By Michael H. Crawford, MD, Editor

SYNOPSIS: The five Mayo Clinic Doppler echocardiographic criteria for the diagnosis of constrictive pericarditis (CP) were tested in a group of surgically confirmed CP compared to a group of cardiac biopsy proven amyloid restrictive cardiomyopathy (RC). Interventricular septal bounce and an elevated mitral annular medial e’ were found to be the most accurate criteria, especially if used together.

SOURCE: Qamruddin S, Alkharabsheh SK, Sato K, et al. Differentiating constriction from restriction (from the Mayo Clinic Echocardiographic Criteria). Am J Cardiol 2019;124:932-938.

The differentiation of cardiac constriction vs. cardiac restriction is important because constriction could be curable. The Mayo Clinic has promulgated echocardiographic criteria for this diagnostic distinction based on their experience. The purpose of this study from the Cleveland Clinic was to evaluate the diagnostic accuracy of these criteria in 107 patients with surgically proven constrictive pericarditis (CP) as distinguished from 30 patients with restrictive cardiomyopathy (RC) from biopsy-proven cardiac amyloidosis. Patients with severe mitral or tricuspid regurgitation or mitral valve prostheses were excluded, but patients with atrial fibrillation were not. Per the Mayo criteria, the following five Doppler echo variables were measured: interventricular septal shift with respiration, exaggerated respiratory variation in mitral inflow Doppler E velocity, tissue Doppler medial mitral annular e’ velocity, ratio of medial to lateral Doppler mitral annular e’ velocity, and exaggerated hepatic vein diastolic forward flow reversal during expiration on pulsed Doppler.

Respiratory septal shift was more frequent in CP patients compared to RC patients (92% vs. 27%; P = 0.001). Respiratory percent change in mitral E was higher in CP vs. RC (27 ± 18% vs. 12 ± 11%; P = 0.001). Medial mitral e’ also was higher in CP (11 ± 3 vs. 4 ± 1 cm/s; P = 0.001) as was the ratio of medial to lateral mitral e’ (1 ± 0.3 vs. 0.8 ± 0.2; P = 0.001). Finally, hepatic vein diastolic forward flow in expiration was slower in CP (24 ± 10 vs. 35 ± 15 cm/s; P = 0.001). A multivariate analysis showed that of five variables, only medial e’ velocity ≥ 9 cm/s (odds ratio, 8.0; 95% confidence interval, 2.4-28; P = 0.001) was independently associated with CP. Ventricular septal respiratory shift carried the greatest sensitivity (94%), and medial e’ velocity ≥ 9 cm/s carried the highest specificity (100%). Combining the two gave a sensitivity of 80% and a specificity of 92%. The optimal cutoff for medial e’ was ≥ 7 cm/s (area under the curve, 0.99). The authors concluded that CP can be confused with RC, and highly accurate Doppler echo tools can help diagnose this potentially reversible condition.


This study largely replicates the Mayo experience and validates the utility of Doppler echo to diagnose CP, with a few differences due to patient inclusion and exclusion criteria. The Cleveland Clinic study population excluded those with severe tricuspid valve regurgitation and included patients with atrial fibrillation. The key information here is the prevalence of the five Mayo Doppler echo criteria in CP, as RC from amyloidosis should be diagnosed easily through other parameters. Whether other forms of RC without thick left ventricular myocardial walls would give the same results is not discernable from this study. The results show that the most sensitive Doppler echo finding for CP is respiratory shift of the interventricular septum. This so-called septal bounce usually is the finding that prompts the consideration of CP because it is easily visible without making measurements. Unfortunately, it is not very specific and can be seen with conduction abnormalities and postcardiac surgery, which are more common conditions. However, it should grab one’s attention so that one considers the other Doppler echo findings of CP.

The results of the Cleveland Clinic study suggest cardiologists look at the medical e’ velocity, because if it is ≥ 9 cm/s and there is a septal bounce, the positive predictive valve (PPV) is high (100% in this study). Also, finding that the medial e’ is greater than the lateral annular e’, or so-called annulus reversus, has been considered a useful finding, because it suggests that the septal myocardium is unfettered and the lateral myocardium is tethered to the pericardium. However, in this study, this finding carried only modest sensitivity (67%) and specificity (73%). Other findings, such as late expiratory diastolic flow reversed in the hepatic veins and exaggerated respiratory variation in mitral inflow E velocity, also showed modest performance. However, combining septal shift, medial e’, and hepatic flow increased specificity to 97%, but was not that sensitive (70%).

There were some weaknesses in the study that could have affected the results. For example, high-quality hepatic vein velocity recordings were obtained in only 79% of subjects. Also, a medial e’ of ≥ 9 cm/s can be seen in otherwise normal individuals, so this finding alone is not that useful — unless one suspects CP. The bottom line is that there are no perfect Doppler echo measurements that will diagnose CP accurately. Thus, a high index of suspicion is necessary, with a careful look at the five criteria and with an emphasis on septal bounce and medial mitral annular e’ as a starting point. Suggestive findings in the other three criteria should help, but, unfortunately, are not completely reliable. Ultimately, other imaging modalities are needed to confirm the diagnosis.