Differentiating constrictive pericarditis and restrictive cardiomyopathy

INTRODUCTION — 

Constrictive pericarditis (CP) and restrictive cardiomyopathy (RCM) are both causes of heart failure with normal (or near normal) systolic function and abnormal ventricular filling with similar clinical and hemodynamic features. Differentiating between these two conditions is critical given the marked differences in their causes and management. In some patients, the correct diagnosis may be readily suggested from the history or routine diagnostic testing. In others, however, distinguishing these conditions may be challenging and, in rare cases, require biopsy or even surgical exploration.

●CP is a condition in which an inelastic pericardium restricts ventricular filling in mid and late diastole. As a result, the majority of ventricular filling occurs rapidly in early diastole and the ventricular volume does not significantly increase after the end of the early filling period. The clinical presentation, diagnosis, and management of CP are discussed in detail separately. (See "Constrictive pericarditis: Clinical features and causes" and "Constrictive pericarditis: Diagnostic evaluation" and "Constrictive pericarditis: Management and prognosis".)

●RCM is characterized by nondilated, severely noncompliant ventricle(s), resulting in severe diastolic dysfunction and restrictive filling that produces hemodynamic changes resembling those of CP. The clinical manifestations, diagnosis, and management of RCM are discussed in detail separately. (See "Restrictive cardiomyopathies".)

The distinction between CP and RCM will be reviewed here.

CLINICAL MANIFESTATIONS — 

In the evaluation of a patient with suspected CP or RCM, the history can provide important clues to the cause. The physical examination may also be helpful; while most patients with CP and RCM present with similar manifestations of heart failure, subtle differences in physical findings may suggest a particular diagnosis.

History — In the evaluation of CP versus RCM, the history is most valuable in identifying a systemic disorder which can predispose to either CP or RCM:

●Causes of constrictive pericarditis – A prior history of pericarditis, trauma, cardiac surgery, or a systemic disease that affects the pericardium (eg, tuberculosis, connective tissue disease, malignancy) favors a diagnosis of CP. (See "Etiology of pericardial disease" and "Constrictive pericarditis: Clinical features and causes", section on 'Constrictive pericarditis'.)

●Causes of restrictive cardiomyopathy – A history of an infiltrative disease that may involve the heart muscle (eg, amyloidosis, sarcoidosis) favors a diagnosis of RCM. (See "Clinical manifestations and diagnosis of cardiac sarcoidosis" and "Cardiac amyloidosis: Epidemiology, clinical manifestations, and diagnosis" and "Restrictive cardiomyopathies", section on 'Clinical features'.)

●Causes of either or both conditions – Prior thoracic radiation treatment can result in either CP, RCM, or a condition with features of both CP and RCM. (See "Cardiotoxicity of radiation therapy for breast cancer and other malignancies" and "Constrictive pericarditis: Diagnostic evaluation".)

Amyloidosis is a common cause of RCM (see "Restrictive cardiomyopathies", section on 'Etiology'). In contrast, amyloidosis is a rare cause of CP [1].

Physical examination — There are similarities and differences in the physical examination findings for patients with CP and RCM.

●Findings common to both conditions – Most patients with CP or RCM display elevated jugular venous pressure (JVP) on physical examination (figure 1). From the JVP waveform alone, it is not generally possible to distinguish between CP, RCM, tricuspid regurgitation with an enlarged compliant right atrium, or right heart failure (eg, due to right ventricular [RV] infarction or pulmonary hypertension). The contour of the jugular venous pulse in all these conditions is dominated by a deep, steep Y descent. (See "Examination of jugular venous waveforms", section on 'Abnormalities of the y-descent'.)

Additional physical examination findings that occur in patients with CP or RCM include:

•Kussmaul sign (the lack of an inspiratory decline in JVP)

•Peripheral edema

•Ascites and hepatomegaly

•Pleural effusions

Pulsus paradoxus (drop in systolic blood pressure greater than 10 mmHg during inspiration) is most commonly associated with cardiac tamponade. Pulsus paradoxus occurs in a minority of patients with CP but is found more commonly in the subgroup with effusive-constrictive pericarditis. Pulsus paradoxus is an infrequent finding in patients with RCM.

