Tell me your anti-NMDA receptor encephalitis stories!

[wellbutrin.girlfriend]

Advertisement - Members don't see this ad

Hi all, trainee here wondering about this community's clinical anecdotes on anti-NMDA receptor encephalitis (or really any good stories of primary neuro/other medical problems masquerading as primary psych).

What presentations made you suspicious that your ostensible schizophrenia spectrum disorder etc patient had something more going on? Pls feel free to brag about your superior clinical intuition here-- I'm all ears!

 

[splik]

As a neuropsychiatrist, I've seen many, many cases of autoimmune encephalitis of various stripes as well as other autoimmune causes of psychiatric symptoms. The last time I checked, I had seen a wider range of autoimmune cases (in terms of underlying pathology) than most even in neuropsych. It isn't my main specialty and there are some people who do specialize in this population. As a resident, I diagnosed one of my outpatients with neuropsychiatric lupus, and a patient in the ED with voltage-gated potassium channel encephalitis. I won't go into all these cases but will share some clinical pearls that will hopefully be helpful after seeing these patients for over 10 years.

1. NMDA-R encephalitis became a bit of a faddish diagnosis and there were some people who were just testing for this without doing a full workup. That is not appropriate. If you think the patient might have an autoimmune or other cause for their symptoms as indicated by their presentation (e.g. subacute course, episodic memory loss, seizures, MRI findings) or risk factors (e.g. prior HSV infection, past COVID, autoimmunity, malignancy) then they need to have a proper workup considering all the potential causes autoimmune or otherwise. I seem to be seeing less of this in recent years at least.

2. Something like 5% of patients with schizophrenia have positive antibodies and these are innocent bystanders in the majority of cases. Positive serum labs are not necessary pathogenic, whereas if found in the CSF they are. Fewer than 1% of these patients have positive CSF antibodies.

3. Every patient I have seen with NMDA-R encephalitis has had catatonic features, often overlooked or eclipsed by their other symptoms (e.g. seizures, coma, delirium). If you read cases reported in the literature, there is usually catatonia often undiagnosed. Catatonia, rather than delusions or hallucinations, should be the key feature that would prompt further workup.

4. It would be very unusual for these patients to have normal cognitive testing. Isolated psychiatric symptoms with normal cognitive testing would be highly unlikely to be an autoimmune etc case. At the minimum ask your patients to draw a clock!

5. Delusions of misidentification have a higher pre-test probability of being related to an underlying neurological condition.

6. Patients can respond very well to psychotropic medications (or ECT) without immunotherapy. This can sometimes lead the diagnosis to be missed until the patient relapses or dramatically deteriorates.

7. Patients often need high doses of psychotropics in addition to immunotherapy. Some people have this fantasy that you just treat the underlying condition and all will be well, but that is not the case. If it were I'd be out of a job.

8. Be careful with antipsychotics, especially high potency neuroleptics, as these patients are at high risk of NMS

9. A paraneoplastic condition can precede the onset of detectable malignancy by years in some cases. On my cases with paraneoplastic cerebellitis was diagnosed with lymphoma only years after her symptoms began.

10. These auto-antibodies not only affect the limbic system. Neuropsychiatric presentations can also occur with antobodies affecting the basal ganglia, brainstem, and cerebellum. With the basal ganglia, you can see significant hypersomnia and akinetic mutism, with the brain stem you can see prominent visual hallucinations and delirium, and with the cerebellum you can see frank psychosis, emotional lability and executive dysfunction in addition to traditional cerebellar signs.

11. I have seen many patients who have been misdiagnosed with an autoimmune condition who actually had a personality disorder or hysterical condition. Treatment is not benign and can hamper real diagnosis, put patients in enormous debt, and cause significant harm. I had patients develop hyperosmolar coma and steroid-induced mania from steroids they should never have received just because of an irrelevant antibody titre.
12.Nowadays, this diagnosis is often sought after by patients and families. It has become part of the PANS/PANDAS family for some groups who have a loved one with hysterical symptoms or symptoms due to severe family pathology. I had one patient with factitious disorder by proxy who was receiving unnecessary IVIG and rituximab. On the other side, some patients who have a primary psychotic disorder end up getting treatment because the family would rather they have a "neurological disorder."

