1. Background
Clinically, presentations of acute pulmonary embolism typically include acute progressive shortness of breath with associated pleuritic chest pain, with or without cough/hemoptysis. Physiologic derangements manifest as tachycardia, syncope/presyncope, and in severe cases may result in systemic hypotension, hemodynamic instability, or shock. Physiologically, this phenomenon occurs due to a sudden impairment of blood flow from the right heart through the lungs to the left heart with a consequent reduction in systemic blood flow. Echocardiographically, the abrupt increase in trans-pulmonary gradient leads to acute right ventricular dysfunction (often with the classic distribution of McConnell’s sign), with or without increased right ventricular systolic pressure. Obstructed trans-pulmonary flow leads to a decrease in left atrial filling pressure, with consequently reduced left ventricular filling that may result systemic hypo-perfusion and/or cardiogenic shock.
Acute pulmonary embolism (PE) presentations carry a mortality rate as a high as 25% in the first 30 days [
1]. As such, prompt and thorough evaluation and management is crucial in minimizing the potential morbidity and mortality associated [
2]. The “gold standard” diagnosis of acute pulmonary embolism is via computed tomography pulmonary angiogram (CTPA), to anatomically assess clot burden and distribution [
3]. Ventilation perfusion scintigraphy (V/Q scan) has comparative utility, via assessment of the burden of mismatched perfusion defects, as a diagnostic and prognostic measure [
4]. The hemodynamic functional effect of PE however is typically assessed by echocardiography [
5]. Specifically, patients are assessed for evidence of right ventricular (RV) dysfunction/enlargement and/or significant pulmonary hypertension, manifest as elevated right ventricular systolic pressures [
6]. Based on European Society of Cardiology Guidelines 2014 [
7], immediate thrombolysis is both justified and recommended in the setting of significant RV dysfunction and associated cardiogenic shock or hemodynamic instability. This includes instances where CTPA may not be feasible prior to echocardiographic evaluation.
In this setting, significant right ventricular dysfunction may be evidenced by reduced right ventricular function, e.g., McConnell’s sign (RV free wall akinesia with apical wall hypercontractility) [
8], reduced TAPSE (tricuspid annular plane systolic excursion) [
9], reduced tricuspid annular Doppler Tissue Imaging S’-velocity (RV S’) [
10], and, more recently, reduced RV free wall longitudinal strain [
11,
12]. All of these are functional assessments, which suggest right ventricular dysfunction in the setting of acutely increased afterload. Importantly, however, the absence of RV dysfunction on initial echocardiography does not definitively exclude significant embolic burden in hemodynamically, normotensive stable patients [
7]. Furthermore, these measurements of RV dysfunction may be confounded by RV pathology (for example with RV wall infarct, or pre-existing pulmonary hypertension) [
13]. Conversely, in hemodynamically unstable patients with suspected high risk PE, the absence of echocardiographic evidence of RV dysfunction or overload can effectively exclude PE [
5].
The pathophysiologic underpinning of the hemodynamic perturbation of acute PE is
pre-capillary obstruction to trans-pulmonary flow. The novel parameter ePLAR (the echocardiographic Pulmonary to Left Atrial Ratio) has been validated as a non-invasive surrogate to trans-pulmonary gradient [
14]. ePLAR, which assesses the relationship between right ventricular systolic pressure and left atrial pressure via the formula,
has been shown to correlate well with invasively-derived trans-pulmonary gradient (TPG). ePLAR non-invasively differentiates between
pre-capillary and
post-capillary pulmonary hypertension in patients being investigated for consideration of specific pulmonary vasodilator therapies. Higher ePLAR values suggest higher trans-pulmonary gradients and
pre-capillary pulmonary hypertension. Lower ePLAR values suggest higher left atrial pressures with minimal increase in trans-pulmonary gradient and
post-capillary pulmonary hypertension (see
Figure 1 [
14]). It has also been shown that with age, as left heart filling pressures naturally rise, ePLAR declines linearly. Thus, consideration of normal vs. abnormal values in a given patient must take into consideration age.
