Deep Brain Stimulation for Epilepsy Explained: Targets, Technology & Future Treatments

1
00:00:08,508 --> 00:00:10,543
Hi, welcome to another episode of STEIN
2
00:00:10,543 --> 00:00:12,479
Fireside Podcast. Today
3
00:00:12,512 --> 00:00:13,380
Kai's going to be giving us
4
00:00:13,380 --> 00:00:14,848
a lecture on brain stimulation for
5
00:00:14,848 --> 00:00:16,683
epilepsy. He'll go over the
6
00:00:16,683 --> 00:00:18,585
history and now the variety
7
00:00:18,585 --> 00:00:21,221
of devices and approaches used to help
8
00:00:21,221 --> 00:00:22,355
patients with that disorder.
9
00:00:22,856 --> 00:00:24,057
So today I'm going to tell you about
10
00:00:24,057 --> 00:00:24,624
things from the
11
00:00:24,624 --> 00:00:26,426
perspective of a neurosurgeon. A
12
00:00:26,426 --> 00:00:28,361
lot of, just as with movement disorders,
13
00:00:28,361 --> 00:00:30,096
a lot of the most
14
00:00:30,096 --> 00:00:31,564
important aspects of the therapy,
15
00:00:31,564 --> 00:00:33,833
if not most of the most important aspects
16
00:00:33,833 --> 00:00:36,703
of the therapy, come from the programming
17
00:00:36,703 --> 00:00:39,172
that my neurology colleagues do with the
18
00:00:39,172 --> 00:00:40,039
devices once we've
19
00:00:40,039 --> 00:00:41,174
implanted them. And we'll
20
00:00:41,174 --> 00:00:42,909
address that in a separate lecture. But
21
00:00:42,909 --> 00:00:44,477
from today, I'll talk to you
22
00:00:44,477 --> 00:00:46,212
about deep brain stimulation
23
00:00:46,212 --> 00:00:47,113
for epilepsy from the
24
00:00:47,113 --> 00:00:48,281
neurosurgeon's perspective.
25
00:00:48,848 --> 00:00:50,683
I don't have any personal financial
26
00:00:50,683 --> 00:00:51,785
disclosures related to
27
00:00:51,785 --> 00:00:53,653
this talk, but it is important
28
00:00:53,653 --> 00:00:54,921
to know that many of the things that I'll
29
00:00:54,921 --> 00:00:55,889
be talking about are
30
00:00:55,889 --> 00:00:57,757
off-label uses of stimulation
31
00:00:57,757 --> 00:01:00,093
devices made for movement disorders
32
00:01:00,460 --> 00:01:01,995
applied to the treatment of epilepsy.
33
00:01:06,099 --> 00:01:08,601
So as we begin, I think there are a few
34
00:01:08,601 --> 00:01:09,936
practical aspects of the
35
00:01:09,936 --> 00:01:11,371
deep brain stimulation implant
36
00:01:11,438 --> 00:01:13,273
that are important to discuss at the
37
00:01:13,273 --> 00:01:15,975
beginning. When I do these
38
00:01:15,975 --> 00:01:17,277
surgeries, I bring the patient
39
00:01:17,277 --> 00:01:20,346
in and I typically do them with a Lexel
40
00:01:20,346 --> 00:01:21,247
frame. There are many
41
00:01:21,247 --> 00:01:22,248
good ways of doing it, but
42
00:01:22,248 --> 00:01:24,584
this is how I do it. I do think that
43
00:01:24,584 --> 00:01:26,386
using a CTV Enogram as
44
00:01:26,386 --> 00:01:27,620
your registration scan can
45
00:01:27,620 --> 00:01:30,423
be very important because that way your
46
00:01:30,423 --> 00:01:32,525
fiducial scan has the
47
00:01:32,559 --> 00:01:34,561
blood vessels on it. And the
48
00:01:34,561 --> 00:01:36,262
first most important aspect in any
49
00:01:36,262 --> 00:01:38,031
surgery is to be safe. The
50
00:01:38,031 --> 00:01:39,699
second is to be efficacious.
