-1with myocardial fibrosis in hypertensive patientswith heart failure. Hypertension 2014;63:483–9.

7. Echegaray K, Andreu I, Lazkano A, et al. Role ofmyocardial collagen in severe aortic stenosis withpreserved ejection fraction and symptoms of heartfailure. Rev Esp Cardiol (English ed) 2017;70:832–40.

8. Mukherjee D, Sen S. Alteration of collagenphenotypes in ischemic cardiomyopathy. J ClinInvest 1991;88:1141–6.

9. Shoulders MD, Raines RT. Collagen structureand stability. Annu Rev Biochem 2009;78:929–58.

10. Kasner M, Westermann D, López B, et al.Diastolic tissue Doppler indexes correlate with thedegree of collagen expression and cross-linking in

heart failure and normal ejection fraction. J AmColl Cardiol 2011;57:977–85.

11. López B, Querejeta R, González A, et al.Collagen cross-linking but not collagen amountassociates with elevated filling pressures in hy-pertensive patients with stage C heart failure:potential role of lysyl oxidase. Hypertension 2012;60:677–83.

12. Zile MR, Baicu CF, Ikonomidis J, et al.Myocardial stiffness in patients with heart failureand a preserved ejection fraction: contributions ofcollagen and titin. Circulation 2015;131:1247–59.

13. Kong P, Christia P, Frangogiannis NG. Thepathogenesis of cardiac fibrosis. Cell Mol Life Sci2014;71:549–74.

14. Weber KT, Sun Y, Bhattacharya SK, Ahokas RA,Gerling IC. Myofibroblast-mediated mechanismsof pathological remodelling of the heart. Nat RevCardiol 2013;10:15–26.

15. Piccoli MT, Bär C, Thum T. Non-coding RNAs asmodulators of the cardiac fibroblast phenotype.J Mol Cell Cardiol 2016;92:75–81.

16. Bomb R, Heckle MR, Sun Y, et al. Myofibro-blast secretome and its auto-paracrine signaling.Expert Rev Cardiovasc Ther 2016;14:591–8.

17. Bishop JE, Laurent GJ. Collagen turnover andits regulation in the normal and hypertrophyingheart. Eur Heart J 1995;16 Suppl C:38–44.

18. Trackman PC. Lysyl oxidase isoforms andpotential therapeutic opportunities for fibrosisand cancer. Expert Opin Ther Targets 2016;20:935–45.

19. van der Slot-Verhoven AJ, van Dura EA,Attema J, et al. The type of collagen cross-link

determines the reversibility of experimentalskin fibrosis. Biochim Biophys Acta 2005;1740:60–7.

20. Aoki T, Fukumoto Y, Sugimura K, et al. Prog-nostic impact of myocardial interstitial fibrosis innon-ischemic heart failure. Comparison betweenpreserved and reduced ejection fraction heartfailure. Circ J 2011;75:2605–13.

21. Villari B, Campbell SE, Hess OM, et al. Influ-ence of collagen network on left ventricular sys-tolic and diastolic function in aortic valve disease.J Am Coll Cardiol 1993;22:1477–84.

22. Nguyen MN, Kiriazis H, Gao XM, Du XJ. Cardiacfibrosis and arrhythmogenesis. Compr Physiol2017;7:1009–49.

23. McLenachan JM, Dargie HJ. Ventricular ar-rhythmias in hypertensive left ventricular hyper-trophy. Relationship to coronary artery disease,left ventricular dysfunction, and myocardialfibrosis. Am J Hypertens 1990;3:735–40.

24. Sabbah HN, Sharov VG, Lesch M, Goldstein S.Progression of heart failure: a role for interstitialfibrosis. Mol Cell Biochem 1995;147:29–34.

25. Dai Z, Aoki T, Fukumoto Y, Shimokawa H.Coronary perivascular fibrosis is associated withimpairment of coronary blood flow in patientswith non-ischemic heart failure. J Cardiol 2012;60:416–21.

26. Azevedo CF, Nigri M, Higuchi ML, et al.Prognostic significance of myocardial fibrosisquantification by histopathology and magneticresonance imaging in patients with severe aorticvalve disease. J Am Coll Cardiol 2010;56:278–87.

González et al. J A C C V O L . 7 1 , N O . 1 5 , 2 0 1 8

Myocardial Interstitial Fibrosis in HF A P R I L 1 7 , 2 0 1 8 : 1 6 9 6 – 7 0 6

1706

27. Yamada T, Fukunami M, Ohmori M, et al.Which subgroup of patients with dilated cardio-myopathy would benefit from long-term beta-blocker therapy? A histologic viewpoint. J Am CollCardiol 1993;21:628–33.

