Skip to main content
  1. Courses
  2. NCS Podcast Series
  3. Episode 95: Hot Topics - Ea...

Episode 95: Hot Topics - Early MiNimally-invasive Removal of Intracerebral hemorrhage (ENRICH)

Loading audio, please wait...

Description

The ENRICH trial compared the outcomes between early surgical intervention using the BrainPath® Approach (i.e., MIPS) and a medically managed cohort. The integrated surgical approach included a combination of available technologies, including the FDA-cleared NICO BrainPath® for non-disruptive access and NICO Myriad® to achieve the goal of maximum clot evacuation. Alex Reynolds is joined by the ENRICH trial PIs, Gustavo Pradilla, Jonathan Ratcliff, and Alex Hall plus guests Chris Kellner and Kara Melmed to discuss the findings of the ENRICH clinical trial.

Contributors

  • Christopher Kellner, MD

  • Alex Hall, MD

  • Gustavo Padilla, MD

  • Jonathan Ratcliff

  • Alexandra Reynolds, MD

    Alexandra Reynolds, MD, is an Assistant Professor of Neurosurgery and Neurology and a board-certified neurointensivist in the Neurocritical Care Unit at The Mount Sinai Hospital. Dr. Reynolds also serves as the Director of TeleNeurocritical Care for the Mount Sinai Health System.

    A native New Yorker, she was educated at Princeton University and the Columbia University College of Physicians & Surgeons before completing a neurology residency at NewYork-Presbyterian/Columbia University Medical Center and a Neurocritical Care fellowship at NewYork-Presbyterian/Columbia University Medical Center and Weill Cornell Medical Center.

    Dr. Reynolds is an expert in the treatment of neurological emergencies such as subarachnoid hemorrhage, intracerebral hemorrhage, status epilepticus, and neurotrauma. Her research centers on assessment, treatment and neurological prognostication of survivors of cardiac arrest and the neurological management of patients in fulminant hepatic failure.

  • Kara Melmed, MD

    Positions
    Clinical Assistant Professor, Department of Neurology at NYU Grossman School of Medicine
    Clinical Assistant Professor, Department of Neurosurgery at NYU Grossman School of Medicine
    Education and Training
    Fellowship, New York-Presbyterian/Columbia University Medical Center, Neurocritical Care, 2019
    Residency, Cedars-Sinai Medical Center, Neurology, 2017
    MD from University Of Arizona, 2013