●Findings that differ

•Constrictive pericarditis - Approximately 50 percent of patients with CP may present with a pericardial knock (an accentuated heart sound occurring slightly earlier than a third heart sound, which may be audible and rarely is palpable [2]. This finding is not expected in RCM.

•Restrictive cardiomyopathy - An audible third heart sound (S3) is frequently present in individuals with RCM because of the abrupt cessation of rapid ventricular filling; this is not usually present in CP. However, it can be difficult to distinguish an S3 from a pericardial knock in some patients. (See "Restrictive cardiomyopathies", section on 'Clinical features' and "Constrictive pericarditis: Clinical features and causes", section on 'Key signs of constrictive pericarditis'.)

INITIAL TESTS — 

Patients suspected of having either CP or RCM based on history and physical examination should undergo initial evaluation with electrocardiography (ECG), B-type natriuretic peptide (BNP) testing, chest radiography, and echocardiography (table 1 and algorithm 1).

Electrocardiogram — The ECG may suggest CP or RCM, although the ECG findings are not diagnostic.

●Constrictive pericarditis – While low voltage and isolated repolarization abnormalities can occur in CP or RCM, repolarization abnormalities are more common in CP.

●Restrictive cardiomyopathy – Bundle branch blocks, ventricular hypertrophy, pathologic Q waves, or impaired atrioventricular conduction strongly favor RCM.

●Both conditions – Atrial fibrillation is common in the late stages of both diseases. (See "Constrictive pericarditis: Diagnostic evaluation", section on 'Electrocardiogram' and "Restrictive cardiomyopathies", section on 'Electrocardiogram'.)

Natriuretic peptides — Given the available evidence and physiologic rationale for BNP testing, plasma N-terminal pro-BNP (NT-proBNP) or BNP should be measured early in the evaluation of patients with suspected CP or RCM. Natriuretic peptide levels are generally higher in patients with RCM than in those with CP [3]. A normal BNP or NT-proBNP level essentially excludes RCM.

BNP is released in response to ventricular wall stress, which is a product of intracardiac cavity size and filling pressure [4]. In RCM, there is severe diastolic dysfunction and wall stress is increased. In contrast, in CP, ventricular function is generally normal and cavity size is limited by the thickened pericardium. These physiologic differences suggest that measurement of plasma BNP or NT-proBNP might aid in distinguishing between these two disorders.

The utility of natriuretic peptides in differentiating CP from RCM has been evaluated in several small studies [3,5-8]. While mean BNP and NT-proBNP levels in patients with RCM are generally higher than in patients with CP, diagnostic cut-points have not been identified given wide variation in levels among patients with CP and among patients with RCM.

●BNP – A BNP level >400 pg/mL is commonly identified in patients with RCM, and a level <300 pg/mL is commonly seen in patients with CP [3]. However, there is considerable overlap between BNP levels for RCM and CP.

●NT-proBNP – An NT-proBNP level >2000 pg/mL is commonly seen in patients with RCM, and a level <700 pg/mL is commonly seen in patients with CP [3,8]. However, there is wide overlap between NT-proBNP levels for RCM and CP.

Of note, in patients with kidney disease, natriuretic peptide levels are less helpful in distinguishing CP from RCM [9]. Decreased glomerular filtration rate causes elevation in BNP levels and even greater elevation in NT-proBNP levels, such that the differences in natriuretic peptide levels between RCM and CP are reduced. (See "Natriuretic peptide measurement in heart failure", section on 'Renal failure'.)

