Incremental value of left ventricular shape parameters measured by gated SPECT MPI in predicting the super-response to CRT

Abstract

Background.

The purpose of this study was to evaluate the predictive value of left ventricular (LV) shape parameters measured by gated SPECT myocardial perfusion imaging (MPI) in super-responders enrolled in the VISION-CRT trial.

Methods.

One hundred and ninety-nine patients who met standard criteria for CRT from multiple centers were enrolled in this study. End-systolic eccentricity (ESE) and end-diastolic eccentricity (EDE) were measures of LV shape. Super-responders were the patients who had a relative increase in left ventricular ejection fraction (LVEF) ≥ 15%.

Results.

Complete data were obtained in 165 patients, and 43.6% of them were classified as super-responders. ESE was an independent predictor of CRT super-responders in univariate (OR 12.59, 95% CI 1.56–101.35, P = .017) and multivariate analysis (OR 35.71, 95% CI 1.66–766.03, P = .006). ESE had an incremental value over significant clinical and SPECT imaging variables, including angiotensin-converting enzyme inhibitors or angiotensin II receptor blocker, coronary artery disease, myocardial infarction, LVEF, end-diastolic volume index, and scar burden (AUC 0.82 vs. 0.80, sensitivity 0.68 vs. 0.65, specificity 0.82 vs. 0.78).

Conclusions.

LV shape parameters derived from gated SPECT MPI have the promise to improve the prediction of the super-response to CRT. Moreover, ESE provides incremental value over existing clinical and nuclear imaging variables. (J Nucl Cardiol 2022;29:1537–46.)

INTRODUCTION

Ventricular remodeling is characterized by a group of molecular, cellular, and interstitial changes, which occurs after cardiac injury and is clinically manifested by changes in size, shape, and function.^1–3 In addition, a remodeled ventricle is associated with the development and progression of ventricular dysfunction, arrhythmias, and poor prognosis.⁴ Studies have found that LV shape is related to the cardiac function; the normal elliptical LV shape would change to a spherical shape due to the development of eccentric hypertrophy or post myocardial infarction (MI).^5,6 Some shape parameters, such as LV shape index, eccentricity, and elongation, have been reported to have predictive value for congestive heart failure, diabetes, and patients with significant cardiac structural and functional abnormalities.^5,7–9

ECG-gated single-photon emission computer tomography (SPECT) myocardial perfusion imaging (MPI) can assess LV structure by the tracer counts of myocardial perfusion, instead of geometric changes in the myocardium, which is more objective and reproducible. Gimelli et al found that the LV eccentricity measured by gated SPECT MPI predicted the presence of multivessel coronary artery disease (CAD).¹⁰ The left ventricular shape parameters also have a close relationship with LV volumes and function in HF patients,^11,12 even in healthy subjects.⁸ Moreover, LV shape index has been applied to detect the presence of adverse LV remodeling in patients with structural and functional cardiac alterations due to diabetes mellitus.⁷ However, very few works have been proposed in the literature for CRT considering LV shape measured by gated SPECT MPI.¹³

Super-responders are the patients who have a significant improvement in functional capacity, quality of life, HF symptoms, left ventricular function, and reverse remodeling after CRT.¹⁴ Several studies found that LV shape characteristics, such as smaller LV, less LV end-diastolic diameter, and greater LV strain, measured by echocardiography, were associated with super-response to CRT.^14,15 Gated SPECT MPI, which is widely accepted in the clinic, has been validated to effectively access LV shape,^9,10,12 due to its high repeatability and reproducibility in the evaluation of myocardial perfusion, global and regional ventricular function, and mechanical dyssynchrony for patients with CRT.^16,17 This study is a post hoc retrospective analysis of the shape variables in patients obtained from the IAEA VISION-CRT trial. It is aimed to evaluate the clinical role of LV eccentricity measured by gated SPECT MPI in the prediction of super-response to CRT.

