Volumetric measurements in patients with corrected tetralogy of Fallot: comparison of short-axis versus axial cardiac MRI and echocardiography

Abstract

Background

Patients with corrected tetralogy of Fallot (cToF) are prone to develop pulmonary regurgitation and right ventricular enlargement resulting in long-term complications, thus correct right ventricular volumetric monitoring is crucial. However, it remains controversial which cardiovascular magnetic resonance imaging (CMRI) slice orientation is most appropriate in cToF for the analysis of the right ventricular volume.

Purpose

To investigate which slice orientation is most suited for right ventricular volumetry in cToF we compared short-axis and axial slices, and furthermore we compared right ventricular data between CMRI and echocardiography.

Material and Methods

Thirty CMRI examinations of 27 patients with cToF were included retrospectively. Right ventricular end-diastolic (EDV) and end-systolic volume (ESV) were derived from short-axis and axial cine CMRI planes. Furthermore, pulmonary trunk forward flow in phase-contrast CMRI and right ventricular inner diastolic diameter in echocardiography (R VIDdiast) were measured. By Bland-Altman and variance analysis intra- and inter-observer agreement were assessed for cine CMRI data. By Pearson correlation CMRI cine and phase-contrast data and CMRI cine and echocardiographic data were compared.

Results

Intra- and inter-observer variability for right ventricular EDV were significantly lower in axial slices (P = 0.016, P = 0.010). For right ventricular ESV a trend towards a lower intra- and inter-observer variability in axial slices was found (P = 0.063, P = 0.138). Right ventricular stroke volume in short-axis (r = 0.872, P < 0.001) and in axial (r = 0.914, P < 0.001) planes correlated highly, respectively very highly with pulmonary trunk forward flow in phase-contrast CMRI. R VIDdiast correlated highly with right ventricular EDV assessed by short-axis and axial CMRI (P < 0.001, P < 0.001).

Conclusion

Due to lower intra- and inter-observer variability, axial slices are recommended for right ventricular volumetry in cToF.

Keywords

Tetralogy of Fallot cardiac MRI echocardiography right ventricular volumetry

Introduction

Tetralogy of Fallot is the most common form of cyanotic congenital heart diseases. It represents about 5.5% of all congenital heart defects in live-born children (1). Due to surgical correction within the first year of life, nowadays the long-term outcome has considerably improved. Nevertheless, in the majority of Fallot patients anatomical and functional cardiac abnormalities persist postoperatively resulting in long-term complications. Hence, patients with corrected tetralogy of Fallot (cToF) are prone to develop severe pulmonary valve regurgitation, biventricular dysfunction, malignant arrhythmia, and sudden cardiac death (2 –5). In particular, frequent chronic pulmonary regurgitation leads to a right ventricular volume load. This right ventricular volume load may primarily be tolerated quite well, but over the years it causes irreversible cardiac damage (6 –9), which might be prevented by early pulmonary valve replacement (5,10 –12). However for decision-making, whether a pulmonary valve replacement is indicated in an asymptomatic cToF patient, Geva et al. have recently suggested different volumetric criteria such as right ventricular end-diastolic volume (EDV) exceeding 150 mL/m² of body surface area, right ventricular ejection fraction <47%, or left ventricular ejection fraction <55% (6). Therefore, in order to reveal relevant cardiac changes timely, cToF patients must be monitored lifelong for significant pulmonary valve regurgitation, right and left ventricular dysfunction, and in enlargement of the right ventricle in particular.

Currently cardiovascular magnetic resonance imaging (CMRI) is the accepted gold standard for right and left ventricular volumetry (13). In contrast to left ventricular volumetry with broadly accepted short-axis slices, no real consensus exists so far which slice orientation in cine CMRI is most suited for right ventricular volumetric measurements. Different plane orientations like short-axis (14), modified short-axis (15), and axial planes (13,16,17) have been suggested. Hence, the aim of this study was to compare short-axis and axial imaging for assessment of right ventricular volumes in cToF patients. To introduce a gold standard for the right ventricular stroke volume the pulmonary trunk forward flow was measured additionally in phase-contrast CMRI (18).

