Rapidly Progressive Dementia

Identifying Treatable Causes of Rapidly Progressive Dementia

Clinical Takeaway

• In a Dutch prospective multicenter cohort, 95 of 147 adults with rapidly progressive dementia (RPD), or 65%, received a diagnosis classified as potentially treatment-responsive.¹
• Autoimmune encephalitis (AE) was the largest diagnostic group, accounting for 58 of 147 patients (39%) and 58 of 95 treatment-responsive diagnoses (61%).
• This finding supports a rapid, structured search for reversible and immune-mediated causes. It does not mean that every patient with RPD should receive empiric immunotherapy.
• Seizures, mesiotemporal MRI abnormalities, cerebrospinal fluid (CSF) inflammation, and hyponatremia increased suspicion for AE, whereas characteristic cortical or basal ganglia restricted diffusion favored Creutzfeldt-Jakob disease (CJD) in this cohort.
• If AE is highly suspected, the patient meets an appropriate clinical level of evidence, infection has been reasonably excluded, and a diagnosis requiring biopsy is not being considered, expert recommendations support starting immunotherapy before antibody results return.^2,3

What Is Rapidly Progressive Dementia?

• RPD is a syndrome rather than a single disease.
• It is generally defined as progression from symptom onset to dementia within 1 or 2 years. The study used a stricter definition: dementia causing substantial functional interference within 1 year.¹
• Early etiologic classification matters because several immune-mediated, infectious, toxic, metabolic, neoplastic, vascular, epileptic, nutritional, and psychiatric causes may improve with targeted treatment.
• Early recognition of a nonresponsive disorder such as CJD is also clinically important for counseling, infection-control procedures, care planning, and possible trial enrollment.

Study Design

• Design: prospective, multicenter observational cohort.
• Setting: the Netherlands, including the national referral center for neuroinflammation and neuronal autoantibody testing, an academic memory clinic, and nationwide neurology consultations.
• Recruitment: December 2019 through December 2024.
• Screening: 286 patients with suspected RPD were screened; 147 met inclusion criteria.
• Participants:
  • Median age: 67 years.
  • Female: 67 of 147 (46%).
  • Median time from symptom onset to first presentation: 7 weeks.
  • Median follow-up: 33 months.
• Evaluation:
  • Brain MRI was performed in 146 of 147 patients (99%).
  • Paired serum and CSF were tested for neuronal and glial autoantibodies in 146 of 147 patients (99%).
  • Three neurologists with expertise in neurodegenerative or neuroinflammatory disease established final diagnoses by consensus.
  • Autopsy was performed in 8 patients (5%).

Major Diagnostic Findings

Potentially Treatment-Responsive Causes

• Ninety-five of 147 patients (65%) had a diagnosis classified as potentially treatment-responsive.¹
• AE was the most common category:
  • 58 of 147 patients (39% of the full cohort).
  • 58 of 95 patients (61% of the treatment-responsive group).
• Other treatment-responsive categories included additional neuroinflammatory disorders, toxic or metabolic disorders, delirium, central nervous system infection, malignancy, epilepsy, psychiatric disorders, and a dural arteriovenous fistula.

Nonresponsive Causes

• Fifty-two of 147 patients (35%) had a diagnosis classified as nonresponsive.
• The leading categories were:
  • Neurodegenerative disease: 20 of 147 (14%).
  • CJD: 17 of 147 (12%).
• The responsive versus nonresponsive label was based primarily on whether published evidence suggested that disease-specific therapy could produce sustained, clinically meaningful cognitive improvement. It was not an observed treatment-response endpoint for every participant.

Autoimmune Encephalitis Subtypes

• The 3 most common AE subtypes were:
  • Anti-LGI1 encephalitis: 19 of 58 AE cases (33%).
  • Seronegative AE: 11 of 58 (19%).
  • Autoimmune glial fibrillary acidic protein (GFAP) astrocytopathy: 8 of 58 (14%).
• Together, these diagnoses accounted for approximately two thirds of AE-RPD cases.
• Anti-LGI1 encephalitis was especially common when RPD presented with seizures.
• GFAP astrocytopathy was the most common AE subtype among patients without seizures. Psychosis, early movement disorders, cerebellar ataxia, CSF pleocytosis, and characteristic contrast-enhancement patterns should increase suspicion.
• The investigators recommend including GFAP antibody testing in CSF. GFAP testing is not included in every commercial panel, and serum testing alone has limited specificity.

