Right Ventricular Dysfunction Impairs Effort Tolerance Independent of Left Ventricular Function Among Patients Undergoing Exercise Stress Myocardial Perfusion Imaging

Abstract

Background

RV and LV function are closely linked due to a variety of factors, including common coronary blood supply. Altered LV perfusion holds the potential to affect the RV, but links between LV ischemia and RV performance, as well as independent impact of RV dysfunction on effort tolerance are unknown.

Methods and Results

The population comprised 2051 patients who underwent exercise stress MPI and echo (5.5 ± 7.9 days), among whom 6% had echo-evidenced RV dysfunction. Global summed stress scores were nearly 3-fold higher among patients with RV dysfunction, attributable to increments in inducible and fixed LV perfusion defects (all p≤0.001). Regional inferior and lateral wall ischemia was greater among patients with RV dysfunction (both p<0.01), without difference in corresponding anterior defects (p=0.13). In multivariable analysis, inducible inferior and lateral wall perfusion defects increased the likelihood for RV dysfunction (both p<0.05) independent of LV function, fixed perfusion defects, and PA pressure. Patients with RV dysfunction demonstrated lesser effort tolerance whether measured by exercise duration (6.7±2.8 vs. 7.9±2.9 min, p<0.001) or peak treadmill stage (2.6±0.9 vs. 3.1±1.0, p<0.001), paralleling results among patients with LV dysfunction (7.0±2.9 vs. 8.0±2.9, p<0.001 |2.7±1.0 vs. 3.1±1.0, p<0.001 respectively). Exercise time decreased stepwise in relation to both RV and LV dysfunction (p<0.001), and was associated with each parameter independent of age or medication regimen.

Conclusions

Among patients with known or suspected CAD, regional LV ischemia involving the inferior and lateral walls confers increased likelihood of RV dysfunction. RV dysfunction impairs exercise tolerance independent of LV dysfunction.

Right ventricular (RV) systolic dysfunction is an adverse prognostic marker that has been shown to confer increased risk for heart failure and death.^1–3 RV and left ventricular (LV) performance are closely linked due to a variety of factors, including common coronary arterial blood supply. Patients with LV injury have been shown to develop RV dysfunction in the extreme case of myocardial infarction, for which risk for RV dysfunction varies in relation to LV infarct pattern.^{2, 4} By extension, LV ischemia holds the potential to impact RV performance, both due to direct (i.e. altered RV perfusion) and indirect (i.e. increased RV afterload) mechanisms. However, magnitude and distribution of LV ischemia have yet to be studied as markers of RV dysfunction.

Radionuclide stress myocardial perfusion imaging (MPI) is widely used to assess LV ischemic burden, as supported by data showing extent and severity of perfusion deficits to correspond with anatomic burden of atherosclerotic plaque and predict adverse prognosis.^{5, 6} MPI is often performed adjunctively with exercise treadmill testing, enabling perfusion data to be acquired together with physiologic exercise parameters that are known to predict adverse prognosis.^7–10 Paralleling uncertainty regarding perfusion-based markers of RV dysfunction, the independent impact of RV performance on exercise tolerance is unknown. Given that nearly half of all patients with symptomatic heart failure have preserved LV systolic function,¹¹ identification of novel parameters that impact effort tolerance – as offered by readily assessable indices of RV function – is of widespread clinical significance.

This study assessed RV function - as well as clinical, exercise, and myocardial perfusion indices - among a broad cohort undergoing echocardiography and exercise MPI so as to test the primary hypothesis that RV dysfunction impairs exercise tolerance independent of LV contractile function. To do so, multimodality imaging was employed, whereby MPI was used as the reference for LV perfusion pattern and echo as the reference for RV (and LV) structure/function. Study goals were three-fold - (1) to assess prevalence and clinical markers of RV dysfunction among patients with known or suspected CAD undergoing exercise stress testing; (2) to determine whether pattern and severity of LV perfusion deficits modifies risk for RV dysfunction; and (3) to test the impact of RV dysfunction on exercise tolerance.

Methods

Population

The population comprised consecutive patients who underwent clinically indicated transthoracic echocardiography (echo) within 1 month of exercise single photon emission computed tomography (SPECT) MPI: A maximum interval of 1 month was selected so as to minimize the impact of changes in loading conditions on RV performance during the interval between tests. Imaging was performed February 2011 – January 2016 at Weill Cornell Medical College (WCMC). To ensure accurate echo categorization of RV performance, both tricuspid annular plane systolic excursion (TAPSE) and peak tricuspid annular systolic velocity (S’) were required for study inclusion: 7% (n=166) of otherwise eligible patients were excluded due to absence of TAPSE or S’. No patients were excluded based on clinical characteristics or MPI results. This retrospective protocol was approved by the WCMC Institutional Review Board (IRB).

