- Email: [email protected]
Relation of Right Ventricular Mass and Volume to Functional Health Status in Repaired Tetralogy of Fallot
Relation of Right Ventricular Mass and Volume to Functional Health Status in Repaired Tetralogy of Fallot Jimmy C. Lu, MDa,b,*, Jason T. Christensen, MDa, Sunkyung Yu, MSa, Janet E. Donohue, BSa, Maryam Ghadimi Mahani, MDb, Prachi P. Agarwal, MBBSc, and Adam L. Dorfman, MDa,b After repair of tetralogy of Fallot, right ventricular (RV) mass and mass:volume ratio may reflect RV remodeling and adverse outcomes. This study aimed to evaluate the relation of RV mass to functional health status and subsequent adverse RV remodeling and to determine whether RV mass measurement in systole could improve reproducibility. In 53 patients with tetralogy of Fallot (median 29 years old) who previously underwent cardiovascular magnetic resonance and completed the Short Form 36, version 2 (Optum, Eden Prairie, MN), short-axis images were analyzed for RV end-diastolic volume and diastolic and systolic mass, indexed to body surface area. The most recent subsequent cardiovascular magnetic resonance study (before pulmonary valve or conduit replacement) was evaluated for change in RV end-diastolic volume and ejection fraction. Diastolic indexed mass ‡37.3 g/m2 (odds ratio 7.6, p [ 0.02) predicted decreased general health scores. In patients with normal RV ejection fraction, indexed mass correlated with Physical Component Summary and general health scores. RV diastolic mass:volume ratio >0.2 had a strong association with subsequent increase in RV end-diastolic volume (odds ratio 26.1, p [ 0.002). Systolic RV mass measurement had excellent correlation with diastolic measurement (r [ 0.97, p <0.0001), but did not improve intraobserver or interobserver variability. In conclusion, RV mass relates to functional health status and adverse RV remodeling and can be measured with good reproducibility. RV mass should be routinely evaluated in this population and is best measured in diastole; further study is necessary to evaluate longitudinal changes in functional health status and RV parameters. Ó 2014 Elsevier Inc. All rights reserved. (Am J Cardiol 2014;114:1896e1901) After repair of tetralogy of Fallot (TOF), patients often have free pulmonary insufficiency and a varying degree of right ventricular (RV) outflow tract obstruction, but the optimal timing of pulmonary valve replacement is unclear. Much attention has focused on ventricular size and function,1e3 but recent data suggest that RV mass, specifically mass:volume ratio, may be an important marker of ventricular remodeling, as it predicts adverse outcomes in patients with repaired TOF.4 The potential relation between RV mass and functional health status has not been demonstrated. In addition, although RV mass measurement by cardiovascular magnetic resonance (CMR) has been shown to accurately reflect actual RV weight on autopsy,5 several studies have shown lesser reproducibility of RV mass measurement than other parameters,6e11 which may be related to the thin wall of the RV in diastole. It is unclear whether this measurement would be more reliable in systole,12 when the RV thickens, and borders may be more clearly defined. This study aimed to determine whether RV
a Division of Pediatric Cardiology, Department of Pediatrics and Communicable Diseases, bSection of Pediatric Radiology, Department of Radiology and cDivision of Cardiothoracic Radiology, Department of Radiology, University of Michigan, Ann Arbor, Michigan. Manuscript received July 3, 2014; revised manuscript received and accepted September 17, 2014. See page 1900 for disclosure information. *Corresponding author: Tel: (734) 647-8924; fax: (734) 936-9470. E-mail address: [email protected] (J.C. Lu).
0002-9149/14/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2014.09.027
mass in patients with repaired TOF is related to functional health status and RV remodeling in this population, and whether measurement variability would improve with measurement in systole rather than diastole. Methods This study included patients with repaired TOF, age 14 years, who were referred for clinically indicated CMR from June 2008 to December 2009 and were prospectively enrolled as part of a previous cross-sectional study evaluating CMR predictors of functional health status.13 Patients were excluded if they had pulmonary atresia or absent pulmonary valve, additional significant cardiac malformations (e.g., atrioventricular septal defect), contraindications to CMR or gadolinium, CMR performed under anesthesia, or cognitive impairment preventing completion of the health status assessment. Additionally, patients were also excluded if the health status assessment was completed >3 months from time of CMR or if short-axis images were not available for analysis. CMR was performed using a commercially available 1.5 T scanner (Philips Intera Achieva, Best, The Netherlands). Cine images were obtained with a breathhold, electrocardiographic-gated, segmented k-space, steady-state free precession sequence. End-diastolic and end-systolic phases were manually chosen for the RV. RV endocardial and epicardial contours were manually drawn in both end-diastole and end-systole on short-axis images, using QMass software (Medis, Leiden, The Netherlands). www.ajconline.org
Congenital Heart Disease/RV Mass and Health Status in Tetralogy of Fallot Table 1 Clinical characteristics of the cohort (N¼53). Data are presented as number (%) for categorical variables and mean standard deviation or median (interquartile range) for continuous variables, as appropriate Male Age at CMR (years) Age at repair (years) Body surface area (m2) Type of TOF Pulmonary stenosis Unknown Type of TOF repair Transannular patch Valve-sparing Conduit Unknown Left ventricular ejection fraction (%) RV end-diastolic indexed volume (ml/m2) RV end-systolic indexed volume (ml/m2) RV ejection fraction (%) RV diastolic indexed mass (g/m2) RV systolic indexed mass (g/m2) RV diastolic mass:volume ratio (g/ml)
31 (58.5%) 29.4 (20.8-41.3) 3.5 (0.8-5.4) 1.88 0.27 42 (72.4%) 11 (19.0%) 31 (58.5%) 10 (18.9%) 4 (7.5%) 8 (15.1%) 53.6 7.5 148.7 49.5 83.9 36.4 44.8 9.0 27.7 7.7 24.9 7.4 0.19 0.04
CMR ¼ cardiovascular magnetic resonance; RV ¼ right ventricular; TOF ¼ tetralogy of Fallot.
