Pellikka, Patricia A.

Importance Timely identification of aortic stenosis (AS) is essential for appropriate clinical management, yet screening remains limited by dependence on comprehensive echocardiography and trained imaging personnel. Objective To develop and validate a deep learning algorithm for detection of moderate or greater AS and prospectively evaluate its performance using artificial intelligence (AI)–guided focused cardiac ultrasound (FoCUS) acquired by novice operators. Design, Setting, and Participants This diagnostic study included retrospective algorithm development and validation and prospective evaluation of AI-guided FoCUS across Mayo Clinic sites in the Midwest, Arizona, and Florida. The model was developed using 6753 patients and evaluated in internal validation (n = 852), internal test (n = 844), and validation (n = 1912) cohorts. Performance was assessed on FoCUS acquired by experienced sonographers (n = 602) and prospectively by novice operators (n = 1302). The retrospective model development and validation cohorts comprised studies performed from January 2005 through September 2022. Prospective study was conducted in 2 enrollment periods from June to August 2024 and from June to September 2025. Participants from both periods were combined to comprise the final prospective cohort. Exposure AI-guided FoCUS acquisition and automated deep learning–based assessment for detection of moderate or greater AS. Main Outcomes and Measures Detection of moderate or greater AS. Performance was assessed using area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, and positive predictive value. Results The model demonstrated excellent discrimination in the internal test cohort (AUROC, 0.99; 95% CI, 0.98-1.00) and geographically distinct validation cohorts in Arizona (AUROC, 0.99; 95% CI, 0.97-1.00) and Florida (AUROC, 0.99; 95% CI, 0.96-1.00). Among FoCUS examinations acquired by experienced sonographers, sensitivity was 95% (95% CI, 82-99) and specificity was 97% (95% CI, 95-98). In the prospective novice-operator cohort, 1258 of 1302 examinations (96.6%) were suitable for automated analysis. Sensitivity was 93% (95% CI, 82-99) and specificity was 96% (95% CI, 95-97). Expert review of AI-positive and uninterpretable examinations increased the positive predictive value from 49.4% to 91.1%, with sensitivity of 85.4%. Conclusions and Relevance A deep learning algorithm accurately detected moderate or greater AS across validation cohorts in this study. In prospective evaluation, novice operators were able to acquire AI-guided FoCUS examinations that enabled accurate detection of moderate or greater AS. These findings support a scalable strategy that may expand access to AS detection in settings with limited echocardiography resources.

Abstract Background and Aims Exercise stress echocardiography is recommended as an alternative to invasive testing for diagnosing heart failure with preserved ejection fraction (HFpEF), but an evidence-based operational framework guiding its application is lacking. Methods Patients with chronic unexplained dyspnoea underwent invasive haemodynamic exercise testing with simultaneous echocardiography to test the hypotheses that (i) the current diagnostic algorithms (H2FPEF, HFA-PEFF, and HFpEF-ABA scores) could be enhanced when combined with exercise echocardiography; and (ii) incorporating resting left atrial (LA) compliance (LA reservoir strain divided by E/e′) could further improve diagnostic triage, using separate cut points optimizing sensitivity and specificity. Findings were then validated in an international multicentre cohort. Results Of 482 patients, HFpEF was present in 386 and non-cardiac dyspnoea in 96. Sensitivity to detect HFpEF was only 55%–60% and accuracy 61%–67% using individual diagnostic scores with currently recommended exercise echocardiography. Addition of abnormal resting LA compliance to exercise echocardiography increased sensitivity to 84%–85% but increased the false-positive rate to 31%–43%. Applying separate cut points that optimize specificity and sensitivity (either exercise E/e′ ≥ 13.8 or resting LA compliance ≤1.6% to rule-in HFpEF; both exercise E/e′ < 7.2 and resting LA compliance >4.4% to rule-out HFpEF, remaining patients indeterminate who require invasive testing) improved sensitivity to 95%–99% among definitively classified patients and reduced the number of patients that require invasive exercise testing from ∼60% to ∼30%. Findings were replicated in a multicentre, international validation cohort of patients undergoing rest–exercise echocardiography and invasive exercise testing. Conclusions Currently used non-invasive approaches to HFpEF diagnosis have inadequate sensitivity, indicating that exercise echocardiography cannot supplant invasive testing. Addition of LA strain imaging with optimized rule-in or rule-out values improves diagnostic triage by expanding non-invasive rule-in/rule-out classifications while identifying patients who still require invasive testing in the evaluation of HFpEF.