METHODS

Patient Population

The VISION-CRT trial was a prospective multicenter trial that enrolled subjects from 10 cardiological centers in 8 countries (Brazil, Chile, Colombia, Cuba, India, Mexico, Pakistan, and Spain). The complete study design and primary results of VISION-CRT were previously published.^13,18,19 The inclusion criteria were as follows: symptomatic HF patients over 18 years old with NYHA functional class II, III or ambulatory IV HF for at least 3 months before enrollment, despite optimal medical treatment according to the current guidelines; LVEF ≤ 35% from ischemic or non-ischemic causes, measured according to the usual procedure at the participating center for inclusion, whereas LVEFs used for analysis came from nuclear core lab; sinus rhythm with LBBB configuration defined as a wide QRSd (≥ 120 ms). Exclusion criteria were as follows: right bundle branch block, pregnancy or breast-feeding, acute coronary syndromes, coronary artery bypass grafting, or percutaneous coronary intervention in the last 3 months before enrolment and within 6 months of CRT implantation. The CRT devices were implanted by standard procedures. The LV lead was implanted in posterolateral coronary vein, depending on vein availability.

Definitions of all clinical variables were outlined before the start of VISION-CRT. Subjects underwent a detailed clinical and gated SPECT MPI evaluation just before recruitment to the study. All patients provided written informed consent, and all procedures were done according to the Declaration of Helsinki.

Clinical parameters and gated SPECT MPI were assessed before CRT (baseline) and at 6 ± 1 month after (follow-up). The patients were classified as “super-responders” to CRT if they had a relative increase of LVEF ≥ 15% as measured by gated SPECT MPI at follow-up. Others were classified as non-super-responders.

SPECT MPI Assessment

Gated SPECT MPI was performed around 30 min post injection using 20 to 30 mCi of 99mTc-sestamibi or tetrofosmin of 740 to 1110 MBq. All the images were acquired in gamma cameras using 180° orbits with a complimentary 8 or 16 frames ECG gating. Image reconstruction and reorientation were completed by Emory Reconstruction Toolbox (ERToolbox; Atlanta, GA) using the OSEM method with three iterations and ten subsets and filtered by a Butterworth filter with a power of 10 and a cut-off frequency of 0.3 cycles/mm. The resulting short-axis images were sent to Emory Cardiac Toolbox (ECTb4, Atlanta, GA) for automatized accessing of LV function, including LVEF, left ventricular end-systolic volume (LVESV), left ventricular end-diastolic volume (LVEDV), LV shape; end-systolic eccentricity (ESE) and end-diastolic eccentricity (EDE), and LV mechanical dyssynchrony; and phase standard deviation (PSD) and phase bandwidth (PBW). ESV Index (ESVi) and EDV Index (EDVi) (in milliliters per square meter) are obtained by dividing LVESV and LVEDV by the body surface area (BSA), respectively. Moreover, the latest contracting viable sites could be identified by Emory Cardiac Toolbox as a recommendation of the optimal LV lead position.²⁰ Concordance is defined as the agreement between CRT LV lead position recorded and the recommended site.

Shape Parameters Measured by Gated SPECT MPI

Measurement techniques of myocardial shape variables have been well established. They calculate the 3-dimensional LV shape parameters from gated SPECT MPI.¹² Eccentricity is a measure of the elongation of LV, which is derived from an ellipsoid fitted from the LV endocardial surface according to Eq. (1), where $x$ and $\mathrm{y}$ are the minor axes, and $\mathrm{z}$ is the major axis of the ellipsoid.¹⁰

$$\text{Eccentricity}=\sqrt{1-\frac{x\times y}{z^2}}.$$ (1)

Based on the assumption that the minor axes have the same length ($x=y$), the ellipsoid can be considered as a spheroid, as shown in Figure 1. The LV eccentricity is closer to 0 if the shape of LV is closer to a sphere and closer to 1 if the shape of LV is closer to a line.

Figure 1.

The best-fitted ellipsoid based on the LV endocardial surface from a gated SPECT MPI study measured by Emory Cardiac Toolbox (ECTb4, Atlanta, GA).

