Abbreviations ACS acute coronary syndrome CI confidence interval CK creatine kinase CMR cardiac magnetic resonance tomography DAPT dual antiplatelet therapy DAT dual antithrombotic therapy HR hazard ratio IQR interquartile range LAT left atrial thrombus LVEF left ventricular ejection fraction LVT left ventricular thrombus MACE major adverse cardiac events MI myocardial infarction NOACS non-vitamin-K antagonists NSTEMI non-ST-elevation myocardial infarction OAC oral anticoagulation STEMI ST-elevation myocardial infarction TAT triple antithrombotic therapy TTE transthoracic echocardiography VKA vitamin K antagonists 1 INTRODUCTION

Left ventricular thrombus (LVT) is a rare complication after acute coronary syndrome (ACS), especially occurring in patients presenting late with ST-elevation myocardial infarction (STEMI).1 Incidence rates differ among the observational studies from 1.6% up to 39% indicating that many LVT cases might remain undetected.2-7 This substantial variation in the incidence rate, is caused by varieties in the imaging modality used for diagnosis and the timing and frequency of screening. Additionally, the use of modern revascularization therapies has reduced the occurrence of LV-thrombus formation.2-7 While the prognosis of patients presenting with LVT after ACS has been controversially discussed, it seems intuitive that individuals without thrombus resolution have an increased risk for cardiovascular events and mortality. Thrombus formation is significantly associated with anterior myocardial infarction (MI) and confers an increased risk for thromboembolic events (mostly cerebrovascular).8-10 In the pre-thrombolytic era, these complications were described in approximately 10% of cases, whereas in the era of thrombolytic therapy embolic events occurred in 2%–3% of cases.8-10 Until now there is scarce evidence regarding the incidence of embolic events in patients treated by primary percutaneous coronary intervention (PCI) that receive dual antiplatelet therapy (DAPT) or even triple antithrombotic therapy (TAT; oral anticoagulation [OAC] plus DAPT) or dual antithrombotic therapy (DAT) as the combination of OAC with only one antiplatelet agent. However, optimal pharmacological therapy—used to reduce complications of LVT—remains challenging. While patients post MI requires DAPT for reduction of atherothrombotic risk, they also need OAC in case of LVT formation for reduction of related complications, with subsequent high risk of bleeding.11 For patients with large anterior STEMI, DAPT with a potent P2Y12 inhibitor may even be an attractive option, as potent DAPT has been shown to contribute to a lower incidence of LVT.12

Considering a strong impact of LVT on patient outcome and the notion that many LVTs remain undetected in clinical practice, patient characteristics that help to identify ACS individuals at risk for the development of LVT with an adverse outcome should be considered in terms of a personalized secondary prevention. However, profound data on long-term outcome of this highly vulnerable patient population are scarce in current literature. Therefore, we aimed to investigate the impact of LVT resolution and associated antithrombotic treatment strategies on patient's outcome from a long-term perspective.

2 METHODS 2.1 Study population and patient selection

Patients presenting with ACS (n = 2011) who underwent treatment at the Vienna General Hospital, a university affiliated tertiary care center with a high-volume cardiac catheterization unit in the time period between January 2015 and September 2019 were screened for presence of LVT. Out of the source population a total of 52 patients (2.6%) presented with LVT after MI. Six individuals died before hospital discharge and three did not receive follow-up imaging and were subsequently excluded for the final analysis—resulting in a total study population of 43 patients for the present long-term analysis. All patients were older than 18 years. The study was conducted in accordance to the current criteria of the Declaration of Helsinki and was approved by the ethics committee of the Medical University of Vienna (1702/2019). Based on the study design the investigation was conducted without informed consent due to the minimal risk of study inclusion. After completion of follow-up, the study population was stratified in patients with thrombus resolution and individuals without thrombus resolution.

2.2 Data acquisition and patient follow-up

Patient-relevant characteristics were assessed via the patients' electronic medical records of the Vienna General Hospital, as well during a standardized follow-up procedure. Data assessment was performed by specially trained chart reviewers that inserted predefined patient characteristics into a record abstraction form for further analysis of the registry at the time of hospitalization and re-evaluation during the entire hospitalization. Discharge letters of all participants were screened for the antithrombotic treatment approach at the time of discharge.

