Cardiovascular magnetic resonance imaging-guided right heart catheterizations for cardiac pressure–volume loop analyses

Cardiovascular magnetic resonance (CMR) imaging provides radiation-free, three-dimensional visualization of the cardiovascular anatomy with the ability to quantify haemodynamic features and perform tissue characterization, making it a promising tool for guiding transcatheter interventions. Right heart catheterization (RHC) is the gold standard for measuring cardiac and pulmonary pressures.1 However, conventional RHC guided by X-ray fluoroscopy is limited by poor visualization of tissues and anatomy. RHC guided by CMR (RHC-CMR) was pioneered 20 years ago and is currently routinely used in several centres worldwide.1 RHC-CMR has the potential to enable synchronous pressure–volume (PV) loop analysis,2 which is considered the reference method to study the load-independent function of the cardiac chambers and ventricular–vascular interactions (i.e. ventriculo-arterial coupling). Therefore, PV analyses can elucidate pathophysiological mechanisms in cardiac diseases and may be used for monitoring therapeutic effects.3 Notably, RHC-CMR offers the opportunity to relate PV analyses to quantitative blood flow and tissue characterization related to myocardial stiffness and contractility, including assessment of oedema, inflammation, and fibrosis. In this article, we provide a comprehensive overview on how to perform cardiac PV loop analyses during RHC-CMR. RHC-CMR is typically performed in a 1.5 T interventional CMR suite. It requires a dedicated, multidisciplinary team. Mitral regurgitation (MR) Safe 7F single-lumen, air-balloon tipped, Swan-Ganz catheters and MR safe monitoring systems are required to obtain and visualize pressure curves. The suite set-up and patient preparation have been described before.4 After conventional CMR localizers, the catheter is inserted via the femoral or jugular vein and guided by real-time balanced steady-state free precession (bSSFP) imaging The catheter is visualized by the air-filled balloon at the tip, causing a signal void (Figure 1). Subsequently, reference views of the right atrium (RA), right ventricle (RV), and pulmonary artery (PA) are acquired4 to navigate the catheter through the vena cava, RA, RV, and ultimately positioned in the PA, reaching the wedge position. Simultaneously, a stack of bSSFP cine images covering the atria and ventricles are acquired to measure cardiac volumes and function. Alternatively, biplane cine images of the atria and ventricles can be used (i.e. volumes and function are derived from the two-chamber cine images of the left and right side of the heart and four-chamber cine images). Measurements have to be obtained during end-expiration to limit the effects of intra-thoracic pressures. CMR-RHC for PV loop analyses. (A) The catheter placed in the pulmonary artery wedge position. (B) PV curves of the LA over the cardiac cycle. The a-wave is produced during atrial contraction. The c-wave represents bulging of the mitral valve into the LA during early-systolic ventricular contraction. The x-descent represents atrial relaxation (mid-systolic dip). The v-wave represents passive LA filling before the opening of the mitral valve (directly after ECG T-wave). The y-descent represents the rapid fall in pressure during ventricular filling after opening of the mitral valve. (C) The smaller LA PV loop of a healthy control in the left and of PH patient on the right. The slope of the dotted line represents LA passive stiffness. (D) The catheter in the RA. (E) RA PV-curves. a-wave: during atrial contraction. c-wave: during bulging of the tricuspid valve into the RA during early systole (directly after the ECG QRS-complex). x-descent represents atrial relaxation. v-wave represents passive RA filling. y-descent represents the rapid fall in pressure after tricuspid valve opening. (F) RA PV loops. The slope of the orange line represents RA passive stiffness. (G) The catheter in the RV. (H) RV PV curves. (I) RV PV loops. Effective arterial elastance (Ea) represents RV afterload. CMR, cardiovascular magnetic resonance; EDPVR, end-diastolic pressure–volume relationship; EDV, end-diastolic volume; ESPVR, end-systolic pressure–volume relationship; ESV, end-systolic volume; P-iso, maximum isovolumic pressure; V-max, maximal volume; V-min, minimal volume. Finally, a stack of sagittal bSSFP images is obtained to visualize