We collected original articles in this field by searching PubMed, Google Scholar, Web of Science, and Scopus databases for English‑language literature published up to May 2024. The search was conducted based on (“cardiogenic cerebral embolism”) or (“CCE”), AND (“therapy”) or (“treatment”) and a combination of a list of drugs and treatment modalities related to cardiogenic cerebral embolism as keywords. The full search strategy is reported in (Supplementary Table 1; Supplementary Digital Content is available online at https://doi.org/10.26676/jevtm.40591).

Furthermore, duplicate records were omitted using EndNote version 21 and the Rayyan intelligence tool for systematic reviews. To identify other suitable studies, we also reviewed the references of relevant papers and reviews on the topic of safety of current therapies for cardiogenic cerebral embolism (Supplementary Table 1).

Two reviewers (MA, AM) employed the Rayyan intelligence tool for systematic reviews to evaluate and filter titles and abstracts in a blinded manner, identifying analogous works. The texts were acquired to evaluate the qualifications of the “Yes” and “Maybe” groups. In the event of conflicts, a third reviewer was engaged to facilitate consensus and resolve discrepancies. Conflicts were addressed by dialogue between the parties. Quality assessment and risk of bias for each included study were conducted utilizing JBI’s critical appraisal methods.

A total of 14,236 studies were found in the screening database search, 5,740 of which were duplicate records. Two reviewers (MA, SH) examined article titles and/or abstracts. After screening 8,496 records (Figure 1), we excluded 8,443. In total, 53 studies were selected for full‑text review. Forty‑three reports were not retrieved due to the unavailability of full text. Ten studies met the inclusion criteria and were included in this review.

Fu et al. demonstrated that patients undergoing embolectomy, either alone or in conjunction with thrombolytic therapy, had superior short‑ and long‑term functional outcomes compared to patients receiving IV rtPA therapy alone [9]. Compared to the group that received thrombolytic treatment alone, the NIHSS score improvement for the Solitaire stent embolectomy group was noticeably more significant. The group with embolectomy showed a significantly higher GCS improvement than the group that received IV rtPA therapy alone. The long‑term outcome that was assessed by measuring the mRS score was determined to be significantly better in stent patients [9]. The GCS is typically performed upon arrival at the emergency department to assess the patient’s initial level of consciousness. This is a crucial early evaluation to determine the severity of neurological impairment and guide immediate treatment decisions.

Matsukawa et al. demonstrated noticeably lower NIHSS scores in the EVT group [10]. Patients in the EVT group had a better clinical course (mRS score 0–2) than those in the no‑EVT group. The EVT group had a significantly lower mortality rate within 90 days than the no‑EVT group. The proportions of any intracranial bleeding and symptomatic intracranial bleeding within 72 hours and recurrence of stroke or transient ischemic attack within 90 days were similar between the two groups [10].

According to Cao et al., direct mechanical thrombectomy may be more beneficial for individuals with mild to moderate cardioembolic stroke than bridging therapy [11]. There were no significant differences in the outcome and mortality rate of CE stroke patients with an NIHSS >15 between the two treatment groups, but patients with an NIHSS ≤15 in the direct MT group were linked to better outcomes and lower mortality than those in the bridging therapy group.

The results from Zhao et al. showed that patients who received tirofiban experienced fewer decompressive craniotomies and cerebral hernia than those who did not [18]. There was a significant difference in the three‑month mortality rates between patients who received tirofiban (20.5%) and those who did not (31.6%).

Yaghi et al. discovered that individuals with AIS in the context of AF who were not treated with MT had a lower mortality rate when they received IV alteplase [12]. Among patients undergoing MT, there was a non‑significant reduction in the number of passes and deaths in subjects treated with intravenous alteplase compared to those who did not receive intravenous alteplase. Alteplase‑treated patients were more likely to have an initial NIHSS score with a higher median (interquartile range) (8 [4–15] versus 6 [2–14], P = 0.002).

According to Vaclavik et al., there is no association between cardioembolic strokes (CS) and higher mortality rates [13]. After intravenous thrombolysis (IVT), patients with CS had better outcomes and were less likely to have sICH [13].

Frisullo et al. found that 53 (70.7%) of the 75 patients had excellent functional outcomes at three months (defined as mRS score 0–1), two (2.7%) patients had gastrointestinal bleeding, and 11 (14.7%) patients had minor bleeding (five epistaxis, three gingival bleedings, and three cutaneous hematomas) [14]. None of the 75 patients had significant intracranial hemorrhage at three months.

Cappellari found no correlation between the early introduction of DOAC and intracranial bleeding [16]. No patients experienced recurrent ischemic stroke.

The study conducted by Berge provides no evidence that high‑dose LMWH is superior to aspirin for the prevention of recurrent ischemic stroke during the first 14 days [17]. The study showed no significant increase in sICH or improved outcome on LWMH compared with aspirin.

According to Labovitz et al., the early use of the DOAC apixaban did not increase the number of intracranial hemorrhages, including hemorrhagic transformation (HT) and intracranial hemorrhages (ICH) [15].

The summary of the outcomes is shown in Table 1.

