Skip to main content
  1. Courses
  2. NCS Podcast Series
  3. Live at Annual Episode 7: N...

Live at Annual Episode 7: Non-Invasive ICP Monitoring with Mohammad Hirzallah

Loading audio, please wait...

Description

Mohammad Hirzallah joins Nick Morris to discuss improving the utility of optic nerve sheath diameter as a non-invasive intracranial pressure measurement through standardization and automation.

Contributors

  • Nicholas A. Morris, MD

    Assistant Professor and Fellowship Director
    Division of Neurocritical Care and Emergency Neurology
    Department of Neurology
    University of Maryland School of Medicine Medicine

  • Mohammad Hirzallah, MD, MMSc

    Assistant Professor
    Departments of Neurology, Neurosurgery, and Center for Space Medicine
    Section of Neurocritical care and Vascular Neurology
    Baylor College of Medicine

  1. 00:00:20
    Hi, this is Nick Morris for the NCS
  2. 00:00:22
    podcast. We are live at the annual
  3. 00:00:24
    meeting and we just finished the
  4. 00:00:26
    non-invasive ICP session. I'm here with
  5. 00:00:28
    Dr. Mohamed Herzella from Baylor
  6. 00:00:31
    College of Medicine who really gave a
  7. 00:00:32
    wonderful talk on optic nerve sheath and
  8. 00:00:34
    it was clear to me that Dr. Herzella is
  9. 00:00:36
    probably thought more about optic nerve
  10. 00:00:38
    sheath than just about anyone I've ever
  11. 00:00:40
    met and I thought it would be wonderful
  12. 00:00:42
    to have him on the pod. So welcome Dr.
  13. 00:00:43
    Herzella. Thank you. Thank you for
  14. 00:00:45
    having me Now, you know, I think a lot
  15. 00:00:48
    of us employ optic nerve sheath diameter
  16. 00:00:51
    as a sort of easily obtainable
  17. 00:00:53
    non-invasive ICP measurement, but no
  18. 00:00:56
    one's quite clear on what the proper cut
  19. 00:00:58
    off or threshold should be for
  20. 00:00:59
    determining high ICP and you did a
  21. 00:01:02
    wonderful job in your talk and
  22. 00:01:03
    explaining some of the reasons why we
  23. 00:01:05
    don't understand this. Can you
  24. 00:01:07
    summarize those? Yeah, absolutely So
  25. 00:01:10
    to the first point, the cut off, the
  26. 00:01:14
    cut offs are tricky because the meta
  27. 00:01:18
    analysis often allows studies to use
  28. 00:01:20
    their own cutoffs. So you'll see
  29. 00:01:23
    cutoffs as low as four millimeters and
  30. 00:01:25
    others as high as 65 millimeters. So
  31. 00:01:27
    that's a huge variation. I would say a
  32. 00:01:30
    good rule of thumb is above five and
  33. 00:01:33
    someone that is unconscious, you likely
  34. 00:01:37
    have CT changes.
  35. 00:01:41
    Above six is a higher probability that
  36. 00:01:44
    you have ICP changes. So kind of use
  37. 00:01:47
    your pretest probability to adjust the
  38. 00:01:49
    cutoff. So if someone, if you have a
  39. 00:01:51
    high suspicion, use five millimeters as
  40. 00:01:53
    a cutoff. If your suspicion is low, I
  41. 00:01:55
    would use six millimeters as a cutoff.
  42. 00:01:57
    Great, and you pointed out in your talk
  43. 00:02:00
    some of the technical factors that vary
  44. 00:02:03
    from study to study. In fact, some
  45. 00:02:04
    studies you pointed out didn't even
  46. 00:02:06
    measure the optic nerve sheet at all.
  47. 00:02:07
    They actually measured just the optic
  48. 00:02:08
    nerve. Can you kind of go through some
  49. 00:02:10
    of these factors and what we should be
  50. 00:02:12
    thinking about and aware of when we're
  51. 00:02:13
    getting these measurements? Yeah,
  52. 00:02:14
    absolutely I think part of this is maybe
  53. 00:02:19
    we need to do this. to upgrade how we
  54. 00:02:21
    do this measurement, right? If I come
  55. 00:02:23
    to you and say I have this thing called
  56. 00:02:25
    echocardiography and I'm not exactly
  57. 00:02:28