the vertical distance between the catheter insertion site and the cardiac chambers, which is required for catheter nulling (see below). To ensure accurate pressure curves, several precautions must be taken. It is essential to adequately null the system, prevent air bubbles, and minimize catheter whipping. Conventional catheter system nulling is based on a reference point chosen at the mid-thoracic line, but its relationship with the actual position of the heart may vary. Alternatively, a reference point may be chosen at the level of the catheter sheath and related to the position of the catheter in the cardiac chambers, both visualized by sagittal CMR images. The nulling method is important, since a reference point chosen 1.0 cm too high may result in pressure underestimation by 0.74 mmHg. The RHC-CMR environment poses several challenges in order to acquire accurate electrocardiogram (ECG) and pressure curves: (i) resonance frequency of the system, (ii) damping, (iii) signal noise, and (iv) a phase shift of the pressure curves. It is important to optimize the system (for example by using non-elongated catheters and conductor loops, a damping system, preventing air bubbles, and performing optimal flushing) and to ensure that the resonance frequency of the system remains outside physiological ranges (i.e. above ∼20 Hz) to limit Challenges 1 and 2. Determining the resonance frequency and damping of the pressure curves demands performing pressure step tests that involve the evaluation of pressure variations during flushing. Storage and low-pass filtering of the simultaneous ECG and pressure curves are required to limit signal noise. The pressure signals and ECG signals are obtained with a sample rate of 1000 Hz. Volumes are measured over the entire cardiac cycle and typically result in 25 volume points. In order to reduce Challenge 4, it is important to average the signals over 10 heart beats and two respiratory cycles. Pressure and volume signals have to be smoothed and aligned by fitting a polynomial through the original data points.5 This is performed to derive an equal number of samples for the pressure and volume signals to obtain a PV loop. Despite the synchronous PV measurements, small differences in heart rate may have occurred during the measurements. Therefore, the pressure and volume signals have to be aligned before the creation of the PV loops. For the atria, this can be done based on the v-wave (Figure 1) that occurs at the end of ventricular contraction. For the RV, this can be done on the maximum PV point.5 Using PV analyses, end-systolic elastance (Ees) and diastolic stiffness (ß) of the cardiac chambers, which are considered load-independent parameters for systolic contractility and diastolic function, respectively, are measured.3 Figure 1 illustrates the PV curves and PV loops of the left atrium (LA), RA, and RV of a healthy control (black lines) and of a patient with severe primary MR and secondary pulmonary hypertension (PH) (purple lines). The PA wedge pressure (PAWP) provides an accurate reflection of LA pressures. To generate a LA PV loop, the PAWP is plotted against the LA volume at multiple time points throughout a cardiac cycle (Figure 1A–C). Using ‘single-beat’ PV analysis, LA passive stiffness is calculated as the difference between the change in pressures and volumes during the atrial filling phase (reservoir phase). A line is fitted between the x-descent (lowest pressure during the reservoir phase) and the v-wave (end of reservoir phase corresponds to maximal LA pressure). The slope of this line represents LA passive stiffness (Figure 1C).6 Currently, there are no validated ‘single-beat’ methods for the measurement of LA contractility (Ees). Instead, a ‘multi-beat’ method has to be used.7,8 The ‘multi-beat’ method requires multiple PV loops of the atrium during pre-load manipulation. The slope of the line that connects the atrial end-systolic (ES) PV points (minimum volume after ECG P-wave) at various loading conditions represents Ees. In addition, LA ß is represented by the slope of the line that connects the end-diastolic (ED) PV points.8 In the early stages of chronic heart disease, LA contractility is augmented while LA stiffness is increased (Figure 1). However, during progressive disease, LA contractility decreases.7 RA passive stiffness is calculated as the difference between the change in pressures and volumes during the atrial filling