seizures after an LP

[llogg]

yesterday a pt was admitted to me on my medicine service. 21yo AAF w/SLE presenting with 4 days of fevers (subjective), headache, myalgias, and nausea but no vomiting. She had had a missed abortion with a D&C two weeks prior. Gyn exam performed by ob-gyn ER resident normal. On my exam febrile to 38.4C, tachy to 100s, other vitals normal. Pt crying, +photophobia, neck stiffness, no papilledema, OP clear, single small mobile right ant cervical lymph node. Pt not cooperative with neuro exam, but no focal deficit noted. Labs unremarkable (CBC and basic metabolic panel). Other workup pending (C3/C4, ESR, cultures). I performed a lumbar puncture in sitting position (after lateral position was unsuccessful). No complications immediately after procedure, clear CSF obtained, no opening pressure as she was sitting and I was afraid she would move (she had jumped several times already). Two hours later called that patient is seizing. Seizure stopped spontaneously but pt given ativan and loaded on dilantin. Three hours later paged by cross cover to tell me she is continuing ot seize, having decorticate posturing and is being transferred to ICU. Did my LP cause this?

---

Members don't see this ad.

 

[Strokeguy]

Was no neuroimaging done, like CT or MRI ?

 

[typhoonegator]

I agree, it's a bit difficult to put all this together without a bit more info. Imaging would be useful, particularly with the posturing. However, it is a little hard to understand why the patient would cone herself from your LP five hours after you did it. Did she have any symptoms like diplopia, nausea, or vomiting immediately after the LP? Herniation syndromes also don't typically (or even atypically) present with seizure anyway. Did someone reliable actually see this "seizure", or could this have been something else, like asterixis or myoclonus? Also, what did the LP actually show? Was it bland CSF? Could she have a venous sinus thrombosis from possible APLAS and be seizing from that, with no contribution from your LP?

I've seen a case or two of patients dying or hemorrhaging after LP, usually because they have some other structural disease. It is usually fairly immediate and quite dramatic. I know of one patient who was LP'd by a non-neurologist and found to have a very high opening pressure, but still they took off four full tubes. The patient had a nearly immediate PEA arrest as her brainstem herniated downward. The case you have presented doesn't sound anything like that.

Don't be so quick to blame yourself. Besides, it sounds like this was a well-indicated diagnostic procedure for this patient. Imaging, or at least a very good funduscopic exam with positive venous pulsations, would have been nice to have gotten before the procedure.

 

[llogg]

Imaging: CT head w/o contrast was done and was read as essentially normal study.

At time of my initial post I did not have CSF studies back, but this is what they showed: Tube 1 -- 12 nucl cells, 120 rbcs; Tube 4 -- 10 nucl cells, 22 rbcs, 410 prot, 66 gluc.

While sweating it out this morning I felt I should have waited for the CT head, but at the time I felt she was low risk: young, no focal deficit, no papilledema.

She had no new symptoms immediately after the LP. She was AAOx3, no n/v, no visual complaints.

Seizure was apparently witnessed by her mom.

 

[typhoonegator]

That sounds like some thick and chunky CSF. How much did you take off? CSF with that high protein is so viscous, you can actually get a quick large pressure differential along the spinal column as you drain it, which can cause some...problems.

However (and again), it sounds like she has a real CNS process going on that you did not cause, which is likely causing the bulk of her presenting symptoms.

I'll leave the remaining workup of pleiocytosis and elevated CSF protein in the context of SLE to you, but at first blush it doesn't seem likely that your LP caused or exacerbated her neurologic deterioration. Imagine how bad you would feel if you hadn't tapped her and this process continued unabated.

 

[Strokeguy]

It does look like it was a coincidence that she had seizures after the LP and then developed decorticate posturing. I would also consider a urine drug screen. I have seen quite a few patients with cocaine induced neurologic insult. I would consider an MRI (may not always be possible) or at least a repeat CT (portable if available) to see if there was an evolution of the neurologic process not seen on CT previously or a different neurologic pathology (since seizures started 'a few hours' after the LP). Also most importantly, decortication or decerebration can also be seen in status epilepticus. It does look like she went into status. So these signs infact do not necessarily indicate herniation, but may be a consequence of status. I feel that even though she stopped 'seizing' for 3 hours after dilantin load, she was in 'nonconvulsive status'. So next important steps should be- CT, EEG, and further management as per status epilepticus protocol. I would also start empirical treatment for presumed meningitis with broad spectrum anti-biotics untill cultures results are available.

 

[llogg]

Utox was negative. MRI was done but I don't know the results. (she is now on the ICU service and I'm only following peripherally). EEG is also pending, though neurology consultant did not feel she was in status. Infectious disease consultant feels this may be listeria meningitis vs viral.

I had started her empirically on cefotaxime and vancomycin. The ICU team has added acyclovir and ampicillin. She is also getting dexamethasone now.

I took off between 8 and 12 cc's.

I'm feeling better about my role in all this. In fact, I'm shocked that I was the first one to stick a needle in her back with this presentation. She was a patient I picked up on our "short call" day, which means that an ER resident, an ER attending, the AOD medicine resident, and the night float medicine resident all saw her before I did. Maybe her exam continued to evolve over that time, but the presentation I got from the night float included fever, headache and neck stiffness. I think it was too easy for everyone to chalk this up to a lupus flare.

NOTE: I'm probably taking an undue amount of credit as it was my attending who suggested I tap her.

