Invasive Assessment of Right Ventricular Dysfunction Predicts Periprocedural Outcomes in Patients Undergoing Unplanned Ventricular Tachycardia Ablation

ABSTRACT

Background

Risk stratification in patients undergoing urgent ventricular tachycardia (VT) ablation is limited, particularly regarding right ventricular dysfunction, which is often underrecognized in this setting.

Objectives

We investigated whether the right atrial pressure (RAP) to pulmonary capillary wedge pressure (PCWP) ratio or pulmonary artery pulsatility index (PAPi) predicts short‐term outcomes following urgent VT ablation.

Methods

We retrospectively analyzed 102 consecutive patients who underwent unplanned inpatient VT ablation and had right heart catheterization (RHC) within the preceding 12 months. All were admitted with recurrent VT; 67 (66%) had VT storm. Hemodynamic stratification was performed using two established indices of right‐sided function: RAP:PCWP > 0.6 versus ≤ 0.6 and pulmonary artery pulsatility index (PAPi) < 2 versus ≥ 2, calculated as (pulmonary artery systolic pressure − pulmonary artery diastolic pressure)/right atrial pressure. Outcomes included acute kidney injury (AKI), intraprocedural hemodynamic instability, 30‐day VT recurrence, 30‐day mortality/LVAD/transplant, delayed extubation, postoperative length of stay > 7 days, and same‐day complications. Multivariable logistic regression adjusted for age, sex, LVEF, NYHA class, ischemic cardiomyopathy, RHC type, and moderate‐to‐severe MR/TR.

Results

The cohort was 88% male, mean age 64 ± 11.9 years; 47% had LVEF < 25%, and 58% NYHA III–IV symptoms. RAP:PCWP > 0.6 was independently associated with higher risk of AKI (OR 10.4, 95% CI 2.4–14.1, p = 0.002) and intraprocedural hemodynamic instability (OR 6.1, 95% CI 1.5–16.3, p = 0.050). Age > 60 also predicted AKI (OR 12.8, 95% CI 1.3–21.3). PAPi < 2 and LVEF < 25% were not associated with adverse outcomes.

Conclusion

Among urgent VT ablation patients, elevated RAP:PCWP, but not PAPi, identifies those at increased risk for AKI and intraprocedural hemodynamic instability. Incorporating RAP:PCWP into preprocedural assessment may enhance risk stratification.

Displays bar graphs showing outcome rates stratified by clinical and hemodynamic variables. Asterisks (*) denote statistically significant differences (p ≤ 0.05). “NS” indicates non‐significant comparisons. Percentages are shown above each bar. The top panel corresponds to acute kidney injury (blue bars), and the bottom panel corresponds to acute hemodynamic decompensation (orange bars). RAP:PCWP ratio, LVEF, NYHA class, and age are the variables used for subgrouping. Abbreviations: RAP (right atrial pressure), PCWP (pulmonary capillary wedge pressure), LVEF (left ventricular ejection fraction), NYHA (New York Heart Association), AKI (acute kidney injury), NS (not significant).

Abbreviations

ACC

American College of Cardiology

AHA

American Heart Association

AKI

Acute Kidney Injury

ARVC

Arrhythmogenic Right Ventricular Cardiomyopathy

CI

Cardiac Index

KDIGO

Kidney Disease: Improving Global Outcomes

LV

Left Ventricle/Left Ventricular

LVAD

Left Ventricular Assist Device

LVEF

Left Ventricular Ejection Fraction

MR

Mitral Regurgitation

NYHA

New York Heart Association

PAPi

Pulmonary Artery Pulsatility Index

PCWP

Pulmonary Capillary Wedge Pressure

PEA

Pulseless Electrical Activity

RAP

Right Atrial Pressure

RHC

Right Heart Catheterization

RV

Right Ventricle/Right Ventricular

TR

Tricuspid Regurgitation

VF

Ventricular Fibrillation

VT

Ventricular Tachycardia

1. Introduction

Outcomes following catheter ablation for ventricular tachycardia (VT) in patients with structural heart disease are highly variable. While most individuals recover without complication, some face prolonged hospitalization, procedural complications, or early mortality. This heterogeneity likely reflects underlying differences in myocardial substrate and hemodynamic reserve—of which both the left ventricle (LV) as well as the right ventricle (RV) are key determinants.

Indeed, RV dysfunction has emerged as a key prognostic factor in this setting. A 2013 study demonstrated that echocardiographic RV dysfunction was independently associated with increased mortality after VT ablation [1]. In our own institutional experience, we previously found that moderate‐to‐severe RV dysfunction assessed by echocardiography correlated with prolonged postprocedural hospitalization, especially among patients otherwise thought to have low/medium periprocedural risk (PAINESD score < 17) [2].

However, both studies relied on echocardiographic definitions of RV dysfunction, which can be limited in acutely ill patients—particularly those receiving mechanical ventilation or experiencing elevated left‐sided filling pressures [3, 4].

Given the increasing availability of right heart catheterization (RHC) data, we sought to investigate whether invasive hemodynamic parameters—specifically the right atrial pressure to pulmonary capillary wedge pressure (RAP:PCWP) ratio [5] and the pulmonary artery pulsatility index (PAPi) [6] could provide more accurate markers of RV dysfunction. We hope to expand on previous findings to more precisely define RV dysfunction and evaluate its relationship to procedural complications and clinical outcomes following VT ablation.

