Imaging of the pulmonary vasculature in congenital heart disease without gadolinium contrast: Intraindividual comparison of a novel Compressed SENSE accelerated 3D modified REACT with 4D contrast-enhanced magnetic resonance angiography

Background Patients with Congenital heart disease (CHD) require repetitive imaging of the pulmonary vasculature throughout their life. In this study, we compared a novel Compressed SENSE accelerated (factor 9) electrocardiogram (ECG)- and respiratory-triggered 3D modified Relaxation-Enhanced Angiography without Contrast and Triggering (modified REACT-non-contrast-enhanced magnetic resonance angiography (modified REACT-non-CE-MRA)) with standard non-ECG-triggered time-resolved 4D CE-MRA for imaging of the pulmonary arteries and veins in patients with CHD. Methods This retrospective analysis of 25 patients (June 2018–April 2019) with known or suspected CHD was independently conducted by two radiologists executing measurements on modified REACT-non-CE-MRA and 4D CE-MRA on seven dedicated points (inner edge): Main pulmonary artery (MPA), right and left pulmonary artery, right superior and inferior pulmonary vein, left superior (LSPV) and inferior pulmonary vein. Image quality for arteries and veins was evaluated on a four-point scale in consensus. Results Twenty-three of the 25 included patients presented a CHD. There was a high interobserver agreement for both methods of imaging at the pulmonary arteries (ICC ≥ 0.96); at the pulmonary veins, modified REACT-non-CE-MRA showed a slightly higher agreement, pronounced at LSPV (ICC 0.946 vs. 0.895). Measurements in 4D CE-MRA showed higher diameter values compared to modified REACT-non-CE-MRA, at the pulmonary arteries reaching significant difference (e.g. MPA: mean 0.408 mm, p = 0.002). Modified REACT-non-CE-MRA (average acquisition time 07:01 ± 02:44 min) showed significant better image quality than 4D CE-MRA at the pulmonary arteries (3.84 vs. 3.32, p < 0.001) and veins (3.32 vs. 2.72, p = 0.015). Conclusions Compressed SENSE accelerated (factor 9) ECG- and respiratory-triggered 3D modified REACT-non-CE-MRA allows for reliable and fast imaging of the pulmonary arteries and veins with higher image quality and slightly higher interobserver agreement than 4D CE-MRA without contrast agent and associated disadvantages. Therefore, it represents a clinically suitable technique for patients requiring repetitive imaging of the pulmonary vasculature, e.g. patients with CHD.


(Continued from previous page)
Conclusions: Compressed SENSE accelerated (factor 9) ECG-and respiratory-triggered 3D modified REACT-non-CE-MRA allows for reliable and fast imaging of the pulmonary arteries and veins with higher image quality and slightly higher interobserver agreement than 4D CE-MRA without contrast agent and associated disadvantages. Therefore, it represents a clinically suitable technique for patients requiring repetitive imaging of the pulmonary vasculature, e.g. patients with CHD.

Background
With an incidence of 6-8/1000 at birth, congenital heart disease (CHD) comprises a wide range of different manifestations regarding the cardiovascular system potentially leading to death if left untreated. CHD has shown a serious improvement of survival over the past decades due to advancement in surgical techniques and early diagnosis, mostly owing to the widely use of fetal echocardiography [1][2][3]. Echocardiography represents the primary imaging modality of choice as it allows for fast, accurate and non-invasive imaging of cardiac function and vessel morphology [3][4][5]. However, it suffers from limitations such as user dependency and limited field of view (FOV) in growing patients [6,7].
Given the radiation dose as well as the use of iodinated contrast agent in computed tomography (CT)-angiography (CTA) and digital subtraction angiography (DSA) with the invasiveness of the latter, cardiovascular magnetic resonance (CMR) has been established as the non-invasive imaging of choice to evaluate the different vascular territories of the thorax in patients with CHD and has to be regarded as the gold standard [6,[8][9][10]. Besides 4D flow CMR, contrast-enhanced MR-angiography (CE-MRA) has proven to sufficiently detect vascular abnormalities and has shown technical progress over the past decades with the development of time-resolved 4D CE-MRA [8,[11][12][13]. However, the accurate application of CE-MRA and 4D CE-MRA is technically demanding and shows further limitations such as nephrogenic systemic fibrosis (NSF) [14] and long term retention of gadolinium.
Therefore, many non-CE-MRA techniques have been developed in the past, including sequences based on turbo spin echo (TSE), spoiled gradient echo sequences, steadystate free precession (SSFP), and balanced SSFP (bSSFP) with SSFP and bSSFP being the most widely used for the assessment of the pulmonary vasculature in patients with CHD as well as other diseases affecting the pulmonary vessels, e.g. pulmonary hypertension (PH) [15][16][17][18][19]. Recently, a novel 3D Relaxation-Enhanced Angiography without Contrast and Triggering (REACT) sequence, a combination of non-volume-selective short tau inversion recovery (STIR) pulse, a T2 preparation (T2 prep) pulse, and dual gradient echo Dixon (mDIXON XD), was introduced. It combines the benefits of SSFP with robust fat and background suppression [20].
The purpose of this study was to investigate the feasibility of a novel electrocardiogram (ECG)-and navigator-triggered 3D non-CE-MRA based on a modified REACT approach (modified REACT-non-CE-MRA) for the imaging of the pulmonary arteries and veins in patients with CHD and to compare measurement values and image quality to standard 4D CE-MRA.