Contours were compared with long-axis images to confirm the plane of the tricuspid valve annulus. RV mass was indexed to body surface area or to end-diastolic volume (mass:volume ratio). On enrollment, typically on the day of CMR, subjects completed the Short Form 36, version 2 (SF-36), which provides a measure of multiple domains of health and functioning.14 The Physical Component Summary score and the subscales of physical functioning (ability to perform physical activities), role-physical (participation in work or usual activities without physical limitations), and general health (perception of general health and its likelihood to change) were chosen to represent the physical impact of the disease. Scores were normalized for age based on population-based normative data. As previous reports demonstrated only small differences between patients with TOF and controls,13,15 an age-adjusted z-score 1 was considered a clinically significant decrease in functional health status. To evaluate the relation of RV mass to subsequent remodeling, a subset of patients from the initial cohort with an additional clinically indicated CMR study was evaluated. Measurements were taken from the most recent CMR, before any pulmonary valve or conduit replacement. A relative increase in RV end-diastolic volume of 10% from the initial measurement or an absolute decrease in RV ejection fraction of 5% was considered clinically significant. RV indexed mass and mass:volume ratio on the initial CMR study was compared with subsequent increase in RV volume or decrease in RV ejection fraction. For reproducibility of diastolic and systolic measurements, all CMR studies were analyzed by a single observer (JCL) and reanalyzed at least 1 month later for intraobserver variability. For interobserver variability, a subset of 29 patients was analyzed by a second observer (JTC), blinded to the initial measurements.
1897
Data are presented as frequency with percent for categorical variables and mean SD or median with interquartile range as appropriate for continuous variables. Receiver operating characteristic curves were used to determine the cutoffs of each RV mass measurement for predicting SF-36 scores, optimizing specificity. Sensitivity, specificity, and positive predictive value for the cutoffs were reported. Odds ratios with 95% confidence intervals were calculated for clinically significant decreased health status or changes in RV end-diastolic volume or ejection fraction. Changes in RV end-diastolic volume or ejection fraction from initial assessment to follow-up measurement were examined using the paired t test. Correlations between systolic and diastolic RV measurements, as well as between RV mass measurements and SF-36 scores in the patients with normal RV ejection fraction, were also evaluated by Pearson or Spearman correlation coefficient, as appropriate. RV ejection fraction was previously shown to be the best predictor of functional health status,13 and the subset of patients with normal RV ejection fraction was chosen to determine whether RV mass could be an adjunctive measure for further risk stratification. Values of p <0.05 were considered statistically significant. Intraclass correlation coefficients, coefficients of variation, and limits of agreement were used to evaluate interobserver and intraobserver variability on each RV measurement. Results In the initial cohort, 62 of 65 consecutive patients met inclusion criteria (1 declined participation, and an investigator was not available for consent for 2). Of these patients, 5 had pulmonary atresia or absent pulmonary valve and were excluded; 3 did not complete the SF-36 within 3 months of CMR. In 1 patient, short-axis images were not available for analysis. Patient and clinical characteristics of the remaining 53 patients are described in Table 1. SF-36 scores in this population were previously reported13 and were similar to the population norm, except for a small but statistically significant decrease in general health. RV diastolic indexed mass 37.3 g/m2 conveyed increased odds of decreased general health scores (Figure 1), although Physical Component Summary scores did not reach significance. Diastolic indexed mass 37.3 g/m2 had 95% specificity, 31% sensitivity, and 71% positive predictive value for decreased general health scores. RV systolic indexed mass did not reach significance for decreased general health scores. In the subset of patients with normal RV systolic function (ejection fraction 45%), indexed RV mass (either diastolic or systolic) correlated with Physical Component Summary scores and general health subscale scores (Table 2). RV enddiastolic volume did not correlate with functional health status. Only 1 patient (2%) in this cohort had RV diastolic mass:volume ratio >0.3. Neither RV diastolic mass:volume ratio 0.2 nor systolic mass:volume ratio 0.19 predicted decreased SF-36 scores. A subset of 24 patients had subsequent CMR before pulmonary valve or conduit replacement, a median 3.7 years (interquartile range 2.5 to 4.0) after initial study. There was
1898
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Odds ratios (and 95% confidence intervals) for decreased SF-36 scores (age-adjusted z-score 1), with elevated RV diastolic indexed mass (A) or RV systolic indexed mass (B). GH ¼ general health; RP ¼ role-physical; PCS ¼ Physical Component Summary; PF ¼ physical functioning.