    sure but anything from an ejection
  58. 00:02:29
    fraction of 30 to 60 might be abnormal,
  59. 00:02:33
    no one would use it, right? So we need
  60. 00:02:36
    to kind of start becoming more rigorous
  61. 00:02:39
    about how we do the measurement. So
  62. 00:02:40
    think about where you place your probe
  63. 00:02:44
    and really making those small
  64. 00:02:46
    adjustments with your probe until you
  65. 00:02:48
    get a nice cut with that thin interface
  66. 00:02:52
    between the nerve and the eyeball and
  67. 00:02:54
    the clear and atomic differentiation
  68. 00:02:57
    between the subarachnoid space, the
  69. 00:02:58
    nerve,
  70. 00:02:59
    the dura, and the retroorptophat, you
  71. 00:03:01
    need to see those clean lines. And when
  72. 00:03:04
    you're doing that, remember, you're
  73. 00:03:05
    scanning a four to six millimeter
  74. 00:03:07
    structure, right? So that's maybe
  75. 00:03:10
    thinner than a pencil. And so when you
  76. 00:03:13
    are moving your probe around, these are
  77. 00:03:15
    really small adjustments to get that
  78. 00:03:17
    longitudinal cut through the nerve. So,
  79. 00:03:19
    once you get that nice image, then
  80. 00:03:22
    understanding the anatomy and placing
  81. 00:03:24
    the markers correctly is going to be the
  82. 00:03:26
    most impactful thing to get an accurate
  83. 00:03:30
    measurement. And you mentioned some
  84. 00:03:32
    studies clearly are not all using a
  85. 00:03:34
    linear probe, but your recommendation
  86. 00:03:36
    is right to always use a linear probe.
  87. 00:03:38
    Yes, so classic ultrasound teaching is
  88. 00:03:41
    linear probe for more superficial
  89. 00:03:43
    structures where we need the resolution
  90. 00:03:45
    and then
  91. 00:03:47
    your phased array or your curvilinear
  92. 00:03:49
    probe for when you need deeper
  93. 00:03:52
    penetration with less resolution, right?
  94. 00:03:54
    So linear probe is for this use case,
  95. 00:03:58
    so we care about that accurate lateral
  96. 00:04:02
    resolution so you should use a linear
  97. 00:04:04
    probe at least 75 megahertz, probably
  98. 00:04:07
    closer to 10 as ideal Great, and my
  99. 00:04:11
    trainees often know that they should be
  100. 00:04:12
    measuring 3 millimeters back, but they
  101. 00:04:14
    don't always know back from what, so
  102. 00:04:16
    what is your recommendation for that?
  103. 00:04:18
    So that was. That was one of the big
  104. 00:04:20
    sorts of variations. So you see some
  105. 00:04:22
    studies measuring from the papella,
  106. 00:04:24
    even with papilladema, right? So if
  107. 00:04:26
    your papella is already one or two
  108. 00:04:28
    millimeters elevated, you're barely
  109. 00:04:30
    getting to the optic nerve sheath if you
  110. 00:04:32
    only go three millimeters from that.
  111. 00:04:34
    You have studies measuring from the
  112. 00:04:35
    retina, and you have studies measuring
  113. 00:04:38
    from the sclera or behind the eyeball.
  114. 00:04:43
    So the retina seems to be the most
  115. 00:04:45
    consistent structure that people agree
  116. 00:04:47
    on It's also easier to spot because of
  117. 00:04:50
    that equagenicity difference between the
  118. 00:04:53
    retina and the rest of the eyeball. So,
  119. 00:04:56
    I would say a retina, you place your
  120. 00:04:59
    line right at that
  121. 00:05:02
    equagenicity change between the
  122. 00:05:03
    hyperacolic retina and the hyperacolic
  123. 00:05:05
    bitters fluid glacial marker there, go
  124. 00:05:08
    three millimeters depth along the axis
  125. 00:05:10
    of the optic nerve, right? So if your
  126. 00:05:12
    image is tilted, you need to adjust
  127. 00:05:14
    that depth marker, and that's where you
  128. 00:05:16
    perform the measurement And then you
