Abstract
Persistent left atrial appendage thrombus despite anticoagulation is common. Current guidelines advise against percutaneous mitral balloon valvuloplasty (PMBV) because of perceived thromboembolic risk, leaving surgical mitral valve replacement—an option associated with substantial morbidity—as the primary alternative. However, the procedural risk of PMBV in the setting of left atrial appendage thrombus remains poorly established. We conducted a systematic review and meta-analysis to assess this. Of the 2,136 studies identified in the initial search, 17 were included in the analysis, comprising 386 patients undergoing PMBV with left atrial appendage (LAA) thrombus. The rate of stroke or embolic complication was 2.8% (95% CI 1.4% to 5.5%). No stroke or embolic complications occurred in Type Ia thrombus. In conclusion, LAA-confined thrombus (Type 1a) was not associated with embolic events, supporting a morphology-based approach to patient selection and emphasizing the need for prospective data to refine the current guidelines.
Percutaneous mitral balloon valvuloplasty (PMBV) is an established treatment for symptomatic severe rheumatic mitral stenosis (MS) in patients with pliable leaflets, <moderate mitral regurgitation, and absence of left atrial thrombus. Although no randomized trials have compared PMBV with surgical mitral valve replacement (MVR), PMBV has been widely adopted owing to its safety, efficacy, and minimally invasive nature. When optimal results are achieved, long-term durability post PMBV is favorable, with reintervention-free survival of 80% at 10 years and 30% at 20 years. Left atrial appendage (LAA) thrombus is common in rheumatic MS, occurring in up to 30% with atrial fibrillation and 14% in sinus rhythm. Despite guideline-directed anticoagulation, 20% to 47% of thrombi fail to resolve. , Thrombus morphology influences embolic risk, with shape, mobility, and protrusion being key determinants ( Figure 1 ). Type IA thrombus—fully confined within the appendage without protrusion—is the most common subtype and carries the lowest embolic potential. Current guidelines advise against PMBV in the presence of left atrial thrombus, without accounting for thrombus morphology or quantifying a procedural risk. Consequently, patients with persistent thrombus are typically referred for surgical MVR, which carries a notable risk. To address this knowledge gap, we conducted a systematic review and meta-analysis of published studies evaluating the procedural safety and short-term outcomes of PMBV in patients with persistent LAA thrombus.
Types of left atrial thrombus. Type Ia thrombus is confined to the LAA. Type Ib thrombus protrudes out from the LAA. Type IIa thrombus is adherent to the roof of the LA without crossing the fossa ovalis. Type IIb thrombus is adherent to the roof of the LA extending past the fossa ovalis. Type III thrombus is layered clot over the interatrial septum. Type IV thrombus is mobile clot attached to the LA free wall, roof, or interatrial septum. Type V LA thrombus is free-floating.
Methods and Material
Study protocol and search strategy
Two independent reviewers (PD, MM) conducted a comprehensive search of Ovid MEDLINE, EMBASE, PubMed, and EBM Reviews from database inception through November 2025. Search terms included combinations of mitral stenosis, balloon valvotomy, percutaneous mitral valvuloplasty, left atrial thrombus, and related MeSH keywords. Conference abstracts and unpublished data indexed in the searched databases were also included. The systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
Inclusion and exclusion criteria
Studies were eligible for inclusion if they met all of the following criteria: (1) observational design (prospective or retrospective cohorts, comparative studies, single-arm studies, or case series), case reports, or randomized trials evaluating PMBV; (2) adult patients (≥18 years) with rheumatic MS; (3) documented persistent LA or LAA thrombus at the time of PMBV; and (4) reporting of procedural outcomes, patient characteristics, or follow-up clinical events. Case reports were included in the descriptive synthesis but excluded from quantitative meta-analysis due to the absence of denominators and inherent publication bias.
Studies were excluded if: (1) the thrombus had fully resolved with anticoagulation before PMBV; (2) patients did not have confirmed LA or LAA thrombus; or (3) insufficient data were available to extract outcomes of interest. Conference abstracts were included only if they provided adequate extractable data.