Imaging Protocol

SPECT

MPI was performed in accordance with a previously described protocol using a dual headed scintillation camera system with a low-energy high-resolution collimator.^{12, 13} In brief, thallium-201 (~3 mCi) or technetium-99m sestamibi (~10 mCi) was injected intravenously; baseline (i.e. rest) perfusion images were acquired approximately 10 minutes after Tl-201, and 60 minutes after Tc-99m sestamibi injection. Symptom limited exercise treadmill testing was performed using a Bruce protocol: Tc-99m (~30 mCi) was intravenously administered at peak stress, for which a minimum target heart rate response to exercise (≥85% predicted maximum heart rate [220-age]) was used as the criterion for adequate workload required for radionuclide injection.¹⁴ Serial 12-lead electrocardiograms (ECGs) and blood pressure measurements were obtained (together with assessment of clinical symptoms) at baseline and at each stage of the exercise treadmill protocol. For patients unable to attain target heart rate during exercise, pharmacological (adenosine or regadenoson) stress was performed after treadmill testing – patients were instructed to refrain from caffeine intake >12 hours prior to MPI. Attenuation correction imaging and/or prone reposition imaging (as clinically tolerated) was used to differentiate between pathologic perfusion deficits and attenuation artifact in accordance with previously employed methods.^{12, 15} Post-stress images were acquired approximately 30 minutes following exercise, and 1–2 hours following pharmacologic stress.

Echocardiography

Non-contrast transthoracic echoes were performed by experienced sonographers using commercially available equipment. Images were acquired in standard parasternal, as well as apical 2-, 3-, and 4- chamber orientations. RV systolic function assessment was performed using M mode (for TAPSE) and tissue Doppler (for S’) imaging. In accordance with consensus standards, TAPSE and S’ data were acquired in apical 4-chamber orientation.¹⁶

Image Analysis

SPECT

MPI was interpreted by American Heart Association / American College of Cardiology (AHA/ACC) level III trained readers, for whom high reproducibility has been previously reported.¹⁷ Visual interpretation was confirmed by review of polar plots with comparison of segmental radiotracer intensity to computer-generated, gender-matched datasets. Perfusion defect severity on a per segment basis was graded using a standard 17 segment, 5-point per-segment scoring system (0 = normal, 1 = equivocal or mildly reduced, 2 = moderately reduced, 3 = severely reduced, 4 = absence of detectable radioisotope uptake).¹⁸ Fixed segments were defined as those with matched perfusion abnormalities (matched scores) on stress and rest; ischemic segments were defined as those with inducible abnormalities (stress>rest). Summed stress and rest scores were calculated by adding per-segment defect severity for all segments. Inducible perfusion abnormalities (summed difference score) were assessed as the difference between rest and stress images. MPI exams were interpreted blinded to echo-based quantification of RV function.

Regional LV perfusion was assessed using uniform partitions (see Figure 1), whereby LV was partitioned into three equal size (anterior, inferior and lateral) territories, in accordance with prior methods as employed by our group^{15, 19}: Summed stress, rest, and difference scores in each territory were calculated as the sum of individual perfusion scores within encompassed segments. Perfusion territories were not mutually exclusive; patients with defects in multiple territories were scored for each territory based on extent and severity of constitutive segmental deficits.

Figure 1. LV Perfusion Territories.

Bullseye plots illustrating regional LV perfusion territories (highlighted). Each category comprised five segments, such that the total myocardium subtended by each was equivalent.

Echocardiography

Echoes were interpreted by experienced investigators within a high-volume laboratory, for which expertise and reproducibility for quantitative LV and RV indices have been documented.^{20, 21} RV systolic function was visually scored and quantified via TAPSE and S’, which were acquired in accordance with consensus guidelines: TAPSE was measured (on M-mode) as the distance of systolic excursion of the lateral tricuspid annulus along its longitudinal plane. S’ was measured (on tissue Doppler) as the peak tricuspid annular longitudinal velocity of excursion. Established cutoffs (TAPSE < 1.6 cm, S’<10 mm/s) were used for each parameter¹⁶; cutoffs were defined prospectively for data analysis purposes. To reduce false positive classification, RV dysfunction was defined by impairment of both TAPSE and S’.

Additional analyses were performed to assess RV remodeling and ancillary indices relevant to effort tolerance. RV size was measured as the basal linear end-diastolic diameter in a 4-chamber orientation²². Pulmonary artery (PA) systolic pressure was calculated based on tricuspid regurgitant velocity and inferior vena cava caliber. LV systolic function, geometry, and mass were quantified based on linear dimensions in parasternal long axis, consistent with quantitative methods previously validated in necropsy-comparison and population-based outcomes studies.^23–27 MR was graded in accordance with consensus guidelines;²⁸ severity was categorized using a 5-point (0–4+) scale based on aggregate data yielded by vena contracta, volumetric indices, jet depth as well as mitral and pulmonary vein flow pattern^{15, 29}.