  1. 00:00:20
    Hi, everyone. Welcome to the Hot
  2. 00:00:22
    Topics podcast of the Neurocritical Care
  3. 00:00:25
    Society. My name is Alex Reynolds, and
  4. 00:00:28
    I'm a neuro intensivist at Mount Sinai
  5. 00:00:30
    Health System in New York City. And I'm
  6. 00:00:33
    really excited today to present to you a
  7. 00:00:37
    group of people who are going to be
  8. 00:00:39
    discussing minimally invasive ICH
  9. 00:00:43
    evacuation. So we're lucky enough to
  10. 00:00:45
    have the Enrich Group from Emory who
  11. 00:00:50
    will be speaking about their latest
  12. 00:00:53
    study. And we'll also have two people
  13. 00:00:59
    commenting on the results of the study.
  14. 00:01:01
    So we'll start off by introducing Dr.
  15. 00:01:05
    Gustavo Pradea. He is an associate
  16. 00:01:08
    professor of neurosurgery at Emory and
  17. 00:01:10
    chief of neurosurgery at Grady Memorial.
  18. 00:01:13
    We also have Jonathan Radcliffe, who is
  19. 00:01:15
    associate professor of emergency
  20. 00:01:17
    medicine. at Emory and a
  21. 00:01:21
    neuro-intensivist. And we have Alex
  22. 00:01:24
    Hall, Assistant Professor of Emergency
  23. 00:01:26
    Medicine and Director of Clinical Trials,
  24. 00:01:28
    also from Emory. So the three of them
  25. 00:01:31
    are gonna be presenting the results of
  26. 00:01:35
    the enriched trial. And then we'll have
  27. 00:01:38
    Dr. Christopher Kalner, who is
  28. 00:01:40
    Assistant Professor of Neurosurgery and
  29. 00:01:43
    Director of the Interest Reeval
  30. 00:01:45
    Hemorrhage Program at Mount Sinai Health
  31. 00:01:47
    System, to comment on the neurosurgical
  32. 00:01:50
    perspective on this trial. And we'll
  33. 00:01:53
    have Dr. Kara Melmed, who is an
  34. 00:01:56
    Assistant Professor of Neurology and
  35. 00:01:58
    Neurosurgery at NYU Langone, to provide
  36. 00:02:01
    the neuro-intensivist perspective on how
  37. 00:02:04
    this trial might be changing the scope
  38. 00:02:06
    of medical and surgical care for ICH
  39. 00:02:09
    patients. So with that, I want to
  40. 00:02:12
    invite the Emory team you speak to us
  41. 00:02:14
    about their trial or. Early, minimally
  42. 00:02:17
    invasive removal of ICH or enriched.
  43. 00:02:24
    Okay, we haven't had a chance to
  44. 00:02:26
    discuss how we would proceed with that.
  45. 00:02:29
    So I think the first thing to talk about
  46. 00:02:34
    before we talk about Enrich is why we
  47. 00:02:37
    wanted to do Enrich. So Enrich was
  48. 00:02:41
    first conceived approximately seven
  49. 00:02:44
    years ago when we started discussing the
  50. 00:02:47
    trial design and the rationale for the
  51. 00:02:50
    trial And that came about because we had
  52. 00:02:55
    a group of neurosurgeons primarily who
  53. 00:03:00
    were starting to use this minimal access
  54. 00:03:04
    port for different surgeries, not just
  55. 00:03:07
    for Enrich's arrival hemorrhage, a lot
  56. 00:03:09
    of tumor work was being done in that.
  57. 00:03:11
    And ITH became one of
  58. 00:03:15
    the uses of that technology that seemed
  59. 00:03:18
    to make sense and a few people started
  60. 00:03:20
    to do it and. still some promising
  61. 00:03:22
    results. So there was a prospective
  62. 00:03:25
    registry that was put together by a
  63. 00:03:29
    neurosurgeon named Muhammad Labib, who
  64. 00:03:31
    is now the University of Maryland. And
  65. 00:03:33
    that had 10 centers that contributed to
  66. 00:03:36
    patients to that. It was a very small
  67. 00:03:38
    study, 50 patients. But it was the
  68. 00:03:42
    first organized attempt at seeing if
  69. 00:03:44
    this was something that made sense that
  70. 00:03:47
    we should pursue further. And we had
  71. 00:03:49
    some promising results there with
  72. 00:03:51
    regards to the amount of hemorrhage that
  73. 00:03:53
    was evacuated and the safety of the
  74. 00:03:57
    technique. And we saw some very
  75. 00:03:59
    preliminary reports on improvements in
  76. 00:04:04
    Glasgow, Coma Scale III and post-op.
  77. 00:04:07
    So that was
  78. 00:04:12
    the
  79. 00:04:16
    initial interest in it, was this
  80. 00:04:16
    Myspace perspective registry. And then
  81. 00:04:16
    following that, we decided to, put it
  82. 00:04:20
    to the test in a more rigorous way.
  83. 00:04:23
    Once much space was published, a couple
  84. 00:04:25
    of single institution centers were
  85. 00:04:27
    published. One was the Cleveland Clinic
  86. 00:04:30
    experience that Martyr Vain presented.
  87. 00:04:31
    It was a small series, but it showed
  88. 00:04:34
    the same type of results with good cloud
  89. 00:04:37
    evacuation rates over 95 in March series
  90. 00:04:42
    that were also interesting. And one of
  91. 00:04:45
    the things that we liked a lot about
  92. 00:04:46
    March in the Cleveland Clinic Team
  93. 00:04:49
    series was that they saw a lot of
  94. 00:04:51
    positive spot signs in
  95. 00:04:54
    their initial experience. And they were
  96. 00:04:56
    able to get human stasis comfortably and
  97. 00:04:59
    definitively in all the cases they saw
  98. 00:05:02
    spot signs. And it was about 65 or so.
  99. 00:05:06
    So that was a nice thing. There was
  100. 00:05:09
    another group from St. Louis University.
  101. 00:05:13
    That was led by Giro and Co-Pence and
  102. 00:05:15
    they all tried to small share his set.
  103. 00:05:18
    was a single institution series. So all
  104. 00:05:20
    of that was put together as part of the
  105. 00:05:26
    initial experience for Enrich. So when
  106. 00:05:28
    we decided to set up Enrich, Jonathan
  107. 00:05:31
    and I recruited our research team. And
  108. 00:05:36
    Alex Paul was the architect of putting
  109. 00:05:40
    all of the trial
  110. 00:05:43
    logistics together
  111. 00:05:46
    The trial was funded by the manufacturer,
  112. 00:05:51
    the device, that's Niko Corporation.
  113. 00:05:53
    They make the brain pad device, which
  114. 00:05:55
    is the port. And they make the
  115. 00:05:58
    evacuation tool. And that's called the
  116. 00:06:01
    myriad device. So they decided to fund
  117. 00:06:05
    a study. And then we reached out to our
  118. 00:06:09
    colleagues from neuroendology, from
  119. 00:06:14
    a
  120. 00:06:16
    health economics. group to design the
  121. 00:06:20
    quality outcome measures for that side.
  122. 00:06:24
    And we recruited a strip neurologist,
  123. 00:06:27
    Mike Trankel, who's the head of our
  124. 00:06:29
    neurologic group, at Greater Memorial
  125. 00:06:31
    Hospital. David Wright, who's our
  126. 00:06:34
    chair of emergency medicine and leads
  127. 00:06:36
    the neuroscience emergency group here.
  128. 00:06:40
    And then we reached out to the best team
  129. 00:06:44
    we could find for basin statistics,
  130. 00:06:47
    that was the berry consulting group.
  131. 00:06:51
    And then we put ourselves in their hands.
  132. 00:06:56
    And Alex, if you want to talk about the
  133. 00:06:59
    initial phases of the design and how
  134. 00:07:01
    those simulations were put together.
  135. 00:07:04
    Yeah, so basically, we sit all around
  136. 00:07:08
    in a room, talk about what the current
  137. 00:07:10
    existing,
  138. 00:07:13
    knowledge and literature is talk about
  139. 00:07:15
    what we expect performance-wise in a
  140. 00:07:20
    clinical trial through mortality and
  141. 00:07:21
    functional benefit and sort of what sort
  142. 00:07:23
    of is seen in general medical management.
  143. 00:07:28
    And then the statisticians do thousands
  144. 00:07:32
    of simulations
  145. 00:07:36
    of trials and different possibilities to
  146. 00:07:39
    identify what's likely the most
  147. 00:07:42
    successful way of doing that, whether
  148. 00:07:43
    that's looking at location,
  149. 00:07:47
    be it in basal ganglia versus low bar
  150. 00:07:49
    versus something else, timing. So all
  151. 00:07:52
    these go into different simulations that
  152. 00:07:54
    are run ahead of time. And then we sort
  153. 00:07:59
    of whittle down the inclusion,
  154. 00:08:02
    exclusion criteria to try to figure out
  155. 00:08:04
    what's feasible across the possibilities
  156. 00:08:07
    that we can achieve based on our
  157. 00:08:11
    simulations. That's what led to the
  158. 00:08:14
    enriched population and using an
  159. 00:08:16
    adaptive design where the location could
  160. 00:08:20
    be enriched at multiple interim time
  161. 00:08:23
    points, depending on what the data
  162. 00:08:27
    showed to the DSMB. So that was our
  163. 00:08:30
    general premise starting. And once we
  164. 00:08:33
    selected and finalized the design, then
  165. 00:08:35
    it's kind of out of our hands and runs
  166. 00:08:42
    Yeah,
  167. 00:08:44
    one of the critical things about putting
  168. 00:08:47
    this in context of time is that when we
  169. 00:08:49
    started the sign in the trial, MISTI2
  170. 00:08:52
    was just published and MISTI3 was
  171. 00:08:54
    ongoing. And collectively, I think in
  172. 00:08:59
    their critical care, neurology and
  173. 00:09:00
    neurosurgery, we all expected MISTI3 to
  174. 00:09:04
    be positive based on how promising
  175. 00:09:06
    MISTI2
  176. 00:09:09
    looked And it was a very strong team
  177. 00:09:11
    with. a lot of very seasoned
  178. 00:09:13
    investigators. So that was something we
  179. 00:09:18
    had in the back of our minds. We didn't
  180. 00:09:20
    know what that was going to look like,
  181. 00:09:22
    but we had an expectation that was going
  182. 00:09:24
    to be positive. And we were rooting for
  183. 00:09:27
    it, because the only other trials that