Because echocardiography remains the mainstay in the routinely annual imaging follow-up in cToF patients, we additionally compared right ventricular volumetric measurements between cine CMRI and echocardiography.

Material and Methods

Patients

Retrospective analysis and use of data were approved by the local ethic committee. All patients or their legal guardians had given informed consent for CMRI and echocardiography. Retrospectively 30 CMRI examinations (from October 2010 until November 2012) of 27 patients with cToF (17 male patients, 10 female patients) were included. The patients were referred to CMRI as part of their clinical follow-up, reasons of referral to CMRI were clinical deterioration that may be heart related, echocardiographically detected new or progressive right ventricular dilation, and preoperative planning (to confirm indication for cardiac surgery). Mean patient age was 14.6 ± 6.4 years (range, 11 months to 41 years), mean size 154.5 ± 21.6 cm (range, 66–188 cm), and mean weight 48.4 ± 20.9 kg (range, 6–101 kg). Three of our patients received two CMRIs and the remaining patients received one single CMRI resulting in a total number of 30 CMRIs. Time interval between routine echocardiography and CMRI was 49.7 ± 26.8 days (range, 0–92 days).

Cardiovascular magnetic resonance imaging (CMRI)

The CMRI examinations were performed on a 1.5-Tesla scanner (Magnetom Avanto, Siemens, Erlangen, Germany). CMRIs of children aged less than 9 years were carried out in intubation anesthesia. Image acquisition was undertaken in inspiration breath-hold. Balanced steady-state free precession cine sequences with TE of 1.2 ms, TR of 2.9 ms, flip angle of 72°, TA of 8.8 s, a matrix of 192 × 156, a field of view (FOV) in between 300 × 300 and 340 × 340, depending on the subjects size, 6 mm slice thickness, and 0% interslice gap were acquired in short-axis as well as in strict axial orientation including the pulmonary trunk (resulting in complete biventricular coverage in both slice orientations). Additionally flow measurements of the proximal pulmonary trunk were acquired using a velocity-encoded phase-contrast sequence with TE of 3.5 ms, TR of 47.4 ms, flip angle of 20°, a matrix of 192 × 156, a FOV in between 300 × 300 and 340 × 340, depending on the subjects size, 6 mm slice thickness, and a velocity encoding between 130 and 320 cm/s. All cine CMRI studies were re-evaluated twice by one reader (with 7 years of CMRI practice) and a third time by a second reader (with 2 years of CMRI practice), who was trained by the first reader appropriately. To ensure blinding to the first reading the interval between both readings in reader one was 6 months. All endocardial contours of the right ventricle were drawn manually including moderator band and trabeculae into the cavity (Fig. 1). Volume calculation was performed using the disc summation method (Argus software, Siemens Healthcare, Erlangen, Germany). The pulmonary trunk forward flow was measured by reader one by manual definition of the pulmonary trunk as region of interest (Argus Software, Siemens Healthcare, Erlangen, Germany).

Fig. 1.

A 15-year-old male patient with corrected tetralogy of Fallot. Balanced steady-state free precession cine sequence in short-axis (a) and axial (b) orientation with manually drawn regions of interest for right ventricular volumetry using the disc summation method. Echocardiographic M-mode, transthoracic long-axis view: the diastolic inner right ventricular diameter (R VIDdiast) is marked.

Echocardiography

As part of the annual clinical follow-up a transthoracic echocardiography was performed in all patients by an advanced pediatric cardiologist (>1000 echocardiographies). Investigations were performed on a Philips IE33 (Royal Philips Electronics N.V., Amsterdam, The Netherlands). In M-mode the right ventricular inner diastolic diameter (R VIDdiast) in the transthoracic long-axis view was measured in all patients (19,20) (Fig. 1).

