Citations

BACKGROUND: Cavotricuspid isthmus (CTI) ablation is frequently performed either as a standalone procedure or in combination with pulmonary vein isolation. With the rapid adoption of pulsed field ablation for atrial fibrillation, it is essential to delineate the utility of this modality in treating CTI-dependent atrial flutter (AFL). This study aims to evaluate the procedural and clinical outcomes of CTI ablation using pulsed field energy. METHODS: We conducted a retrospective analysis of consecutive patients who underwent pulsed field ablation for CTI-dependent AFL between January 2024 and March 2025. The primary end points were acute procedural success, periprocedural complications, and CTI-dependent AFL recurrence during follow-up. RESULTS: A total of 132 patients underwent CTI nonthermal ablation. The median age was 69.5 years, and 27.3% were female. The Farawave catheter was used in 93.9% of cases, PulseSelect in 4.5%, and Sphere-9 in 1.5%. Acute block was achieved in 99.2% of patients, although 8 required adjunctive radiofrequency ablation to complete the line. Periprocedural complications included transient ST-segment elevation in 2 patients and transient conduction disturbances in 3. During a median follow-up of 114 days (n=131), 5 patients (3.8%) experienced recurrence of typical AFL. The 6-month typical AFL-free survival estimate was 93.6%. CONCLUSIONS: Pulsed field ablation appears to be a feasible and effective strategy for CTI-dependent AFL. However, anatomic variability may limit its universal applicability with current catheter designs. Although acute procedural success is high, the long-term durability of the CTI block and its comparative efficacy versus conventional thermal ablation remain areas requiring further investigation.

ABSTRACT Background While radiofrequency ablation (RFA) and cryoablation (CBA) are established treatments for atrial fibrillation (AF), pulsed field ablation (PFA) offers potential benefits. This study compares PFA's procedural outcomes with RFA and CBA. Methods A PRISMA‐guided systematic review searched PubMed, Cochrane, Embase, and Scopus through September 2024. Studies comparing PFA to RFA or CBA in paroxysmal and persistent AF were included. Results From 1413 titles, 28 studies (7944 patients, 4–13 months follow‐up) met inclusion criteria. PFA showed lower recurrence versus RFA (RR = 0.85, p < 0.05) and CBA (RR = 0.81, p < 0.01). Procedure times were shorter versus RFA (‐46.39 min, p < 0.01) and CBA (‐9.26 min, p < 0.01). PFA had fewer complications than CBA (RR = 0.45, p < 0.01) and thermal ablation (RR = 0.55, p < 0.01), with similar rates to RFA (RR = 0.81, p = 0.25). Fluoroscopy times were higher. Conclusion Promising results with PFA suggest it as a viable alternative to thermal ablation.

Background: Pulsed Field Ablation (PFA) has gained widespread adoption for pulmonary vein isolation (PVI) due to its tissue selectivity and safety profile. However, its role in cavotricuspid isthmus (CTI) ablation for typical atrial flutter is less defined. This study compares CTI ablation using PFA versus radiofrequency ablation (RFA) in patients undergoing PVI exclusively with PFA. Research Question: We hypothesized that in patients undergoing PFA-based PVI, CTI ablation using PFA would reduce procedure and fluoroscopy times without increasing complication or flutter recurrence rates compared to RFA. Methods: We retrospectively analyzed patients who underwent CTI ablation with either PFA or RFA alongside PFA PVI between 2024 – 2025 across all Mayo Clinic sites. Baseline characteristics, procedural parameters, acute success, and 12-month arrhythmia recurrence were compared between groups. All patients received standard follow-up, including ECGs and Holter monitoring. Results: A total of 350 patients were included (CTI PFA: n = 126; CTI RFA: n = 224). Patients in the PFA group were older (68.8 ± 8.5 vs. 65.7 ± 10.1 years; p <0.01) and had higher CHA2DS2-VASc scores (2.84 ± 1.57 vs. 2.45 ± 1.45; p = 0.02). Fluoroscopy (18.9 ± 9.1 vs. 25.2 ± 23.3 min; p < 0.01 ) and procedure times (172.6 ± 22.4 vs. 224.7 ± 74.2 min; p < 0.01) were significantly lower in the PFA group. Complication rates (8.7% vs. 4.9%; p = 0.15) and typical flutter recurrence (3.1% vs. 1.7%; p = 0.46) were similar between groups during a mean follow-up of 125 days. Conclusion: In patients undergoing PVI with PFA, CTI ablation using PFA is associated with significantly reduced procedure and fluoroscopy times compared to RFA, without increased complications or arrhythmia recurrence. These findings support the feasibility and efficiency of PFA for typical flutter ablation when performed during PFA-based AF procedures.