Statistical Methods

Comparisons between groups with or without RV dysfunction were made using the Student t test (expressed as mean ± SD) for continuous variables. Categorical variables were compared using chi-square tests or, when <5 expected outcomes per cell, the Fisher exact test was used. Multiple group comparisons were performed using 2-way analysis of variance (ANOVA); ranked variables (i.e. global perfusion scores) were compared using the Kruskal-Wallis test. Multivariable logistic regression analysis was used to test whether LV perfusion deficit pattern was independently associated with RV dysfunction after controlling for magnitude of LV systolic dysfunction and RV afterload (i.e. PA pressure); variables were concomitantly entered into the multivariable model after each were confirmed to be associated with RV dysfunction in univariable analysis. Linear regression was used to test variables associated with exercise time. Statistical calculations were performed using SPSS 22.0 (SPSS Inc. [Chicago, IL]). Two-sided p<0.05 was considered indicative of statistical significance.

Results

Population characteristics

The study population comprised 2051 patients who underwent echo and MPI within 1 month (mean 5.5±7.9 days [73% within 7 days]) reflecting 15.9% (2051/12,841) of all patients who underwent MPI during study period. Among the study population, 6.0% (n=123) had echo-evidenced RV dysfunction as manifested by both impaired TAPSE and RV S’.

Table 1 details population characteristics stratified by RV functional status, including clinical conditions and patient reported baseline medications at time of MPI. As shown, diabetes mellitus and hypertension were more common among patients with RV dysfunction (p<0.05), paralleling increased prevalence of clinically reported obstructive CAD at time of stress testing (p<0.001). Consistent with the concept that adverse LV remodeling impacts RV performance, both MPI and echo demonstrated that LV systolic function was lower, LV chamber size larger, and PA pressure higher among patients with RV dysfunction (all p<0.01). Of note, Table 1 also demonstrates that less than half (40%) of patients with quantitative RV dysfunction as measured via TAPSE and S’ were identified based on qualitative visual assessment alone.

Table 1.

Population Characteristics

	Overall (n=2051)	RV Dysfunction − (n=1928)*	RV Dysfunction + (n=123)	p
Clinical
Age (years)	64 ± 12	63 ± 12	66 ± 12	0.01
Male gender	1211 (59%)	1118 (58%)	93 (76%)	<0.001
Body Mass Index (kg/m2)	28.5 ± 5.8	28.5 ± 5.9	28.0 ± 5.2	0.36
Atherosclerosis risk factors
Diabetes mellitus	528 (26%)	477 (25%)	51 (41%)	<0.001
Hypertension	1311 (64%)	1221 (63%)	90 (73%)	0.03
Tobacco use	166 (8%)	161 (8%)	5 (4%)	0.09
Hypercholesterolemia	1339 (65%)	1252 (65%)	87 (71%)	0.19
Family history of coronary artery disease	636 (31%)	602 (31%)	34 (28%)	0.41
Known coronary artery disease	566 (28%)	489 (25%)	77 (63%)	<0.001
Prior myocardial infarction	215 (10%)	197 (10%)	18 (15%)	0.12
Prior coronary revascularization (PCI / CABG)	459 (22%)	389 (20%)	70 (57%)	<0.001
Chronic Obstructive Pulmonary Disease (COPD)	84 (4%)	78 (4%)	6 (5%)	0.65
Pulmonary Embolism (prior history)	27 (1%)	25 (1%)	2 (2%)	0.68
Sleep Apnea	131 (6%)	129 (7%)	2 (2%)	0.02
Indication for stress testing
Chest Pain	1169 (57%)	1113 (58%)	56 (46%)	0.008
Dyspnea	662 (32%)	622 (32%)	40 (33%)	0.95
Medication Regimen
Aspirin	1089 (53%)	1003 (52%)	86 (70%)	<0.001
Thienopyridines	239 (12%)	223 (12%)	16 (13%)	0.63
Beta blocker	809 (39%)	726 (38%)	83 (67%)	<0.001
ACE Inhibitor/Angiotensin Receptor Blocker	830 (40%)	769 (40%)	61 (50%)	0.03
HMG CoA Reductase Inhibitor	1092 (53%)	1003 (52%)	89 (72%)	<0.001
Calcium channel blockers	419 (20%)	394 (20%)	25 (20%)	0.98
Nitrates	72 (4%)	65 (3%)	7 (6%)	0.18
Electrocardiography
Sinus Rhythm	1987 (97%)	1880 (98%)	107 (87%)	<0.001
Atrial Fibrillation or Flutter	45 (2%)	34 (2%)	11(9%)	<0.001
Q-Wave Myocardial Infarction
Anterior	31 (2%)	29 (2%)	2 (2%)	0.71
Lateral	12 (1%)	11 (1%)	1 (1%)	0.53
Inferior	46 (2%)	37 (2%)	9 (7%)	<0.001
Imaging
Right ventricular function/morphology
Echocardiography
TAPSE	2.1 ± 0.4	2.2 ± 0.4	1.3 ± 0.2	<0.001
RV S’	12.5 ± 2.5	12.8 ± 2.3	8.0 ± 1.2	<0.001
RV end-diastolic diameter (cm/m2)	1.7 ± 0.3	1.7 ± 0.3	1.8 ± 0.4	0.31
Qualitative RV Dilation	85 (4%)	62 (3%)	23 (19%)	<0.001
Qualitative RV Dysfunction	70 (3%)	21 (1%)	49 (40%)	<0.001
Left ventricular function/morphology
SPECT
Post-stress ejection fraction (%)	64 ± 11	64 ± 10	56 ± 13	<0.001
LV end-diastolic volume (ml/m2)	48 ± 17	48 ± 16	55 ± 23	<0.001
Echocardiography
Ejection fraction (%)	62 ± 9	62 ± 9	53 ± 14	<0.001
Advanced LV Dysfunction (EF < 35%)	2.6 ± 0.3	2.6 ± 0.3	2.7 ± 0.4	0.12
LV end-diastolic diameter (cm/m2)	1.7 ± 0.3	1.7 ± 0.3	1.9 ± 0.4	<0.001
LV end-systolic diameter (cm/m2)	85 ± 24	85 ± 24	90 ± 25	0.04
LV mass (g/m2)	62 ± 9	62 ± 9	53 ± 14	<0.001
Left atrial size (cm2/m2)	9.8 ± 2.3	9.8 ± 2.2	11.1 ± 3.1	<0.001
Left atrial diameter (cm/m2)	2.0 ± 0.3	2.0 ± 0.3	2.2 ± 0.4	<0.001
Mitral regurgitation (≥ moderate)	64 (3%)	55 (3%)	9 (7%)	0.01
Tricuspid regurgitation (≥ moderate)	64 (3%)	54 (3%)	10 (8%)	0.001
PASP**	31 ± 8	30 ± 7	34 ± 12	0.008
Pulmonary hypertension (PASP ≥35 mmHg)	377 (22%)	341 (21%)	36 (32%)	0.006

