BackgroundMachine learning tools to facilitate literature screening have existed for over a decade, yet adoption remains limited. Available approaches range from active learning tools that require hundreds of interactive screening decisions before they can prioritise citations effectively, to commercial large language model (LLM) chatbots repurposed for screening without task-specific training.ObjectiveWe aimed to develop TITAN-SR, a screening tool built on biomedical BERT models, and to compare its performance against ASReview, a widely used active-learning tool for systematic-review screening; four commercial LLM chatbots evaluated on a 500-review benchmark (Claude Sonnet 4, GPT-4o, DeepSeek-V3, Gemini 2.0 Flash); and GPT-4o-mini evaluated separately on all 3,434 evaluable internal test reviews.MethodsWe assembled 762,934 citation–label pairs from 19,787 completed reviews and trained TITAN-SR on a prespecified review-level training split, with validation and testing on held-out reviews. The tool uses an ensemble of two biomedical BERT models (PubMedBERT and BioLinkBERT) that read a review’s eligibility criteria alongside each citation’s title and abstract and outputs a raw inclusion score. Training penalised missing a relevant study 10 times more heavily than retaining an irrelevant one. We compared TITAN-SR with four commercial LLM chatbots applied without task-specific training on a stratified 500-review benchmark, and additionally evaluated GPT-4o-mini on the 3,434 evaluable internal test internal test set. External validation and the ASReview comparison used a temporally independent holdout of 22 Cochrane reviews comprising 142,504 records, with no review-level overlap with the Cochrane reviews used during model development. To ensure a fair comparison with an active-learning tool that improves with reviewer input, ASReview received a warm-up period (20% of known includes plus proportional excludes).ResultsAcross 3,434 test reviews, TITAN-SR achieved a median specificity at 99% recall (S@99R) of 0.902 and median AUC of 0.974; at the threshold required to achieve 99% recall within each review, TITAN-SR excluded a median of 90.2% of irrelevant citations. TITAN-SR outperformed all four chatbots evaluated on the 500-review benchmark (Bonferroni-corrected paired Wilcoxon signed-rank p ≤ 5 × 10−31 for every comparison) and also outperformed GPT-4o-mini on the full 3,434-review test set. The chatbots exhibited calibration collapse: 82–99% of their self-reported confidence scores were ≥ 0.9 regardless of accuracy. Temporal validation on 22 held-out Cochrane reviews showed performance virtually identical to internal results (median AUC 0.976, S@95R 0.897). In head-to-head comparison on those 22 reviews, TITAN-SR outperformed ASReview, winning 18 of 22 (81.8%; median paired S@95R difference +0.145, bootstrap 95% CI +0.101 to +0.201).ConclusionsA purpose-built screening tool based on biomedical BERT models, developed using a corpus of nearly 20,000 completed reviews, outperformed ASReview, a widely used active-learning tool, and four commercial LLM chatbots evaluated without task-specific training (plus GPT-4o-mini on all 3,434 evaluable internal test reviews). Temporal validation on 22 held-out Cochrane reviews confirmed that performance generalised to new Cochrane reviews published after the model-development period.