28. López B, Ravassa S, González A, et al.Myocardial collagen cross-linking is associatedwith heart failure hospitalization in patients withhypertensive heart failure. J Am Coll Cardiol 2016;67:251–60.

29. Ravassa S, López B, Querejeta R, et al. Phe-notyping of myocardial fibrosis in hypertensivepatients with heart failure. Influence on clinicaloutcome. J Hypertens 2017;35:853–61.

30. López B, González A, Ravassa S, et al. Circulatingbiomarkers of myocardial fibrosis: the need for areappraisal. J Am Coll Cardiol 2015;65:2449–56.

31. Prockop DJ, Kivirikko KI. Collagens: molecularbiology, diseases, and potentials for therapy. AnnuRev Biochem 1995;64:403–34.

32. Querejeta R, López B, González A, et al.Increased collagen type I synthesis in patients withheart failure of hypertensive origin: relation tomyocardial fibrosis. Circulation 2004;110:1263–8.

33. López B, Querejeta R, González A, Sánchez E,Larman M, Díez J. Effects of loop diuretics onmyocardial fibrosis and collagen type I turnover inchronic heart failure. J Am Coll Cardiol 2004;43:2028–35.

34. Izawa H, Murohara T, Nagata K, et al. Miner-alocorticoid receptor antagonism ameliorates leftventricular diastolic dysfunction and myocardialfibrosis in mildly symptomatic patients with idio-pathic dilated cardiomyopathy: a pilot study. Cir-culation 2005;112:2940–5.

35. Löfsjögård J, Kahan T, Díez J, et al. Biomarkersof collagen type I metabolism are related to B-type natriuretic peptide, left ventricular size, anddiastolic function in heart failure. J CardiovascMed (Hagerstown) 2014;15:463–9.

36. Krum H, Elsik M, Schneider HG, et al. Relationof peripheral collagen markers to death and hos-pitalization in patients with heart failure and pre-served ejection fraction: results of the I-PRESERVEcollagen substudy. Circ Heart Fail 2011;4:561–8.

37. Löfsjögård J, Kahan T, Díez J, et al. Usefulnessof collagen carboxy-terminal propeptide andtelopeptide to predict disturbances of long-termmortality in patients $60 years with heart fail-ure and reduced ejection fraction. Am J Cardiol2017;119:2042–8.

38. Flevari P, Theodorakis G, Leftheriotis D, et al.Serum markers of deranged myocardial collagenturnover: their relation to malignant ventriculararrhythmias in cardioverter-defibrillator recipientswith heart failure. Am Heart J 2012;164:530–7.

39. Klappacher G, Franzen P, Haab D, et al.Measuring extracellular matrix turnover in theserum of patients with idiopathic or ischemicdilated cardiomyopathy and impact on diagnosisand prognosis. Am J Cardiol 1995;75:913–8.

40. Zannad F, Alla F, Dousset B, Perez A, Pitt B.Limitation of excessive extracellular matrix

turnover may contribute to survival benefit ofspironolactone therapy in patients with congestiveheart failure: insights from the randomizedaldactone evaluation study (RALES). Rales In-vestigators. Circulation 2000;102:2700–6.

41. Gulati A, Jabbour A, Ismail TF, et al. Associa-tion of fibrosis with mortality and sudden cardiacdeath in patients with nonischemic dilated car-diomyopathy. JAMA 2013;309:896–908.

42. Schelbert EB, Sabbah HN, Butler J,Gheorghiade M. Employing extracellular volumecardiovascular magnetic resonance measures ofmyocardial fibrosis to foster novel therapeutics.Circ Cardiovasc Imaging 2017;10:e005619.

43. Chin CW, Everett RJ, Kwiecinski J, et al.Myocardial fibrosis and cardiac decompensation inaortic stenosis. J Am Coll Cardiol Img 2017;10:1320–33.

44. Schelbert EB, Piehler KM, Zareba KM, et al.Myocardial fibrosis quantified by extracellularvolume is associated with subsequent hospitali-zation for heart failure, death, or both across thespectrum of ejection fraction and heart failurestage. J Am Heart Assoc 2015;4:e002613.

45. Diao K, Yang Z, Xu H, et al. Histologic valida-tion of myocardial fibrosis measured by T1 map-ping: a systematic review and meta-analysis.J Cardiovasc Magn Reson 2016;18:92.