  184. 00:09:29
    we had to look at were the stitch trials,
  185. 00:09:33
    or conventional craniotomy trials. And
  186. 00:09:35
    those were negative.
  187. 00:09:38
    Both stitch one and stitch two were
  188. 00:09:40
    negative studies But we incorporated the
  189. 00:09:44
    data that we had at the time. So stitch
  190. 00:09:46
    one, stitch two, very small in this
  191. 00:09:50
    topic trial that was done in Austria
  192. 00:09:52
    several years ago. We used also all the
  193. 00:09:57
    MISTI data that was available at the
  194. 00:09:59
    time. And we incorporated that into our
  195. 00:10:01
    simulations and into the design of the
  196. 00:10:04
    trial When we started, there were a few
  197. 00:10:10
    -
  198. 00:10:13
    very small experiences on endoscopic ICH
  199. 00:10:16
    evacuation, not to the extent that it
  200. 00:10:20
    is today and what the Sinai Group and
  201. 00:10:22
    Chris has done to bring that technology
  202. 00:10:26
    to the forefront. It was a very, seven
  203. 00:10:28
    years ago, it was just starting to
  204. 00:10:30
    build that experience.
  205. 00:10:33
    There
  206. 00:10:44
    was a big experience in using endoscopic
  207. 00:10:45
    techniques for interventional hemorrhage
  208. 00:10:46
    evacuation. So that was part of what we
  209. 00:10:47
    were also considering was
  210. 00:10:49
    technology-wise, what was it, seven
  211. 00:10:52
    years ago, what was it, the forefront
  212. 00:10:53
    of technology? That was one big part
  213. 00:10:56
    that we wanted to incorporate into
  214. 00:10:57
    Enrich was the best possible technology
  215. 00:11:00
    we can bring to the procedure. And that
  216. 00:11:03
    included several fronts. One was
  217. 00:11:06
    imaging, we wanted to make sure we had
  218. 00:11:09
    the right pre-operative imaging
  219. 00:11:11
    assessment. to determine eligible
  220. 00:11:15
    patients, to determine adequate
  221. 00:11:18
    surgical trajectories. We wanted to
  222. 00:11:21
    have the best tools for access that were
  223. 00:11:24
    the least disruptive possible. We
  224. 00:11:27
    wanted to have the best visualization in
  225. 00:11:29
    surgery, what was the best optics that
  226. 00:11:31
    we could get at the time, and the best
  227. 00:11:34
    tools for efficient clot evacuation,
  228. 00:11:36
    get the clots out in a quick and
  229. 00:11:39
    proficient way And also the best
  230. 00:11:42
    chemostasis, some of the concerns with
  231. 00:11:48
    doing cases that have positive spot
  232. 00:11:50
    signs and doing ultra-early surgery, of
  233. 00:11:53
    course, primarily relate to your
  234. 00:11:55
    ability to get the chemostasis. So that
  235. 00:11:57
    was one of the key factors we consider
  236. 00:12:01
    when we put the approach together. And
  237. 00:12:05
    while in the early phases, We did a lot
  238. 00:12:09
    of photography to validate this.
  239. 00:12:12
    trajectories that we were going to pick.
  240. 00:12:14
    As we did more and more cases, the need
  241. 00:12:17
    to use tractography decrees for ICH
  242. 00:12:20
    primarily in two of the three most
  243. 00:12:25
    common locations. So for anterior base
  244. 00:12:27
    of the angular hemorrhages and for
  245. 00:12:29
    the majority of lower hemorrhages that
  246. 00:12:31
    reach the surface, we had a pretty good
  247. 00:12:33
    idea what the trajectories were going to
  248. 00:12:35
    be like without having to use
  249. 00:12:36
    tractography. With the laminic
  250. 00:12:38
    hemorrhages,
  251. 00:12:40
    those were the ones that we were relying
  252. 00:12:42
    most heavily on tractography. And those
  253. 00:12:44
    were the ones that, at the end of the
  254. 00:12:47
    day, we decided to exclude one for that
  255. 00:12:50
    reason. But second, because looking at
  256. 00:12:52
    my space and other studies, the number
  257. 00:12:56
    of patients that were able to achieve a
  258. 00:12:59
    comparable neurological recovery was the
  259. 00:13:03
    same. But the timeline was much more
  260. 00:13:06
    extended we would see good. recoveries
  261. 00:13:09
    in MRS at six months in basal ganglia
  262. 00:13:12
    and lower hemorrhages. Alamic
  263. 00:13:14
    hemorrhages,
  264. 00:13:16
    probably because of the extension into
  265. 00:13:18
    the mid-brain, took nine to 12 months
  266. 00:13:21
    or sometimes longer to get to the same
  267. 00:13:23
    level. And since we had an limited
  268. 00:13:25
    amount of follow-up based in the design
  269. 00:13:28
    at six months, we think we were going
  270. 00:13:31
    to get to see the full extent of the
  271. 00:13:34
    possible recovery for the thalamic, so
  272. 00:13:36
    those salamis were excluded So that's
  273. 00:13:39
    also part of the context as to why we
  274. 00:13:41
    selected that type of technology and
  275. 00:13:44
    what the inclusion experience here were
  276. 00:13:46
    based on. Jonathan, do you want to
  277. 00:13:49
    speak to the full inclusion, inclusion
  278. 00:13:51
    criteria that we selected?
  279. 00:14:00
    Yeah, sure, some of the most important
  280. 00:14:02
    ones that you spoke about was that
  281. 00:14:03
    exclusion of the thalamic, the thalamic
  282. 00:14:05
    hemorrhages, and the decision to focus
  283. 00:14:08
    on the low bar and the anterior base of
  284. 00:14:10
    ganglia. Obviously, the volume of the
  285. 00:14:12
    stroke, we wanted to, to keep it in a
  286. 00:14:15
    tight range of that, of that 30 to 80
  287. 00:14:19
    Ccs, because, you know, we've seen
  288. 00:14:21
    clinically and, and from the data that
  289. 00:14:22
    we have, that if you're less than 30
  290. 00:14:24
    Ccs, that the risk of surgery might not,
  291. 00:14:27
    might not actually be beneficial. And
  292. 00:14:29
    over 80 Ccs, there's often significant
  293. 00:14:32
    structural damage at the time of the
  294. 00:14:34
    original ictus, or at least at the time
  295. 00:14:35
    that the stroke is that sized.
  296. 00:14:41
    We also wanted to get patients
  297. 00:14:44
    that were, that were actually having
  298. 00:14:45
    some degree of significant neurologic
  299. 00:14:48
    impairment And so the simplest tool that
  300. 00:14:49
    we could use for that was the Glasgow
  301. 00:14:51
    Coma Scale. Limited that too, an upper
  302. 00:14:54
    limit of 14, we wanted to have them be
  303. 00:14:56
    somewhat impaired as we know that some
  304. 00:14:57
    of these low bar hemorrhages patients
  305. 00:14:59
    can be pretty awake, depending on the
  306. 00:15:02
    location of the hemorrhage. But wanted
  307. 00:15:04
    to have some obvious impairment, but
  308. 00:15:06
    also have a extent of the impairment not
  309. 00:15:10
    be so significant that they weren't
  310. 00:15:12
    patients that people wouldn't want to
  311. 00:15:13
    ultimately operate on. So we cut that
  312. 00:15:15
    at five for total GCS score.
  313. 00:15:19
    And the same is true of using the NIH
  314. 00:15:21
    stroke scales So they had to have an NIH
  315. 00:15:23
    stroke scale of greater than five. And
  316. 00:15:26
    then it was some degree of functional
  317. 00:15:29
    baseline. So we landed the MRS at one
  318. 00:15:32
    or better for a baseline MRS. One of
  319. 00:15:35
    the things that we had the most
  320. 00:15:37
    discussion about was what do you, what
  321. 00:15:39
    do we define as early surgery? And when
  322. 00:15:43
    we first were planning to study out and
  323. 00:15:45
    we, our partner, Mike Frankl,
  324. 00:15:50
    was the greatest advocate for trying to
  325. 00:15:51
    push the time as early as absolutely
  326. 00:15:53
    possible. And so he was the big
  327. 00:15:57
    advocate for that. And so we ended up
  328. 00:15:59
    making a cut point at 24 hours. And
  329. 00:16:01
    that was really driven by what is
  330. 00:16:04
    practical to some extent in our current
  331. 00:16:06
    world. And also being able to activate
  332. 00:16:10
    an operating room, identify patient,
  333. 00:16:12
    enroll a patient, et cetera. However,
  334. 00:16:15
    at the time, the only data that we had
  335. 00:16:18
    was from Stitch and from Misty, which
  336. 00:16:21
    we're obviously allowing much later
  337. 00:16:22
    enrollments. And from a technical
  338. 00:16:25
    standpoint, Gustavo and Chris could
  339. 00:16:27
    probably speak to this better than me.
  340. 00:16:29
    But there is a practical component that
  341. 00:16:32
    people were concerned about related to
  342. 00:16:33
    time. Gustavo, do you want to talk
  343. 00:16:35
    about that real quick?
  344. 00:16:37
    Sure, when we talked about getting
  345. 00:16:40
    imaging that we could use for
  346. 00:16:42
    intraoperative navigation and then to
  347. 00:16:43
    get the patient into the operating room,
  348. 00:16:46
    It really required a. a culture change
  349. 00:16:49
    because most centers did not consider an
  350. 00:16:53
    interest review of hemorrhage procedure
  351. 00:16:55
    as an emergency. Maybe an urgent case,
  352. 00:16:58
    and those definitions vary from hospital
  353. 00:17:01
    to hospital, but for our site, an
  354. 00:17:03
    urgent case is within six hours of
  355. 00:17:06
    posting the case. But we wanted to have
  356. 00:17:09
    this as the most
  357. 00:17:13
    rapid emergency response case So it
  358. 00:17:16
    would be equal to an epidural hematoma
  359. 00:17:19
    or something like that. So it required
  360. 00:17:22
    a culture change, and we were able to
  361. 00:17:26
    achieve that in our center, but we saw
  362. 00:17:30
    a lot of pushback in other places. So
  363. 00:17:33
    when we did our site selection, we had