51
00:01:39,866 --> 00:01:42,402
To ensure safety, CTV Enogram on your
52
00:01:42,402 --> 00:01:43,636
registration scan helps
53
00:01:43,636 --> 00:01:46,439
you. Efficacy really comes from
54
00:01:46,473 --> 00:01:48,174
appropriate targeting and that comes from
55
00:01:48,174 --> 00:01:50,376
the MRI. Obviously,
56
00:01:51,377 --> 00:01:53,413
you have to merge your
57
00:01:53,446 --> 00:01:56,115
scan, which in my case again is the CTV
58
00:01:56,115 --> 00:01:58,017
Enogram with your MRI.
59
00:01:58,084 --> 00:01:59,786
I'll be showing you white
60
00:01:59,786 --> 00:02:02,055
matter null sequences like F-Gator, T2
61
00:02:02,055 --> 00:02:04,057
sequences, and T1 post-contrast
62
00:02:04,057 --> 00:02:05,525
sequences. I think it's
63
00:02:05,525 --> 00:02:07,360
important to take the time to do the
64
00:02:07,360 --> 00:02:08,561
merge between all of these
65
00:02:08,561 --> 00:02:09,996
different imaging modalities
66
00:02:10,363 --> 00:02:12,999
carefully and appropriately. If you're
67
00:02:12,999 --> 00:02:13,633
doing an asleep
68
00:02:13,633 --> 00:02:16,135
surgery, as we do for epilepsy,
69
00:02:16,503 --> 00:02:18,171
I think it's important to do volumetric
70
00:02:18,171 --> 00:02:20,406
imaging. So before you close your
71
00:02:20,406 --> 00:02:22,542
incisions and potentially
72
00:02:22,542 --> 00:02:24,277
after you place each lead, I think it's
73
00:02:24,277 --> 00:02:25,778
important to do this. If
74
00:02:25,778 --> 00:02:27,580
you wait until after you're
75
00:02:27,580 --> 00:02:30,150
completely closed, you close the OR and
76
00:02:30,150 --> 00:02:32,752
then get imaging to verify
77
00:02:32,752 --> 00:02:34,721
your accuracy, you run into the
78
00:02:34,754 --> 00:02:36,322
problem that you're only going to revise
79
00:02:36,322 --> 00:02:37,490
the most egregious of
80
00:02:37,490 --> 00:02:38,558
leads. And you want to keep the
81
00:02:38,558 --> 00:02:40,860
barrier very low to improve your
82
00:02:40,860 --> 00:02:43,963
placement. If you're using an OR for
83
00:02:43,963 --> 00:02:45,098
intraoperative imaging,
84
00:02:45,298 --> 00:02:47,200
because it gives good skull resolution
85
00:02:47,200 --> 00:02:48,168
and good resolution of your
86
00:02:48,168 --> 00:02:49,269
hardware, it can tell you where
87
00:02:49,269 --> 00:02:51,604
your hardware is, but I don't believe in
88
00:02:51,604 --> 00:02:53,206
fusing it to the MRI. The
89
00:02:53,206 --> 00:02:55,208
reason for that is that your MRI
90
00:02:55,208 --> 00:02:57,243
sequences don't have the best skull
91
00:02:57,243 --> 00:02:59,746
resolution. And I do think
92
00:02:59,746 --> 00:03:02,649
it's helpful to merge brain to
93
00:03:02,649 --> 00:03:06,452
brain rather than skull on an MRI to
94
00:03:06,452 --> 00:03:08,755
skull on a CT scan. So using
95
00:03:08,755 --> 00:03:10,490
a pre-op CT, which has tissue
96
00:03:10,490 --> 00:03:14,127
resolution, can bridge the MRI to the
97
00:03:14,127 --> 00:03:17,397
intraoperative ORMs. For
98
00:03:17,397 --> 00:03:19,232
taking care of incisions, I think
99
00:03:19,265 --> 00:03:21,200
it's important to have a protocol. I'm
100
00:03:21,200 --> 00:03:21,901
going to show you what my
101
00:03:21,901 --> 00:03:24,504
protocol is. I don't, I wouldn't
102
00:03:24,504 --> 00:03:26,472
propose that that's the only or the best
103
00:03:26,673 --> 00:03:27,874
protocol, but I do think
104
00:03:27,874 --> 00:03:29,275
that whatever your protocol is,
105
00:03:29,609 --> 00:03:31,144
you want to stick to it. That way you
106
00:03:31,144 --> 00:03:32,612
have a very regimented way of doing
107
00:03:32,612 --> 00:03:33,479
things. And if you start
108
00:03:33,479 --> 00:03:35,114
having problems, you can improve what
109
00:03:35,114 --> 00:03:38,084
you're doing and adjust accordingly. For
110
00:03:38,084 --> 00:03:39,385
prepping the incisions,
111
00:03:39,852 --> 00:03:42,922
either I do it or one surgical assistant
112
00:03:42,922 --> 00:03:43,923
in particular who I work
113
00:03:43,923 --> 00:03:45,058
with regularly does it.
