Pericardial Disease

Diagnostic Concordance of Echocardiography and

Cardiac Magnetic Resonance–Based Tissue Tracking
for Differentiating Constrictive Pericarditis From
Restrictive Cardiomyopathy
Makoto Amaki, MD, PhD; John Savino, MD; David L. Ain, MD; Javier Sanz, MD;
Gianni Pedrizzetti, PhD; Hemant Kulkarni, MD; Jagat Narula, MD, PhD; Partho P. Sengupta, MD

Background—Variations in longitudinal deformation of the left ventricle have been suggested to be useful for differentiating
chronic constrictive pericarditis (CP) and restrictive cardiomyopathy (RCM). We assessed left ventricular mechanics
derived from cardiac magnetic resonance (CMR) cine–based and 2-dimensional echocardiography–based tissue tracking
to determine intermodality consistency of diagnostic information for differentiating CP from RCM.
Methods and Results—We retrospectively identified 92 patients who underwent both CMR and 2-dimensional echocardiography
and who had a final diagnosis of CP (n=28), RCM (n=30), or no structural heart disease (n=34). Global longitudinal strain from
long-axis views and circumferential strain from short-axis views were measured on 2-dimensional echocardiographic and
CMR cine images using the same offline software. Logistic regression models with receiver operating characteristics curves,
continuous net reclassification improvement, and the integrated discrimination improvement (IDI) were used for assessing
the incremental predictive performance. Global longitudinal strain was higher in patients with CP than in those with RCM
(P<0.001), and both techniques were found to have similar diagnostic value (area under the curve, 0.84 versus 0.88 for CMR
and echocardiography, respectively). For echocardiography, the addition of global longitudinal strain to respiratory septal
shift and early diastolic mitral annular velocity resulted in improved continuous net reclassification improvement (P<0.001
for both) and integrated discrimination improvement (P=0.005 and 0.024) for both models. Similarly, for CMR, the addition
of global longitudinal strain to septal shift and pericardial thickness resulted in improved continuous net reclassification
improvement (P<0.001 for both) and integrated discrimination improvement (P=0.003 and <0.001).
Downloaded from http://ahajournals.org by on April 12, 2020

Conclusions—CMR and echocardiography tissue tracking–derived left ventricular mechanics provide comparable diagnostic
information for differentiating CP from RCM.  (Circ Cardiovasc Imaging. 2014;7:819-827.)

Key Words: cardiomyopathy, restrictive ◼ constrictive pericarditis ◼ echocardiography ◼ magnetic resonance imaging

D ifferentiating constrictive pericarditis (CP) from restrictive

cardiomyopathy (RCM) in patients with right heart fail-
ure can be a challenge. Cardiac catheterization, although often
performed for hemodynamic confirmation, may not always
be conclusive.1,2 This has led to continued interest in other
diagnostic modalities. Several echocardiographic measure-
ments have been proposed to differentiate myocardial diseases
from pericardial constriction.3,4 Cardiac magnetic resonance
(CMR) imaging has superior contrast-to-noise and signal-to-
noise ratios, permitting accurate quantification of pericardial
thickening and providing assessment of the entire thorax.5
Because noninvasive imaging techniques continue to advance,
the clinical differentiation of CP from RCM is based on the
recognition of a cluster of structural, mechanical, and hemody-
namic aberrations rather than a single structural or functional
variable used in isolation. Specifically the recent growth in
tissue tracking software allows myocardial functional data to
be extracted from both echocardiography and CMR images
with equal ease and accuracy.6–8 An emphasis on intermodal-
ity comparison may permit consistent data collection for serial
comparisons. We, therefore, sought to compare the diagnostic
value of strain measurements derived from 2-dimensional (2D)
echocardiography-based and CMR cine–based tissue tracking
methods for differentiating CP from RCM.
Clinical Perspective on p 827
Methods
Patient Population
The institutional review board at the Mount Sinai Medical Center
approved the protocol. We retrospectively identified 92 patients who
underwent both transthoracic echocardiography and CMR imaging

Received February 25, 2014; accepted August 1, 2014.

From the Zena and Michael A. Wiener Cardiovascular Institute (M.A., D.L.A., J.S., G.P., J.N., P.P.S.) and Department of Medicine (J.S.), Icahn School
of Medicine at Mount Sinai, New York, NY; Department of Civil Engineering and Architecture, University of Trieste, Italy (G.P.); and Department of
Medicine, University of Texas Health Science Center, San Antonio (H.K.).
Correspondence to Partho P. Sengupta, MD, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place, PO Box 1030, New York, NY 10029.
E-mail partho.sengupta@mountsinai.org
© 2014 American Heart Association, Inc.
Circ Cardiovasc Imaging is available at http://circimaging.ahajournals.org DOI: 10.1161/CIRCIMAGING.114.002103