Abstract Background Right ventricular (RV) function predicts survival in pulmonary hypertension (PH) and other cardiovascular diseases, yet echocardiographic AI has largely focused on the left ventricle (LV). Objectives To develop and evaluate PH-ECHO-AI, a unified deep learning model performing four-chamber segmentation, landmark localisation, biventricular ejection fraction (EF) estimation, deformation analysis, and PH prediction from a single apical four-chamber (A4C) clip. Methods We developed the model using 8,416 clips from four public datasets and no institutional data: EchoNet-Dynamic, CAMUS, RVENet (apical four-chamber clips paired with 3D-echocardiographic right ventricular ejection fraction, RVEF), and MIMIC-IV-ECHO. Evaluation used held-out, training-excluded data with expert-reviewed reference standards and a per-cohort audit of patient-level separation: 1,416 clips for segmentation; 600 clips for function and deformation (350 referenced to 3D-echocardiographic RVEF, 250 to the EchoNet LVEF); and 1,076 MIMIC-IV patients for PH prediction, with five-fold cross-validation. Performance measures were Dice, correlation, mean absolute error (MAE), Bland-Altman agreement, and area under the receiver operating characteristic curve (AUC). Results Four-chamber segmentation generalised robustly across all datasets (pooled Dice: LV 0.925, RV 0.836, LA 0.910, RA 0.904). Left ventricular ejection fraction (LVEF) was estimated with r=0.845 (95% CI 0.786 to 0.886) and MAE 4.67%. RVEF, regressed directly from the clip by a supervised head trained on 3D-echocardiographic labels with no geometric assumption, reached r=0.754 (95% CI 0.690 to 0.806) and MAE 4.98%, matching published single-view RVEF ceilings and exceeding geometric RV fractional area change (RVFAC; r=0.278). Deformation and excursion metrics, namely RV free-wall and LV A4C longitudinal strain and tricuspid and mitral annular plane systolic excursion (TAPSE, MAPSE), proved physiologically coherent. Segmentation generalised to the external MIMIC-IV cohort, and PH prediction was developed and evaluated entirely within it; RVEF evaluation was clip-disjoint and same-source, so cross-centre RVEF validation remains outstanding. Using echocardiographic geometry alone, confirmed PH was detected with an AUC of 0.697 and strong calibration (Brier 0.061). Conclusions A single, reproducible model provides comprehensive right-heart-focused interpretation from one A4C view. It achieves RVEF accuracy competitive with dedicated RV models while simultaneously delivering segmentation, deformation, annular excursion (TAPSE and MAPSE), and PH prediction. Registration This retrospective study used existing datasets. Code is openly released, and trained model weights are available to credentialed investigators, for independent evaluation.

BACKGROUND:Mesenchymal stromal cells (MSCs) have the capacity to self-renew and exert immunomodulatory and paracrine effects. Our previously published systematic review of 55 randomized controlled trials (RCTs) of all MSC sources found an overall favorable safety profile for MSCs aside from more frequent fever in the MSC group. OBJECTIVE:Umbilical cord MSCs (UC-MSCs) possess potential biological and economic advantages over other MSC sources. Since several UC-MSC RCTs have been recently published in a heterogenous group of disorders, we performed a systematic review to specifically examine the safety profile of adult UC-MSC RCTs, while also providing quantitative syntheses (meta-analyses) of prespecified adverse events. METHODS:A systematic search was performed up to August 2025. RCTs were identified that compared intravascular delivery of UC-MSCs in adults to a control group. The primary outcome was safety categorized by the following prespecified adverse events with potential relevance to UC-MSCs: (1) fever and (2) non-fever acute infusional toxicity within 24 hours of study product administration, (3) infection, (4) death, (5) malignancy/ectopic tissue formation, (6) thrombotic/thromboembolic events, and (7) mycotic aneurysms. Prespecified adverse events were pooled and meta-analyzed. RESULTS:A total of 9160 studies were screened and 42 UC-MSC RCTs (n = 2280 patients) were included. UC-MSCs were associated with increased risk for fever (relative risk [RR] =2.34, 95% confidence interval [CI] =1.02-5.35). None of the other prespecified adverse events were associated with signals for harm. No included trials were stopped early due to safety concerns. CONCLUSIONS:UC-MSC therapy appears safe for study in heterogeneous patient populations.