114
00:03:45,491 --> 00:03:47,393
He does it the exact same way that I do.
115
00:03:48,094 --> 00:03:49,829
If he's not there, I do it myself. And
116
00:03:49,829 --> 00:03:50,697
that way, if we start
117
00:03:50,697 --> 00:03:52,532
having any problems with infection or
118
00:03:53,132 --> 00:03:54,400
wound integrity, we can go
119
00:03:54,400 --> 00:03:55,802
back and revise accordingly.
120
00:03:56,603 --> 00:03:58,137
I think it's important to preserve scalp
121
00:03:58,137 --> 00:03:59,472
vasculature and to
122
00:03:59,472 --> 00:04:00,974
preserve it whenever possible.
123
00:04:01,241 --> 00:04:03,610
Obviously, part of this is taking into
124
00:04:03,610 --> 00:04:04,310
account previous
125
00:04:04,310 --> 00:04:06,279
incisions that patients often have
126
00:04:06,312 --> 00:04:08,715
when you do epilepsy surgery, making sure
127
00:04:08,715 --> 00:04:09,916
you incorporate those
128
00:04:09,916 --> 00:04:11,284
and you don't leave any,
129
00:04:12,252 --> 00:04:14,087
what I would call an isthmus of tissue
130
00:04:14,454 --> 00:04:15,488
where there's really poor
131
00:04:15,488 --> 00:04:17,056
vascularization. I think
132
00:04:17,056 --> 00:04:19,292
it's important to really take into
133
00:04:19,292 --> 00:04:21,327
account previous incisions with your new
134
00:04:21,327 --> 00:04:22,629
incisions in a way that
135
00:04:23,096 --> 00:04:25,131
is optimized for preserving vasculature.
136
00:04:25,965 --> 00:04:26,866
I countersink and I'll
137
00:04:26,866 --> 00:04:27,634
show you what I mean by
138
00:04:27,634 --> 00:04:28,901
that in a moment in order to keep the
139
00:04:28,901 --> 00:04:30,069
hardware profile low,
140
00:04:30,603 --> 00:04:32,071
which reduces the erosion risk,
141
00:04:32,071 --> 00:04:33,973
which in turn reduces the infection risk.
142
00:04:35,141 --> 00:04:35,742
I try to keep my
143
00:04:35,742 --> 00:04:37,343
incisions off of the hardware when
144
00:04:37,377 --> 00:04:40,413
possible. I use vancomycin powder in all
145
00:04:40,413 --> 00:04:42,281
of my incisions. I know
146
00:04:42,281 --> 00:04:43,283
that people have different
147
00:04:43,283 --> 00:04:44,817
opinions about this. For me, I feel like
148
00:04:44,817 --> 00:04:47,020
this can really help reduce infection
149
00:04:47,020 --> 00:04:48,688
risk. And then really
150
00:04:48,688 --> 00:04:50,657
take the time to do a solid closure. So
151
00:04:50,657 --> 00:04:51,758
this means, you know, five
152
00:04:51,758 --> 00:04:53,026
to seven millimeters between
153
00:04:53,026 --> 00:04:55,928
galial sutures and using a running
154
00:04:55,928 --> 00:04:58,431
surface suture. For countersinking,
155
00:04:58,931 --> 00:04:59,932
that's making sure that your
156
00:04:59,932 --> 00:05:02,935
hardware, as you see here, sits at the
157
00:05:03,202 --> 00:05:05,271
level of the outer gall
158
00:05:05,271 --> 00:05:07,707
cortex. And so what you do is, is
159
00:05:07,774 --> 00:05:09,676
the hardware of your anchoring device is
160
00:05:09,976 --> 00:05:11,911
set down into the bone by drilling out a
161
00:05:11,911 --> 00:05:13,146
countersunk profile.
162
00:05:13,413 --> 00:05:15,782
This slide just shows a tool that we
163
00:05:15,782 --> 00:05:16,949
developed to use for this.
