Background
The grading of pulmonary regurgitation (PR) severity by two-dimensional (2D) and Doppler echocardiography is not standardized. Cardiovascular magnetic resonance imaging is the clinical gold standard for PR quantification. The purpose of this study was to determine the best 2D and Doppler echocardiographic predictors of severe PR.
Methods
Thirty-six patients with tetralogy of Fallot or pulmonary valve stenosis with prior pulmonary valvuloplasty or transannular or subannular patch repair underwent 2D and Doppler echocardiography and cardiovascular magnetic resonance. Two-dimensional and Doppler echocardiographic measurements used to predict severe PR included diastolic flow reversal in the main or branch pulmonary arteries, PR jet width ≥ 50% of the pulmonary annulus, PR pressure half-time < 100 ms, and PR index < 0.77.
Results
With the exception of PR index, all indices were significant independent predictors of severe PR. The best univariate predictor of severe PR was branch pulmonary artery diastolic flow reversal.
Conclusion
Two-dimensional and Doppler echocardiography reliably identified severe PR in this cohort.
Pulmonary regurgitation (PR) is common following the repair of some forms of congenital heart disease, including Tetralogy of Fallot (TOF) and isolated pulmonary stenosis (PS). Chronic PR can lead to right ventricular (RV) and left ventricular (LV) dysfunction, a decline in exercise tolerance, an increased incidence of arrhythmias, and an increased risk for sudden cardiac death. Conversely, the relief of PR via timely pulmonary valve replacement is associated with clinical stabilization or improvement. In fact, moderate to severe PR is one of the clinical parameters that may be used to dictate the timing of pulmonary valve replacement in these patients. Because echocardiographic imaging of the pulmonary valve can be difficult, two-dimensional (2D) and Doppler echocardiographic quantification of PR severity has not been standardized. Therefore, the grading of PR severity by 2D and Doppler echocardiography is most often performed qualitatively.
The limitations of 2D and Doppler echocardiography in the assessment of PR have been largely overcome by advances in cardiac magnetic resonance imaging (CMR). Indeed, CMR is now considered the clinical gold-standard method for the noninvasive quantitation of PR. However, CMR is relatively expensive and not universally available. In contrast, easily obtained and readily available 2D and Doppler echocardiographic measurements may be useful in determining PR severity and following its progression over time.
Multiple techniques have been studied for the 2D and Doppler echocardiographic estimation of PR severity. Some of these methods can detect hemodynamically significant PR, while others have been able to predict “severe” PR specifically. The 2D and Doppler measurements that appear to distinguish severe PR (as quantified by CMR) are (1) the ratio of PR jet width to the pulmonary valve annulus and (2) diastolic reversal of flow in the branch pulmonary arteries. However, the effect of combining multiple 2D and Doppler echocardiographic indices for the prediction of severe PR has not been fully elucidated.
We sought to determine the best 2D and Doppler echocardiographic predictors of severe PR, defined as a PR fraction (PRF) > 40% by CMR. Moreover, we sought to evaluate the utility of combining of several 2D and Doppler echocardiographic parameters together, as well as to construct a clinical decision tree to help guide their practical use.
Methods
Patients
The institutional review board at the David Geffen School of Medicine at the University of California, Los Angeles, approved the performance of this retrospective review. Thirty-six nonconsecutive adult and pediatric patients with repaired TOF or PS who had undergone both 2D and Doppler echocardiography and CMR were included in this analysis. The mean age was 33 years (range, 10-67 years), with 21 male patients. All patients had undergone previous pulmonary valvuloplasty or transannular or subannular patch repair. No patient had undergone pulmonary valve replacement or right ventricle–to–pulmonary artery conduit placement. The median time between 2D and Doppler echocardiographic and CMR studies was 0.4 months (range, 0-17 months). Twenty-seven patients (75%) underwent 2D and Doppler echocardiographic and CMR studies within 3 months of each other. No patients underwent interventions or had any significant changes in their clinical status between examinations. All 2D and Doppler echocardiographic and CMR data were abstracted from the original clinical reports. No additional interpretations or corrections were performed by the study investigators.
CMR
All examinations were performed on a 1.5-T magnetic resonance scanner (Magnetom Avanto; Siemens Medical Solutions, Malvern, PA) equipped with 32 independent receiver channels and a rapid 3-axis gradient system with peak gradient amplitudes of 40 mT/m and maximum slew rate of 200 mT/m/ms. A 12-element (6 × 2) wraparound torso coil provided optimal signal reception.
