Restrictive cardiomyopathy (RCM) is a heart muscle disease characterized by abnormal diastolic function with restrictive filling and normal ventricular size, wall thickness, and ejection fraction. Differently from hypertrophic, dilatated and right ventricle arrhythmogenic cardiomyopathy, where definition is based on morphology, RCM is defined based on physiology. However, under this common denominator, a wide spectrum of diseases are enclosed, with different causes, natural history, prognosis, and management. Furthermore, a restrictive hemodynamic profile can appear during the natural course of dilated and hypertrophic cardiomyopathy (HCM), being predictor of poor prognosis.

Among pediatric cardiomyopathies, RCM is the least common phenotype, representing only 2.5–3% of cardiomyopathies diagnosed during childhood (1). Unfortunately, compared to other pediatric cardiomyopathies, RCM is less amenable to medical or surgical treatment with higher mortality rates: 63% within 3 years of diagnosis and 75% within 6 years of diagnosis (2). As a consequence, rate for heart transplantation is relatively higher. Accordingly, within the Pediatric Heart Transplant Study Group, patients affected by RCM represents 12% of whole group of patients with cardiomyopathy undergoing heart transplantation.

Purpose of the present review is to summarize the causes of pediatric RCMs, their pathophysiology, clinical presentation and management.

According to European Society of Cardiology position statement, RCM is defined as a myocardial disease characterized by impaired ventricular filling and normal/reduced diastolic volumes in the presence of (near) normal ejection fraction and myocardial thickness (3). Decreased active relaxation and increased parietal stiffness cause pressure within the ventricles to rise precipitously during diastolic filling, with only small increases in volumes. Although systolic function was classically said to be preserved in RCM, contractility is rarely normal, indeed. Furthermore, restrictive physiology can also occur in other scenarios, namely end stage HCM and dilated cardiomyopathy (DCM). However, it is suggested to consider these two entities apart. Restrictive cardiomyopathy is the least common among pediatric cardiomyopathies, accounting for only 2–5% of all cases, although the incidence may be higher in tropical areas of Africa, Asia, and South America, where endomyocardial fibrosis (EMF) is endemic (4). Its prognosis is poor with a 2-year survival <50%, being heart transplantation the only effective treatment (5, 6). Restrictive cardiomyopathy has been described in children of all ages, with mean age at diagnosis ranging from 6 to 11 years old in recent studies.

Restrictive cardiomyopathies represent a heterogeneous group of cardiomyopathies which recognize several etiologies, including inherited and acquired causes. The term of idiopathic RCM is probably no longer appropriate in a large group of patients: in fact, genetics has identified mutations in various genes, above all sarcomeric genes. Figure 1 summarizes the main causes of pediatric RCMs.

Restrictive cardiomyopathy recognizes a unique hemodynamic profile, independently from the specific cause at the basis of diastolic dysfunction. In RCM impairment of diastole can be related both to the abnormal myocardial relaxation (i.e., the active actin-myosin cross-bridge detachment) and to the increased myocardial stiffness due to the myocardial cells (e.g., titin) and the interstitial matrix (fibrosis) alterations, determining elevated left and right-sided filling pressure. Although left ventricular ejection fraction is typically preserved, systolic contractility is often impaired as showed by tissue Doppler imaging and speckle tracking. Systo-diastolic dysfunction leads to reduced stroke volume. In the protodiastole—despite delayed active relaxation—there is an unusually early rapid filling of the ventricles, due to high atrial pressures, halted by incompliant ventricular walls from the end of the first third of diastole onward—reflecting myocardial stiffness. This results in a prominent “y” descent on the atrial pressure curves and, sometimes, in the square root or dip and plateau sign on ventricular pressure curves, consisting in an early decrease in ventricular diastolic pressure followed by a rapid rise to a plateau phase. During the following atrial contraction, the stiff ventricles are unable to easily accept additional blood volume, and thus the contribution from atrial contraction is often minimal. Differently from constrictive pericarditis (CP), in patients with RCM there is not enhanced ventricular interdependence, with concordant left and right ventricular pressures during the respiratory cycle and parallel changes in their pressure curve areas. Moreover, atria progressively enlarge, due to high intracavitary pressures and thin and distensible walls, predisposing to atrial arrhythmias and thromboembolic episodes. A relevant proportion of patients develop pulmonary hypertension (PH) with elevated pulmonary vascular resistances, unresponsive to vasodilator testing, precluding them form orthotopic heart transplantation (47).