164
00:05:16,949 --> 00:05:18,017
This tool is now commercially
165
00:05:18,017 --> 00:05:20,687
available, but it acts as a stencil to
166
00:05:20,687 --> 00:05:22,255
identify the optimal shape for that
167
00:05:22,255 --> 00:05:23,256
countersinking profile.
168
00:05:24,090 --> 00:05:26,326
I countersink the extensions as well, as
169
00:05:26,326 --> 00:05:27,827
you see here. That does
170
00:05:27,827 --> 00:05:29,462
require a little bit larger
171
00:05:29,696 --> 00:05:32,031
on the posterior aspect of the scalp. But
172
00:05:32,131 --> 00:05:33,366
I picked this up actually
173
00:05:33,366 --> 00:05:35,134
from talking to Kelly Foote,
174
00:05:35,802 --> 00:05:37,904
Wyell, Assad, and Ziv Williams. And I
175
00:05:37,904 --> 00:05:40,440
think that this really is nice for
176
00:05:40,473 --> 00:05:42,075
avoiding erosion related
177
00:05:42,075 --> 00:05:44,077
to things like hats. Patients like it a
178
00:05:44,077 --> 00:05:45,511
lot better, but it does require a
179
00:05:45,511 --> 00:05:46,713
slightly larger incision.
180
00:05:47,547 --> 00:05:49,682
This shows where my hardware sits with
181
00:05:49,682 --> 00:05:51,617
respect to my incisions.
182
00:05:52,085 --> 00:05:53,953
Obviously, these are frontal
183
00:05:53,953 --> 00:05:56,189
approach incisions. The incisions that I
184
00:05:56,189 --> 00:05:57,256
use for epilepsy are
185
00:05:57,256 --> 00:05:59,192
typically posterior, and I'll show
186
00:05:59,225 --> 00:06:01,060
those. But for some surgeons,
187
00:06:01,227 --> 00:06:03,329
particularly those who use anterior
188
00:06:03,329 --> 00:06:04,597
approaches to the central median
189
00:06:04,597 --> 00:06:07,333
nucleus and anterior transventricular
190
00:06:07,333 --> 00:06:09,202
approaches to the anterior nucleus of the
191
00:06:09,202 --> 00:06:10,203
thalamus, your incisions
192
00:06:10,203 --> 00:06:12,338
will sit somewhere around here. And the
193
00:06:12,338 --> 00:06:13,606
reason why that's been done
194
00:06:13,606 --> 00:06:14,807
traditionally is because this
195
00:06:14,807 --> 00:06:16,709
is where the coronal suture is, which has
196
00:06:16,709 --> 00:06:19,679
been a sort of standard landmark for
197
00:06:19,679 --> 00:06:21,581
placement of insertion
198
00:06:21,681 --> 00:06:25,184
sites coming down through the superior
199
00:06:25,351 --> 00:06:27,086
middle frontal gyri. And
200
00:06:27,086 --> 00:06:28,988
then I have it so, if possible,
201
00:06:29,288 --> 00:06:31,724
wiring extends away from the incisions.
202
00:06:32,091 --> 00:06:32,959
And then as you come
203
00:06:32,959 --> 00:06:34,794
posteriorly down to where you make your
204
00:06:34,794 --> 00:06:36,362
skipped incision for the extensions, you
205
00:06:36,362 --> 00:06:37,330
try to have your wiring
206
00:06:37,730 --> 00:06:39,232
anterior to that skipped incision,
207
00:06:39,232 --> 00:06:40,767
and then you pull things forward as
208
00:06:40,767 --> 00:06:41,300
you're making that
209
00:06:41,300 --> 00:06:43,803
countersunk profile for the extensions.
210
00:06:44,203 --> 00:06:46,472
This is for responsive neurostimulation,
211
00:06:46,472 --> 00:06:47,974
just showing how I
212
00:06:48,007 --> 00:06:49,742
countersink the neuropaste device.