13. Bear in mind diagnostic overshadowing too - it is possible for patients to have BOTH a primary psych disorder (e.g. bipolar or schizophrenia) AND develop autoimmune encephalitis. Similarly, it is possible for patients to have functional neurological symptoms in the setting of AE. It can sometimes be hard to tease apart things, and some patients may have a signficant unrelated psych history. Some patients (or more often, family members) may wish to ascribe everything to some antibodies, but the reality is often more complicated.

14. Seronegative autoimmune encephalitis is a thing. However if the patient only has psych symptoms with normal or non-specific MRI findings, bland CSF, no seizures, no cognitive symptoms etc then they don't have a compelling profile for seronegative AE. Some people think the dx means that we should just presumptive treat classic psych issues as if they could be seronegative AE which is misguided.

15. The APE-2 score is really a very useful tool in assessing likelihood a patient has AE.

16. People often have this idea that patients magically get better with immunotherapy, and the reality is in most cases it is a long recovery, often requiring rehab and risk of relapsing symptoms (unless there is a clear tumor etc that can be resected). Neuropsych symptoms can also persistent even when the disease has been effectively treated with immunotherapies. Just as with viral encephalitis, then can be lasting damage to the brain. That said, it is a much better diagnosis to have in terms of treatability than a lot of the neurodegenerative disorders that can present w/ psychiatric symptoms.

 

[Merovinge]

APE-2 is new since I was in training, really appreciate that reference. I don't imagine I will see much of this in the adolescent space, but always great to know about.

 

[splik]

APE-2 is new since I was in training, really appreciate that reference. I don't imagine I will see much of this in the adolescent space, but always great to know about.

It's not common, but NMDA-R encephalitis is more common in younger patients with a significant minority being under 18. I'm not CAP of course but I've still be consulting on some CAP cases (including where ECT was needed to treat the catatonia) so definitely something to be aware of in this population. We also had some kids develop autoimmune encephalitis post-COVID during the pandemic. Also, it's helpful to be able to explain to parents why you don't think their kid has limbic encephalitis if they are wanting that diagnosis. On the pediatric side, I feel peds neuro is far more willing to give ritux and IVIG to kids with nebulous symptoms (especially if they have wealthy white parents) compared to the adult neurologists. I come across this issue when the pts have to transition over with age. Then there's the PANS/PANDAS crowd* and this has become lumped in. Honestly, it's the new chronic lyme disease.

*I do believe there are legit PANS/PANDAS cases but there is no doubt that in a sizable majority of cases, general family dysfunction etc gets wrongly labeled as such.

 

[Merovinge]

*I do believe there are legit PANS/PANDAS cases but there is no doubt that in a sizable majority of cases, general family dysfunction etc gets wrongly labeled as such.

Can you tell me more about legit PANDAS cases? I almost never see a case where IVIG correlates directly to change in OCD symptoms and I see it as a way to avoid ExRP/SSRI treatment in affluent white families. Interestingly, I have never seen a non-white patient report a history of PANDAS. I can think of a single patient where IVIG was very helpful, but in this case it appeared that symptoms would flare with any infection rather than anything strep related and they had already had 3-4 discrete rounds of treatment with later rounds having less prominent improvements.

And yes certainly it is possible to have NMDA-R/autoimmune encephalitis in kids. My organization, despite doing PHP/IOP work, does find kids who are catatonic on a not completely irregular basis and something I need to keep in mind. I have never seen this actually treated with IVIG but I have seen plenty of reported PANDAS patient's who receive IVIG from peds neuro. I had excellent neuropsych training in adult residency but it's been 10+ years so I love getting this type of refresher on SDN .

 

[Wilf]

...to avoid ExRP/SSRI treatment in affluent white families. Interestingly, I have never seen a non-white patient report a history of PANDAS.

Interesting. What do you and/or splik believe is the reason for this?

