Electrical cardioversion in patients with left ventricular thrombus: A systematic review of safety and thromboembolic risk

Introduction

Cardiovascular disease remains the leading cause of death globally, accounting for an estimated 19.8 million deaths in 2022, thereby underscoring the importance of optimizing decision-making around arrhythmia management and thromboembolic risk in complex cardiac patients. ¹

Electrical cardioversion remains a cornerstone of rhythm-control therapy for atrial fibrillation (AF) and other clinically important arrhythmias, but it carries well-recognized thromboembolic risks when undertaken in the presence of intracardiac thrombus. In particular, left-atrial appendage (LAA) thrombus is an established contraindication to elective cardioversion or left-sided ablation because of the high risk of systemic embolization; contemporary guidelines therefore recommend exclusion of LAA thrombus and adequate anticoagulation before cardioversion.^2,3

By contrast, the magnitude and clinical relevance of thromboembolic risk attributable to left-ventricular (LV) thrombus at the time of cardioversion are poorly defined. LV thrombi occur most commonly in the setting of large anterior myocardial infarction or severe LV systolic dysfunction and are heterogeneous in appearance and behavior; thrombus morphology (mobility, protrusion, and attachment site), and adequacy of anticoagulation can strongly influence embolic potential. ⁴ Because LV thrombus is less frequently encountered in routine practice than LAA thrombus, clinical practice around electrical cardioversion in the presence of LV thrombus has been inconsistent.

To our knowledge, no prior systematic review has specifically synthesized thromboembolic outcomes after electrical cardioversion or internal procedural shocks in patients with imaging-confirmed LV thrombus. This review, therefore, addresses an important evidence gap by systematically summarizing the available data, characterizing thrombus morphology and anticoagulation patterns in published cohorts, and providing a clinically focused framework for individualized, imaging-guided decision-making in a scenario for which guideline-directed evidence is sparse.

Methods

Data extraction and synthesis

Two reviewers independently screened titles/abstracts and full texts for eligibility and performed data extraction using a standardized proforma that captured study design, period, setting, patient demographics, LV thrombus characteristics (morphology, location), anticoagulation status at the time of cardioversion, procedural details (external versus internal shocks, indication), follow-up duration, and thromboembolic and mortality outcomes. Disagreements were resolved by discussion and consensus, with adjudication by a third reviewer when necessary. Given the small number and heterogeneity of included studies, we summarized study characteristics and outcomes descriptively (counts, percentages, means ± SD where reported) and synthesized findings narratively rather than conducting a meta-analysis. The review is reported in accordance with guidance for narrative reviews and systematic reporting standards; details of screening and reasons for exclusion are provided in the PRISMA-style flow diagram (Figure 1).

OPEN IN VIEWER

Figure 1.

Flow chart for study selection.

Results

Three retrospective studies met the inclusion criteria (Bangalore et al., ⁶ Rao et al., ⁷ and Dabbagh et al. ⁸ ). An abstract-only report (Alkharabsheh et al. ⁹ ) that preceded and was later superseded by the completed Dabbagh study was excluded. None of the articles from gray literature search met final inclusion. Characteristics of the included studies are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Characteristics of included studies.