Statistical analysis

For statistical analysis IBM SPSS Statistics 19 (version 19.0, IBM Corporation, Armonk, NY, USA) and MedCalc for Windows (version 12.3.0.0, MedCalc Software, Mariakierke, Belgium) were used. Normal distribution was proven in all investigated parameters by Kolmogorov-Smirnov test. Accordingly, mean values and standard deviations were calculated. Analysis for mean value differences of volumetric data between short-axis and axial orientation was done by paired t-test. For assessment of intra- and inter-observer agreement, Bland-Altman plots were performed and for variance analysis (short-axis versus axial method) f-test was utilized. Pearson correlation analysis between stroke volume (in short-axis and axial cine CMRI) and pulmonary trunk forward flow in phase-contrast imaging was carried out. Furthermore, Pearson correlation analysis between right ventricular EDV in cine CMRI and R VIDdiast in echocardiography was performed.

Results

Right ventricular volumetry in CMRI

The right ventricular stroke volume in short-axis cine CMRI (Table 1) correlated highly with the pulmonary trunk forward flow in phase-contrast CMRI (pulmonary trunk forward flow: 85.3 ± 28.2 mL [range, 14.1–161.8 mL], r = 0.872, P < 0.001, Fig. 2). The right ventricular stroke volume in axial planes (Table 1) correlated even very highly with the pulmonary trunk forward flow (r = 0.914, P < 0.001, Fig. 2). No significant mean value difference between short-axis and axial planes was found for right ventricular EDV, EDV corrected to body surface area, ESV, stroke volume, and ejection fraction (Table 1). Bland-Altman analysis for intra-observer agreement of right ventricular EDV showed a significantly lower variance using axial orientation compared to short-axis (P = 0.016, Table 2, Fig. 3). For intra-observer variability of right ventricular ESV there was a trend towards a lower variance in axial images, which was not significant (P = 0.063, Table 2). Bland-Altman analysis for inter-observer variability of right ventricular EDV revealed a significantly lower variance using axial orientation compared to short-axis as well (P = 0.010, Table 3, Fig. 3). For inter-observer variability of right ventricular ESV, again a trend towards a lower variance in axial images was found which was not significant (P = 0.138, Table 3).

Fig. 2.

Scatter diagrams with integrated correlation line of right ventricular EDV (in short-axis and axial cine MRI planes) and the pulmonary trunk forward volume in phase-contrast imaging demonstrating a high correlation for short-axis and a very high correlation for axial slices.

Fig. 3.

Bland-Altman plots for intra- and inter-observer agreement of right ventricular end-diastolic volume (EDV) comparing short-axis and axial planes in MRI.

Table 1.

Basic right ventricular volumetric data of the first reading (reader one) in short-axis and axial cine MRI: mean ± standard deviation (range).

	Short-axis	Axial	P value*
EDV (mL)	210.7 ± 75.5 (35.2–362.4)	213.6 ± 77.1 (50.1–429.6)	0.578
EDV/BSA (mL/m²)	143.6 ± 32.1 (68.3–203.0)	146.8 ± 31.8 (76.7–210.6)	0.377
ESV (mL)	119.7 ± 47.7 (18.1–235.2)	120.4 ± 51.1 (23.4–270.5)	0.884
SV (mL)	91.0 ± 33.4 (17.1–169.0)	93.3 ± 31.5 (25.6–160.0)	0.500
EF (%)	43.6 ± 6.8 (22.7–55.0)	44.9 ± 8.3 (24.7–71.2)	0.402

P values calculated using paired t-test.

BSA, body surface area; EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume; SV, stroke volume.

Table 2.

Bland-Altman analysis: Intra-observer agreement of right ventricular volumetric data comparing short-axis versus axial planes in cardiovascular MRI.

	Bias	Standard deviation	Limits of agreement	Variance (mL²)
EDV (mL)
Short-axis	8.1	21.6	[−34.3, 50.5]	468.0
Axial	11.5	15.9	[−19.8, 42.7]	253.4
ESV (mL)
Short-axis	7.7	24.2	[−39.6, 55.1]	584.9
Axial	14.7	17.9	[−20.3, 49.8]	320.7

EDV, end-diastolic volume; ESV, end-systolic volume.