Background: Chronic thromboembolic disease is a recognised sequela of acute pulmonary embolism and may occur with or without pulmonary hypertension. The purpose of this study was to evaluate differences in chronic thromboembolic burden, lesion type, and distribution between chronic thromboembolic pulmonary disease without pulmonary hypertension (CTEPD) and chronic thromboembolic pulmonary hypertension (CTEPH). Methods: This retrospective single-centre study included patients with CTEPD or CTEPH by right heart catheterisation using 2022 ESC/ERS criteria between 2021 and 2022. Two CTEPH patients were randomly selected per CTEPD patient. CTPA studies were reviewed by a blinded thoracic radiologist, assessing 32 pulmonary vessels from the main to segmental arteries for chronic thromboembolic lesions. Groups were compared by disease distribution, most proximal lesion level, Qanadli obstruction index, lesion type, and location. Exploratory receiver operating characteristic (ROC) analysis assessed discriminatory performance. Results: The 44 CTEPD and 88 CTEPH patients included had no differences in age, BMI, or sex. CTEPH patients had shorter six-minute walk distance (p = 0.001), greater right ventricular dilation/dysfunction (p < 0.001), and lower prevalence of deep vein thrombosis (p = 0.03). CTPA identified more lesions in CTEPH (21.2 vs. 10.0 lesions/case, p < 0.001) across more lobes (4.8 vs. 3.4, p < 0.001), with more proximal main/lobar disease (p < 0.001). CT obstruction index was 51% in CTEPH and 26% in CTEPD (p < 0.001). Total lesion number best discriminated CTEPH from CTEPD (AUC 0.91; optimal cutoff, 16 lesions). Conclusions: CTEPH demonstrates a greater, more diffuse, and more proximal chronic thromboembolic lesion burden on CTPA than CTEPD.

INTRODUCTION:Optimal management for patients with acute pulmonary embolism (PE) at elevated risk of decompensation remains unclear given the lack of predictors and the limitations of existing risk stratification tools. Thus, we aimed to evaluate the prognostic association between clinical variables and short-term mortality in patients with acute PE. METHODS:Medline and EMBASE were searched from inception through 19 May 2025. We included English-language studies that described adult inpatients with a confirmed acute PE and evaluated prognostic factors associated with short-term mortality. Two authors performed citation screening and data extraction. Only adjusted OR were included in the data synthesis. We performed pooling using a random-effects model and inverse variance weighting. Risk of bias was evaluated using the Quality in Prognosis Studies tool. We assessed certainty of evidence using the Grading of Recommendations, Assessment, Development, and Evaluations approach. RESULTS:We included 70 studies involving 1 446 190 participants in our review. Variables with a moderate or high certainty association with short-term mortality in acute PE include increasing age, an elevated Pulmonary Embolism Severity Index (PESI) Score, an abnormal simplified PESI Score, an elevated heart rate, a reduced systolic blood pressure, a lower oxygen saturation, a higher respiratory rate, active malignancy, heart failure, chronic lung disease, an elevated troponin, a larger right ventricle (RV) to left ventricle ratio and evidence of RV dysfunction. DISCUSSION:We identified multiple variables associated with early mortality in acute PE. Notably, we demonstrate that RV function, a marker omitted from commonly used risk stratification tools, represents an important prognostic variable that is supported by multiple society guidelines. REGISTRATION:We registered the protocol with the Open Science Framework (https://osf.io/9w2jg).

Positive pressure ventilation has multiple adverse cardiopulmonary effects leading to impaired lung mechanics and hemodynamic performance. Controlled levels of diaphragm activity may mitigate these. The objectives of the study were to assess the effect of varying levels of diaphragm neurostimulation on respiratory mechanics and hemodynamics during positive pressure ventilation and to establish whether these effects depend on the level of applied positive end-expiratory pressure (PEEP). Patients enrolled in a clinical trial of diaphragm neurostimulation underwent a nested randomized cross-over study assessing the effect of passive ventilation alone or in combination with three levels of diaphragm neurostimulation, performed at two levels of PEEP. Respiratory mechanics, distribution of ventilation, and hemodynamics were assessed. Sixteen patients were included. Diaphragm neurostimulation increased ventilation to the dorsal lung regions in a dose-dependent fashion (median 63

Pulmonary arterial hypertension (PAH) is a rare and progressive disease resulting in increased workload of the right ventricle (RV). Despite advances in therapies, PAH remains highly morbid from a progressive vasculopathy and RV failure. For patients admitted to hospital with RV failure, transplant remains the only option for a select few. Sotatercept, a novel activin signaling inhibitor approved for the treatment of PAH, has demonstrated significant improvement in clinical outcomes across a wide spectrum of disease severity. However, descriptions of its use to treat acute RV failure in an intensive care unit (ICU) setting are limited. We report a case of cardiogenic shock and respiratory failure secondary to decompensated PAH requiring veno-arterial extracorporeal membrane oxygenation (VA ECMO), with rapid improvement following sotatercept initiation.

