Thread by @caseyalbin

1/🧵
In the early days of fellowship, I remember checking our SAH patients’ transcranial dopplers (TCD), scanning the Vmeans & if they were ~<70 cm/sec throughout thinking:

“Great. Perfect. TCDs globally low. Nothing to worry about here!”

Right?

A #tweetorial on TCDs

2/
Right? Sort of.

🚨Note. This is not a #tweetorial about if large vessel vasospasm is the cause of DCI or just an epiphenomenon OR if treating vasospasm is the way to improve functional outcomes …That is important!... but that is not this tweetorial.
pubmed.ncbi.nlm.nih.gov/21285966/

3/
Given #TCDs is a pretty large topic, this @medtweetorial will be told in 3 parts:
Part 1⃣:
⭐️Basic principles of TCDs
⭐️Use of TCDs to detect Vasospasm

Part 2⃣: The Pulsatility Index - why it matters
Part 3⃣: The Utility of TCDs as an ancillary test in BDT

4/
1⃣st up: Basics
A TCD examination is performed using a 2MHz ultrasound probe.

It transmits u/s to the flowing intracranial vessels.

The change in freq. between the waves emitted and those reflected back is directly proportional to the blood’s velocity (Doppler Shift)

5/
Important to note that the Reflector Speed “Velocity” (cm/s) involves some math… most importantly, it is dependent on the cosine of the angle of insonation 🔽

6/
We assume this angle is 0, as in the pic below⬇️. The prob is directly aligned with the blood flow:

In that case. Cos(0)=1.

Image from: Role of transcranial Doppler ultrasonography in stroke pmj.bmj.com/content/83/985…

7/
But if the probe is not directly in line with the blood vessel, the cos(x) does not = 1.

And the speed reported is ⬇️⬇️ than actual flow.

Thus, the flow is *AT LEAST* as fast as what’s reported & may be FASTER!

8/
Since blood flow is laminar the reflected arrays have a spectral display representing the mixture of velocities.
Each waveform can be analyzed to calc:
✅Peak Systolic Velocity
✅End Diastolic Velocity
✅Pulsatility Index
✅Time-average mean velocity (Vmean)(MFV)

9/
For a standard vasospasm study, the MCA, ACA and PCA are insonated through the transtemporal widow (A).

The suboccipital window (D) is used for the basilar and vertebral arteries.

Image: TCD ultrasonography: clinical applications in CV disease pubmed.ncbi.nlm.nih.gov/2214882/

10/
If two views are being used to give the details of 5 arteries, how do you know which is which?

Because each vessel has a characteristic waveform; and can be ID’ed by the speed and director of flow + characteristic depth.

11/
Ok. Got it.
🔊TCDs = form of ultrasonography
📈Track the speed of blood by the doppler equation
🌊Generate a waveform b/c of laminar flow
🏄‍♂️Vessels have characteristic waveforms

12/
🚨Section 2: TCDs in Vasospasm (VSP) Monitoring

Although TCDs have many uses, they really take center stage in aSAH, where (at least in all the institutions I’ve practiced in), everyone follows the TCD tech around like:

13/
Why because TCDs help detect VSP.

VSP causes narrowing of the arteries (diffusely or focally)➡️an increase in the flow velocity [Poiseuille’s law (🙀)🔽].

All things being otherwise constant (a BIG if, as we’ll see)*, if the Vmean ⬆️, our concern for VSP ⬆️.

14/
At what point is the velocity fast enough to be concerning?

Honestly for how much we all worry about these #⃣, how well the speed correlates w/ DCI is investigated & debated.

Each artery is different.
ExMCA
🔥Mild = 120-149 cm/s
🔥🔥Moderate =150-199
🔥🔥🔥Severe>200

15/
An example:
Post-Bleed Day 1: Nice plump vessels in the DSA, Low Vmean (in the 50s cm/sec)

16/
Compared to Post-Bleed Day 9
Sig narrowing of communicating segment of ICA, M1 and PComm, and PCA. Elevated TCD VMeans (in the 120s/130s cm/sec).

17/
A nice analysis by @samBsnider showed that patients that developed DCI were enriched with earlier increased in mean velocities, and that the risk of infarction increases with vasospasm severity
medrxiv viewer disq.us/t/3wesh7

18/
@neuro_intensive's analysis of our SAH population showed that ⬆️TCD velocities were correlated to the risk of DCI. thejns.org/view/journals/…

19/
That all said, it’s INCREDIBLY important to recognize many variables affect blood's velocity:
Such as
✨Age
✨Gender
(not likely to change during hospitalization)

✨Hematocrit
✨Blood viscosity
✨pCO2
✨BP
(highly likely to change during hospitalization)

20/
🚨 Lots of patients develop anemia during hospitalization.
🚨we manipulate the pCO2 to reduce ICP via vasoconstriction… (hopefully not when a patient is in vasospasm tho!)
🚨Induced hypertension (something we may try in vasospasm) may also ⬆️velocity

21/
How can we account for these changes?

The Lindegaard ratio (LR)

LR = MCA Vmean / ICA Vmean

The above changes should affect the ICA and MCA. But vasospasm ⬆️ flow in the MCA and not the proximal ICA.

22/
LR between 3-6 is a sign of mild VSP
LR > 6 should concern you for severe VSP.

Remember though that a lower but abrupt change from baseline is also concerning!

23/
If you are interested in a more in-depth look at the principles of TCD monitoring, I highly recommend this review by Dr. Farzaneh Sorond (@SorondLab), which is a beautiful overview of all the applications and science behind TCDs.
pubmed.ncbi.nlm.nih.gov/23361485/

24/
Finally, started all of this saying “Vmeans <70, we all good!” … so far everything in this tread would confirm that narrative, yes??

Enter Pulsatility Index.

For next time.

25/
♥️ thoughts / comments / feedback! Also - a special shout out to Dr. Aaron Anderson of @EmoryNeurohosp1 for always helping review challenging TCDs!!!
@judyhtchang @drdangayach @emcrit @Capt_Ammonia @namorrismd @rwregen @AvrahamCooperMD @pouyeah @KP_MD2018 @EmoryNeuroCrit