Patient population
Patients were retrospectively selected from our internal database of 26 consecutive patients over a ten-month study period (June 2018-April 2019) receiving a dedicated clinical protocol regarding known or suspected CHD including both, 4D CE-MRA and modified REACT-non-CE-MRA. Insufficient contrast in 4D CE-MRA led to exclusion of patients. There were no exclusions regarding pathologies or operative treatment. Due to the retrospective design of the study, the local ethics committee waived written informed consent requirement in the patient cohort.

Image acquisition
All scans were performed on a clinical whole body 1.5 T CMR system (Philips Ingenia, Philips Healthcare, Best, The Netherlands) equipped with a dedicated 28-channel coil for cardiac imaging. The protocol comprised a non-CE-MRA using a modified REACT approach, a 4D CE-MRA, and 2D bSSFP breath hold cine imaging in standard orientations (4-chamber, 2-chamber, 3-chamber, short axis, transversal, left ventricular outflow tract, right ventricular outflow tract (RVOT)) as well as phase contrast velocity measurements of the main pulmonary artery (MPA) and the ascending aorta.
For 4D CE-MRA, a 3D spoiled gradient echo sequence was used. Gadobutrol (Gadovist, Bayer HealthCare Pharmaceuticals, Berlin, Germany; 0.1 ml/kg body weight) was injected at a flow-rate of 2 ml/second into an antecubital vein. Patients were asked to perform a breath hold during the acquisition. To allow for high spatiotemporal resolution, the acquisition was combined with parallel imaging using SENSitivity Encoding (SENSE) and a keyhole technique where 20% of the central kspace was acquired in each dynamic (4D Track, Philips Healthcare). The keyhole data was then combined with outer k-space data from a reference scan during image reconstruction.
For non-CE-MRA, imaging was based on a modified flow-independent REACT sequence. 3D magnetizationprepared mDIXON XD (Philips Healthcare) was combined with a 30 ms T2 prep sequence. Since it was verifiable that background suppression of mDIXON XD in combination with T2 prep was sufficient for cardiovascular applications, no STIR preparation was applied, contrary to the original REACT sequence as introduced by Yoneyama et al. [20]. To compensate for cardiac and respiratory motion, ECG-triggering (end-diastolic) and respiratory navigator-triggering were added to the originally proposed REACT sequence, therefore being referred to as "modified" REACT-non-CE-MRA throughout the manuscript. Data were acquired in the coronal plane. For acceleration of image acquisition, Compressed SENSE (Philips Healthcare), a combination of compressed sensing and parallel imaging using SENSE, was used [21,22]. For data acquisition, a variable density incoherent sampling pattern with highdensity sampling in the center and continuously increased undersampling towards the k-space periphery was employed. Image reconstruction was based on an iterative L1 norm minimization, assuring data consistency and image sparsity in the wavelet domain. Additionally, the reconstruction was regularized by coil sensitivity distribution and SENSE parallel imaging. Reconstruction was done online on the standard hardware as provided by the manufacturer of the CMR system. An acceleration factor of 9 was used, resulting in a nominal scan time of 02:11 min.
Detailed imaging parameters are given in Table 1.

Measurement
Anonymized images of 4D CE-MRA and modified REACT-non-CE-MRA were presented in random order to two radiologists (L.P., A.W.), each with at least two years of experience in cardiovascular imaging, who independently conducted the measurement on seven distinct measurement points: For each point, the measurement (inner diameter approach) was conducted on source images using the manual Multiplanar-Reconstruction-(MPR) tool in IMPAX EE (Agfa Healthcare N.V., Mortsel, Belgium) in manual perpendicular alignment. Maximum intensity projection images were not used since they lead to an apparent reduction in vessel diameter resulting in an underestimation [23]. Measurement points were excluded when they could not be assessed due to susceptibility or severe pulsation artifacts.