Chest radiography — The chest radiograph is generally not diagnostic for CP or RCM but may reveal findings favoring one of these diagnoses:

●Constrictive pericarditis – Calcification of the pericardium strongly suggests CP (image 1), but it can also be seen in other conditions, such as asbestosis, which may or may not be associated with CP. In one retrospective review of 135 patients with CP confirmed surgically or at autopsy, only 27 percent had pericardial calcification [10]. The cause of CP was more likely to be indeterminant when calcification was seen (67 versus 21 percent in the absence of calcification). (See "Constrictive pericarditis: Diagnostic evaluation", section on 'Chest radiograph'.)

●Restrictive cardiomyopathy – Pulmonary venous congestion with or without pleural effusions can be seen in RCM but is not generally seen in CP. (See "Restrictive cardiomyopathies", section on 'Chest imaging'.)

Echocardiography — An echocardiogram is generally the initial cardiac imaging test in patients with suspected CP or RCM. Echocardiographic findings guide further diagnostic testing (table 1 and algorithm 1).

Role of ventricular interaction — Markedly enhanced ventricular interaction is a key characteristic of CP not present in RCM. In patients with CP, the noncompliant pericardium limits total cardiac volume and impairs transmission of intrathoracic pressure to the cardiac chambers [11]. Ventricular septal motion is not limited, and, as a result, ventricular interdependence is greatly enhanced. Manifestations of ventricular interaction with CP include respirophasic ventricular filling and respirophasic ventricular septal shift. These findings are not present with RCM. (See 'Respirophasic ventricular filling' below and 'Respirophasic ventricular septal shift' below.)

Doppler echocardiography — CP and RCM share some Doppler findings, most notably a restrictive mitral inflow pattern, with striking E-wave dominance and a short deceleration time, indicating early rapid ventricular filling (figure 2) [12]. However, Doppler echocardiography provides the following clues that differentiate CP from RCM (table 1) [13], as discussed in the following sections:

Respirophasic ventricular filling

●Cause of respirophasic ventricular filling in CP – The fall in intrathoracic pressure with inspiration induces a physiologic fall in pulmonary capillary wedge pressure, while left ventricular (LV) pressure is shielded from respiratory pressure variations by the noncompliant pericardium. Thus, inspiration lowers the pulmonary capillary wedge pressure and left atrial pressure, but not LV diastolic pressure, thereby decreasing the pressure gradient for ventricular filling. The reduced LV filling pressure gradient causes a lower LV filling velocity and also delays mitral valve opening, so isovolumic relaxation time is lengthened during inspiration. Reciprocal changes occur in the velocity of RV filling [14,15]. These changes in RV filling are mediated by the ventricular septal shift. (See 'Respirophasic ventricular septal shift' below.)

●Role in distinguishing CP from RCM – The respiratory variation in ventricular filling velocity in RCM is usually minimal (less than 10 percent), while patients with CP may have respiratory variations as high as 30 to 40 percent in ventricular filling velocity (similar to that seen in cardiac tamponade). A meta-analysis including five studies with a total of 365 patients found that respiratory variation in mitral inflow ≥14.6 percent had a pooled sensitivity of 71 percent (95% CI 51-85 percent) and a pooled specificity of 82 percent (95% CI 66-91 percent) for differentiating CP from RCM [3].

●Conditions that may obscure respirophasic ventricular filling

•High left atrial pressure – Ventricular filling velocity is highly influenced by the level of the left atrial pressure. When left atrial pressure is greatly elevated in a patient with CP, respiratory variation in ventricular filling may not be observed, whereas patients with lower left atrial pressure (ie, due to volume depletion or earlier stage of disease) may have more noticeable changes in ventricular filling velocities with respiration. Left atrial pressure can be reduced by asking a patient to move from the supine to seated position; such a maneuver may elicit the abnormality [16].

•Atrial fibrillation – The ability to assess respirophasic changes in ventricular filling velocities is challenging in patients with atrial fibrillation, due to the presence of variable cardiac cycle length from beat to beat. While some investigators have suggested that respiratory variation can be accurately assessed in patients with atrial fibrillation [17], interpretation is often difficult in clinical practice [13].

