Cardiac-MRI Predicts Clinical Worsening and Mortality in Pulmonary Arterial Hypertension

Objectives This meta-analysis evaluates assessment of pulmonary arterial hypertension (PAH), with a focus on clinical worsening and mortality. Background Cardiac magnetic resonance (CMR) has prognostic value in the assessment of patients with PAH. However, there are limited data on the prediction of clinical worsening, an important composite endpoint used in PAH therapy trials. Methods The Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, and Web of Science databases were searched in May 2020. All CMR studies assessing clinical worsening and the prognosis of patients with PAH were included. Pooled hazard ratios of univariate regression analyses for CMR measurements, for prediction of clinical worsening and mortality, were calculated. Results Twenty-two studies with 1,938 participants were included in the meta-analysis. There were 18 clinical worsening events and 8 deaths per 100 patient-years. The pooled hazard ratios show that every 1% decrease in right ventricular (RV) ejection fraction is associated with a 4.9% increase in the risk of clinical worsening over 22 months of follow-up and a 2.1% increase in the risk of death over 54 months. For every 1 ml/m2 increase in RV end-systolic volume index or RV end-diastolic volume index, the risk of clinical worsening increases by 1.3% and 1%, respectively, and the risk of mortality increases by 0.9% and 0.6%. Every 1 ml/m2 decrease in left ventricular stroke volume index or left ventricular end-diastolic volume index increased the risk of death by 2.5% and 1.8%. Left ventricular parameters were not associated with clinical worsening. Conclusions This review confirms CMR as a powerful prognostic marker in PAH in a large cohort of patients. In addition to confirming previous observations that RV function and RV and left ventricular volumes predict mortality, RV function and volumes also predict clinical worsening. This study provides a strong rationale for considering CMR as a clinically relevant endpoint for trials of PAH therapies.


P ulmonary arterial hypertension (PAH)
is characterized by remodeling of the distal pulmonary arteries, leading to an increase in pulmonary vasculature resistance, reduced compliance, and elevated pulmonary artery pressure (1)(2)(3). Untreated, PAH has high morbidity and mortality that are closely linked to right ventricular (RV) dysfunction (2). A new diagnosis of PAH increases the risk of death at 1 year fivefold (4). However, over the last 20 years, treatment advancements have led to an increase in median survival from 3 to 7 years (4-6), although development of new therapies is needed.
Recently, clinical studies of PAH therapies have moved from assessing exercise capacity and pulmonary hemodynamics to using composite endpoints. One such approach uses the time to clinical worsening. Clinical worsening events include hospitalization, disease progression, and unsatisfactory longterm clinical response, in addition to mortality (7). However, given the large number of patients required and the expense of conducting such event-driven studies, cardiac magnetic resonance (CMR) has recently been explored as a primary endpoint to evaluate PAH therapies (8).
CMR is the gold standard method of measuring RV function, volumes, and mass, and it is an established prognostic and therapy response tool (1,9,10). In 2015, a meta-analysis assessed the prognostic value of CMR measurements in 5 studies with 332 participants (9); however, no data were reported on clinical worsening. Since then, multiple new studies assessing clinical worsening in addition to mortality have been published in PAH.
The current meta-analysis also includes unpublished supplemental data on CMR metrics from 16 previously published studies, which allowed us to provide new data on the utility of CMR to predict clinical worsening in addition to mortality. The goal of the current study therefore was to review the evidence for CMR metrics to predict clinical worsening and mortality in patients with PAH.

METHODS
The review was prospectively registered with The International Prospective Register of Systematic Reviews on December 12, 2019 (ID: CRD42019160296).
The Preferred Reporting Items for Systematic reviews and Meta-Analysis guidelines were followed (11).
Ethical approval was not required for this metaanalysis because it was based on published literature and did not recruit patients. CRITERIA     All studies were performed at PH referral centers and are therefore at risk for referral bias. The    Table 2.  There may, however, be sources of heterogeneity that could not be assessed in a meta-regression analysis where not enough data were available. There were differences in the types of clinical worsening events

Higher increases event Lower increases event Higher increases event Lower increases event Higher increases event Lower increases event Higher increases event
The meta-analyses of right ventricular (RV) and left ventricular (LV) function and mass showed that RVEF and RVMI are significant prognostic markers. RVEF could predict clinical worsening separate from mortality, whereas RVMI is a nonspecific prognostic marker. Unpublished data are indicated by (þ). CI ¼ confidence interval; other abbreviations as in Figure 1.

DISCUSSION
To the best of our knowledge, this paper is the largest meta-analysis of CMR imaging in patients with PAH and the first to report on clinical worsening in addition to mortality. We have confirmed that CMR imaging is a powerful prognostic marker in a large cohort of patients from multiple institutions, across several continents and using different imaging platforms. In addition, we have shown that CMR imaging predicts clinical worsening in patients with PAH. Our findings highlight the clinical utility of CMR imaging and support further evaluation of this modality as a clinically meaningful trial endpoint for the assessment of new therapies for PAH (Central Illustration, Table 3).
Clinical worsening as a composite endpoint has been shown to predict mortality (34) and has established itself as a primary efficacy endpoint in trials of PAH therapies (7,35). Although heart failure and all-cause mortality are included in all PAH trials using a composite clinical worsening endpoint, these trials vary in their inclusion and definition of progression markers, such as change in exercise tolerance or functional capacity (36), and they may use different thresholds to define a meaningful change (37). Nonetheless, study designs using time to clinical worsening have been increasingly adopted to evaluate PAH therapies. However, such studies require large numbers of participants and a prolonged period of follow-up, usually lasting for several years. As a consequence and given recent events such as the coronavirus disease-2019 pandemic, there has been a focus on considering clinical trial endpoints, which allow the impact of candidate therapies to be assessed over a shorter period and using an endpoint that correlates with clinically meaningful events.
In this meta-analysis, we have shown for the first time in a large cohort of patients that CMR-derived RV volumetric and functional metrics but not LV  in RV volumes is smaller than the 1.8% associated with a decrease in LV end-diastolic volume, the overall risk of mortality is more linked to RV volume, previously highlighted in large cohort studies (26,38).

CENTRAL ILLUSTRATION Cardiac Magnetic Resonance Imaging for Prediction of Clinical
Specifically, the increase in RV volume due to dilation in response to an increase in afterload is substantially larger than the change in LV volume, occurring as a consequence of ventricular interaction (39), particularly in advanced disease in PAH when uncoupling of the right ventricle and its load occurs (40).
This meta-analysis has shown that an increased RV mass has prognostic value but does not predict clinical worsening. In PAH, an increase in RV mass and RV hypertrophy is likely to represent an appropriate adaptive response to an increase in afterload (41).