819
 820  Circ Cardiovasc Imaging  September 2014
between May 2006 and May 2013. This included 28 patients with CP,
30 patients with RCM, and 34 control patients who had no structural
or functional abnormality on echocardiography or CMR. The control
subjects were age- and sex-balanced to the mean values obtained for
both CP and RCM. Among these subjects, 80 (87%) had both imag-
ing tests with a median interval between the tests of 2.0 days and
interquartile range of 12 to 21 days, whereas 12 subjects (13%) had
both examinations on the same day.
Cardiac MRI
CMR was performed using either a 1.5-T magnet (Magnetom Sonata;
Siemens Medical Solutions, Erlangen, Germany; Optima MR450w;
General Electric, Milwaukee) or a 3.0-T magnet (Ingenuity TF scan-
ner; Philips, Best, The Netherlands) and a dedicated surface coil with
retrospective electrocardiographic gating, as previously described.9–11
Contiguous cine short-axis views were acquired using steady-state
free precession imaging at end-expiratory breathholds. Left ven-
tricular (LV) end-diastolic volume and end-systolic volume, ejection
fraction (LVEF), and mass were obtained according to the Simpson
method using specialized software and indexed to body surface area.
The presence of increased pericardial thickness (>4 mm)12 and septal
shifts during respiration13 were reported. Constrictive physiology on
CMR was defined by increased LV-right ventricle (RV) coupling as
defined by visual determination of septal shifts during respiration.14
Findings consistent with RCM in CMR included normal or thick-
ened RV and LV walls, normal or small LV cavity size, and enlarged
atria. Gadolinium-enhanced CMR was also performed, and the pres-
ence of circumferential delayed enhancement involving the entire
subendocardium and extending into the surrounding myocardium
was considered typical for amyloidosis,15 the most common under-
lying cause for RCM. Other patterns including localized enhance-
ment with a vascular distribution and focal or subepicardial delayed
enhancement were excluded. Three experienced, blinded readers (J.
Sanz and two colleagues) reviewed the CMR images.
Downloaded from http://ahajournals.org by on April 12, 2020
Echocardiographic Doppler Studies
All patients had comprehensive 2D echocardiogram and Doppler
studies, performed and interpreted according to American Society
of Echocardiography standards. Three different commercially avail-
able ultrasound machines were used (Philips IE33 system; Philips
Medical Systems, Andover, MA; Acuson 512; Siemens Medical
Solutions, Inc, Mountain View, CA; Vivid 7; GE Vingmed Ultrasound
AS, Horten, Norway). From the apical 4- and 2-chamber views, LV
end-diastolic volume, LV end-systolic volume, and LVEF were cal-
culated using the biplane modified Simpson method, and left atrial
volume index was calculated using the biplane area length method.16
LV mass index was derived from the parasternal long-axis view us-
ing M-mode measurements.17,18 Findings consistent with CP on echo-
cardiography included plethora of the inferior vena cava, abnormal
motion of the interventricular septum, respiratory variation of LV
inflow velocities, and increased expiratory hepatic vein flow reversal,
as has been previously described.19 The echocardiographic findings
of RCM were similar to those described for CMR (with the exception
of delayed enhancement imaging), as well as restrictive filling pattern
of transmitral flow and reduced tissue Doppler peak early-diastolic
mitral-annular velocities.20–22
Diagnosis of CP and RCM
The identification of subjects with CP for inclusion was based on a
previously described diagnostic paradigm. We included patients pre-
senting with heart failure with preserved EF (>50%) in whom the ini-
tial echocardiographic assessment suggested CP and who fulfilled at
least 1 of 4 additional criteria: (1) surgical confirmation during peri-
cardiectomy; (2) cardiac catheterization findings consistent with CP;
(3) evidence of thickened pericardium (thickness >4 mm by CMR);
or (4) evidence of increased LV-RV coupling (septal shift with res-
piration) by both echocardiography and CMR.4,23,24 The catheteriza-
tion criteria included ≥2 of the following criteria: (1) a difference
between LV end-diastolic pressure and RV end-diastolic pressure of
≤5 mm Hg; (2) pulmonary arterial systolic pressure <55 mm Hg; (3) a
ratio of RV end-diastolic pressure to RV systolic pressure of >1/3; (4)
inspiratory decrease in pulmonary capillary wedge pressure/LV end-
diastolic pressure difference of >5 mm Hg; and (5) systolic area index
>1.1.24 A total of 24 patients (86%) were referred for pericardiec-
tomy, with preoperative cardiac catheterization in 22 patients (79%).
Four patients without surgical referral were managed with medical
therapy because of end-stage malignancy. Of the patients who were
referred for pericardiectomy, 16 (57%) underwent partial or complete
removal of the pericardium, 3 (11%) refused the operation, 2 (7%)
were deemed too high risk for the procedure, and 3 (11%) were lost to
follow-up. Of the 12 patients who did not undergo pericardiectomy,
the criteria for diagnosing CP were fulfilled using cardiac catheter-
ization in 6 (21%) patients, by demonstration of echocardiographic
feature of CP and thickened pericardium in 4 (14%) and by demon-
stration of increased LV-RV coupling on echocardiography and CMR
in 2 (7%) patients.
The diagnosis of RCM was made by presence of all of the fol-
lowing echocardiographic features: (1) an interventricular septum
>12 mm, (2) preserved EF (>50%), (3) biatrial enlargement, and (4)
restrictive filling pattern, as well as delayed gadolinium-enhanced
CMR evidence of myocardial involvement.
Strain Analysis by Tissue Tracking in CMR and in
Echocardiography
Gray-scale echocardiographic images were saved with a frame rate
of 25 to 40 frames/second, and CMR cine images were obtained with
a temporal resolution <50 ms and reconstructed into 20 to 25 phases
per cardiac cycle. The images were digitally stored in the Digital
Imaging and Communications in Medicine platform (The National
Electrical Manufacturers Association).
Strain measurements were performed using the vendor-customized
offline 2D Cardiac Performance Analysis software for echocardiog-
raphy (2D-CPA, version 1.1.3) and CMR (2D-CPA MR, version
1.1.2.36; TomTec Imaging Systems GmbH, Unterschleissheim,
Germany). The software enabled measurement of the angle-indepen-
dent 2D strain previously validated with sonomicrometry and tagging
CMR both in echocardiography25,26 and CMR.27 2D-CPA determines
myocardial motion based on the user-defined myocardial border; en-
docardial borders are traced throughout 1 cardiac cycle. From this
motion, myocardial velocity and strain components are calculated for
endocardial regions along the trace.28 Measurements of LV longitu-
dinal strain by CMR were derived from the 2-, 3-, and 4-chamber
views, and circumferential strain from short-axis views at the level
of the papillary muscles.29 Global longitudinal strain (GLS) for both
techniques was expressed as the average endocardial strain value
from apical 4- and 2-chamber views as previously described because
the apical 3-chamber view in patients with CP was frequently sub-
optimal.4 To obtain differences in septal versus lateral wall deforma-
tion, LV septal wall strains and lateral wall strains were measured
by averaging strains from 3 segments in each wall. The peak values
for all strain parameters were recorded and analyzed. Figure 1 shows
examples of longitudinal strains in CP, RCM, and controls in both
modalities. The offline analysis was performed independently by 1
observer (M. Amaki) who was not involved in image acquisition. The
retrospective review of electronic medical records and offline analy-
sis of stored echocardiography and CMR images for the present study
was approved by institutional review board of Mount Siani Hospital.
Statistical Analysis
Continuous data were reported as the median and interquartile range
within each group, and categorical data as numbers and percent-
ages. For continuous variables, the differences between the groups
were evaluated with Kruskal–Wallis test because the majority of the
variables showed skewed distribution in ≥1 groups. For categorical
variables, the χ2 test was used. Mann–Whitney tests for pairwise com-
parisons with Bonferroni-adjusted levels were used to assess differ-
ences among groups. We performed a logistic regression analysis to
determine univariate echocardiographic predictors for differentiating
RCM from CP. The univariate predictors were subsequently entered
 Amaki et al   LV Mechanics in Constriction and Restriction   821
Downloaded from http://ahajournals.org by on April 12, 2020
Figure 1. Longitudinal strain (LS) measured from both echocardiogram and cardiac MR images using apical 4-chamber view in a patient
with constrictive pericarditis (A and B), restrictive cardiomyopathy (C and D), and control (E and F).
into stepwise multivariable logistic regression analysis. The compari-
son of receiver operating characteristics curves was performed ac-
cording to the method described by DeLong et al30 using Medcalc
Software (version 12.7.1.0; Mariakerke, Belgium), with P value re-
ported for a 2-sided test. We also determined the incremental value of
GLS by comparing logistic regression models with and without GLS
in addition to the known echocardiographic and CMR discriminators
between CP and RCM. For this, we used 4 characteristics that capture
conceptually differing domains of incremental value: (1) information
content (quantified as Akaike information criterion), (2) diagnostic
accuracy (measured as area under receiver operating characteristics
curve), (3) improved discrimination (measured as integrated discrimi-
nation improvement), and (4) reclassification (measured as net reclas-
sification index). We used the R package PredictABEL31 to estimate
continuous net reclassification index and integrated discrimination
improvement,32,33 and MedCalc to determine Akaike information cri-
terion and area under the receiver operating characteristics curves.
All significance values for improvement were 1-sided. Interobserver
agreement and intraobserver consistency were presented using inter-
class correlation coefficients and a 95% confidence interval (95% CI).
A P value <0.05 was considered significant.
Results
Clinical Characteristics, Echocardiographic, and
CMR Measurements
The cause of CP was previous cardiac surgery in 7 (25%),
postpericarditis in 6 (21%), malignancy in 2 (7%), and post-
radiation in 1 patient (4%). The remaining 12 patients (43%)
had idiopathic CP. Of the patients with RCM, 11 (37%) had
cardiac biopsy proven amyloidosis and 11 (37%) had extra-
cardiac biopsy suggesting cardiac amyloidosis. The cause of
myocardial restriction could not be determined in the remain-
ing 8 patients (27%).
Table 1 summarizes the clinical characteristics, conven-
tional echocardiographic measurements, and CMR mea-
surements of the 3 groups. CP patients had higher heart
rates compared with RCM patients (P=0.02). Other param-
eters including sex, body surface area, blood pressure, and
the percentage of atrial fibrillation and coronary artery
disease risk factors were comparable between the CP and
RCM groups.
With respect to the echocardiographic findings, in com-
parison with CP patients, RCM patients had significantly
higher LV wall thickness and LV mass index, lower sep-
tal early diastolic mitral annular velocity (e′), and higher
E/e′ (P<0.001 for all). Other echocardiographic parameters
including LV dimensions, transmitral flow characteristics,
and left atrial volume index demonstrated no significant dif-
ference between the 2 disease groups. Both CP and RCM
had preserved LVEF (>50%), although the absolute values
for LVEF were marginally lower for RCM patients. Septal
shifts and >25% respiratory variation in LV inflow veloci-
ties were seen in 25 (89%) and 11 (39%) patients with CP,
 822  Circ Cardiovasc Imaging  September 2014