Data collection and quality assessment
Two reviewers (MM and PD) independently screened titles and abstracts, followed by full-text evaluation. Data extraction was performed in duplicate using a standardized data collection form. Extracted variables included: study design and setting, sample size, patient demographics, thrombus location, type if specified, and morphology, procedural approach (Inoue vs over-the-wire), hemodynamic outcomes (pre and post-PMBV gradients and mitral valve area), postprocedural complications (death, embolic events, bleeding)
Endpoint definitions
Primary endpoints: All-cause mortality; stroke or systemic embolism, need for emergency mitral valve surgery, or cardiac tamponade within the hospital admission.
Secondary outcomes included procedural success, changes in mitral valve area, and changes in mean mitral gradient, when reported.
Risk of bias and quality of evidence
Case series were assessed using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Case Series, evaluating inclusion clarity, diagnostic reliability, completeness of follow-up, and reporting quality. Studies were classified as low, moderate, or high risk of bias.
( Supplementary Figure : risk of bias)
Data synthesis and statistical analysis
Quantitative synthesis was performed using Comprehensive Meta-Analysis (CMA) Version 4 (Biostat, Englewood, NJ). Pooled event rates for mortality, stroke, and systemic embolism were calculated using a random-effects model with corresponding 95% confidence intervals (CIs), given expected heterogeneity across study designs and populations. Case reports were included in the descriptive synthesis but excluded from quantitative meta-analysis.
Results
Studies and patients characteristics
From 2,312 studies identified in the initial search, 17 studies (10 case series and 7 case reports) ,,,,,,,,,,,,,,,, met the inclusion criteria and were included in the analysis. Across these reports, 386 patients underwent PMBV in the presence of LA thrombus, 90% confined within the left atrial appendage and 10% involving the left atrium. The weighted mean age was 40 ± 12.0 years, and 58% were female. The thrombus was uniformly defined on preprocedural transesophageal echocardiography (TEE). The duration between thrombus diagnosis and PMBV was inconsistently reported. Only 6 studies provided detailed information regarding anticoagulation (warfarin or acenocoumarol) following LAA thrombus diagnosis, including duration of therapy and reported changes in thrombus characteristics. ,,,,,,, In studies detailing intraprocedural management, anticoagulation consisted of unfractionated heparin, administered either at unspecified dose (1 study ), weight-based protocol (1 study ) or as a fixed dose ranging from 3,000 to 5,000 units (3 studies ,, ). No studies reported the use of alternative agents, such as bivalirudin, or the continuation of warfarin during the procedure.
Among the 10 case series (n = 386), all reported in-hospital outcomes, and 3 ,, provided follow-up data extending beyond discharge (n = 120). One case series included a comparative cohort, evaluating outcomes of patients with LAA thrombus treated with PMBV versus surgical mitral valve intervention ( Table 1 ).
Table 1
Studies characteristics
| Total number of studies | 17 |
|---|---|
| Total number of patients | 386 |
| Age (years), mean ± SD | 40.1 ± 12.0 |
| Female (%) | 58 |
| LAA thrombus types (%) (n = 288) | |
| IA | 90.0 |
| IIA | 2.4 |
| IB | 7.4 |
| LAA thrombus size (mm) (n = 5), mean ± SD | 24.5 ± 4.5 × 1.31 ± 25.7 |
| PMBV technique (%) | |
| Inoue | 71% |
| Over-the-wire | 28% |
| Follow-up | |
| In hospital | 10 |
| 30-days | 2 |
| 6 mo | 1 |
| 12 mo | 1 |
LAA = left atrial appendage thrombus;PBMV = percutaneous balloon mitral valvuloplasty; MG = mean gradient; MR = mitral regurgitation; MVA = mitral valve area; SD = standard deviation.
LAA thrombus characteristics
All included studies either reported thrombus classification directly or provided sufficient echocardiographic detail to allow assignment of thrombus morphology. Type IA thrombus—confined entirely to the appendage—was the most common, representing 90% of cases. Type IB accounted for 7.5%, and Type IIA for 2.5% ( Figure 1 ). Thrombus size was reported in 5 studies, ranging from 18 × 8 to 35 × 28 mm. ,,,, All studies utilized transesophageal echocardiography (TEE) as the diagnostic modality for thrombus characterization.