  129. 00:05:18
    mentioned in your talk that the there
  130. 00:05:19
    can be great variability in the
  131. 00:05:21
    measurements from the interior aspect of
  132. 00:05:23
    the optic nerve sheath versus the
  133. 00:05:25
    exterior aspect. And so what do you
  134. 00:05:27
    recommend we measure? So the, and this
  135. 00:05:31
    is one of those things where I'll try to
  136. 00:05:33
    be descriptive to kind of convey an
  137. 00:05:35
    image with words. So the inner more
  138. 00:05:36
    structure, that hypoacobic line,
  139. 00:05:37
    that's the nerve, then you have two
  140. 00:05:37
    structures outside of that and people
  141. 00:05:37
    describe them as the stripped,
  142. 00:05:37
    hyperacolic bands, so they're like
  143. 00:05:37
    horizontal strips, they're hyperacolic
  144. 00:05:37
    and they surround the nerve. That's
  145. 00:05:37
    most likely the subarachnoid space.
  146. 00:05:37
    Then you have the outer black line
  147. 00:05:37
    outside that, so an outer hypoacolic
  148. 00:05:37
    band surrounding that, that's most
  149. 00:05:37
    likely the dura. Then you have the
  150. 00:05:37
    retroorbital fat that is hyperacolic.
  151. 00:05:38
    And that's the
  152. 00:06:06
    challenge where it's important to get
  153. 00:06:08
    that nice resolution because the
  154. 00:06:11
    equigenicity of the subarachnoid space
  155. 00:06:13
    and the retroorbital fat can be very
  156. 00:06:16
    similar, so you really don't get that
  157. 00:06:18
    nice line. in between. Now the
  158. 00:06:22
    difference between internal and external
  159. 00:06:23
    is whether you place that marker inside
  160. 00:06:27
    the black line, which is the dura, and
  161. 00:06:30
    you measure the inside diameter of the
  162. 00:06:32
    optic nerve sheet, or you place that
  163. 00:06:34
    line outside the dura, and then it's
  164. 00:06:37
    the external diameter of the optic nerve
  165. 00:06:39
    sheet. And the reason that's important
  166. 00:06:41
    is they vary by around 15 millimeters.
  167. 00:06:46
    So, and that's statistically
  168. 00:06:48
    significantly different in some studies
  169. 00:06:50
    that we showed today, right? So, that
  170. 00:06:53
    would be the challenge is if you're not
  171. 00:06:55
    measuring the same structure
  172. 00:06:56
    consistently, you will get variable
  173. 00:06:58
    results, and then your measurements are
  174. 00:07:00
    not accurate. Both measurements are
  175. 00:07:02
    sensitive and specific for elevated ICP.
  176. 00:07:05
    So, whichever measurement you do in
  177. 00:07:07
    your practice for now, until we have
  178. 00:07:09
    standardization, just make sure that
  179. 00:07:10
    you're making that same measurement
  180. 00:07:13
    Great. And lastly, I don't think you
  181. 00:07:16
    mentioned it in your talk, but we've
  182. 00:07:17
    seen papers looking at. What is the
  183. 00:07:20
    right plane to look at the optic nerve
  184. 00:07:22
    and some papers suggest that we should
  185. 00:07:24
    be looking in multiple planes and taking
  186. 00:07:25
    an average, what's your practice? So
  187. 00:07:28
    yeah, we didn't mention that in the
  188. 00:07:30
    talk, but we discussed it extensively
  189. 00:07:32
    in the panel. So there are, the planes
  190. 00:07:36
    that are published are, there is an
  191. 00:07:39
    infraorbital plane that almost tries to
  192. 00:07:41
    get a cross-section that is not feasible
  193. 00:07:43
    in all people, then the more common
  194. 00:07:46
    planes that we see is the axial or the
  195. 00:07:48
    coronal plane.
  196. 00:07:51
    There are studies that compared them,
  197. 00:07:53
    and a lot of the times they don't
  198. 00:07:55
    correlate, but those weren't very
  199. 00:07:56
    extensive studies. From a, most
  200. 00:08:00