Abstract Carcinoid heart disease (CHD) is associated with advanced neuroendocrine tumor liver metastases (NETLM) and may preclude surgical cytoreduction. We assessed perioperative and long‐term outcomes of hepatectomy in patients with CHD. We retrospectively analyzed 311 patients undergoing cytoreductive hepatectomy for intestinal NETLM: non‐functional ( n = 163), carcinoid syndrome (CS) without CHD ( n = 110), and CHD ( n = 38), including patients undergoing pre‐hepatectomy valve replacement. CHD patients more frequently had >10 liver metastases (78%) and larger lesions (median 9.5 cm) and required major hepatectomy more often (58%). Major morbidity was higher in CHD (up to 47%), yet 90‐day mortality remained low (≤4%). Median overall survival after hepatectomy was comparable across groups (12.5 vs. 9.1 vs. 11.4 years; p = .19), including matched analyses. With optimal cardiac management, cytoreductive hepatectomy in CHD is feasible and provides long‐term survival comparable to patients without CHD.

Abstract Introduction Transthoracic echocardiography (TTE) identifies a hypercontractile phenotype (HP) in chronic coronary syndromes (CCS), characterized by elevated resting left ventricular (LV) elastance (force = systolic blood pressure/end-systolic volume). To evaluate the prognostic significance and functional correlates of HP. Methods In a prospective multicentre study, 10 677 patients with CCS underwent resting TTE to assess LV ejection fraction (EF), stroke volume, and force by quantitative volumetric echocardiography. All patients were followed for the endpoint of all-cause mortality. In a subset of 5834 patients, stress echocardiography (exercise or dobutamine) was performed for LV contractile reserve and heart rate reserve. Results Patients were stratified into Force quintiles (Q1–Q5). Patients with hypercontractile phenotype exhibited lower stroke volume at rest (Q5 = 34.8 ± 12.3 vs Q1–Q4 = 57.4 ± 19.1 mL; P < .01) and higher EF at rest (Q5 = 64.8 ± 6.9% vs Q1–Q4 = 58.1 ± 8.7%, P < .01). During a median follow-up of 24 months (interquartile range = 12–40 months), 509 deaths occurred. The exposure-adjusted death rate was lowest in Q3 (3.53–4.51 mmHg/mL; 1.03 per 100 person/years) and higher in Q1 (≤2.62 mmHg/mL, 2.88), Q2 (2.63–3.52 mmHg/mL, 1.86), Q4 (4.52–6.11 mmHg/mL, 1.56), and Q5 (HP, >6.11 mmHg/mL; 1.88; P < .0001 vs Q1 and Q3). Multivariable analysis identified HP (Q5; HR 1.531 vs Q3, 95% CI 1.116–2.099; P = .006) and EF (HR 0.963, 95% CI 0.953–0.972; P < .0001) as independent predictors of death. During exercise or dobutamine stress, HP showed reduced LV contractile reserve (ΔEF: Q5 = 4.3 ± 9.4% vs Q1–Q4 = 7.0 ± 9.3%; P < .001) and blunted heart rate reserve (Q5 = 1.77 ± 0.33 vs Q1–Q4 = 1.85 ± 0.39; P < .01). All patients with force-based LV contractile reserve >4.1 (present in 185, 3.2% of the population) survived. Conclusion Patients with CCS with HP assessed by resting TTE demonstrate higher mortality and multilayered functional impairment, including reduced LV contractile and chronotropic reserves. Hypercontractile phenotype improved the prediction of mortality by EF. A ‘stronger’ heart is, in fact, functionally and prognostically weaker.