Numbers in boldface indicate p values < 0.05

^*Defined based on established cutoffs as employed in consensus guidelines (TAPSE < 1.6 cm, S’<10 mm/s)¹⁶

^**Available in 83% of study population (n=1711)

Myocardial Perfusion Pattern

Table 2 compares perfusion indices in relation to RV function, including its composite parameters of TAPSE and S’. As shown, global LV summed stress scores were on average nearly 3-fold higher among patients with RV dysfunction, attributable to increments in inducible (summed difference score) and fixed perfusion defects (all p≤0.001). Perfusion defect severity was paralleled by increased defect size, as measured based on greater number of segments with inducible and fixed perfusion defects among patients with RV dysfunction (both p<0.001). Regarding regionality, both inferior and lateral wall ischemia was greater among patients with RV dysfunction (both p≤0.01), whereas corresponding anterior wall defects (whether measured based on summed difference score or number of ischemic segments) were similar (p≥0.13). Table 2 also demonstrates that perfusion defect findings associated with the composite definition of RV dysfunction by both TAPSE and S’ were similar with RV dysfunction defined by each of the two individual parameters. Multivariable logistic regression analysis (Table 3) was used to test whether LV perfusion deficit pattern was independently associated with RV dysfunction after controlling for conventional factors.

Table 2.