  364. 00:17:39
    these conversations very early on, and
  365. 00:17:43
    we asked our investigators at every site,
  366. 00:17:46
    to meet with their OR directors and OR
  367. 00:17:47
    managers, and go to the OR committee
  368. 00:17:51
    and change the priority level allowed
  369. 00:17:54
    for an interest rate of hemorrhage so
  370. 00:17:56
    that it would be done as quickly as
  371. 00:17:58
    possible. That was a significant
  372. 00:18:00
    culture change. It's a little bit
  373. 00:18:02
    different to how the workflow happens
  374. 00:18:08
    for Chris and the sign I team, because
  375. 00:18:13
    having control of the industry really
  376. 00:18:15
    allows them to go whenever they want,
  377. 00:18:18
    but when you're sharing the operating
  378. 00:18:19
    room with all the other hospital
  379. 00:18:22
    surgical specialties, you have a lot
  380. 00:18:24
    less control and you need to have this
  381. 00:18:27
    apiary agreement that you will go and
  382. 00:18:30
    you will bump whatever elective surgery
  383. 00:18:33
    is supposed to happen to get these
  384. 00:18:35
    patients into the operating room.
  385. 00:18:39
    Yeah, so the other issue that we
  386. 00:18:41
    addressed was uncorrectable coagulopathy.
  387. 00:18:44
    If you had a coagulopathy that
  388. 00:18:46
    couldn't be corrected, then we didn't
  389. 00:18:48
    want to enroll you. We knew at this
  390. 00:18:50
    time, we had Doax that they were on the
  391. 00:18:52
    market when we planned the study, and
  392. 00:18:54
    Idris's a map had just become available,
  393. 00:18:57
    and Dexnet Alpha was still actually in
  394. 00:18:59
    clinical trials and hadn't yet been
  395. 00:19:01
    approved. So we just said that they
  396. 00:19:04
    couldn't be enrolled if we couldn't
  397. 00:19:06
    correct the Quagulopathy regardless of
  398. 00:19:07
    the etiology of that Quagulopathy.
  399. 00:19:11
    And then we also excluded patients with
  400. 00:19:14
    profound IVH or significant IVH, and we
  401. 00:19:17
    defined that at the 50th percentile just
  402. 00:19:19
    because we had to provide a metric per
  403. 00:19:20
    cut point. So if greater than 50 of a
  404. 00:19:23
    ventricle was occluded, then we
  405. 00:19:25
    excluded those patients. And that's
  406. 00:19:28
    just related to poorer outcome and more
  407. 00:19:31
    difficult difficulty in managing the
  408. 00:19:33
    patient.
  409. 00:19:36
    And that really rounds out, I think,
  410. 00:19:37
    the major inclusion and exclusion
  411. 00:19:38
    criteria that Alex or Gustavo, did I
  412. 00:19:42
    forget any?
  413. 00:19:44
    No, I think that's the key in there.
  414. 00:19:47
    We had one of the key differentiators
  415. 00:19:49
    was time, because we wanted four hours,
  416. 00:19:52
    but we encouraged all of our
  417. 00:19:54
    investigators to enroll within eight
  418. 00:19:56
    hours of last no normal. And what for
  419. 00:19:59
    us, in practical terms, it meant that
  420. 00:20:01
    you were able to cross the door to the
  421. 00:20:04
    operating room within eight hours of
  422. 00:20:06
    last no normal, and within maximum of
  423. 00:20:09
    24 hours. Yeah, that's what we asked
  424. 00:20:12
    people to target that. People did
  425. 00:20:14
    pretty well with that, honestly,
  426. 00:20:16
    pretty amazingly throughout the trial,
  427. 00:20:18
    and I'm sorry for stepping on you and
  428. 00:20:19
    Gustavo, but the randomization from
  429. 00:20:22
    randomization to surgery throughout the
  430. 00:20:24
    trial was less than two hours, and most
  431. 00:20:26
    of the time it was about an hour and a
  432. 00:20:27
    half, which is we were pretty happy
  433. 00:20:31
    with that, given that many of these
  434. 00:20:32
    places had to deal with the complexities
  435. 00:20:35
    that biggest I was just describing in
  436. 00:20:37
    terms of their activating and operating
  437. 00:20:40
    room on an emergent case and changing
  438. 00:20:41
    the culture. But the investigators at
  439. 00:20:43
    each of the 37 sites were amazing at
  440. 00:20:46
    producing that effect.
  441. 00:20:49
    So, Alex mentioned, you know, the
  442. 00:20:52
    study was, was designed with adaptation
  443. 00:20:56
    in mind, like all these adaptive trials.
  444. 00:21:00
    We had a sample size that could
  445. 00:21:01
    fluctuate between 150 and 300 patients
  446. 00:21:05
    And we were doing block randomization
  447. 00:21:07
    based in two factors. The first one was
  448. 00:21:10
    a GCS, and we used greater or less than
  449. 00:21:14
    nine. And the second one was the
  450. 00:21:17
    location. So, until you're based again,
  451. 00:21:20
    yeah, or lower, where are the two
  452. 00:21:21
    locations And Alex, you want to talk
  453. 00:21:24
    about our interim analysis.
  454. 00:21:30
    Yeah, so at 150 patients and every 25
  455. 00:21:34
    subsequent patients, there was an
  456. 00:21:35
    interim analysis of the primary endpoint,
  457. 00:21:42
    the Bayesian analysis of the utility
  458. 00:21:44
    weighted MRS endpoint that was presented
  459. 00:21:48
    to the DSMB.
  460. 00:21:51
    And if it met a certain threshold for
  461. 00:21:54
    efficacy or futility, which in that
  462. 00:21:56
    case, futility would be stopping and
  463. 00:21:59
    enriching the other to one location only,
  464. 00:22:02
    the DSMB would make a recommendation
  465. 00:22:05
    based on an unblinded review of the
  466. 00:22:09
    analysis. In addition, they, of
  467. 00:22:10
    course, looked at safety across both
  468. 00:22:13
    groups and
  469. 00:22:16
    performance and several other metrics as
  470. 00:22:18
    sort of standard with any DSMB
  471. 00:22:21
    And so, at 175 patients, that's when
  472. 00:22:24
    we enriched to the low bar location only.
  473. 00:22:29
    which was pre-planned possibility in any
  474. 00:22:32
    25 increment of patients. And so they
  475. 00:22:35
    had to have at least their
  476. 00:22:39
    90-day outcome complete to be included
  477. 00:22:41
    in the interim analyses. And, you know,
  478. 00:22:46
    we had to have at least 60 patients with
  479. 00:22:49
    full outcomes at 180 days to be able to
  480. 00:22:52
    make any decision rules happen So that's
  481. 00:22:56
    what happened in the remaining of the
  482. 00:22:57
    patients. We enrolled in just the low
  483. 00:23:00
    bar locations. So
  484. 00:23:03
    at the end of the day, or at the end of
  485. 00:23:06
    the trial,
  486. 00:23:11
    we would have 180 days on everybody.
  487. 00:23:14
    And to claim superiority had to be
  488. 00:23:17
    greater than 0975 of the posterior
  489. 00:23:19
    probability So that was the target, the
  490. 00:23:22
    inner analysis for efficacy was much
  491. 00:23:24
    higher as
  492. 00:23:26
    099 greater than 099 so.
  493. 00:23:29
    And that was our model we were working
  494. 00:23:31
    with throughout the entire trial.
  495. 00:23:36
    We, are we audio recorded all the MRS
  496. 00:23:39
    interviews at 30, 90 and under 80 days.
  497. 00:23:44
    And then we had a central adjudicator,
  498. 00:23:46
    a neuropsychologist, and depending on
  499. 00:23:49
    the review, the
  500. 00:23:53
    audio file and listen to that and then
  501. 00:23:55
    provide a unblinded
  502. 00:24:00
    adjudication to the MRS based on the
  503. 00:24:01
    interview by the site. And so we had
  504. 00:24:04
    some background things that happened.
  505. 00:24:07
    They didn't get to see what the site
  506. 00:24:10
    entered until the neuropsychologist
  507. 00:24:13
    entered it. And if there was a mismatch,
  508. 00:24:14
    then that's when there had to have a
  509. 00:24:16
    conflict resolution of the two values.
  510. 00:24:18
    But because if you're doing an interview
  511. 00:24:21
    with a patient and they have a surgical
  512. 00:24:23
    scar on the forehead or whatever, it's
  513. 00:24:25
    pretty obvious and the interviewer can't
  514. 00:24:28
    stay blinded. So this was our best case
  515. 00:24:32
    scenario to try to do a blinded outcome.
  516. 00:24:39
    Can you actually comment on the decision
  517. 00:24:42
    to use the utility weighted MRS rather
  518. 00:24:45
    than just a plain MRS
  519. 00:24:50
    You immediately, Jonathan.
  520. 00:24:54
    So, sure, the, the decision to use
  521. 00:24:56
    utility weighted MRS gave us several
  522. 00:24:58
    advantages. It had been previously
  523. 00:25:01
    described with, with its utility
  524. 00:25:03
    weights and so it's, it's a patient
  525. 00:25:05
    centered and clinician centered
  526. 00:25:07
    adaptation of the MRS that provides
  527. 00:25:10
    these utility weights. The nice thing
  528. 00:25:12
    about the utility weighted MRS, unlike
  529. 00:25:14
    the standard MRS is that is comes in
  530. 00:25:16
    through in an analysis It provides you a
  531. 00:25:18
    little bit more power because it changes
  532. 00:25:20
    the variable to a continuous variable.
  533. 00:25:23
    So your analysis is a little bit more
  534. 00:25:25
    powerful in that you're not losing
  535. 00:25:26
    things between categories as much.
  536. 00:25:29
    And that is the major, that's the major
  537. 00:25:31
    differences from the analysis. We still
  538. 00:25:34
    collected the MRS and then, and then
  539. 00:25:36
    use the utility weights of the MRS to do
  540. 00:25:38
    the analysis
  541. 00:25:40
    We also did the logistic regression for
  542. 00:25:44
    the ordinal MRS at all the time points.
  543. 00:25:47
    We'll show you the results of that.
  544. 00:25:49
    Yeah, that was a secondary analysis.
  545. 00:25:51
    Yeah, but the primary outcome measure