All patients underwent phase-contrast flow quantification, performed 1 cm distal (and perpendicular) to the pulmonary valve annulus. A through-plane, phase-sensitive, velocity-encoded pulse sequence was used for this purpose ( Figure 1 A). Images were vectorcardiographically gated. The lowest velocity encoding value not associated with aliasing was chosen. Using the optimal velocity encoding value, imaging was performed with a breath-hold sequence lasting 12 to 15 seconds per slice level. Typical pixel dimensions were 2 × 1.4 mm, and the slice thickness was 6 mm. Velocity maps were generated at 20 equidistant phases of the cardiac cycle, from which time-resolved volume flow curves were generated ( Figure 1 B).
Flow data were processed on a 3-dimensional workstation (Leonardo; Siemens Medical Solutions), using commercial flow analysis software (Argus; Siemens Medical Solutions), whereby spatially resolved flow maps were calculated from the phase difference images. By default, the flow-sensitized phase spectrum was assigned equal bias (±180°) to antegrade and retrograde flow. Severe PR by CMR was defined as ≥ 40% PRF. The PRF was calculated by dividing the regurgitant volume across the pulmonary annulus by the total forward volume.
Two-Dimensional and Doppler Echocardiography
All 2D and Doppler echocardiographic studies were performed using Acuson Sequoia C256 equipment (Siemens Medical Solutions). Two-dimensional and Doppler echocardiographic measurements used to predict severe PR were diastolic flow reversal of color flow (or spectral pulsed-wave Doppler) in the main or in the branch pulmonary arteries, a PR color flow Doppler jet width ≥ 50% of the pulmonary annular diameter, PR pressure half-time (PHT) <100 ms, and a PR index (PRi) < 0.77.
All 2D and Doppler echocardiographic studies reviewed had continuous-wave spectral Doppler recordings across the main pulmonary artery. Sixty percent had spectral pulsed-wave Doppler recordings in the main and/or branch pulmonary arteries. Diastolic flow reversal in the main pulmonary artery, and either the right or the left branch pulmonary arteries, was categorized as present or not present by color flow mapping and/or pulsed-wave spectral Doppler, when available ( Figure 2 ).
Using color flow Doppler in the main pulmonary artery, the width of the PR jet at the level of the pulmonary annulus was measured. The result was expressed as a percentage of the annulus, measured on standard 2D images. A jet width > 50% of the annular diameter was selected as a marker of severe PR and expressed as a categorical variable.
The PR PHT was determined using the continuous or pulsed-wave Doppler tracing of the PR jet. It is defined as the derivative of the rate of deceleration of the regurgitant velocity to 50% of the maximum regurgitant velocity ( Figure 3 ). As the volume of the regurgitating blood increases, the pressure both distal and proximal to the pulmonary valve annulus equilibrates more quickly. This manifests on the PR spectral Doppler tracing as a steeper slope of the line drawn along the right side of the regurgitation flow signal ( Figure 3 ). Therefore, a shorter PHT would suggest more severe regurgitation. A PR PHT < 100 ms has been shown to indicate the presence of hemodynamically significant (but not necessarily severe) PR.
Last, PRi was measured from continuous pulsed-wave spectral Doppler signals in the main pulmonary artery. PRi is defined as the ratio of the time occupied by the regurgitation signal divided by the total diastolic time interval ( Figure 4 ). Similar to the mechanism behind PHT, the PRi is expected to decrease with increasing severity of regurgitation. This is because the faster the regurgitating volume equilibrates in diastole, the shorter the PRi. On the basis of previously published data, a PRi < 0.77 identifies patients with PRFs > 24.5% (ie, hemodynamically “significant” but not necessarily severe). All the above measurements were made in a standard parasternal short-axis view.
Statistical Analysis
Demographic data were analyzed using Fisher’s exact test, and PR volumes were analyzed using the Mann-Whitney rank-sum test (SPSS 12.0 Student Version; SPSS, Inc, Chicago, IL). The sensitivities, specificities, positive predictive values (PPVs), and negative predictive values (NPVs) of the 5 2D and Doppler echocardiographic indices for the detection of severe PR were determined manually.
Univariate analysis was performed using Stata version 10 (StataCorp LP, College Station, TX) to delineate which 2D and Doppler echocardiographic indices individually predicted severe PR. P values < .05 were considered statistically significant. Multivariate analysis was also conducted but was not of value given the high correlation of the variables to the outcome. Pearson’s correlation coefficient calculation for PRF and PR volumes was also conducted using Stata.