Patients were invited to the local department for screening of thrombus resolution. The presence of LVT was validated by contrast transthoracic echocardiography (TTE) and/or cardiac magnetic resonance tomography (CMR) of all enrolled individuals. Clinically relevant data, including the antithrombotic treatment and adherence to medication was assessed during the follow-up visit. Thrombus resolution was defined as lack of evidence of thrombus mass in follow-up contrast TTE or CMR.

2.3 Endpoint definition

All-cause death (including death from cardiovascular, renal, and cancer disease) was chosen as primary study endpoint. The composite of major adverse cardiac events (MACE), defined as a nonfatal MI, nonfatal stroke, and cardiovascular death as well as thromboembolic events were chosen as secondary endpoints and assessed during follow-up. The patients' cause and date of death was assessed by screening the national registry of death until December 2019 via the Austrian Registry of Death (Statistics Austria, Vienna, Austria). Causes of death were defined according to the International Statistical Classification of Disease and Related Health Problems 10th Revision.

2.4 Statistical analysis

Continuous data are presented as median and the respective interquartile range and analyzed using Mann Whitney U test. Categorical parameters are presented as counts and percentages and analyzed using Chi-square test. Univariate and multivariate Cox proportional hazard models were applied to assess the influence of thrombus resolution on primary and secondary endpoints and to assess the impact of a TAT with newer P2Y12 antagonists on LVT resolution. Results were presented as hazard ratio (HR) and the respective 95% confidence interval (CI). A three-step adjustment approach was followed within the multivariate regression model including comprehensive adjustment for patient characteristics (= Model 1: age and sex), clinical presentation (= Model 2: STEMI and heart failure), and laboratory values (= Model 3: NT-proBNP and creatine kinase [CK]).

Continuous variables were log-transformed prior to inclusion in the regression analysis. Kaplan Maier charts were plotted to graphically illustrate the impact of LVT resolution on all-cause death, MACE and thromboembolic events and compared using log-rank test. Statistical significance was defined by two-sided p-values <.05. Statistical analyses were performed using SPSS 26.0 (IBM SPSS, NY, USA).

3 RESULTS

Detailed baseline characteristics for the study population presenting with LVT (n = 43), stratified in individuals with thrombus resolution and without thrombus resolution are summarized in Table 1.