Table 2 Correlation of RV diastolic or systolic indexed mass with Short Form 36 age-adjusted z-scores, in patients with normal RV systolic function (N¼28). Data are presented as r, Spearman correlation coefficient (p-value) Diastolic Physical Component Summary Physical Functioning Role Physical General Health
0.41 0.18 0.20 0.47
(0.03) (0.37) (0.31) (0.01)
Systolic 0.48 0.30 0.30 0.59
(0.01) (0.13) (0.12) (0.001)
no significant difference in pulmonary regurgitant fraction at follow-up (33.7 19.7% vs 32.0 21.0%, p ¼ 0.38), but indexed RV end-diastolic volume increased at the follow-up study (148 42 vs 137 43 ml/m2, p ¼ 0.02), with 9 patients having an increase >10% from the initial measurement. There was also a small but statistically significant decrease in RV ejection fraction (47 8% vs 45 9%, p ¼ 0.03), with only 1 patient having an absolute decrease of 5% in ejection fraction. RV diastolic mass:volume ratio >0.2 was associated with subsequent increase in RV end-diastolic volume (odds ratio 26.1, 95% confidence interval 2.9 to 765, p ¼ 0.002). RV indexed mass measurement in systole correlated very well with diastolic measurement (Figure 2), but mildly underestimated indexed mass relative to diastole (mean difference 2.8 g/m2, limits of agreement 0.9 to 6.5). RV systolic mass:diastolic volume ratio also correlated very well with diastolic measurement (Figure 2), with minimal underestimation relative to diastole (mean difference 0.02, limits of agreement 0.01 to 0.05). Intraobserver and interobserver reproducibility of all RV measurements was excellent (Table 3). Bland-Altman plots for limits of agreement are presented in Figures 3 and 4. Reproducibility of systolic mass was similar to minimally worse than reproducibility of diastolic mass measurements. Discussion In patients with repaired TOF, elevated RV mass relates to decreased functional health status and predicts progressive RV dilatation. RV mass should be routinely measured in this population, and can be measured accurately and reproducibly in either systole or diastole, although diastolic measurements are recommended, as they are slightly more reproducible, with a better relation to outcomes. The relation of RV indexed mass to some but not all SF-36 subscores underscores the multifactorial nature of
functional health status in this population. As RV mass is unlikely to be used in isolation in determining intervention thresholds, sensitivity was less essential, and specificity was prioritized in selecting a cutoff value, to minimize false positives. This led to a greater indexed mass cutoff, with few patients in this cohort at extremes of elevated mass. However, the correlation of indexed mass with functional health status in the context of normal RV ejection fraction suggests this may be a useful adjunctive measure for risk stratification and should be routinely evaluated in this population. The relation to functional health status is consistent with findings in aortic valve disease, in which left ventricular indexed mass predicts left ventricular systolic dysfunction16 and outcomes post aortic valve replacement.17,18 RV indexed mass likely is a marker of adverse remodeling, either due to residual RV outflow tract obstruction, volume load from dilatation, or systolic or diastolic dysfunction. This is also consistent with the International Multicenter TOF Registry (INDICATOR) cohort, which found RV mass:volume ratio 0.3 as a predictor for death or sustained ventricular tachycardia.4 In that study, 41% of patients with adverse outcomes had pulmonary atresia, a subgroup at risk for conduit stenosis or pulmonary hypertension, which was excluded in this study. This may explain why only 1 patient in this cohort had mass:volume ratio 0.3 and why we could not resolve mass:volume ratio as a significant predictor. However, the consistent association of RV mass to outcome, even when excluding patients with pulmonary atresia, underscores the potential applicability of RV mass measurement in this population. We do not believe that elevated indexed mass is simply a surrogate for dilatation, as we had previously shown no relation of indexed end-diastolic volume with functional health status.13 Similarly, RV volume did not relate to outcomes in the INDICATOR cohort. However, the trend toward increased progressive dilatation in patients with elevated mass suggests an interplay between wall thickness and volume. Increased RV mass appears to reflect not only the result of volume and/or pressure overload but may also portend further adverse remodeling. Due to conservation of mass, systolic and diastolic mass should be identical. We did not have pathologic specimens or a gold standard, and thus it is not clear which of these methods is truly more accurate. Given the thin RV wall in diastole, it is conceivable that the diastolic mass measurement is actually overestimated, as very small overestimation in wall thickness would increase the RV mass calculation, with small errors magnified over a dilated ventricle. Blalock
Congenital Heart Disease/RV Mass and Health Status in Tetralogy of Fallot
1899
Figure 2. Correlation of systolic versus diastolic measurement of RV indexed mass (A) and mass:volume ratio (B). Table 3 Intraobserver and interobserver variability of right ventricular measurements. RV EDV, g/m2
Intra-observer COV (%) ICC (95% CI) Inter-observer COV (%) ICC (95% CI)
RV ESV, g/m2
RVEF, %
RV indexed mass, g/m2
RV mass:volume ratio
diastolic
systolic
diastolic
systolic
2.4 1.00 (1.00-1.00)
3.8 1.00 (0.99-1.00)
7.2 0.93 (0.89-0.96)
8.0 0.96 (0.93-0.98)
9.9 0.94 (0.91-0.97)
9.7 0.87 (0.79-0.92)
10.0 0.91 (0.85-0.95)
12.1 0.96 (0.91-0.98)
11.4 0.97 (0.94-0.99)
9.5 0.81 (0.66-0.91)
11.1 0.93 (0.85-0.96)
13.0 0.88 (0.77-0.94)
15.9 0.78 (0.60-0.89)
18.0 0.75 (0.56-0.87)
CI ¼ confidence interval; COV ¼ coefficient of variation; EDV ¼ end-diastolic volume; EF ¼ ejection fraction; ESV ¼ end-systolic volume; ICC ¼ intraclass correlation coefficient; RV ¼ right ventricular.
Figure 3. Intraobserver limits of agreement for measurement of RV diastolic indexed mass (A), RV systolic indexed mass (B), RV diastolic mass:volume ratio (C), and RV systolic mass:diastolic volume ratio (D). Solid lines represent mean difference and dashed lines represent 1.96 SDs.