In child training, I had few patients' parents try to go for the PANS/PANDAS diagnosis but one of the few cases I do recall was an African American mother/patient.

 

[clausewitz2]

Interesting. What do you and/or splik believe is the reason for this?

E/RP is extremely uncomfortable for the patient undergoing it, is time intensive, and requires a fair amount of change on the part of the family to not accommodate the patient. It asks parents who want to do the humane, empathetic thing for the child to stop doing that and let them be distressed without trying to fix it. Many people have trouble tolerating their children's distress and so if there is another explanation, particularly if it is pitched as "but your child's problem is special and not like those mentally ill people", it's not surprising people opt for it.

 

[oliversacks4thewin]

Thank you to all who have posted, the information is wonderful!

I happen to work as a consulting psychologist in the interconnection between these areas and have worked with neuroimmunology collaboratively on these types of neuropsychiatric cases. My work has been more in the cognitive assessment portion and some associated supportive therapeutic practices for recovery. All this to say, yes I have also seen these confirmed cases (and the preponderance of providers who too quickly leap to PANDAS, PANS, and other autoimmune related conditions), but will defer to my more expert colleagues in this thread.

I will say that sleep is a real big thing to also be checking. Anti-NMDA-RE has a real interesting sleep presentation and in the two most recent cases I encountered, this was one of the defining things that got us all thinking about the workup for it. We got the spidey sense, when the young ladies (yes, both below 18 years old) told us they were sleeping like 30 minutes to an hour a night, consistently, for way too long. We also had one whose initial presentation correlated with a real high fever associated with psychiatric and cognitive symptom manifestations, so it made it even more clear, which direction to take the evaluation.

Quick reference link to the sleep pattern:

Sleep disorders in anti-NMDAR encephalitis - PubMed

Sleep disturbances are frequent in anti-NMDARe. They show a temporal pattern (predominantly insomnia at onset; hypersomnia during recovery), are associated with behavioral and cognitive changes, and can occur with confusional arousals during NREM sleep.

pubmed.ncbi.nlm.nih.gov

 

[Wilf]

E/RP is extremely uncomfortable for the patient undergoing it, is time intensive, and requires a fair amount of change on the part of the family to not accommodate the patient. It asks parents who want to do the humane, empathetic thing for the child to stop doing that and let them be distressed without trying to fix it. Many people have trouble tolerating their children's distress and so if there is another explanation, particularly if it is pitched as "but your child's problem is special and not like those mentally ill people", it's not surprising people opt for it.

Haha naughty calusewitz! Purposely not answering my question and instead giving an explanation of E/RP, and parents' reaction to it, that a resident or fellow would know. 🤣

What I was asking is for splik's or merovinge's thoughts on why it is only white parents of patients that do this, which is obviously dubious. I could understand a correlation with SES, but I do not see why it would correlate with race.

 

[comp1]

I've seen it once. This is a really hard one because it's definitely real, but geeze do the tests involved have poor specificity AND sensitivity.

 

[Merovinge]

Haha naughty calusewitz! Purposely not answering my question and instead giving an explanation of E/RP, and parents' reaction to it, that a resident or fellow would know. 🤣

What I was asking is for splik's or merovinge's thoughts on why it is only white parents of patients that do this, which is obviously dubious. I could understand a correlation with SES, but I do not see why it would correlate with race.

I don't have any knowledge of why that is and it could just be small n (although I do work, and have worked, in very racially diverse areas except for a 3 year stint in rural America). It is likely due to a combination of SES and the ALGORITHM deciding on something different for my white patients' families vs PoC patients's families. There could be cultural reasons why white people are more comforted with autoimmune/neurologic vs psychiatric issues. I do believe (but am not a rheumatologist) that white people have higher incidents of some autoimmune conditions.

 

[Jules A]

I’ve seen a few on CL service. Most recent late 30’s female, no psych or medical hx, educated, employed to ED with disorganization, poor self care, agitation and perceptual disturbances. She was admitted to inpatient psych and almost immediately began having seizures. Off to medicine she went.