PBMV technique
PMBV was performed in 395 patients. The classic Inoue balloon technique was used in 71% (N: 275 patients) of reported cases, an over-the-wire (OTW) technique in 28% (n: 108 patients), and a modified wiretracking technique with veno-arterial loop in 1 case report (n: 1 patient). Active intraprocedural TEE guidance was used in 59% of cases (n: 229). One study described the use of intraprocedural ICE guidance in 7 patients. Additionally, 1 study described the use of intraprocedural transthoracic echocardiography. Tables 2 and 3 summarize individual study characteristics and procedural outcomes.
Table 2
Summary of case series of patients with symptomatic rheumatic MS and concomitant LA thrombus who underwent PMBV
| MVG (mmHg) | MVA (cm 2) | LA thrombus | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Study author and year | Patients (N) | Age (years) | Female (%) | Pre | Post | Pre | Post | Type | Size (mm) | PMBV technique | Follow-up | Reported outcomes |
| Chen (1992) | 6 | 50 ± 5 | 83 | 15 ± 4 | 7 ± 4 | 0.85 ± 0.32 | 1.77 ± 0.23 | IA | <2.5 cm | Inoue | 2 to 3 weeks | Mortality, emergency surgency, cardiac tamponade, thromboembolism, and mitral regurgitation |
| Tessier (1994) | 7 | 56 ± 12 | 86 | 10 ± 5 | 7 ± 2 | 0.90 ± 0.20 | 1.8 ± 0.3 | IA | 17.9 ± 8.4 × 6.56 ± 1.8 mm | OTW | 12.4 ± 11.9 mo | Embolic event, peripheral emboli |
| Yeh (1995) | 49 | 46 | – | – | – | – | – | – | Inoue | Immediate | Embolism | |
| 7 | IB | |||||||||||
| 42 | IA | |||||||||||
| Koca (2000) | 9 | 40 ± 9 | 88 | 14 ± 4 | 4 ± 2 | 0.97 ± 0.22 | 1.94 ± 0.27 | IA | – | Inoue | Immediate | Embolism |
| Tansuphaswadiku (2001) | 111 | 44 ± 10 | 67 | 19 ± 7 | 12 ± 6 | 0.80 ± 0.21 | 1.47 ± 0.31 | IA | 28 × 35 | Inoue | Immediate | Mortality, cardiac tamponade, mitral regurgitation, and TIA/CVA |
| Kandpal (2002) | 38 | 33 ± 10 | 62 | 17 ± 5 | 7 ± 3 | 0.80 ± 0.16 | 1.65 ± 0.17 | 27.6 ± 9.1 mm | Inoue | Immediate | Embolism | |
| 30 | IA | |||||||||||
| 8 | IB | |||||||||||
| Shaw (2005) | 32 | 65 ± 11 | 76 | 12 ± 4 | 7 ± 1 | 0.78 ± 0.23 | – | Inoue | Immediate | – | ||
| 24 | IA | |||||||||||
| 6 | IIA | |||||||||||
| 2 | IB | |||||||||||
| Manjunath (2009) | 106 | 27 ± 8 | 58 | 18 ± 4 | – | 0.80 ± 0.20 | 1.8 ± 0.2 | IA, Ib, IIa | 20 ± 6 × 11 ± 4 mm | OTW | 1 mo | CVA (stroke and TIA), peripheral embolic episodes, mod/severe MR, MR requiring surgery, cardiac tamponade, in hospital death, and suboptimal results |
| Rajbhandari (2016) | 20 | 31 ± 9 | 65 | – | – | 0.90 ± 0.24 | 1.5 ± 0.21 | IA | – | Inoue | Immediate and 1 week | Mortality, mitral regurgitation, CVA, and emergency surgery |
| Moey (2024) | 7 | 63 ± 2.7 | 86 | 10 ± 5 | 6 ± 2 | – | – | IA | – | Inoue | Outpatient 30days | Mortality, Stroke, emergency surgency, thromboembolism,and mitral regurgitation |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