Although sedation is commonly used in patients undergoing venovenous extracorporeal membrane oxygenation (V-V ECMO), the use of volatile anesthetics is less frequent, and there is limited data on the pharmacokinetics/pharmacodynamics (PK/PD) of these sedatives (e.g., sevoflurane) directly vaporized into ECMO circuits. We aimed to develop a kinetic model for sevoflurane vaporized in ex vivo ECMO oxygenators to inform the design of a pilot clinical trial to evaluate the safety, feasibility, and in vivo PK/PD model of such a strategy in patients supported with V-V ECMO. We conducted ex vivo trials using three ECMO circuits primed with Ringer’s lactate, and the ECMO pump speed set to achieve a fluid flow rate of 5 L/min. Sevoflurane was vaporized at concentrations of 1–8

Background:Reduced exercise capacity and compromised pulmonary function are common findings following Fontan palliation. Although yoga practice has been associated with improved respiratory function in individuals with acquired heart disease, little is known about the potential impact in the Fontan population. Methods:We assessed the safety, feasibility, and efficacy of a 12-week in-person yoga program in adult Fontan patients. Participants with stable Fontan physiology and without limiting comorbidities were included. Study patients were encouraged to continue yoga practice regularly after study completion. Study outcomes included cardiopulmonary exercise testing (CPET) and pulmonary function testing (PFT). Comparisons were made within group (pre/post intervention) at 12 weeks and between group (study patients compared with non-intervention controls) at last follow-up. Results:Ten patients were enrolled and 5 completed the program. No adverse events were reported. There were no statistically significant changes in CPET or PFT measures after the 12-week intervention. At last follow-up [median 77 months (IQR 61-91)], delta forced vital capacity (FVC) differed between the yoga and usual care groups (-0.1L [-0.1, 0.2]) versus -0.2L [-0.4, -0.2], p = 0.032). There were no other significant between-group differences. Conclusions:A 12-week yoga program was well-tolerated in adult Fontan patients. While there was no apparent short-term improvement in CPET or PFT parameters, regular yoga practice may contribute to stabilization of FVC at longer term follow-up. Further studies are warranted to evaluate the clinical impact of yoga post Fontan palliation.

A 70-year-old man with transfusion-dependent myelodysplastic syndrome was treated with luspatercept. Nine months into treatment and 2 months after dose increase, he developed new hypoxemia with a room air resting oxygen saturation of 88%. He was admitted, requiring up to 90% Fio2 by high-flow nasal cannula to maintain saturations ≥ 92%. CT scan demonstrated stable mild probable usual interstitial pneumonia, without ground-glass, pulmonary embolism, or arteriovenous malformation. Agitated saline contrast-enhanced echocardiogram was late positive, consistent with intrapulmonary shunt. Macroaggregated albumin (MAA) shunt scintigraphy demonstrated a shunt of 7.5%. Extensive investigations ruled out congenital portosystemic shunt and hepatopulmonary syndrome. We diagnosed hypoxemia from intrapulmonary vascular dilatation (IPVD), with luspatercept as a possible cause. Oxygen requirements improved, and supplemental oxygen was discontinued by 7 weeks after cessation. Repeat MAA 10 weeks after cessation confirmed shunt resolution (3.6% shunt). This case of reversible IPVD attributable to luspatercept highlights important monitoring implications in treated patients.