First author (year)	Dabbagh et al. (2025)N = 39 (%)	Bangalore et al. (2006)N = 21 (%)	Rao et al. (2016)N = 8 (%)
Data collection period	2010–2019	1996–2001	2010–2013
Country	USA	USA	USA and India
Mean age ± SD (years)	60 ± 14	66 ± 10	60 ± 11
Male	36 (92%)	20 (95%)	7 (87.5%)
Ejection fraction %, mean ± SD	24 ± 11	22 ± 10	25 ± 7
LV Thrombus diagnosis
New diagnosis	25 (64%)	—	—
Chronic	14 (26%)	—	—
Risk factors
Hypertension	23 (60%)	—	—
Diabetes	11 (28%)	—	—
Hyperlipidemia	24 (61%)	—	—
Coronary artery disease	24 (61%)	16 (76%)	—
Stroke	7 (18%)	—	—
Atrial fibrillation/flutter	19 (49%)	—	—
Chronic kidney disease	9 (23%)	—	—
Non-ischemic cardiomyopathy	—	5 (24%)	—
Congestive heart failure	—	19 (90%)	—
Previous embolism/cerebrovascular accident	—	3 (14%)	—
Anticoagulation/antiplatelets use prior to cardioversion
Use of adjuvant antiplatelets	17 (44%)	—	—
Aspirin	—	11 (52%)	—
Warfarin	29 (74%)	—	—
Direct oral anticoagulants	2 (5%)	—	—
Heparin/warfarin	—	17 (81%)	8 (100%)
None	8 (21%)	—	—
Indication for cardioversion
Ventricular tachycardia	14 (36%)	13 (62%)	—
Atrial fibrillation/atrial flutter	14 (36%)	8 (38%)	—
Defibrillation threshold testing	11 (28%)	—	—
Intra-procedural VT-ablation instability	—	—	8 (100%)
Cardioversion urgency
Emergent	—	5 (24%)	—
Elective	—	16 (76%)	—
Description of thrombi
Fixed	36 (92%)	—	—
Mobile	3 (8%)	—	—
Laminated	—	15 (71%)	8 (100%)
Protruding	—	6 (29%)	—
Size of Thrombi	—	0.7 ± 0.4 × 1.6 ± 0.8 cm	443 ± 233 mm2 (area)
Apical LV thrombus location	—	21 (100%)	8 (100%)
LV thrombus diagnosis-to-cardioversion interval	6 ± 6.3 days	6 ± 5 (range 1–18 days)	350 ± 764 days
Follow-up	81 ± 17 days	153 ± 150 days	810 ± 318 days
Thrombo-embolic events	0	0	1

Across the three studies, 68 patients with a known LV thrombus either underwent direct-current cardioversion (60 patients^6,8) or received internal electrical shocks (8 patients ⁷ ). Indications for cardioversion/shock therapy were predominantly ventricular arrhythmias (including ventricular tachycardia and electrical storm with implantable cardioverter-defibrillator therapies) and AF /atrial flutter. Thrombus morphology was variably reported: in Dabbagh et al. LV thrombus was described as fixed in 92% and mobile in 8% of cases, ⁸ whereas Bangalore et al. reported 71% laminated and 29% protruding thrombi. ⁶ Rao et al. reported on patients with laminated LV thrombus presenting with electrical storm and ICD shocks who underwent VT catheter ablation and received cardioversions during the procedure. ⁷

Anticoagulation duration was incompletely reported across studies. Bangalore et al. and Dabbagh et al. described antithrombotic use before cardioversion but did not provide patient-level treatment duration, whereas Rao et al. reported that the patient who later developed ischemic stroke had been receiving warfarin for 1 month before ablation; however, therapeutic adequacy immediately before the procedure was not reported. Of note, up to 21% of patients in the Dabbagh et al. cohort had no anticoagulation prior to cardioversion. ⁸ Follow-up durations differed between studies (Dabbagh et al.: 81 ± 17 days ⁸ ; Bangalore et al.: 153 ± 150 days ⁶ ; Rao et al.: clinical follow-up reported to 810 ± 318 days for surviving patients ⁷ ).

Thromboembolic outcomes were grossly non-existent in these series. Neither Dabbagh et al. nor Bangalore et al. observed any thromboembolic complications (clinically apparent stroke, transient ischemic attack, myocardial infarction, systemic embolism, or other objective embolic events) during their respective follow-up periods.^6,8 In the Rao cohort, ⁷ one patient experienced an ischemic stroke with right-sided hemiplegia and global aphasia on post-VT ablation day 9; however, this patient was also found to have a LAA thrombus, representing a likely potential alternative source for the event. At the 3-month follow-up, he had partially recovered but had persistent residual aphasia. No additional thromboembolic events were reported among the remaining Rao patients during >1 year of follow-up.

All three studies reported rare mortality events that were largely unrelated to acute thromboembolic complications. Dabbagh et al. observed no deaths. ⁸ Bangalore et al. reported two deaths: one patient died of septic shock on day 1 after cardioversion and a second patient died of cardiogenic shock three months after cardioversion. ⁶ Rao et al. reported two deaths during longer follow-up: one patient underwent orthotopic heart transplant 10 days after ablation for refractory VT and subsequently died 286 days after transplantation from rejection, and another patient died of progressive heart failure 373 days after ablation. ⁷

Discussion

There is clinician hesitancy with performing electrical cardioversion in the presence of a left ventricular thrombus. This is likely due to the fact that, in contrast to AF in the presence of an LAA thrombus, ² the magnitude and nature of the thromboembolic risk posed by a LV thrombus at the time of electrical cardioversion are poorly defined. It could be hypothesized that cardioversion in those with left ventricular thrombi is associated with the risk of embolization from the left ventricle, which can cause catastrophic cerebral or peripheral ischemia. ⁴