  546. 00:25:53
    was analyzed using adivation methodology
  547. 00:25:57
    and the posterior probability of success.
  548. 00:26:00
    One of the things that it's hard for
  549. 00:26:04
    people to digest is what futility means
  550. 00:26:09
    in this studies. And it's a very
  551. 00:26:13
    statistical definition So when we got to
  552. 00:26:16
    175 patients and the
  553. 00:26:22
    statisticians,
  554. 00:26:24
    the projection based on the 90-day
  555. 00:26:26
    outcomes, they looked at all the
  556. 00:26:27
    scenarios and they saw that it was
  557. 00:26:30
    statistically unlikely that we would get
  558. 00:26:32
    to a positive trial if we kept enrolling
  559. 00:26:35
    both the basal gameness and the low
  560. 00:26:37
    birth. Now it doesn't, so there was
  561. 00:26:41
    statistical futility Now, that doesn't
  562. 00:26:44
    mean that there's a procedure for
  563. 00:26:46
    basically, and it's futile. As you
  564. 00:26:48
    will see when we share the results,
  565. 00:26:51
    we don't know. That's still an
  566. 00:26:53
    unanswered question. And I'm sure Chris
  567. 00:26:57
    has a lot of very, very positive cases
  568. 00:27:01
    in basal ganglia evacuations that do
  569. 00:27:03
    really well. We had the same. We just
  570. 00:27:06
    had to make it sort of an executive
  571. 00:27:09
    decision when you have a limited sample
  572. 00:27:10
    size and limited time. You have to
  573. 00:27:14
    focus on the population that's going to
  574. 00:27:16
    give you the highest yield And that's
  575. 00:27:18
    what that adaptation took place.
  576. 00:27:24
    anything else we should talk about
  577. 00:27:27
    before we get into the results. We're
  578. 00:27:29
    at 37 minutes.
  579. 00:27:33
    I think that's good. That's pretty
  580. 00:27:34
    thorough.
  581. 00:27:37
    I just have one
  582. 00:27:40
    question
  583. 00:27:42
    that I was wondering about as far as the
  584. 00:27:45
    inclusion exclusion is patients in other
  585. 00:27:47
    trials that had to wait for stability of
  586. 00:27:49
    the ACH And going so fast that you
  587. 00:27:52
    didn't get that opportunity, how did
  588. 00:27:54
    that affect? That's a great, I love
  589. 00:27:57
    that question. I love that question too.
  590. 00:28:00
    We saw that actually as a
  591. 00:28:03
    differentiating factor and a strength.
  592. 00:28:06
    So whenever we saw a spot sign, we
  593. 00:28:09
    actually used that as a target. We went
  594. 00:28:11
    into a land right on top of the spot
  595. 00:28:14
    sign because we could control it and get
  596. 00:28:17
    human stasis and stop it. So it was
  597. 00:28:20
    never a deterrent. It was actually an
  598. 00:28:22
    opportunity. We thought that
  599. 00:28:25
    we had the greatest possibility of
  600. 00:28:28
    changing the outcome of the patient in a
  601. 00:28:30
    spot sign.
  602. 00:28:32
    Yeah, remember that part of our
  603. 00:28:37
    planning for early surgery was obviously
  604. 00:28:39
    we wanna relieve the mechanical pressure.
  605. 00:28:41
    We wanted to interrupt the inflammatory
  606. 00:28:43
    cascade, but really a big one was early
  607. 00:28:45
    hemostasis and prevent clot expansion
  608. 00:28:48
    And that was a primary motivators for us.
  609. 00:28:57
    All right, so I think we can start
  610. 00:29:00
    sharing with you our results. So
  611. 00:29:06
    when we looked at assessment of patients,
  612. 00:29:09
    we had 37 centers in the US that
  613. 00:29:12
    participated in the study, and every
  614. 00:29:15
    site had a neurosurgeon and a neuro
  615. 00:29:17
    intensivist as co-site investigators for
  616. 00:29:20
    the study, and that was instrumental
  617. 00:29:23
    for the success of the trial in terms of
  618. 00:29:26
    the logistics. We assessed 11, 600
  619. 00:29:30
    patients for eligibility. Alex, you
  620. 00:29:33
    want to run through the assessment and
  621. 00:29:36
    the numbers real quick?
  622. 00:29:39
    Sure, the majority of we sort of had an
  623. 00:29:43
    all we didn't clean any of the screening
  624. 00:29:45
    data, it's whatever the site submitted
  625. 00:29:46
    to us. The overwhelming majority of the
  626. 00:29:50
    screening data weren't our target
  627. 00:29:53
    population around
  628. 00:29:55
    3, 700 of those weren't spontaneous,
  629. 00:30:00
    intraprenechimal hemorrhages.
  630. 00:30:03
    Some of them, another 2, 700 or so
  631. 00:30:05
    where the volumes were too large and
  632. 00:30:08
    then age was another big exclusion
  633. 00:30:10
    criteria factor that screened them out.
  634. 00:30:13
    So of those that were screened, 300 are
  635. 00:30:16
    randomized, 150 to the medical
  636. 00:30:19
    management group and 150 to the surgical
  637. 00:30:21
    group or MIPS
  638. 00:30:25
    And at 175 patients that we enriched.
  639. 00:30:27
    And so the final location count, there
  640. 00:30:32
    were 48 in the
  641. 00:30:35
    anterior basal ganglia group for the
  642. 00:30:36
    surgery and 102 for low var and surgery.
  643. 00:30:40
    44
  644. 00:30:43
    basal ganglia in the medical management
  645. 00:30:44
    and 106 in the low var for medical
  646. 00:30:47
    management.
  647. 00:30:49
    There were three lost to follow ups and
  648. 00:30:51
    the surgical group
  649. 00:30:54
    11 lost the follow-ups or withdrawals of
  650. 00:30:56
    consent in the medical management group.
  651. 00:31:00
    So our observed MRS at six months was
  652. 00:31:03
    147 in the surgical group and 139 in the
  653. 00:31:07
    medical group. So we had a pretty good
  654. 00:31:09
    retention rate overall and certainly
  655. 00:31:11
    maintained above our threshold for
  656. 00:31:30
    I mean, you provide for context the
  657. 00:31:33
    number of patients that were screened
  658. 00:31:36
    for MISTI.
  659. 00:31:39
    I don't remember on the top of my head,
  660. 00:31:41
    but I think it was in the 14, 000 in
  661. 00:31:43
    the range It was almost 15, 000. So
  662. 00:31:46
    this is a lot less patient. I mean,
  663. 00:31:48
    MISTI had more patients was
  664. 00:31:52
    500, right? Yeah, 500. So when you
  665. 00:31:55
    look through those,
  666. 00:31:58
    it's very comparable based on the sample
  667. 00:32:01
    size that we had of the patient
  668. 00:32:02
    selection
  669. 00:32:05
    And it's a frequent question that we got
  670. 00:32:08
    in the presentation of ESOC was
  671. 00:32:12
    all the patients that were excluded and
  672. 00:32:15
    what those reasons are. So I think Alex
  673. 00:32:18
    right into those really well.
  674. 00:32:22
    I think next we can talk about our
  675. 00:32:26
    primary outcome measure result. I
  676. 00:32:33
    guess we can add that, basically for
  677. 00:32:36
    all of our baseline characteristics,
  678. 00:32:37
    there really wasn't any difference
  679. 00:32:39
    between groups. It was all
  680. 00:32:41
    well-balanced. And so there's nothing
  681. 00:32:46
    substantial. Our median age for both
  682. 00:32:51
    groups was around 64. Years old, and
  683. 00:32:56
    so other than that, that was everything
  684. 00:32:59
    was well-balanced.
  685. 00:33:03
    Yeah, when we look at the primary
  686. 00:33:05
    outcome measure, which is the
  687. 00:33:06
    difference in utility-weighted modified
  688. 00:33:08
    rankings course, the difference between
  689. 00:33:14
    the
  690. 00:33:17
    surgical group and the medical group was
  691. 00:33:19
    0084.
  692. 00:33:23
    that was significant with a posterior
  693. 00:33:27
    probability of 098 or
  694. 00:33:32
    98 probability. So that was a very
  695. 00:33:36
    overwhelming, positive result for the
  696. 00:33:40
    entire study population. When we looked
  697. 00:33:43
    at the two locations, the basic anglia
  698. 00:33:47
    had a neutral effect, was at negative
  699. 00:33:51
    0013, with patients almost right down
  700. 00:33:54
    the middle in between the right and left
  701. 00:33:57
    side of the treatment effect. But the
  702. 00:34:00
    low bar location had a difference of
  703. 00:34:02
    0127 with a posterior probability of
  704. 00:34:05
    0996
  705. 00:34:08
    or 997 probability of benefit of the
  706. 00:34:12
    intervention. So the results were
  707. 00:34:15
    strongly driven by the lower location,
  708. 00:34:18
    and
  709. 00:34:20
    that was the primary result. So, That
  710. 00:34:23
    was the most significant and where we
  711. 00:34:26
    were most excited about. When we look
  712. 00:34:30
    at the distribution of those across the
  713. 00:34:33
    different MRS categories, we saw those
  714. 00:34:37
    results represented, of course, in
  715. 00:34:39
    more patients in the MRS categories of
  716. 00:34:42
    zero, one, and two in the surgical
  717. 00:34:46
    group and more patients with the five
  718. 00:34:49
    and six outcomes in the medical
  719. 00:34:51
    management group as well.
  720. 00:34:54
    We looked at the results of the MRS at
  721. 00:35:00
    all the other time points, including
  722. 00:35:07
    day seven or discharge, day 30, day 90,
  723. 00:35:10
    and day 180. And we did audiologist
  724. 00:35:14
    regression for all of those time points
  725. 00:35:16
    And across all of those surgical
  726. 00:35:20
    treatment was statistically severe. the
  727. 00:35:22
    area to medical management.
  728. 00:35:28
    Can I ask you there? Was that only in
  729. 00:35:29
    the low bar groups across those? Or was
  730. 00:35:31
    that in? This includes both those.
  731. 00:35:35
    Both cohorts. And that's a good
  732. 00:35:37
    question. Any data we talk about is
  733. 00:35:38