Rationale: Despite the potential risks associated with sedation, there is a paucity of pharmacokinetic/pharmacodynamic (PK/PD) data for propofol and fentanyl in patients supported with venovenous extracorporeal membrane oxygenation (V-V ECMO). Objective: Describe propofol and fentanyl PK/PD profiles in patients receiving V-V ECMO. Methods: Prospective, single-center, open-label PK/PD study at the Toronto General Hospital intensive care unit between July 2022 and January 2023. Using high-performance liquid chromatography/tandem mass spectrometry, propofol and fentanyl total concentrations were measured during V-V ECMO. Sequential PK/PD modeling, using sex as a covariate, was conducted with processed electroencephalography (patient state index [PSI]) for sedation and expiratory occlusion pressure and airway occlusion pressure during the first 0.1 second for respiratory effort. Results: Eleven patients underwent 106 evaluations over a median follow-up of 146 (interquartile ranges, 116-146) hours. Patients' average age was 43 (standard deviation, 13) years, and 55% were female. Propofol and fentanyl PKs were best described by a two-compartment model. Propofol PSI PD were described using an effect compartment, with a coefficient of determination of 0.78. There was a significant increase in propofol (P = 0.01) and fentanyl (P = 0.03) clearance within 10 minutes of ECMO initiation, plateauing after 8 hours of ECMO support. Despite this, patient oversedation (PSI <40) occurred in 74% of the observations. Female patients exhibited a higher sedative central volume of distribution and lower propofol clearance. Conclusions: ECMO initiation resulted in a time-limited increased sedative clearance. PSI accurately described sedative PD, but variable respiratory effort was observed irrespective of sedative plasma concentrations. Sex-based differences were found in sedative PK/PD parameters.

Pulmonary hypertension (PH) is a highly prevalent hemodynamic condition that occurs as a complication of circulatory, pulmonary, and systemic disorders, increases right ventricular (RV) afterload, and confers adverse prognosis. For patients experiencing chronic dyspnea, echocardiographic screening may raise suspicion for PH. Untangling its cause(s) can then be challenging, as the circulatory and pulmonary systems are functionally interlinked and diseases in both often co-exist, but is essential to appropriately select therapies that may unload the RV. Here, we highlight advances in the hemodynamic assessment of PH, with a focus on using exercise to clarify the underlying pathophysiology to better individualize management.

Pulmonary arterial hypertension (PAH) is characterized by widespread pruning of the distal pulmonary arterial bed, leading to increased pulmonary vascular resistance, right heart failure, and death. PAH patients receiving background therapy were randomized to receive a course of 4 monthly infusions of placebo or angiogenic endothelial progenitor cells (EPCs) transfected with endothelial NO synthase (eNOS) over the first 6-month period for a cumulative dose of 80 M cells. Due to the COVID-19 pandemic, the trial was stopped after 12 of the planned 45 patients were enrolled (five placebo, seven eNOS-EPCs). While the small sample size precluded appropriately powered analyses, a possible difference was seen in 6MWD between the eNOS-EPC and placebo groups over 6 months, with a greater proportion of patients in the cell therapy group achieving a minimal clinically important increase in 6MWD of 33-m (71.4% vs 40%, respectively). Treatment of PAH patients with a course of 4 monthly infusions of eNOS-EPCs was well tolerated and may have resulted in clinical benefit.

The aim of this study was to develop machine learning (ML) models to explore the relationship between chronic pulmonary embolism (PE) burden and severe pulmonary hypertension (PH) in surgical chronic thromboembolic pulmonary hypertension (CTEPH). CTEPH patients with a preoperative CT pulmonary angiogram and pulmonary endarterectomy between 01/2017 and 06/2022 were included. A mean pulmonary artery pressure of > 50 mmHg was classified as severe. CTs were scored by a blinded radiologist who recorded chronic pulmonary embolism extent in detail, and measured the right ventricle (RV), left ventricle (LV), main pulmonary artery (PA) and ascending aorta (Ao) diameters. XGBoost models were developed to identify CTEPH feature importance and compared to a logistic regression model. There were 184 patients included; 54.9

Lung transplantation is indicated for selected patients with advanced pulmonary arterial hypertension (PAH). We used a modified Delphi process to develop recommendations on care of patients with PAH undergoing lung transplantation. This Delphi panel was recruited from the Pulmonary Vascular Research Institute's Innovative Drug Discovery Initiative - Lung Transplantation Workstream, consisting of clinical and research experts in PAH and lung transplantation. In this process, 29 panelists were given open-ended questions, querying topics related to lung transplantation in PAH. A steering group converted the responses into discrete statements. Panelists then rated agreement using a Likert scale in two further survey rounds: -5 (strongly disagree) to 5 (strongly agree). Consensus was defined as mean ≥ 2.5 or ≤ -2.5, with a standard deviation not crossing zero. Consensus was reached on 141 of 223 statements. Notable areas of consensus were for early discussions about transplantation, and agreement with previously published referral and listing criteria. There was agreement that lung transplantation could be offered in sick candidates, including those with concurrent renal or hepatic insufficiency. Bilateral lung transplantation was considered the procedure of choice for most patients, with rare indications for heart-lung transplantation. Consensus on bridging strategies included use of veno-arterial extracorporeal membrane oxygenation and preemptive awake cannulation in those with severe right ventricular dysfunction. Consensus was also achieved on intraoperative use of invasive hemodynamic monitoring, and prolonged postoperative circulatory support guided by hemodynamic response and echocardiography. Patients with PAH undergoing transplantation require specialized management, which differs somewhat from other candidates.

