Complications of carotid revascularization: Hyperperfusion syndrome

Your latest CEA patient – a left ICA embolic stroke with mild  sensorimotor disturbance – deteriorates neurologically 2 days after the procedure – what is the differential?

We came up with the following reasons for late complications:

  • Hyperperfusion syndrome
  • Yet another stroke from a different source (the presumed etiology wasn’t correct)
  • Hemodynamic instability because of restenosis or occlusion of the vessel
  • Embolic complication after the fact (rare)
  • Intracerebral hemorrhage into the original stroke
  • Seizure

So how do you deal with that? Of course, you get the history and do a physical. But then?

  • Ultrasound of the ICA – is it patent?
  • TCD of the media on both sides: is there hyperperfusion (raised mean systolic velocity?) or hypoperfusion (e.g. distal stenosis) or occlusion of the MCA due to embolic stroke?
  • CT: stroke? ICH? signs of hyperperfusion?
  • EEG: PLEDs? Status?

If you cannot do emergency ultrasound turn the CT order into CT+CTA+CTP or – even better (and easier if the original procedure was done by your neuroradiologist) – MRI with PWI.

Next we talked about the hyperperfusion syndrome – or reperfusion syndrome (sounds better to me).

Here are the most important facts about it:

  • Distinguish between hyperperfusion (raised CBF or – as surrogate marker – mean systolic velocity) and the hyperperfusion syndrome (headache, confusion, seizures, neurologic deterioration, in the worst case ICH)
  • Pathophysiology: you need raised CBF (due e.g. to hypertension, often due to baroreflex disturbance) and autoregulatory failure, which leeds to endothelial damage and vasogenic, later cytotoxic edema, just as in PRES.
  • Stent vs. TEA: Probably similar rates of hyperperfusion, but more ICHs with stents due to double platelet inhibition and balloon dilatation damage to the baroreflex. Thus in high risk patients (see below), you might think about balloon-free-stents (a recent hype)
  • Predisposing factors are:
    • Hypertension pre-procedure
    • Blood pressure changes during the procedure
    • Hypertension post-procedure
    • Diabetes mellitus
    • Age
    • S/p contralateral CEA
    • High grade stenosis (poststenotic hypotension) with or without
    • Bad collaterals (incomplete Circulus Willisii)
    • Volatile anesthetics?
    • Perioperative stroke
    • Symptomatic status
    • Anticoagulation, platelet inhibition
    • Impaired baroreflex [since this often coincides with impaired CO2 reactivity, many such patients have Cheyne-Stokes-respiration, so the latter might be a risk factor]
  • Predictive numeric parameters
    • MCA ipsilateral < 75% of contralateral side
    • MCA normal, yet impaired autoregulation in acetazolamide test
    • Low (< 40 mmHg) distal ICA pressure
    • Cerebral transit time > 2,7sec (time for contrast from distal ICA to cortical veins)
    • Impaired baroreflex (might be sensible to screen this)
  • Diagnosis: Headache + Hyperperfusion. The most useful parameter apart from the above seems to be TCD about 2h after the procedure
  • Epidemiology: roughly 1-2% of carotid revascularizations, about 0,5% bleed… Interestingly, stenting patients seem to develop hyperperfusion earlier (< 36h) than CEA patients (3-7d).
  • Prevention: Statins preop might be an option (no real evidence yet, but some rationale and a recent retrospective study), RR < 140 mmHg, avoid predisposing factors, screen high risk patients with TCD 2h after the procedure
  • Therapy:
    • Transfer to ICU, arterial line
    • RR-goal < 100 mmHg (this is my personal recommendation, no evidence), using Urapidil/Labetalol or Clonidine, or even Dexmedetomidine, if delirium is bad. Next line agents would be Esmolol/Metoprolol. Enalapril iv might work, but could be dangerous. Dihydralazine, Ca antagonists, Nitro derivates should be avoided.
    • If neurologic symptoms come up, consider Mannitol and steroids (again, no evidence)

References: There are many reviews on the topic, but not much has happened in the last two years.


Yet another delirium

Your stroke patient changes his mental status. As for instance these 2 recent studies (Neurology 2011, Stroke 2012) tell us, delirium is quite frequent in stroke, yet it is not clear (to me) whether it is more or less frequent in stroke than in other hospitalized patients, of which about 30% might develop altered mental status, aka delirium.

