We conducted a systematic review and meta-analysis of studies on the effects of DOACs on LVT and outcomes in comparison with VKAs. The review was registered with PROSPERO (CRD42021248170) and reported according to PRISMA guidelines.

The quality appraisal was established according to Le Floch et al. criteria, by two independent assessors (HF and JA).9 This tool appraises the quality of the study based on the following questions: Did this article give an answer to the research question? Did the article focus clearly on the research question? Was the methodology appropriate? Do you believe the results? (Can it be due to chance, bias, or confounding?). To be included, the article had to score “yes” on every question.

Quantitative meta-analysis was performed for an outcome when more than one study presented relevant data. We excluded individual outcomes from studies reporting no adverse outcomes in one or both groups, and studies not satisfying the normality assumption for continuous variables.

A random-effects estimate of the pooled odds of each outcome was generated with the use of the Mantel-Haenszel method. Between-study heterogeneity was explored using the I2 statistic, with substantial heterogeneity defined as an I2 value greater than 50%. We reported p values and the amount of accounted heterogeneity for each covariate. Potential publication bias was assessed with Egger's test and funnel plots for visual inspection when sufficient studies (n>10) were available. Statistical analyses were performed using the Revman software package (Review Manager, Version 5.4. Copenhagen, The Nordic Cochrane Center, the Cochrane Collaboration).

The literature retrieval process from this meta-analysis is presented in Figure 1. A total of 368 papers were initially identified, of which 279 records nonrelated to the theme were excluded. After review of the 38 remaining studies, 24 were excluded with reasons: systematic reviews and meta-analysis (n 9); case reports and case series (n 7); five review articles (n 5); two single-arm studies (2); one trial design (1). Finally, a total of 14 studies that were published between 2018 and 2021 involving 2498 patients with left ventricular thrombus (n=631 for DOACs and n=1867 for VKAs) were included in this meta-analysis.10–23 Of the 14 studies, only one was prospective (23); and nine retrospective, observational cohort studies with the full text.10–13,15,16,18,20,21 The other four studies were published as a conference abstract.14,17,19,22

Figure 1.

Forest plot of efficacy outcomes with DOACs versus VKAs. DOACs: direct oral anticoagulants; VKAs: vitamin K antagonists.

(1.02MB).

The baseline characteristics of the included studies are shown in Table 1. Overall, there was a predominance of males, constituting 74–90% of cases. The patients’ mean age ranged from 56 to 63±11. The most common cause for LVT was ischemic cardiomyopathy and the main comorbidities were hypertension and diabetes mellitus, followed by dyslipidemia. History of smoking was reported from 15 to 59% of cases. The average follow-up period ranged from 3 to 36 months. Nine studies described the type of DOACs, totaling 407 of the 631 patients. Apixaban was used in 170 (41.7%) patients and rivaroxaban in 200 (49.1%) patients. Dabigatran and edoxaban were used in 35 (8.5%) and 2 (0.4%) patients, respectively. Triple therapy was reported in five studies, 214 patients (8.5% of the total population).

The outcomes analyzed were efficacy and safety of oral anticoagulation (DOACs vs. VAKs) in the treatment of left ventricular thrombus LVT. Thirteen studies involved the outcome of stroke and/or systemic embolism10–13,15–23; twelve studies mentioned the thrombus resolution,10,12–18,20–23 but only four studies involved the outcome of all-cause mortality.12,16,20,22 Eleven studies reported bleeding events.10–12,15–18,20–23

Effect of NOACs versus that of VKAs and, on stroke or systemic embolism A total of 13 included studies reported the stroke or systemic embolism event rate.10–13,15–23 As shown in (Figure 1), fewer thromboembolic (stroke or systemic embolism) events were reposted in the DOACs-treated patients compared with those receiving VKA (109/618 vs. 386/1814 patients (OR 0.86; 95% CI, 0.55–1.33; p=0.50; I2=32%)), however, it was statistically non-significant. For ischemic stroke, there were no significant differences among the groups, (46/571) DOACs vs. (175/1737) VKA patients (OR 0.86; 95% CI, 0.51–1.42; p=0.56; I2=27%). Systemic embolism rate was (50/320) DOACs vs. (64/1008) VKA patients (OR 0.71; 95% CI, 0.13–3.58; p=0.70; I2=80%).