Abstract Randomized controlled trials (RCTs) provide high internal validity but often rely on restrictive eligibility criteria that limit generalizability and complicate real-world trial emulation. We propose AERO (AI Agent for Adaptive Eligibility Refinement and Optimization), an agentic framework that systematically adapts clinical trial eligibility criteria for application to electronic health record data. AERO integrates external clinical knowledge sources and large language model–based reasoning to classify criteria as strict inclusion, safety exclusion, confounder, or operational artifact. We evaluated AERO by emulating the WARCEF trial using Mayo Clinic Platform data restricted to the pre-trial completion period. Emulation with optimized criteria yielded a hazard ratio of 1.561 (p = 0.0605), consistent with the original neutral trial finding (HR = 1.01, p = 0.91). An ablation analysis demonstrated that eligibility handling decisions materially influence observed treatment effects. These results highlight the importance of systematic, knowledge-informed eligibility refinement in real-world evidence generation.

Abstract Aims Heart failure is prevalent; however, there is no cost-effective screening option. Against formal echocardiography, we assessed the diagnostic performance of focused cardiac ultrasound (FoCUS), performed by novice users guided and interpreted by artificial intelligence (AI) for the assessment of left ventricular ejection fraction (LVEF). Methods and results We prospectively enrolled 496 adults referred for diagnostic echocardiography. Novice operators (without clinical or imaging experience) underwent a 4-h imaging workshop, then performed FoCUS using a point-of-care device (Philips Lumify) utilizing real-time AI-image guidance (UltraSight). Images were assessed by AI-image interpretation software (Mayo Clinic), and two expert blinded echocardiologists. Median age was 67, 39% female, and body mass index range 17–56 kg/m2. Forty-one subjects (8.3%) exhibited moderate or greater LV dilatation and 28 (5.6%) had LVEF <40%. The median scan time was 4 min (IQR 3–5). Adequate views were achieved in 95.0% and 97.4% of subjects for AI interpretation and expert analysis, respectively. A two-step diagnostic screening process for low EF in which only abnormal AI reads or uninterpretable scans were reviewed by experts (15.1%) had a sensitivity 96.2%, specificity 95.4%, PPV 54.3%, and NPV 99.8%. A subsequent prospective validation study of 344 subjects demonstrated similar results (sensitivity 100%, specificity 99.4%). Conclusion In this early feasibility investigation, AI-guidance technology embedded on a handheld device, enabled novice users without clinical or imaging experience to acquire sufficient quality images to accurately assess LVEF with minimal training. The technology evaluated in this study may hold promise for AI-guidance and interpretation to facilitate low-cost screening for cardiac dysfunction.

Abstract Aims Early detection of structural heart disease (SHD) improves patient outcomes. However, population-based screening is not recommended due to the lack of accurate and cost-effective tools. We evaluated the costs of artificial intelligence–enabled electrocardiogram (AI-ECG) alone vs. AI-ECG followed by handheld cardiac ultrasound (HCU) for SHD screening. Methods and results We performed a model-based cost analysis using data from 286 adult patients who underwent ECG and same-day HCU performed by a novice operator. Transthoracic echocardiogram (TTE) was the reference standard. We compared two screening strategies: (i) AI-ECG alone and (ii) a stepwise approach (AI-ECG followed by HCU). We assessed costs per diagnosis of aortic stenosis (AS), increased left ventricular wall thickness (ILVWT), and left ventricular systolic dysfunction (LVSD). Sensitivity analyses were conducted for varying disease prevalence. The stepwise approach decreased the cost per diagnosis of AS from $6386 (AI-ECG alone) to $2746 (57.0% savings), ILVWT from $4448 to $2895 (34.9% savings), and LVSD from $1469 to $1296 (11.8% savings). Overall, the cost per diagnosis for all SHDs combined decreased from $1940 to $1570 (19.1% savings). Sensitivity analysis demonstrated that cost savings were inversely proportional to disease prevalence. Nevertheless, stepwise screening remained cost-saving compared with AI-ECG alone until prevalence exceeded ∼55.9% for AS, 28.9% for ILVWT, 20.7% for LVSD, and 40.8% for all SHDs combined. Conclusion A stepwise screening strategy incorporating HCU after a positive AI-ECG reduces the immediate costs of SHD detection by minimizing unnecessary TTEs. This approach may enhance the feasibility of population-based SHD screening, particularly in lower-prevalence settings.