Classification and regression tree (CART) analysis was then performed using Splus version 6.2 (Insightful Corporation, Seattle, WA) to construct a decision tree that might serve as a guideline for the detection of severe PR on 2D and Doppler echocardiography in the clinical setting. CART analysis is an empirical and nonparametric technique that is able to analyze a set of variables and their various interactions with one another to identify the variables that best discriminate between patient groups with respect to the outcome. The analysis uses repetitive binary partitioning of groups of patients on the basis of the “best” possible variable to split the group initially into 2 groups. The process is repeated with the “next best” variable to further subdivide the cohort group into 4 groups, and so on until all possibilities are exhausted. Each binary split is assigned a probability. As such, CART analysis has the potential to uncover complex interactions between various predictors that may be overlooked by traditional multivariate techniques. A particularly attractive feature of CART analysis is that the decision trees it generates can be relatively simple to interpret and to apply in clinical practice.
Results
Demographics
The demographic characteristics of our patient sample are enumerated in Table 1 . Approximately 200 patients with TOF or PS were seen at our institution during the study period. Thirty-six patients with both CMR and 2D and Doppler echocardiographic data were included in this study. Thirty-three patients (97%) carried a diagnosis of TOF, and 3 patients (8%) had isolated PS. Severe PR was detected using CMR in 17 of the 36 patients (47%). The degree of tricuspid regurgitation (by 2D and Doppler echocardiography) was more than mild in only 2 patients.
| Variable | Value |
|---|---|
| Male patients | 21 (58%) |
| Age (y) | 33 ± 15 |
| Diagnosis | |
| TOF | 33 (92%) |
| PS (isolated) | 3 (8%) |
| Previous palliative procedure | |
| Blalock-Taussig shunt | 15 (42%) |
| Other systemic to pulmonary shunt | 3 (8%) |
| Surgical valvotomy | 4 (11%) |
| Transcatheter valvotomy | 1 (3%) |
| Not known | 1 (3%) |
CMR-Derived RV Indices
RV ejection fraction, stroke volume, end-diastolic and end-systolic volumes, PR volumes, and cardiac output from CMR data are presented in Table 2 . The mean RV end-diastolic volume of the cohort was 191 ± 68 mL, or 107 ± 40 mL/m 2 indexed. The average RV ejection fraction was 46 ± 10%. The mean PR regurgitation volume was 40.5 ± 29 mL.
| RV index | Value |
|---|---|
| Ejection fraction (%) | 46 ± 10 |
| EDV (mL) | 191 ± 68 |
| EDVi (mL/m 2 ) | 107 ± 40 |
| ESV (mL) | 105 ± 44 |
| ESVi (mL/m 2 ) | 60 ± 23 |
| Stroke volume (mL) | 86 ± 35 |
| Regurgitation volume (mL) | 40.5 ± 29 |
| Cardiac output (L/min) | 5.8 ± 2.1 |
| Cardiac index (L/min/m 2 ) | 3.4 ± 1.2 |
Sensitivity, Specificity, and Predictive Values of 2D and Doppler Echocardiographic Measurements
Diastolic flow reversal in the main pulmonary artery had 100% sensitivity, but low specificity (39%), for detecting severe PR ( Table 3 ). Diastolic flow reversal in the branch pulmonary arteries remained quite sensitive (87%) but was much more specific (87%) for severe PR compared with main pulmonary artery flow reversal. The presence of diastolic flow reversal in the branch pulmonary arteries also had high NPV (87%) and PPV (87%). Diastolic flow reversal in the main pulmonary artery demonstrated excellent NPV (100%) but low PPV (59%). PR jet width ≥ 50% of the pulmonary annular diameter was sensitive (93%) and had a good NPV (93%) for detecting severe PR. A PR PHT < 100 ms was a sensitive (90%) but not specific (64%) indicator of severe PR. Finally, a PRi < 0.77 was neither sensitive (73%) nor specific (47%).
| Parameter | Sensitivity | Specificity | PPV | NPV |
|---|---|---|---|---|
| MPA diastolic flow reversal | 100% | 39% | 59% | 100% |
| BPA diastolic flow reversal | 87% | 87% | 87% | 87% |
| PR jet/pulmonary annular diameter ratio ≥ 50% | 94% | 74% | 76% | 93% |
| PHT < 100 ms | 90% | 64% | 78% | 82% |
| PRi < 0.77 | 73% | 47% | 58% | 64% |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