2. Methods

2.1. Study Design and Population

We performed a retrospective cohort study using a Cleveland Clinic registry of patients who underwent an urgent (i.e., non‐elective) catheter ablation for scar‐mediated VT between January 2022 and October 2024. Only one ablation per patient was included, using the most recent urgent procedure. For patients who underwent more than one VT ablation during a single hospitalization, only the first procedure was analyzed.

Inclusion criteria required that patients underwent urgent ablation and had a right heart catheterization—either via an indwelling Swan‐Ganz catheter (i.e., intensive care unit setting) or a one‐time measurement in the catheterization lab—within 1 year prior to the procedure. Elective admissions were excluded. The registry was used to obtain demographic, clinical, procedural, hospitalization, and follow‐up data. Intraprocedural and anesthesia records, along with follow‐up assessments and any entries flagged for inconsistencies, were reviewed manually for accuracy.

2.2. Definitions and Outcomes

Right ventricular (RV) dysfunction was defined using a ratio of right atrial pressure (RAP) to pulmonary capillary wedge pressure (PCWP) > 0.6. This threshold is supported in the literature as a marker of RV failure, particularly in advanced heart failure and peri‐left ventricular assist device (LVAD) populations. A normal RAP/PCWP ratio is approximately 0.5; values exceeding 0.6 indicate disproportionate right‐sided pressures and impaired RV filling mechanics [4, 7, 8]. Pulmonary artery pulsatility index (PAPi) was also used to assess RV function and was calculated as (pulmonary artery systolic pressure—pulmonary artery diastolic pressure)/right atrial pressure. A PAPi < 2 has been shown to correlate with reduced RV contractile reserve and is associated with adverse outcomes in cardiogenic shock, advanced heart failure, and pre‐LVAD evaluation [9, 10, 11].

Significant mitral regurgitation (MR) and tricuspid regurgitation (TR) were defined as moderate‐to‐severe or severe ( ≥ 3 + ) as reported on transthoracic echocardiography.

Outcomes of interest included intraprocedural hemodynamic instability, acute kidney injury (AKI), 30‐day VT recurrence, 30‐day mortality or LVAD/transplant, delayed extubation, prolonged length of stay (> 7 days), and same‐day complications.

Hemodynamic instability was defined as significant intraprocedural decompensation requiring premature termination of VT ablation. Based on manual chart review, we identified the proximate cause of instability in each case.

AKI was defined according to KDIGO 2024 criteria as an increase in serum creatinine by ≥ 0.3 mg/dL or ≥ 1.5 times the baseline within 48 h of the procedure [12].

Delayed extubation was defined as extubation occurring after electrophysiology lab exit, excluding patients who were already intubated pre‐procedure.

Prolonged length of stay was defined as hospitalization > 7 days, consistent with previously published thresholds for increased post‐ablation morbidity specifically in patients undergoing urgent VT ablations [13]. Same‐day complications included pericardial complications such as tamponade or other acute events prompting early case termination. These were adjudicated separately and were not included under the primary definition of hemodynamic instability.

2.3. Mapping and Ablation Strategy

All VT ablation procedures were conducted following institutional protocols at Cleveland Clinic and tailored to operator discretion and patient‐specific factors. When feasible, antiarrhythmic medications were withheld at least 48 h prior, and anticoagulation was paused 24 h before ablation. Mechanical circulatory support was employed when clinically necessary. Ultrasound‐guided femoral access was routinely used, and epicardial access, when indicated, was achieved via the modified Sosa technique [14], with or without carbon dioxide insufflation [15]. Catheters were advanced into the left ventricle (LV) using either a transseptal or retrograde aortic approach, depending on operator preference. Mapping and ablation were performed using CARTO 3 or EnSite Precision systems with the following catheters: PentaRay, OctaRay, Optrell, DecaNav, ThermoCool ST/ST‐SF, Advisor HD Grid, TactiFlex, and TactiCath (Biosense Webster, Irvine, CA; Abbott, Chicago, IL). Operators generally followed a structured workflow that began with baseline programmed ventricular stimulation to assess inducibility, followed by substrate mapping during sinus rhythm or atrial/ventricular pacing. When VT was reinduced, activation and entrainment mapping were conducted if feasible. Radiofrequency energy (typically 30–50 W for 30–120 s) was then applied to putative critical isthmus sites and areas of abnormal substrate identified on mapping; targeted substrate included local abnormal ventricular activity (LAVA) potentials [16] or deceleration zones identified with isochronal later activation mapping (ILAM) [17]. Post‐ablation, remapping was performed to confirm substrate elimination and/or myocardial noncapture using high‐output pacing. Final programmed stimulation, using two drive cycle lengths and up to three extrastimuli (down to 200 ms or the ventricular effective refractory period), was repeated to confirm noninducibility of clinical or any sustained VT with a cycle length of > 240 ms. All procedures were conducted in collaboration with anesthesia services, who determined sedation strategy and airway management (including use of endotracheal or laryngeal intubation) based on each patient's clinical status.