Image quality evaluation
Image quality was evaluated by both observers in consensus on a four-point scale regarding sharpness, presence of pulsation artifacts at the levels of measurement, and anatomic delineation. Quality was rated on a Likert scale of 1 to 4: 1 non-diagnostic, 2 poor image quality with substantial blurring impairing diagnostic confidence, 3 intermediate image quality with mild blurring, and 4 good image quality without any blurring and resulting high diagnostic confidence.

Statistical analysis
Data are shown as mean ± standard deviation (SD), unless noted otherwise. Statistical significance was set at p < 0.05. For each point of measurement, the average diameter of two tangential measurements was used for analysis. To evaluate interobserver reliability, intraclass correlation coefficients (ICCs) were calculated. Bland-Altman analysis was conducted to assess differences regarding measurement values of the pulmonary arteries

Study population and baseline characteristics
Of the 26 patients, one patient was excluded due to insufficient contrast in 4D CE-MRA, resulting in a study population of 25 patients (39 ± 20 years, body mass index 23.8 ± 5.2; 15 male subjects). Twenty-three patients presented a CHD, the most frequent being an atrial septal defect (n = 7), followed by ventricular septal defect (n = 5), transposition of the great arteries (n = 3), tetralogy of Fallot (TOF, n = 2) and pulmonary atresia (n = 2). 10 patients had surgery for CHD prior to the examination.

Imaging
All included imaging studies were executed without any complications. Modified REACT-non-CE-MRA showed an average total acquisition time of 7:01 ± 2: 44 min (depending on the patient's breathing frequency as well as heart rate), 4D CE-MRA of 2:14 ± 1:01 min.  Table 2.  Due to impaired image quality, measurement was not possible at the LSPV in one patient and at the LIPV in three patients using 4D CE-MRA. In modified REACT-non-CE-MRA, all measurements were conducted sufficiently. Comparison

Discussion
In our study, we retrospectively compared a novel Compressed SENSE accelerated navigator-and ECGtriggered 3D modified REACT-non-CE-MRA with standard non-ECG-triggered 4D CE-MRA for the imaging of the pulmonary vessels in patients with CHD. The major findings of this study are the following: 1. 4D CE-MRA showed greater diameters for the pulmonary vessels in comparison to modified REACTnon-CE-MRA, at the arteries with significant difference. 2. Modified REACT-non-CE-MRA offers a significantly higher image quality of the pulmonary vasculature and a slightly higher interobserver agreement at the pulmonary veins than 4D CE-MRA.
In line with previous works, which compared non-ECG-triggered CE-MRA with ECG-triggered non-CE-MRA (SSFP, end-diastolic) for the imaging of pulmonary arteries and veins, 4D CE-MRA showed higher measurement values of the pulmonary vessels compared to modified REACT-non-CE-MRA, at the arteries with a significant difference [16,24]. These differences are mainly due to pulsation and breathing artifacts in 4D CE-MRA and the resulting hampered vessel delineation. These pulsation artifacts are pronounced in patients with CHD such as TOF who show highly pulsatile circulations of the pulmonary arteries [23]. Furthermore, the motion and the Furthermore, the above-mentioned factors also implicate the slightly higher interobserver agreement of modified REACT-non-CE-MRA at the pulmonary veins. In line with above-mentioned studies, which have shown that other ECG-triggered and respiratorygated non-CE-MRA sequences such as 3D SSFP outperform untriggered CE-MRA in terms of image   quality, modified REACT-non-CE-MRA with respective triggering provided significant higher image quality scores of the pulmonary arteries and veins than 4D CE-MRA [15,16,18]. Patients with CHD require repetitive imaging of the pulmonary vasculature throughout their life and CMR has to be regarded as the gold standard of imaging with CE-MRA being used to evaluate the different vascular territories of the thorax and to detect its pathologies [6,[8][9][10]. However, the accurate execution of CE-MRA and 4D CE-MRA is technically demanding, as image quality depends on good coordination among the contrast injection, exact timing of data acquisition and patient cooperation for breath holding, subsequently limiting its use in incompliant patients [9,16]. Recently, 4D CE-MRA has been developed, which offers the possibility to acquire a series of volume angiograms in quick succession therefore simplifying the timing of acquisition in relation to the passage of the contrast bolus [9]. Nevertheless, the majority of time-resolved MRA uses some form of data sharing across time, making them sensitive to respiratory motion artifacts; therefore the majority is performed during a breath-hold. New techniques of free-breathing time-resolved MRAs are technically feasible, but show decreased signal-and contrast-tonoise ratios [9].
Furthermore, the use of gadolinium contrast poses a drawback potentially leading to NSF in end-stage renal disease, anaphylactic reactions or extravasation. Contrast agents show high costs and require an intravenous access, potentially limiting its use in the clinical routine [14,[27][28][29][30]. Given the growing literature on long-term gadolinium deposition within the brain and the uncertainty regarding its long-term effects, repetitive application of gadolinium contrast, especially in children, should be executed cautiously [31][32][33].
Consequently, many non-CE-MRA methods have been developed over the past decades with SSFP and bSSFP being routinely used for thoracoabdominal vessels. 3D bSSFP/SSFP show advantages such as high signal-tonoise ratios, high blood-to-tissue contrast due to its bright-blood signal and flow independence [34][35][36][37]. However, they are highly sensitive to off-resonance effects caused by B0 heterogeneities in the main magnetic field and disruptions of the steady state due to highly pulsatile flow or motion and show high background signals [15,38,39]. Therefore, image quality can be impaired by signal loss, banding artifacts and insufficient fat suppression. These effects are pronounced in higher magnetic fields such as 3 T and in large FOVs. When applied in large FOVs, a long acquisition time is required, consequently limiting its use in clinical routine [15,37].
Recently, a Compressed SENSE accelerated 3D REACTnon-CE-MRA was introduced, which overcomes these limitations by the following: On the one hand, mDIXON XD combines the above-mentioned benefits of SSFP with reduced sensitivity to inhomogeneities in the magnetic field [20,40]. mDIXON XD provides robust suppression of fat and background and allows for separation of water and fat, consequently leading to insensitivity of REACTnon-CE-MRA to inhomogeneities in the magnetic field, even in large FOVs [41]. Therefore, it provides highresolution scans in large FOVs and allows the application in higher magnetic fields such as 3 T, where inhomogeneities are expected to be higher. On the other hand, with the advent of new acceleration techniques such as compressed sensing, shorter  [22,42]. In this work, Compressed SENSE was employed, allowing for image-acquisition acceleration factors currently not achievable by compressed sensing or parallel imaging alone. Consequently, this directly addresses shortcomings of non-CE-MRA techniques as identified in previous work [15,36]. The Compressed SENSE technique was fully integrated on the clinical system, resulting in reconstruction times below one minute and an overall scan time of 7:01 ± 2:44 min, lower than 3D SSFP for the same kind of investigation and FOV (10:01 ± 4.5 min) [15]. REACT-non-CE-MRA was combined with respiratory navigator-triggering and ECG-triggering to compensate for respiratory and cardiac motion, therefore being referred to as "modified" REACT-non-CE-MRA in this study. Given its flow-independency, REACT-non-CE-MRA can in principle also be used without triggering, making it a versatile alternative to CE-MRA, especially for patients who are unable to perform a breath-hold, e.g. children.
Modified REACT-non-CE-MRA only delivers a stationary depiction of the vessels and does not include dynamic flow conditions as 4D CE-MRA [8]. Given the fact that modified REACT-non-CE-MRA enables a simultaneous display of arterial and venous vessels of the thorax unlike 4D CE-MRA, resulting images might appear overloaded. However, since the vessels are displayed sharply and in good quality, a sufficient differentiation is possible.