In our qualitative synthesis of three retrospective series spanning different eras and clinical settings, we found no temporally related, clinically apparent thromboembolic events after direct-current cardioversion or internal shocks in patients with documented LV thrombus. Across the pooled cohort of 68 patients (60 who underwent external cardioversion^6,8 and 8 who received internal shocks during VT procedures ⁷ ), there were no thromboembolic complications directly attributable to the procedure. The single ischemic stroke reported in the pooled studies occurred on post-procedure day 9 in a patient who was also found to have a LAA thrombus, ⁷ making causal attribution to the LV thrombus or the cardioversion itself unlikely. Importantly, even the subsets of patients with reported mobile thrombi and those not receiving any anticoagulation prior to cardioversion did not experience any clinically evident thromboembolic events in the reviewed series ⁸ ; however, these subgroup sizes were small and must be interpreted cautiously. Overall, these concordant findings emerged across differing clinical scenarios (AF and ventricular arrhythmias), procedural contexts (external cardioversion and intra-procedural/internal shocks), and varying anticoagulation practices, suggesting that electrical cardioversion in the presence of an LV thrombus, particularly when patients are anticoagulated, may indeed carry little-to-no risk of clinically overt thromboembolism than commonly presumed.

What should clinicians take from these data? First and foremost, the evidence favors a nuanced, individualized approach rather than a routine deferral of cardioversion in the presence of LV thrombus. For patients with laminated, non-mobile LV thrombus who are therapeutically anticoagulated, proceeding with cardioversion, or performing urgent internal shocks during VT ablation can be a reasonably performed with a low risk for thromboembolic complications. In these situations, clinicians should optimize and document anticoagulation, confirm thrombus morphology with the available imaging (contrast transthoracic echo, transesophageal echo, or cardiac MRI as indicated), and discuss the risk–benefit balance with the patient (or surrogate) as part of informed consent. Similarly, in situations where urgent cardioversion is required in the presence of a mobile LV thrombus, proceeding with the procedure on anticoagulation after risk–benefit discussion is a viable option when the anticipated clinical benefit (restoration of sinus rhythm, improvement in hemodynamics, termination of electrical storm, or meaningful symptom relief) clearly outweighs the theoretical thromboembolic risk.

Future research should focus on multicenter prospective registries that capture thrombus morphology, imaging modality, timing and adequacy of anticoagulation, cardioversion urgency, and both clinical and subclinical embolic outcomes. Of note, although contrast-enhanced transthoracic echocardiography and cardiac MRI remain the preferred modalities for definitive LV thrombus characterization, the expanding use of cardiac point-of-care ultrasound may facilitate earlier bedside recognition of major ventricular structural abnormalities and prompt timely confirmatory imaging when LV thrombus is suspected.^10,11

Our review has important limitations that constrain the confidence and generalizability of its findings. Only three retrospective studies met our inclusion criteria, yielding small and heterogeneous cohorts that are vulnerable to selection bias; centers are likely to avoid cardioversion when high-risk, mobile thrombi are seen, producing a study population enriched for lower-risk, laminated thrombi. Consequently, the absence of clinically apparent embolic events in these series may reflect careful patient selection and limited statistical power rather than a true absence of procedural risk. Importantly, silent cerebral infarction has been documented after cardioversion and other catheter procedures even in the absence of clinical stroke, ¹² and because routine diffusion-weighted MRI or systematic neurologic follow-up was not performed in these retrospective cohorts, subclinical brain injury cannot be excluded. In addition to potentially missed studies and only two database queries, the small number of studies and sample sizes also precluded a meta-analysis as originally planned, denying us greater precision and statistical power. Finally, because randomized trials that deliberately expose patients with known LV thrombus to cardioversion may likely be ethically challenging, the evidence base will likely remain observational and uncertain; in this context, as previously noted by Dabbagh et al., ⁸ we must recognize that even a single thromboembolic catastrophe would be devastating for the patient and could carry disproportionate medicolegal consequences for treating clinicians, and that reassurance from small retrospective series may not fully mitigate that risk.

Conclusion

Limited and available retrospective data suggest that electrical cardioversion in the presence of an LV thrombus has not been associated with a measurable increase in clinically overt thromboembolic events. Clinical decisions should, however, be individualized, supported by clinical and hemodynamic data, detailed imaging, and periprocedural anticoagulation.

Supplemental Material