The right ventricle is no longer forgotten; rather, it is now simply misunderstood. This is a leap forward compared with a time as recent as 5 years ago, but being misunderstood is still not a desirable state. Our approach to the assessment of the right ventricle reminds us of the Indian parable of the blind men and the elephant, captured in poetic form by John Godfrey Saxe in 1872. Each man palpates a different but single part of the elephant, and they then compare notes. Although each is accurate in describing the part that he examined, none of them describes what would sound to a sighted man as an elephant: “Though each was partly in the right, and all were in the wrong.”
As we have benefitted from more evolved echocardiographic metrics to describe the size and function of the right ventricle, the focus has now begun to shift toward the harmonious integration of these metrics. Given the ultimate mandate of providing an accurate, cohesive assessment of the right ventricle, there are several factors relating to the echocardiographic approach and the complex nature of the right ventricle that must be carefully considered.
As we are aware, compared with the prolated ellipse or “bullet shape” of the left ventricle, the right ventricle is difficult to model by a single simple geometric shape. The right ventricle is made up of three functional regions: the inflow region, the body/apex, and the outflow tract. The sum of these parts forms a shape that loosely resembles a pyramid (or a seated elephant facing toward the side), but this pyramid would hardly stand on its own, and it defies simple modeling. Further complicating our ability to describe the right ventricle is the observation that the right ventricle remodels differently, both geometrically and functionally, in different disease states. For example, in acute pressure overload states such as pulmonary embolism, and in contradistinction to chronic pulmonary hypertension, the right ventricular (RV) apex contracts (or moves) disproportionately to the adjoining regions of the right ventricle. Notwithstanding, the observed pattern of RV dilation or dysfunction for a given pathologic state is often highly unpredictable.
The muscle bands of the right ventricle are arranged in a different configuration as opposed to the left ventricle. The right ventricle contains two layers of fibers, the predominant longitudinal and less important radial fibers, and lacks a layer of spiral fibers. This contributes to the very noticeable base-to-apex shortening and less noticeable inward movement. The left ventricle appears to contract toward a cavitary centroid, while the right ventricle contracts from base to apex. Finally, given the low impedance against which the right ventricle contracts, the RV wall is much thinner than that of the left ventricle. This results in additional challenges in the echocardiographic evaluation of RV systolic function, as we discuss below. The net effect of these features is that the untrained right ventricle is well suited to serve as a volume-loaded pump and ill-suited to serve as a pressure-loaded pump.
The echocardiographic parameters that we use to measure RV systolic function reflect these RV characteristics, but each has its own limitations. Systolic annular velocity (S′) and tricuspid annular plane systolic excursion (TAPSE; also termed “tricuspid annular motion”) reflect the velocity and the integrated velocity over time of the RV free wall motion, one of the most highly recognizable cardiac motions. These two measures, however, are angle dependent and also assume that the motion of this basal-most region of the right ventricle is representative of the function of the entire chamber. Fractional area change (FAC) is a two-dimensional surrogate of RV volume change but is technically more challenging to learn and only partially representative of the entire right ventricle. Three-dimensionally derived RV ejection fraction (RVEF) avoids many of the geometric limitations of the other measures but requires technically superb image quality and has the same fundamental shortcoming as left ventricular ejection fraction in representing true contractility. Unlike left ventricular ejection fraction, RVEF by echocardiography lacks prognostic data. Despite the availability of these parameters, the large majority of clinical echocardiographic assessments of the right ventricle are limited to the “eyeball” or visual estimate. Visual acuity aside, it is evident that one person’s eyeball is not the same as another’s.
To standardize the approach and establish minimal standards, the American Society of Echocardiography published guidelines for the echocardiographic assessment of the right heart. These guidelines were adopted by the European Association of Echocardiography and Canadian Society of Echocardiography and have become an internationally recognized document. The document strongly endorses an integrative approach to the assessment of the right heart, recommending several complementary parameters to measure RV size and function. These parameters were recommended on the basis of robustness of data available and technical advantages and limitations. A number of quantitative parameters were recommended, including TAPSE, S′, FAC, and right-sided index of myocardial performance (also known as the RV myocardial performance index or the RV Tei index). It was further recommended that at least one measure of quantitation be present in the global assessment. What was lacking in the available body of evidence and consequently in our recommendations at that time was a systematic approach to integrate and resolve situations in which the quantitative parameters were not necessarily concordant.