Background: Diagnosing pulmonary arterial hypertension (PAH) versus pulmonary hypertension associated with left hear disease (PH-LHD) can be challenging in patients with risk factors for both conditions. When resting pulmonary artery wedge pressure (PAWP) is proximate to a threshold of 15mmHg, exercise has been recommended to differentiate pre- versus post-capillary contributions to PH. To improve our understanding of this practice recommendation, we studied relationships between resting PAWP and the balance of pre- and post-capillary contributions to exercise PH. Methods: Patients suspected of PAH (n=29, 72±2y, 52% F) with risk factors for LHD were prospectively recruited to undergo cycle ergometry at time of diagnostic right-heart catheterization. Hemodynamic data, including pressure-flow slopes and contributions of transpulmonary gradient (TPG) and PAWP to mPAP, were analyzed to evaluate pre- and post-capillary contributions, respectively, at rest and during exercise. Results: PAWP ranged from 0 to 20 mmHg. Exercise pressure-flow slopes demonstrated 62% with post-capillary PH, and 31% with pre-capillary PH only. The relationship between resting PAWP and the pre- versus post-capillary contributions to exercise PH was not straightforward. Of patients with PAWP<12mmHg, 67% had post-capillary contributions to exercise PH. Conversely, 50% of patients with PAWP>15mmHg had pre-capillary contributions to exercise PH. Exercise-associated increases in pulmonary artery pressures were more strongly associated with pre-capillary contributions regardless of post-capillary contributions or the value of resting PAWP. Conclusion: In this population, post-capillary contributions to exercise PH were commonly disclosed over a range of resting PAWP, including <12mmHg. The severity of exercise PH was determined by the pre-capillary contributions. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement The study was funded by Heart and Stroke Foundation of Canada Grant in Aid (G-18-0022220). Conflict of interest: All other authors have no relevant financial disclosures. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Institutional research ethics boards at Mount Sinai Hospital (#18-0257-A) and the University Health Network (19-5069.0) approved this study. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes The data supporting this study's findings are available from the corresponding author upon reasonable request.

ObjectivesFrailty is associated with increased morbidity and mortality in chronic lung disease, but its prognosis has not been evaluated in pulmonary arterial hypertension (PAH). This study aimed to assess: (1) impact of frailty on hospital length of stay (LOS) and health-care utilization in PAH; (2) association of frailty with 1-year post-discharge outcomes.MethodsRetrospective, single-centered cohort study of consecutive PAH patients admitted non-electively (January 2009-December 2018), predominantly for right heart failure (57%). Frailty was defined as ≥ 0.25 using a cumulative deficits frailty index. Disease characteristics, hospital factors, and mortality were compared using univariate analysis and multivariable regression, adjusting for age and sex.Results44/96 (46%) PAH patients were frail. Frailty was associated with older age, greater comorbidities, and lower six-minute walk distance pre-admission (p < 0.05). Frail patients had a longer hospital LOS (4 days 95% (0.4-6.3), p = 0.04) and were more likely to receive social work consultation (36% vs 13%, p = 0.01), independent of age and sex. There were no adjusted differences (frail vs non-frail) in hospital mortality (OR:1.01 95% (0.28-3.72) or 12-months mortality post-discharge (HR:1.26 95% (0.48-3.29).ConclusionFrailty was associated with greater hospital LOS and interdisciplinary support, but not 1-year mortality. Future studies should explore whether alternative frailty models may be more informative of longer-term PAH outcomes.