The german terminology tends to equate “delirium” with the american terminus “hyperactive delirium”, i.e. altered mental status with vegetative and psychomotor activation, leaving the “hypoactive delirium” without a proper label – perhaps this explains the divergent epidemiologic data.

So here is my algorithm for altered mental status:

History (A4D5): The following strongly predispose to delirium

  • Alcohol
  • Angst
  • Atemnot (dyspnea)
  • Aua (pain)
  • Deprivation (sensory deprivation, no hearing/visual aids, no contact)
  • Dehydration
  • Depression
  • Degeneration (dementia and other degenerative disease than Alzheimer’s)
  • Drugs (a long list, all psychotropic, see below)

Labs: SCHNELL Vitamines + Tox

  • Sepsis (CRP, Leukos, PCT, Urine, chest xray)
  • Cardiac (cTnI – MIs can lead to delirium curiously)
  • Heme (anemia)
  • Nephro (Uremia)
  • Endocrine (TSH, Elytes – Addison’s/Cushing)
  • Liver
  • eLytes
  • Vitamins
  • Toxicology

CT/MRI: Stroke, SAE, Tumor etc. (in the Stroke Unit we usually have such images ready, but think of neurologic complications)

EEG: r/o Status and distinguish between withdrawal (fast EEG) and the rest (slowing, metabolic)

LP: r/o encephalitis

As for management,

  • I can only stress the importance of reducing the triggers, in particular psychotropic drugs – remember that opiates are rarely delirogenic, yet Metamizol is (it is one of the most lipophilic of NSAIDs).
  • Withdrawal has to be dealt with.

Think of the non-pharmacologic treatment:

  • get relatives to the bed to hold hands and soothe the patient, if not possible, do it yourself.
  • Provide hearing aids, glasses, a warm and friendly environment without too much but also without too few cues – beware of sensory deprivation.
  • Mobilize, if possible in the corridor, so that they see and hear people.
  • If seeing and hearing is difficult, use a pacifier to provide at least oral sensory stimuli.
  • Give them something to do, to work on.
  • Reorientation: Have calenders and clocks everywhere. Tell them where they are and what they are here for.
  • Reduce physical restraints as much as possible.

What about restraints and drugs? There is not much data about the use of either in the treatment of delirium. I think of most of the therapy as self defense – if they are harming you or themselves, you need to protect them without harming too much.

  • Haloperidol is the standard except for extrapyramidal degenerative patients where it triggers hypoactive delirium. Has few side effects, can be given iv. Risperidone is just like Haloperidol only younger, so more data for delirium in dementia.
  • Quetiapine is good for sedating without too many side effects (except orthostatic hypotension), helps resynchronize day/night, but start slow.
  • Olanzapine is good for rough patients, can be given im.
  • Cholinesterase inhibitors might be helpful, but to do a proper study you would have to define the subgroups that really profit. IMHO it is the sensory deprived semi-demented patients that are kicked over the edge by too much information for a too slow braing that benefit.
  • I personally do like Circadin (retarded L-Melatonin) to resynchronize the diurnal rhythm, but there is obviously no study for it.
  • Please refrain from sedating typical neuroleptics, antihistaminics and antidepressants because of their usually quite pronounced anticholinergic properties [it’s not fair to pull the handbrake during the race].


  • As usual, uptodate has a good article on the topic.
  • Use a pharmacology book to study receptor profiles of neuroleptics.

ECG visit

In our Stroke Unit, either our internal medicine specialist or the dayshift neurologist performs the ECG visit for all patients whose ischemia is etiologically undetermined. This involves the following steps on our Philips monitor system:

  • Call up the patient’s window
  • Make sure the “Afib” alarm is set on the “Arrhythmiealarme” page
  • Alarms: Look at all alarms in the last 24-72 hours
  • For any interesting alarm call up the rhythm strip. If necessary, compare RR-intervals using the ruler gadget (left mouse click in the rhythm strip, hold and measure, let go, then click on the new ruler, moving it over the other QRS-complexes)
  • Trends: Call up the “Trends” page: look at sudden surges in heart rate. If you found one, click into the timeline to mark the time, then move over to the “Kurve” page to look at the rhythm strip for that time.
  • Rhythm strips: Using the “Kurve” page, click arbitrarily into the timeline to sample the ECG, looking out for arrhythmia. If necessary, perform the RR-check

The cholesterol controversy

Risk factors for stroke? Hypertension, smoking, cholesterol, heart problems, positive family history, maybe adipositas, ethanol.