Statistical Analysis

Differences between the super-responders and non-super-responders were compared by the Student t test for continuous variables, expressed as mean ± standard deviation, and Pearson χ² test for categorical variables, expressed in number and percentage. The univariate binary logistic regression analysis was applied to estimate potential predictors for super-responders. The multivariable binary logistic regression was performed to analyze the independent predictors of super-responders, and the variables with P < .05 in the univariate analysis were included. Moreover, the incremental values of shape parameters were evaluated by comparing the receiver operator characteristic (ROC) curve of binary logistic regression from the clinical variables alone, from the combination of clinical variables with ESE or EDE, and from the combination of clinical variables with all shape parameters. P < .05 was considered to be statistically significant. Statistical analysis was performed by the Python Statsmodels package.²¹

RESULTS

Baseline Characteristics

A total of 199 patients underwent CRT, but complete data of clinical assessment, baseline SPECT MPI, and clinical 6-month follow-up data were obtained in 177 patients. Among these patients, 11 of them died before the follow-up and 1 patient had an extremely small ESV (8 mL), which was an outlier caused by the low resolution of gated SPECT MPI when measuring a small heart. Finally, 165 patients were included in the statistic analysis in this research (Figure 2).

Figure 2.

Study flow chart. CRT cardiac resynchronization therapy, SPECT gated single-photon emission-computed tomography, MPI myocardium perfusion imaging.

The baseline characteristics of the study population are shown in Table 1. For all the patients, the age was 60.3 ± 10.9 years, and 98 (59.4%) patients were male. Fifty-one (30.9%) patients had a previous history of CAD. Hypertension (58.8%), smoking (16.4%), and diabetes (24.8%) were also shown in the baseline data.

Table 1.

Baseline characteristics and left ventricular parameters of the enrolled patients

Variables	All (n = 165)	Super-responders (n = 72, 43.6%)	Non-super-responders (n = 93, 56.4%)	P value
ACEI/ARB	136 (82.4%)	65 (90.3%)	71 (76.3%)	.034
Age	60.3 ± 10.9	61.2 ± 9.9	59.6 ± 11.6	.354
BSA	1.8 ± 0.2	1.8 ± 0.2	1.8 ± 0.2	.984
CAD	51 (30.9%)	14 (19.4%)	37 (39.8%)	.008
Diabetes mellitus	41 (24.8%)	16 (22.2%)	25 (26.9%)	.613
Gender	98 (59.4%)	40 (55.6%)	58 (62.4%)	.469
Hypertension	97 (58.8%)	39 (54.2%)	58 (62.4%)	.367
MI	35 (21.2%)	7 (9.7%)	28 (30.1%)	.003
NYHA				.014
II	46 (27.9%)	27 (37.5%)	19 (20.4%)
III	101 (61.2%)	35 (48.6%)	66 (71.0%)
IV	18 (10.9%)	10 (13.9%)	8 (8.6%)
Race				.362
African	17 (10.3%)	8 (11.1%)	9 (9.7%)
Asian	6 (3.6%)	1 (1.4%)	5 (5.4%)
Caucasian	23 (13.9%)	10 (13.9%)	13 (14.0%)
Hispanic	87 (52.7%)	35 (48.6%)	52 (55.9%)
Indian	32 (19.4%)	18 (25.0%)	14 (15.1%)
QRS duration	160.9 ± 25.1	162.9 ± 22.5	159.4 ± 26.9	.378
Smoking	27 (16.4%)	14 (19.4%)	13 (14.0%)	.466
SPECT variables
Concordance	40 (24.2%)	15 (20.8%)	25 (26.9%)	.474
EDVi	143.1 ± 56.7	132.4 ± 54.5	151.4 ± 57.0	.034
ESVi	106.9 ± 52.4	101.6 ± 52.8	111.0 ± 51.7	.252
EDE	0.5 ± 0.2	0.6 ± 0.2	0.5 ± 0.2	.085
ESE	0.6 ± 0.2	0.6 ± 0.1	0.5 ± 0.2	.014
LVEF	27.7 ± 10.3	25.8 ± 10.3	29.2 ± 10.1	.035
LVEDV	257.6 ± 104.7	239.6 ± 103.7	271.6 ± 103.4	.052
LVESV	192.7 ± 96.2	184.3 ± 98.8	199.2 ± 93.6	.330
PBW	152.4 ± 73.5	145.7 ± 74.4	157.6 ± 72.4	.306
PSD	48.8 ± 19.7	47.6 ± 20.6	49.6 ± 19.9	.520
Scar	24.5 ± 14.4	20.0 ± 11.6	27.9 ± 15.4	< .001