The clinical presentation of RCM can be highly variable in children population, ranging from asymptomatic to right and/or left heart failure with PH.

Biventricular systolic function is typically normal until advanced stages of the disease, leading to heart failure with preserved ejection fraction (HFpEF). Consequently, clinical presentation is characterized by dyspnoea, poor appetite, ascites, peripheral edema, and hepatomegaly. However, while in adults with RCM symptoms and signs of heart failure are easy to detect, clinical evaluation in children is challenging because of the non-specific findings, resulting in some delay in correct diagnosis. Besides, children with RCM may have a history of frequent respiratory infections. Progressive atrial enlargement can also lead to atrial arrhythmias such as atrial fibrillation and thromboembolic complications, with mitral and tricuspid functional regurgitation frequently associated, due to anulus dilatation. Wolff Parkinson-White syndrome with supraventricular tachycardia has also been reported.

Sudden death occurs in ≈25% of pediatric RCM patients, with an annual mortality rate reported of 7%, being cardiac ischemia, arrhythmias, and thromboembolic events the main responsible. Various risk factors for sudden death in pediatric RCM have been identified in previous studies: cardiomegaly, thromboembolism, raised pulmonary vascular resistance, pulmonary venous congestion, syncope, chest pain, left atrial size, PR and QRS duration. Albeit inconsistently sometimes, the main limitation of these studies is their retrospective nature and the intrinsic bias associated. Rivenes et al. evaluated a cohort of 18 pediatric patients with RCM who had sudden, unanticipated cardiac arrests, identifying chest pain, and syncope as risk factors for sudden death (48). Histopathologic evidence for ischemia was found in the majority of patients who died and the evidence of ischemia at Holter monitoring (i.e., ST depression) predicted death within several months. They proposed lethal ventricular arrhythmias as cause of death, showing examples of ventricular tachycardia/fibrillation recorded during resuscitation attempts (48). Complementary, Walsh et al. in a 16 pediatric RCMs cohort reported five sudden cardiac events, with three patients having complete heart block. In one of them, ST-segment elevation was documented before the onset of complete heart block, suggesting an underlying ischemic process as trigger of the bradyarrhythmia. Older age at presentation, longer PR interval and QRS duration were associated with sudden cardiac events (49).

Approximately 98% of RCM patients have an abnormal electrocardiogram (ECG) (50). The most common abnormalities are right and/or left atrial enlargement (91% of patients) (Figure 4). ST-T segment and T waves abnormalities are also frequently present and may be most evident at higher heart rates. In a small cohort of 12 children affected by RCM, Hayashi et al. found that obliquely elevated ST-T segments and notched or biphasic T waves were the most frequent ventricular repolarization abnormalities (67% of patients). Besides, the criteria for biventricular hypertrophy based on QRS voltage were achieved in seven patients (although three of them actually had some degree of hypertrophy at echocardiogram) (51). ST-T depression is usually mild and non-specific, but in some cases can be pronounced, mimicking a left main or proximal left anterior descending artery occlusion or a multivessel coronary disease. This was associated with high risk of sudden cardiac death (48) and Selvaganesh et al. hypothesized that marked ST depression can be caused by high end-diastolic pressure, impairing perfusion in the subendocardial region or stretching the myocardium and activating stretch sensitive channels (52). Conduction abnormalities can also be seen, as well as right or left ventricular hypertrophy signs (48). Furthermore, serial ECG-Holter monitoring is useful for the evaluation of rhythm disturbances and ST segment analysis.

Chest X-ray is abnormal in nearly 90% of cases and usually show cardiomegaly and pulmonary venous congestion (51).