Table 3.

Bland-Altman analysis: Inter-observer agreement of right ventricular volumetric data comparing short-axis versus axial planes in cardiovascular MRI.

	Bias	Standard deviation	Limits of agreement	Variance (mL²)
EDV (mL)
Short-axis	10.7	26.4	[−40.9, 62.4]	695.8
Axial	14.4	20.7	[−26.1, 55.0]	429.1
ESV (mL)
Short-axis	−2.2	29.2	[−59.5, 55.1]	854.7
Axial	−0.1	17.7	[−34.6, 34.5]	311.6

EDV, end-diastolic volume; ESV, end-systolic volume.

Right ventricular end-diastolic volume comparing CMRI and echocardiography

R VIDdiast in echocardiography (28.0 ± 8.1 mm; range, 17.0–51.6 mm) showed a high correlation with the right ventricular EDV in short-axis (r = 0.725, P < 0.001) and axial planes (r = 0.804, P < 0.001) with a trend towards a higher correlation in axial planes (Fig. 4).

Fig. 4.

Scatter diagrams with integrated correlation line of right ventricular inner diastolic diameter in echocardiography (R VIDdiast) versus right ventricular EDV in short-axis and axial planes in MRI demonstrating a high correlation in both cases with a trend for a higher correlation in axial slice.

Discussion

Short-axis and axial planes of 30 cine CMRI studies of patients with cToF were compared to evaluate which slice orientation is more suited for right ventricular volumetry in cToF. The pulmonary trunk forward flow, representing the gold standard of right ventricular stroke volume, correlated very highly with the right ventricular stroke volume derived from axial cine CMRI and highly with the right ventricular stroke volume derived from short-axis slices. Furthermore, intra- and inter-observer variability of right ventricular EDV was significantly lower in axial planes compared to short-axis. For right ventricular ESV, the intra- and inter-observer variability did not significantly differ, but showed a trend towards lower values in axial planes. Although the bias of right ventricular EDV and ESV was higher in axial planes compared to short-axis, the lower intra- and inter-observer variability stands for fewer outliers of right ventricular EDV and ESV derived from axial planes, resulting in robuster overall results. Therefore, we consider analysis of axial images as more reliable for volumetric assessment of the right ventricle in patients with cToF. The higher reproducibility of right ventricular volumes using axial images is presumably caused by a clearer demarcation of the right ventricular borders (level of the atrio-ventricular valve, right ventricular outflow tract) in axial in contrast to short-axis planes and seems to outbalance the more challenging delineation of the blood/myocardial boundary on the inferior right ventricular wall in axial slices (13).

Recently, James et al. compared volumetry of the right and left ventricle in short-axis and axial cine CMRI in 30 unselected patients. They did not find a significant impact of the slice orientation on biventricular stroke volume measurements (21) and concluded that in patients without complex cardiac diseases no additional acquisition of axial slices is necessary (21). However, considering the peculiar cardiac anatomy in patients with cToF, who present often with an enlarged right ventricle, we consider the current and James et al. results as not directly comparable. Atalay et al. investigated 20 patients with different cardiac diseases not focusing especially on congenital heart diseases. Comparing short-axis, axial, and horizontal long-axis slices they found the axial method to be the most reliable and accurate one for right ventricular volumetry (22). Clarke et al. investigated 50 patients with congenital heart disease by CMRI and did not find any significant difference in reproducibility of right ventricular volumes comparing short-axis and axial slice in their total evaluation (23). However, subgroup analysis of patients with right ventricular EDV ≥ 150 mL/m² revealed a higher agreement of pulmonary trunk flow measurements with right ventricular stroke volume derived from axial slices. Therefore, they recommended the usage of axial slices for right ventricular volumetry in this subgroup. In the current study the mean right ventricular EDV derived from axial slices was 146.8 ± 31.8 mL/m², therefore Clarke et al. would have recommended the axial method for right ventricular volumetry in our cohort.