 

[llogg]

Here is the official MRI read, the comparison is to a study from two weeks prior:
1. Interval development of multiple areas of abnormal T2 hyperintensity
involving the cortical white matter and bilateral cerebellar hemispheres.
Questionable enhancement of the subarachnoid space is also seen. Several of
these areas of abnormal signal also reveal restricted diffusion. These
findings are nonspecific, however, leading considerations include
vasculitis, meningitis with developing cerebritis, atypical ADEM, and less
likely posterior reversible encephalopathy syndrome.
2. MRA of the head is essentially nondiagnostic with multiple
irregularities in the vessels at all be explained by motion artifact.
3. MRA of the neck is unremarkable.

 

[Strokeguy]

So imaging did show interval development of a neurologic pathology. Restricted diffusion needs more clarification. It is important in this situation to differentiate this as cytotoxic or vasogenic edema (on the basis of DWI and ADC maps). This can be useful to differentiate conditions as reversible post leucoencephalopathy (vasogenic), cerebritis (initially vasogenic) from ADEM (cytotoxic). Vasculitis could produce both.
Patients with uncontrolled seizures can also produce same findings and conversely seizures can be the manifestation of RPLE.
The EEG will also be very important in this case. If the first is negative or shows diffuse slowing, it will be very worthwhile to repeat it in the next few days. It would be very important to control any seizure activity.
I saw two patients during residency, one was treated for cryptococcal meningitis and subsequently developed vasogenic edema in the posterior areas which evolved into a cryptococcal abscess. The other patient had lupus but was subsequently treated as focal cerebritis from vasculitis.

 

[neurodoc]

Cerebritis due to her SLE could explain the seizures. Without a mass on imaging, I doubt that the LP caused her to cone down. That very high protein is a problem. Such high protein levels raise the question of a CSF block. If no structural cause of block is apparent on the brain MRI, you need to start looking at lower levels along the spinal neuroaxis.

 

[cliff]

what were coags before LP?

 

[Docxter]

Strokeguy,

I've noticed from this post and another older one that you haven't fully gotten the diffusion imaging terminology down yet.

Restricted diffusion needs more clarification.
It is important in this situation to differentiate this as cytotoxic or vasogenic edema (on the basis of DWI and ADC maps).

Actually it does not need more clarification. For all clinical intents and purposes in this context, restricted diffusion is indicative of cytotoxic edema. It seems that you erroneously equate the term restricted diffusion with high signal on DWI, and you're asking for clarification by looking at the ADC map. That is incorrect. You should only call something restricted (decreased) diffusion if it is both high signal on DWI and low on ADC. In other words the term "restricted diffusion" automatically implies both high DWI and low ADC, hence cytotoxic edema for most intents and purposes.

I said for all clinical intents and purposes because I don't want to get into all the different exceptions, confounders and all the different theories behind decreased brownian motion detection by the Stejskal-Tanner family of MR pulse sequences.

This can be useful to differentiate conditions as reversible post leucoencephalopathy (vasogenic), cerebritis (initially vasogenic) from ADEM (cytotoxic).

These are all generally true in most cases, but you need to realize that PRES/RPL can sometimes give areas of cytotoxic edema (the partially nonreversible cases) and ADEM does show cytotoxic edema in a large proportion of cases.

 

[danielmd06]

I'm going to hold my breath and voice a dumb question here for Docxter or Strokeguy.

I ask that all forgive the temporary hijacking of the thread.

Can one of you please explain the techincal difference between vasogenic edema and cytotoxic edema...how they are differentiated on MRI sequencing and which pathologies they respectively appear with (not just those mentioned above).

I am happy to look them up myself, but figured you guys could give me a much more terse and cleaner outline to keep in mind.

Thanks in advance.

 

[Strokeguy]

I would agree with Docxter with the 'definition' of restricted diffusion. In my post I mentioned the clinical entities which produce cytotoxic or vasogenic edema or both. I did also mention that ADEM produces cytotoxic edema.
With regards to RPLE, it is 'vasogenic' in most cases. Some may have cytotoxic edema. We have to see the etiology of HTN encephalopathy - high blood pressure and exceeding the upper limit of cerebral autoregulation. High BP is also a risk factor for cerebral ischemia. Some of these pts may have small infarcts (cytotoxic edema) as a consequence of high BP which has also produced RPLE. The cause-effect relationship then becomes more controversial and this is a well accepted fact among most vascular neurologists. It is not surprising to see the 'non reversible' forms of RPLE to have cytotoxic edema. From a clinical perspective, the longer you take to treat the cause (HTN enceph or eclampsia in most cases), worse becomes the prognosis. Sustained high BP in such cases produces ischemia (a vicious cycle).
We have to realize that no test is absolute (except for confirmation at autopsy ??). If a test is 95% sensitive and specific it is clinically useful.
The other issue is how do we use the currently available radiologic technology to make clinical decisions. Then we have to go by the more commonly accepted concepts and not exceptions.
The most common differential diag of HTN enceph is ischemic stroke, which may need TPA and certainly requires BP to be maintained at higher values to maintain cerebral perfusion; unlike HTN enceph which would worsen with this approach (and has higher risk of bleed with TPA).
A lot of radiology has to be evaluated with the clinical perspective. There are a number of neurologic diseases producing cytotoxic or vasogenic edema or both (these by the way are neuropathologic entities). In addition there is also interstitial edema from acute hydrocephalus.