 

[Stagg737]

I have too many stories to tell. I'm at a large academic center and I see several Anti-NMDAr/autoimmune cases per month. We currently have 3 on our service right now including an anti-LG1. I'm not an expert, but I'm surrounded by a department of them and we have an autoimmune encephalitis clinic here where patients follow longitudinally after treatment. We see quite a few seronegative AIE cases as well and work with our neuroimmunology department and fellows pretty regularly. I'll post some cases later, but some quick thoughts from stuff above.

A few things I'd disagree with:

3. Every patient I have seen with NMDA-R encephalitis has had catatonic features, often overlooked or eclipsed by their other symptoms (e.g. seizures, coma, delirium). If you read cases reported in the literature, there is usually catatonia often undiagnosed. Catatonia, rather than delusions or hallucinations, should be the key feature that would prompt further workup.

While catatonia is very common with AIE (likely somewhere between 50-60% from larger studies) it is not uncommon to see patients with AIE without catatonia. However, I do agree that catatonia without an obvious cause (psychiatric or medical) should raise concerns for possible AIE.

6. Patients can respond very well to psychotropic medications (or ECT) without immunotherapy. This can sometimes lead the diagnosis to be missed until the patient relapses or dramatically deteriorates.

I don't think I've seen any cases of verified AIE where a patient got better without steroids or immunotherapy. I've seen some cases where we did ECT and patients improved enough that certain neurologists here said, "See, it was psych all along!", but Idk that I've seen anyone with verified AIE improve without direct treatment. As you mentioned above, catatonia is very common with AIE and ECT should absolutely be considered, but data as to whether catatonia actually benefits the underlying AIE is still dubious, though imo is worth plenty of further investigation. I've tried to see if there was any possibility of this where I'm at, but our PACU/anesthesia department already limits us severely in terms of how much ECT we're allowed to do, so I've basically been told there's no chance it would happen.

7. Patients often need high doses of psychotropics in addition to immunotherapy. Some people have this fantasy that you just treat the underlying condition and all will be well, but that is not the case. If it were I'd be out of a job.

Yes and no? Ime the psychiatric symptoms are much like treating delirium. Some patients do require significant psychotropics to manage symptoms and treating the underlying issue doesn't magically make them better. However, I'd question what it means when we say they "need" psychotropics other than agitation management.

15. The APE-2 score is really a very useful tool in assessing likelihood a patient has AE.

APE-2 is fine, certainly a decent tool if one isn't using anything, but RITE-2 has much better specificity and is nearly as sensitive.


Things I agree with that I'd emphasize further:

If you think the patient might have an autoimmune or other cause for their symptoms as indicated by their presentation (e.g. subacute course, episodic memory loss, seizures, MRI findings) or risk factors (e.g. prior HSV infection, past COVID, autoimmunity, malignancy) then they need to have a proper workup considering all the potential causes autoimmune or otherwise. I seem to be seeing less of this in recent years at least.

Just ordering an autoimmune panel and shipping it off to Mayo isn't an appropriate work-up. There needs to be a full reversible-cause of encephalopathy work-up which includes EEG monitoring, MRI w/ and w/o contrast, nurtitional panel, and LP at the very least. We let neuro handle this where I'm at, but we follow these patients closely.

2. Something like 5% of patients with schizophrenia have positive antibodies and these are innocent bystanders in the majority of cases. Positive serum labs are not necessary pathogenic, whereas if found in the CSF they are. Fewer than 1% of these patients have positive CSF antibodies.

Serum labs don't matter and are pointless other than confirming you should get an LP. Anyone where we legitimately suspect has AIE gets tapped. Period.

4. It would be very unusual for these patients to have normal cognitive testing. Isolated psychiatric symptoms with normal cognitive testing would be highly unlikely to be an autoimmune etc case. At the minimum ask your patients to draw a clock!

Completely agree but will add a caveat. I've had 2-3 patients who we (psych) was convinced should get a full AIE work-up but neuro pushed back because EEG and MRI were negative and then patient had several hours or a day where they were lucid before becoming encephalopathic again. Neuro was only convinced to tap them after we started steroids and patient improved. Every one of those cases came back seropositive. That said, even those patients were encephalopathic >95% of the time.