Fratz et al. also focused on patients with cToF and compared short-axis versus axial planes for right ventricular volumetry (24). In the 46 patients they examined, a significantly lower intra-observer variability for right and left ventricular EDV in the axial orientation was reported as well. The intra-observer variability for right ventricular ESV in their study similarly showed a not significant trend towards a lower intra-observer variability using axial slices (24). In accordance with our results they recommended the use of axial slices for right ventricular volumetry in patients with cToF. The reproducibility of right ventricular volumes in the current study was not excellent. However, CMRI is the gold standard for left and right ventricular volumetry, because it has been proven to be more accurate than other methods like echocardiography (13,19). And although Rominger et al. demonstrated a better reproducibility of pulmonary trunk flow measurements compared to volumetric measurements based on cine CMRI, flow measurements do only provide information about stroke volumes but not about the EDV and ejection fraction which are important parameters in the follow up of cToF patients (6,18). As demonstrated, these parameters can actually be calculated most accurate from axial cine CMRI. And, although the reproducibility of this method is not perfect, pediatric cardiologists rely on this as the best method. In order to improve the reproducibility of right ventricular volumetric measurements in patients with cToF consented training of the CMRI readers is recommended, which was proven by Beerbaum et al. to improve the reproducibility (25). Furthermore, the reproducibility might be considerably improved by cross-referencing in the postprocessing of cine CMRI data, as it decreases inaccuracies caused by the often challenging delineation of the pulmonary valve in axial planes and the atrio-ventricular valve in short-axis planes.

Moreover, we found a high correlation between R VIDdiast assessed by echocardiography and right ventricular EDV in short-axis and axial slices cine CMRI, with a trend towards a higher correlation in axial planes. These results are in accordance with those of Helbing et al. who reported correlation coefficients from 0.57 to 0.86 between right ventricular volumetric measurements based on transthoracic two-dimensional echocardiography and axial orientated cine CMRI data (16). In a further study Ladouceur et al. investigated 94 patients, who had undergone right ventricular outflow tract repair, and likewise found a significant correlation between right ventricular EDV derived from axial CMRI and the right ventricular diameter in echocardiography (26). Because of the reported high correlation with right ventricular EDV we consider R VIDdiast in echocardiography as a sufficient screening parameter in the clinical routine. Since echocardiography is easily acquirable and less expensive than CMRI we suggest it as a useful close-meshed follow-up examination in cToF patients, aiming for selecting those who need a more accurate right ventricular volumetry performed by axial cine CMRI.

One limitation of this study is the patient selection which might introduce a bias and impair the generalization of the current results. Also the study cohort was relatively heterogeneous regarding age, size, and weight of the patients resulting in a high variation of right ventricular parameters.

In conclusion, due to a higher intra- and inter-observer agreement, we recommend the use of axial slices for right ventricular volumetry by cine CMRI in patients after repair of tetralogy of Fallot.

Footnotes

Conflict of interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Hoffman

. Incidence of congenital heart disease: I. postnatal incidence. Pediatr Cardiol 1995; 16: 103–113.

Gallego

Gonzalez

Sanchez-Recalde

. Incidence and predictors of sudden cardiac arrest in adults with congenital heart defects repaired before adult life. Am J Cardiol 2012; 110: 109–117.

Koyak

Harris

de Groot

. Sudden cardiac death in adult congenital heart disease. Circulation 2012; 126: 1944–1954.

Broberg

Aboulhosn

Mongeon

. Prevalence of left ventricular systolic dysfunction in adults with repaired tetralogy of fallot. Am J Cardiol 2011; 107: 1215–1220.

Gatzoulis

Balaji

Webber

. Risk factors for arrhythmia and sudden cardiac death late after repair of tetralogy of fallot: A multicentre study. Lancet 2000; 356: 975–981.

Geva

. 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: 9–9.

Geva

Sandweiss

Gauvreau

. Factors associated with impaired clinical status in long-term survivors of tetralogy of fallot repair evaluated by magnetic resonance imaging. J Am Coll Cardiol 2004; 43: 1068–1074.