    panelists prefer the coronal axis, also
  201. 00:08:05
    from just a logistic standpoint, unless
  202. 00:08:07
    you have a specialized small footprint
  203. 00:08:10
    probe, it's really hard to fit that
  204. 00:08:12
    over the eye if you're going coronally,
  205. 00:08:15
    so that's why most people prefer axial
  206. 00:08:16
    and
  207. 00:08:19
    the classical that were first performed
  208. 00:08:21
    or axial. A lot of our normative
  209. 00:08:23
    databases, the good studies that kind
  210. 00:08:26
    of give you the normal values also used
  211. 00:08:28
    coronal. So it seems like most people
  212. 00:08:31
    prefer coronal and that's the one I
  213. 00:08:33
    would encourage people to use for
  214. 00:08:34
    clinical applications. But if you have
  215. 00:08:37
    a research application where you want to
  216. 00:08:39
    investigate different axes against each
  217. 00:08:41
    other or average them for to see if that
  218. 00:08:43
    improves the measurement that's that's
  219. 00:08:45
    valid but the standard is coronal.
  220. 00:08:48
    Great. And then lastly you mentioned it
  221. 00:08:51
    but we haven't explicitly said it so the
  222. 00:08:53
    the panel that you're describing can you
  223. 00:08:55
    tell the listeners who that is and how
  224. 00:08:57
    you're working with this group in order
  225. 00:08:58
    to standardize these measurements? Yeah
  226. 00:09:00
    absolutely. So we the panel is right
  227. 00:09:03
    now 60 panelists we have 10 steering
  228. 00:09:07
    committee members and these are people
  229. 00:09:09
    that published on variability and how to
  230. 00:09:11
    improve the measurement and then
  231. 00:09:14
    together we reviewed the literature
  232. 00:09:19
    and we classified the sources of
  233. 00:09:19
    variation. recruited our expert
  234. 00:09:21
    panelists outside the steering committee
  235. 00:09:23
    so that's our 50 panelists and we had
  236. 00:09:26
    selection criteria for our panelists so
  237. 00:09:29
    we picked the most published up
  238. 00:09:31
    technology damter experts by number of
  239. 00:09:33
    publications we reached out to them we
  240. 00:09:36
    reached out to education experts
  241. 00:09:40
    basically ultrasound fellowship program
  242. 00:09:42
    directors and asked them if they teach
  243. 00:09:47
    up technology damter frequently and if
  244. 00:09:50
    they'll be interested in participating
  245. 00:09:54
    and then we asked them to nominate
  246. 00:09:58
    clinicians that use this measurement and
  247. 00:10:02
    their clinical practice so we tried to
  248. 00:10:03
    keep it balanced between clinicians
  249. 00:10:03
    educators and researchers we have 50
  250. 00:10:03
    panelists Canada US Latin America Europe
  251. 00:10:03
    Africa Middle East and China so it's a
  252. 00:10:03
    diverse panel great well thank you so
  253. 00:10:03
    much I'm really excited to see what this
  254. 00:10:11
    panel produces and I think it
  255. 00:10:16
    will go a long way into standardizing
  256. 00:10:19
    this practice so that we're all on the
  257. 00:10:21
    same page when we talk about optic nurse,
  258. 00:10:22
    she's diameter and it's relation to ICP.
  259. 00:10:24
    So thanks for your work and we look
  260. 00:10:26
    forward to seeing the eventual
  261. 00:10:27
    publications that will come from this.
  262. 00:10:29
    Awesome. Thank you for having me. All
  263. 00:10:30
    right. Thank you.
  264. 00:10:33
    Thanks everyone for joining us here on
  265. 00:10:35
    the neurocritical care podcast. It's
  266. 00:10:38
    available anywhere you get your podcast
  267. 00:10:40
    and it comes along with some continuing
  268. 00:10:42
    education credits so check it out. This
  269. 00:10:44
    is Nick Morris live from the annual
  270. 00:10:45
    meeting