phase (Figure 1D–F). Recently, the ‘single-beat’ method was validated against the slope of the PV loop at maximal volume (end of reservoir phase), which corresponds to RA ß (Figure 1F).5 A ‘multi-beat’ method can also be applied to the RA to determine RA Ees and RA ß. This requires a similar method as described for the LA.8 The behaviour of the RA, however, differs from that of the LA. While RA passive stiffness remains unchanged during chronic right-sided pressure overload, RA contractility is increased5 (Figure 1). Figure 1G and H depicts RV PV curves. A ‘single-beat’ PV method has been validated for the RV.9 This method estimates the ES PV relationship (ESPVR) based on the prediction of the maximal pressure (Pmax) that would have been achieved if ventricular contraction had remained isovolumic. The slope of the line between Pmax and ES PV point represents RV Ees. Measuring RV diastolic stiffness based on the non-linear ED PV relationship (EDPVR) requires an exponential fit between the RV begin diastolic and ED PV points. RV afterload is estimated using effective arterial elastance (Ea), which represents the slope of the line intersecting the ES and ED PV points (Figure 1I). RV Ees/Ea ratio has become widely used as a means to characterize RV-PA coupling. In the presented PH patient, we found that RV Ees showed a 3.5-fold increase, which is sufficient to compensate for the increased afterload (Ea), leading to preserved RV-to-PA coupling (optimal Ees/Ea ratio = 1.5–2.0) (Figure 1). Similarly to the atria, the load-independent properties of the RV can be determined using a ‘multi-beat’ method, examining the slope of the ESPVR and EDPVR to measure RV Ees and RV ß, respectively, during pre-load manipulation.3 According to the ‘multi-beat’ method, PV loops have to be obtained during different loading conditions. Pre-load can be manipulated through an intravenous saline fluid challenge or inferior vena cava occlusions. Alternatively, in our RHC-CMR protocol, we perform a passive leg-raise (PLR) manoeuvre to increase the cardiac pre-load. The PLR involves raising a patient’s legs to 45° resulting in an increase in cardiac venous return of ∼150–300 mL. After PLR, the average time to maximum PAWP is 20 s followed by a short period of steady state before PAWP decreases.10 This limited time frame makes it challenging to measure synchronous volumes, leading to suggested alternative methods of biplane volume measurements, chamber areas, or real-time imaging at rest and during PLR.11 It is not recommended to utilize Valsalva manoeuvres, since Phase IIB of the manoeuvre augments ventricular contractility. RHC-CMR can be safely performed in patients with adverse cardiopulmonary haemodynamics. RHC-CMR procedure times are faster than the total duration of the component investigation performed individually.4 A common indication for RHC-CMR is (suspected) PH.1 RHC-CMR may be of advantage to discriminate between Group 1 pulmonary arterial hypertension (PAH) and PH associated with left heart disease, based on PAWP measurements during rest and PLR10 and simultaneous CMR measurements of LA and left ventricular structure and function.12 Most frequently, RHC-CMR has been performed in congenital heart disease.1 Other indications are heart transplant screening and surveillance.1 RHC-CMR is used to understand the pathophysiology and outcomes in heart failure and valvular heart diseases. In addition, atrial characterization and studying atrioventricular interactions may provide insights in arrhythmias. Advanced insight into RV dysfunction is of particular importance since it is a strong predictor of prognosis in cardiac diseases. RHC-CMR offers a unique opportunity to obtain the acquisition of synchronous cardiac PV loop measurements. This provides valuable insights into intrinsic cardiac function and ventriculo-arterial coupling that surpass the information gained from individual and sequential cardiac PV measurements. Synchronous RHC-CMR demonstrates that the collective analysis of these measurements yields a more comprehensive understanding of cardiac physiology—truly exemplifying the notion that ‘the whole is greater than the sum of its parts’. Only a few data were generated in support of this research. The data underlying this article cannot be shared publicly due to the privacy of individuals who participated in the study. The data will be shared on reasonable request to the corresponding author.