    going to include both cohorts unless we
  734. 00:35:40
    specifically say it. Yeah, Chris,
  735. 00:35:42
    good question. And that's a good, the
  736. 00:35:42
    mean I. point
  737. 00:35:45
    adaptive design and the analysis, you
  738. 00:35:48
    do the Bayesian analysis is the overall
  739. 00:35:51
    population. Which is really to control
  740. 00:35:55
    the false positive rates or type one
  741. 00:35:57
    error.
  742. 00:35:59
    So you have to maintain your primary
  743. 00:36:01
    analysis plan, which is all comers in
  744. 00:36:04
    the trial. And so all of our normal
  745. 00:36:06
    logistic regression was also, as I just
  746. 00:36:09
    said, that way. But it goes to the
  747. 00:36:13
    questions of, well, what shifts that
  748. 00:36:18
    the overall effect is primarily, seen
  749. 00:36:21
    from low bar, which is expected if you
  750. 00:36:23
    think about the design of the trial.
  751. 00:36:24
    When you stop one, what you really have
  752. 00:36:27
    is about 50 that are beneficial and 50
  753. 00:36:30
    that are harm in the basal ganglia group.
  754. 00:36:32
    So it's somewhere right in the middle.
  755. 00:36:33
    So it stopped at a reasonable spot and
  756. 00:36:38
    resource allocated to the low bar group.
  757. 00:36:41
    And so, you know, if you think about
  758. 00:36:43
    the where they all set and sort of
  759. 00:36:47
    treatment benefit that, you know, sort
  760. 00:36:50
    of intuitively makes sense of the
  761. 00:36:51
    observed data. And then when you
  762. 00:36:54
    articulate the conclusion, do you have
  763. 00:36:57
    to stipulate that the conclusion applies
  764. 00:36:59
    only to low bar hemorrhages? Or do you
  765. 00:37:02
    still state in the conclusion this
  766. 00:37:04
    applies to surgically treated
  767. 00:37:06
    hemorrhages? How do we understand that
  768. 00:37:09
    after the fact when we've got the data
  769. 00:37:11
    and we're thinking, are we applying
  770. 00:37:13
    this to surgical treatment of basal
  771. 00:37:15
    ganglion low bar? Or are we just
  772. 00:37:17
    applying it to low bar in the end?
  773. 00:37:20
    I think, and I'll let, I'll let both,
  774. 00:37:21
    Jonathan, you can't comment. The
  775. 00:37:23
    conclusion is.
  776. 00:37:29
    Sorry. Sorry, I don't know. You're
  777. 00:37:30
    breaking up there. You're breaking up
  778. 00:37:32
    this time. Like we were losing or
  779. 00:37:35
    breaking up, yeah. So the conclusion
  780. 00:37:38
    is for both populations. The study was
  781. 00:37:41
    positive overall for the surgical group.
  782. 00:37:45
    But we do put a comment at the end that
  783. 00:37:47
    the results appear to be driven by the
  784. 00:37:50
    strongly positive effect seeing in the
  785. 00:37:52
    lower location. Yeah, when you break
  786. 00:37:54
    down. Go ahead, go ahead. I'm sorry.
  787. 00:37:57
    No I, was gonna say, and for the basic
  788. 00:37:59
    thing there, all we say is that it
  789. 00:38:02
    remains on an answer question. Right.
  790. 00:38:06
    Yeah, it's the post-year probability of
  791. 00:38:07
    success for the basal ganglia group is
  792. 00:38:09
    right at about 50, 48 or something like
  793. 00:38:12
    that, which means it's the flip of a
  794. 00:38:14
    coin still, right? that we do, we
  795. 00:38:16
    don't know.
  796. 00:38:18
    I think that the way to interpret it is
  797. 00:38:20
    exactly what Gustavo just described and
  798. 00:38:21
    how we're describing it in the paper and
  799. 00:38:23
    how we're describing it when we talk
  800. 00:38:24
    about it is that the whole trial is
  801. 00:38:26
    positive, but it's important to
  802. 00:38:29
    understand that that's driven by these
  803. 00:38:31
    low bar groups, which we then also
  804. 00:38:32
    selected because we were seeing that
  805. 00:38:35
    that was going to lead because they were
  806. 00:38:38
    meeting the predefined criteria for
  807. 00:38:41
    stopping so that we had the most likely
  808. 00:38:44
    successful trial
  809. 00:38:47
    Yeah, it seems like you guys did a very
  810. 00:38:49
    good job planning the trial in a way,
  811. 00:38:52
    'cause you're coming into a space where
  812. 00:38:54
    so many trials have been negative.
  813. 00:38:56
    Missed he was ongoing and had a certain
  814. 00:38:57
    strategy. So it seems like along all
  815. 00:39:01
    the decision-making points about how to
  816. 00:39:03
    design the trial, you designed it in
  817. 00:39:04
    the right way to maximize the chance of
  818. 00:39:06
    getting an answer that we could build on
  819. 00:39:09
    in the ICH community. And then in the
  820. 00:39:12
    end, you know, that all worked out So
  821. 00:39:15
    really commend you on the design.
  822. 00:39:21
    Thank you. I think we had seen the
  823. 00:39:28
    situation that evolved from Stitch,
  824. 00:39:30
    where Lovar gave so many good hints on
  825. 00:39:33
    Stitch 1, and they were very close from
  826. 00:39:37
    Stitch 2. They were, you know, many
  827. 00:39:41
    design.
  828. 00:39:44
    I want to say flaws, but just
  829. 00:39:46
    characteristics that maybe prevented
  830. 00:39:50
    them from reaching the statistical
  831. 00:39:53
    significance that was needed, but they
  832. 00:39:55
    were close, and we were very aware of
  833. 00:39:57
    that. That's what we thought are
  834. 00:39:59
    selling out the locations and analyzing
  835. 00:40:01
    them separately was going to be
  836. 00:40:04
    important to the success of the overall
  837. 00:40:08
    study We can add
  838. 00:40:12
    the secondary endpoints that
  839. 00:40:17
    we're not going to tell you all of them
  840. 00:40:18
    because we need to reserve some for the
  841. 00:40:20
    many manuscript results, there are some
  842. 00:40:22
    that we've presented in the past.
  843. 00:40:25
    Jonathan, you want to go over, I guess,
  844. 00:40:28
    you get a pic. Yeah, I'll highlight
  845. 00:40:32
    what we presented, Esauch. You know,
  846. 00:40:34
    one of the things that we were most
  847. 00:40:37
    interested in looking at, and the
  848. 00:40:39
    mechanism through which we hypothesize
  849. 00:40:40
    that surgical evacuation is effective,
  850. 00:40:43
    as we've described, it's releasing
  851. 00:40:46
    mechanical pressure, interrupting of
  852. 00:40:48
    inflammatory cascade, et cetera. And
  853. 00:40:50
    so we were very concerned about the end
  854. 00:40:51
    of treatment volume. And so in our
  855. 00:40:55
    treatment group, the median end of
  856. 00:40:58
    treatment volume was 14 CCs. And we had
  857. 00:41:03
    a change in ICH volume of about 44 CCs.
  858. 00:41:07
    And so that resulted in a percentile
  859. 00:41:10
    reduction of about 87, 88 of all of the
  860. 00:41:14
    ICH. So. very, you know, I think
  861. 00:41:17
    that we can say we feel it so that the
  862. 00:41:20
    the surgery was very effective at
  863. 00:41:22
    removing the volume of the hemorrhage
  864. 00:41:25
    and that seems to be associated with the
  865. 00:41:27
    final outcome as well that the the
  866. 00:41:29
    greater the extent of hemorrhage removal
  867. 00:41:32
    the better those patients seem to do
  868. 00:41:33
    which I think is at this point not
  869. 00:41:36
    surprising but a bit perhaps a bit of a
  870. 00:41:38
    relief to see because it adds biologic
  871. 00:41:40
    plausibility to the surgical evacuation
  872. 00:41:43
    and further pursuit of it as we as we
  873. 00:41:45
    move forward in with these the low bar
  874. 00:41:47
    and the thalamic hemorrhages as well.
  875. 00:41:50
    We also saw a pretty significant
  876. 00:41:52
    reduction in mortality. The mortality
  877. 00:41:56
    in the medical management group was very
  878. 00:41:58
    similar to what we'd seen in the MISTI
  879. 00:42:02
    group in their placebo arm at about 18
  880. 00:42:05
    percent whereas in the surgically
  881. 00:42:07
    treated group it was a nine nine percent.
  882. 00:42:10
    And that's a 30 days. That's a 30 days
  883. 00:42:13
    I'm sorry. Thank you, Alex. Yeah,
  884. 00:42:14
    that's a 30-day mortality Um, in
  885. 00:42:18
    addition to that, though, we also saw
  886. 00:42:20
    production in the mean ventilator days,
  887. 00:42:23
    uh, the surgical group was only at
  888. 00:42:26
    about five days, whereas the medical
  889. 00:42:27
    management group was about nine days of,
  890. 00:42:29
    of ventilator usage, and then an
  891. 00:42:31
    associated decrease in, uh, overall
  892. 00:42:34
    hospital length of stay and ICU length
  893. 00:42:35
    of stay. So all of these kind of
  894. 00:42:37
    pointed to the patients were getting
  895. 00:42:39
    better quicker, uh, even early in
  896. 00:42:41
    their course, uh, in the, in the
  897. 00:42:43
    surgical group
  898. 00:42:46
    I wonder, I don't know if you actually
  899. 00:42:49
    looked at this, but did you find
  900. 00:42:50
    significant variability in the amount of
  901. 00:42:53
    evacuation amongst different procedure
  902. 00:42:55
    lists, or did you feel like, I think
  903. 00:42:58
    with all sort of procedural studies, we
  904. 00:43:01
    sort of wonder if the, um, the
  905. 00:43:05
    experience of the procedure list sort of
  906. 00:43:08
    impacts your outcomes? Sure. And
  907. 00:43:12
    especially, I think for something that,
  908. 00:43:14
    um, is not a, very
  909. 00:43:19
    common procedure to be done. I sort of
  910. 00:43:21
    wonder what it looked like amongst the
  911. 00:43:24
    different centers. Yeah, that's a
  912. 00:43:27
    great question. So when we looked at
  913. 00:43:31
    performance, we only had, I'll explore,
  914. 00:43:35
    I think if I'm Roman, I think it's
