By Michael T. Lin, MD, PhD
Associate Professor of Neurology and Neuroscience, Weill Cornell Medical College
SYNOPSIS: Many cases of rapidly progressive dementia are caused by prion diseases and have no effective treatments. But, with the greater awareness of the presentation for autoimmune encephalitis, these disorders make up an increasing percentage of presenting cases and can be aggressively and successfully treated. The STAM3P score helps to identify potentially treatable cases of this disorder.
SOURCE: Satyadev N, Tipton PW, Martens Y, et al. Improving early recognition of treatment-responsive causes of rapidly progressive dementia: The STAM3P score. Ann Neurol 2023; Oct 2. doi: 10.1002/ana.26812. [Online ahead of print].
Rapidly progressive dementia (RPD), defined as progression from first symptom to frank dementia in less than one year or to full-time care in less than two years, accounts for 3% to 4% of dementia cases. Although early studies emphasized the role of irreversible conditions, such as prion diseases, more recent studies show that a significant proportion of RPD is caused by potentially treatable conditions, particularly autoimmune encephalitis. It is critical that such cases be identified and treated expeditiously. In their recent article, Satyadev and colleagues found that 86 of 155 (55%) of prospectively collected RPD cases were potentially treatable, and they produced a simple clinical score to assist with identifying such cases.
From 2016 to 2022, 226 patients with suspected RPD were referred to two specialty memory clinics in the United States. Of these, 155 met the definition for RPD, and cases with obvious causes (large territory strokes, infectious encephalitis, trauma) were excluded. Workup consisted of history, examination, blood work, magnetic resonance imaging (MRI), electroencephalogram (EEG), and cerebrospinal fluid (CSF) analysis (including Alzheimer’s disease [AD] biomarkers [Athena Diagnostics or Mayo Clinic], prion evaluation [National Prion Disease Pathology Surveillance Center, Case Western Reserve], and autoimmune panel [Mayo Clinic]).
Suspicion for autoimmune encephalitis was high if there was rapid onset (< 3 months) of short-term memory loss, altered consciousness, or psychiatric symptoms; there were seizures, new focal findings, CSF pleocytosis, or MRI features of encephalitis; and other causes were excluded. MRI was considered suggestive for autoimmune encephalitis if there were characteristic T2-fluid attenuated inversion recovery (T2-FLAIR) abnormalities in the medial temporal lobes, supratentorial white matter, basal ganglia, brainstem, or cerebellum. MRI was suggestive for prion disease when there were characteristic diffusion or T2-FLAIR abnormalities in the basal ganglia or cortical ribboning.
Out of 155 cases, the majority (n = 86) had potentially treatable causes: autoimmune (52), vascular (inflammatory [4], dural arteriovenous fistula [1], sarcoidosis [1], Susac syndrome [1]), neoplastic (lymphoma [5], craniopharyngioma [1]), toxic metabolic (Wernicke-Korsakoff syndrome [4], hepatic encephalopathy [1], medications [2]), psychiatric (4), amyloid angiopathy with inflammation (8), and status epilepticus (2). The treatment-unresponsive conditions (n = 69) consisted primarily of Creutzfeldt-Jakob disease (25), neurodegenerative diseases (AD [18], frontotemporal degeneration [10], dementia with Lewy bodies [6], vascular dementia [7], glioblastoma multiforme [1]), and idiopathic leukoencephalopathy (2).
Regression analysis identified seven features associated with treatment-responsiveness, encapsulated in the mnemonic STAM3P: Seizures early in the course; Tumor (lymphoma or paraneoplastic); Age < 50 years; MRI suggestive of autoimmune encephalitis; Mania; Movement disorder (ataxia, chorea, dystonia, myoclonus, parkinsonism, opsoclonus); and CSF Pleocytosis (≥ 10 white blood cells/mm3). The STAM3P score is obtained by giving one point for each item present, with the total score ranging from 0 to 7. Detection of any one feature (STAM3P ≥ 1) at presentation captured 82/86 cases of potentially treatment-responsive RPD (sensitivity = 95.3%). All patients with at least three features (STAM3P ≥ 3) had potentially treatment-responsive causes of RPD (specificity = 100%).
The authors suggested that any RPD patient with STAM3P ≥ 1 be aggressively worked up for treatment-responsive etiologies, including blood and CSF testing for autoantibodies, brain and body imaging, and EEG. In patients with STAM3P = 2 (positive predictive value [PPV] = 91.3% for treatment responsiveness) or STAM3P ≥ 3 (PPV = 100% for treatment responsiveness), it would be reasonable to consider empiric therapy with steroids, intravenous immunoglobulin, or plasma exchange while awaiting diagnostic results.
These treatments have proven efficacy in common causes of RPD, and they are well tolerated in most non-treatment-responsive cases of RPD (prion disease, neurodegenerative disease). Exceptions include suspected lymphoma, sarcoidosis, or infection, in which early immunosuppressive treatment may compromise subsequent diagnosis or treatment. In contrast, a STAM3P score of 0 has a high negative predictive value (91.5%) for treatment responsiveness, and such cases most likely are the result of prion disease or AD.
COMMENTARY
Weaknesses of this study include the relative lack of diversity (84% participants were non-Hispanic white and all were enrolled in the United States), debatable inclusion of lymphoma as a treatment-responsive condition, and the relatively small number of autopsy confirmations (n = 24). However, based on this study, the STAM3P items form an easy, practical score that may be determined clinically close to the time of presentation, assisting in identifying and expediting workup and treatment of potentially reversible conditions.
