Detection of Passive Liver Congestion in Pericarditis Patients With Constrictive Physiology Using MR Relaxometry of the Liver.


 Constrictive pericarditis is a common cause of right heart failure, characterized by increased filling pressures and passive liver congestion. As magnetic resonance (MR) T1 and T2 mapping allow quantification of liver relaxation times, we hypothesized liver mapping may allow depiction of passive liver congestion. We studied 45 pericarditis patients, of whom 25 presented constrictive physiology (CP+) and 20 normal physiology (CP-), and 30 control subjects. Pericarditis patients were predominantly male, but CP+ patients were on average 13 years older than CP- patients (p=0.003). Native T1 and T2 values as well as extracellular volume (ECV) values of the liver were significantly higher in CP+ than in CP- patients and controls, i.e. native T1: 765±102ms vs 581±56ms and 537±30ms; T2: 63±13ms vs 50±4ms and 46±4ms; ECV: 42±7% vs 31±3% and 30±3% control (all p<0.001). Using a cut-off right atrial (RA) pressure of >5 mm Hg to discriminate between normal and increased RA pressure, native T1 liver yielded the highest AUC (0.926) at ROC analysis with a sensitivity of 79.3% and specificity of 95.6%. Moreover, liver mapping showed excellent intra- and inter-reader agreement. In conclusion, liver mapping as part of a comprehensive cardiovascular MR exam is valuable to depict liver congestion in pericarditis patients.


Introduction
Congestive heart failure in patients with constrictive pericardial physiology is caused by increased right heart lling pressure leading to passive liver congestion induced by impaired hepatic venous out ow, a condition also called congestive hepatopathy [1,2]. Timely recognition of liver congestion is crucial as this condition may cause liver brosis and eventually cirrhosis. As the patients' symptoms and clinical ndings in congestive hepatopathy vary and may appear only at later stages, early detection of this condition is challenging on clinical grounds only. A combination of right atrial (RA) pressure measurements at transthoracic echography (TTE) or cardiac catheterization coupled with liver markers of cholestasis are currently considered the mainstay feature of congestive hepatopathy [2,4,5]. Abnormal liver ultrasound or contrast-enhanced computed tomography may be used to con rm this condition [3]. Novel techniques, such as assessment of the elastic properties of the liver using magnetic resonance (MR) elastography, are appealing not only to study patients with primary liver diseases but also to study liver stiffness in patients with heart failure [6][7][8]. However, MR elastography is not yet integrated in routine clinical practice given the complex acquisition, time-consuming post-processing and the requirement of an additional vibration device [6].
Cardiovascular MR imaging allows to comprehensively assess the heterogeneous spectrum of patients with pericarditis, ranging from acute pericarditis to end-stage constrictive pericarditis [9,10]. It is important to emphasize that constrictive physiology may occur in any condition where the pericardial compliance is decreased, thus not only in healed, brotic stages of pericarditis but also in acute conditions with active or ongoing pericardial in ammation. A typical, almost pathognomonic, hallmark of constrictive pericardial physiology is an increased ventricular coupling, a phenomenon which can be studied using free-breathing real-time cine CMR [11]. In recent years, quantitative, parametric imaging of the heart, using T1 and T2 MR relaxometry (the so-called mapping techniques), has become part of a standard CMR exam in ischemic and non-ischemic myocardial diseases [12]. Moreover, information with regard to the myocardial extracellular volume (ECV) can be obtained when T1 mapping is repeated post-contrast [12].
Clinical conditions with increased water content, such in patients with acute myocardial infarction, can be accurately depicted at T1 and T2 mapping [13].
In this monocentric, observational study, we hypothesized that in patients with pericardial disease MR mapping of the liver can complement cardiac imaging by unravelling co-existing congestive hepatopathy.
To test this hypothesis, we determined native and post-contrast T1 and T2 values of the liver parenchyma, obtained during a routine cardiovascular MR imaging study, and correlated these ndings to constrictive physiology pattern, RA pressures and other features of congestive hepatopathy.