  915. 00:43:36
    seven outliers in terms of the
  916. 00:43:39
    volume evacuation. And those were
  917. 00:43:44
    pretty diluted in terms of the sites.
  918. 00:43:48
    There was not one side that had more
  919. 00:43:50
    than the others. They were isolated
  920. 00:43:52
    events. And across the different
  921. 00:43:55
    centers, there were different numbers
  922. 00:43:57
    of surgeons that were part of the study.
  923. 00:43:59
    Some centers had one, some center had
  924. 00:44:01
    three or four. So we had a large number
  925. 00:44:03
    of surgeons that were participating in
  926. 00:44:05
    the study. And I think overall, the
  927. 00:44:08
    execution of the surgical protocol was
  928. 00:44:11
    very well done. We were recording all
  929. 00:44:13
    of the procedures. quality, we
  930. 00:44:16
    reviewed the first two that were done to
  931. 00:44:19
    make sure that they adhere to the
  932. 00:44:20
    surgical protocol. And then we did ad
  933. 00:44:22
    hoc reviews of the videos. If there was
  934. 00:44:25
    a concern of the volume at the end of
  935. 00:44:28
    the procedure not being good enough, we
  936. 00:44:30
    would go back and see if there were
  937. 00:44:32
    technical issues. It looked like it was
  938. 00:44:35
    difficult to get human statuses. And we
  939. 00:44:37
    also have the feedback from the surgeons
  940. 00:44:40
    about the procedure, you know, what
  941. 00:44:42
    they use, what they use for human
  942. 00:44:44
    statuses. I was developed after during
  943. 00:44:45
    the procedure, was this a case that was
  944. 00:44:48
    anti-regulation was reversed. We looked
  945. 00:44:51
    at the surgical performance in the
  946. 00:44:53
    context of all of those features to be
  947. 00:44:56
    able to see that. And then another
  948. 00:44:58
    important thing is that we had people
  949. 00:45:00
    with different levels of experience and
  950. 00:45:02
    we also had community hospitals within
  951. 00:45:05
    the 37 sites. Not all were academic
  952. 00:45:08
    centers and we saw good performance
  953. 00:45:11
    across all of those sites. Do you see
  954. 00:45:14
    any features that predicted maybe worse
  955. 00:45:18
    evacuation percentage like presence of a
  956. 00:45:21
    spot sign, you're targeting the spot
  957. 00:45:22
    sign, you're coagulating it, but maybe
  958. 00:45:24
    that makes it a little harder to do the
  959. 00:45:26
    procedure and get everything out, or
  960. 00:45:27
    maybe you encounter bleeding a little
  961. 00:45:28
    more often, or maybe early evacuation
  962. 00:45:31
    or deeper slow bar, maybe deeper, it's
  963. 00:45:35
    harder to really see the whole cavity
  964. 00:45:36
    when you're in there with the end of
  965. 00:45:37
    port. Did you see anything like that?
  966. 00:45:41
    Not yet, we haven't finalized looking
  967. 00:45:45
    at all of the surgical factors that I
  968. 00:45:49
    think is study number four in our list
  969. 00:45:53
    of things. But while we looked at
  970. 00:45:55
    preliminary spot signs was not a problem,
  971. 00:45:59
    it was actually very favorable
  972. 00:46:04
    plot evacuations with spot signs. In
  973. 00:46:08
    the overall study, I think we Thank you
  974. 00:46:10
    Ahh.
  975. 00:46:13
    What was our group leading rate, Alex?
  976. 00:46:17
    3 is 3 for the surgical group, so was
  977. 00:46:22
    there a level? Yeah, so 3
  978. 00:46:26
    which included spot signs, I think it's
  979. 00:46:29
    a very nice number to see and speaks to
  980. 00:46:29
    the ease of
  981. 00:46:32
    the endostasis. And this is
  982. 00:46:39
    a moving target, so as a side of being
  983. 00:46:42
    better and better. You know, I think
  984. 00:46:45
    when we started doing this in CRIS for
  985. 00:46:48
    endoscopic techniques as well, people
  986. 00:46:50
    say this devices and adjuvants weren't
  987. 00:46:54
    as good. Things keep getting better,
  988. 00:46:57
    coagulation systems, bipolar cottery
  989. 00:46:59
    devices for a different methodologist,
  990. 00:47:02
    keep getting better year after year, so
  991. 00:47:05
    I only expect that number to keep
  992. 00:47:07
    getting better.
  993. 00:47:09
    I think that's really nice thing about
  994. 00:47:10
    this technique you can apply to a broad
  995. 00:47:13
    set of ICH patients. You don't have to
  996. 00:47:15
    have stability. You don't have to
  997. 00:47:16
    exclude spot-side patients. And then
  998. 00:47:18
    when you treat those patients who are
  999. 00:47:20
    particularly high risk, you're still
  1000. 00:47:22
    able to handle it in the case. And you
  1001. 00:47:23
    guys have shown that. So that's, I
  1002. 00:47:25
    think definitely one of the things that
  1003. 00:47:27
    maybe unlock the positive result here is
  1004. 00:47:30
    being able to take patients earlier,
  1005. 00:47:31
    being able to handle challenging cases.
  1006. 00:47:35
    Can someone comment on the parapecicular
  1007. 00:47:37
    approach, the surgical approach? I'm
  1008. 00:47:39
    not as familiar with that And I don't
  1009. 00:47:41
    know how much we think that might factor
  1010. 00:47:44
    in to the improved outcomes.
  1011. 00:47:48
    Yeah, that's a great question. So that
  1012. 00:47:52
    has a lot to do with
  1013. 00:47:56
    the post-mortem of Stitch. So when we
  1014. 00:47:59
    looked at how the surgical protocol was
  1015. 00:48:02
    defined in Stitch, there was a lot of
  1016. 00:48:05
    variability. There were some centers
  1017. 00:48:07
    that did very big operations. including
  1018. 00:48:11
    the hemichrymium, and a partial frontal
  1019. 00:48:14
    temporal lobectomy, and as aggressive a
  1020. 00:48:17
    clot evacuation as they could get. And
  1021. 00:48:19
    they were other centers that did a small
  1022. 00:48:22
    craniotomy, so both left way back on,
  1023. 00:48:24
    and they did a small institution in the
  1024. 00:48:27
    cortex, a cortisectomy, and did a
  1025. 00:48:28
    conservative clot evacuation. And when
  1026. 00:48:28
    we looked at
  1027. 00:48:36
    several of the studies that Dr.
  1028. 00:48:37
    Mendelow has published subsequently,
  1029. 00:48:41
    there was not a surgical awareness at
  1030. 00:48:44
    the time those studies were designed on
  1031. 00:48:48
    the impact of disconnection syndromes
  1032. 00:48:51
    when you're doing this type of outcome
  1033. 00:48:53
    measures, like a modified Rankine scale,
  1034. 00:48:56
    or Glasgow outcomes scores, and many of
  1035. 00:48:59
    these other outcome measures, those
  1036. 00:49:01
    disconnection syndromes that we create
  1037. 00:49:04
    surgically when we do a lobectomy, come
  1038. 00:49:06
    to bear and patients don't perform as
  1039. 00:49:08
    well. And
  1040. 00:49:12
    when we started using this minimal axis
  1041. 00:49:16
    sports, one of the important principles
  1042. 00:49:20
    in their design was to pick a trajectory
  1043. 00:49:23
    to the region that may not necessarily
  1044. 00:49:25
    be the shortest distance to the hematoma.
  1045. 00:49:29
    But it goes along the long axis of the
  1046. 00:49:32
    fiber tracks that are in the vicinity of
  1047. 00:49:35
    where the hematoma is happening so that
  1048. 00:49:38
    you can set the least amount of white
  1049. 00:49:40
    matter fibers. And you try to decrease
  1050. 00:49:43
    these disconnection syndromes as much as
  1051. 00:49:46
    possible.
  1052. 00:49:48
    For certain regions, that is less
  1053. 00:49:52
    critical. But if you're talking about
  1054. 00:49:55
    pulsating a super parietal lobe syndrome
  1055. 00:49:59
    or disconnecting the unsungent
  1056. 00:50:02
    fasciculus
  1057. 00:50:04
    at the frontal basal region, the impact
  1058. 00:50:07
    of those disconnection signals coming.
  1059. 00:50:09
    really profound. So, so that was the,
  1060. 00:50:12
    the concept behind these parapsicular
  1061. 00:50:15
    approaches is making the trajectory that
  1062. 00:50:17
    is most respectful to the white matter,
  1063. 00:50:20
    although it may not be the quickest,
  1064. 00:50:23
    shortest way to get to the pathology
  1065. 00:50:30
    And Gustavo, do you want to talk a
  1066. 00:50:31
    little bit to about the MRI studies that
  1067. 00:50:35
    you did and that we did before the trial
  1068. 00:50:38
    at this trajectories. Yeah, so we did
  1069. 00:50:42
    like space and, and several of the
  1070. 00:50:46
    studies for too many sections in the
  1071. 00:50:49
    same devices Because all of those
  1072. 00:50:52
    required diffusion tensor imaging we had
  1073. 00:50:57
    photography for all of those, and we
  1074. 00:51:01
    used it to find a trajectory and we also
  1075. 00:51:03
    used to test the footprint of those
  1076. 00:51:06
    devices after surgery. And the
  1077. 00:51:09
    trajectory that we came up with over and
  1078. 00:51:11
    over again, for certain locations, we
  1079. 00:51:14
    are pretty standard. So for the base of
  1080. 00:51:16
    ganglia, that trajectory was
  1081. 00:51:22
    pretty universal. It was a trajectory
  1082. 00:51:24
    that split two regions, the singular
  1083. 00:51:29
    fasciculus and superior longitudinal
  1084. 00:51:31
    fasciculus. And he provided a
  1085. 00:51:32
    longitudinal axis along the vector of
  1086. 00:51:38
    orientation of those two fascicles. So
  1087. 00:51:41
    the MR data was very good for that. For
  1088. 00:51:44
    some of the lower locations, we were
  1089. 00:51:47
    able to avoid disrupting the arkwood
  1090. 00:51:51
    fasciculus or the ascending Raymai of
  1091. 00:51:54
    the superior longitudinal fasciculus and
  1092. 00:51:56
    avoid the superior cryolobular