Table 1.  Clinical Characteristics and Standard Measurements of the Study Subjects
CP (n=28) RCM (n=30) Control (n=34) P Value
Age, y 60 (48–69) 64 (59–76) 61 (55–66) 0.182
Men, n (%) 19 (68) 20 (67) 19 (56) 0.550
Body surface area, cm/m2 1.88 (1.70–2.06) 1.93 (1.78–2.05) 1.92 (1.78–2.07) 0.687
Systolic BP, mm Hg 117 (110.0–124.0) 109.5 (102.0–130.0) 123.0 (110.0–136.0)* 0.045
Diastolic BP, mm Hg 70.0 (65.0–73.0) 67.0 (60.0–74.0) 72.0 (67.0–83.0)† 0.016
Heart rate, beats per min 83.5 (72.0–94.5) 73.0 (68.0–81.0)‡ 70.0 (65.0–80.0)‡ 0.009
Atrial fibrillation, n (%) 4 (14) 3 (10) 0 (0) 0.090
Hypertension, n (%) 11 (39) 14 (47) 18 (53) 0.474
Hyperlipidemia, n (%) 10 (36) 14 (47) 17 (50) 0.510
Diabetes mellitus, n (%) 8 (29) 5 (17) 8 (24) 0.354
Smoking, n (%) 8 (29) 7 (23) 8 (24) 0.872
Familial history of CAD, n (%) 3 (11) 3 (10) 5 (15) 0.610
Echocardiogram
 LV septum, mm 8.0 (7.0–9.5) 13.5 (12.0–15.0)‡ 10.0 (8.2–10.8)†‡ <0.001
 LVDd, mm 42.0 (37.0–46.0) 41.0 (38.0–45.0) 46.6 (42.0–50.2)†‡ 0.002
 LVDs, mm 26.0 (23.0–28.0) 28.0 (25.0–29.0) 27.6 (25.6–32.2) 0.103
 LVEF, % 64.5 (55.5–69.0) 58.5 (53.0–64.0)§ 65.0 (62.0–69.0)† 0.010
 LVMI, g/m2 60.5 (50.5–72.5) 124.5 (107.0–147.0)‡ 81.0 (75.0–119.0)†‡ <0.001
 E wave, m/s 79.5 (70.0–97.5) 99.0 (79.5–116.3) 69.0 (57.6–83.3)* <0.001
 A wave, m/s 49.0 (39.0–66.5) 38.0 (28.0–49.0) 67.5 (56.5–76.5)‡ <0.001
 E/A wave 1.55 (1.30–2.00) 2.50 (2.10–2.90) 0.95 (0.80–1.30)†‡ <0.001
 DT, ms 184.5 (136.5–202.0) 138.5 (127.0–159.5) 230.0 (188.3–270.3)†‡ <0.001
 e′, cm/s 9.9 (9.0–13.3) 5.2 (4.1–6.4)‡ 7.4 (6.1–9.8)†‡ <0.001
Downloaded from http://ahajournals.org by on April 12, 2020