●Distinguishing CP from COPD – Respiratory variation in mitral E velocity, as with pulsus paradoxus on physical examination, is not specific to CP and is frequently seen in patients with chronic obstructive pulmonary disease (COPD). In an attempt to distinguish between CP and COPD, the pulsed wave Doppler recordings of mitral and superior vena cava flow velocities were compared in 20 patients with COPD who had a 25 percent respiratory variation in mitral E-wave velocity and 20 patients with surgically proven CP [18]. The patients with COPD had a marked increase in inspiratory superior vena cava systolic flow velocity, which was not seen in those with CP (waveform 1).

Hepatic vein flow — Measurement of hepatic venous flow can help to distinguish CP from RCM.

●Constrictive pericarditis – In patients with CP, there is a reversal of hepatic vein forward flow during expiration, since the RV becomes less compliant with greater LV filling. The presence of hepatic vein expiratory diastolic flow reversal favors the presence of CP. In a meta-analysis of two studies with a total of 303 patients, a ratio of hepatic vein expiratory diastolic reversal velocity to forward velocity ≥0.79 had a pooled sensitivity of 73 percent (95% CI 65-81 percent) and a pooled specificity of 71 percent (95% CI 19-96 percent) for distinguishing CP from RCM [19].

●Restrictive cardiomyopathy – In contrast, reversal of hepatic vein flow occurs during inspiration in RCM [20].

Color M-mode flow propagation — Early diastolic color M-mode Doppler shows an excessively rapid transit of blood flow ≥100 cm/s for the first aliasing contour from the mitral orifice to the apex in CP, whereas the transit is slower than normal in RCM [21].

Tissue Doppler imaging — Tissue Doppler imaging provides information about myocardial function that can help to distinguish CP from RCM (table 1).

●Mitral annular velocities – In healthy individuals, lateral early diastolic mitral annular velocity (e’) is greater than septal e’. With CP, lateral e’ is reduced due to pericardial adhesions and septal e’ is enhanced, so the ratio of septal e’ to lateral e’ is elevated (known as annulus reversus) (waveform 2) [13,16,22]. In contrast, with RCM, both lateral and septal e’ are reduced due to intrinsic impairment in myocardial contraction and relaxation, and lateral e’ commonly remains higher than septal e’.

•Septal e’ ≥8 cm/s – In a meta-analysis including 12 studies with a total of 705 patients, septal e’ ≥8 cm/s had a pooled sensitivity of 83 percent (95% CI 80-87 percent) and pooled specificity of 90 percent (95% CI 83-95 percent) for differentiating CP from RCM [23].

•Septal e’/lateral e’ ≥0.88 – In a meta-analysis including three studies with a total of 375 patients, a septal e’ to lateral e’ ratio ≥0.88 had a pooled sensitivity of 74 percent (95% CI 64-82 percent) and specificity of 81 percent (95% CI 70-88 percent) for differentiating CP from RCM [23].

A reduced septal e’ (<9 cm/s) in a patient with confirmed CP suggests the presence of concurrent myocardial impairment (ie, mixed disease) [13].

Respirophasic ventricular septal shift — In patients with CP, the ventricular septum shifts posteriorly toward the LV (reflecting LV underfilling) with inspiration and the septum shifts anteriorly toward the RV (reflecting recovery of LV filling) with expiration (see 'Respirophasic ventricular filling' above). This respirophasic ventricular septal shift is commonly identified by echocardiography; it can also be visualized by cardiovascular magnetic resonance (CMR) imaging or cardiac computed tomography (CT) [24]. (See 'Cardiovascular magnetic resonance (CMR)' below and 'Cardiac computed tomography' below.)

In a meta-analysis including four studies with a total of 383 patients, respirophasic ventricular septal shift on echocardiogram had a pooled sensitivity of 82 percent (95% CI 60-94 percent) and a pooled specificity of 78 percent (95% CI 65-87 percent) for distinguishing CP from RCM [23].