Myocardial Perfusion Pattern

	Normal RV Function* (n=1928)	Impaired RV Function** (123)	P	Normal TAPSE (1886)	Abnormal TAPSE (165)	P	Normal RVS’ (1767)	Abnormal RVS’ (284)	P
A. GLOBAL LV PERFUSION
Summed Stress Score	2.7 ± 6.1	6.9 ±8.5	<0.001	2.6 ± 6.0	6.5 ± 8.5	<0.001	2.5 ± 5.8	5.6 ± 8.5	<0.001
(0–3|4–8 |9–13|>13)	1528 (79%)|179 (9%) |88 (5%)|133 (7%)	60 (49%)|23 (19%) |14 (11%)|26 (21%)	<0.001	1502(80%)|172(9%) |85(5%)|127(7%)	86(52%)|30(18%) |17(10%)|32(19%)	<0.001	1420(80%)|155(9%) |79(5%)|113(6%)	168(59%)|47(17%) |23(8%)|46(16%)	<0.001
Summed Rest Score	1.9 ± 5.0	4.4 ± 6.7	<0.001	1.8 ± 4.9	4.4 ± 7.1	<0.001	1.7 ± 4.7	3.9 ± 7.2	<0.001
0–3|4–8 |9–13|>13)	1620 (84%)|160 (8%) |71 (4%)|77 (4%)	75 (61%)|22 (18%) |15 (12%)|11 (9%)	<0.001	1592 (84%)|155 (8%) |65(3%)|74 (4%)	103(62%)|27 (16%) |21 (13%)|14 (9%)	<0.001	1501 (85%)|139 (8%) |65 (4%)|62 (4%)	194(68%)|43 (15%) |21 (7%)|26 (9%)	<0.001
Summed Difference Score	0.8 ± 2.7	2.5 ± 5.3	0.001	0.8 ± 2.7	2.1 ± 4.8	0.001	0.8 ± 2.7	1.8 ± 4.2	<0.001
(0|1–5 |6–10|>10)	1629(85%)|204(11%) |63(3%)|32(2%)	79(64%)|24(20%) |13(11%)|7(6%)	<0.001	1595(85%)|198(11%)| 61(3%)|32(2%)	113(69%)|30(18%) |15(9%)|7(4%)	<0.001	1510(86%)|172(10%) |56(3%)|29(2%)	198(70%)|56(20%) |20 (7%)|10(4%)	<0.001
# Ischemic Segments	0.6± 1.6	1.6± 2.8	<0.001	0.5 ± 1.6	1.3 ± 2.6	<0.001	0.5 ± 1.6	1.2± 2.3	<0.001
# Fixed Segments	0.7 ± 1.8	1.7 ± 2.5	<0.001	0.7 ± 1.8	1.7 ± 2.7	<0.001	0.7± 1.2	1.5 ± 2.6	<0.001
Normal Perfusion	1393 (72%)	50 (41%)	<0.001	1370 (73%)	72 (44%)	<0.001	1298 (74%)	145 (51%)	<0.001
Fully Reversible Defects	113 (6%)	12 (10%)	0.08	110 (6%)	15 (9 %)	0.09	99 (6%)	26 (9%)	0.02
Partially Reversible Defects	236 (12%)	29 (24%)	<0.001	225 (12%)	40 (24%)	<0.001	212(12%)	53 (19%)	0.002
Fixed Defects	186 (10%)	32 (26%)	<0.001	181 (10%)	37 (22%)	<0.001	158 (9%)	50 (21%)	<0.001
B. REGIONAL PERFUSION
Perfusion Defect Severity
Anterior
Summed Stress Score	0.6 ± 2.1	1.3 ± 2.9	0.01	0.6 ± 2.1	1.1 ± 2.7	0.02	0.5 ± 2.0	1.2 ± 3.0	0.001
Summed Rest Score	0.3 ± 1.6	0.8 ± 2.5	0.03	0.3 ± 1.6	0.7 ± 2.3	0.02	0.3 ± 1.4	0.8 ± 2.7	0.003
Summed Difference Score	0.3 ± 1.2	0.5 ± 1.4	0.14	0.3 ± 1.2	0.4 ± 1.2	0.39	0.3 ± 1.2	0.4 ± 1.2	0.06
# Ischemic Segments	0.2 ± 0.7	0.3 ± 0.8	0.13	0.2 ± 0.7	0.2 ± 0.7	0.42	0.2 ± 0.7	0.3 ± 0.8	0.02
# Fixed Segments	0.1 ± 0.6	0.4 ± 1.0	0.009	0.1 ± 0.6	0.3 ± 0.9	0.003	0.1 ± 0.6	0.3 ± 0.9	0.002
Inferior
Summed Stress Score	1.2 ± 2.9	3.0 ± 3.9	<0.001	1.2 ± 2.9	2.9 ± 4.0	<0.001	1.2 ± 2.9	2.4 ± 3.6	<0.001
Summed Rest Score	1.0 ± 2.6	2.3 ± 3.6	<0.001	1.0 ± 2.6	2.3 ± 3.7	<0.001	0.9 ± 2.6	1.9 ± 3.4	<0.001
Summed Difference Score	0.2 ± 1.0	0.7 ± 1.7	0.007	0.2 ± 1.0	0.6 ± 1.5	0.008	0.2 ± 1.0	0.4 ± 1.4	0.009
# Ischemic Segments	0.2 ± 0.7	0.4 ± 1.0	0.005	0.2 ± 0.7	0.4 ± 0.9	0.006	0.2 ± 0.7	0.3 ± 0.8	0.006
# Fixed Segments	0.4 ± 1.0	0.9 ± 1.4	<0.001	0.4 ± 1.0	0.9 ± 1.4	<0.001	0.4 ± 1.0	0.8 ± 1.3	<0.001
Lateral
Summed Stress Score	0.5 ± 2.0	2.0 ± 3.3	<0.001	0.5 ± 2.0	1.8 ± 3.4	<0.001	0.4 ± 1.9	1.4 ± 3.1	<0.001
Summed Rest Score	0.3 ± 1.6	0.7 ± 1.9	0.008	0.3 ± 1.5	0.8 ± 2.3	0.005	0.2 ± 1.5	0.6 ± 2.1	0.004
Summed Difference Score	0.2 ± 1.1	1.2 ± 2.6	<0.001	0.2 ± 1.1	1.0 ± 2.4	<0.001	0.2 ± 1.1	0.8 ± 2.1	<0.001
# Ischemic Segments	0.1 ± 0.6	0.8 ± 1.5	<0.001	0.1 ± 0.6	0.6 ± 1.4	<0.001	0.1 ± 0.6	0.5 ± 1.2	<0.001
# Fixed Segments	0.1 ± 0.5	0.2 ± 0.7	0.01	0.1 ± 0.4	0.3 ± 0.8	0.004	0.1 ± 0.4	0.2 ± 0.7	0.004

^*Defined as both normal TAPSE and S’, via established cutoffs as employed in consensus guidelines (TAPSE ≥1.6 cm, S’≥10 mm/s)¹⁶

^**Defined as both abnormal TAPSE and S’ (TAPSE < 1.6 cm, S’<10 mm/s)

Table 3.

