OUP user menu

Regional right ventricular dysfunction in acute pulmonary embolism and right ventricular infarction

Franco Casazza, Amedeo Bongarzoni, Angela Capozi, Ornella Agostoni
DOI: http://dx.doi.org/10.1016/j.euje.2004.06.002 11-14 First published online: 1 January 2005

Abstract

Background: A normally contracting right ventricular apex associated to a severe hypokinesia of the mid-free wall (‘McConnell sign’) has been considered a distinct echocardiographic pattern of acute pulmonary embolism.

Objective: To evaluate the clinical utility of the ‘McConnell sign’ in the bedside diagnostic work-up of patients presenting to the Emergency Department with an acute right ventricular dysfunction due to pulmonary embolism or right ventricular infarction.

Design: Among 201 patients, consecutively selected from our clinical database and diagnosed as having massive or submassive pulmonary embolism or right ventricular infarction, 161 were suitable for an echocardiographic review of regional right ventricular contraction and were included in the study. There were 107 cases with pulmonary embolism (group 1) and 54 cases with right ventricular infarction (group 2). All echocardiographic studies were randomly examined by two experienced and independent echocardiographers, blinded to the patient diagnosis and without Doppler informations.

Results: The McConnell sign was detected in 75 of 107 patients in group 1 (70%) and in 36 of 54 patients in group 2 (67%); the finding was absent in 32 cases in group 1 and in 18 cases in group 2 (P=0.657). The sensitivity, specificity, positive and negative predictive values of the McConnell sign for the diagnosis of pulmonary embolism were respectively 70, 33, 67 and 36%.

Conclusions: In a clinical setting of patients with acute right ventricular dysfunction the McConnell sign cannot be considered a specific marker of pulmonary embolism.

Keywords
  • Echocardiography
  • Right ventricular wall motion
  • Pulmonary embolism
  • Right ventricular infarction

Introduction

Echocardiography is recognized as an important tool in the rapid diagnostic work-up and in prognostic stratification of pulmonary embolism (PE).1–6 In association with a high or intermediate clinical probability, the findings of right ventricular (RV) dysfunction and pulmonary hypertension are crucial in increasing the clinical suspicion of PE and in initiating the appropriate therapy, including thrombolysis in unstable cases.7,8 Echocardiographic approach is especially worthwhile in the Emergency Department or Coronary Care Unit to investigate patients with circulatory failure of uncertain cause, when the electrocardiogram is not conclusive.

In 1996, by means of quantitative and qualitative analyses, McConnell et al. described in pulmonary embolism a ‘distinct echocardiographic pattern of RV dysfunction’, characterized by a severe hypokinesia of RV mid-free wall, with a normal contraction of the apical segment9 (Fig. 1a,b). In comparison with several conditions involving the RV, the finding showed a 77% sensitivity and a 94% specificity, with a positive and negative predictive values of 71 and 96%, respectively The authors therefore suggested that the presence of this echocardiographic sign should raise the level of clinical suspicion for the diagnosis of PE and prompt further diagnostic evaluation. To the best of our knowledge, no other study has been carried out to confirm these single observations; yet this abnormality has become widely known as the ‘McConnell sign’ and holds a position in the scenario of 2D echocardiographic hallmarks of PE10,11 The aim of the present study was to evaluate the diagnostic yield of the ‘McConnell sign’ in patients exclusively affected by an acute RV dysfunction.

Figure 1

Echocardiographic 4-chamber views of a patient with massive pulmonary embolism and right ventricular dysfunction in end-diastole (a); and in end-systole (b). Arrows indicate normally contracting apical segments in a globally hypokinetic right ventricle (Mc Connell sign). RV, right ventricle; RA, right atrium; LV, left ventricle; and LA, left atrium.

Methods

A total of 201 patients with massive or submassive PE12 or a first RV infarction were consecutively selected from our clinical database between January 1998 and May 2003. The echocardiographic recordings, obtained within 24 h of admission (mainly within 2 h), were initially reviewed by one of us in order to select appropriate views from videotapes for a subsequent blind evaluation. Forty cases were excluded due to: poor quality images in 20 cases, absence of a clear RV hypokinesia in 12 cases with RV infarction, free floating right heart thrombi in nine cases. A total of 161 patients were therefore included in this retrospective study: 107 cases with PE, all confirmed by spiral CT (group 1) and 54 cases with RV infarction, confirmed by ST-segment elevation ≥1 mm in lead V4r (group 2). No patient in group 2 had known arterial or venous pulmonary hypertension. In group 1, there were 44 males and 63 females with a mean age of 68.4±23 years (range 24–88 years); in group 2 there were 32 males and 22 females with a mean age of 69.2±10 years (range 48–95 years).