Strain analysis — Strain analysis with two-dimensional speckle tracking may be helpful in distinguishing between CP and RCM [13,22,25]. This method may avoid the limitations of tissue Doppler annular velocities when the annulus is calcified.

With CP, lateral longitudinal strain is usually reduced more than septal strain; this finding is known as strain reversus. Greater depression in lateral versus septal longitudinal strain is not generally seen in RCM. As an example, the ratios of LV lateral/septal strain and RV free wall strain/septal strain were significantly lower in 52 patients with CP (both 0.8) than in 35 patients with RCM (1.1 and 1.4) and 26 control subjects (1.0 and 1.2) [25]. A LV lateral to septal longitudinal wall strain ratio <0.96 had a sensitivity of 89 percent and a specificity of 96 percent for differentiating CP from RCM. The strain ratios were more robust than the ratios of annular tissue velocities. In patients with CP, the regional myocardial mechanics significantly correlated with regional pericardial thickness and improved after pericardiectomy.

In CP, circumferential strain and torsion (which arise chiefly from the subepicardial myocardium) are reduced compared with RCM and healthy controls [13,26]. Conversely, global longitudinal strain (which arises chiefly from the subendocardial myocardium) tends to be lower in RCM than in CP.

Pericardial imaging — The pericardium is thickened in most patients with CP but is not thickened in RCM.

While transthoracic echocardiography (TTE) is the most commonly used imaging modality for suspected pericardial disease, it is not a reliable tool for assessing pericardial thickness [13,24], as illustrated by the following study. In a review of 143 patients with surgically confirmed CP, a pathologically thickened pericardium (>2 mm) was seen in 82 percent [27]. Among patients with thickened pericardium on pathologic examination, 43 percent had pericardial thickening identified by TTE, and 86 percent had thickened pericardium identified by cardiac CT. Transesophageal echocardiography (TEE) measurements of pericardial thickness may better correlate with measurements on cardiac CT [28], but data are limited and TEE is rarely performed solely for this indication.

Pericardial thickness is most accurately imaged by cardiac CT or CMR imaging. (See 'Cardiovascular magnetic resonance (CMR)' below and 'Cardiac computed tomography' below.)

Signs of pericardial tethering are occasionally detected by echocardiography. Detailed assessment of pericardial tethering can be performed with tagged CMR. (See 'Cardiovascular magnetic resonance (CMR)' below.)

ADDITIONAL CARDIAC TESTING — 

While a diagnosis is often made following echocardiography, patients commonly undergo cardiac catheterization, during which invasive hemodynamic evaluation can help elucidate the correct diagnosis [29]. In most patients, there is a role for CT or CMR imaging [29]. Both CT and CMR provide accurate measurement of pericardial thickness and additional detailed anatomic information about cardiac and adjacent structures (eg, lymph nodes).

Cardiac computed tomography — Cardiac CT enables the following assessment [24]:

●Pericardial characterization – Cardiac CT provide accurate quantification of pericardial thickness (image 2). A thickened pericardium is suggestive of CP but is not specific or sensitive for this diagnosis.

Cardiac CT is a key test for identification of pericardial calcification (which may not be detected on chest radiograph). However, the presence of pericardial calcification is also not sufficiently sensitive or specific for CP. Pericardial calcification is present in a minority of patients treated surgically for CP.

●IVC dilation – The inferior vena cava (IVC) is typically dilated in patients with CP. Thus, a diagnosis of CP is unlikely if IVC dilation is absent.

Cardiovascular magnetic resonance (CMR) — CMR may help differentiate CP from RCM (table 1) [30].

●Pericardial characterization – Pericardial thickness can be accurately quantified by CMR [31,32]. As noted above, the presence of thickened pericardium is not sufficient to establish a diagnosis of CP, and absence of pericardial thickening does not exclude the presence of CP [27,31]. (See 'Pericardial imaging' above.)