It is hypothesized that ePLAR (as a measure of trans-pulmonary gradient) may have a higher diagnostic yield than other previously utilized echocardiographic markers of elevated pulmonary pressures (as assessed by TRVmax/RVSP) and RV dysfunction (reduced TAPSE or RV S’) in detecting hemodynamic perturbations in patients with acute sub-massive pulmonary embolism. Abnormally high ePLAR levels would likely be comprised of a small (if any) increase in TRVmax with significantly reduced mitral E/e’ (consistent with reduced left atrial filling pressures). Increased ePLAR in these patients will suggest increased TPG even in the absence of actual elevated right heart pressures or right ventricular dysfunction.
3. Results
There were 223 patients referred for acute echocardiography with the diagnosis of possible/probable/proven sub-massive pulmonary embolism. Of these, 88 patients were subsequently shown to either not have sub-massive pulmonary emboli (32 patients with small/lobar pulmonary emboli and 56 with negative CTPA/VQ scans for PE) or had a delay of >72 h from CTPA/VQ to echocardiography. A further 25 patients were excluded with incomplete datasets. The remaining 110 patients (64 male, aged 57.4 ± 17.6 years) were shown to have bilateral pulmonary emboli (
n = 97) or saddle emboli (
n = 13) by CTPA/VQ scan. Echocardiography was performed at 0.3 ± 0.9 days from CTPA/VQ (range −3 to +3 days). Demographics of the patients are shown in
Table 1, with statistical comparison to the aged matched control cohort.
Based on recent guideline definitions of pulmonary hypertension [
12], a TRV
max > 2.9 m/s was defined as indicative of elevated right ventricular systolic pressure. Based on the previous validation of ePLAR [
14], values > 0.28 m/s were considered to be indicative of elevated TPG. Normal values for TAPSE were considered excursions distances ≥17 mm. Normal values for tricuspid annular Doppler Tissue Imaging systolic velocity (RV S’) were considered ≥9.5 cm/s [
16].
Using these cut-off values, patients were separated analyzed in dichotomous assessments of ePLAR vs. TRV
max (
Table 2 and
Figure 3A), ePLAR vs. TAPSE (
Table 3 and
Figure 3B), and ePLAR vs. RV S’ (
Table 4 and
Figure 3C). For each respective analysis, four groups were defined: Group 1 (abnormal TRV
max or TAPSE or RV S’, elevated ePLAR ≥ 0.28 m/s), Group 2 (normal TRV
max or TAPSE or RVS’, elevated Eplar ≥ 0.28 m/s), Group 3 (abnormal TRV
max or TAPSE or RVS’, normal ePLAR < 0.28 m/s), and Group 4 (normal TRV
max or TAPSE or RVS’, normal ePLAR < 0.28 m/s).
Using the standardized cut–off values listed above, sensitivity and specificity data were calculated for the prediction of pulmonary embolism, in comparison to the age matched control cohort. As shown in
Table 5, ePLAR performed most strongly with sensitivity 72% (confidence interval (CI) 62–80%), specificity 66% (CI 57–75%), positive predictive value 68% (CI 62–74%), and negative predictive value 70% (CI 63–75%). TRV
max demonstrated sensitivity 29% (CI 21–39%), specificity 98% (CI 94–100%), positive predictive value 94% (CI 80–98%), and negative predictive value 58% (CI 57–70%). Right ventricular function as assessed by TAPSE demonstrated sensitivity 22% (CI 14–33%), specificity 85% (CI 77–91%), positive predictive value 52% (CI 36–66%), and negative predictive value 61% (CI 58–65%). Right ventricular function as assessed by RV S’ demonstrated sensitivity 13% (CI 6–22%), specificity 85% (CI 76–91%), positive predictive value 37% (CI 22–55%), and negative predictive value 57% (CI 55–60%). Comparison of the predictive power of each parameter was assessed using ROC curves (see
Figure 4). ePLAR performed strongly (AUC 0.74,
p < 0.05) compared with TRV
max (AUC 0.62,
p < 0.05), TAPSE (AUC 0.54,
p = 0.33), and RV S’ (AUC 0.76,
p < 0.05).
4. Discussion
Sub-massive acute pulmonary embolism is associated with significant morbidity and mortality, in both the acute and chronic setting. It has been well documented that early treatment with anticoagulation, and in severe cases thrombolysis, is critical in the care of these patients [
2]. Anatomical assessment of clot location and burden is well achieved with CTPA and nuclear medicine V/Q scan in most cases. Echocardiography is unlikely to ever replace these tests for definitive diagnosis of PE [
5,
17].