 E/e′ 8.0 (5.7–11.3) 18.4 (14.8–23.6)‡ 8.5 (7.1–12.5)† <0.001

 LAV index, mL/m2 41.0 (30.0–53.0) 45.0 (31.0–52.0) 28.0 (23.8–36.0)†‡ <0.001
CMR
 LVEDV, mL 114.5 (95.5–147.0) 146.5 (124.0–166.0)‡ 125.5 (113.0–147.5)* 0.011
 LVESV, mL 41.5 (34.0–68.5) 63.5 (53.0–75.0)‡ 49.0 (40.0–57.5)† 0.009
 LVEF, % 60.5 (56.5–67.0) 54.5 (50.0–61.0) 63.0 (56.8–66.0)* 0.066
 LVMI, g/m2 44.3 (40.4–62.9) 87.3 (63.0–113.6)‡ 53.8 (44.2–63.9)† <0.001
 Myocardial enhancement║ 2 (11) 30 (100) … …
Values are median (interquartile range); P value across 3 groups by Kruskal–Wallis test and χ test for categorical value. BP indicates blood
2

pressure; CAD, coronary artery disease; CMR, cardiac MR; CP, constrictive pericarditis; Dd, diastolic dimension; Ds, systolic dimension; DT,
deceleration time; EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume; LAV, left atrial volume; LV, left ventricle; MI, mass
index; and RCM, restrictive cardiomyopathy.
*P<0.05 vs RCM by Mann-Whitney tests.
†P<0.01 vs RCM by Mann–Whitney tests.
‡P<0.01 vs CP by Mann-Whitney tests.
§P<0.05 vs CP by Mann-Whitney tests.
║Focal enhancement in CP patients and diffuse enhancement in RCM patients.

respectively. Mild-to-moderate mitral regurgitation was
found in 3 CP patients and 6 RCM patients. Six patients
with CP and 2 patients with RCM had mild to moderate
tricuspid regurgitation.
CMR measurements also revealed higher LV mass index
in RCM, but LVEF was comparable between the 2 groups.
Respiratory septal shifts were seen in 22 patients (79%) with
CP. Delayed contrast CMR was performed in 19 patients
(68%) with CP and revealed pericardial enhancements in 10
patients and focal subsegmental myocardial enhancements in
2. In comparison, multisegmental diffuse enhancement was
seen in all RCM patients.
Strain Analysis by Echocardiogram and CMR
Echocardiography-derived GLS was significantly reduced in
the RCM group compared with the CP (P<0.001) and con-
trol (P<0.001) groups (Table 2; Figure 2). The ratio between
lateral and septal longitudinal strain was not significantly dif-
ferent among the 3 groups (P=0.78). Similarly, patients with
RCM had marginally lower circumferential strains when com-
pared with CP (P=0.07).
Similar to echocardiography, CMR-derived GLS was
reduced in the RCM group compared with the CP (P<0.001)
and control groups (P<0.001). The ratio between lateral to
septal longitudinal strain showed no difference among the 3
 Amaki et al   LV Mechanics in Constriction and Restriction   823

Table 2.  Echocardiography and CMR-Derived Strain Characteristics

CP (n=28) RCM (n=30) Control (n=34) P Value
Echocardiogram
 LS (4ch), % −17.2 (−20.3 to −15.0) −13.0 (−14.7 to −9.0)* −17.2 (−21.7 to −12.9)† <0.001
 LS (2ch), % −17.0 (−21.6 to −15.5) −11.0 (−14.3 to −9.4)* −18.0 (−23.4 to −16.8)† <0.001
 GLS‡, % −18.5 (−20.1 to −15.2) −11.6 (−14.6 to −9.3)* −18.2 (−20.4 to −16.2)† <0.001
 Septal LS (4ch), % −17.6 (−21.7 to −13.8) −12.3 (−14.6 to −8.5)* −16.8 (−20.9 to −12.4)† <0.001
 Lateral LS (4ch), % −17.5 (−22.1 to −15.7) −12.9 (−15.2 to −7.5)* −17.4 (−23.2 to −14.2)† <0.001
 Lateral/Septal LS ratio (4ch) 0.90 (0.85 to 1.25) 1.00 (0.90 to 1.15) 1.10 (0.90 to 1.30) 0.781
 Mid CS, % −23.9 (−28.3 to −20.2) −19.3 (−23.3 to −16.0) −26.2 (−30.3 to −18.8) 0.070
CMR
 LS (4ch), % −17.6 (−21.7 to −15.0) −12.2 (−14.2 to −10.0)* −17.2 (−20.4 to −15.0)† <0.001
 LS (2ch), % −20.0 (−23.6 to −15.0) −13.9 (−17.0 to −11.6)* −19.6 (−23.0 to −14.1)† <0.001
 LS (3ch), % −18.6 (−21.3 to −14.6) −13.9 (16.7 to −10.0)* −19.6 (−22.6 to −14.6)† <0.001
 GLS‡, % −18.8 (−22.3 to −15.4) −13.3 (−14.7 to −11.8)* −18.3 (−20.4 to −16.5)† <0.001
 Septal LS (4ch), % −19.3 (−22.0 to −14.8) −12.6 (−15.4 to −8.8)* −17.0 (−19.4 to −11.6)† <0.001
 Lateral LS (4ch), % −16.5 (−20.9 to −14.3) −11.3 (−14.9 to −9.6)* −18.8 (−22.4 to −14.9)† <0.001
 Lateral/Septal LS ratio (4ch) 0.99 (0.76 to 1.14) 0.93 (0.77 to 1.50) 1.19 (0.79 to 1.73) 0.204
 Mid CS, % −26.5 (−30.3 to −21.7) −23.2 (−26.1 to −18.5) −26.8 (−31.2 to −20.9)† 0.024
Values are median (interquartile range); P value across 3 groups by Kruskal–Wallis test and χ2 test for categorical value. ch indicates chamber;
CP, constrictive pericarditis; CS, circumferential strain; CMR, cardiac MR; GLS, global longitudinal strain; LS, longitudinal strain; and RCM,
restrictive cardiomyopathy.
*P<0.01 vs CP vs CP by Mann-Whitney tests.
†P<0.01 vs RCM by Mann–Whitney tests.
‡Global strain was derived by the average of 4ch and 2ch.

groups, as with echocardiography (P=0.20). Circumferential Diagnostic Value of GLS