46. Miller CA, Naish J, Bishop P, et al. Compre-hensive validation of cardiovascular magneticresonance techniques for the assessment ofmyocardial extracellular volume. Circ CardiovascImaging 2013;6:373–83.

47. Fontana M, White SK, Banypersad SM, et al.Comparison of T1 mapping techniques for ECVquantification. Histological validation and repro-ducibility of ShMOLLI versus multibreath-hold T1quantification equilibrium contrast CMR.J Cardiovasc Magn Reson 2012;14:88.

48. Schelbert EB, Testa SM, Meier CG, et al.Myocardial extravascular extracellular volumefraction measurement by gadolinium cardiovas-cular magnetic resonance in humans: slow infu-sion versus bolus. J Cardiovasc Magn Reson 2011;13:16.

49. Chin CW, Semple S, Malley T, et al. Optimi-zation and comparison of myocardial T1 tech-niques at 3T in patients with aortic stenosis. EurHeart J Cardiovasc Imaging 2014;15:556–65.

50. Singh A, Horsfield MA, Bekele S, Khan J,Greiser A, McCann GP. Myocardial T1 and extracel-lular volume fraction measurement in asymptomaticpatients with aortic stenosis: reproducibility andcomparison with age-matched controls. Eur Heart JCardiovasc Imaging 2015;16:763–70.

51. Schelbert EB, Fridman Y, Wong TC, et al.Temporal relation between myocardial fibrosisand heart failure with preserved ejectionfraction: association with baseline diseaseseverity and subsequent outcome JAMA. Cardiol2017;2:1–12.

52. Treibel TA, Fontana M, Steeden JA, et al.Automatic quantification of the myocardial

extracellular volume by cardiac computed to-mography: synthetic ECV by CCT. J CardiovascComput Tomogr 2017;11:221–6.

53. Puntmann VO, Carr-White G, Jabbour A, et al.T1-mapping and outcome in nonischemic cardio-myopathy: All-cause mortality and heart failure.J Am Coll Cardiol Img 2016;9:40–50.

54. BrillaCG,FunckRC,RuppH. Lisinopril-mediatedregression of myocardial fibrosis in patients withhypertensive heart disease. Circulation 2000;102:1388–93.

55. Díez J, Querejeta R, López B, et al. Losartan-dependent regression of myocardial fibrosis isassociated with reduction of left ventricularchamber stiffness in hypertensive patients. Circu-lation 2002;105:2512–7.

56. López B, Querejeta R, González A,Beaumont J, Larman M, Díez J. Impact of treat-ment on myocardial lysyl oxidase expression andcollagen cross-linking in patients with heart fail-ure. Hypertension 2009;53:236–42.

57. Gourdie RG, Dimmeler S, Kohl P. Novel ther-apeutic strategies targeting fibroblasts andfibrosis in heart disease. Nat Rev Drug Discov2016;15:620–38.

58. Roubille F, Busseuil D, Merlet N, Kritikou EA,Rhéaume E, Tardif JC. Investigational drugs tar-geting cardiac fibrosis. Expert Rev Cardiovasc Ther2014;12:111–25.

59. Gyöngyösi M, Winkler J, Ramos I, et al.Myocardial fibrosis: biomedical research frombench to bedside. Eur J Heart Fail 2017;19:177–91.

60. Heydari B, Abdullah S, Pottala JV, et al. Effectof omega-3 acid ethyl esters on left ventricularremodeling after acute myocardial infarction: theOMEGA-REMODEL randomized clinical trial. Cir-culation 2016;134:378–91.

61. Treibel TA, Kozor R, Schofield R, et al.Reverse myocardial remodeling in aortic stenosis:Cellular hypertrophy and extracellular volumebut not focal fibrosis regress following aorticvalve replacement. J Am Coll Cardiol 2018;71:860–71.

62. Zannad F, Radauceanu A. Effect of MRblockade on collagen formation and cardiovasculardisease with a specific emphasis on heart failure.Heart Fail Rev 2005;10:71–8.

63. Zile MR, Jhund PS, Baicu CF, et al., for theProspective Comparison of ARNI with ARB onmanagement of heart failure with preservedejection fraction (PARAMOUNT) Investigators.Plasma biomarkers reflecting profibrotic processesin heart failure with a preserved ejection fraction:data from the Prospective Comparison of ARNIWith ARB on Management of Heart Failure WithPreserved Ejection Fraction study. Circ Heart Fail2016;9:e002551.

KEY WORDS diagnosis, fibrosis, heartfailure, mechanisms, myocardium,therapeutics