1900
The American Journal of Cardiology (www.ajconline.org)
Figure 4. Interobserver limits of agreement for measurement of RV diastolic indexed mass (A), RV systolic indexed mass (B), RV diastolic mass:volume ratio (C), and RV systolic mass:diastolic volume ratio (D). Solid lines represent mean difference and dashed lines represent 1.96 SDs.
et al suggested that systolic measurement may improve reproducibility, but did not reach statistical significance.12 The current study had greater intraclass correlation coefficients than in the previous report and thus would be less likely to detect a difference; the intraclass correlation coefficient for systolic mass was actually slightly lesser than that for diastolic mass. Although not seen in the current study, others have reported decreased reproducibility of systolic volumes,6,8 likely related to wall thickening and difficulty in distinguishing the blood-endocardial border, which may counteract any gain in tracing a thicker wall. With no clear advantage to measurement in systole, we recommend RV mass measurement in diastole. Several limitations should be acknowledged. The sample size was limited to the previous cohort that completed the SF-36 survey. We did not evaluate differences in RV measurements from axial or other planes, as our laboratory measures ventricular volumes from a short-axis plane, and full axial stacks through the ventricles were not available for analysis. A limited definition of functional health status was used, focusing on patients’ functional ability and perception of their health status. Although this is subjective, patientcentered metrics are an important consideration in clinical decision-making. Objective measurements such as exercise performance should also be considered, but few patients in this cohort had a contemporaneous exercise test. The overall population with TOF is heterogeneous, and this cohort with a later age of repair than current standards may differ from a more contemporary cohort. However, this cohort is similar
to current adult populations for whom intervention is being considered. The cross-sectional nature limits evaluation of whether potential markers predict changes in functional health status. We do not believe that the inclusion criterion of clinical referral for CMR was a significant source of sampling bias because CMR is a routine part of clinical care for this population at our center. Disclosures The authors have no conflicts of interest to disclose. 1. Therrien J, Siu SC, McLaughlin PR, Liu PP, Williams WG, Webb GD. Pulmonary valve replacement in adults late after repair of tetralogy of Fallot: are we operating too late? J Am Coll Cardiol 2000;36: 1670e1675. 2. Knauth AL, Gauvreau K, Powell AJ, Landzberg MJ, Walsh EP, Lock JE, del Nido PJ, Geva T. Ventricular size and function assessed by cardiac MRI predict major adverse clinical outcomes late after tetralogy of Fallot repair. Heart 2008;94:211e216. 3. Geva T. Repaired tetralogy of Fallot: the roles of cardiovascular magnetic resonance in evaluating pathophysiology and for pulmonary valve replacement decision support. J Cardiovasc Magn Reson 2011;13:9. 4. Valente AM, Gauvreau K, Assenza GE, Babu-Narayan SV, Schreier J, Gatzoulis MA, Groenink M, Inuzuka R, Kilner PJ, Koyak Z, Landzberg MJ, Mulder B, Powell AJ, Wald R, Geva T. Contemporary predictors of death and sustained ventricular tachycardia in patients with repaired tetralogy of Fallot enrolled in the INDICATOR cohort. Heart 2014;100:247e253. 5. Beygui F, Furber A, Delepine S, Helft G, Metzger JP, Geslin P, Le Jeune JJ. Routine breath-hold gradient echo MRI-derived right
Congenital Heart Disease/RV Mass and Health Status in Tetralogy of Fallot
6. 7. 8.
9.
10.
11.
12.
ventricular mass, volumes and function: Accuracy, reproducibility and coherence study. Int J Cardiovasc Imaging 2004;20:509e516. Mooij CF, de Wit CJ, Graham DA, Powell AJ, Geva T. Reproducibility of MRI measurements of right ventricular size and function in patients with normal and dilated ventricles. J Magn Reson Imaging 2008;28:67e73. Grothues F, Moon JC, Bellenger NG, Smith GS, Klein HU, Pennell DJ. Interstudy reproducibility of right ventricular volumes, function, and mass with cardiovascular magnetic resonance. Am Heart J 2004;147:218e223. Luijnenburg SE, Robbers-Visser D, Moelker A, Vliegen HW, Mulder BJ, Helbing WA. Intra-observer and interobserver variability of biventricular function, volumes and mass in patients with congenital heart disease measured by CMR imaging. Int J Cardiovasc Imaging 2010;26:57e64. Buechel EV, Kaiser T, Jackson C, Schmitz A, Kellenberger CJ. Normal right- and left ventricular volumes and myocardial mass in children measured by steady state free precession cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2009;11:19. Catalano O, Antonaci S, Opasich C, Moro G, Mussida M, Perotti M, Calsamiglia G, Frascaroli M, Baldi M, Cobelli F. Intra-observer and interobserver reproducibility of right ventricle volumes, function and mass by cardiac magnetic resonance. J Cardiovasc Med (Hagerstown) 2007;8:807e814. Bradlow WM, Hughes ML, Keenan NG, Bucciarelli-Ducci C, Assomull R, Gibbs JS, Mohiaddin RH. Measuring the heart in pulmonary arterial hypertension (PAH): Implications for trial study size. J Magn Reson Imaging 2010;31:117e124. Blalock SE, Banka P, Geva T, Powell AJ, Zhou J, Prakash A. Interstudy variability in cardiac magnetic resonance imaging measurements
13.
14. 15.
16. 17.
18.