2.4. Statistical Analysis

Continuous variables were reported as means with standard deviations or as medians with interquartile ranges, based on their distribution. Categorical data were expressed as counts and corresponding percentages. Baseline characteristics between groups were compared using either a two‐tailed Student's t‐test, Wilcoxon rank‐sum test, or Fisher's exact test, as appropriate. Associations were testing using logistic regression and reported using odds ratios (ORs), each with 95% confidence intervals (CIs). To compare crude event rates across RHC timing tertiles, we used one‐way analysis of variance (ANOVA) to calculate p‐values for each outcome within RAP:PCWP subgroups. Significance testing was based on differences in proportions across the 9–12 month, 3–9 month, and < 3 month intervals. Receiver‐operating characteristic (ROC) and the corresponding area under the curve (AUC) was used to quantify the overall predictive accuracy of RAP:PCWP ratio. All analyses were performed using SPSS software (IBM SPSS Statistics, IBM Corp., Armonk, NY). A p‐value ≤ 0.05 was considered significant.

2.5. Data Availability

All data supporting this study's findings are available from the corresponding author upon reasonable request. Tess Calcagno had full access to all data and takes responsibility for its integrity and the analysis performed.

2.6. Ethics Statement

This study was conducted in accordance with the Declaration of Helsinki and institutional guidelines [18]. The study received approval from the Cleveland Clinic Institutional Review Board, with informed consent waived due to the retrospective design. No sex‐ or race/ethnicity‐based differences were identified in the outcomes of interest.

3. Results

3.1. Baseline Characteristics and Procedural Details

A total of 102 patients met inclusion criteria, all of whom underwent urgent VT ablation during hospitalization for VT storm or cardiogenic shock. Right heart catheterization (RHC) was performed either as a one‐time diagnostic procedure in the catheterization lab (n = 54 [52.9%]) or via an indwelling Swan‐Ganz catheter (n = 48 [47.1%]) a median of 11 (IQR 4–85) days pre‐ablation.

The population was dichotomized into 2 groups based on RV function: preserved RV function (RAP:PCWP ≤ 0.6, n = 55 [53.9%]) versus reduced RV function (RAP:PCWP > 0.6, n = 47 [46.1%]). Baseline characteristics were generally similar between the two groups (Table 1). The mean age was 65.6 ± 10.4 years in the RAP:PCWP ≤ 0.6 group and 63.0 ± 13.1 years in the > 0.6 group (p = 0.289). Most patients were male, including 49 (89.1%) in the preserved RV function group and 41 (87.2%) in the reduced RV function group (p = 0.783). VT storm was the presenting indication in 36 (65.5%) versus 31 (66.0%) patients, respectively (p = 0.961). Prior VT ablation was noted in 19 (34.5%) versus 13 (27.7%) (p = 0.445), and NYHA functional class > II was present in 31 (56.4%) versus 32 (68.1%) (p = 0.232) patients, respectively. A PAINESD score > 17 was observed in 32 (58.2%) of patients with preserved RV function and 29 (61.7%) of those with reduced RV function (p = 0.723). LVEF < 25% was seen in 28 (50.9%) versus 20 (42.6%) patients (p = 0.386), respectively.

TABLE 1.

Baseline characteristics.

Category	RAP: PCWP ≤ 0.6 (N = 55)	RAP: PCWP > 0.6 (N = 47)	p‐value
Patient Characteristics
Male gender, n (%)	49 (89.1%)	41 (87.2%)	0.783
Age, years, mean ± SD	65.6 ± 10.4	63.0 ± 13.1	0.289
VT storm, n (%)	36 (65.5%)	31 (66.0%)	0.961
History of prior VT ablation, n (%)	19 (34.5%)	13 (27.7%)	0.445
eGFR < 30 mL/min/1.73 m2, n (%)	7 (12.7%)	5 (10.6%)	0.733
Baseline creatinine, mg/dL, median [IQR]	1.32 [1.00–1.75]	1.40 [0.90–1.90]	0.411
NYHA Class > II, n (%)	31 (56.4%)	32 (68.1%)	0.232
PAINESD > 17, n (%)	32 (58.2%)	29 (61.7%)	0.723
LVEF < 25%, n (%)	28 (50.9%)	20 (42.6%)	0.386
Moderate‐severe/severe MR, n (%)	10 (18.2%)	6 (12.8%)	0.448
Primary MR, n (%)	8 (14.5%)	6 (12.8%)	0.801
Moderate‐severe/severe TR, n (%)	3 (5.5%)	11 (23.4%)	0.014
Primary TR, n (%)	2 (3.6%)	2 (4.3%)	0.842
Ischemic cardiomyopathy, n (%)	30 (54.5%)	17 (36.2%)	0.071
Prior cardiac surgery, n (%)	24 (43.6%)	20 (42.6%)	0.924
LVAD at baseline, n (%)	3 (5.5%)	6 (12.8%)	0.194

Note: Values are presented as mean ± standard deviation, median [interquartile range], or number (percentage), as appropriate. Comparisons were made between patients stratified by right atrial pressure to pulmonary capillary wedge pressure (RAP:PCWP) ratio < 0.6 and > 0.6.

Abbreviations: ARVC, arrhythmogenic right ventricular cardiomyopathy; eGFR, estimated glomerular filtration rate; IQR, interquartile range; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; NICM, non‐ischemic cardiomyopathy; NYHA, New York Heart Association; PAINESD, Pulmonary disease, Age, Ischemic heart disease, NYHA class, EF, Structural heart disease, Diabetes risk score; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure; SD, standard deviation; TR, tricuspid regurgitation; VT, ventricular tachycardia.