Background Tricuspid regurgitation (TR) is a common valvular disorder that can affect patients’ quality of life and survival. The impact of TR etiology on overall survival and the associated risk factors in each subgroup are not well studied. Methods Multisite, retrospective study evaluating survival based on TR etiology using Kaplan–Meier and Cox regression models. Stepwise approach was used to define TR etiology (≥moderate): primary (PTR; presence of a primary valvular pathology), lead associated (LTR; lead interfering with closure), and secondary TR (STR) in the setting of left‐sided valvular disease, left ventricular disease, other causes of pulmonary hypertension, right ventricular disease, and atrial secondary TR. Results A total of 12 899 patients were included (3% PTR, 90% STR, 7% LTR). The mean age was 68±14, 73±13, and 71±13 years, respectively (44%–55% female patients). Adjusting for age, sex, and TR severity, patients with PTR had better overall survival (PTR: reference; STR hazard ratio [HR], 1.4 [1.2–1.6], LTR HR, 1.5 [1.1–1.7]). On multivariable Cox regression, the Tricuspid Regurgitation Impact on Outcomes score was associated with mortality across all groups ( P <0.05 for all). In addition, other risk factors associated with worse outcomes included pulmonary hypertension and ≥moderate right ventricular enlargement in PTR; left‐sided valvular disease, pulmonary hypertension, and ≥moderate right ventricular dysfunction in STR; and left ventricular disease and ≥moderate right ventricular dysfunction in LTR. Conclusions Patients with PTR have better survival than STR and LTR. Tricuspid Regurgitation Impact on Outcomes score was associated with mortality across etiologies, whereas other risk factors are etiology specific. Improved risk profiling may enhance patient selection, treatment decisions, and outcomes.

BACKGROUND: Guidelines acknowledge that discordant low-gradient (LG) aortic stenosis (AS) may be severe, but verifying this can be challenging. Right heart catheterization during exercise is considered the gold standard for evaluating ventricular hemodynamics. No invasive studies have compared the hemodynamic response of discordant LG and severe AS during exercise. The aim of this observational study was to describe exercise hemodynamics in patients with asymptomatic discordant AS and left ventricular ejection fraction ≥50%. METHODS: Patients with aortic valve area ≤1.5 cm 2 underwent right heart catheterization at rest and during maximal exercise, measuring pulmonary capillary wedge pressure (PCWP), cardiac output (CO), and the PCWP/CO-slope. Patients were stratified into 3 groups: discordant LG AS (aortic valve area ≤1.0 cm 2 and mean gradient <40 mm Hg); moderate AS (aortic valve area >1.0 cm 2 ); and high-gradient (HG) severe AS (aortic valve area ≤1.0 cm 2 and mean gradient ≥40 mm Hg). RESULTS: Among 86 patients, 17 (20%) had discordant LG, 49 (57%) moderate, and 20 (23%) HG severe AS. The median PCWP/CO-slope was significantly steeper in discordant LG (3.3 [interquartile range, 2.1–4.3] mm Hg/L/min) and HG severe AS (2.7 [1.9–3.4] mm Hg/L per minute) compared with moderate AS (1.9 [0.7–2.8] mm Hg/L per minute), P =0.004. In a regression model adjusted for age, sex, and rest PCWP, systemic arterial compliance and AS severity were significantly associated with the PCWP/CO-slope. Furthermore, patients with discordant LG AS had a leftward-upward shift in the PCWP/CO-curve. CONCLUSIONS: Discordant LG and HG severe AS had similar hemodynamic responses to exercise with steeper PCWP/CO-slope than in moderate AS, suggesting that discordant LG AS is a severe form of AS. In addition, the left upwards shift in PCWP/CO-curve for discordant LG compared with HG severe AS indicates that this group also has heart failure with preserved ejection fraction physiology. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT04913870 and NCT02395107.