e21598 Background: While large language models (LLMs) like GPT-4 (OpenAI) and Claude Opus (Anthropic) perform well on general medical exams, limited data exist on their application in oncology, particularly in multiple-choice question (MCQ) scenarios. This study benchmarks the accuracy and assesses the clinical reasoning of LLMs on cutaneous oncology MCQs from the American Society of Clinical Oncology (ASCO) question bank. Methods: Skin oncology MCQs from the ASCO question bank were tested using GPT-4 (OpenAI) and Claude Opus (Anthropic). Questions were entered into OpenAI and Anthropic API playgrounds without additional prompting, with identical settings for both models (Temperature = 0, Token = Max). Each question was tested three times per model to assess response consistency and precision. Initial responses were collected, and chain-of-thought (COT) prompting was applied to encourage stepwise reasoning mimicking an oncologist’s thought process. The percentage of revised answers after COT prompting was analyzed. Accuracy before and after COT prompting was compared. Incorrect responses were reviewed by board-certified oncologists specializing in skin cancer at National Cancer Institute (NCI)-designated centers. Each oncologist scored the responses on clarity, clinical reasoning, bias, and confidence in the models’ utility in clinical practice. Qualitative feedback was collected via free-text comments and analyzed descriptively. Results: Of 66 skin oncology MCQs, Claude correctly answered 55 questions (83% accuracy, 95% CI: 72%–91%), while ChatGPT correctly answered 53 (80% accuracy, 95% CI: 68%–89%). Both models performed comparably, with overlapping confidence intervals. A chi-square test showed no significant difference between their performances (p = 0.65). After COT prompting, both tools remained at their respective accuracy level. Thematic analysis of incorrect answers revealed that LLM responses frequently deviated from clinical guidelines and failed to reflect evidence-based oncology practice, often neglecting key trial data like DREAMSeq and frontline therapeutic strategies. The most frequent errors were Treatment Planning Errors, where recommendations misaligned with NCCN guidelines, and Interpretation Errors, involving misinterpretation of clinical trials or patient data. Anchoring bias emerged as a recurring issue, with models fixating on details (e.g., BRAF mutation status) at the expense of broader reasoning. Conclusions: This study demonstrates that while large language models like GPT-4 and Claude Opus show promise in the field of oncology, they still face significant challenges in clinical reasoning and adherence to evidence-based practices. Continued refinement and training on specialized medical data are essential before these models can be reliably integrated into clinical workflows.

e13637 Background: Large language models (LLMs) like GPT-4 (OpenAI) and Claude Opus (Anthropic) showed high accuracy in medical multiple-choice exams, but data on their oncology-specific clinical reasoning and performance is limited. This study evaluates their accuracy and clinical reasoning on breast oncology multiple-choice questions (MCQs) from the American Society of Clinical Oncology (ASCO) question bank. Methods: Using OpenAI and Anthropic Application Programming Interface (APIs), questions were tested without additional prompts under consistent settings (GPT-4 and Claude Opus; Temperature = 0, Tokens = Max). Each question was tested three times to assess precision. Then, Chain-of-thought (COT) prompting was applied to promote LLMs stepwise reasoning to increase their accuracy. Accuracy before and after COT prompting was compared. Incorrect responses were reviewed by board-certified medical oncologists speclized in breast cancer, who scored reasoning clarity, bias, and clinical relevance. Qualitative feedback was descriptively analyzed. Results: A total of 273 breast oncology MCQs were evaluated across the two LLMs. GPT-4 achieved an initial accuracy of 81.3% (222/273; 95% CI: 76.3%-85.5%), compared to Claude Opus, which achieved an accuracy of 79.5% (217/273; 95% CI: 74.3%-83.9%). The Chi-squared test for difference between the models before chain-of-thought (COT) prompting yielded a p-value of 0.59, indicating no statistically significant difference in accuracy between GPT-4 and Claude Opus prior to COT prompting. After COT prompting, the performance of the two models diverged significantly. GPT-4 saw a net decline in accuracy, decreasing by a net of 1 correct answer, resulting in an overall accuracy of 80.95% (221/273). In contrast, Claude Opus experienced a notable improvement, with a net of 19 additional correct answers, leading to an accuracy of 86.4% (236/273). A statistical analysis using a Chi-squared test revealed a difference in accuracy between the two models of 5.5% (p = 0.08), demonstrating that the improvement in accuracy for Claude Opus after COT prompting was borderline statistically significant compared to GPT-4. Thematic analysis of oncologists' feedback revealed that the most common reasons for incorrect answers were reliance on outdated guidelines, misinterpretation of clinical trial data, and failure to consider multidisciplinary or patient-specific approaches in clinical decision-making. Conclusions: Although AI models can achieve high scores on multiple-choice exams, they still require human supervision. These models rely on potentially outdated training data and lack the ability to individualize patient care or apply data from clinical trials. especially in unique or unconventional/non textbook scenarios.