Results
General patient data (Table 1) CP+ and CP-patients were predominantly male, but CP+ patients were on average 13 years older than CPpatients (p=0.003). The majority of patients (75%) in the CP+ group had predisposing factors for developing constricting, i.e., history of pericarditis (n=17), cardiac surgery (n=2), radiation therapy (n=3), systemic in ammatory disease (n=2). Eleven patients underwent pericardiectomy while the others received optimized anti-in ammatory treatment. In the CP-group, nine patients had a history of previous/recurrent pericarditis, two patients had cardiac surgery. All CP-patients received antiin ammatory therapy. Typical pericarditis-like chest pain was present in 9(36%) and 19(95%) of CP+ and CP-patients, respectively. CP+ patients had a higher heart rate than CP-patients (p<0.001). Dyspnea with NYHA class was present in 20 CP+ patients versus 3 CP-patients (p<0.001). Clinical examination showed raised central venous pressure in 13(52%) and lower limb edema in 9(36%) of CP+ patients, but in none of the CP-patients (both p<0.001). Atrial brillation was present in 6 CP+, 2 CP-patients and in none of the controls (p=0.004). Prior to CMR, pericardiocentesis was performed in 5 CP+ patients and 1 CPpatient.
Intra-and inter-observer agreement for native T1, T2 liver and ECV liver was excellent with all ICC values above 0.9, and a CV ranging from 3.1% to 8.3% and 2.2% to 10.0% for intra-observer and inter-observer agreement, respectively (table 3).