1901
of ventricular volume, mass, and ejection fraction in repaired tetralogy of Fallot: a prospective observational study. J Magn Reson Imaging 2013;38:829e835. Lu JC, Cotts TB, Agarwal PP, Attili AK, Dorfman AL. Relation of right ventricular dilation, age of repair, and restrictive right ventricular physiology with patient-reported quality of life in adolescents and adults with repaired tetralogy of Fallot. Am J Cardiol 2010;106: 1798e1802. Ware JEJ, Kosinski M, Bjorner JB, Turner-Bowker DM, Gandek B, Maruish ME. User’s Manual for the Sf-36v2 Health Survey. Lincoln, RI: QualityMetric Incorporated, 2007. Hickey EJ, Veldtman G, Bradley TJ, Gengsakul A, Webb G, Williams WG, Manlhiot C, McCrindle BW. Functional health status in adult survivors of operative repair of tetralogy of Fallot. Am J Cardiol 2012;109:873e880. Kupari M, Turto H, Lommi J. Left ventricular hypertrophy in aortic valve stenosis: preventive or promotive of systolic dysfunction and heart failure? Eur Heart J 2005;26:1790e1796. Fuster RG, Argudo JA, Albarova OG, Sos FH, Lopez SC, Sorli MJ, Codoner MB, Minano JA. Left ventricular mass index in aortic valve surgery: a new index for early valve replacement? Eur J Cardiothorac Surg 2003;23:696e702. Mehta RH, Bruckman D, Das S, Tsai T, Russman P, Karavite D, Monaghan H, Sonnad S, Shea MJ, Eagle KA, Deeb GM. Implications of increased left ventricular mass index on in-hospital outcomes in patients undergoing aortic valve surgery. J Thorac Cardiovasc Surg 2001;122:919e928.
a Division of Pediatric Cardiology, Department of Pediatrics and Communicable Diseases, bSection of Pediatric Radiology, Department of Radiology and cDivision of Cardiothoracic Radiology, Department of Radiology, University of Michigan, Ann Arbor, Michigan. Manuscript received July 3, 2014; revised manuscript received and accepted September 17, 2014. See page 1900 for disclosure information. *Corresponding author: Tel: (734) 647-8924; fax: (734) 936-9470. E-mail address: [email protected] (J.C. Lu).
0002-9149/14/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2014.09.027
mass in patients with repaired TOF is related to functional health status and RV remodeling in this population, and whether measurement variability would improve with measurement in systole rather than diastole. Methods This study included patients with repaired TOF, age 14 years, who were referred for clinically indicated CMR from June 2008 to December 2009 and were prospectively enrolled as part of a previous cross-sectional study evaluating CMR predictors of functional health status.13 Patients were excluded if they had pulmonary atresia or absent pulmonary valve, additional significant cardiac malformations (e.g., atrioventricular septal defect), contraindications to CMR or gadolinium, CMR performed under anesthesia, or cognitive impairment preventing completion of the health status assessment. Additionally, patients were also excluded if the health status assessment was completed >3 months from time of CMR or if short-axis images were not available for analysis. CMR was performed using a commercially available 1.5 T scanner (Philips Intera Achieva, Best, The Netherlands). Cine images were obtained with a breathhold, electrocardiographic-gated, segmented k-space, steady-state free precession sequence. End-diastolic and end-systolic phases were manually chosen for the RV. RV endocardial and epicardial contours were manually drawn in both end-diastole and end-systole on short-axis images, using QMass software (Medis, Leiden, The Netherlands). www.ajconline.org
Congenital Heart Disease/RV Mass and Health Status in Tetralogy of Fallot Table 1 Clinical characteristics of the cohort (N¼53). Data are presented as number (%) for categorical variables and mean standard deviation or median (interquartile range) for continuous variables, as appropriate Male Age at CMR (years) Age at repair (years) Body surface area (m2) Type of TOF Pulmonary stenosis Unknown Type of TOF repair Transannular patch Valve-sparing Conduit Unknown Left ventricular ejection fraction (%) RV end-diastolic indexed volume (ml/m2) RV end-systolic indexed volume (ml/m2) RV ejection fraction (%) RV diastolic indexed mass (g/m2) RV systolic indexed mass (g/m2) RV diastolic mass:volume ratio (g/ml)
31 (58.5%) 29.4 (20.8-41.3) 3.5 (0.8-5.4) 1.88 0.27 42 (72.4%) 11 (19.0%) 31 (58.5%) 10 (18.9%) 4 (7.5%) 8 (15.1%) 53.6 7.5 148.7 49.5 83.9 36.4 44.8 9.0 27.7 7.7 24.9 7.4 0.19 0.04
CMR ¼ cardiovascular magnetic resonance; RV ¼ right ventricular; TOF ¼ tetralogy of Fallot.
Contours were compared with long-axis images to confirm the plane of the tricuspid valve annulus. RV mass was indexed to body surface area or to end-diastolic volume (mass:volume ratio). On enrollment, typically on the day of CMR, subjects completed the Short Form 36, version 2 (SF-36), which provides a measure of multiple domains of health and functioning.14 The Physical Component Summary score and the subscales of physical functioning (ability to perform physical activities), role-physical (participation in work or usual activities without physical limitations), and general health (perception of general health and its likelihood to change) were chosen to represent the physical impact of the disease. Scores were normalized for age based on population-based normative data. As previous reports demonstrated only small differences between patients with TOF and controls,13,15 an age-adjusted z-score 1 was considered a clinically significant decrease in functional health status. To evaluate the relation of RV mass to subsequent remodeling, a subset of patients from the initial cohort with an additional clinically indicated CMR study was evaluated. Measurements were taken from the most recent CMR, before any pulmonary valve or conduit replacement. A relative increase in RV end-diastolic volume of 10% from the initial measurement or an absolute decrease in RV ejection fraction of 5% was considered clinically significant. RV indexed mass and mass:volume ratio on the initial CMR study was compared with subsequent increase in RV volume or decrease in RV ejection fraction. For reproducibility of diastolic and systolic measurements, all CMR studies were analyzed by a single observer (JCL) and reanalyzed at least 1 month later for intraobserver variability. For interobserver variability, a subset of 29 patients was analyzed by a second observer (JTC), blinded to the initial measurements.