Echocardiography plays a key role in RCM diagnosis. Marked biatrial enlargement with a normal or slightly decreased LV ejection fraction are widely considered as pathognomonic findings (Figure 5). Regardless of etiology, the considerable elevation in filling pressures in patients with RCM is reflected in abnormal mitral inflow and tissue Doppler variables. A short (<140 ms) mitral deceleration time, increased pulsed wave (PW) Doppler mitral E/A ratio (>2.5) and E/e′ >15 are markers of significantly elevated left filling pressures (53). Although IVRT (isovolumic relaxation time) is prolonged when myocardial relaxation is impaired, due to delayed LV pressure falling during the isovolumic relaxation, a short LV IVRT of <50 ms is frequently detected in RCM, as consequence of the high LA pressure (53). Typically plethoric inferior vena cava and hepatic veins are seen and, with inspiration, diastolic flow reversal in the hepatic veins is documented due to the inability of a non-compliant right ventricle to accommodate the increased venous return. Echocardiography is helpful in differential diagnosis between RCM and CP. Although both conditions shares E-wave predominance and short deceleration time, respirophasic shifting of the interventricular septum, caused by exaggerated interventricular dependence, is characteristic of CP. Accordingly respiratory flow variations consisting in increasing >25% in mitral inflow during expiration and >40% in tricuspid inflow after inspiration are absent in RCM but frequently noted in CP (54). However, among all echocardiographic parameters, the most useful to distinguish RCM from CP are those of tissue Doppler imaging. In fact a normal tissue Doppler e′ velocity (>8 cm/s) indicating normal LV relaxation virtually excludes RCM. In patients affected by CP, where diastolic dysfunction is due to pericardial constraint, e′ is normal or even increased since the longitudinal movement of the myocardium is enhanced because of constricted radial motion. Furthermore, in patients with pericardial constriction, lateral mitral annular e′ is usually lower than e′ from the medial annulus (so called “annulus reversus”) (55). This finding, absent in RCM, reflects the tethering of the lateral mitral annulus to the adjacent fibrotic and scarred pericardium. In this regard, speckle tracking may add even higher diagnostic accuracy in differentiating constriction from restriction: in fact while in RCM both radial and longitudinal strains are reduced due to a diseased myocardium, in CP reduction mainly involve circumferential strain, reflecting the subepicardial tethering offered by the fibrous pericardium (55, 56). The myocardial performance index (MPI) or Tei index may provide further information of both LV diastole and systole, with normal value of 0.33 ± 0.02 from 3 to 18 years old. It is defined as the sum of the isovolumic contraction and relaxation times divided by the ejection time, and it can be calculated from PW Doppler at the mitral and aortic valve simultaneously or using TDI at mitral valve annulus (57). Moreover, lateral a′ velocity (cut-off ≤ 0.042 m/s) and pulmonary vein A wave duration (cut-off ≥156 m/s) both have sensitivity and specificity ≥80% for LVEDP ≥ 20 mmHg measured on right heart catheterization (58). Ancillary methods such as pulmonary regurgitant flow velocity, color M-mode flow propagation, and myocardial velocity gradient have been proposed for differential diagnosis between RCM and CP, however data in pediatric age are lacking. Finally, the performance and interpretation of diastolic measurements in children are challenging, given the higher heart rates, potential need for sedation, together with conflicting and limited data on the relationships among the diastolic variables and the degree of dysfunction (1). Indeed, in the assessment of diastolic dysfunction among 175 children with cardiomyopathy, the percentage of normal diastolic variables in children with overt cardiac dysfunction was high, with discordance between e′ and left atrium (LA) volume criteria. Patients with RCM were best identified with mitral E deceleration time, which was found to be abnormal in 75% of patients (59). In another study, Sasaki et al. described the LA area indexed to body surface area as the most useful measurement to differentiate between healthy children and RCM patients (60).

