Abstract Altered mental status in the elderly often represents overlapping metabolic, infectious, autoimmune, or hypoxic processes. When neuroimaging is unrevealing, the distinction between reversible and irreversible encephalopathy becomes particularly challenging. This case of subacute encephalopathy with GPEDs and inflammatory cerebrospinal fluid findings after cardiac arrest highlights the difficulty of differentiating autoimmune or infectious encephalitis from post-anoxic injury. A 75-year-old man with hypertension, type 2 diabetes, and urothelial carcinoma of the bladder presented with three days of generalized weakness 266 RBC/μL; protein 300 mg/dL; normal glucose), inconsistent with isolated HIBI and suggestive of underlying encephalitis. Repeat EEG findings were unchanged. A multidisciplinary team initiated empiric therapy with acyclovir, ceftriaxone, doxycycline, and high-dose IV methylprednisolone (1 g daily ×5). Comprehensive infectious (West Nile, arbovirus, Lyme) and autoimmune (anti-NMDA, Hu, LGI1, GABA, AMPA) panels were negative. Despite aggressive management, he failed to recover meaningful neurologic function and required tracheostomy and PEG for long-term care. This case underscores the diagnostic and prognostic complexity of encephalopathy after cardiac arrest. Persistent GPEDs with inflammatory CSF findings created a diagnostic crossroads between possible Lyme-associated neuroinfection, autoimmune encephalitis, and post-anoxic injury. Graus et al. propose that possible autoimmune encephalitis may be diagnosed when subacute cognitive decline in less than 3 months coexists with CSF pleocytosis or new focal findings once alternative causes are excluded. Thus, even after cardiac arrest, clinicians should not attribute encephalopathy solely to hypoxia when CSF inflammation is present. Early CSF analysis, integration of EEG findings, and a parallel empiric “treat-while-testing” approach are essential in evaluating ambiguous encephalopathy, as timely recognition of reversible etiologies may alter prognosis. This abstract is funded by: None
Ahsan et al. (Fri,) studied this question.