Artificial intelligence to predict individualized outcome of acute ischemic stroke patients: The SIBILLA project.

INTRODUCTION: Formulating reliable prognosis for ischemic stroke patients remains a challenging task. We aimed to develop an artificial intelligence model able to formulate in the first 24 h after stroke an individualized prognosis in terms of NIHSS. PATIENTS AND METHODS: Seven hundred ninety four acute ischemic stroke patients were divided into a training (597) and testing (197) cohort. Clinical and instrumental data were collected in the first 24 h. We evaluated the performance of four machine-learning models (Random Forest, K-Nearest Neighbors, Support Vector Machine, XGBoost) in predicting NIHSS at discharge both in terms of variation between discharge and admission (regressor approach) and in terms of severity class namely NIHSS 0-5, 6-10, 11-20, >20 (classifier approach). We used Shapley Additive exPlanations values to weight features impact on predictions. RESULTS: XGBoost emerged as the best performing model. The classifier and regressor approaches perform similarly in terms of accuracy (80% vs 75%) and f1-score (79% vs 77%) respectively. However, the regressor has higher precision (85% vs 68%) in predicting prognosis of very severe stroke patients (NIHSS > 20). NIHSS at admission and 24 hours, GCS at 24 hours, heart rate, acute ischemic lesion on CT-scan and TICI score were the most impacting features on the prediction. DISCUSSION: Our approach, which employs an artificial intelligence based-tool, inherently able to continuously learn and improve its performance, could improve care pathway and support stroke physicians in the communication with patients and caregivers. CONCLUSION: XGBoost reliably predicts individualized outcome in terms of NIHSS at discharge in the first 24 hours after stroke.

Anticoagulation in acute ischemic stroke patients with mechanical heart valves: To bridge or not with heparin. The ESTREM study.

INTRODUCTION: The best therapeutic strategy for patients with mechanical heart valves (MHVs) having acute ischemic stroke during treatment with vitamin K antagonists (VKAs) remain unclear. Being so, we compared the outcomes for: (i) full dose heparin along with VKA (bridging therapy group) and (ii) restarting VKA without heparin (nonbridging group). PATIENTS AND METHODS: For this multicenter observational cohort study, data on consecutive acute ischemic stroke patients with MHV was retrospectively collected from prospective registries. Propensity score matching (PSM) was adopted to adjust for any treatment allocation confounders. The primary outcome was the composite of stroke, systemic embolism, symptomatic cerebral bleeding, and major extracerebral bleeding at 90 days. RESULTS: Overall, 255 out of 603 patients (41.3%) received bridging therapy: 36 (14.1%) had combined outcome, compared with 28 (8.0%) in the nonbridging group (adjusted OR 1.83; 95% CI 1.05-3.18; p = 0.03). Within the bridging group, 13 patients (5.1%) compared to 12 (3.4%) in the nonbridging group had an ischemic outcome (adjusted OR 1.71; 95% CI 0.84-3.47; p = 0.2); major bleedings were recorded in 23 (9.0%) in the bridging group and 16 (4.6%) in the nonbridging group (adjusted OR 1.88; 95% CI 0.95-3.73; p = 0.07). After PSM, 36 (14.2%) of the 254 bridging patients had combined outcome, compared with 23 (9.1%) of 254 patients in the nonbridging group (OR 1.66; 95% CI 0.95-2.85; p = 0.07). CONCLUSION: Acute ischemic stroke patients with MHV undergoing bridging therapy had a marginally higher risk of ischemic or hemorrhagic events, compared to nonbridging patients.

Timing of Antithrombotic Secondary Prevention in Patients with Intracranial Hemorrhage after Stroke Thrombolysis and Thrombectomy.

In patients with acute ischemic stroke, hemorrhagic transformation (HT) of infarcted tissue frequently occurs after reperfusion treatment. We aimed to assess whether HT and its severity influences the start of secondary prevention therapy and increases the risk of stroke recurrence. In this retrospective dual-center study, we recruited ischemic stroke patients treated with thrombolysis, thrombectomy or both. Our primary outcome was the time between revascularization and the start of any secondary prevention therapy. The secondary outcome was ischemic stroke recurrence within three months. We compared patients with vs. without HT and no (n = 653), minor (n = 158) and major (n = 51) HT patients using propensity score matching. The delay in the start of antithrombotics or anticoagulants was median 24 h in no HT, 26 h in minor HT and 39 h in major HT. No and minor HT patients had similar rates of any stroke recurrence (3.4% (all ischemic) vs. 2.5% (1.6% ischemic plus 0.9% hemorrhagic)). Major HT patients had a higher stroke recurrence at 7.8% (3.9% ischemic, 3.9% hemorrhagic), but this difference did not reach significance. A total of 22% of major HT patients did not start any antithrombotic treatment during the three-month follow-up. In conclusion, the presence of HT influences the timing of secondary prevention in ischemic stroke patients undergoing reperfusion treatments. Minor HT did not delay the start of antithrombotics or anticoagulants compared to no HT, with no significant difference in safety outcomes. Major HT patients remain a clinical challenge with both a delayed or lacking start of treatment. In this group, we did not see a higher rate of ischemic recurrence; however, this may have been censored by elevated early mortality. While not reaching statistical significance, hemorrhagic recurrence was somewhat more common in this group, warranting further study using larger datasets.