However, echocardiography has two major roles in this disease spectrum [
7]. Firstly, in some situations, these anatomic tests are logistically not feasible in the emergent setting, and echocardiography is used to assess for “supportive” or “surrogate” evidence to confirm or refute the clinical suspicion of PE [
18]. Indeed, while clearly not the optimal test for the definitive diagnosis of PE, echocardiography is frequently requested (often before CTPA or V/Q scans can logistically be obtained) to guide emergent therapy. Clinicians have long recognized the constellation of visual cues (McConnell’s sign with preserved apical RV function [
8] and reduced free wall function), in the setting of minimally elevated TRV
max/RVSP values, and underfilled left-sided chambers. Quantitative measures of RV function help reinforce these subjective impressions of the constellation of findings of significant PE [
10].
In this study, patients with proven acute PE had mean TAPSE values of 21.1 mm and RV S’ of 13.5 cm/s, with 59/110 and 69/110 false negative results, respectively. As a binary diagnostic tool for RV dysfunction, TRVmax performed poorly, with only 32 out of 110 patients having appropriately elevated TRVmax ≥ 2.9 m/s. Clinically, all three parameters had very low sensitivity for the detection of acute PE (TRVmax 29%, TAPSE 22%, RV S’ 13%), though high specificity. ePLAR on the other hand was found to have much higher sensitivity (72%) with modest specificity (66%), 31/110 results being falsely normal. Although RV S’ was found to have a greater ROC AUC than ePLAR, ePLAR was found to have the highest negative predictive value of all parameters, and much higher positive predictive value than RV S’. Clinically, these data suggest that ePLAR, as a marker of impaired trans-pulmonary flow, is much more likely to register as abnormal than measures of right heart pressure (RVSP/TRVmax), or right heart function (TAPSE and RV S’), with a negative predictive value superior to these other parameters.
Secondly, and more logically, echocardiography in the setting of suspected or proven acute PE plays a major role in risk stratification and the assessment of hemodynamic distress of the right-sided circulation. Current Guidelines (both European and American) for thrombolysis or catheter-based intervention for PE recommend echocardiographic assessment of RV dysfunction [
7,
19], and subsequent treatment in the setting of RV dysfunction + hemodynamic instability, with or without CTPA or V/Q scan confirmation of PE [
7]. Prognostic advantage has been demonstrated when these criteria lead to lysis/intervention [
20]. However, in this study, only 32% had reduced TAPSE while 70% had abnormal ePLAR. It is hypothesized that ePLAR is a subtler indicator of hemodynamic disturbance. Future studies may well show that using this marker of impaired transpulmonary flow may better predict response to therapy.
Finally, this study offers mechanistic insights into the hemodynamics of acute PE, and, specifically, the physiologic underpinning of systemic hypotension and shock [
21]. In this cohort, the mitral E/e’ (as a marker of left atrial pressure) was substantially lower than age-matched controls (8.2 ± 3.8 vs. 10.8 ± 5.1,
p = 0.013). Even though, as a group, these pulmonary embolism patients were not pulmonary hypertensive (TRV
max 2.61 ± 0.61 m/s, RVSP 34.18 ± 13.49 mmHg), ePLAR was markedly elevated, suggesting for the first time non-invasively, elevated TPG levels. Interestingly, for each of the three echocardiographic parameters assessed dichotomously with ePLAR (
Table 2,
Table 3 and
Table 4), patients in Group 2 (ePLAR true positive, TRV
max/TAPSE/RV S’ false negative) were significant younger than patients in Group 1 or 3 (TRV
max/TAPSE/RV S’ true positive). This may suggest an element of age related physiologic compensation or pulmonary vasculature reserve in the setting of an acute obstruction to trans-pulmonary flow [
22]. Consequently, it is hypothesized that ePLAR may offer mechanistic understanding into normalization of trans-pulmonary flow in acute PE following therapy [
23]. Follow up echocardiographic analysis may also provide insight into the long-term physiologic impacts of acute PE, in particular in the setting of pre-existing cardiac systolic or diastolic dysfunction.