Discussion
To summarize, in the present study we showed that in constrictive pericarditis patients, assessment of relaxation values of the liver parenchyma on routinely acquired T1 and T2 maps of the heart may provide valuable and reproducible information with regard to the presence of passive liver congestion, re ecting increased right heart lling pressures caused by pericardial constriction. Findings at liver mapping correlated signi cantly with serum liver biomarkers, in particular GGT, which is deemed to re ect elevated RA pressures (4,5). As lling pressure pressures cannot be measured at cardiovascular MR imaging, information obtained by liver mapping may be valuable to indirectly depict right heart failure in pericarditis patients.
The diagnosis of constrictive pericarditis is typically two-fold, i.e. rstly, depiction of a pathologic noncompliant pericardium; secondly, demonstration of constrictive physiology. Presence of a rigid pericardium, impeding cardiac lling, causes increased lling pressures and symptoms of congestive heart failure [9][10][11]. Moreover, the ventricular (inter)dependence (or coupling) is increased and enhanced by respiration. The latter is a pathognomonic hallmark of constrictive pericarditis allowing differentiation with restrictive cardiomyopathy. In recent years, cardiovascular MR imaging has become an important modality in the diagnosis of constrictive pericarditis, as it allows not only an excellent assessment of the pathologic pericardium (e.g. thickness, presence of pericardial edema, pericardial enhancement, effusion) but also depicting of abnormal ventricular coupling [9][10][11]. However, assessment of congestive heart failure is more challenging as lling pressures cannot be measured at cardiovascular MR imaging. As the liver immediately feels the adverse effects of increased right heart pressures, assessment of this target organ provides valuable information on the presence of congestive heart failure. As shown in the present study, assessment of T1 and T2 relaxation times and ECV maps of the liver parenchyma using the T1, T2 and ECV maps of the myocardium provides valuable information on the presence of passive liver congestion, without prolonging scanning time or adding time-consuming post-processing. On the contrary, it enhances the clinical role of cardiovascular MR imaging as it allows a better appreciation of the patient's hemodynamic status.
As mentioned above, MR relaxometry has dramatically changed our understanding of cardiac diseases by allowing for quantitative and precise imaging of myocardial tissue abnormalities which enable to evaluate disease progression and response to therapy [12,13]. So far, the main focus of MR relaxometry has been myocardial texture analysis, though some preliminary reports suggest the utility to assess other organs. For instance, splenic relaxation times have been applied with success to assess of the response to adenosine during stress perfusion cardiovascular MR imaging [14]. In contrast, relaxometry of the liver in cardiac patients has received limited interest. Preliminary results suggested that mapping of the liver may be useful in patients with tetralogy of Fallot and in Fontan circulation [15,16]. Our ndings not only endorse those ndings but suggest a broader use of liver MR relaxometry offering valuable information in cardiac conditions which affect the liver with potential deleterious consequences, such as cardiac cirrhosis.
The liver-to-heart axis becomes important into two main contexts: 1) severely impaired cardiac output leading to peripheral organs hypoperfusion (e.g. cardiogenic shock); 2) increased central venous pressure due to inability of the right heart to accommodate the venous return [1,2,17]. In our study, none of the patients had impaired cardiac output, thereby liver hypoperfusion can be easily ruled out. Moreover, primary liver diseases (e.g., hepatitis) were careful excluded by thorough assessment of the clinical records. Thus, the increased values at MR relaxometry of the liver re ect an increased central venous pressure due to constrictive physiology prompted by pericarditis. Patients with pericarditis and increased ventricular coupling, deemed to re ect constrictive physiology, show not only signs of increased lling pressures at TTE (dilated IVC with diminished/absent respiratory motion, increased RA pressure) and cardiovascular MR imaging (dilated IVC, with increased IVC/aorta ratio, dilated hepatic veins, ascites), but show signi cantly longer T1 and T2 relaxation time of the liver coupled with higher ECV as compared to pericarditis patients with normal ventricular coupling or controls. Of interest, T1 value of the liver, and to a lesser extent ECV, correlated well with liver enzymes, in particular cholestatic enzymes such as GGT. These observations point toward the use of MR relaxometry as imaging biomarker to depict congestive hepatopathy.
Whereas MR relaxometry of the liver has hardly been used in cardiac patients, use of this modality is more common in patients with 'primary' liver pathology such as to liver brosis in the setting of chronic hepatitis and cirrhosis, and hemochromatosis [18,19]. Of note, the values in our healthy control group agree well with normal values in literature [20,21]. Moreover, MR relaxometry has been compared to MR elastography in patients with liver brosis/cirrhosis and shown a good correlation between these techniques [18,19]. Of interest, in an animal model of liver congestion, changes in RA pressure have an immediate effect on the liver [7]. In patients with constrictive pericarditis, Fenstad et al. showed increased liver stiffness, correlating with septal bounce, IVC size and abnormal septal motion [22]. These ndings suggest that liver relaxation times can be used as an alternative to MR elastography to appreciate hepatic venous congestion and/or brosis, which is appealing as the time-consuming MR elastography (i.e. extra measurement time about 20 minutes) may be too long to add in patients with right heart failure [6].
Limitations. Firstly, because of the retrospective nature, this paper is to a large extent descriptive. As congestive hepatopathy re ects a wide spectrum ranging from pure congestion to onset of liver brosis eventually leading to cardiac cirrhosis [1][2][3] the current study ndings do not provide information on the histologic correlate as no liver biopsy was performed, nor on the reversibility of ndings as most patients did not undergo a follow-up cardiovascular MR imaging. Secondly, in a small number of patients routinely cardiac mapping did not su ciently encompass the liver parenchyma (i.e. less than 10% in our study). In these patients additional imaging, e.g. coronal of transverse liver imaging, may be recommended. Thirdly, positioning of the T1 and T2 sequences for cardiac mapping and local shim volume may be suboptimal to study the liver parenchyma. To deal with this potential issue, the region of interest to measure liver relaxation times was selected as close as possible to the heart. Finally, only in constrictive pericarditis patients who underwent pericardiectomy right heart catheterization with pressure measurements was performed, while in the other subjects, the RA pressure was estimated by echocardiography which may be less accurate. descriptive.
In conclusion, liver mapping as part of a comprehensive cardiovascular MR imaging exam may provide valuable information with regard to the presence of congestive hepatopathy in constrictive pericarditis patients. The current study ndings open perspectives towards a more generalized use of liver MR relaxometry in cardiac patients, in particular in those at risk of developing right heart failure.