Pericardial late gadolinium enhancement (LGE) is present in a subset of patients with CP. Small studies suggest that moderate or severe pericardial LGE and greater pericardial thickness detected by CMR are characteristics associated with transient CP [33,34].

Radiofrequency tissue tagging enables assessment of the presence and distribution of pericardial tethering [35]. Tag lines maintain continuity throughout the cardiac cycle in regions of pericardial tethering, in contrast to the discontinuity in tag lines seen with normal slippage between the visceral and parietal pericardium.

●Global longitudinal strain – In a cohort of 92 patients (28 with CP, 30 with RCM, and 34 controls), both echocardiographic speckle tracking and CMR-derived measures of global longitudinal strain using tissue tracking were significantly higher in patients with CP than RCM; diagnostic discrimination was comparable with echocardiography and CMR [22]. (See 'Strain analysis' above.)

●Respirophasic ventricular septal shift – Similar to echocardiography, real-time cine CMR of ventricular septal position and shape can be used to distinguish CP from RCM. This was demonstrated in a study including patients with CP (n = 18), inflammatory pericarditis (n = 6), or RCM (n = 15) and healthy controls (n = 17) [36]. The degree of ventricular coupling, measured as the difference in the maximal septal excursion between inspiration and expiration (normalized to the biventricular diameter), was significantly greater in patients with CP (20 percent) and inflammatory pericarditis (14.8 percent) than in normal subjects (7 percent) and patients with RCM (4.2 percent). A partition value of 11.8 percent completely differentiated CP from RCM.

●Relative atrial volume ratio – Limited data suggest that the ratio between left and right atrial volume is higher with CP than with RCM. A study found that the relative atrial volume ratio in 23 patients with CP was significantly higher than in 22 patients with RCM (1.5±0.29 versus 1.12±0.33). A partition value of 1.32 for the relative atrial ratio yielded a sensitivity of 82.6 percent and specificity of 86.4 percent for the diagnosis of CP [32].

●Myocardial LGE – Myocardial LGE is not present in patients with isolated CP. In the above-cited study of 45 patients with either CP or RCM, myocardial LGE was present in 31.8 percent of patients with RCM and absent in all patients with CP and in normal subjects [32]. In another study, patients with biopsy-proven cardiac amyloidosis had a pattern of LGE involving the entire subendocardial circumference; the positive and negative predictive values for diagnosing cardiac amyloidosis were 92 and 85 percent, respectively [37].

Nuclear imaging — The role of nuclear imaging in patients with suspected RCM to evaluate for amyloidosis and sarcoidosis is discussed separately. (See "Restrictive cardiomyopathies", section on 'Nuclear imaging' and "Clinical manifestations and diagnosis of cardiac sarcoidosis", section on 'FDG-PET'.)

Cardiac catheterization — In selected cases, particularly when there are multiple potential etiologies for heart failure, invasive cardiac catheterization is indicated to determine or confirm the diagnosis.

Dip and plateau sign — In patients with CP or RCM, there is a "dip and plateau" or "square root" pattern in the RV and LV pressure waveforms (waveform 3), which reflects rapid ventricular filling during early diastole, which is halted by pericardial constraint (with CP) [38] or noncompliant ventricles (with RCM). While a dip and plateau sign is seen in both conditions, an LV rapid filling wave height >7 mmHg favors CP [39].

Equalization of pressures — Equalization of filling pressures in all four cardiac chambers is commonly described as a key criterion for CP. However, this is an oversimplification since pulmonary wedge pressure falls during inspiration while systemic venous pressure remains constant, so equalization may be present only during inspiration. Also, equalization of pressures may be seen in other disease states, including RCM, other causes of heart failure, and acute volume overload (eg, acute mitral regurgitation).

Some patients with RCM have LV end-diastolic pressure (LVEDP) that exceeds RV end-diastolic pressure (RVEDP) by ≥5 mmHg and pulmonary hypertension [40]. However, these findings are not sufficient to reliably distinguish RCM from CP [11,41,42].