Cardiac magnetic resonance (CMR) offers a better spatial resolution than echocardiography, providing detailed information about anatomic structures, ventricular function, perfusion, and tissue characterization (61). For instance, T2^*-weighted CMR is the diagnostic gold standard to detect and quantify myocardial iron content in IOC (62). Late gadolinium enhancement (LGE) can show peculiar patterns of replacement and reactive fibrosis, which can direct the diagnosis to specific subtypes of RCM. In AFD, LGE is typically localized in the infero-lateral mid-basal wall of left ventricle, and because of the distinctive fatty nature of the intracellular deposits, native T1 mapping has typically a low value, in contrast to most of other infiltrative or storage cardiomyopathies (63). In cardiac sarcoidosis LGE distribution is patchy, often with multifocal distribution, not following a coronary artery topography, sparing the endomyocardial layer, and involving mainly the basal and lateral LV walls (64). When performing CMR, it must be considered that the young patients must hold still in the scanner and follow the instructions to minimize motion artifacts during image acquisition. Whereas, this is possible in older children (more than 6–8 years of age), it requires sedation and anesthesia for younger patients, with different possible strategies, often depending on institutional preference and availability of resources such as pediatric anesthesiologists (65).

Whole-body scintigraphy or SPECT with bone-seeking tracers [(99mTc)-labeled bisphosphonate compounds: pyrophosphate (PYP); 3,3-diphosphono-1,2-propanodicarboxylic acid (DPD), and hydroxydiphosphonate (HDP)] can reveal amyloid deposits (especially in ATTR subtype) in the heart, as well as PET with [¹⁸F] FDG can detect inflammatory cells in some pathological processes such as cardiac sarcoidosis. However, due to the rarity of the aforementioned diseases and due to the concern for radiation exposure in childhood, nuclear imaging has very limited applications in the diagnostic work-up of pediatric RCMs.

Cardiac catheterization is usually not necessary for RCM diagnosis, but it can be useful to distinguish between restrictive and constrictive physiology and to determine the severity of the diastolic dysfunction by directly measuring the filling pressures of both ventricles (Figure 6). In RCM left ventricular end-diastolic pressures are usually higher than right end-diastolic pressure, whereas are equal or very nearly equal in CP. Furthermore, specular discordance between RV and LV peak systolic pressures during inspiration are typical of CP, with an increase in RV pressure occurring during peak inspiration, when LV pressure is lowest. Cardiac catheterization can also reveal the presence of PH, detect the presence of elevated pulmonary vascular resistance, evaluate the cardiac index and test for pulmonary vasculature reactivity.

Plasma levels of natriuretic peptides can be helpful in the diagnostic pathway of RCM, especially in the differential diagnosis with CP. A study of 49 adults (20 with RCM and 29 with CP) showed that median plasma NT-proBNP was 1,775 (208–7,500) pg/ml in those with RCM vs. 124 (68–718) pg/ml in those with CP (P = 0.001) (66). Specific etiologies may require additional laboratory exams such as angiotensin converting enzyme dosage in sarcoidosis, complete blood count to establish eosinophilia in HESs, serum iron concentrations, total iron-binding capacity, and ferritin levels in hemochromatosis, alpha-galactosidase activity, and lyso-Gb3 levels in AFD, immunoglobulin free light chain testing, and serum and urine immunofixation in AL amyloidosis.

The endomyocardial biopsy can be valuable in doubtful cases, when non-invasive tests are inconclusive. Unfortunately, in idiopathic RCM it often demonstrates non-specific findings such as myocyte hypertrophy, interstitial and/or endocardial fibrosis. Furthermore, periprocedural risks in fragile affected children should be considered (50).

Restrictive cardiomyopathy should be distinguished from CP, since the two diseases have different treatments and outcomes. In some cases—namely, radiation induced cardiac disease—restriction and constriction may coexist in the same patient, making final diagnosis even more challenging. In Table 1 the main instrumental features of each condition are summarized.