Study population.
A search in our cardiovascular MR imaging database (2014-2020) yielded 337 patients with MR diagnosis suggestive of pericarditis. Incomplete studies (n=51) or studies with insu cient coverage of the liver parenchyma of the T1/T2 maps (n=25) were excluded. Next, patients with concomitant liver pathology (e.g. liver cirrhosis) were excluded (n=12). This yielded 249 cardiovascular MR studies. In 25 studies (10%), increased ventricular coupling was found at free-breathing real-time cine MR imaging, de ned as inspiratory septal inversion, and/or increased excursion of the ventricular septum between inspiration and expiration [11]. These patients were de ned as pericarditis with constrictive physiology (CP+group). We randomly selected 20 patients in the cohort without increased ventricular coupling and de ned them as pericarditis without constrictive physiology (CP-group). Finally, we selected as control group 30 subjects with normal cardiovascular MR imaging ndings, lack of increased central venous pressure at TTE and normal liver enzymes (control group). The study was conducted in accordance with the Declaration of Helsinki and was approved by the research ethics committee of the hospital. Because of the retrospective design, informed consent by the patient was waived.

MR analysis.
Images were analyzed off-line using IntelliSpace Portal software (Philips Healthcare) by an European Association Cardiovascular Imaging (EACVI) level-III CMR reader. Pericardial thickness was measured, and the images were assessed for presence of effusion, edema (T2-weighted imaging) and in ammation (LGE imaging). Cine images were used to calculate LV and right ventricular (RV) volumes, function, and LV mass. Left atrial and right atrial area was contoured on the end-systolic horizontal long-axis image.
The diameter of the IVC and descending aorta were measured at the level of the diaphragm. Hepatic vein size was visually assessed as normal or dilated. Presence of pleural uid and ascites were evaluated. The early-diastolic inspiratory shape of the ventricular septum was described at real-time cine imaging as normal, attened or inverted, and the total shift of the septum between inspiration and expiration was measured [11]. As mentioned above, this information was used to de ne the CP+ and CP-group.
On the T1 and T2 map, myocardial T1 and T2 was measured in the ventricular septum, and a standardized method was used to calculate myocardial ECV. For the liver, a region of interest (ROI) was drawn freehand with a minimum number of 500 pixels. The ROI was typically chosen in the subcapsular liver parenchyma in the vicinity of the heart. Care was taken not to include the centrally located liver vessels. A similar formula as for the myocardium was used to calculate ECV liver (23), but the RV cavity instead of the LV cavity was used to measure the T1 of blood. Intra-and inter-reader reproducibility of T1/T2 liver mapping was assessed in 30 randomly chosen patients (10 from each group).

Statistical analysis
Statistical analysis was performed using R programming language for statistical computing v.4.0.0. (The R-Foundation for Statistical Computing). Histograms and Q-Q plots with graphical bootstrap were used to test the normality of the data distributions. Continuous variables were expressed as mean ± SD or medians with interquartile ranges (IQR) as appropriate. Categorical variables were expressed as frequency with percentage. One-way analysis of variance (ANOVA) or the Kruskal-Wallis test were used to assess differences among the different groups, and post hoc pairwise comparisons with Bonferroni adjustment were performed by using the Student t test or Wilcoxon Rank Sum test, respectively. The Chisquare test or Fisher's exact test was used to compare categorical variables. ROC curve analysis was performed to examine differences in performance of each variable for prediction of clinical symptoms of RHF (i.e., elevated central venous pressure, lower limb edema, NYHA class >2). The 95% con dence interval (CI) of each area under the curve (AUC) was estimated and compared with the R "pROC" package using the DeLong method [24]. The optimal cut-off points for each continuous predicator was de ned from the ROC analysis in terms of maximum sensitivity and speci city by using the Youden index [25]. The Youden index (J) is de ned as , with J=1 indicating complete separation and J=0 indicating complete overlap of distributions. The amount of overlap in values among different patient group was assessed using density plots and quanti ed using the R "overlapping" package with bootstrap estimates [26]. An adjusted 2-tailed p-value <0.05 was considered to indicate a statistically signi cant difference. Intrareader and inter-reader agreement of multiparametric liver mapping were assessed by using intraclass correlation coe cients (ICCs) and Bland-Altman analysis with 95% limits of agreement (LOAs). ICCs of greater than 0.75 and of 0.4-0.75 indicate strong and average agreement, respectively. A difference between ICCs was considered to be statistically signi cant when there was no overlap between their respective 95% CI limits.
Human subject research.