Perfusion-Based Correlates of Right Ventricular Systolic Dysfunction

				Model chi-square= 93.96, p < 0.001
	Univariable			Multivariable
Variable	Odds Ratio	95% Confidence Interval	P	Odds Ratio	95% Confidence Interval	P
Inferior/Lateral Ischemic Perfusion Defect Size* (# segments)	1.40	1.29 – 1.56	<0.001	1.39	1.24 – 1.57	<0.001
Inferior/Lateral Fixed Perfusion Defect Size* (# segments)	1.31	1.19 – 1.45	<0.001	1.15	0.99 – 1.32	0.06
LV Ejection Fraction (per 10 point decrement)	1.71	1.54 – 1.88	<0.001	1.62	1.41 – 1.83	<0.001
Pulmonary Arterial Systolic Pressure (per 10 mmHg)	1.44	1.22 – 1.67	<0.001	1.19	0.93 – 1.46	0.16

^*Inferior/lateral defect size was calculated by summing aggregate number of affected segments in these territories

As shown in Table 3, inducible myocardial perfusion defects (as measured in the inferior or lateral walls) increased the likelihood for RV dysfunction (both p<0.05). Of note, modeling data demonstrated that the relationship between RV dysfunction and LV perfusion deficits was continuous, such that likelihood for RV dysfunction increased in relation to size of LV perfusion deficit. Applied clinically, results indicate that ischemic perfusion deficits involving 3 LV segments in the inferior or lateral walls would be expected to more than double likelihood for RV dysfunction even after controlling for magnitude of LV dysfunction and pulmonary hypertension.

RV Function in Relation to Effort Tolerance

All patients initiated a standardized Bruce exercise treadmill protocol, for which a target of 85% predicted maximal heart rate (220-age) was required for test completion. As shown in Table 4, exercise-induced hemodynamic indices, including change in heart rate and systolic blood pressure were lower among patients with RV dysfunction (both p<0.001): Results paralleled differences among patients with LV systolic dysfunction defined by EF<50% (Δ heart rate: 60±27 vs. 70±24 bpm |Δ blood pressure: 33±31 vs. 42±27 mmHg, both p≤0.001). Patients with RV dysfunction also demonstrated lesser effort tolerance whether measured by exercise duration or peak Bruce stage achieved (both p<0.001), again paralleling results among patients with LV dysfunction (7.0±2.9 vs. 8.0±2.9, | 2.7±1.0 vs. 3.1±1.0; both p<0.001). Accordingly, both groups were more likely to require conversion of exercise testing to pharmacologic stress (RV dysfunction: 35% vs. 19%, p<0.001 | LV dysfunction: 27% vs. 19%, p=0.03).

Table 4.

Exercise Physiological Parameters in Relation to RV Function

	Overall (n=2051)	RV Dysfunction − (n=1928)	RV Dysfunction + (n=123)	p
Effort Tolerance
Exercise Duration (time)	7.8 ± 2.9	7.9± 2.9	6.7 ± 2.8	<0.001
Peak Treadmill Stage Achieved	2.6 ± 0.96	2.6 ± 0.96	2.2 ± 0.91	<0.001
Workload (METS)	9.4 ± 2.1	9.5 ± 2.1	8.5 ± 2.3	<0.001
Heart Rate
Rest	70.4 ± 12.3	70.3 ± 12.2	72.1± 13.3	0.15
Peak Stress	138.4 ± 25.2	139.1 ± 25.0	128.3± 26.9	<0.001
Change in Heart Rate (stress - rest)	68.0 ± 24.4	68.8 ± 24.1	56.2± 26.4	<0.001
Predicted maximum heart rate	156.5±11.7	156.7 ± 11.7	154.0± 11.5	0.014
% Converted to pharm stress testing*	415 (20%)	372 (19%)	43 (35%)	<0.001
Systolic Blood Pressure
Rest	126.4 ± 17.0	126.3 ± 16.9	127.7 ± 18.0	0.40
Peak Stress	166.6 ± 29.2	167.3 ± 28.4	155.5 ± 32.3	0.001
Change in Blood Pressure (stress - rest)	40.2 ± 27.7	41.0 ± 27.3	27.9 ± 31.0	<0.001
Exercise ECG Response
≥ 1 mm ST Depression (%)	389 (19%)	371(19%)	18(15%)	0.21
Maximal ST Depression (mm)	0.32 ± 0.75	0.33 ± 0.74	0.30 ± 0.84	0.67
Exercise Clinical Response
Chest pain	104(5%)	94 (5%)	10(8%)	0.11
Shortness of breath	273(13%)	254 (13%)	19 (15%)	0.47
Duke Treadmill Score	4.9 ±5.5	5.0 ± 5.4	4.2 ± 6.5	0.13

^*Failure to attain target heart rate (220-age)