Moderate‐to‐severe tricuspid regurgitation (TR) was significantly more common in the RAP:PCWP > 0.6 group, affecting 11 (23.4%) versus 3 (5.5%) patients (p = 0.014). A trend toward greater prevalence of ischemic cardiomyopathy was seen in the preserved RV function group (30 [54.5%] vs. 17 [36.2%], p = 0.071). The RAP:PCWP > 0.6 group included 2 (4.3%) patients with arrhythmogenic right ventricular cardiomyopathy (ARVC) and 2 (4.3%) with cardiac sarcoidosis, while none were observed in the RAP:PCWP ≤ 0.6 group.

Hemodynamic values obtained from RHC are shown in Table 2. The RAP:PCWP > 0.6 group had significantly higher right atrial pressures: 16.3 mmHg (IQR 5.6–14.0) versus 11.2 mmHg (IQR 4.4–13.0) (p = 0.019). Most of the other major RHC indices were similar between the reduced RV function and preserved RV function groups, respectively. This included the mean RV pressure (21.7 mmHg [IQR 12.8–29.3] vs. 21.7 mmHg [IQR 8.9–30.6], p = 0.931), the mean pulmonary artery pressure (29.1 mmHg [IQR 19.5–34.0] vs. 29.0 mmHg [IQR 21.3–36.3], p = 0.931) and the pulmonary capillary wedge pressure (17.1 mmHg [IQR 11.0–22.0] vs. 19.0 mmHg [IQR 12.0–24.0], p = 0.262). PAPi was numerically lower in the RAP:PCWP > 0.6 group at 2.2 (IQR 1.3–2.7) versus 2.7 (IQR 1.4–2.9) (p = 0.095).

TABLE 2.

Right heart cathetrization data.

Category	RAP: PCWP ≤ 0.6 (N = 55)	RAP: PCWP > 0.6 (N = 47)	p‐value
Days between RHC and ablation, median [IQR]	8 [2–82] days	19 [10–68] days	0.281
Indwelling catheter, n (%)	23 (42.4%)	25 (53.2%)	0.981
Mean RAP, mmHg, median [IQR]	11.2 [4.4–13.0]	16.3 [5.6–14.0]	0.019
Mean RVP, mmHg, median [IQR]	21.7 [8.9–30.6]	21.7 [12.8–29.3]	0.974
Mean PAP, mmHg, median [IQR]	29.0 [21.3–36.3]	29.1 [19.5–34.0]	0.931
PAPi, median [IQR]	2.7 [1.4–2.9]	2.2 [1.3–2.7]	0.095
PCWP, mmHg, median [IQR]	19.0 [12.0–24.0]	17.1 [11.0–22.0]	0.262

Note: Values are presented as mean ± standard deviation, median [interquartile range], or number (percentage), as appropriate. Comparisons were made between patients stratified by right atrial pressure to pulmonary capillary wedge pressure (RAP:PCWP) ratio ≤ 0.6 and > 0.6.

Abbreviations: IQR, interquartile range; PAP, pulmonary artery pressure; PAPi, pulmonary artery pulsatility index; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure; RVP, right ventricular pressure.

Procedural characteristics were largely similar (Table 3). Mechanical ventilation was used during the case in 33 (70.2%) patients with reduced RV function versus 37 (67.3%) in the preserved RV function group (p = 0.761). Maintenance propofol was more frequently employed in the reduced RV function group, 17 (36.2%) versus 11 (20.0%) patients (p = 0.073), though the difference was not statistically significant. High dose propofol ( > 5 mg/kg/h) was used in 1 patient (1.8%–2.1%) in each group (p = 0.922). Norepinephrine equivalent > 0.2 μg/kg/min was administered in 5 (10.6%) in the RV dysfunction group versus 3 (5.5%) patients with preserved RV function (p = 0.302). Epicardial ablation was performed in 8 (17.0%) versus 7 (12.7%) patients (p = 0.564). Median procedure duration was similar between groups: 279 min (IQR 227–305) in the reduced RV function group and 265 min (IQR 211–310) in the preserved RV function group (p = 0.376).

TABLE 3.

Procedural characteristics.

Variable	RAP: PCWP ≤ 0.6 (N = 55)	RAP: PCWP > 0.6 (N = 47)	p‐value
Mechanical ventilation, n (%)	37 (67.3%)	33 (70.2%)	0.761
Maintenance propofol, n (%)	11 (20.0%)	17 (36.2%)	0.073
Propofol > 5 mg/kg/hr, n (%)	1 (1.8%)	1 (2.1%)	0.922
NE equivalent > 0.2 μg/kg/min, n (%)	3 (5.5%)	5 (10.6%)	0.302
Epicardial ablation, n (%)	7 (12.7%)	8 (17.0%)	0.564
Procedural duration, min, median [IQR]	265 [211–310]	279 [227–305]	0.376

Note: Values are presented as mean ± standard deviation, median [interquartile range], or number (percentage), as appropriate. Comparisons were made between patients stratified by right atrial pressure to pulmonary capillary wedge pressure (RAP:PCWP) ratio ≤ 0.6 and > 0.6.

Abbreviations: IQR, interquartile range; NE, norepinephrine; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure.

3.2. Clinical Outcomes Measured by Univariate Analysis

Unadjusted outcomes are summarized in Table 4. Acute kidney injury occurred in 13 patients (27.7%) with RAP:PCWP > 0.6, compared to three patients (5.5%) with RAP:PCWP ≤ 0.6 (p = 0.002). Hemodynamic instability requiring early termination of the case was observed in 5 (10.6%) versus 2 (3.6%) patients (p = 0.019). Post‐operative hospital length of stay greater than 7 days was more prevalent in the RV dysfunction group compared to the preserved RV function group (9 [19.15%] vs. 4 [7.27%], p = 0.045).