8. Be careful with antipsychotics, especially high potency neuroleptics, as these patients are at high risk of NMS

Not only NMS, but may worsen psychotic symptoms as well. Classic AIE patient is tried on 2-3 antipsychotics and just seems to get worse on them (especially those which are highly anti-dopaminergic like risperdal or haldol) paradoxically. Antipsychotics worsening psychosis should be a huge red flag.

13. Bear in mind diagnostic overshadowing too - it is possible for patients to have BOTH a primary psych disorder (e.g. bipolar or schizophrenia) AND develop autoimmune encephalitis. Similarly, it is possible for patients to have functional neurological symptoms in the setting of AE. It can sometimes be hard to tease apart things, and some patients may have a signficant unrelated psych history. Some patients (or more often, family members) may wish to ascribe everything to some antibodies, but the reality is often more complicated.

I've seen this too and it is not only very hard to untangle the co-morbidities, but it's also often very hard to convince neurology that it's not a pure psych problem.

14. Seronegative autoimmune encephalitis is a thing. However if the patient only has psych symptoms with normal or non-specific MRI findings, bland CSF, no seizures, no cognitive symptoms etc then they don't have a compelling profile for seronegative AE. Some people think the dx means that we should just presumptive treat classic psych issues as if they could be seronegative AE which is misguided.

Imo it's probably a much more common thing than we realize. When I first worked with consults in 2019, Mayo's AIE panel only had 6 auto-antibodies on it. Today the panel has around 30 auto-antibodies they test for and they've added 4 more since the last time I met with the Mayo Lab reps last fall. However, per a recent meta-analysis I saw (will try and find it again) 95% of seronegative cases have significant clinical symptoms and other findings (MRI, EEG, etc). If the clinical signs aren't there, pushing for aggressive treatment should be questioned.

16. People often have this idea that patients magically get better with immunotherapy, and the reality is in most cases it is a long recovery, often requiring rehab and risk of relapsing symptoms (unless there is a clear tumor etc that can be resected). Neuropsych symptoms can also persistent even when the disease has been effectively treated with immunotherapies. Just as with viral encephalitis, then can be lasting damage to the brain. That said, it is a much better diagnosis to have in terms of treatability than a lot of the neurodegenerative disorders that can present w/ psychiatric symptoms.

Some patients do make great, full recoveries, but it rarely resolves quickly and often takes months or over a year to fully recover. Also agree that some patients also never fully recover. Plenty of the patients in our f/up clinic still have sequelae 3-4 years after resolution of symptoms (I'm not aware of patients we have who recovered further in the past than that as the clinic here has only been around a few years).

 

[Stagg737]

A few clinical pearls other than what Splik has mentioned above that should raise suspicion:

1. Atypical or bizarre clinical symptoms, especially neurological findings. New and unexplained seizures are a big red flag. Acute psychosis where the patient has good insight that what they're experiencing isn't real and is significantly distressing raises concerns. Along with that if the agitation seems to be related to severe fear (not panic/anxiety, legitimate terror) it's worth asking more questions.

2. Rapid onset and progression. Almost every AIE patient I've seen had an acute or subacute (max a couple of weeks) onset of significant symptoms. First break mania can happen fast, but first break psychosis typically has the smoldering negative symptoms for months before the obvious positive symptoms hit. If there's no prodrome or obvious inciting factor for the psychosis (drugs, medications, trauma, severe illness, etc) and it just comes out of nowhere over a few days or a week, that's a big red flag. This is not specific for AIE, but should warrant a medical work-up before shipping someone to inpatient psych.

3. Relevant medical co-morbidities. Obviously, cancer as Splik said above but also other autoimmune disorders. A good number of our AIE patients have other chronic autoimmune or endocrine-like conditions long before developing AIE.

4. Maybe the biggest pearl/flag, if you start them on a strong dopamine blocker and the psychosis gets dramatically worse over the next 24 hours it should be a huge red flag. Neuroleptics may not always be effective for schizophrenia or primary psych problems and can certainly have significant side effects, but they should NOT make psychosis significantly worse.