1897
Data are presented as frequency with percent for categorical variables and mean SD or median with interquartile range as appropriate for continuous variables. Receiver operating characteristic curves were used to determine the cutoffs of each RV mass measurement for predicting SF-36 scores, optimizing specificity. Sensitivity, specificity, and positive predictive value for the cutoffs were reported. Odds ratios with 95% confidence intervals were calculated for clinically significant decreased health status or changes in RV end-diastolic volume or ejection fraction. Changes in RV end-diastolic volume or ejection fraction from initial assessment to follow-up measurement were examined using the paired t test. Correlations between systolic and diastolic RV measurements, as well as between RV mass measurements and SF-36 scores in the patients with normal RV ejection fraction, were also evaluated by Pearson or Spearman correlation coefficient, as appropriate. RV ejection fraction was previously shown to be the best predictor of functional health status,13 and the subset of patients with normal RV ejection fraction was chosen to determine whether RV mass could be an adjunctive measure for further risk stratification. Values of p <0.05 were considered statistically significant. Intraclass correlation coefficients, coefficients of variation, and limits of agreement were used to evaluate interobserver and intraobserver variability on each RV measurement. Results In the initial cohort, 62 of 65 consecutive patients met inclusion criteria (1 declined participation, and an investigator was not available for consent for 2). Of these patients, 5 had pulmonary atresia or absent pulmonary valve and were excluded; 3 did not complete the SF-36 within 3 months of CMR. In 1 patient, short-axis images were not available for analysis. Patient and clinical characteristics of the remaining 53 patients are described in Table 1. SF-36 scores in this population were previously reported13 and were similar to the population norm, except for a small but statistically significant decrease in general health. RV diastolic indexed mass 37.3 g/m2 conveyed increased odds of decreased general health scores (Figure 1), although Physical Component Summary scores did not reach significance. Diastolic indexed mass 37.3 g/m2 had 95% specificity, 31% sensitivity, and 71% positive predictive value for decreased general health scores. RV systolic indexed mass did not reach significance for decreased general health scores. In the subset of patients with normal RV systolic function (ejection fraction 45%), indexed RV mass (either diastolic or systolic) correlated with Physical Component Summary scores and general health subscale scores (Table 2). RV enddiastolic volume did not correlate with functional health status. Only 1 patient (2%) in this cohort had RV diastolic mass:volume ratio >0.3. Neither RV diastolic mass:volume ratio 0.2 nor systolic mass:volume ratio 0.19 predicted decreased SF-36 scores. A subset of 24 patients had subsequent CMR before pulmonary valve or conduit replacement, a median 3.7 years (interquartile range 2.5 to 4.0) after initial study. There was
1898
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Odds ratios (and 95% confidence intervals) for decreased SF-36 scores (age-adjusted z-score 1), with elevated RV diastolic indexed mass (A) or RV systolic indexed mass (B). GH ¼ general health; RP ¼ role-physical; PCS ¼ Physical Component Summary; PF ¼ physical functioning.
Table 2 Correlation of RV diastolic or systolic indexed mass with Short Form 36 age-adjusted z-scores, in patients with normal RV systolic function (N¼28). Data are presented as r, Spearman correlation coefficient (p-value) Diastolic Physical Component Summary Physical Functioning Role Physical General Health
0.41 0.18 0.20 0.47
(0.03) (0.37) (0.31) (0.01)
Systolic 0.48 0.30 0.30 0.59
(0.01) (0.13) (0.12) (0.001)
no significant difference in pulmonary regurgitant fraction at follow-up (33.7 19.7% vs 32.0 21.0%, p ¼ 0.38), but indexed RV end-diastolic volume increased at the follow-up study (148 42 vs 137 43 ml/m2, p ¼ 0.02), with 9 patients having an increase >10% from the initial measurement. There was also a small but statistically significant decrease in RV ejection fraction (47 8% vs 45 9%, p ¼ 0.03), with only 1 patient having an absolute decrease of 5% in ejection fraction. RV diastolic mass:volume ratio >0.2 was associated with subsequent increase in RV end-diastolic volume (odds ratio 26.1, 95% confidence interval 2.9 to 765, p ¼ 0.002). RV indexed mass measurement in systole correlated very well with diastolic measurement (Figure 2), but mildly underestimated indexed mass relative to diastole (mean difference 2.8 g/m2, limits of agreement 0.9 to 6.5). RV systolic mass:diastolic volume ratio also correlated very well with diastolic measurement (Figure 2), with minimal underestimation relative to diastole (mean difference 0.02, limits of agreement 0.01 to 0.05). Intraobserver and interobserver reproducibility of all RV measurements was excellent (Table 3). Bland-Altman plots for limits of agreement are presented in Figures 3 and 4. Reproducibility of systolic mass was similar to minimally worse than reproducibility of diastolic mass measurements. Discussion In patients with repaired TOF, elevated RV mass relates to decreased functional health status and predicts progressive RV dilatation. RV mass should be routinely measured in this population, and can be measured accurately and reproducibly in either systole or diastole, although diastolic measurements are recommended, as they are slightly more reproducible, with a better relation to outcomes. The relation of RV indexed mass to some but not all SF-36 subscores underscores the multifactorial nature of