Downloaded from http://ahajournals.org by on April 12, 2020

strain was also comparable between the CP and RCM
groups (P=0.07).
Echocardiography- and CMR-derived GLS were correlated
and did not significantly differ (mean difference, 0.7; limits of
agreement, −6.3% to 7.7%; Figure 3).
Multivariable Analysis
Using logistic regression analysis, we identified LVEF (odds
ratio [OR], 0.93; 95% CI, 0.87–0.99; P=0.036), LVMI (OR,
1.075; 95% CI, 1.039–1.112; P<0.001), e′ (OR, 0.45; 95%
CI, 0.29–0.68; P<0.001), respiratory septal shift (OR, 0.004;
95% CI, 0.000–0.043; P<0.001), and GLS (OR, 1.72; CI,
1.29–2.30; P<0.001) as echocardiographic predictors for dif-
ferentiating RCM from CP. On stepwise multivariable logistic
regression, GLS (OR, 2.44; CI, 1.06–5.60; P=0.034) and sep-
tal shift (OR, 0.002; 95% CI, 0–0.195, P=0.009) were inde-
pendent predictors for differentiating CP from RCM.
The receiver operating curve for GLS derived by echocardiog-
raphy and CMR for differentiating RCM from CP is shown in
Figure 4. The area under the curve was comparable for both
CMR- and echocardiography-derived GLS. Table 3 shows
summary measures of different aspects of logistic regression
models exploring the incremental value of GLS over known
echocardiographic and CMR discriminators. The overall
model fit (Akaike information criterion) was significantly
improved when echocardiographically derived GLS was
added individually to septal shift and e′ (P=0.001 and 0.0020,
respectively). Addition of GLS to septal shift and e′ also
resulted in significant improvement in continuous net reclas-
sification index (P<0.001 for both models) and integrated dis-
crimination improvement (P=0.0057 and 0.024, respectively).
Similarly, for CMR, the addition of GLS to respiratory
septal shift and pericardial thickness resulted in significant
improvement in Akaike information criterion (P=0.0022 and
0.0014, respectively), continuous net reclassification index
(P<0.001 for both models), and integrated discrimination
improvement (P=0.003 and <0.001, respectively).

Reproducibility
Interclass correlation coefficients are shown in Table 4. All
significant parameters demonstrated good reproducibility
when tested for interobserver and intraobserver variability.

Figure 2. Box plot showing median global longitudinal strain Discussion

(GLS) derived by echocardiogram (Echo) and cardiac MR (CMR:
A and B) in patients with constrictive pericarditis (CP), restrictive The diagnosis of CP34 and RCM35,36 by CMR has previ-
cardiomyopathy (RCM), and controls. ously been described. Aside from pericardial thickening,
 824  Circ Cardiovasc Imaging  September 2014
exaggerated respiratory-related LV-RV coupling, defined as
the difference in the maximal septal excursion between inspi-
ration and expiration, has been demonstrated to discriminate
CP from RCM and from normal hearts.13 However, the use of
CMR tissue tracking strain for differentiating LV mechanics in
CP and RCM has not been previously reported. In this study,
CMR-measured GLS had similar diagnostic value as echo-
cardiography-derived GLS for distinguishing CP from RCM.
The results of our work underscore the value of extracting
more diagnostic variables from a single modality, particularly
for conditions in which a constellation of diagnostic param-
eters (structural, mechanics, or hemodynamic) are needed to
recognize a specific pattern of disease such as CP and RCM.37
Such assessment increases the cost efficacy and conviction in
following a patient or identifying an interval change in disease
severity using a single technique.
Downloaded from http://ahajournals.org by on April 12, 2020
Myocardial Deformation in CP and RCM
Several previous investigations have shown that LV longitu-
dinal motion and diastolic lengthening velocities are impaired
to a greater degree in RCM compared with CP.21,38–40 We
have also previously demonstrated that longitudinal short-
ening strain is significantly reduced in RCM compared with
CP, reflecting marked impairment of endocardial function.3
Similar observations were reported by Kusunose et al,4 who
additionally illustrated the incremental value of the lateral-to-
septal strain ratio over regional velocities in discriminating CP
Figure 4. Area under the curve (AUC) to discriminate constrictive
pericarditis from restrictive cardiomyopathy using global longitu-
dinal strain (GLS) derived by echocardiography and cardiac MR
(CMR) images.
Figure 3. Correlation plot showing global
longitudinal strain (GLS) from echocar-
diogram (Echo) and cardiac MR (CMR) in
patients with constrictive pericarditis (CP)
and restrictive cardiomyopathy (RCM;
A). Brand–Altman plot of Echo and CMR
derived GLS (B).
from RCM. Because the free wall of the LV is tethered in CP,
LV lateral wall shortening strains are seen to be lower than
septal shortening strains. However, the prior studies have ana-
lyzed LV deformation using techniques that are not resolved
for the layers of the LV wall. In the present investigation, we
derived strain selectively from the subendocardial region of
the LV (average 4-pixel spatial resolution). The addition of
GLS using such a uniform strategy for both echocardiography
and CMR showed incremental value. In particular, the use of
GLS using cine CMR images showed incremental value over
pericardial thickness and assessment of LV-RV coupling by
septal shift during respiration. Such assessments may have
value for situations where gadolinium contrast enhancement
is not used or are relatively contraindicated.41
Both echocardiography and CMR data showed concordant
diagnostic information and incremental value for differen-
tiating CP from RCM. However, in contrast to previously
reported findings, we found that among CP patients the sep-
tal shortening strains were similar to those in the lateral wall,
and the endocardial shortening strains in the circumferential
direction were similar for all 3 groups. These differences from
prior studies may be related to the selective sampling from
the subendocardial region, which is spared from myocardial
tethering. Indeed, in a previous experimental model of peri-
cardial adhesion, we demonstrated that myocardial motion is
reduced because of pericardial adhesions over the epicardial
surface; however, the epi-to-endocardial gradient of myocar-
dial function remained unaltered, suggesting that endocardial
shortening proceeds relatively normally despite the process of
epicardial adhesions.42
Our results suggest that the pericardial–myocardial func-
tional relationships are more complex, and that a constrictive
pericardium may alter myocardial function differentially in
the myocardial layers. Perhaps with improved spatial resolu-
tion, in the near future, investigations could be undertaken to
resolve the transmural gradient of strains in patients with CP
and RCM.
Study Limitations
The present study has limitations. First, our patients were
evaluated in a tertiary referral center and were referred for
CMR; therefore, a selection bias may be present. Although
patients with structural heart disease were excluded from the
control population, these patients still had risk factors for heart
disease, such as hypertension and diabetes mellitus. This may
partly explain the lower GLS values in the control subjects.
 Amaki et al   LV Mechanics in Constriction and Restriction   825

Table 3.  Incremental Value of Global Longitudinal Strain Over Echocardiographic and CMR Variables for Differentiating
CP From RCM
Echo CMR
LV-RV vs LV-RV Plus GLS e′ vs e′ Plus GLS LV-RV vs LV-RV Plus GLS PT vs PT Plus GLS
Characteristic Difference P Value Difference P Value Difference P Value Difference P Value
AIC 8.49 0.0018 8.29 0.0020 8.13 0.0022 8.94 0.0014
∆AUC 0.053 0.0282 0.037 0.1056 0.046 0.0590 0.070 0.044
NRI (95% CI) 1.40 (1.03– 1.78) <0.001 0.94 (0.47 – 1.42) <0.001 1.16 (0.73 – 1.58) <0.001 1.16 (0.73 – 1.58) <0.001
IDI (95% CI) 0.09 (0.02 – 0.16) 0.0057 0.09 (0.00 – 0.18) 0.0240 0.11 (0.03 – 0.20) 0.0032 0.15 (0.07 – 0.24) <0.001
∆AUC indicates difference in the area under the receiver operating characteristics curve; AIC, Akaike information criterion; CI, confidence interval;
CMR, cardiac MR; GLS, global longitudinal strain; IDI, integrated discrimination improvement; LV-RV, left ventricular–right ventricular coupling; NRI, net
reclassification index; and PT, pericardial thickening.