Clinical Clues: AE Versus Other Causes

Seizures

• New-onset seizures at presentation occurred in 20 of 58 AE cases (34%) versus 9 of 89 non-AE cases (10%), with p less than 0.001.¹
• AE accounted for 20 of the 29 patients with RPD and seizures (69%).
• Seizures may be subtle. In the AE group, nonmotor, autonomic, dyscognitive, or faciobrachial dystonic seizures were common, and seizures were missed at the first presentation in 5 of 20 seizure-positive cases.
• A normal initial history does not exclude seizures. Obtain collateral history, review videos when available, and use EEG when suspicion persists.

Cerebellar and Brainstem Findings

• Cerebellar or brainstem findings were more common with non-AE diagnoses than with AE (21% versus 9%).
• Of the 19 non-AE patients with these findings, 11 (58%) had CJD. This is not the same as saying that 58% of all patients with CJD had these findings.
• Movement disorders alone did not clearly separate AE from CJD at first presentation. Timing was more informative: movement disorders tended to occur within 3 months in AE and later in CJD.

MRI

• Mesiotemporal T2/FLAIR hyperintensities were present in 30 of 57 patients with AE who underwent MRI (53%) versus 7 of 89 patients with other diagnoses (8%).
• Restricted diffusion was more common in non-AE conditions (18% versus 4%).
• Restricted diffusion involving at least 2 cortical regions or the caudate or putamen occurred in 7 non-AE patients and no AE patients; all 7 had CJD.
• These patterns were exclusive within this cohort, not universally pathognomonic. MRI must be interpreted with the clinical syndrome, ADC maps, EEG, CSF, and prion testing.

CSF and Laboratory Clues

• CSF pleocytosis and hyponatremia were more common in AE than in other diagnoses.
• Markedly elevated CSF total tau or a high total-tau-to-phosphorylated-tau ratio can occur in AE and should not be treated as specific for CJD.
• CSF real-time quaking-induced conversion (RT-QuIC) was positive only in CJD cases, but it was negative in 3 patients with autopsy-confirmed sporadic CJD. A negative result therefore does not completely exclude CJD when the phenotype and MRI remain compelling.

A Structured Diagnostic Approach

Immediate Assessment

• Establish the tempo, baseline function, medication exposure, toxic risk, immune status, malignancy history, recent infection, travel, and family history.
• Seek collateral history for subtle seizures, faciobrachial dystonic events, psychiatric change, fluctuation, sleep disturbance, autonomic symptoms, and rapid loss of activities of daily living.
• Stabilize seizures, delirium, electrolyte disorders, respiratory compromise, and other immediate threats while the etiologic evaluation proceeds.

Core Testing

• MRI brain with diffusion-weighted imaging, apparent diffusion coefficient maps, FLAIR sequences, and contrast when appropriate.
• EEG, with prolonged or continuous monitoring if subtle focal seizures or nonconvulsive status epilepticus remain possible.
• Lumbar puncture tailored to the differential:
  • Cell count, protein, glucose, cultures, and pathogen testing.
  • Oligoclonal bands and immunoglobulin G index when inflammation is suspected.
  • Paired serum and CSF neuronal and glial antibody testing.
  • CSF GFAP testing when the phenotype is compatible.
  • Prion studies, including RT-QuIC, when CJD is suspected.
• Blood testing should rapidly address common metabolic, toxic, endocrine, nutritional, infectious, and autoimmune causes according to the clinical context.
• Perform malignancy screening when paraneoplastic AE or another neoplastic process is plausible.

When Can Treatment Start Before Antibody Results?

