Abstract 26: Timing The Optimal Transfer Modality For Suspected Large-vessel Stroke Patients: A Post-hoc Analysis Of The Racecat Trial

Introduction: Current recommendations for regional stroke destination suggest that patients with an acute severe stroke should be triaged based on estimated time to arrival to a thrombectomy-capable center. We aimed to evaluate which time period available at the time that patient is triaged is able to discriminate which transfer modality should be chosen.

Methods: We built and ordered logistic regression model adjusted for multiple comparisons with the RACECAT trial population using time periods available during triage: time from onset to emergency medical services (EMS) evaluation, estimated time of arrival to the thrombectomy-capable center and between centers distance. Estimated times were computed using a distance matrix API. Primary outcome was disability at 90 days, as assessed by the shift analysis on the modified Rankin score.

Results: Of the 1369 patients evaluated, median time from onset to EMS evaluation, estimated time to arrival to the thrombectomy-capable center and between centers distances were 65 minutes (interquartile ratio (IQR) 43 to 138), 61 minutes (IQR 36 to 80) and 62 minutes (IQR 36 to 73), respectively. In patients transferred to local stroke centers, delay in EMS evaluation was associated with higher degrees of disability (for each 30 minutes delay, adjusted common odds ratio (acOR) 1.035, 97.5% confidence interval (CI) 1.005 to 1.066), with no influence in patients directly transferred to thrombectomy-capable centers (for each 30 minutes delay, acOR 0.999, 97.5% CI 0.981 to 1.018) (p interaction =0.048). In patients evaluated by EMS above 120 minutes after stroke onset, direct transfer to a thrombectomy-capable center was associated with lower degrees of disability (acOR 1.494, 95% CI 1.026 to 2.174).

Conclusion: In the RACECAT trial, delay in EMS evaluation was associated with higher degrees of disability in patients transferred to local stroke centers and may serve as a potential biomarker for prehospital triage optimization.

Abstract WMP66: Multiparametric Neuroimaging And Its Association With Non-Contrast Computed Tomography In Late Window Large Vessel Occlusion Acute Stroke

Background: Endovascular treatment (EVT) for acute ischemic stroke (AIS) between 6 to 24 hours is established as a standard of care among patients selected by multiparametric neuroimaging. Therefore, we aimed to explore neuroimaging parameters in late window AIS large vessel occlusion (LVO) patients and the association between findings in non-contrast computed tomography (NCCT) and multiparametric CT.

Methods: We included consecutive AIS patients within 6-24 hours from symptoms onset with CTA-LVO. We studied potential associations between computed tomography mismatch defined by DAWN and/or DEFUSE-3 neuroimaging criteria (CTP-MM), infarct volume on CTP, and ASPECTS on NCCT. We also analyzed the association between neuroimaging parameters and outcome determined by 90-day mRS.

Results: We included 206 patients, of which 176 (85.4%) presented CTP-MM and 184 (89.3%) presented with an ASPECTS ≥ 6 on admission. The rate of CTP-MM was 90.8% in patients with ASPECTS ≥ 6, as compared with 40.9% in those with low ASPECTS [Figure 1A] . The ASPECTS correlated with infarct core, determined by Cerebral Blood Flow <30% volume (rP=-0.575, P<0.001). In EVT-treated patients (185, 89.8%), after adjusting for identifiable confounders, the presence of CTP-MM was a predictor of 90-day functional independence (OR 3.38; 95%CI 1.01-11.29; P=0.048). We did not find an association between CTP-MM and 90-day functional disability (ordinal mRS shift, aOR 1.39; 95% CI 0.58-3.34; P=0.459) [Figure 1B] .

Conclusions: A great majority of patients who presented a LVO in late window fulfilled guidelines imaging criteria to undergo EVT, especially those with high ASPECTS (≥ 6). Our data suggest that NCCT with CT angiography is a reasonable approach for acute ischemic stroke treatment selection also in the late window.

Abstract 94: Prospective Clinical Evaluation Of The Modified CDC Criteria For Diagnosis And Management Of Stroke Associated Pneumonia

Objective: Stroke-associated infections (SAI) and specially pneumonia (SAP) frequently complicate stroke, leading to poor outcome. The modified Center of Disease Control and Prevention (mCDC) diagnostic criteria for SAP have been proposed, but not yet validated. We aimed to analyze the fulfillment of these criteria for SAP diagnosis and antibiotic guidance, as well as compare clinical outcomes between mCDC-defined SAP and physician’s diagnosed pneumonia.

Methods: The PROTEUS study was a prospective, multicenter, observational study including ischemic stroke patients with baseline stroke severity (National Institutes of Health Stroke Scale) NIHSS≥4. Patients with concurrent infection or antibiotic use were excluded. During the first week after stroke, the mCDC criteria for SAP were daily assessed by the study investigators. Clinical diagnosis of pneumonia and other infections, and antibiotic use were recorded by treating physicians. Clinical outcome was assessed with the modified Rankin scale (mRS) at three months, and categorized into poor (mRS>2) and favorable (mRS≤2) outcome. The effect of criteria-defined SAP and physician-diagnosed SAP on clinical outcome was evaluated by logistic regression analysis.

Results: From July-2018 to May-2021, 340 patients were included in 9 stroke units from Spain. Of them, any SAI was diagnosed in 75 patients (22.1%), being 41 (12.1%) diagnosed as pneumonia by physicians. SAP according to mCDC criteria was diagnosed in 26 cases (7.6% of the cohort, 63.4% of physician’s diagnoses). Antibiotics were used in 88.5% of criteria-defined SAP and 73.3% in physician-diagnosed pneumonia (50% in patients with pneumonia not fulfilling mCDC criteria). In logistic regression analysis, criteria-defined SAP was an independent predictor of 3-month poor outcome (OR=7.13 (1.58-32.12)). However, physician-diagnosed pneumonia not fulfilling mCDC criteria was not an independent predictor of poor clinical outcome (OR=1.40 (0.38-5.08)).

Conclusions: Despite being associated with clinical outcome, mCDC criteria are not used in clinical practice, neither for SAP diagnosis nor to guide antibiotic therapy. Adherence to these criteria might reduce variability in clinical diagnosis and antibiotic overuse in stroke units.