Data are expressed as mean ± SD or number (percentage)

ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin II receptor blocker, BSA body surface area, CAD coronary artery disease, MI myocardial infarction, NYHA New York Heart Association, LVEDV end-diastolic volume, LVESV end-systolic column, EDVi end-diastolic volume index, ESVi end-systolic volume index, EDE end-diastolic eccentricity, ESE end-systolic eccentricity, LVEF left ventricular ejection fraction, PBW Phase bandwidth, PSD phase standard deviation

After 6-month follow-up, 72 of the 165 patients (43.6%) were considered as a super-responder to CRT, and 93 of the 165 patients (56.4%) were considered as non-super-responders. Significant differences of ESE and EDE (0.6 ± 0.1 vs. 0.5 ± 0.2, P = .014; 0.5 ± 0.2 vs. 0.6 ± 0.2, P = .045) were noted between the two groups, as well as other clinical variables, including ACE inhibitors or ARB (65 patients [90.3%] vs. 71 [76.3%], P = .034), CAD (14 [19.4%] vs. 37 [39.8%], P = .008), myocardial infarction (7 [9.7%] vs. 28 [30.1%], P = .003), EDVi (132.4 ± 54.5 vs. 151.4 ± 57.0, P = .034), NYHA (P = .014), LVEF (25.8 ± 10.3 vs. 29.2 ± 10.1, P = .035), and myocardial scar (20.0% ± 11.6% vs. 27.9% ± 15.4%, P < .001). However, the concordance was not statistically significant to distinguish super-response and non-superresponse (P = .474). If the significance level was set at 0.1, EDE (0.6 ± 0.2 vs. 0.5 ± 0.2, P = .085) could also be statistically significant. Representative examples of super-responders and non-super-responders are depicted in Figure 3.

Figure 3.

Illustrations of the end-systolic frames of a super-responder and a non-super-responder. (a, b) are the baseline and follow-up vertical long axis (VLA) and horizontal long axis (HLA) images of a 57-year-old male as an example of super-responder. (c, d) are the baseline and follow-up VLA and HLA images of a 54-year-old male as an example of non-super-responder.

Prediction of Super-Response to CRT

In the univariate analysis, ACEI or ARB (OR 2.88, 95% CI 1.15–7.18, P = .024), CAD (OR 0.37, 95% CI 0.18–0.75, P = .006), EDVi (OR 0.99, 95% CI 0.99–1.0, P = .036), myocardial scar (OR 0.96, 95% CI 0.94–0.98, P < .001), myocardial ischemia (OR 0.25, 95% CI 0.10–0.61, P = .002), LVEF (OR 0.97, 95% CI 0.94–1.00, P = .037), and ESE (OR 12.59, 95% CI 1.56–101.35, P = .017) were associated with super-response. However, the baseline mechanical dyssynchrony (PSD, P = .518; PBW, P = .304) and concordance (P = .369) were not significantly associated with super-response. In the multivariate analysis, ESE was also an independent predictor (OR 35.71; 95% CI, 1.66–766.03; P = .006). The results of the univariate and multivariate analysis are shown in Table 2.

Table 2.

