{
  "abstract": "Mechanical thrombectomy is the standard of care for acute ischemic stroke from large vessel occlusion (LVO). Recent trials have demonstrated a significant treatment effect of thrombectomy despite large ischemic cores (≥70 mL), prompting reconsideration of residual functional estimation within infarcted tissue. We propose that one component of this treatment effect relates to differences in depth of infarction within affected tissue, which traditional volumetric assessments alone fail to capture. Apparent diffusion coefficient (ADC) values provide a per-voxel quantitative measure of diffusion restriction and are reduced in acute infarction, with lower values potentially reflecting greater cytotoxic or ionic edema. We hypothesize that lower ADC values within the infarct bed additively predict worse long-term neurologic outcomes alongside infarct volume in large core infarct patients.We retrospectively reviewed a prospectively maintained thrombectomy database at one comprehensive stroke center (2022-2025). Inclusion criteria were thrombectomy patients >18 years old with anterior circulation LVO who had a 24-hour post-procedure diffusion-weighted imaging (DWI) study of sufficient quality. Clinical data included pre-procedure modified Rankin Scale (mRS) and 90-day mRS. DWI sequences were manually segmented to generate infarct maps, which were superimposed onto corresponding ADC maps. Overall infarct volume and per-voxel ADC values were collated for each patient. Outcomes (mRS 0-6) were modeled using generalized linear mixed models with observations nested within patients (SAS Software). ADC values were additionally grouped into eight bins (0-120 as bin 1 to 721-820 as bin 8) for comparative analysis.Ninety-eight patients were included: 38 with large core infarct (71-100 cc) and 60 with very large core infarcts (>101 cc). For every 100,000 mm3 increase in infarct volume, the odds of a 1-point increase in 90-day mRS were 3.03 (95% CI 1.11-8.26; p=0.032). For every 1000-unit decrease in ADC by volume, the odds of a higher 90-day mRS increased 3.6-fold (OR 3.6, 95% CI 3.5-3.72; p<0.0001), though this effect was smaller than that of volume alone. When comparing binned ADC values, bin 1 was associated with 3.7-fold higher odds of a 1-point mRS increase relative to bin 8 (95% CI 3.6-3.8; p<0.0001). Qualitative color mapping demonstrated ADC heterogeneity within infarcts.In this cohort enriched for very large infarcts, infarct volume remained the dominant predictor of outcome. However, per-voxel ADC mapping contributed independently, with deeper infarction associated with poorer functional outcomes. Data truncation limits generalizability of this conclusion across the spectrum of moderate-to-large core infarcts.Disclosures L. Bi: None. X. Yang: None. Z. Jiao: None. %M. Usher: None. C. Nguyen: None. R. Kim: None. E. Karayi: None.Abstract O-048 Figure 1",
  "authors": [
    {
      "affiliations": [
        "Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "L Bi"
    },
    {
      "affiliations": [
        "Rhode Island Hospital, Providence, RI"
      ],
      "name": "X Yang"
    },
    {
      "affiliations": [
        "Rhode Island Hospital, Providence, RI"
      ],
      "name": "Z Jiao"
    },
    {
      "affiliations": [
        "Department of Imaging, The Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "G Baird"
    },
    {
      "affiliations": [
        "Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "M Usher"
    },
    {
      "affiliations": [
        "Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "C Nguyen"
    },
    {
      "affiliations": [
        "Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "R Kim"
    },
    {
      "affiliations": [
        "Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "E Karayi"
    },
    {
      "affiliations": [
        "Department of Imaging, The Warren Alpert Medical School of Brown University, Providence, RI"
      ],
      "name": "D Hong"
    },
    {
      "affiliations": [
        "The Warren Alpert Medical School, Providence, RI"
      ],
      "name": "MR Gonzales"
    },
    {
      "affiliations": [
        "Department of Neurosurgery, Brown University, Providence, RI"
      ],
      "name": "JR Feler"
    },
    {
      "affiliations": [
        "Neurointervention, Semmes-Murphey Clinic, Memphis, TN"
      ],
      "name": "R Torabi"
    },
    {
      "affiliations": [
        "Rhode Island Hospital, Providence, RI"
      ],
      "name": "M Jayaraman"
    },
    {
      "affiliations": [
        "Diagnostic Imaging, The Warren Alpert School of Medicine at Brown University, Providence, RI"
      ],
      "name": "DN Wolman"
    }
  ],
  "title": "O-048 Per-voxel quantitative ADC mapping of large core infarction following mechanical thrombectomy as a measure of infarct depth",
  "uid": "49fcc752-6159-530b-a919-1956212d72ed"
}