Of note, diastolic equalization may not be present in patients with CP or RCM who have early disease or are volume depleted. In patients with a low-volume state, the cardiac output will be low. Therefore, the presence of normal cardiac output with normal filling pressures precludes the presence of hemodynamically significant CP.

In patients who are volume depleted, fluid loading at the time of catheterization is a way to determine if equalization of pressures occurs in a euvolemic state.

Signs of ventricular interaction — The role of ventricular interaction in causing reciprocal respirophasic ventricular filling is described above. (See 'Role of ventricular interaction' above and 'Respirophasic ventricular filling' above.)

Mirror image discordance — In CP, respiratory changes in LV and RV filling reflecting enhanced ventricular interaction result in distinctive changes in the LV and RV pressures during respiration (table 1). Mirror image discordance in LV and RV pressures during respiration (figure 3), measured as a ratio of pressure-time areas under the RV and LV systolic pressure curves during inspiration versus expiration (systolic area index) >1.1, was 97 percent sensitive and 100 percent specific for differentiating 59 cases of surgically proven CP from 41 cases of RCM [39]. (See 'Role of ventricular interaction' above.)

Discordant aortic and pulmonary ejection times — A simplified method to detect enhanced ventricular interaction in the catheterization laboratory using aortic and pulmonic ejection times (ie, surrogates for stroke volumes) has been described [43]. Ventricular interaction was assessed by evaluating the difference of aortic and pulmonary artery ejection times from expiration to inspiration. The difference was significant in a 10-patient cohort with surgically proven CP versus 10 patients with RCM or severe tricuspid regurgitation (50.8±22.5 ms versus 5.4±15.2 ms). Although the method is potentially useful, discordant ejection times are not specific for constriction and further validation is needed.

Ratio of right atrial pressure to pulmonary artery wedge pressure — Right atrial pressure (RAP) fairly closely approximates pericardial pressure, while pulmonary artery wedge pressure (PAWP) approximates LVEDP (in the absence of mitral stenosis). It has been hypothesized that a larger contribution of pericardial restraint, as shown by a greater RAP/PAWP ratio, could help identify patients with "pure" constriction (ie, minimal contribution of myocardial disease to symptoms) in whom pericardiectomy is more likely to be beneficial. In a single-center retrospective study of 113 patients with surgically confirmed CP, those with values above the median (mean RAP/PAWP ratio 0.86) had greater pericardial thickness (6.6 versus 4.5 mm) and greater postoperative survival [15,29]. This approach is similar to examining for diastolic equalization of intracardiac pressures. While potentially useful, the RAP/PCWP ratio should not be used in isolation and requires further validation.

Biopsy — Endomyocardial or, less commonly, pericardial biopsy may be helpful when hemodynamic and imaging studies fail to establish the diagnosis [44,45]. (See "Restrictive cardiomyopathies", section on 'Endomyocardial biopsy'.)

Proposed use of artificial intelligence — A deep-learning model of echocardiographic analysis has been tested as a potential means of aiding differentiation of CP from RCM [46]. The apical four-chamber views from transthoracic echocardiograms from 381 patients (184 with CP and 197 with cardiac amyloidosis) were split into training, validation, and test sets and input into a deep learning model. The model differentiated the two conditions with an area under the curve of 0.97. The Gradient-Weighted Class Activation Mapping (GradCAM) heatmap showed activation around the ventricular septum.

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pericardial disease".)

SUMMARY AND RECOMMENDATIONS

●Approach to distinguishing constrictive pericarditis from restrictive cardiomyopathy – Patients with both constrictive pericarditis (CP) and restrictive cardiomyopathy (RCM) have elevated left- and right-sided filling pressures, often of equal magnitude, and usually have normal systolic ventricular function. A number of findings that can aid in distinguishing the two conditions include (table 1 and algorithm 1):

•History – The history may provide helpful clues, such as prior pericarditis or a systemic disease predisposing to CP, or a cause of RCM (eg, sarcoidosis or amyloidosis). (See 'History' above.)