e20011 Background: The therapeutic landscape of lung cancer is rapidly evolving, presenting oncologists with the challenge of staying updated amidst an overwhelming influx of data. Clinical decision support (CDS) tools, including artificial intelligence (AI) and large language models (LLMs), may help bridge this gap. Evaluating the accuracy of LLMs in complex, real-world oncology scenarios is crucial to understanding their potential. Methods: Twenty-five de-identified lung cancer cases from the fellows’ clinic at Karmanos Cancer Institute, Detroit, MI, were analyzed. Two LLMs, GPT-4 (OpenAI) and Claude Opus (Anthropic), were assessed using advanced prompting techniques like persona-based and chain-of-thought prompting. Five board-certified lung cancer oncologists from NCI-designated centers evaluated LLM-generated responses based on accuracy, treatment recommendation comprehensiveness, and supportive care planning, using a 1–5 scale. Novel insights, the presence of fabricated information, and harmful recommendations were flagged as binary outcomes. Oncologists were blinded to the LLM source and actual treatment decisions. Results: Table 1 presents patient characteristics. GPT-4 achieved an average accuracy score of 4.2 (95% CI, 3.9–4.4), with 3.7 for comprehensiveness of medical/surgical treatment recommendations and 3.7 for supportive care planning. Six responses (32%) were flagged as potentially harmful, and two (8%) contained inaccuracies. Sixteen GPT-4 responses (64%) were rated trustworthy as a CDS tool. Claude Opus had an average accuracy score of 3.6 (95% CI, 3.1–4.1), scoring 3.6 for treatment recommendation comprehensiveness and 3.5 for supportive care planning. Nine responses (36%) were flagged for potential harm, and five (20%) included inaccuracies. Eleven Claude responses (44%) were deemed trustworthy. Significant differences were observed in accuracy (p=0.04) and trustworthiness (p=0.03) between models using McNemar's test. Other factors showed no statistical significance. Conclusions: GPT-4 outperformed Claude Opus in accuracy and trustworthiness, but both models demonstrated limitations, including harmful recommendations and inaccuracies. These findings highlight the need for improved LLM refinement before routine use as CDS tools in lung cancer treatment. Patient demographics and clinical characteristics. Category subcategory Number Median Age (range)- Yr 65 (26-78) Female 7 Male 18 Histology Adenocarcinoma 10 Squamous Cell Carcinoma (SCC) 7 Small Cell Carcinoma 6 Poorly Differentiated 2 Total 25 Stage NSCLC Stage 3 7 NSCLC Stage 4 13 Small Cell limited stage 3 Small Cell Extensive Stage 2

ABSTRACT Cavotricuspid isthmus ablation (CTI) is a first‐line therapy in patients with typical atrial flutter. With the advent of pulsed field (PF) as a new energy source, we sought to evaluate the use of PF for CTI ablation. A systematic literature search was conducted on the use of PF for CTI‐flutter up to December 2024. A meta‐analysis was performed for studies reporting pooled data, while individual case reports were reviewed and summarized. The mean number of PF applications, acute success rate, and prevalence of coronary vasospasm were evaluated. Eleven studies with pooled data from 155 patients were included. All patients had an acute block of the CTI. The mean number of PFA applications was 7.78 (95% CI 6.53–9.48). The incidence of ST‐elevation was 0.04% (95% CI 0–2.23%). Subclinical vasospasm was documented in 45% (95% CI 32%–59%) of patients who underwent periprocedural coronary angiography. Prophylactic use of nitrates showed a trend toward reducing the incidence of subclinical vasospasm (RR 0.24, 95% CI 0.06–1.06, p = 0.059). Twelve cases with patient‐level data were included; six reported complications, including ST elevation and conduction disturbances. PFA for CTI flutter demonstrates high acute success; however, evidence regarding the durability of the block is limited. Clinical vasospasm with ST segment elevation is uncommon but can lead to life‐threatening complications. The incidence of subclinical vasospasm is high, and nitrates tend toward reducing this phenomenon. To date, the role of PFA for this condition appears to be limited.