Univariate and multivariate logistic regression analyses of super-responders defined by a relative increase in LVEF ≥ 15%

Variables	Univariate analysis			Multivariate analysis
	OR	95% CI	P value	OR	95% CI	P value
ACEI/ARB	2.88	1.15–7.18	.024	5.27	1.73–16.00	.003
CAD	0.37	0.18–0.75	.006	0.83	0.28–2.45	.730
Concordance	0.72	0.35–1.49	.369
EDVi	0.994	0.99–1.00	.036	0.98	0.98–0.99	.001
ESVi	1.00	0.99–1.00	.252
EDE	4.89	0.79–30.44	.089
ESE	12.59	1.56–101.34	.017	35.71	1.66–766.03	.006
Gender	0.75	0.40–1.41	.377
LVEF	0.97	0.94–1.00	.037	0.87	0.83–0.92	< .001
LVEDV	1.00	0.99–1.00	.055
LVESV	1.00	0.99–1.00	.329
MI	0.25	0.10–0.61	.002	0.25	0.07–0.91	.036
NYHA	0.72	0.43–1.20	.212
PSD	1.00	0.98–1.01	.518
PBW	1.00	0.99–1.00	.304
QRS duration	1.01	0.99–1.02	.376
Scar	0.96	0.94–0.98	.001	0.97	0.94–1.00	.005

ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin II receptor blocker, CAD coronary artery disease, MI myocardial infarction, EDE end-diastolic eccentricity, ESE end-systolic eccentricity, LVEF left ventricular ejection fraction, LVEDV left ventricular end-diastolic volume, LVESV left ventricular end-systolic volume, EDVi end-diastolic volume index, ESVi end-systolic volume index, PBW Phase bandwidth, PSD phase standard deviation

In the ROC analysis of LV shape parameters, the area under the curve (AUC) of clinical variables alone (sensitivity 0.65, specificity 0.78, AUC 0.8), the combination of clinical variables with EDE (sensitivity 0.65, specificity 0.8, AUC 0.8), the combination of clinical variables with ESE (sensitivity 0.68, specificity 0.82, AUC 0.82), and the combination of clinical variables with both ESE and EDE (sensitivity 0.67, specificity 0.81, AUC 0.83) increased sequentially, as shown in Figure 4.

Figure 4.

Receiver operating characteristic curves of clinical characteristic with or without ESE and EDE.

Furthermore, sequential models indicated that the addition of ESE (likelihood 6.41, P = .011) was more strongly associated with CRT super-responders. However, the addition of EDE to a model of clinical characteristics did not provide a significant improvement in association with CRT super-responders. Moreover, due to the collinearity of ESE and EDE (Pearson’s correlation coefficient 0.81, P < .001), EDE reduces the performance of the model of ESE plus clinical characteristics (LH 1.55, P = .213), as shown in Figure 5.

Figure 5.

Incremental adjusted additive value of eccentricity in the prediction of super-responders of CRT.

DISCUSSION

This study demonstrates that the LV shape parameter ESE is a promising variable derived automatically from gated SPECT MPI to predict super-responders of CRT. Moreover, it provides incremental value over clinical and nuclear imaging variables. In patients treated with CRT, presenting a super-response is associated with excellent long-term outcomes.²² The identification of these patients can improve our understanding of pathophysiologic mechanisms linked to reverse remodeling and provide better tools to select the best candidates for CRT.

LV remodeling is associated with progressive worsening of LV function and increased cardiovascular morbidity and mortality in various cardiovascular diseases.^1,3 The most common causes are represented by conditions with an elevated LV hemodynamic load or after myocardial infarction, resulting in an increase in LV chamber volume, muscle mass, and fibrous tissue contents.^6,8 Several reports have suggested that descriptors of LV shape enhance the ability to discriminate normal from pathological LV because the occurrence of abnormal LV eccentricity takes place even before the change in LV systolic function becomes apparent.^23,24