Figure 2 stratifies exercise time in relation to RV and LV systolic function. As shown, effort tolerance decreased in relation to both functional parameters, as evidenced by stepwise decrements in patients with normal biventricular function compared to those with isolated ventricular dysfunction, and biventricular dysfunction. Of note, effort tolerance was more than 1 minute lower among patients with isolated RV dysfunction as compared to those with normal biventricular function (6.9±2.6 vs. 8.1±2.9, p=0.001), paralleling similarly lower exercise duration among patients with isolated LV dysfunction compared to normals (7.1±2.9 vs. 8.1±2.9, p<0.001). RV and LV dysfunction were each associated with impaired exercise time in univariable linear regression (p<0.001): In multivariable analysis, impaired effort tolerance (assessed per minute exercise time) was independently associated with RV dysfunction (regression coefficient 0.73 [95% confidence interval 0.14–1.31], p=0.02) and LV dysfunction (0.71 [0.27–1.16], p=0.002), even after controlling for age (0.66 per decade [0.55–0.77], p<0.001) and β-blocker use (0.66 [0.39–0.93], p<0.001) at time of stress testing.

Figure 2. Exercise Time in Relation to Ventricular Function.

Exercise time (during Bruce treadmill protocol) among groups partitioned based on LV and RV systolic function (LV dysfunction p=0.034, RV dysfunction p=0.008, interaction p=0.38). As shown, exercise time decreased in relation to extent of systolic dysfunction, as evidenced by stepwise decrements between patients with normal biventricular function, isolated LV or RV dysfunction, and those with biventricular dysfunction.

Discussion

This study – performed among a broad cohort of patients undergoing stress MPI and echo - provides new insights concerning links between LV ischemia and RV performance, as well as the independent impact of RV dysfunction on effort tolerance. There are several key findings: First, while RV dysfunction was uncommon (6%) among the overall cohort of patients undergoing MPI, prevalence was higher among patients with CAD risk factors and markedly increased (63%) among those with known CAD (p<0.001). Second, regional LV perfusion impacted likelihood of RV dysfunction, as evidenced by the fact that both inferior and lateral wall ischemia was greater among patients with RV dysfunction (both p≤0.01), whereas corresponding anterior wall deficits (whether measured based on summed difference score or number of ischemic segments) were similar. In multivariable analysis, inducible perfusion defects in the LV inferior or lateral walls were independently associated with RV dysfunction even after controlling for magnitude of LV dysfunction, fixed defects, and PA systolic pressure. Finally, multiple physiologic indices of effort tolerance (e.g. Δ heart rate, Δ blood pressure, effort time) were impaired among patients with RV dysfunction (all p<0.001), paralleling results among patients with LV dysfunction: exercise time decreased stepwise in relation to both RV and LV dysfunction (p<0.001), and was associated with each parameter independent of age or medication regimen.

To the best of our knowledge, this is the first study to demonstrate a link between LV ischemia and RV function. Our observed association between inferior and lateral wall inducible perfusion defects and RV dysfunction is consistent with established concepts regarding LV perfusion pattern and RV blood flow. Prior studies have shown inferior and lateral wall perfusion deficits on MPI to correspond to occlusion of the right coronary and left circumflex arteries,³⁰ each of which would be expected to variably provide blood flow to the RV based on coronary dominance pattern. Of note, anterior ischemia was not associated with RV dysfunction despite the fact that the left anterior descending artery often provides blood flow to the RV apex. One possible explanation for this concerns the measurement technique used to assess RV function: whereas TAPSE and S’ are included in consensus echo guidelines as well-validated indices for RV function,¹⁶ both are measures of tricuspid annular/basal RV excursion and would not necessarily be impacted by regional dysfunction of the RV apex. Concerning our observation that anterior fixed defects were larger among patients with RV dysfunction, it is possible that the distal RV (as may be supplied by the left anterior descending artery) contributes relatively little to global RV performance, such that anterior ischemia itself bears minimal consequences for RV function, but that anterior infarction produces adverse LV remodeling that bears secondary consequences on RV performance. Consistent with this, several prior studies have shown RV dysfunction (measured via echocardiography, radionuclide cine-angiography, or CMR) to be less common among patients with anterior MI,^{2, 4} but present in 17–40% of such cases.^{2, 4, 31} Taken together, these data suggest that our findings are consistent with prior literature, and not primarily attributable to measurement techniques.