Current medical therapy for RCM is primarily supportive and is in large part limited to diuretics in patients with signs and symptoms of systemic or pulmonary venous congestion. The International Society for Heart and Lung Transplantation (ISHLT) guidelines for the management of pediatric heart failure published in 2014 recommend in class I the diuretic therapy to establish a clinically euvolemic state, with a close monitoring of renal function and blood pressure during initiation and up-titration (67). Diuretic therapy reduces signs of systemic congestion, with beneficial effect on symptoms as dyspnea, fatigue, peripheral edema, and cough. However, excessive diuresis should be avoided because these patients are sensitive to alterations in preload. Furthermore, diuretics can mask an underlying PH or increased biventricular filling pressures, therefore this aspect must be considered before performing a hemodynamics invasive assessment (particularly prior to candidacy to heart transplantation). Restrictive cardiomyopathy has a unique pathophysiology: small ventricular cavity dimensions and rapid increase of filling pressures significantly compromise stroke volume (according to pressure-volume loop). Therefore, cardiac output is strongly influenced by heart rate, which must be kept at relatively high values to guarantee adequate systemic perfusion. For these reasons β-blockers and calcium channel blockers are not currently recommended in pediatric RCM, unless for a different indication (67). Angiotensin converting enzyme inhibitors and angiotensin receptor blockers may be considered if coexisting systemic arterial hypertension is present (class IIb recommendation). Similarly, digoxin (unless for rate control of atrial arrhythmias), intravenous inotropes (such as dopamine, dobutamine, and epinephrine), and pulmonary vasodilators (prostaglandins and endothelin receptor antagonists to treat secondary PH) are generally not recommended (Class III).

Although atrial thrombosis has been linked to atrial fibrillation, abnormal hemodynamics, and to a possible hypercoagulable state in adults with RCM, dedicated studies in pediatric populations are lacking (68). The incidence of intracardiac thrombus in the reviewed literature ranges from 0 to 42%, with rates of embolism between 12 and 33%. The risk of embolism appears to be much greater in children with RCM than to DCM; therefore, antithrombotic or anticoagulation therapy should be considered at the time of diagnosis (47, 69). However, there are no studies comparing the effectiveness of antiplatelet agents, vitamin K antagonists or enoxaparin in preventing embolism in children with RCM (70).

Conduction system disease recognize different etiologies, such as ischemic injury of the atrio-ventricular node and His-Purkinje system, mechanical stretching due to atrial and ventricular dilation or genetically determined mechanisms (71–73). In addition, prolonged PR interval and wide QRS complex at ECG were associated with acute cardiac event (49). The 2014 ISHLT guidelines recommend in Class I permanent PM for advanced second- or third-degree atrioventricular block associated with ventricular dysfunction. On the basis of available data, the presence of conduction system disease should trigger increased surveillance through baseline ECG and ECG Holter monitoring with ST-segment analysis together with a routine screening for research of clinical (chest pain or syncope) and instrumental ischemia (ST-segment variations). Some Authors suggest in the presence of PR prolongation, QRS widening, and left bundle-branch block particular attention and to consider prophylactic pacing, eventually as part of an implantable-cardioverter defibrillators (ICD) system (48). However, it should be noticed that there are no studies documenting the efficacy of defibrillator systems in large pediatric cohorts, and case to case evaluation, assessing the individual risk factors (including the specific etiology) should be carried out, in order to avoid inappropriate and detrimental therapies (49). At moment, without unanimous criteria, ICD should be considered in the subset of pediatric RCM patients with evidence of ischemia and ventricular arrhythmia, where also β-blockers may be beneficial (48). In a single center experience of pediatric patients with RCM, 40% of them with an ICD or PM, device therapies were relatively rare and inappropriate therapies were exceedingly rare (6).