• This cohort did not test an empiric-treatment strategy and cannot prove that all patients with suspected RPD benefit from treatment while antibody results are pending.
• The 2016 clinical AE criteria allow a provisional diagnosis without antibody results. Possible AE requires all of the following:²
  • Subacute progression over less than 3 months of working-memory deficits, altered mental status, or psychiatric symptoms.
  • At least one supportive feature: a new focal central nervous system finding, unexplained new-onset seizures, CSF pleocytosis, or MRI features suggestive of encephalitis.
  • Reasonable exclusion of alternative causes.
• Best-practice recommendations support acute immunotherapy when AE is highly suspected after basic CSF evaluation has reasonably excluded infection and when corticosteroids will not compromise a needed biopsy for primary CNS lymphoma or another inflammatory mimic.³
• Send antibody studies before treatment whenever feasible, but do not wait for results when the clinical criteria and risk-benefit assessment support treatment.
• Do not use a positive serum antibody alone as proof of AE. Interpret the antibody, specimen type, assay, phenotype, MRI, EEG, and CSF together.

Study Strengths

• Prospective and consecutive screening.
• Multicenter and nationwide referral pathways.
• Nearly universal paired serum and CSF antibody testing.
• Direct MRI review and consensus diagnosis by 3 specialist neurologists.
• Median follow-up approaching 3 years, which reduced the risk of premature etiologic classification.

Limitations and Generalizability

• Referral bias is substantial. Erasmus University Medical Center is the Dutch national center for AE and neuronal antibody testing, likely enriching the cohort for immune-mediated disease.
• Typical CJD cases may have been referred less often, creating selection toward atypical CJD and reducing the observed proportion of prion disease.
• The distribution of RPD causes varies by setting. Infection may dominate in other countries or resource settings, while neurodegenerative disease or CJD may predominate in specialized memory or prion centers.
• Only 8 patients underwent autopsy. Although clinicopathologic agreement was complete in those 8, most diagnoses lacked neuropathologic confirmation.
• The study focused on AE. It did not systematically include every biomarker useful for neurodegenerative disease, and formal neuropsychological testing was available in only a minority.
• Clinical and MRI associations are not validated standalone diagnostic rules. They should guide prioritization, not replace a complete differential diagnosis.

Practical Summary

• RPD is a diagnostic emergency because a large fraction of referred patients may have a potentially responsive cause.
• Search early for AE, infection, toxic-metabolic disorders, malignancy, epilepsy, vascular lesions, nutritional deficiency, and other reversible conditions while evaluating for CJD and neurodegenerative disease.
• Seizures, mesiotemporal MRI abnormalities, CSF pleocytosis, and hyponatremia raise suspicion for AE.
• Multiregional cortical or striatal restricted diffusion strongly raises suspicion for CJD, but no single test should be interpreted in isolation.
• Test paired serum and CSF for neuronal and glial antibodies, including CSF GFAP when clinically appropriate.
• Start immunotherapy before antibody results only when accepted clinical criteria are met, infection and major mimics have been addressed, and the patient-specific risk-benefit assessment supports treatment.

---

References

1. van Steenhoven RW, Bastiaansen AEM, Kerstens J, et al. Autoimmune Encephalitis as Treatment-Responsive Cause of Rapidly Progressive Dementia: A Multicenter Prospective Cohort Study. Neurology. 2026;106(11):e214933. doi:10.1212/WNL.0000000000214933.
2. Graus F, Titulaer MJ, Balu R, et al. A Clinical Approach to Diagnosis of Autoimmune Encephalitis. Lancet Neurology. 2016;15(4):391-404. doi:10.1016/S1474-4422(15)00401-9.
3. Abboud H, Probasco JC, Irani S, et al. Autoimmune Encephalitis: Proposed Best Practice Recommendations for Diagnosis and Acute Management. Journal of Neurology, Neurosurgery & Psychiatry. 2021;92(7):757-768. doi:10.1136/jnnp-2020-325300.

Educational lecture notes. RPD and suspected autoimmune encephalitis require urgent specialist evaluation. Diagnostic testing and immunotherapy should be individualized to the clinical syndrome and local resources.