Apitz

Webb

Redington

. Tetralogy of fallot. Lancet 2009; 374: 1462–1471.

Knauth

Gauvreau

Powell

. Ventricular size and function assessed by cardiac MRI predict major adverse clinical outcomes late after tetralogy of fallot repair. Heart 2008; 94: 211–216.

10.

Therrien

Siu

McLaughlin

. Pulmonary valve replacement in adults late after repair of tetralogy of fallot: Are we operating too late? J Am Coll Cardiol 2000; 36: 1670–1675.

11.

Therrien

Provost

Merchant

. Optimal timing for pulmonary valve replacement in adults after tetralogy of fallot repair. Am J Cardiol 2005; 95: 779–782.

12.

Theisen

Dalla Pozza

Malec

. MRI for therapy control in patients with tetralogy of fallot. Radiologe 2011; 51: 44–51.

13.

Helbing

de Roos

. Clinical applications of cardiac magnetic resonance imaging after repair of tetralogy of fallot. Pediatr Cardiol 2000; 21: 70–79.

14.

Mooij

de Wit

Graham

. Reproducibility of MRI measurements of right ventricular size and function in patients with normal and dilated ventricles. J Magn Reson Imaging 2008; 28: 67–73.

15.

Strugnell

Slaughter

Riley

. Modified RV short axis series–a new method for cardiac MRI measurement of right ventricular volumes. J Cardiovasc Magn Reson 2005; 7: 769–774.

16.

Helbing

Bosch

Maliepaard

. Comparison of echocardiographic methods with magnetic resonance imaging for assessment of right ventricular function in children. Am J Cardiol 1995; 76: 589–594.

17.

Alfakih

Plein

Bloomer

. Comparison of right ventricular volume measurements between axial and short axis orientation using steady-state free precession magnetic resonance imaging. J Magn Reson Imaging 2003; 18: 25–32.

18.

Rominger

Dinkel

Bachmann

. Comparison between fast MR flow quantification in breathhold technique in ascending aorta and pulmonary trunc with right and left ventricular cine-MRI for the assessment of stroke volumes in healthy volunteers. Rofo 2002; 174: 196–201.

19.

Lai

Gauvreau

Rivera

. Accuracy of guideline recommendations for two-dimensional quantification of the right ventricle by echocardiography. Int J Cardiovasc Imaging 2008; 24: 691–698.

20.

Lang

Bierig

Devereux

. Recommendations for chamber quantification: A report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005; 18: 1440–1463.

21.

James

Wald

Wintersperger

. Accuracy of right and left ventricular functional assessment by short-axis vs axial cine steady-state free-precession magnetic resonance imaging: Intrapatient correlation with main pulmonary artery and ascending aorta phase-contrast flow measurements. Can Assoc Radiol J 2013; 64: 213–219.

22.

Atalay

Chang

Grand

. The transaxial orientation is superior to both the short axis and horizontal long axis orientations for determining right ventricular volume and ejection fraction using simpson’s method with cardiac magnetic resonance. ISRN Cardiol 2013; 2013: 268697–268697.

23.

Clarke

Gurka

Norton

. Assessment of the accuracy and reproducibility of RV volume measurements by CMR in congenital heart disease. JACC Cardiovasc Imaging 2012; 5: 28–37.

24.

Fratz

Schuhbaeck

Buchner

. Comparison of accuracy of axial slices versus short-axis slices for measuring ventricular volumes by cardiac magnetic resonance in patients with corrected tetralogy of fallot. Am J Cardiol 2009; 103: 1764–1769.

25.

Beerbaum

Barth

Kropf

. Cardiac function by MRI in congenital heart disease: Impact of consensus training on interinstitutional variance. J Magn Reson Imaging 2009; 30: 956–966.

26.

Ladouceur

Gillaizeau

Redheuil

. Optimal follow-up in adult patients with congenital heart disease and chronic pulmonary regurgitation: Towards tailored use of cardiac magnetic resonance imaging. Arch Cardiovasc Dis 2013; 106: 27–35.