TABLE 1. Baseline characteristics Overall (n = 43) No resolution (n = 16) Resolution (n = 27) p-value Clinical presentation Age, years (IQR) 63 (58–69) 68 (61–72) 62 (56–67) .074 Gender (male), n (%) 38 (88.4) 14 (87.5) 24 (88.9) .892 BMI, kg/m2 (IQR) 26.2 (24.2–29.5) 26.1 (24.1–29.2) 26.3 (24.2–30.2) .756 PCI, n (%) 33 (76.7) 10 (62.5) 23 (85.2) .093 STEMI, n (%) 26 (60.5) 8 (50.0) 18 (66.7) .289 Anterior wall infarction, n (%) 43 (100) 16 (100) 27 (100) 1.000 Systolic BP, mmHg (IQR) 135.0 (115.8-149.3) 135.5 (109.8–148.3) 135.0 (116.0–150.0) .836 Diastolic BP, mmHg (IQR) 74.0 (68.0–90.0) 74.0 (66.5–89.0) 74.0 (69.5–90.0) .785 LVEF <50% (%) 21 (48.8) 7 (43.8) 14 (51.9) .612 Comorbidities Diabetes, n (%) 7 (16.3) 2 (12.5) 5 (18.5) .610 Hyperlipidemia, n (%) 21 (48.8) 6 (37.5) 15 (55.6) .258 Hypertension, n (%) 31 (72.1) 12 (75.0) 19 (70.4) .746 Nicotine, n (%) 22 (51.2) 6 (37.5) 16 (59.3) .173 Prior MI. n (%) 15 (34.9) 5 (31.3) 10 (37.0) .704 Medication Beta blockers, n (%) 41 (95.3) 16 (100) 25 (92.6) .271 ACEI, n (%) 31 (72.1) 12 (75.0) 19 (70.4) .746 ATI, n (%) 8 (18.6) 1 (6.3) 7 (25.9) .113 Antithrombotic treatment approach Acetylsalicylic acid, n (%) 43 (100) 16 (100) 27 (100) 1.000 Clopidogrel, n (%) 22 (51.2) 9 (56.3) 13 (48.1) .612 Ticagrelor/prasugrel, n (%) 8 (7.0) 0 (−) 8 (29.6) .016 DAT, n (%) 13 (30.2) 7 (43.8) 6 (22.2) .137 TAT, n (%) 30 (69.8) 9 (56.3) 21 (77.8) .137 TAT with clopidogrel, n(%) 22 (73.3) 9 (100) 13 (61.9) .031 Duration of clopidogrel (months), median (IQR) 12 (12–12) 12 (12–12) 12 (12–12) 1.000 TAT with ticagrelor/prasugrel, n (%) 8 (26.7) 0 (−) 8 (38.1) .031 Duration of ticagrelor/prasugrel (months), median (IQR) 12 (12–12) – 12 (12–12) NA VKA, n (%) 33 (76.7) 13 (81.3) 20 (74.1) .595 NOAC, n (%) 10 (23.3) 3 (18.8) 7 (25.9) .595 Duration of OAK (months), median (IQR) 24 (12–72) 24 (18–64) 12 (12–84) .780 Laboratory variables NTproBNP, pg/ml (IQR) 1526.0 (609.3–7012.3) 2945.0 (1040.3–20355.0) 1308.0 (401.8–4284.8) .089 Troponin T max., μg/ml median (IQR) 2339.5 (224.5–5828.0) 1324.5 (44.5–3583.0) 3502.0 (548.8–7455.5) .431 CK, U/I, median (IQR) 986.0 (119.5–2906.5) 222.0 (68.0–1545.0) 1355.0 (281.5–3431.3) .066 CK-MB, U/I median (IQR) 191.5 (49.0–283.0) 77.0 (26.0–465.0) 200 (61.0–296.0) .455 Note: Categorical data are presented as counts and percentages and analyzed using Chi-square-test. Continuous data are presented as median and the respective interquartile range and analyzed using Mann Whitney U test. Abbreviations: ACEI, angiotensin converting enzyme inhibitor; AF, atrial fibrillation; ATI, angiotensin II receptor inhibitor; BP, blood pressure; BMI, body-mass index; CK, creatinine kinase; DAPT, dual antiplatelet therapy; DAT, dual antithrombotic therapy; INR, international normalized ratio; IQR, interquartile range; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NOAC, non-vitamin-K anticoagulant; NT-proBNP, N-terminal pro b-type natriuretic peptide; PCI, percutaneous coronary intervention; STEMI, ST-elevation myocardial infarction; TAT, triple antithrombotic therapy; VKA, vitamin-K antagonist.

In short, the present study population (median age: 63 years [IQR 58–69]; 88.4% male gender) covered a representative number of participants presenting with STEMI (n = 26; 60.5%) and all patients developed LVT after anterior wall infarction. The remaining 39.5% of patients presented with non-ST-elevation myocardial infarction (NSTEMI). In 97.7% of cases, LVT was diagnosed via TTE and in 2.3% via CMR. The median time of thrombus detection after the acute event was 5 days (IQR 3–15). The median time between thrombus detection and the first follow-up imaging was 14 weeks (IQR 6–22).

Comparing characteristics of patients with thrombus resolution and without thrombus resolution, we observed balanced frequencies of cardiovascular risk factors, such as diabetes mellitus (p = .610), hyperlipidemia (p = .258) and hypertension (p = .746). Established risk factors for the development of LVT—such as reduced left ventricular ejection fraction (LVEF) (48.8%), anterior wall infarction (100%), and elevated NT-proBNP values (median: 1526.0 [IQR 609.3–7012.3]) reflecting cardiac strain were observed with high frequencies within the study population. However, we did not observe any significant group differences. Detailed baseline echocardiographic parameters are shown in Table 2. In short, severe left ventricular dysfunction was common (median LVEF 36.0% [IQR 33.0–45.0]) with no differences between both groups (p = .908). Left ventricular aneurysms were found in 27.9% of patients (n = 12) without any group differences (p = .286). Median time of thrombus resolution was 14 weeks (IQR 6–23).