Abstract Background Epicardial access for ventricular arrhythmia (VA) ablation has become integral, particularly in non-ischemic and arrhythmogenic right ventricular cardiomyopathy. General anesthesia (GA) or conscious sedation (CS) is typically employed to minimize patient discomfort and movement during epicardial access. This study aimed to identify key differences in baseline characteristics and procedural outcomes between patients undergoing GA or CS. Purpose This study seeks to compare baseline characteristics, procedural details, and outcomes between patients receiving GA versus CS during epicardial access for VA ablation. Understanding these differences may inform procedural planning and patient selection to optimize outcomes. Methods Patients who underwent epicardial access for VA ablation from 2004 to 2024 were included. Baseline characteristics, procedural details, and outcomes were collected and compared using t-tests between GA and CS patients. Additionally, Kaplan-Meier survival analysis was conducted to compare survival in GA and CS groups, stratified by VA type (Ventricular Tachycardia (VT) and Premature Ventricular Complexes (PVC)). Results The study included 265 patients, with 148 receiving GA and 117 receiving CS. Significant differences in baseline characteristics were observed between the groups. CS patients had a higher ejection fraction (48.71% vs. 40.97%, p &lt; 0.01) and a greater proportion of structurally normal hearts (35% vs. 8%, p &lt; 0.01). GA patients had a higher comorbidity burden, including a greater prevalence of diabetes (21% vs. 10%, p = 0.02), heart failure (70% vs. 41%, p &lt; 0.01), hypertension (48% vs. 33%, p = 0.01), device in-situ (80% vs. 50%, p &lt; 0.01), and ischemic heart disease (25% vs. 12%, p = 0.01). PVC ablation was more common in CS patients (37% vs. 18%, p &lt; 0.01). Fluoroscopy time was longer in CS patients (74.94 vs. 60.42 mins, p &lt; 0.01), and GA patients required more ablation lesions (34.13 vs. 24.06, p &lt; 0.01). Kaplan-Meier survival analysis showed no significant differences in survival between GA and CS groups for either VT or PVC. Complication rates and the combined endpoint of recurrence, heart transplant, or death also did not differ significantly between groups. Conclusions Patients undergoing epicardial access with GA had a higher burden of comorbidities and required more ablation lesions, while those with CS generally had better baseline cardiac function and were more frequently treated for PVCs. Although fluoroscopy time was longer in CS patients, complication rates and the combined outcome of recurrence, heart transplant, or death did not significantly differ between groups. Kaplan-Meier analysis further demonstrated similar survival outcomes between GA and CS groups for both VT and PVC. These findings suggest that both GA and CS can be safely used for epicardial access in VA ablation, with each approach tailored to the patient’s clinical profile and procedural needs.Table 1 Figure 1

Abstract Background Cardiac sarcoidosis (CS) is a rare condition associated with electrical conduction disturbances. Ventricular tachycardia (VT) may be a life-threatening manifestation of CS. Immunosuppressants and antiarrhythmic drugs (AADs) are considered first-line management, but for refractory cases, an implantable cardioverter-defibrillator (ICD) and VT ablation may be required. Purpose This study aims to document predictors for complete acute success after VT ablation in patients with CS. Methods This was a post hoc analysis from an observational study that included patients who underwent VT ablation for cardiac sarcoidosis at our institution from 2007 to 2024. Patients with suspected, clinically diagnosed, or histologically diagnosed CS, according to the JCS 2016 guidelines, were included. The cohort was divided into two groups: patients with complete acute success, defined as no VT induced at the end of the procedure, and patients with no acute success, which included patients with partial success, unsuccessful procedures, and no inducibility test at the end of the procedure. After univariate analysis, variables with an uneven distribution between the two groups (p &lt; 0.1) were included in a multivariate logistic regression analysis. Results A total of 95 patients were enrolled. Complete acute success was achieved in 40 patients (42%). The mean age was 56.7 ± 10.1 years, and 14 (15%) patients were female. Extracardiac sarcoidosis involvement was reported in 38 (40%) patients. Only one patient in the complete acute success group did not have an ICD. Before VT ablation admission, 40 (42%) patients were on immunomodulators, and 78 (82%) on AADs. The mean ejection fraction in the complete acute success group was 44.5% ± 11.9, compared to 41.7 ± 12.0 in the no-complete success group. Among the 53 patients who underwent a PET scan in the 6 months before ablation, 17 (32%) showed active inflammation based on 18FDG uptake, with no significant difference between the two groups. LGE was present in 57 of 58 patients with a cardiac MR before ablation. Septal involvement was more common in the complete success group (86% versus 58%; OR, 4.41; 95% CI, 1.02-27.5; P = 0.039). A multifocal distribution was also more common in the complete success group (95% versus 67%; OR, 10.17; 95% CI, 1.3-468; P = 0.011). Fluoroscopy and total procedure time were comparable between both groups. The presence of clinical VT with an RBBB morphology was more common in the no-complete success group (54% versus 79%; OR, 0.31; 95% CI, 0.09-1.02; P = 0.03). In the multivariate analysis, epicardial LGE was associated with lower odds of complete acute success (adjusted OR, 0.087; 95% CI, 0.008-0.944; P = 0.04). Conclusions In our population, the presence of epicardial LGE on cardiac MR prior to VT ablation was an independent predictor of no complete acute success.