•Physical Examination – The vast majority of patients with both CP and RCM display elevated jugular venous pressure (JVP). The presence of a pericardial knock (an accentuated heart sound occurring slightly earlier than a third heart sound) favors CP, while an audible S3 is frequently present in persons with RCM. (See 'Physical examination' above.)

•Electrocardiography – ECG findings are nonspecific but may suggest CP or RCM. Repolarization abnormalities are more typical for CP. Depolarization abnormalities (such as bundle branch block), ventricular hypertrophy, pathologic Q waves, or impaired atrioventricular conduction suggest RCM. (See 'Electrocardiogram' above.)

•B-type natriuretic peptide (BNP) – Plasma N-terminal pro-BNP (NT-proBNP) or BNP should be measured early in the evaluation of patients with suspected CP or RCM, as normal values essentially exclude RCM. (See 'Natriuretic peptides' above.)

•Chest radiograph – Calcification of the pericardium (except for scattered calcified plaques which are more consistent with prior asbestos exposure) strongly suggests CP (image 1); however, pericardial calcification is not excluded by chest radiograph. Cardiomegaly on chest radiograph (generally due to atrial rather than ventricular enlargement) is more prominent in RCM. (See 'Chest radiography' above.)

•Echocardiography – An echocardiogram is generally the initial cardiac imaging test in patients with suspected CP or RCM (table 1). Exaggerated ventricular interaction is a key characteristic of CP, which impacts ventricular septal motion and ventricular filling.

•Respirophasic ventricular septal shift – With CP, the ventricular septum shifts posteriorly toward the left ventricle (LV) with inspiration and anteriorly toward the right ventricle (RV) with expiration. (See 'Respirophasic ventricular filling' above.)

•Doppler echocardiography – (See 'Doppler echocardiography' above.)

-Respirophasic variations in ventricular filling velocity are prominent (as high as 30 to 40 percent) with CP and generally minimal (<10 percent) in RCM. (See 'Respirophasic ventricular filling' above.)

-Hepatic venous flow reversal is expiratory in CP, but inspiratory in RCM. (See 'Hepatic vein flow' above.)

-Tissue Doppler imaging enables identification of enhanced septal early diastolic mitral annular velocity (e’) and depressed lateral e’ (annulus reversus) in CP. (See 'Tissue Doppler imaging' above.)

●Pericardial imaging – Increased thickness or calcification of the pericardium favors a diagnosis of CP, although increased pericardial thickness is not present in all patients. Pericardial thickening is most reliably assessed by cardiac computed tomography (CT) or cardiovascular magnetic resonance (CMR) imaging. Pericardial calcification is most reliably assessed by cardiac CT. (See 'Pericardial imaging' above.)

Signs of pericardial tethering are occasionally detected by echocardiography. Detailed assessment of pericardial tethering can be performed with tagged CMR. (See 'Cardiovascular magnetic resonance (CMR)' above.)

●Cardiac catheterization

•Constrictive pericarditis – RV and LV end-diastolic pressures (RVEDP and LVEDP) are equal or very nearly equal in CP. Mirror-image discordance between RV and peak LV systolic pressures during inspiration is a sign of increased ventricular interdependence (table 1); during peak inspiration, an increase in RV pressure occurs when LV pressure is lowest [11] (See 'Cardiac catheterization' above.)

•Restrictive cardiomyopathy – LVEDP is commonly higher than RVEDP in RCM. However, in many cases of RCM, the plateaus of RV and LV diastolic pressures are equally elevated. If the RVEDP and LVEDP are approximately equal, a fluid bolus, exercise, or pharmacologic maneuver may increase LVEDP above RVEDP in RCM.

●Biopsy – Endomyocardial or, less commonly, pericardial biopsy may be helpful when hemodynamic and imaging studies fail to establish the diagnosis. (See 'Cardiac catheterization' above.)