Echocardiograms were randomly evaluated by two experienced and independent echocardiographers, blinded to the patient diagnosis and without Doppler informations. A qualitative wall motion analysis of the presence or absence of the Mc Connell sign was performed in the apical four-chamber view for ≥3 cardiac cycles and discrepancies were resolved by consensus. Interobserver agreement was 89%. For categorical variables, the comparison between the two groups was performed with chi square test.

Results

A pattern of normal wall motion of the apex and abnormal wall motion in the mid-free wall of the right ventricle was detected in 75 of 107 patients (70%) in group 1 and in 36 of 54 patients (67%) in group 2; the finding was absent in 32 cases in group 1 and in 18 cases in group 2 (Table 1). The sensitivity, specificity, positive and negative predictive values of the McConnell sign for the diagnosis of pulmonary embolism were respectively 70, 33, 67 and 36%.

View this table:
Table 1

Prevalence of McConnell sign in patients with acute PE and RV infarction

Pulmonary Embolism N° (%)RV infarction N° (%)Patients N° (%)
McConnell sign present75 (70%)36 (67%)111 (69%)
McConnell sign absent32 (30%)18 (33%)50 (31%)
Total107 (100%)54 (100%)161 (100%)
  • Chi-square statistic: χ2=0.197 (P=0.657).

Discussion

Pulmonary embolism and RV infarction may present with similar clinical and electrocardiographic findings: chest pain, sweatiness, hypotension or shock, ST-segment elevation and/or q waves in inferior leads or in V1. In both diseases 2D-echo may show enlargement of right heart chambers, right ventricular hypokinesia, iv septal diskinesia and inferior vena cava dilation, while ischemic left ventricular inferior wall abnormalities may be subtle when infarction predominantly involves the right ventricle. CW Doppler helps differentiate the two conditions by demonstration of an increased RV-RA pressure gradient in massive or submassive PE, and of a normal or low gradient in RV infarction. In these sometimes confounding conditions, a new echocardiographic pattern of RV contraction, highly specific for the diagnosis of PE, would be very useful, especially in cases of trivial tricuspid insufficiency and/or difficult estimation of velocities and gradients across the tricuspid valve. Unfortunately the results of our study show that the McConnell sign is unable to differentiate acute PE from RV infarction. Actually, this sign was present in 70% of PE and in 67% of RV infarctions, with a very low specificity and negative predictive value. The presence or the absence of the sign cannot therefore confirm or exclude the diagnosis of PE. Indeed our data are at variance with those of McConnell et al.,9 who compared a small group of PE patients with a very heterogeneous population of 85 patients collected on the basis of the presence of echocardiographic signs of RV dysfunction. Causes of RV dysfunction were mainly chronic diseases such as ischemic or dilated cardiomyopathies (n=30), chronic pulmonary hypertension (n=11), valvular heart disease (n=9), congenital heart disease (n=7). Just seven cases of acute RV dysfunction were included, namely six RV infarctions and one acute respiratory distress syndrome. Since the sole false positive case, incidentally quoted in the paper, was the only one suffering from the acute respiratory distress syndrome, the authors stated that ‘patients with other causes of an acute increase in RV afterload may have similar findings’. The results of the present study demonstrate that not only causes of acute increase in RV afterload other than PE may show similar findings, but also diseases directly involving the right ventricle, such as RV infarction, may also account for the onset of the McConnell sign. We are therefore still looking for a specific echocardiographic pattern of acute PE.

Three mechanisms were advocated by McConnell9 to explain the pathogenesis of their findings in pulmonary embolism: tethering of the RV apex to a contracting hyperdynamic left ventricle, a systolic bulging localized to the mid-free wall due to an abrupt increase in afterload which spares the apex and, third, a segmental ischemia of the RV free wall. According to our observations, the first two mechanisms seem quite unlikely because: (1) the RV apex motion is a real inward movement and not just a tethering; and (2) in RV infarction a sudden increase in afterload does not develop in spite of the presence of a normal contracting apex in 67% of our cases. We therefore believe that the most consistent mechanism to explain the McConnell sign could be ischemia: a functional ischemia of the RV free wall in PE, due to a transient increase in wall stress, and a necrotic involvement of the RV free wall in RV infarction, due to mechanical obstruction of the right coronary artery. Yet, the exact reason why the apex is the sole normally contracting segment in an otherwise hypokinetic right ventricle both in acute PE and RV infarction remains still unclear.

In conclusion, the present study confirms the fairly good sensibility but not the high specificity of the McConnell sign for the echocardiographic diagnosis of acute pulmonary embolism, since a normal contracting RV apex could be detected also in two thirds of patients with RV infarction. This regional pattern of RV contraction cannot therefore be used in the bedside diagnostic work-up of critically ill patients with acute right ventricular dysfunction to raise the clinical suspicion of acute cor pulmonale.

References

View Abstract