functional health status in this population. As RV mass is unlikely to be used in isolation in determining intervention thresholds, sensitivity was less essential, and specificity was prioritized in selecting a cutoff value, to minimize false positives. This led to a greater indexed mass cutoff, with few patients in this cohort at extremes of elevated mass. However, the correlation of indexed mass with functional health status in the context of normal RV ejection fraction suggests this may be a useful adjunctive measure for risk stratification and should be routinely evaluated in this population. The relation to functional health status is consistent with findings in aortic valve disease, in which left ventricular indexed mass predicts left ventricular systolic dysfunction16 and outcomes post aortic valve replacement.17,18 RV indexed mass likely is a marker of adverse remodeling, either due to residual RV outflow tract obstruction, volume load from dilatation, or systolic or diastolic dysfunction. This is also consistent with the International Multicenter TOF Registry (INDICATOR) cohort, which found RV mass:volume ratio 0.3 as a predictor for death or sustained ventricular tachycardia.4 In that study, 41% of patients with adverse outcomes had pulmonary atresia, a subgroup at risk for conduit stenosis or pulmonary hypertension, which was excluded in this study. This may explain why only 1 patient in this cohort had mass:volume ratio 0.3 and why we could not resolve mass:volume ratio as a significant predictor. However, the consistent association of RV mass to outcome, even when excluding patients with pulmonary atresia, underscores the potential applicability of RV mass measurement in this population. We do not believe that elevated indexed mass is simply a surrogate for dilatation, as we had previously shown no relation of indexed end-diastolic volume with functional health status.13 Similarly, RV volume did not relate to outcomes in the INDICATOR cohort. However, the trend toward increased progressive dilatation in patients with elevated mass suggests an interplay between wall thickness and volume. Increased RV mass appears to reflect not only the result of volume and/or pressure overload but may also portend further adverse remodeling. Due to conservation of mass, systolic and diastolic mass should be identical. We did not have pathologic specimens or a gold standard, and thus it is not clear which of these methods is truly more accurate. Given the thin RV wall in diastole, it is conceivable that the diastolic mass measurement is actually overestimated, as very small overestimation in wall thickness would increase the RV mass calculation, with small errors magnified over a dilated ventricle. Blalock
Congenital Heart Disease/RV Mass and Health Status in Tetralogy of Fallot
1899
Figure 2. Correlation of systolic versus diastolic measurement of RV indexed mass (A) and mass:volume ratio (B). Table 3 Intraobserver and interobserver variability of right ventricular measurements. RV EDV, g/m2
Intra-observer COV (%) ICC (95% CI) Inter-observer COV (%) ICC (95% CI)
RV ESV, g/m2
RVEF, %
RV indexed mass, g/m2
RV mass:volume ratio
diastolic
systolic
diastolic
systolic
2.4 1.00 (1.00-1.00)
3.8 1.00 (0.99-1.00)
7.2 0.93 (0.89-0.96)
8.0 0.96 (0.93-0.98)
9.9 0.94 (0.91-0.97)
9.7 0.87 (0.79-0.92)
10.0 0.91 (0.85-0.95)
12.1 0.96 (0.91-0.98)
11.4 0.97 (0.94-0.99)
9.5 0.81 (0.66-0.91)
11.1 0.93 (0.85-0.96)
13.0 0.88 (0.77-0.94)
15.9 0.78 (0.60-0.89)
18.0 0.75 (0.56-0.87)
CI ¼ confidence interval; COV ¼ coefficient of variation; EDV ¼ end-diastolic volume; EF ¼ ejection fraction; ESV ¼ end-systolic volume; ICC ¼ intraclass correlation coefficient; RV ¼ right ventricular.
Figure 3. Intraobserver limits of agreement for measurement of RV diastolic indexed mass (A), RV systolic indexed mass (B), RV diastolic mass:volume ratio (C), and RV systolic mass:diastolic volume ratio (D). Solid lines represent mean difference and dashed lines represent 1.96 SDs.
1900
The American Journal of Cardiology (www.ajconline.org)
Figure 4. Interobserver limits of agreement for measurement of RV diastolic indexed mass (A), RV systolic indexed mass (B), RV diastolic mass:volume ratio (C), and RV systolic mass:diastolic volume ratio (D). Solid lines represent mean difference and dashed lines represent 1.96 SDs.
