Background: Cerebral amyloid angiopathy–related inflammation (CAA-RI), a potentially reversible yet life-threatening manifestation of CAA, causes subacute cognitive decline and contributes to vascular cognitive impairment (VCI). Relapse after corticosteroid taper is frequent, but predictors of recurrence and optimal long-term immunosuppression remain undefined. We hypothesized that MRI patterns predict relapse and that mycophenolate mofetil may reduce recurrence risk and cognitive deterioration. Methods: We conducted a retrospective cohort study of 36 consecutive patients with CAA-RI treated at the Cleveland Clinic from January 1, 2020, to December 31, 2024. The primary outcome was relapse, defined as recurrent vasogenic edema after initial corticosteroid response. Secondary outcomes included mortality, time to second relapse, and implications for cognition. Imaging predictors included hemorrhagic and white matter markers, with edema categorized as focal (1 lobar region), multifocal (2–3 lobar regions), or diffuse (≥4 lobar regions). Cause-specific Cox models and Gray’s test assessed predictors and treatment effects, including mycophenolate. Results: Of 36 patients (mean age 71.6 SD 8.2 years; 66.7% female; 19.4% Black), 11 (30.6%) relapsed, 11 (30.6%) died (9 without relapse), and 16 (44.4%) remained stable. After excluding 5 patients, relapse rates were 22.4%, 26.3%, and 44.0% at 6, 12, and 24 months. Diffuse edema was the strongest predictor of relapse (HR 12.26; 95% CI, 1.95–76.94; P = .007). Mycophenolate was used in 11 patients (7 at taper, 4 post-relapse); relapse occurred in 18.2% vs 30.0% without treatment, with mortality 9.1% vs 30.0%. Mycophenolate was associated with a nonsignificant reduction in relapse (HR 0.36; 95% CI, 0.07–1.09). Conclusions: Relapse is common in CAA-RI and persists beyond the early post-treatment period. Diffuse edema pattern strongly predicts recurrence, supporting its role in risk stratification. Mycophenolate shows promise as a steroid-sparing therapy to reduce relapse, cognitive decline, and mortality, but requires prospective trials to confirm benefit and assess safety, including effects on hemorrhagic risks and cognitive outcomes. These findings not only refine prognostication in CAA-RI but also have broader implications for amyloid-related imaging abnormalities with edema (ARIA-E) in Alzheimer’s disease, where overlapping vascular-inflammatory pathways may inform risk stratification and therapeutic strategies.
Kharal et al. (Thu,) studied this question.