Only 2 patients with CP had minimal myocardial involve- values from both techniques is similar to previously pub-
ment on delayed gadolinium-enhanced CMR. This may not lished data for both modalities.6,49 Moreover, the substantial
be representative of all subgroups of CP, such as radiation- concordance in patient characterization suggests potential
related CP where concomitant myocardial involvement may diagnostic value.
also affect myocardial deformation.43 Further studies that
provide more in-depth assessment of role of late gadolinium Conclusions
enhancement, a powerful technique to identify the pattern and CMR and echocardiography tissue tracking–derived LV
the extent of myocardial abnormalities for differentiating CP mechanics provide concordant information that is useful for
from RCM, are needed. differentiating CP from RCM.
The RCM group predominantly comprised amyloid car-
diomyopathy. Consequently, ventricular wall thickness was Acknowledgments
statistically different between the 2 groups, with different val- We acknowledge Chan Seok Park, MD, and Karen Modesto, MD, for
ues by CMR and echocardiography. The overestimation of LV the help in data analysis.
mass index by echocardiography may be related to geomet-
ric assumptions.44,45 CMR does not require cardiac geometry Disclosures
Downloaded from http://ahajournals.org by on April 12, 2020

assumptions, as opposed to linear measurements and formulae Dr Sengupta has received research support from TomTec Imaging
used in echocardiography. Despite these differences, the CP Systems GmbH. Dr Pedrizzetti has R&D relationship with TomTec
and RCM patients in our study had classic diagnostic features Imaging Systems GmbH. Dr Amaki is supported by a grant from
and are comparable to those in recent studies.4,46 Moreover, Japan Heart Foundation/Bayer Yakuhin Research Grant Abroad. The
other authors report no conflicts.
the use of both CMR and echocardiography helped to clearly
define these patient populations and enabled the exploration
of differences in LV mechanics. References
A final limitation concerns the quantitative estimates of 1. Vaitkus PT, Kussmaul WG. Constrictive pericarditis versus restrictive car-
diomyopathy: a reappraisal and update of diagnostic criteria. Am Heart J.
strain obtained from echocardiography and CMR. There 1991;122:1431–1441.
were differences in the type of imaging data, differences in 2. Shabetai R. Pathophysiology and differential diagnosis of restrictive car-
views (less foreshortening on CMR), and the frame rates diomyopathy. Cardiovasc Clin. 1988;19:123–132.
3. Sengupta PP, Krishnamoorthy VK, Abhayaratna WP, Korinek J,
obtained by both techniques. The feature tracking software Belohlavek M, Sundt TM III, Chandrasekaran K, Mookadam F, Seward
used for computing strains has been shown to work at lower JB, Tajik AJ, Khandheria BK. Disparate patterns of left ventricular me-
frame rates because of the combination of speckle-tracking, chanics differentiate constrictive pericarditis from restrictive cardiomy-
endocardial tissue–blood border (edge detection), and the opathy. J Am Coll Cardiol Cardiovasc Imaging. 2008;1:29–38.
4. Kusunose K, Dahiya A, Popović ZB, Motoki H, Alraies MC, Zurick AO,
periodicity of the cardiac cycle used in this feature-tracking Bolen MA, Kwon DH, Flamm SD, Klein AL. Biventricular mechanics in
algorithm (also used in the current study).47,48 Despite these constrictive pericarditis comparison with restrictive cardiomyopathy and
limitations, the correlation between tissue tracking strain impact of pericardiectomy. Circ Cardiovasc Imaging. 2013;6:399–406.
5. Frank H, Globits S. Magnetic resonance imaging evaluation of myocardial
and pericardial disease. J Magn Reson Imaging. 1999;10:617–626.
Table 4.   Interobserver and Intraobserver Variability 6. Azam S, Desjardins CL, Schluchter M, Liner A, Stelzer JE, Yu X, Hoit
BD. Comparison of velocity vector imaging echocardiography with
Interobserver Variability Intraobserver Variability magnetic resonance imaging in mouse models of cardiomyopathy. Circ
ICC 95% CI ICC 95% CI Cardiovasc Imaging. 2012;5:776–781.
7. Kempny A, Fernández-Jiménez R, Orwat S, Schuler P, Bunck AC, Maintz
Echo GLS 0.93 0.75–0.98 0.94 0.81–0.98 D, Baumgartner H, Diller GP. Quantification of biventricular myocardial
function using cardiac magnetic resonance feature tracking, endocardial
Echo mid CS 0.89 0.61–0.97 0.90 0.69–0.97
border delineation and echocardiographic speckle tracking in patients
CMR GLS 0.93 0.76–0.98 0.96 0.85–0.99 with repaired tetralogy of Fallot and healthy controls. J Cardiovasc Magn
CMR mid CS 0.94 0.79–0.98 0.96 0.96–0.99 Reson. 2012;14:32.
8. Schmidt R, Orwat S, Kempny A, Schuler P, Radke R, Kahr PC, Hellige A,
CI indicates confidence interval; CMR, cardiac MR; CS, circumferential strain; Baumgartner H, Diller GP. Value of speckle-tracking echocardiography
GLS, global longitudinal strain; and ICC, intraclass correlation coefficient. and mri-based feature tracking analysis in adult patients after fontan-type
 826  Circ Cardiovasc Imaging  September 2014
palliation [published online ahead of print December 24, 2013]. Congenit
Heart Dis. doi: 10.1111/chd.12156.
9. Stevens GR, Lala A, Sanz J, Garcia MJ, Fuster V, Pinney S. Exercise per-
formance in patients with pulmonary hypertension linked to cardiac mag-
netic resonance measures. J Heart Lung Transplant. 2009;28:899–905.
10. Sanz J, Kuschnir P, Rius T, Salguero R, Sulica R, Einstein AJ,
Dellegrottaglie S, Fuster V, Rajagopalan S, Poon M. Pulmonary arterial
hypertension: noninvasive detection with phase-contrast MR imaging.
Radiology. 2007;243:70–79.
11. Stevens GR, Garcia-Alvarez A, Sahni S, Garcia MJ, Fuster V, Sanz J.
RV dysfunction in pulmonary hypertension is independently related
to pulmonary artery stiffness. J Am Coll Cardiol Cardiovasc Imaging.
2012;5:378–387.
12. Talreja DR, Edwards WD, Danielson GK, Schaff HV, Tajik AJ, Tazelaar
HD, Breen JF, Oh JK. Constrictive pericarditis in 26 patients with
histologically normal pericardial thickness. Circulation. 2003;108:
1852–1857.
13. Francone M, Dymarkowski S, Kalantzi M, Rademakers FE, Bogaert J.
Assessment of ventricular coupling with real-time cine MRI and its value
to differentiate constrictive pericarditis from restrictive cardiomyopathy.
Eur Radiol. 2006;16:944–951.
14. Francone M, Dymarkowski S, Kalantzi M, Bogaert J. Real-time cine MRI
of ventricular septal motion: a novel approach to assess ventricular cou-
pling. J Magn Reson Imaging. 2005;21:305–309.
15. Vogelsberg H, Mahrholdt H, Deluigi CC, Yilmaz A, Kispert EM, Greulich
S, Klingel K, Kandolf R, Sechtem U. Cardiovascular magnetic resonance
in clinically suspected cardiac amyloidosis: noninvasive imaging com-
pared to endomyocardial biopsy. J Am Coll Cardiol. 2008;51:1022–1030.
16. Lester SJ, Ryan EW, Schiller NB, Foster E. Best method in clinical prac-
tice and in research studies to determine left atrial size. Am J Cardiol.
1999;84:829–832.
17. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka
PA, Picard MH, Roman MJ, Seward J, Shanewise JS, Solomon SD,
Spencer KT, Sutton MS, Stewart WJ; Chamber Quantification Writing