There is, however, a slight problem with this list – cholesterol. It turns out that the evidence for the following things is not really impressive:

  • that high cholesterol is a risk factor for stroke
  • that low fat or low cholesterol diet helps reducing risk for stroke
  • that lowering cholesterol is generally helpful for stroke risk

[One can argue that, yes, statins have been shown to reduce stroke (albeit with a very high NNT), but then we have all been told that statins are so very pleiotropic that maybe their antiinflammatory or other properties reduce stroke much more than their cholesterol-lowering properties.

Bear in mind that the evidence for statins (and all other cholesterol lowering drugs) in primary prevention is less than weak, as this metanalysis and a Cochrane analysis shows. Remember that statins might even lead to (discovery of) diabetes.]

The above list of arguments against the so-called cholesterol hypothesis (invented in the 1920s and highly publicized since the 1950ies) is raised by cholesterol agnosticians such as

As so often in history, it is quite difficult to separate the crazy from the serious critics of contemporary science. Myself, I am not too convinced of all the megatrial hype that the statins brought with them and, in general, I don’t like NNTs of > 100, because I know that being a good physician has a NNT of about 5. I am not (yet) an outspoken critic of the cholesterol myth but recommend being conscious and follow the discussions – remember hormone replacement therapy! Still I follow the guidelines recommending a statin for any stroke/TIA patient, even the young ones.

Anterior cerebral artery stroke

There are things you cannot learn from articles or emedicine. In the case of stroke syndromes, the one and only bible is Bogousslavsky’s The Stroke Syndromes, which served as the reference for today’s case of a severly abulic ACA stroke patient. Apart from the clinical pictures, there is a lot to take home:

  • Most of ACA strokes are embolic, nowadays a lot of them are due to endovascular procedures.
  • Proximal A1 emboli can occlude Heubner’s perforating arteries, more distal ones don’t bother the patient if the anterior communicating artery is patent.
  • The corpus callosum is often supplied by both sides, so that you have to occlude either A2 to result in corpus callosum strokes (or have a unilateral supply of both ACAs). In this case one of the famous callotomized patients can result.

Lumbar puncture – the gory details

I am currently revamping our spinal tap SOP, so I went through all the horrible details and tried to find a simple 3-page exposition of how to do it in our department. We discussed these recommendations and corrected a few of the prejudices about LP. Here are the highlights

  • If you cannot do it, you call your resident or attending. So what are his tricks? – Here are my 5 most valuable hints: Sedate, position your patient, be daring – nothing except the aorta is sacred!, use ultrasound
  • Remember that pressure taking is easy – you can always lay down the sitting patient in order to do it. You just need the special “perfusor line”.
  • Take your time for desinfection.
  • Sedate! It’s worth it.

Dyspnea and acute left heart decompensation

What are the essential steps of managing a dyspneic patient in the Stroke Unit? The differential is long but important. Here are the basic steps

  • Take a history
  • Ask your nurses
  • Look at the vitals
  • Do a physical exam: the lung, the heart, the JVP, the abdomen, the skin

Usually, this is enough to manage your patient. Yet, if in doubt:

  • get an ABG
  • get a chest xray
  • get an EKG
  • do labs: CRP, CBC, D-Dimer, BNP. TnI

If still in doubt, ask the ICU resident. He is quite interested in helping you, because if not successful, he has to deal with the patient; and as usual, he has no beds.

Finally, if we arrive at acute left heart decompensation:

  • Determine the nature of CHF (left + right? systolic/diastolic? forward/backward? etiology?)
  • Give Lasix (start with 40 migs) – did you know why it is called Lasix? Ask wikipedia – it is surprisingly simple and has a lot to do with pharmakokinetics…
  • Consider MSI, Nitro
  • Consider non-invasive ventilation