The degree of LV enlargement and dysfunction is known to be directly related to the risk of death.²⁵ However, in gated SPECT and echocardiography, the measurement method of sphericity is different. In two-dimensional echocardiography, LV geometry is calculated from the manual measurements, which is subject to variability and depends on the experience of the operator. The three-dimensional measurement of ventricular shape by gated SPECT MPI derives automatically and it has been demonstrated to be highly reproducible.¹² Two similar shape variables, end-systolic shape index (ESSI) and end-diastolic shape index (EDSI), were also tested in our research. Left ventricle shape index, defined as the ratio of the maximum 3D short- and long-axis dimensions of the left ventricle at the end-systolic or end-diastolic frame, are measured by gated SPECT MPI and have a great correlation with ESE (Pearson correlation coefficient, 0.98) and EDE (Pearson correlation coefficient, 0.99), respectively. ESSI was a significant independent predictor of CRT super-responders (OR, 0.036, 95% CI 0.00–0.77, P value .033). EDSI could also predict CRT super-responders (OR, 0.063, 95% CI 0.00–1.70, P value .10). In addition, ESSI had an incremental value to predict CRT responders, but EDSI did not. Therefore, no matter how the LV shape is defined, it has great potential to predict CRT super-responders.

The improvements in LVEF and the reductions in LVEDV are generally modest for patients with heart failure who undergo CRT. But a part of patients had a dramatic response to CRT (super-responders). Bulava et al were the first to describe this phenomenon that a 72-year-old woman with ischemic cardiomyopathy, whose LVEF increased from 15% to 60% at 1-year follow-up after CRT.²⁶ Rickard et al identified the super-responders as an absolute LVEF improved by ≥ 20%.²⁷ Killu et al defined the patients with an absolute change in LVEF of > 15% as super-responders.²⁸ Ypenburg et al used the decrease in LVESV ≥ 30% as the definition of super-responders.²⁹ Both of them seem reasonable because LVESV, as a volumetric assessment, is an objective measure that provides the information of LV reverse remodeling and predicts long-term clinical outcomes¹⁴; compared to the improvement in LVESV, LVEF is the most widely used variable in echocardiography and can present the LV function while provide prognostic benefit. António et al proposed a definition of super-responders: if patients have a reduction of one or more NYHA functional classes, a two-fold increasement or an absolute change in LVEF of > 45%, and decrease in the LVESV > 15%.¹⁴ However, this definition of super-responder has limitations, because (1) reduction of one NYHA functional class is a subjective evaluation of the improvement of CRT, which might have a placebo effect in our non-randomized clinical trial; (2) two-fold increasement or absolute change in LVEF of > 45% might be too strict to the super-responders.

In our study, the super-response was defined as a relative increase in the LVEF ≥ 15%. In general responders (an increase in LVEF ≥ 5%), the ESE (0.6 ± 0.1 vs. 0.5 ± 0.2, P < .001) and EDE (0.6 ± 0.2 vs. 0.5 ± 0.2, P = .025) also had differences between responders and non-responders. Both of them were independent predictors for general CRT responders in univariate analysis, but not significant in multivariate analysis. Additionally, a combination of a relative increase in LVEF ≥ 15% or a decrease in LVESV ≥ 30% was analyzed. In our study, 74 (44.8%) patients had the combined outcome and were classified into super-responders. Significant differences were found between all end-systolic and end-diastolic eccentricity (ESE: 0.6 ± 0.1 vs. 0.5 ± 0.2, P = .005; EDE: 0.6 ± 0.2 vs. 0.5 ± 0.2, P = .037). Both of them were independent predictors of super-responders in univariate analysis (ESE: OR 18.07, 95% CI 2.14–152.32, P = .008; EDE: OR 6.94, 95% CI 1.08–44.83, P = .042). Due to the collinearity of ESE and EDE, we preferred multivariate analysis in separate models and got a similar result of ESE (OR 12.16, 95% CI 0.95–156.32, P = .055), as shown in Table 3. Although the P value was greater than 0.05, the OR was large, which means a strong association between ESE and super-responders. Moreover, the lead concordance had no predictive value between super-responders (P = .369) and general responders (P = .895).

Table 3.