These gaps in evidence are partly to ascribe to the low prevalence of RCM and its poor prognosis, with half patients dying or being referred to cardiac transplantation within 3 years from the diagnosis (74). Medical treatment has not shown any significant long term benefit and cardiac transplantation is the only effective therapy with survival rates between 70 and 60% at 5 and 10 year, respectively, in children listed for heart transplant (75). Orthotopic heart transplantation is preferred to heart-lung or heterotopic heart transplant and has a better survival rate than the other two options (76). In 2012 Singh et al. analyzed 1,436 children <18 years of age with a diagnosis of cardiomyopathy listed for heart transplant in the United States between 2004 and 2010, of which 167 with RCM. In adjusted analysis, children with non-DCM (83%) had a higher risk of wait-list mortality only if supported by a ventilator at listing. Post-transplant 30-days and at 1-year survival were similar between children with dilated and non-DCM (p = 0.17) (77). In another cohort of children with RCM from the American Pediatric Cardiomyopathy Registry database, about two thirds of children had a pure RCM phenotype, and the rest had a mixed RCM/HCM phenotype. Rate of survival at 5 year was 20 and 28%, respectively, but patients with pure RMC phenotype underwent heart transplantation more frequently (58 vs. 30%) (5).

The majority of deaths in children awaiting heart transplantation is due to progressive heart and multi-organ failure. Hence, in this context, mechanical circulatory support (MCS), as bridge to transplantation or candidacy, may be a precious weapon. At the moment the experience with MCS in children is quite limited, and this is particularly true for RCM. The most used pediatric long-term support device, is the pneumatically driven, pulsatile EXCOR^® Berlin Heart, with a variety of pump sizes, covering almost all pediatric patients, and the only long-term device for neonates and infants approved in Europe and USA. Few case series of successful bridging to cardiac transplantation with the Berlin EXCOR^® left ventricular assist device (LVAD) have been described (78, 79). Conventional LV apical cannulation of a non-compliant left ventricle often results in insufficient drainage and poor pump performance, with residual high left atrial pressures and consequent pulmonary congestion. Left atrial cannulation (such as in EXCOR^® Berlin Heart) is therefore an interesting option in patient with small ventricular cavity, preserved systolic function, and enlarged atria (80). A recent review of the American registry of EXCOR^® Berlin Heart implantation in pediatric patients affected by RCM showed a survival rate of 50%, which is significantly less than that of the overall EXCOR ^® pediatric population (75%) (81). Primary causes of death included stroke, infection, acidosis, multisystem organ failure, and bleeding. It is of note that these patients tended to be sicker than the whole population (INTERMACS class 1 and with ECMO support) and this can explain at least partly, the worst outcome. In summary, long-term MCS implantation is a high-risk procedure that can be considered in advanced stages of the disease as bridge to transplantation or to candidacy by improving hemodynamics, with the reduction of post-capillary PH. Further studies are needed to determine the best timing for the procedure and the best anticoagulant strategy to reduce the risk of thromboembolic events that are the main cause of adverse outcome in these patients.

Considering all these major concerns about the poor efficacy of medical treatment, the identification of the right time to list for heart transplantation a patient affected by RCM during the clinical follow-up becomes both challenging and crucial. Although some clinical and instrumental factors such as pulmonary congestion at diagnosis, severe left atrial dilation, or increased ventricular wall thickness have been identified as potential predictors of poor prognosis, there are no established criteria for listing and the decision is often dependent on individual experienced centers (82). A multi-modal instrumental approach is essential, particularly based on a regular assessment of right heart hemodynamics with or without use of inotropic agents and vasodilators, to detect at the proper time the development of irreversible PH. Some pediatric institutions consider the development of PH an indication for listing, regardless of heart failure symptoms. Accordingly, right heart catheterization is mandatory at first evaluation, since up to 50% of patients have PH at diagnosis (47, 83). The flow-chart illustrated in Figure 7 summarizes the experienced approach for pediatric RCM, developed in our tertiary center, with a dedicated program for pediatric cardiomyopathies and pediatric heart transplantation.

Pediatric RCM is a rare disorder, due to a large heterogeneous group of causes. As a result of its poor prognosis, RCM contributes disproportionately to mortality in children with cardiomyopathy, being heart transplantation the only effective treatment. Therefore, early referral to a third-level cardiomyopathy center is warranted for careful observation, to avoid the development of irreversible PH and to avoid a delay in listing for heart transplantation when indicated.

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.