TABLE 2. Baseline echocardiographic parameters Overall (n = 43) No resolution (n = 16) Resolution (n = 27) p-value Left ventricular ejection fraction, % 36.0 (33.0–45.0) 39.0 (22.0–50.0) 36.0 (35.0–40.0) .908 End-diastolic left ventricular diameter, mm 49.0 (43.0–55.0) 49.0 (43.0–57.0) 48.0 (42.8–53.3) .586 End-diastolic right ventricular diameter, mm 33.0 (29.0–36.0) 35.0 (29.5–40.0) 31.0 (28.0–34.0) .059 Interventricular septum thickness, mm 13.0 (11.5–14.3) 13.0 (11.3–14.0) 13.0 (11.5–14.5) .617 Left ventricular thrombus Area, cm2 2.1 (0.9–4.2) 3.1 (0.8–5.0) 2.1 (1.6–4.1) .956 Volume, cm3 1.7 (1.1–4.2) 2.1 (0.4–4.8) 1.7 (1.1–4.2) .977 Apical thrombus, n (%) 43 (100) 16 (100) 27 (100) 1.000 Left ventricular aneurysm, n (%) 12 (27.9) 6 (37.5) 6 (22.2) .286 Note: Categorical data are presented as counts and percentages and analyzed using Chi-square-test. Continuous data are presented as median and the respective interquartile range and analyzed using Mann Whitney U test. 3.1 Follow-up and outcome analysis

After a median follow-up time of 108 weeks (IQR 68–173), 16.3% of patients died (n = 7), with 31.3% of individuals (n = 5) in the no LVT resolution subgroup and 7.4% (n = 2) in the LVT resolution subgroup, respectively (p = .022) (Table 3). Cardiovascular death occurred in 9.3% of patients with LVT (n = 4) with a non-significant lower event rate in the LVT resolution group (3.7% vs. 18.8%; p = .062). In total, MACE occurred in 32.6% (n = 14) of cases with a LVT, resulting in a significantly lower rate of MACE in the resolution group compared to the no resolution group (18.5% vs. 56.3%; p = .005). Thromboembolic events occurred in 27.9% of cases (n = 12), including 14.8% (n = 4) in the LVT resolution subgroup and 50.0% (n = 8) in the no LVT resolution subgroup (p = .008). Major bleeding events occurred in 9.3% of individuals (n = 4) presenting with LVT, without significant subgroup differences (p = .296). (Table 3) LVT resolution proved to be inversely associated with long-term mortality, presenting with a crude HR of 0.18 (95% CI: 0.03–0.93; p = .041). Three different multivariate models (1. patient characteristics, 2. clinical presentation, and 3. laboratory values) were used to analyze whether the prognostic value of LVT was independently associated with mortality, MACE, and thromboembolic events (Table 4). Within the multivariate model 2, LVT resolution remained inversely associated with long-term mortality with an adjusted HR of 0.14 (95% CI: 0.03–0.75; p = .021) (Table 4).

TABLE 3. All outcomes Overall (n = 43) No resolution (n = 16) Resolution (n = 27) Log rank test p-value MACE, n (%) 14 (32.6) 9 (56.3) 5 (18.5) .005 CV death, n (%) 4 (9.3) 3 (18.8) 1 (3.7) .062 All-cause death, n (%) 7 (16.3) 5 (31.3) 2 (7.4) .022 Thromboembolic events, n (%) 12 (27.9) 8 (50.0) 4 (14.8) .008 Major bleeding (BARC 2/3), n (%) 4 (9.3) 2 (12.5) 2 (7.4) .296 Note: Categorical data are presented as counts and percentages and analyzed using Log rank test. Abbreviations: CV death, cardiovascular death; MACE, major adverse cardiac events. TABLE 4. Unadjusted and adjusted effects of LVT resolution on long-term mortality, MACE and thromboembolic events All-cause death MACE Thromboembolic events HR (95% CI) p-value HR (95% CI) p-value HR (95% CI) p-value Univariate 0.18 (0.03–0.93) .041 0.26 (0.09–0.77) .015 0.23 (0.07–0.78) .018 Multivariate Model 1a 0.22 (0.04–1.21) .081 0.24 (0.08–0.71) .010 0.21 (0.06–0.72) .013 Model 2b 0.14 (0.03–0.75) .021 0.22 (0.07–0.68) .008 0.22 (0.06–0.75) .015 Model 3c 1.10 (0.12–9.73) .932 0.31 (0.09–1.03) .052 0.24 (0.07–0.86) .029 Note: Univariate and multivariate Cox proportional hazard models were applied to assess the effect of LVT resolution on all-cause death, MACE and thromboembolic events. The p values in bold indicate a value of <.05. Abbreviations: CI, confidence interval; HR, hazard ratio. a Model 1 was adjusted for age and sex. b Model 2 was adjusted for STEMI and heart failure. c Model 3 was adjusted for NT-proBNP and CK values.