TABLE 4.

Summary of univariate and multivariable models using RAP: PCWP > 0.6.

Outcome	RAP: PCWP ≤ 0.6 (N = 55)	RAP: PCWP > 0.6 (N = 47)	Univariate Analysis		Multivariable Analysis
			OR (95% CI)	p‐value	OR (95% CI)	p‐value
30‐day mortality/LVAD/tx	6 (11.1%)	9 (19.1%)	1.93 (0.63‐5.91)	0.691	1.50 (0.43‐5.27)	0.529
30‐day VT	3 (5.5%)	2 (4.3%)	0.77 (0.12‐4.82)	0.764	1.32 (0.66‐4.46)	0.141
AKI	3 (5.5%)	13 (27.7%)	6.63 (1.76‐25.00)	0.002	10.38 (2.38‐14.10)	0.002
Delayed extubation	7/48 (14.6%)	12/40 (30.0%)	2.51 (0.88‐7.16)	0.092	2.33 (0.74‐7.28)	0.147
Hemodynamic instability	2 (3.6%)	5 (10.6%)	3.15 (1.58‐17.08)	0.019	6.14 (1.51‐16.3)	0.050
Post‐operative length of stay > 7 days	4 (7.27%)	9 (19.15%)	3.02 (1.86‐10.54)	0.045	2.31 (0.98‐4.70)	0.080
Same‐day complications	8 (14.5%)	11 (23.4%)	1.80 (0.65‐4.92)	0.261	1.88 (0.62‐5.66)	0.262

Note: Summary of univariate and multivariable logistic regression models testing the association of elevated RAP:PCWP ratio (> 0.6) with key in‐hospital outcomes. Odds ratio (OR), 95% confidence intervals (CI), and p‐values are shown, adjusted for age, sex, left ventricular ejection fraction (LVEF < 25%), New York Heart Association (NYHA) class III and IV, ischemic cardiomyopathy, moderate‐severe mitral regurgitation (MR), and moderate‐severe tricuspid regurgitation (TR), and use of indwelling catheter.

Abbreviations: CI, confidence interval; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; NYHA, New York Heart Association; OR, odds ratio; PAPi, pulmonary artery pulsatility index; TR, tricuspid regurgitation.

Comparing subjects with reduced RV function versus preserved RV function, other outcomes did not differ significantly between the two groups. Delayed extubation occurred in 12/40 (30.0%) versus 7/48 (14.6%) patients (respectively) who were spontaneously ventilating at the time of procedure initiation, excluding those who were intubated on arrival to the electrophysiology lab (p = 0.092). Same‐day procedural complications were noted in 11 (23.4%) versus 8 (14.5%) patients (p = 0.261), respectively. Thirty‐day VT recurrence was documented in 2 (4.3%) versus 3 (5.5%) patients (p = 0.764), respectively. The composite of 30‐day mortality, LVAD implantation, or transplant occurred in 9 (19.1%) versus 6 (11.1%) patients (p = 0.691), respectively.

We looked at outcomes in patient subpopulations stratified based on the relative timing between RHC and the VT ablation procedure—since the variability in this timing may introduce unwanted confounders. Crude event rates for acute kidney injury, hemodynamic instability, prolonged hospitalization, 30‐day mortality or LVAD/transplant, and 30‐day VT recurrence did not differ across RHC timing tertiles in either RAP:PCWP group (Supporting Information S1: Table 3).

Receiver operating characteristic analysis demonstrated good discrimination of RAP:PCWP for predicting acute kidney injury (AUC = 0.75) and modest discrimination for intraprocedural hemodynamic instability (AUC = 0.66) (Supporting Information S1: Figure 1).

3.3. Intraprocedural Hemodynamic Instability Subgroup

Among the seven patients who experienced intraprocedural hemodynamic instability, five had RV dysfunction (RAP:PCWP >0.6), and two had preserved RV function ( ≤ 0.6). Based on detailed procedural review, we identified three phenotypes of hemodynamic instability:

• 1.Progressive instability after multiple rounds of VT induction (n = 2, both RV dysfunction): Both patients developed worsening hemodynamics over time following multiple rounds of VT induction and mapping.
• 2.Immediate collapse after first programmed ventricular stimulation and/or ventricular arrhythmia induction (n = 3, two with RV dysfunction): In all 3 cases, hemodynamics were initially stable but rapidly declined and failed to sufficiently recover after just one round of ventricular arrhythmia induction.
• 3.3. Instability without sustained VT (n = 2, one with RV dysfunction): In these patients, hemodynamic decline was not directly attributable to ventricular pacing, programmed stimulation or sustained arrhythmias.

3.4. Multivariable Models

Multivariable logistic regression was performed for each of the seven outcomes: 30‐day mortality or need for LVAD/transplant, 30‐day VT recurrence, acute kidney injury (AKI), delayed extubation, hemodynamic instability, prolonged post‐operative hospital length of stay ( > 7 days), and same‐day complications. All models were adjusted for age > 60, male sex, LVEF < 25%, NYHA class III or IV, ischemic cardiomyopathy, significant mitral regurgitation (MR) and significant tricuspid regurgitation (TR). To mitigate for confounders, the model also included a variable controlling for the setting in which RHC was performed (i.e., indwelling RHC used to guide inpatient therapy versus one‐time RHC performed in the catheterization lab). In these adjusted models, RAP:PCWP > 0.6 remained independently associated with both AKI (OR 10.38, 95% CI 2.38–14.11, p = 0.002) and intraprocedural hemodynamic instability (OR 6.14, 95% CI 1.51–16.30, p = 0.050). There was also a trend toward prolonged post‐operative hospitalization > 7 days (OR 2.31, 95% CI 0.98–4.70, p = 0.080) in patients with reduced RV function.