Group; American Society of Echocardiography’s Guidelines and
Standards Committee; European Association of Echocardiography.
Recommendations for chamber quantification: a report from the American
Downloaded from http://ahajournals.org by on April 12, 2020
Society of Echocardiography’s Guidelines and Standards Committee and
the Chamber Quantification Writing Group, developed in conjunction with
the European Association of Echocardiography, a branch of the European
Society of Cardiology. J Am Soc Echocardiogr. 2005;18:1440–1463.
18. Devereux RB, Reichek N. Echocardiographic determination of left ven-
tricular mass in man. Anatomic validation of the method. Circulation.
1977;55:613–618.
19. Hatle LK, Appleton CP, Popp RL. Differentiation of constrictive peri-
carditis and restrictive cardiomyopathy by Doppler echocardiography.
Circulation. 1989;79:357–370.
20. Appleton CP, Hatle LK, Popp RL. Demonstration of restrictive ven-
tricular physiology by Doppler echocardiography. J Am Coll Cardiol.
1988;11:757–768.
21. Ha JW, Ommen SR, Tajik AJ, Barnes ME, Ammash NM, Gertz MA,
Seward JB, Oh JK. Differentiation of constrictive pericarditis from restric-
tive cardiomyopathy using mitral annular velocity by tissue Doppler echo-
cardiography. Am J Cardiol. 2004;94:316–319.
22. Appleton CP, Hatle LK, Popp RL. Relation of transmitral flow velocity
patterns to left ventricular diastolic function: new insights from a com-
bined hemodynamic and Doppler echocardiographic study. J Am Coll
Cardiol. 1988;12:426–440.
23. Hurrell DG, Nishimura RA, Higano ST, Appleton CP, Danielson GK,
Holmes DR Jr, Tajik AJ. Value of dynamic respiratory changes in left and
right ventricular pressures for the diagnosis of constrictive pericarditis.
Circulation. 1996;93:2007–2013.
24. Talreja DR, Nishimura RA, Oh JK, Holmes DR. Constrictive pericarditis
in the modern era: novel criteria for diagnosis in the cardiac catheteriza-
tion laboratory. J Am Coll Cardiol. 2008;51:315–319.
25. Korinek J, Wang J, Sengupta PP, Miyazaki C, Kjaergaard J, McMahon
E, Abraham TP, Belohlavek M. Two-dimensional strain–a Doppler-
independent ultrasound method for quantitation of regional defor-
mation: validation in vitro and in vivo. J Am Soc Echocardiogr.
2005;18:1247–1253.
26. Korinek J, Kjaergaard J, Sengupta PP, Yoshifuku S, McMahon EM, Cha
SS, Khandheria BK, Belohlavek M. High spatial resolution speckle track-
ing improves accuracy of 2-dimensional strain measurements: an update
on a new method in functional echocardiography. J Am Soc Echocardiogr.
2007;20:165–170.
27. Augustine D, Lewandowski AJ, Lazdam M, Rai A, Francis J, Myerson
S, Noble A, Becher H, Neubauer S, Petersen SE, Leeson P. Global and
regional left ventricular myocardial deformation measures by magnetic
resonance feature tracking in healthy volunteers: comparison with tagging
and relevance of gender. J Cardiovasc Magn Reson. 2013;15:8.
28. Geyer H, Caracciolo G, Abe H, Wilansky S, Carerj S, Gentile F, Nesser
HJ, Khandheria B, Narula J, Sengupta PP. Assessment of myocardial
mechanics using speckle tracking echocardiography: fundamentals and
clinical applications. J Am Soc Echocardiogr. 2010;23:351–369; quiz
453.
29. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum
H, Gutgesell H, Reichek N, Sahn D, Schnittger I. Recommendations
for quantitation of the left ventricle by two-dimensional echocardiogra-
phy. American Society of Echocardiography Committee on Standards,
Subcommittee on Quantitation of Two-Dimensional Echocardiograms.
J Am Soc Echocardiogr. 1989;2:358–367.
30. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas un-
der two or more correlated receiver operating characteristic curves: a non-
parametric approach. Biometrics. 1988;44:837–845.
31. Kundu S, Aulchenko YS, van Duijn CM, Janssens AC. PredictABEL: an
R package for the assessment of risk prediction models. Eur J Epidemiol.
2011;26:261–264.
32. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating
the added predictive ability of a new marker: from area under the ROC
curve to reclassification and beyond. Stat Med. 2008;27:157–172; discus-
sion 207.
33. Pencina MJ, D’Agostino RB Sr, Steyerberg EW. Extensions of net reclas-
sification improvement calculations to measure usefulness of new bio-
markers. Stat Med. 2011;30:11–21.
34. Giorgi B, Mollet NR, Dymarkowski S, Rademakers FE, Bogaert J.
Clinically suspected constrictive pericarditis: MR imaging assessment of
ventricular septal motion and configuration in patients and healthy sub-
jects. Radiology. 2003;228:417–424.
35. Maceira AM, Joshi J, Prasad SK, Moon JC, Perugini E, Harding
I, Sheppard MN, Poole-Wilson PA, Hawkins PN, Pennell DJ.
Cardiovascular magnetic resonance in cardiac amyloidosis. Circulation.
2005;111:186–193.
36. Perugini E, Rapezzi C, Piva T, Leone O, Bacchi-Reggiani L, Riva L, Salvi
F, Lovato L, Branzi A, Fattori R. Non-invasive evaluation of the myo-
cardial substrate of cardiac amyloidosis by gadolinium cardiac magnetic
resonance. Heart. 2006;92:343–349.
37. Sengupta PP, Marwick TH, Narula J. Adding dimensions to unimodal car-
diac images. J Am Coll Cardiol Cardiovasc Imaging. 2011;4:816–818.
38. Ha JW, Oh JK, Ommen SR, Ling LH, Tajik AJ. Diagnostic value of
mitral annular velocity for constrictive pericarditis in the absence of
respiratory variation in mitral inflow velocity. J Am Soc Echocardiogr.
2002;15:1468–1471.
39. Sengupta PP, Mohan JC, Mehta V, Arora R, Pandian NG, Khandheria
BK. Accuracy and pitfalls of early diastolic motion of the mitral annulus
for diagnosing constrictive pericarditis by tissue Doppler imaging. Am J
Cardiol. 2004;93:886–890.
40. Rajagopalan N, Garcia MJ, Rodriguez L, Murray RD, Apperson-Hansen
C, Stugaard M, Thomas JD, Klein AL. Comparison of new Doppler echo-
cardiographic methods to differentiate constrictive pericardial heart dis-
ease and restrictive cardiomyopathy. Am J Cardiol. 2001;87:86–94.
41. Thomsen HS. How to avoid nephrogenic systemic fibrosis: current
guidelines in Europe and the United States. Radiol Clin North Am.
2009;47:871–5, vii.
42. Alharthi MS, Jiamsripong P, Calleja A, Sengupta PP, McMahon EM,
Khandheria B, Tajik AJ, Belohlavek M. Selective echocardiographic
analysis of epicardial and endocardial left ventricular rotational mechan-
ics in an animal model of pericardial adhesions. Eur J Echocardiogr.
2009;10:357–362.
43. Veinot JP, Edwards WD. Pathology of radiation-induced heart disease: a
surgical and autopsy study of 27 cases. Hum Pathol. 1996;27:766–773.
44. Missouris CG, Forbat SM, Singer DR, Markandu ND, Underwood R,
MacGregor GA. Echocardiography overestimates left ventricular mass:
a comparative study with magnetic resonance imaging in patients with
hypertension. J Hypertens. 1996;14:1005–1010.
45. Armstrong AC, Gidding S, Gjesdal O, Wu C, Bluemke DA, Lima JA.
LV mass assessed by echocardiography and CMR, cardiovascular out-
comes, and medical practice. J Am Coll Cardiol Cardiovasc Imaging.
2012;5:837–848.
46. Choi JH, Choi JO, Ryu DR, Lee SC, Park SW, Choe YH, Oh JK. Mitral
and tricuspid annular velocities in constrictive pericarditis and restrictive
 Amaki et al   LV Mechanics in Constriction and Restriction   827