In addition, LVT resolution was also associated with a significant lower risk of MACE with a crude HR of 0.26 (95% CI: 0.09–0.77; p = .015) and thromboembolic events with a crude HR of 0.23 (95% CI: 0.07–0.78; p = .018). LVT resolution remained inversely associated with MACE, after adjustment for model 1 (adj. HR of 0.24 [95% CI: 0.08–0.71]; p = .010) and model 2 (adj. HR of 0.22 [95% CI: 0.07–0.68]; p = .008) and thromboembolic events after adjustment for model 1 (adj. HR of 0.21 [95% CI: 0.06–0.72]; p = .013), model 2 (adj. HR of 0.22 [95% CI: 0.06–0.75]; p = .015) and model 3 (adj. HR of 0.24 [95% CI: 0.07–0.86]; p = .029). (Table 4).

Event rates for all-cause death, MACE, and thromboembolic events at 1 year were 3.7%, 14.8%, and 11.1% in the thrombus resolution group, compared to 25.0%, 43.8%, and 43.8% in the no thrombus resolution group, respectively.

The Kaplan Meier survival plot and log-rank test indicated a higher risk of long-term death (p = .022), MACE (p = .009) and thromboembolic events (p = .010) for individuals without LVT resolution as compared to patients with LVT resolution. (see Figure 1).

Survival curves of (A) all-cause death, (B) MACE, and (C) thromboembolic events. Comparison of survival, MACE, and thromboembolic events between patients with left ventricular thrombus resolution and without left ventricular thrombus resolution. MACE, major adverse cardiac events

3.2 Antithrombotic treatment strategies

Considering antithrombotic treatment strategies, we observed that all patients received OAC including 10 patients (23.3%) receiving non-vitamin-K oral anticoagulants (NOACs) and 33 patients (76.7%) vitamin K antagonists (VKA) respectively immediately after diagnosis. Median duration of anticoagulation therapy was 24 weeks (12–72) without significant difference between both groups. The fraction of individuals receiving DAT or TAT did also not differ significantly between both groups (p = .137). However, the use of TAT with a more potent P2Y12 inhibitor such as ticagrelor or prasugrel was observed only in the thrombus resolution group (38.1% vs. 0%; p = .031). Median duration of antiplatelet therapy with ticagrelor/prasugrel was 12 months, with a similar duration of 12 months for antiplatelet therapy with clopidogrel—there were no significant differences between both groups. Most importantly TAT with either ticagrelor or prasugrel showed a strong and independent association with thrombus resolution with a crude HR of 3.67 (95% CI: 1.53–8.81; p = .004). Notably, the prognostic impact remained stable after adjustment for model 1 (adj. HR of 3.69 [95% CI: 1.53–8.91; p = .004]), model 2 (adj. HR of 3.25 [95% CI: 1.22–8.68; p = .019]), and model 3 (adj. HR of 2.69 [95% CI: 1.10–6.58; p = .030]). (Table 5).

TABLE 5. Unadjusted and adjusted effects of TAT with ticagrelor/prasugrel on thrombus resolution HR (95% CI) p-value Univariate 3.67 (1.53–8.81) .004 Multivariate Model 1a 3.69 (1.53–8.91) .004 Model 2b 3.25 (1.22–8.68) .019 Model 3c 2.69 (1.10–6.58) .030 Note: TAT with ticagrelor/prasugrel. Univariate and multivariate Cox proportional hazard models were applied to assess the effect of TAT with ticagrelor/prasugrel on LVT resolution. The p values in bold indicate a value of <.05. Abbreviations: CI, confidence interval; HR, hazard ratio. a Model 1 was adjusted for age and sex. b Model 2 was adjusted for STEMI and heart failure. c Model 3 was adjusted for NT-proBNP and CK values. 4 DISCUSSION

The current analysis is—to the best of our knowledge—one of the first in literature that investigated the impact of LVT resolution after ACS on cardiovascular events and mortality. The present data illustrates that LVT resolution was independently associated with a favorable long-term outcome and survival free of MACE and thromboembolic events. In addition, our data indicates that TAT with potent P2Y12 inhibitor might be considered in patients with LVT.

Within the present investigation, we observed an incidence rate of LVT after ACS of 2.5%. Reported incidence rates vary from 1.6% up to 39%.2-