Endovascular mechanical recanalization technology is rapidly being used as a first‑line treatment for acute cerebral infarction since the development of catheter and neural intervention techniques, especially in patients with occluded major cerebral arteries. Large intracranial vessels, such as the basilar artery, the middle cerebral artery, the anterior cerebral artery, and the end of the internal carotid artery, frequently become blocked due to cardioembolisms. The recanalization rate of such occluded arteries is relatively low, about 10%, when rtPA thrombolytic therapy is the only treatment used. Furthermore, the bleeding conversion rate is estimated to be 10.23%. Using IV rtPA thrombolysis alone to treat cerebral embolism often does not achieve satisfactory results [28,30].

The MR‑CLEAN trial [31] revealed the first indication that mechanical thrombectomy performed within six hours of symptom onset improved 90‑day clinical outcomes (mRS score) compared to a group receiving standard medical care, with 90.6% receiving IV rtPA within 4.5 hours. According to the Fu et al. study, the stent group had significantly higher NIHSS and GCS scores during admission and discharge than the group that received drug therapy alone. In addition, the group that underwent mechanical thrombectomy had considerably better long‑term clinical outcomes [9].

For the first time, a sub‑analysis of an extensive prospective registry indicated that EVT could reduce the incidence of delayed hemorrhage and had no influence on neither any intracranial hemorrhage nor on symptomatic intracranial hemorrhage within 72 hours in patients with cardioembolic proximal intracranial occlusion in the anterior circulation [10].

The question of whether individuals with acute ischemic stroke and atrial fibrillation should be treated with heparin as an anticoagulant has long been debated [32–34].

The results of the study by Berge et al. provide no evidence that high‑dose LMWH is better than aspirin for improving outcomes at 14 days or three months, or in preventing recurrent ischemic stroke or any other event during the first 14 days [17].

Yaghi et al. discovered that patients with acute ischemic stroke in the setting of AF who were not treated with MT had a lower 90‑day mortality rate when they received intravenous alteplase medication [21]. These results support the findings of the previous extensive cohort studies [21].

When MT patients received IV alteplase treatment as opposed to those who did not, there was a non‑significant reduction in the number of passes and deaths among the former group [12]. This aligns with the pivotal thrombectomy trials, which showed no impact on mortality or outcome [35,36]. This is also consistent with recent trials demonstrating no substantial additional benefit of alteplase usage in patients with proximal blockage undergoing MT [34,37,38]. These findings may be attributable to alteplase’s poor ability to achieve reperfusion in patients with proximal large‑artery occlusion successfully. However, our results emphasize the need for more research on this matter.

In addition, based on the RCT by Cao et al., patients with mild and moderate cardioembolic stroke may benefit more from direct mechanical thrombectomy than from bridging therapy [11].

For individuals with nonvalvular AF, DOACs such as apixaban, dabigatran, edoxaban, and rivaroxaban are now the main treatment option for preventing stroke [39]. One in six stroke patients who are eligible for IVT are expected to have been administered DOACs after the switch from vitamin K antagonists (VKAs) to DOACs. In situations of ischemic stroke, the recommendation suggests against administering IVT to patients who have recently administered DOACs (within the last 48 hours). This advice is based on the assumption that there is a higher risk of sICH. Nevertheless, there is not much data on when oral anticoagulation should be started following an acute stroke [40].

A prospective, non‑randomized study found that early initiation of edoxaban, within five days, does not seem to lead to any symptomatic intracranial bleeding or recurrent stroke after three months. However, two gastrointestinal major bleedings and 11 minor bleedings were reported [14]. A prospective analysis showed no association between early initiation of DOAC (1–3 days after stroke onset) and intracranial bleeding in stroke patients with non‑valvular atrial fibrillation [16]. However, the major weakness in these studies was the absence of a control group with delayed anticoagulant treatment.

Based on our review, a randomized controlled trial was done comparing the safety of early use of apixaban versus warfarin in a 1:1 ratio. It revealed that apixaban had numerically smaller but statistically comparable rates of death, fatal stroke, recurrent strokes/TIA, and symptomatic hemorrhages. Early anticoagulant initiation following TIA or small‑ or medium‑sized AIS from AF does not seem to impair patient safety [15]. However, more extensive pivotal trials are necessary to ascertain the possible effectiveness of the early initiation of DOCA.

The standard treatment for patients with AIS due to large‑vessel obstruction is EVT [41–43]. Conversely, endothelial damage might be an unavoidable consequence that results in early re‑occlusion and infarction extension [44]. Significant interest has been shown in the adjunctive administration of antiplatelet medications to reduce the rate of ischemia complications and increase the rate of favorable reperfusion.

Earlier research demonstrated that tirofiban was linked to a greater possibility of functional independence in AIS patients receiving EVT. It was also associated with a decreased risk of re‑occlusion and an improved rate of favorable reperfusion [45,46].

However, a prospective study of 288 cardioembolic stroke patients found no correlation between the administration of low‑dose tirofiban and three‑month mortality. Additionally, there was no association with hernia, decompressive craniectomy following EVT, or different types of intracranial hemorrhage, including ICH, sICH, or fatal ICH.

Furthermore, there was no association between the use of tirofiban and a low incidence of re‑occlusion after EVT. On the other hand, tirofiban was linked to a lower risk of in‑hospital mortality [18]. The study design and the number of patients in this study may have caused bias. Further studies are needed to confirm these results and improve the best treatment course that can benefit this patient population.

The main strength of this study is that it explicitly addresses cardioembolic stroke patients, which is the first review of different available treatment modalities for these patients. The primary limitation of our investigation is the small sample size of the included studies. More large‑scale multicenter randomized controlled trials are necessary to validate these findings.