Outside of RV dysfunction, a small number of clinical and procedural variables were significantly associated with adverse outcomes. Patients over the age of 60 were more likely to develop acute kidney injury (AKI), with an OR of 12.80 (95% CI 1.30–21.30; p = 0.029). Use of an indwelling pulmonary artery catheter was associated with both delayed extubation (OR 7.35, 95% CI 2.16–25.04; p = 0.001) and prolonged post‐operative hospital length of stay > 7 days (OR 1.31, 95% CI 1.01–9.52; p = 0.049). Higher NYHA functional class (III/IV vs. I/II) was also associated with longer post‐ablation hospitalization (OR 5.17, 95% CI 2.71–9.67; p = 0.001). Interestingly, ischemic cardiomyopathy appeared protective against same‐day complications, including pericardial effusion (OR 0.25, 95% CI 0.07–0.81; p = 0.021).

However, most of the other clinical variables tested in this multivariable model were not independently associated with adverse outcomes. For example, there was no significant relationship between LVEF < 25% and risk of AKI, delayed extubation, prolonged post‐operative hospitalization, or 30‐day events. A complete list of associations is provided in Supporting Information S1: Tables 1 and 2.

In an attempt to further adjust for confounders, we repeated the multivariable analysis in patients who had right heart catheterization within 3 months of ablation (n = 58). Findings were directionally consistent with the main cohort. Patients with RAP:PCWP > 0.6 had higher rates of acute kidney injury (21.4% vs. 6.7%, p = 0.126) and intraprocedural hemodynamic instability (10.7% vs. 3.3%, p = 0.284). On multivariable analysis, both outcomes remained directionally associated with elevated RAP:PCWP (AKI: OR 3.12, 95% CI 0.54–18.20, p = 0.187; hemodynamic instability: OR 2.01, 95% CI 0.15–26.40, p = 0.601). Although underpowered, these results show the same pattern of risk even when hemodynamics were measured within 3 months of ablation (Supporting Information S1: Table 4).

Finally, to account for possible confounding effects of mechanical hemodynamic support, we performed a sensitivity analysis excluding all LVAD patients (Supporting Information S1: Table 5). RAP:PCWP > 0.6 remained a significant predictor of AKI in this subgroup (OR 4.86, 95% CI 1.18–20.04, p = 0.029).

3.5. Pulmonary Artery Pulsatility Index

We re‐stratified the cohort using PAPi as a surrogate for RV function: preserved RV function (PAPi ≥ 2, n = 83 [81.4%]) versus reduced RV function (PAPi < 2, n = 19 [18.6%]). Univariate and multivariable analyses were repeated (details in Supporting Information S1: Table 2; summarized in Table 5). Both models failed to show statistically significant association between PAPi < 2 and any of the six outcomes of interest.

TABLE 5.

Summary of univariate and multivariable models using PAPi < 2.

Outcome	PAPi ≥ 2 (N = 83)	PAPi < 2 (N = 19)	Univariate Analysis		Multivariable Analysis
			OR (95% CI)	p‐value	OR (95% CI)	p‐value
30‐day mortality/LVAD/tx	6 (7.20%)	9 (47.4%)	3.11 (1.44‐35.06)	0.016	1.83 (0.49‐6.76)	0.368
30‐day VT	2 (2.4%)	3 (15.8%)	2.77 (0.53‐8.32)	0.235	1.78 (0.11‐5.63)	0.779
AKI	13 (5.7%)	3 (15.8%)	1.11 (0.28‐4.39)	0.983	2.16 (0.67‐7.01)	0.196
Delayed extubation	12/79 (15.2%)	7/9 (77.8%)	4.95 (0.87‐21.34)	0.671	2.25 (0.69‐7.30)	0.177
Hemodynamic instability	5 (6.0%)	2 (10.5%)	1.84 (0.33‐10.47)	0.591	5.67 (0.57‐14.2)	0.139
Post‐operative length of stay > 7 days	10 (12.0%)	3 (15.8%)	1.10 (0.12‐10.41)	0.541	2.33 (0.74–7.28)	0.147
Same‐day complications	14 (16.9%)	5 (26.3%)	1.28 (0.24‐6.70)	0.917	1.62 (0.54–4.89)	0.392

Note: Summary of univariate and multivariable logistic regression models testing the association of low pulmonary artery pulsatility index (PAPi < 2) with key in‐hospital outcomes. Odds ratios (OR), 95% confidence intervals (CI), and p‐values are shown, adjusted for age, sex, LVEF ( < 25), NYHA class III and IV, ischemic cardiomyopathy, moderate‐severe MR, moderate‐severe TR, and use of indwelling catheter.

Abbreviations: CI, confidence interval; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; NYHA, New York Heart Association; OR, odds ratio; RAP:PCWP, right atrial pressure to pulmonary capillary wedge pressure ratio; TR, tricuspid regurgitation.