5. This is completely anecdotal and I haven't found any literature on it, but one of our med/psych residents noticed that several of our AIE patients who came back positive all had elevated serum IgG levels even several months before AIE symptoms hit. We've seen enough cases of it that one of the future neuroimmunology fellows is actually looking at starting a research project with our resident on this. Curious if anyone else (ahem, @splik ) has seen or heard of this.

6. Early treatment matters. Metanalyses have shown that patients who do not receive treatment within the first MONTH of symptom onset had significantly worse 2 year outcomes and that treatment 12 months after symptom onset was basically completely ineffective. It also suggests that delay of immunotherapy of >1 month is the 2nd strongest indicator of a poor outcome, with infra-tentorial involvement on MRI being more likely to have worse 2-year outcomes (even more than neurological findings including refractory status epilepticus).

Seronegative autoimmune encephalitis: clinical characteristics and factors associated with outcomes - PubMed

Seronegative autoimmune encephalitis is autoimmune encephalitis without any identifiable pathogenic antibody. Although it is a major subtype of autoimmune encephalitis, many unmet clinical needs exist in terms of clinical characteristics, treatments and prognosis. In this institutional cohort...

pubmed.ncbi.nlm.nih.gov

Like a lot of other things in psych and medicine, I've found that after you've seen enough cases you also just kind of start to get a "feel" for when something isn't right. Imo encephalopathy in general typically looks VERY different on the medical floors than psychosis, and sometimes other non-psych specialists seem to be very bad at differentiating between the two. There's also certain times when the encephalopathy just feels atypical and when you sit with the patients long enough it's different from typical delirium or other common encephalopathies we see.

Final thought for this post, while AIE is certainly rare and not something that we should be jumping to when we hear hooves, imo it's much more prevalent than we ever expected it would be when we first started to learn about this. Imo it's not a unicorn diagnosis, and if you work in large medical center you're almost certainly going to see a few of these patients over the course of your career.

 

[Merovinge]

Thank you @Stagg737 and @splik for such extensive discussion of this important topic and frankly for making psychiatry look good for anyone who reads this in passing. Neuropsychiatrists are always such ballers, I have never met one I didn't have a great deal of respect for.

 

[Wilf]

To be clear, by neuropsychiatrist you all are referring to folks who did a combined neurology/psychiatry residency, correct?

Stagg and splik, did you both do this combination? God bless you if so. I could not imagine doing such, but it would be quite helpful to a subset of patients.

 

[Stagg737]

To be clear, by neuropsychiatrist you all are referring to folks who did a combined neurology/psychiatry residency, correct?

Stagg and splik, did you both do this combination? God bless you if so. I could not imagine doing such, but it would be quite helpful to a subset of patients.

I did not, I just work closely with our neurologists on C/L and have a special interest in autoimmune and “hormonal” psychiatry which we see quite frequently due to our size and status as the last stop before Mayo for several states. I also did residency at a program that emphasized robust medical training as well as we had a very strong med/psych presence there (full program with numerous attendings being med/psych trained and dual boarded).

Becoming a “neuropsychiatrist” typically involves combined training or doing a neuropsychiatry fellowship. Fellowships may also be called cognitive neurology or behavioral and cognitive neurology. Most are accredited by UCNS and are through neurology departments, but a few are through psychiatry departments. Ours is through neurology, but we have a separate autoimmune neurology fellowship here that handles many of these cases with us.

 

[Merovinge]

To be clear, by neuropsychiatrist you all are referring to folks who did a combined neurology/psychiatry residency, correct?

Stagg and splik, did you both do this combination? God bless you if so. I could not imagine doing such, but it would be quite helpful to a subset of patients.

Neuropsych fellowship after psychiatry. Can also come at it from neuro residency and then a different behavioral neurology fellowship. I spent infinitely more time with the former than the later since they are psychiatrists first. For whatever reason neuropsych seems to attract people who are cognitively inclined enough for neurology but can handle the ambiguity of psychiatry. It's not easy finding huge nerds who can handle uncertainty, but when you do, it's wonderful.