A total of 12 included studies (n=1474) reported the rate of thrombus resolution.10,12–18,20–23 A total of 1474 patients had imaging follow-up. The vast majority of studies relied on transthoracic echocardiogram (TTE) to evaluate left ventricular thrombus (LVT) resolution. Jones et al. used TTE or cardiac magnetic resonance (CMR),18 and Iqbal et al. used CMR to diagnose LVT at the baseline with subsequent assessments made by TTE.16 LVT resolution was observed in 288/416 in DOACs vs. 732/1085 patients treated with VKA (OR 1.14; 95% CI, 0.77–1.66; p=0.50; I2=33%) (Figure 1).

All-cause mortality was reported by only four studies.12,16,20,22 No significant differences were noted among the groups; all-cause mortality occurred in 30/221 DOAC vs. 82/557 VKA patients (OR 0.92; 95% CI, 0.58–1.45; p=0.74; I2=0%) (Figure 1). Overall, there were no significant differences regarding the efficacy outcomes (OR 0.97; 95% CI, 0.74–1.28; p=0.87; I2=57%) (test for subgroup differences: Chi2=1.34, df=4 (p=0.85), I2=0%).

Eleven of the included studies assessed the effect of DOACs versus that of VKAs on bleeding events using varying criteria to identify and record bleeding incidents.10–12,15–18,20–23 In the study by Guddeti et al., bleeding events were defined as any life-threatening bleeding, drop in hemoglobin ≥2 g, and/or bleeding requiring hospitalization or evaluation by endoscopy.15 Bass et al., defined bleeding events using the Global Use of Strategies to Open Occluded Arteries (GUSTO) criteria, and blood product administration.11 In turn, Iqbal et al. used clinically relevant bleeding and observed three episodes of gastrointestinal bleeding and three epistaxis events none requiring hospitalization or blood transfusion.16 Jones et al. defined bleeding events using the Bleeding Academic Research Consortium (BARC) criteria, and they presented the data of minor bleeding events (BARC 1) and major bleeding events (BARC ≥2).18 Robinson et al. reported bleeding events requiring cessation in anticoagulation;20 Jaidika et al. classified bleeding as minor and major;17 Cochran et al. included minimal, minor, and major bleeding assessed by Thrombolysis in Myocardial Infarction bleeding criteria,12 while no criteria were specified by Ali et al., Willeford et al., and Yunis et al.10,21,22

Any bleeding events were observed in 8.7% of DOACs (46/527) and 8.3% of VKA patients (142/1692) (OR 0.96, 95% CI 0.62–1.48, p=0.88; I2=17%). Clinically relevant bleeding events were reported by 11 of the 14 studies and were observed in 7.01% of DOACs (40/566) and 7.15% of VKA patients (124/1732) (OR 1.09, 95% CI 0.73–1.62, p=0.66; I2=3%). Data for minor bleeding were available in three of 14 studies. Minor bleeding events were reported at a pooled rate of 11% (8/75) in DOAC-treated patients compared with 15% in VKA-treated patients (24/159) (OR 0.59, 95% CI 0.25–1.40, p=0.23; I2=0%). Overall, there were no significant differences regarding the safety outcomes (OR 1.0; 95% CI, 0.78–1.30; p=0.93; I2=2%) (test for subgroup differences: Chi2=1.60, df=2 (p=0.45), I2=0%) (Figure 2).

Figure 2.

Forest plot of safety outcomes with DOACs versus VKAs. DOACs: direct oral anticoagulants; VKAs: vitamin K antagonists.

(0.87MB).

Publication bias assessed using funnel plots is shown in Figures 3 and 4 (supplementary material).