Our observed link between RV dysfunction and reduced exercise supports the notion that RV dysfunction alone can impede effort tolerance – a physiologic parameter widely used to stratify prognosis that has been shown to predict cardiovascular and overall mortality in prior studies.^7–10 Among our cohort, effort tolerance was over 1 minute lower among patients with isolated RV dysfunction compared to those with normal biventricular function, paralleling similar magnitude of difference when patients with isolated LV dysfunction were compared to normals (both p≤0.001). Our data among patients with known or suspected CAD extends upon prior studies in smaller cohorts which have suggested an association between RV dysfunction and exercise capacity, but have not directly tested whether RV dysfunction provides incremental stratification above that yielded by LV function alone. For example, among a cohort of 23 healthy adolescents undergoing cardiopulmonary exercise testing, Pieles et al. reported that both RV and LV strain increased in relation to work rate (p<0.05).³² Moreover, among 44 chronic obstructive pulmonary disease patients, Caminiti et al. reported that those with RV dysfunction (as defined via TAPSE) had lower effort tolerance on baseline 6 minute walk test (p=0.02) and lesser improvement following cardiopulmonary rehabilitation (p<0.001).³³ Other studies conducted among patients with LV systolic dysfunction have linked reduced RVEF to impaired exercise capacity as measured via peak oxygen consumption (V0₂ max) during bicycle ergometry or treadmill exercise.^34–36 Regarding mechanism, it is possible that impaired RV function results in decreased LV preload, thereby altering LV pressure-volume filling curves and impeding LV contractile mechanics. It is also possible that abnormal RV function (as measured by TAPSE or S’) is a marker for subtle changes in LV systolic function not captured by widely employed indices such as LVEF, such that LV contractile function is the primary determinant of effort tolerance and RV dysfunction a secondary marker. Regardless of mechanism, our data indicate RV dysfunction confers increased likelihood for impaired exercise capacity, adding to a growing body of literature demonstrating the importance of RV function as a marker of effort tolerance.^{1–3, 34–36}

Several limitations should be noted. First, it is important to recognize that SPECT does not assess absolute quantitative blood flow, and that this modality does not provide sufficient spatial resolution to assess RV perfusion. On the other hand, SPECT is widely employed to assess ischemic burden, enabling us to study a broad population-based cohort using a perfusion approach that is well validated for prognostic risk stratification.⁵ Second, our study assessed RV performance based on TAPSE and S’ and used established binary cutoffs rather than examining magnitude of dysfunction based on volumetric RVEF. While TAPSE and S’ are included in consensus guidelines¹⁶ and can be readily applied in clinical practice and research, these methods extrapolate global function based on tricuspid annular regional excursions, potentially excluding the effect of RV apical dysfunction. It is thus possible that our results would have differed somewhat were RV function assessed volumetrically using 3-dimensional (3D) echo or CMR, which are known to improve RV assessment compared to 2D imaging.^{3, 37} On the other hand, TAPSE and S’ have been shown to correlate with volumetric RV function,²⁰ and offer a means of RV assessment for patients in whom 3D methods (e.g. CMR) are contraindicated or unavailable, providing a rationale for our imaging approach. An additional limitation concerns the fact that chest pain history was assessed via patient questionnaire at time of exercise MPI (for which available data was insufficient to classify chest pain features) and that exercise treadmill data did not include post exercise variables such as heart rate recovery. It is also important to note that SPECT assessment of ischemia burden was performed via standardized scoring of perfusion defect severity and did not include ancillary indices such as lung-heart-ratio. Finally, due to large number of statistical tests employed, type I error maybe inflated using the pre-specified p value cutoff (<0.05) as the significance threshold per test.

Our results bear several key implications for clinical practice and translational research. First, our finding of an independent association between RV dysfunction and decreased exercise treadmill time supports the notion that RV performance should be evaluated as part of diagnostic testing among patients with impaired effort capacity or associated symptoms. Second, our results add to growing literature supporting use of quantitative RV assessment for stratification of physiologic outcomes, and demonstrate that simple RV visual assessment (an approach widely used in current clinical practice) can be limited - as evidenced by the fact that less than half (40%) of patients in our cohort with quantitative RV dysfunction were identified qualitatively. Finally, given our demonstration of links between LV inferior/lateral ischemia and RV dysfunction, future multicenter studies are warranted to further confirm our findings and test whether targeted coronary revascularization in these regions can be used as a therapeutic means of augmenting RV contractility and clinical performance status among patients with CAD.

Clinical Perspective.

Right ventricular (RV) systolic dysfunction is a strong predictor of adverse clinical outcomes including heart failure and death. The right and left ventricle (LV) are closely linked due to a variety of mechanisms, including coronary blood supply that commonly perfuses regions of the LV and RV. LV ischemia has the potential to affect RV contractile function both by direct (i.e. altered common blood supply) and indirect mechanisms (i.e. increased RV afterload). In this study, 2051 patients underwent exercise myocardial perfusion imaging (MPI) and echocardiography (echo) within a narrow time interval – MPI was used to measure LV ischemia, and echo to assess RV function (via TAPSE and S’). Findings demonstrate global LV ischemic burden to be higher among patients with RV dysfunction. Consistent with the concept that regional coronary perfusion has the potential to impact RV performance, inferior and lateral ischemia were independently associated with RV dysfunction. Regarding exercise stress, results demonstrate multiple indices of effort tolerance to be impaired among patients with RV dysfunction: Exercise time decreased stepwise in relation to both RV and LV dysfunction and was independently associated with each parameter. These findings support the concept that RV performance impacts exercise tolerance, and should be assessed in patients with decreased effort tolerance or related heart failure symptoms. Given our demonstration of a link between regional LV ischemia and RV dysfunction, current findings lay the groundwork for future studies to test the concept of targeted coronary revascularization as a therapeutic strategy for augmenting RV contractile function.