In this issue of JASE , Maffessanti et al. present the results of their cohort study involving 42 patients with severe mitral valve prolapse or flail undergoing isolated mitral valve repair surgery at their institution. The study protocol involved transthoracic echocardiography 1 day before surgery and again 6 months after surgery, and they measured TAPSE, S′, FAC, and longitudinal and radial strain by speckle tracking (using a TomTec algorithm; TomTec Imaging Systems, Munich, Germany). Their reference standard was three-dimensional (3D) RVEF (also using a TomTec algorithm). The primary research question was to what extent, if any, these parameters of RV function would change from the preoperative to the postoperative echocardiographic assessment and whether they would differ between patients and a group of 20 normal controls. The study results are summarized in Table 1 .
| Echocardiographic parameter | Preoperative value vs normal controls | Preoperative value vs postoperative value ∗ |
|---|---|---|
| TAPSE | ↔ | ↓↓ |
| S′ | ↔ | ↓↓ |
| FAC | ↔ | ∼↓ |
| Longitudinal strain | ↔ | ↓ |
| Radial strain | ↔ | ↓ |
| 3D RV volumes | ∼↑ | ↔ |
| 3D RVEF | ∼↓ | ↔ |
∗ TAPSE decreased by 9 mm, S′ by 5 cm/sec, FAC by 4% (although the mean value of 39% was still within the normal range), longitudinal strain by 9%, and radial strain by 7%.
An adjoining correlation matrix analysis attempted to correlate the preoperative values of each parameter with the postoperative changes observed. This analysis was limited by multiple hypothesis testing and relatively weak correlations across the board ( r < 0.50), and the authors acknowledged that postoperative changes in RV function could not be predicted by preoperative values on the baseline echocardiogram. Otherwise said, the mechanism of the postoperative changes did not appear to be related to preexisting RV dysfunction. Although other groups have not necessarily found a decline in RV FAC after cardiac surgery, the decline in RV longitudinal parameters is strikingly similar to studies of patients undergoing mitral valve surgery (a 9-mm decrease in TAPSE and a 6 cm/sec decrease in S′) and coronary artery bypass graft surgery (a 10-mm decrease in TAPSE, a 7 cm/sec decrease in S′, and a 9% decrease in strain), which also found that preexisting RV dysfunction or right coronary artery disease did not explain the changes in RV function. The lack of decline in RVEF is also similar to a magnetic resonance–based study of patients undergoing coronary artery bypass grafting, which found that use of cardiopulmonary bypass or off-pump technique did not correlate with changes in RV size or function.
Other potential mechanisms to explain RV dysfunction were not specifically evaluated, but the authors sided with the pericardial disruption mechanism. They speculated that opening of the pericardium led to disruption of pericardial support and restraint, which in turn led to altered geometry of the right ventricle and an altered contraction pattern. This raised a thought-provoking question: why was RV strain noted to be reduced (traditionally thought to reflect intrinsic contractility) if the mechanism at hand was altered geometry rather than impaired contractility? They went on to propose that altered geometry of the right ventricle may have led to a reduced radius of curvature and consequently reduced wall stress and strain.
The evidence to support a pericardial disruption mechanism is scant but reasonably compelling. In a seminal experiment using real-time intraoperative echocardiographic monitoring, Unsworth et al. demonstrated that the (dramatic) decline in RV longitudinal function temporally coincided with the opening of the pericardium. It has been suggested that pericardial disruption may unmask subclinical RV dysfunction; the lackluster right ventricle cannot handle the increase in preload and the loss of ventricular interdependence that had previously helped it perform. The nearly ubiquitous finding of paradoxical interventricular septal motion after cardiac surgery is viewed as constructive in that it partially compensates for this loss of ventricular interdependence, augmenting the bellows function of the right ventricle by compressing it against the chest wall to compensate for the decline in longitudinal function.
In response to the much maligned issues of feasibility and reliability for RV quantification, Maffessanti et al. diligently report that their measurements of RV strain were feasible in 87% of patients, with intraclass coefficients in the range of 0.68 to 0.84 (good) for interobserver, intraobserver, and test-retest reliability. Measurements of 3D RVEF and volumes were feasible in 88% of patients, with intraclass coefficients in the range of 0.82 to 0.98 (excellent). Questions remain, however, about the real-world implementation of strain imaging of the thin-walled right ventricle at the clinical level.
Also in this issue, Ling et al. present the results of their cohort study involving 12 patients with suspected RV pathology (right-sided valve regurgitation or repair, shunt, or constriction) undergoing standard clinical echocardiography and cardiac magnetic resonance imaging (CMR) examinations within 24 hours at their institution. The following quantitative echocardiographic parameters were measured: TAPSE, S′, FAC, myocardial performance index, basal RV diameter, mid RV diameter, RV length, and proximal and distal RV outflow tract (RVOT) diameters. The gold standard was CMR RVEF and CMR RV volumes (using manual planimetry and the method of disks). Their study design involved two sessions: one in which 15 readers were shown the echocardiographic clips and were asked to rate the RV size and function visually and a later session in which the readers were shown the same clips plus images with quantitative guideline-based measurements and were again asked to rate the RV size and function. The hypothesis was that addition of quantitative measurements to the visual assessment would yield superior sensitivity, specificity, and interobserver reliability to identify abnormal RV size and function. The study results are summarized in Table 2 .