et al suggested that systolic measurement may improve reproducibility, but did not reach statistical significance.12 The current study had greater intraclass correlation coefficients than in the previous report and thus would be less likely to detect a difference; the intraclass correlation coefficient for systolic mass was actually slightly lesser than that for diastolic mass. Although not seen in the current study, others have reported decreased reproducibility of systolic volumes,6,8 likely related to wall thickening and difficulty in distinguishing the blood-endocardial border, which may counteract any gain in tracing a thicker wall. With no clear advantage to measurement in systole, we recommend RV mass measurement in diastole. Several limitations should be acknowledged. The sample size was limited to the previous cohort that completed the SF-36 survey. We did not evaluate differences in RV measurements from axial or other planes, as our laboratory measures ventricular volumes from a short-axis plane, and full axial stacks through the ventricles were not available for analysis. A limited definition of functional health status was used, focusing on patients’ functional ability and perception of their health status. Although this is subjective, patientcentered metrics are an important consideration in clinical decision-making. Objective measurements such as exercise performance should also be considered, but few patients in this cohort had a contemporaneous exercise test. The overall population with TOF is heterogeneous, and this cohort with a later age of repair than current standards may differ from a more contemporary cohort. However, this cohort is similar
to current adult populations for whom intervention is being considered. The cross-sectional nature limits evaluation of whether potential markers predict changes in functional health status. We do not believe that the inclusion criterion of clinical referral for CMR was a significant source of sampling bias because CMR is a routine part of clinical care for this population at our center. Disclosures The authors have no conflicts of interest to disclose. 1. Therrien J, Siu SC, McLaughlin PR, Liu PP, Williams WG, Webb GD. Pulmonary valve replacement in adults late after repair of tetralogy of Fallot: are we operating too late? J Am Coll Cardiol 2000;36: 1670e1675. 2. Knauth AL, Gauvreau K, Powell AJ, Landzberg MJ, Walsh EP, Lock JE, del Nido PJ, Geva T. Ventricular size and function assessed by cardiac MRI predict major adverse clinical outcomes late after tetralogy of Fallot repair. Heart 2008;94:211e216. 3. Geva T. Repaired tetralogy of Fallot: the roles of cardiovascular magnetic resonance in evaluating pathophysiology and for pulmonary valve replacement decision support. J Cardiovasc Magn Reson 2011;13:9. 4. Valente AM, Gauvreau K, Assenza GE, Babu-Narayan SV, Schreier J, Gatzoulis MA, Groenink M, Inuzuka R, Kilner PJ, Koyak Z, Landzberg MJ, Mulder B, Powell AJ, Wald R, Geva T. Contemporary predictors of death and sustained ventricular tachycardia in patients with repaired tetralogy of Fallot enrolled in the INDICATOR cohort. Heart 2014;100:247e253. 5. Beygui F, Furber A, Delepine S, Helft G, Metzger JP, Geslin P, Le Jeune JJ. Routine breath-hold gradient echo MRI-derived right
Congenital Heart Disease/RV Mass and Health Status in Tetralogy of Fallot
6. 7. 8.
9.
10.
11.
12.
ventricular mass, volumes and function: Accuracy, reproducibility and coherence study. Int J Cardiovasc Imaging 2004;20:509e516. Mooij CF, de Wit CJ, Graham DA, Powell AJ, Geva T. Reproducibility of MRI measurements of right ventricular size and function in patients with normal and dilated ventricles. J Magn Reson Imaging 2008;28:67e73. Grothues F, Moon JC, Bellenger NG, Smith GS, Klein HU, Pennell DJ. Interstudy reproducibility of right ventricular volumes, function, and mass with cardiovascular magnetic resonance. Am Heart J 2004;147:218e223. Luijnenburg SE, Robbers-Visser D, Moelker A, Vliegen HW, Mulder BJ, Helbing WA. Intra-observer and interobserver variability of biventricular function, volumes and mass in patients with congenital heart disease measured by CMR imaging. Int J Cardiovasc Imaging 2010;26:57e64. Buechel EV, Kaiser T, Jackson C, Schmitz A, Kellenberger CJ. Normal right- and left ventricular volumes and myocardial mass in children measured by steady state free precession cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2009;11:19. Catalano O, Antonaci S, Opasich C, Moro G, Mussida M, Perotti M, Calsamiglia G, Frascaroli M, Baldi M, Cobelli F. Intra-observer and interobserver reproducibility of right ventricle volumes, function and mass by cardiac magnetic resonance. J Cardiovasc Med (Hagerstown) 2007;8:807e814. Bradlow WM, Hughes ML, Keenan NG, Bucciarelli-Ducci C, Assomull R, Gibbs JS, Mohiaddin RH. Measuring the heart in pulmonary arterial hypertension (PAH): Implications for trial study size. J Magn Reson Imaging 2010;31:117e124. Blalock SE, Banka P, Geva T, Powell AJ, Zhou J, Prakash A. Interstudy variability in cardiac magnetic resonance imaging measurements
13.
14. 15.
16. 17.
18.
1901
of ventricular volume, mass, and ejection fraction in repaired tetralogy of Fallot: a prospective observational study. J Magn Reson Imaging 2013;38:829e835. Lu JC, Cotts TB, Agarwal PP, Attili AK, Dorfman AL. Relation of right ventricular dilation, age of repair, and restrictive right ventricular physiology with patient-reported quality of life in adolescents and adults with repaired tetralogy of Fallot. Am J Cardiol 2010;106: 1798e1802. Ware JEJ, Kosinski M, Bjorner JB, Turner-Bowker DM, Gandek B, Maruish ME. User’s Manual for the Sf-36v2 Health Survey. Lincoln, RI: QualityMetric Incorporated, 2007. Hickey EJ, Veldtman G, Bradley TJ, Gengsakul A, Webb G, Williams WG, Manlhiot C, McCrindle BW. Functional health status in adult survivors of operative repair of tetralogy of Fallot. Am J Cardiol 2012;109:873e880. Kupari M, Turto H, Lommi J. Left ventricular hypertrophy in aortic valve stenosis: preventive or promotive of systolic dysfunction and heart failure? Eur Heart J 2005;26:1790e1796. Fuster RG, Argudo JA, Albarova OG, Sos FH, Lopez SC, Sorli MJ, Codoner MB, Minano JA. Left ventricular mass index in aortic valve surgery: a new index for early valve replacement? Eur J Cardiothorac Surg 2003;23:696e702. Mehta RH, Bruckman D, Das S, Tsai T, Russman P, Karavite D, Monaghan H, Sonnad S, Shea MJ, Eagle KA, Deeb GM. Implications of increased left ventricular mass index on in-hospital outcomes in patients undergoing aortic valve surgery. J Thorac Cardiovasc Surg 2001;122:919e928.