Clinical applications
Besides the depiction of the pulmonary vasculature in CHD, modified REACT-non-CE-MRA may offer an alternative to CTA or other MRA techniques for patients with PH or suspected pulmonary embolism as well as for imaging of the aorta in patients with connective tissue disease. Further, its use in other vascular territories, e.g. the extracranial arteries, might pose an additional clinical application and may warrant future investigation.

Limitations
The size and the heterogeneity of the study population with its wide range of different CHDs and postoperative alterations may be regarded as a drawback of this study.
We did not compare modified REACT-non-CE-MRA to DSA and regarded untriggered breath-hold 4D CE-MRA as the reference standard. Furthermore, no comparison to other ECG-triggered and respiratory-gated non-CE-MRA sequences such as 3D SSFP was conducted in this study, which may nurture future investigations.
Additionally, the lower resolution of 4D CE-MRA consequently leads to an inferior image quality than modified REACT-non-CE-MRA and offers lower image quality than standard CE-MRA [8]. The comparison of the ECGand navigator-triggered modified REACT-non-CE-MRA with an untriggered breath-hold first-pass CE-MRA (4D CE-MRA) may represent a limitation of this work given the fact that ECG-gating as well as navigator-gating with slow-infusion of contrast media have proven to increase the image quality of CE-MRA [43][44][45][46][47]. However, ECGgating with respect to contrast bolus arrival proves to be challenging in daily clinical routine.

Conclusions
Compressed SENSE accelerated (factor 9) ECG-and respiratory navigator-triggered 3D modified REACT-non-CE-MRA allows for robust and reliable imaging of the pulmonary vasculature in CHD with a higher image quality and a slightly higher interobserver agreement than 4D CE-MRA without the need of gadolinium contrast. Given its short acquisition time, it represents a clinically applicable alternative for patients with CHD in need of repetitive imaging of the pulmonary vessels.