cardiomyopathy: correlation with pericardial thickness on computed to-
mography. J Am Coll Cardiol Cardiovasc Imaging. 2011;4:567–575.
47. Onishi T, Saha SK, Ludwig DR, Onishi T, Marek JJ, Cavalcante JL,
Schelbert EB, Schwartzman D, Gorcsan J III. Feature tracking measure-
ment of dyssynchrony from cardiovascular magnetic resonance cine acqui-
sitions: comparison with echocardiographic speckle tracking. J Cardiovasc
Magn Reson. 2013;15:95.
48. Hor KN, Baumann R, Pedrizzetti G, Tonti G, Gottliebson WM, Taylor
M, Benson W, Mazur W. Magnetic resonance derived myocardial strain
assessment using feature tracking. J Vis Exp. 2011;48:2356.
49. Cho GY, Chan J, Leano R, Strudwick M, Marwick TH. Comparison of
two-dimensional speckle and tissue velocity based strain and valida-
tion with harmonic phase magnetic resonance imaging. Am J Cardiol.
2006;97:1661–1666.

Clinical Perspective
Echocardiography and cardiac MR (CMR) are complementary techniques that can aid in the differentiation of constrictive
pericarditis (CP) from restrictive cardiomyopathy (RCM). Previous studies have suggested that pericardial adhesions in CP
attenuate myocardial motion over the epicardial surface, while leaving endocardial shortening relatively well-preserved.
Conversely, patients with RCM show predominantly subendocardial involvement with marked reduction in endocardial
shortening. This study investigated the consistency and incremental diagnostic value of this disparate pattern of left ven-
tricular endocardial mechanics for differentiating CP from RCM. We compared CMR cine–based and echocardiography-
based strain measurements in 58 patients with a diagnosis of CP (n=28) or RCM (n=30), as well as in 34 control patients
who had no structural heart disease. Echocardiography- and CMR-derived global longitudinal strain were correlated, and
both techniques showed good reproducibility. Patients with CP had higher global longitudinal strain compared with patients
with RCM, and both techniques showed similar diagnostic value. In incremental logistic models, the addition of global
longitudinal strain individually to known echocardiographic and CMR discriminators such as respiratory septal shift, early
diastolic mitral annular velocity, and pericardial thickness resulted in significant improvement in reclassification and disease
discrimination. Our data suggest that CMR and echocardiography tissue tracking–derived left ventricular mechanics provide
comparable diagnostic information for differentiating CP from RCM. Further work should incorporate myocardial charac-
terization with late gadolinium enhancement, particularly for conditions in which a constellation of diagnostic parameters
(structural, mechanics, or hemodynamic) are needed to recognize a specific pattern of disease such as CP and RCM.
Downloaded from http://ahajournals.org by on April 12, 2020