  1093. 00:51:58
    disconnection syndrome. We were very
  1094. 00:52:01
    successful of arkwood fasciculus fibers
  1095. 00:52:04
    the amount of speech deficits. was
  1096. 00:52:06
    decreased in those trajectories. And
  1097. 00:52:09
    then for
  1098. 00:52:13
    the access to the posterior phalamus and
  1099. 00:52:15
    the posterior part of the lateral
  1100. 00:52:17
    ventricles, we were able to be
  1101. 00:52:19
    respectful of the optic radiations and
  1102. 00:52:22
    the
  1103. 00:52:24
    connections of the SLF and the arcoid to
  1104. 00:52:28
    the back of the ventricle as well. So
  1105. 00:52:31
    those MR studies were instrumental in
  1106. 00:52:34
    building this, what we call now common
  1107. 00:52:37
    corridors of axis. And we, in certain
  1108. 00:52:41
    pathologies like tumors, we still get
  1109. 00:52:44
    tractography for those. In certain
  1110. 00:52:46
    locations like palliative hemorrhages,
  1111. 00:52:48
    we still like to get tractography to
  1112. 00:52:51
    plan. And in critical lower hemorrhages
  1113. 00:52:54
    adjacent to portacostpartum tract, for
  1114. 00:52:56
    example, we still get some tractography
  1115. 00:53:00
    to be able to plan a very respectful
  1116. 00:53:01
    approach if the hemorrhage is in front
  1117. 00:53:03
    of the surface. another question for
  1118. 00:53:07
    the Emory team. You were talking about
  1119. 00:53:09
    how at the beginning of the trial, some
  1120. 00:53:11
    centers were having a hard time arguing
  1121. 00:53:14
    that this was a emergent or maybe even
  1122. 00:53:15
    urgent type of case. And so now that we
  1123. 00:53:18
    have the results here, do we have an
  1124. 00:53:21
    answer to whether or not this is an
  1125. 00:53:22
    emergent case or an urgent case? Maybe
  1126. 00:53:25
    that comes down in a differential effect
  1127. 00:53:26
    of time that you'll see in your subgroup
  1128. 00:53:29
    analyses. But do we treat this like an
  1129. 00:53:33
    ischemic stroke and get that patient in
  1130. 00:53:34
    there immediately? Or do we go with the
  1131. 00:53:37
    next available room even if it's the
  1132. 00:53:40
    next day at hour 22? So what do you
  1133. 00:53:43
    guys think about that?
  1134. 00:53:48
    So, we can't show everything today. We
  1135. 00:53:52
    have 12 and eight-hour data that's going
  1136. 00:53:55
    to be included in the primary manuscript
  1137. 00:53:58
    that will be published. But the biggest
  1138. 00:54:03
    question in our mind was safety because,
  1139. 00:54:08
    of course, clinical status also bears
  1140. 00:54:11
    heavily into the decision to
  1141. 00:54:15
    intervene immediately in the patient.
  1142. 00:54:18
    And if you have a patient that is
  1143. 00:54:22
    clinically stable and there are some
  1144. 00:54:26
    logistical issues to get the patient to
  1145. 00:54:28
    the afternoon room immediately, we
  1146. 00:54:30
    still still benefit in good results
  1147. 00:54:33
    within that 24-hour enrollment window
  1148. 00:54:38
    We didn't get as many patients as we
  1149. 00:54:41
    wanted in what we would define as the
  1150. 00:54:43
    ultra-early meaning within eight hours
  1151. 00:54:47
    as we wanted to. And I think that is a
  1152. 00:54:50
    group that needs further study. All we
  1153. 00:54:54
    can comment on in regards to time is
  1154. 00:54:56
    that it was safe. There were some
  1155. 00:54:59
    concerns from older studies and maybe
  1156. 00:55:01
    other techniques in the paths about
  1157. 00:55:04
    re-leading with ultra-early intervention.
  1158. 00:55:08
    I think that we can definitely pass the
  1159. 00:55:11
    page on that and say that all of the
  1160. 00:55:14
    available techniques nowadays, what's
  1161. 00:55:16
    our early intervention is safe?
  1162. 00:55:23
    Yeah. And that's so important to know.
  1163. 00:55:24
    I mean, you were able to enroll
  1164. 00:55:26
    patients so much faster than Misty and
  1165. 00:55:28
    hopefully with the knowledge that you've
  1166. 00:55:29
    now brought to the table, the next
  1167. 00:55:30
    trial again, we'll be able to get more
  1168. 00:55:32
    of those ultra-early patients that
  1169. 00:55:33
    you're talking about with the comfort of
  1170. 00:55:35
    knowing that this procedure is safe.
  1171. 00:55:41
    Cara, I was going to have you comment a
  1172. 00:55:42
    little bit, I think Gustavo and
  1173. 00:55:44
    Jonathan both brought up really good
  1174. 00:55:46
    points about how both the timing of
  1175. 00:55:52
    evacuation and also the
  1176. 00:55:56
    short time period in which the patient
  1177. 00:55:59
    actually has active evacuation ongoing,
  1178. 00:56:03
    how do you see that sort of changing the
  1179. 00:56:04
    workflow in the neuro ICU and our
  1180. 00:56:07
    resource utilization for the average ICH
  1181. 00:56:10
    patient? Yeah, I think that's a good
  1182. 00:56:13
    question. I think this is very exciting,
  1183. 00:56:16
    first of all, to see this kind of
  1184. 00:56:17
    result for ICH patients and knowing as
  1185. 00:56:20
    I've always suspected that the faster we
  1186. 00:56:22
    get them to the OR, that the better
  1187. 00:56:24
    that outcomes are going to be. So I
  1188. 00:56:26
    think that it's going to be really, the
  1189. 00:56:28
    onus is going to be on those stroke
  1190. 00:56:31
    residents and the ED with that code
  1191. 00:56:33
    stroke, seeing the patients. Acting
  1192. 00:56:36
    extremely quickly, just like we do for
  1193. 00:56:38
    a scheme of stroke patients and acting
  1194. 00:56:40
    as fast as we can to activate the
  1195. 00:56:43
    surgery teams, get surgery involved
  1196. 00:56:45
    right away, and really understanding
  1197. 00:56:48
    that the time matters here and not just
  1198. 00:56:50
    walking away from the ICH, but staying
  1199. 00:56:54
    on top of the patient and getting them
  1200. 00:56:56
    where they need to go. It'll be
  1201. 00:56:58
    interesting to see how this,
  1202. 00:57:01
    it's nice that community hospitals were
  1203. 00:57:03
    doing as well in this trial, but in
  1204. 00:57:06
    the real world, how that's gonna happen
  1205. 00:57:08
    and are we gonna be having a lot more
  1206. 00:57:09
    transfers now? Are there gonna be ICH
  1207. 00:57:11
    hubs where we have specialists who are
  1208. 00:57:13
    gonna be doing these procedures? And
  1209. 00:57:17
    so it'll be interesting to see how that
  1210. 00:57:19
    plays out in the real world. And I
  1211. 00:57:22
    think,
  1212. 00:57:26
    sorry, was that the last part of your
  1213. 00:57:27
    question?
  1214. 00:57:29
    I don't really remember. Resurization,
  1215. 00:57:31
    workflow.
  1216. 00:57:34
    Yeah, yeah, and so obviously the care
  1217. 00:57:37
    for these patients is hopefully in the
  1218. 00:57:39
    ICU gonna be
  1219. 00:57:42
    less really, having them stay in the
  1220. 00:57:44
    ICU less, having less Venn days,
  1221. 00:57:46
    having less complications that are gonna
  1222. 00:57:47
    arise and seeing that hopefully that's
  1223. 00:57:50
    gonna translate to less costs, less
  1224. 00:57:55
    financial burden on the families of
  1225. 00:57:56
    these patients. And so hopefully we can
  1226. 00:57:59
    get this treatment for all patients,
  1227. 00:58:01
    whether that be in a transfer system or
  1228. 00:58:03
    just having these patients out in the
  1229. 00:58:04
    community and having the surgeons out in
  1230. 00:58:06
    the community be experts at this
  1231. 00:58:07
    procedure.
  1232. 00:58:10
    I also just wanna say, you know, I
  1233. 00:58:11
    think this was a great use of the
  1234. 00:58:12
    patient reported outcomes. I think
  1235. 00:58:14
    that's gonna really change the way we do
  1236. 00:58:16
    trials. Hopefully I think that we know
  1237. 00:58:18
    a modified rank and scale is maybe not
  1238. 00:58:19
    the best scale. And so really thinking
  1239. 00:58:21
    of how we can give the patients a voice
  1240. 00:58:23
    in this trial, in all trials and in all
  1241. 00:58:25
    of our outcomes and understanding that
  1242. 00:58:28
    what really matters is how the patient
  1243. 00:58:30
    feels after the procedure if maybe their
  1244. 00:58:33
    family feels and the burden on the
  1245. 00:58:35
    families that are taking care of these
  1246. 00:58:36
    patients and looking at outcomes that
  1247. 00:58:39
    really matter to the patient and their
  1248. 00:58:40
    families, I think is a really important
  1249. 00:58:42
    way to really start from the get-go and
  1250. 00:58:44
    designing trials that are using those
  1251. 00:58:46
    kinds of outcomes. So thank you so much
  1252. 00:58:48
    for using this user-weighted MRS scale
  1253. 00:58:51
    and bringing this to hopefully more
  1254. 00:58:54
    trials in the future. This was amazing
  1255. 00:58:56
    and it's really incredible work. And I
  1256. 00:59:00
    think like it is underestimated how much
  1257. 00:59:02
    you guys are changing the field. So
  1258. 00:59:06
    thank you very much for taking the time.
  1259. 00:59:09
    Great to meet you guys. And hopefully
  1260. 00:59:12
    we'll see each other at NCS and breath.
  1261. 00:59:18
    I hope to see you in which pretty soon.
  1262. 00:59:21
    Yeah, I'll see you there. And when you
  1263. 00:59:23
    guys do your next study on basal ganglia
  1264. 00:59:27
    and then another study on thalamus,
  1265. 00:59:29
    I'll invite you back for. more Hot
  1266. 00:59:32
    Topics podcasts. I'm
  1267. 00:59:34
    hoping Chris is going to take that.
  1268. 00:59:40
    All right. Thank you everyone so much.
  1269. 00:59:43
    Thanks, everybody. Thanks.