Univariate and multivariate logistic regression analyses of super-responders defined as a relative increase in LVEF ≥ 15% or a decrease in LVESV ≥ 30%

Variable	Univariate analysis			Multivariate analysis (EDE)			Multivariate analysis (ESE)
	OR	95% CI	P value	OR	95% CI	P value	OR	95% CI	P value
ACEI/ARB	2.74	1.03–5.97	.044	2.53	0.98–6.57	.056	2.57	0.99–6.67	.052
CAD	0.34	0.17–0.70	.003	0.60	0.23–1.52	.280	0.62	0.24–1.61	.322
Concordance	0.67	0.32–1.39	.284
EDVi	0.99	0.99–1.00	.020	0.99	0.97–1.01	.311	0.99	0.97–1.01	.315
ESVi	1.00	0.99–1.00	.144
EDE	0.94	1.08–44.83	.042	2.79	0.34–22.72	.337
ESE	18.07	2.14–152.32	.008				12.16	0.95–156.32	.055
Gender	0.74	0.40–1.38	.347
LVEF	0.97	0.95–1.01	.130
LVEDV	1.00	0.99–1.00	.034	1.00	0.99–1.01	.522	1.00	0.99–1.02	.412
LVESV	1.00	0.99–1.00	.206
MI	0.24	0.10–1.58	.002	0.36	0.11–1.15	.084	0.31	0.09–1.00	.051
NYHA	0.72	0.44–1.24	.248
PSD	0.99	0.98–1.01	.292
PBW	1.00	0.99–1.00	.163
QRS duration	1.00	0.99–1.02	.574
Scar	0.96	0.93–0.98	< .001	0.97	0.94–1.01	.110	0.98	0.95–1.01	.155

ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin II receptor blocker, CAD coronary artery disease, MI myocardial infarction, NYHA New York Heart Association, EDVi end-diastolic volume index, ESVi end-systolic volume index, EDE end-diastolic eccentricity, ESE end-systolic eccentricity, LVEF left ventricular ejection fraction, LVEDV left ventricular end-diastolic volume, LVESV left ventricular end-systolic column, PBW Phase bandwidth, PSD phase standard deviation

We also found that scar percentage has a moderate correlation with ESE (Pearson correlation coefficient, − 0.36); other clinical characteristics and nuclear imaging variables had a weak correlation with ESE. The relationship between LV shape and LV size is probably complicated, and LV shape may also depend on other factors. All findings have demonstrated that ESE has incremental value over significant clinical and nuclear imaging variables, including angiotensin-converting enzyme inhibitors (ACEI) or angiotensin II receptor blocker (ARB), CAD, MI, LVEF, and scar burden in predicting CRT super-responders.

There were several limitations in this study. First, this was a post hoc analysis of the VISION-CRT trial that was not a randomized trial. Second, the information provided by the short follow-up period was limited; the prognostic value of LV shape parameters needs further investigation. Third, this study enrolled a relatively small number of patients from multicenters with the inherent limitation of such study design. Fourth, in the design of the VISION-CRT trial, some variables were not included in the data acquisition, such as heart rate that can influence LVEF.³⁰ Further investigation with a longer follow-up period is needed to assess LV shape parameters in CRT super-responders.

NEW KNOWLEDGE GAINED

The role of MPI for heart failure is already known in light of the functional and phase analysis parameters. In brief, this study demonstrates that LV remodeling can be assessed by shape parameters obtained with gated SPECT. In particular, in our sample, ESE was confirmed to be an independent predictive variable for CRT super-response.

CONCLUSIONS

LV shape parameters derived from gated SPECT MPI have the promise to improve the prediction of the super-response to CRT. Moreover, ESE provides incremental value over existing clinical and nuclear imaging variables.

Abbreviations

CRT

Cardiac resynchronization therapy

ECTb4

Emory Cardiac Toolbox Version 4.0

EDE

End-diastolic eccentricity

ESE

End-systolic eccentricity

LV

Left ventricular

LVEDV

Left ventricular end-diastolic volume

LVEF

Left ventricular ejection fraction

LVESV

Left ventricular end-systolic volume

PSD

Phase histogram standard deviation

MPI

Myocardial perfusion imaging