Actigraphic Sensors Describe Stroke Severity in the Acute Phase: Implementing Multi-Parametric Monitoring in Stroke Unit.

Actigraphy is a tool used to describe limb motor activity. Some actigraphic parameters, namely Motor Activity (MA) and Asymmetry Index (AR), correlate with stroke severity. However, a long-lasting actigraphic monitoring was never performed previously. We hypothesized that MA and AR can describe different clinical conditions during the evolution of the acute phase of stroke. We conducted a multicenter study and enrolled 69 stroke patients. NIHSS was assessed every hour and upper limbs' motor activity was continuously recorded. We calculated MA and AR in the first hour after admission, after a significant clinical change (NIHSS ± 4) or at discharge. In a control group of 17 subjects, we calculated MA and AR normative values. We defined the best model to predict clinical status with multiple linear regression and identified actigraphic cut-off values to discriminate minor from major stroke (NIHSS ≥ 5) and NIHSS 5-9 from NIHSS ≥ 10. The AR cut-off value to discriminate between minor and major stroke (namely NIHSS ≥ 5) is 27% (sensitivity = 83%, specificity = 76% (AUC 0.86 p < 0.001), PPV = 89%, NPV = 42%). However, the combination of AR and MA of the non-paretic arm is the best model to predict NIHSS score (R2: 0.482, F: 54.13), discriminating minor from major stroke (sensitivity = 89%, specificity = 82%, PPV = 92%, NPV = 75%). The AR cut-off value of 53% identifies very severe stroke patients (NIHSS ≥ 10) (sensitivity = 82%, specificity = 74% (AUC 0.86 p < 0.001), PPV = 73%, NPV = 82%). Actigraphic parameters can reliably describe the overall severity of stroke patients with motor symptoms, supporting the addition of a wearable actigraphic system to the multi-parametric monitoring in stroke units.

Dual anti-platelet therapy for secondary prevention in intracranial atherosclerotic disease: a network meta-analysis.

Background: Intracranial arterial stenosis (ICAS) is a non-marginal cause of stroke/transient ischemic attacks (TIAs) and is associated with high stroke recurrence rate. Some studies have investigated the best secondary prevention ranging from antithrombotic therapy to endovascular treatment (ET). However, no direct comparison between all the possible treatments is currently available especially between single and dual anti-platelet therapies (SAPT and DAPT). Aim: To establish whether DAPT is more effective than SAPT in preventing the recurrence of ICAS-related stroke, by means of a network meta-analysis (NMA). Design: Systematic review and NMA in accordance to PRISMA guidelines. Data sources and methods: We performed a systematic review of trials investigating secondary prevention (SAPT or DAPT, anticoagulant treatment or ET) in patients with symptomatic ICAS available in MEDLINE, Scopus and Web of Science from January 1989 to May 2021. We defined our primary efficacy outcome as the recurrence of ischemic stroke/TIA. We analysed the extracted data with Bayesian NMA approach. Results: We identified 815 studies and included 5 trials in the NMA. Sequence generation was adequate in all the selected studies while the allocation concealment method was described in one study. All the included studies reported the pre-specified primary outcomes, and outcome assessment was blinded in all the studies. We used the fixed-effect approach as the heterogeneity was not significant (p > 0.1) according to the Cochran's Q statistic. DAPT was superior to SAPT and DAPT + ET in preventing stroke/TIA recurrence [respectively, odds ratio (OR), 0.59; confidence interval (CI), 0.39-0.9; and OR, 0.49, CI, 0.26-0.88], while no difference was found between DAPT and oral anticoagulant therapy (OAC). DAPT was safer than OAC (OR, 0.48; CI, 0.26-0.89) and DAPT + ET (OR, 0.50; CI, 0.35-0.71), while no difference was found between DAPT and SAPT. Conclusion: DAPT is more effective than SAPT for secondary stroke prevention in patients with symptomatic ICAS, without increasing the risk of haemorrhage. Registration: Prospero/CRD42019140033.

Actigraphic Measurement of the Upper Limbs for the Prediction of Ischemic Stroke Prognosis: An Observational Study.

BACKGROUND: It is often challenging to formulate a reliable prognosis for patients with acute ischemic stroke. The most accepted prognostic factors may not be sufficient to predict the recovery process. In this view, describing the evolution of motor deficits over time via sensors might be useful for strengthening the prognostic model. Our aim was to assess whether an actigraphic-based parameter (Asymmetry Rate Index for the 24 h period (AR2_24 h)) obtained in the acute stroke phase could be a predictor of a 90 d prognosis. METHODS: In this observational study, we recorded and analyzed the 24 h upper limb movement asymmetry of 20 consecutive patients with acute ischemic stroke during their stay in a stroke unit. We recorded the motor activity of both arms using two programmable actigraphic systems positioned on patients' wrists. We clinically evaluated the stroke patients by NIHSS in the acute phase and then assessed them across 90 days using the modified Rankin Scale (mRS). RESULTS: We found that the AR2_24 h parameter positively correlates with the 90 d mRS (r = 0.69, p < 0.001). Moreover, we found that an AR2_24 h > 32% predicts a poorer outcome (90 d mRS > 2), with sensitivity = 100% and specificity = 89%. CONCLUSIONS: Sensor-based parameters might provide useful information for predicting ischemic stroke prognosis in the acute phase.