4. Discussion

In this study of patients undergoing urgent VT ablation, we found that invasive hemodynamic evidence of right ventricular dysfunction—defined by a right atrial pressure to pulmonary capillary wedge pressure ratio (RAP:PCWP) > 0.6—was independently associated with both intraprocedural hemodynamic instability and post‐procedural acute kidney injury (AKI). In contrast, the pulmonary artery pulsatility index (PAPi < 2) did not significantly predict these outcomes in adjusted models. These findings support the use of RAP:PCWP as a physiologically grounded marker of RV dysfunction that may reflect venous congestion and impaired preload reserve—two critical contributors to procedural vulnerability in this high‐risk population.

Mechanistically, hemodynamic instability during VT ablation in patients with RV dysfunction may result from impaired preload recruitment and reduced forward flow. In this cohort, four of five patients with RV dysfunction who developed instability did so following VT induction. Rapid monomorphic VT shortens diastole and decreases LV preload. In patients lacking sufficient RV reserve, these effects may not be sufficiently mitigated by compensatory increases in venous return and RV contractility [19]. The reason RAP:PCWP outperformed PAPi is that RAP:PCWP may better reflect the chronic RV afterload mismatch and venous congestion that drive these outcomes. In contrast, PAPi, calculated as (pulmonary artery systolic pressure – diastolic pressure)/RAP, reflects pulsatile RV contractility and may be less reliable in patients with elevated left‐sided pressures or pulmonary hypertension, where the pulsatile component of afterload is distorted [20, 21].

Pre‐operative risk stratification efforts have centered around incorporating multiple clinical markers that are commonly available before VT ablation. PAINESD, the most universally adopted risk assessment tool, enjoys ample evidence supporting its predictive power [22, 23] and has recently been utilized to guide multidisciplinary pre‐operative management of high‐risk patients [24]. However, it is important to keep in mind that the PAINESD algorithm was derived primarily as a predictor of perioperative hemodynamic decompensation in an era when the incidence of such complication was very high. Indeed, even though the PAINESD study subjects (2010–2011) included elective cases, as many as 11% of them suffered from acute decompensation; in contrast, our study (2022–2024) focused only on urgent VT ablation cases, yet the frequency of acute hemodynamic collapse was just under 7%. Conceivably, as the field of VT ablations evolves (shift in focus from extensive VT activation mapping to substrate‐based ablation techniques; use of more powerful tools that require less aggressive fluid irrigation etc.), pre‐operative risk stratification may need to become more nuanced—perhaps including invasive RV functional assessment in select patients.

There is limited existing literature on how to risk stratify patients for VT ablation based on RV dysfunction. As mentioned earlier, the most extensive study on this topic by Ujeyl et al (enrollment 1999–2008) relied on echocardiographic RV assessment and showed association between RV dysfunction and post‐operative mortality [1]; however, more subtle endpoints such as AKI or acute hemodynamic collapse were not specifically explored. Echocardiographic RV dysfunction was also surveyed in a subset of patients (38%) in the original PAINESD study but was not independently associated with outcomes, likely owing to limited statistical power. Conversely, in our own experience, echocardiographic RV dysfunction was a predictor of prolonged post‐operative hospitalization in individuals with a PAINESD score < 17 [2]. To our knowledge, the present study is the first to report the utility of invasive RV functional assessment in patients undergoing urgent VT ablations. Central Illustration 1

Central Illustration 1.

Displays bar graphs showing outcome rates stratified by clinical and hemodynamic variables. Asterisks (*) denote statistically significant differences (p ≤ 0.05). “NS” indicates non‐significant comparisons. Percentages are shown above each bar. The top panel corresponds to acute kidney injury (blue bars), and the bottom panel corresponds to acute hemodynamic decompensation (orange bars). RAP:PCWP ratio, LVEF, NYHA class, and age are the variables used for subgrouping. Abbreviations: RAP (right atrial pressure), PCWP (pulmonary capillary wedge pressure), LVEF (left ventricular ejection fraction), NYHA (New York Heart Association), AKI (acute kidney injury), NS (not significant).

Pre‐operative invasive RV function assessment has been most extensively studied in the LVAD literature, with PAPi [25, 26, 27] and RAP:PCWP [28, 29, 30, 31] being the pre‐eminent predictors of post‐LVAD outcomes including RV failure and mortality. Data are more sparse in other contexts, although there is evidence linking reduced PAPi to the development of AKI after heart transplantation [32] or cardiac surgery, in general [33]. While there is no consensus on what values of RAP:PCWP constitute “high risk” (a concept that, itself, depends on the exact clinical context), most studies use RAP:PCWP > 0.60–0.63 as their dichotomization point. More variability exists in the PAPi literature, although PAPi< 2 is often used as a practical cut‐off point by heart failure specialists in pre‐operative patient evaluation.

Using our findings, it may be challenging to spell out specific recommendations with respect to RV dysfunction‐focused pre‐operative risk stratification and optimization. For example, our data should not be interpreted as supportive of broad adoption of RHC in every patient with structural heart disease slated for a VT ablation. Our findings also do not explicitly prove that pre‐operative RV optimization (e.g., RHC‐guided diuresis) will result in a meaningful change in outcomes. Instead, the association between RAP:PCWP and post‐ablation outcomes may simply contribute to each operator's broader clinical decision making. For example, when the clinical suspicion for RV failure is moderately high, a clinician may wish to consider pre‐op RHC to inform shared decision making (selective risk stratification) and to define which forms of advanced therapies might be available in case of acute hemodynamic decline.

