Another item of interest in Ms. L’s case is her parents’ request for a Neurology consultation and further workup, as there is an association between caregiver request for workup and eventual diagnosis.6 While the etiology of this phenomenon is unclear, the literature suggests individuals with autoimmune encephalitis who initially present to Psychiatry experience longer delays to the appropriate treatment with immunomodulatory therapy than those who first present to Neurology.20
Laboratory and diagnostic testing
Guasp et al2 recommend EEG, MRI, and serum autoimmune antibodies (ie, screening for anti-NMDA receptor antibodies) for patients who present with first-episode psychosis, even in the absence of some of the red flags previously discussed. A recent economic analysis suggested screening all patients with first-episode psychosis for serum antibodies may be cost-effective.21 Since there can be false positives from serum testing, a positive result should be followed by CSF testing. Serum antibody testing will miss cases where anti-NMDA receptor antibodies are present only in CSF, which is why Guasp et al2 recommend ancillary screening with EEG and MRI.Screening all first-episode psychosis patients with EEG and MRI would represent a major change to psychiatric practice and would be beyond the current practical capabilities of many facilities that treat people with new-onset psychosis. Additional evidence is needed before such a change would be required. These suggestions are supported by studies that found most patients with anti-NMDA receptor encephalitis do not initially present with focal neurologic findings, though the majority (95%) do have EEG abnormalities.2,20,22
For patients whose presentations include features concerning for anti-NMDA receptor encephalitis, an EEG and MRI are reasonable. In a review of EEG abnormalities in anti-NMDA receptor encephalitis, Gillinder et al23 noted that while 30% did not have initial findings, 83.6% of those with confirmed anti-NMDA receptor encephalitis demonstrated EEG abnormalities; the most common were generalized slowing, delta slowing, and focal abnormalities. Discovering an extreme delta-brush activity on EEG is specific for anti-NMDA receptor encephalitis, but its absence is not fully informative. Practically, slowing can be a nonspecific manifestation of encephalopathy or a medication effect, and many people who present with first-episode psychosis will have recently received antipsychotics, which alter EEG frequency. In a study of EEG changes with antipsychotics, Centorrino et al24 found that generalized background slowing into the theta range across all antipsychotics was not significantly different from control participants, while theta to delta range slowing occurred in 8.2% of those receiving antipsychotics vs 3.3% of controls. Clozapine and olanzapine may be associated with greater EEG abnormalities, while haloperidol and quetiapine contribute a lower risk.25 For young patients with first-episode psychosis without a clear alternative explanation, we advocate for further autoimmune encephalitis workup among all individuals with generalized theta or delta wave slowing.
Because these medication effects are most likely to decrease specificity but not sensitivity of EEG for autoimmune encephalitis, a normal EEG without slowing can be reassuring.26 Moreover, for patients who receive neuroimaging, an MRI may detect inflammation that is not visible on CT. The concerning findings for anti-NMDA receptor encephalitis are temporal or multifocal T2 hyperintensities, though the MRI is normal in most cases and thus should not be reassuring if other concerning features are present.27
The role of lumbar puncture
Another area of active debate surrounds the usefulness and timing of LP. Guasp et al2 proposed that all individuals with first-episode psychosis and focal neurologic findings should receive LP and CSF antineuronal antibody testing. They recommend that patients with first-episode psychosis without focal neurologic findings also should receive LP and CSF testing if ≥1 of the following is present:
- slowing on EEG
- temporal or multifocal T2 hyperintensities on MRI
- positive anti-NMDA receptor antibody in the serum.2
Continue to: Evidence suggests that basic CSF parameters...