4.1. Limitations

This was a single‐center retrospective study limited by a relatively small sample size. Mortality and LVAD/transplant events were likely too infrequent to detect meaningful differences. Additionally, our cohort included both elective and acute RHCs; we also allowed for a wide time span of RHC data to be included (i.e., subjects were included in the study if their RHC measurements were done up to 12 months pre‐ablation). Although this was necessary to reach an adequate sample size to allow for meaningful analyses, we attempted to address this heterogeneity in several ways. First, we included an indicator variable (indicating acute vs. elective RHC) in all multivariable models; this was designed to (partially) overcome contextual confounders, for example, the fact that patients requiring indwelling RHC may have a poorer prognosis than others. Second, we performed a subanalysis limited to patients with RHC data available within 3 months prior to ablation (i.e., likely a more accurate pre‐operative RV functional assessment) that detected similar numeric trends as in the main analysis—although these findings failed to reach statistical significance. Despite these analytic efforts, the heterogeneity of RHC data may limit the generalizability of our findings.

In addition, the association between RV dysfunction and intraprocedural VT instability in multivariable analysis reached a borderline level of statistical significance (p = 0.050). This finding is directionally consistent with the overall results of the study and the underlying physiologic framework, but should be interpreted in the context of the study size and analytic scope.

It is also important to state that RHC measurements in any single patient can be highly variable and particularly susceptible to ventricular loading conditions, employment of vasopressor/inotropic support, type of ventilation and other factors, which may not be accounted for in this study. Specifically, after additional chart review, we identified 15 of 102 patients (14.7%) whose right heart catheterization measurements were obtained during positive pressure ventilation. In a subset of patients, ventilation status at the time of hemodynamic acquisition was not explicitly documented, particularly in retrospective outpatient or external RHC reports. Mechanical ventilation may influence intracardiac pressures and derived hemodynamic indices and therefore may represent a source of residual confounding. Similarly, the clinical indication prompting RHC referral, as well as the effect that invasive hemodynamic data had on subsequent patient management, were not accounted for in this study, even though these factors may also have influenced our results.

Our use of fixed cutoffs (RAP:PCWP > 0.6 and PAPi < 2) could also be challenged. As discussed earlier, these values are primarily derived from LVAD literature; cutoff boundaries more appropriate for the context of VT ablations may exist. However, given the mixed heart diseases etiologies in our cohort, a data‐driven derivation of new thresholds (e.g., via receiver operating characteristic analysis) would likely be underpowered and beyond the scope of this investigation.

Another limitation of our data is the availability of corresponding assessment of forward flow in the hemodynamic data obtained. While cardiac index (CI) data were easily obtained from elective RHC reports, deducing reliable CI data in patients with indwelling Swan‐Ganz catheters posed a major challenge in our retrospective data collection (CI values were often not directly reported in the ICU notes and, unlike intracardiac pressure measurements, O2 saturation data were not always accessible to allow for Fick‐formula calculation). Because reliable CI data were only available for ~ 60% of our cohort, we chose not to include this metric in our analyses. Furthermore, since both RAP:PCWP and LVEF already incorporate major determinants of cardiac performance, adjusting for CI could even risk overfitting.

Finally, we acknowledge that the diagnosis of acute hemodynamic decompensation is inherently based on a subjective decision of the operator to prematurely terminate the procedure; this may influence the generalizability of our results.

4.2. Future Directions

Risk stratification prior to substrate VT ablation remains challenging. Current recommendations from international societies suggest integrating structural, electrophysiologic, and hemodynamic data to guide management [34, 35]. Our findings support the potential utility of invasive RV measures, such as RAP:PCWP, in this framework. However, additional studies are needed to validate these findings; indeed, a prospective investigation using a standardized RHC protocol may be necessary to help decide whether invasive RV assessment should enter routine practice for every high‐risk VT ablation case. The role of pre‐operative RV optimization may also need be explored.

5. Conclusion

In this cohort of patients undergoing urgent VT ablation, invasive hemodynamic assessment of RV function using RAP:PCWP identified individuals at greater risk for AKI and intraprocedural instability. While PAPi, an alternative metric of RV function, showed similar directional trends, it did not reach statistical significance. RAP:PCWP may offer a more useful reflection of RV dysfunction in this population and could play a key role in pre‐ablation risk stratification moving forward.

Funding

The authors have nothing to report.

Disclosure

Dr. Dresing has received speaking honorarium from Medtronic and Boston Scientific. Dr. Hussein receives research grant support from Boston Scientific. Dr. Callahan has served as a consultant for Abbott, Boston Scientific, Medtronic, Philips and ShockWave. Dr. Kanj has received speaking honorarium from Boston Scientific and has served as a consultant for Boston Scientific. Dr. Martin has served as a consultant for Medtronic. Dr. Saliba has served as a consultant for Boston Scientific and Biosense Webster and as an advisory board member for Boston Scientific. Dr. Taigen has served as a consultant for Biosense Webster and Medtronic. Dr. Varma has served as a consultant for Abbott, Biotronik, Boston Scientific and Medtronic. Drs. Higuchi and Lee have served as consultants for Abbott. Drs. Wazni and Santangeli have served as consultants for Abbott, Biosense Webster, Boston Scientific and Medtronic. Dr. Sroubek has received speaking fees from Biosense Webster and Abbott; he receives research grant support from Abbott. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Supporting information

Supporting File

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.