213
00:06:50,042 --> 00:06:51,677
I should note that the way that I do all
214
00:06:51,677 --> 00:06:53,980
of this, where I keep it away
215
00:06:53,980 --> 00:06:55,281
from the incision margin and
216
00:06:55,314 --> 00:06:57,216
I countersink it down really comes from
217
00:06:57,483 --> 00:06:58,618
discussions I've had with my
218
00:06:58,618 --> 00:06:59,752
colleagues in the field around
219
00:06:59,752 --> 00:07:03,122
the country. For my chest incisions, I
220
00:07:03,256 --> 00:07:04,323
try to put my pulse
221
00:07:04,323 --> 00:07:05,792
generators when possible,
222
00:07:06,092 --> 00:07:08,528
sub-fascial-y on top of the pectoralis
223
00:07:08,528 --> 00:07:10,696
muscle. This isn't possible
224
00:07:10,696 --> 00:07:12,632
if there's large body habitus
225
00:07:12,932 --> 00:07:14,534
or sometimes with female patients who
226
00:07:14,534 --> 00:07:15,668
have more breast tissue, this is
227
00:07:15,668 --> 00:07:16,836
difficult. But if you're
228
00:07:16,836 --> 00:07:19,005
able to get your hardware down under the
229
00:07:19,005 --> 00:07:20,807
fascial layer, this is very
230
00:07:20,807 --> 00:07:22,074
nice because you can then run
231
00:07:22,108 --> 00:07:25,278
fascial closed and that reduces the
232
00:07:25,311 --> 00:07:26,312
likelihood of the pulse
233
00:07:26,312 --> 00:07:27,713
generator coming up and pressing
234
00:07:27,713 --> 00:07:29,549
against the dermis. And then you can
235
00:07:29,549 --> 00:07:32,185
close your dermis independently with
236
00:07:32,185 --> 00:07:33,052
inverted interrupted
237
00:07:33,052 --> 00:07:34,253
sutures followed by a running
238
00:07:34,253 --> 00:07:36,189
subcuticular. I typically use
239
00:07:36,189 --> 00:07:37,957
derma bond in the chest above,
240
00:07:38,357 --> 00:07:39,992
and then I use a silver impregnated
241
00:07:39,992 --> 00:07:41,227
water-resistant bandage
242
00:07:41,327 --> 00:07:42,762
that I keep on for five days.
243
00:07:43,396 --> 00:07:47,066
So getting to epilepsy decision-making,
244
00:07:47,333 --> 00:07:48,100
it's really important to
245
00:07:48,100 --> 00:07:50,336
consider the role of stereo-EEG,
246
00:07:50,470 --> 00:07:52,038
which traditionally, of course, has been
247
00:07:52,038 --> 00:07:53,372
largely for identifying
248
00:07:53,739 --> 00:07:56,042
ablative or resective targets going for
249
00:07:56,042 --> 00:07:57,543
surgical cure. But
250
00:07:57,543 --> 00:07:59,345
increasingly, I incorporate
251
00:07:59,545 --> 00:08:01,948
potential subsequent deep brain
252
00:08:01,948 --> 00:08:03,449
stimulation targets into the
253
00:08:03,449 --> 00:08:05,751
stereo-EEG plan. I think it's
254
00:08:05,751 --> 00:08:08,354
important to do that because, as you see
255
00:08:08,354 --> 00:08:09,388
here, we may start out with
256
00:08:09,388 --> 00:08:10,790
an epilepsy consultation and
257
00:08:10,790 --> 00:08:13,492
then go to a surgical resection or laser
258
00:08:13,492 --> 00:08:15,361
ablation. And many of
259
00:08:15,361 --> 00:08:17,096
those patients for whom we do a
260
00:08:17,096 --> 00:08:19,599
definitive resective or ablative surgery,
261
00:08:20,132 --> 00:08:21,267
they may continue to have
262
00:08:21,267 --> 00:08:23,669
seizures. And I think that
263
00:08:23,669 --> 00:08:25,071
while there has been some debate, and
264
00:08:25,071 --> 00:08:26,105
I'll get back to this later
265
00:08:26,105 --> 00:08:28,107
in this lecture, about the
266
00:08:28,140 --> 00:08:32,378
utility of stereo-EEG in stimulation
267
00:08:32,378 --> 00:08:34,046
decision-making and
268
00:08:34,046 --> 00:08:35,948
incorporation of thalamic stereo-EEG
269
00:08:35,948 --> 00:08:37,450
targets, I think it's really important to
270
00:08:37,450 --> 00:08:39,452
consider the fact that
271
00:08:39,519 --> 00:08:41,754
having a plan as a backup is very
272
00:08:41,754 --> 00:08:43,956
important and knowing what your strategy
273
00:08:43,956 --> 00:08:44,790
is going to be if
274
00:08:44,790 --> 00:08:45,992
those resective or ablative
275
00:08:46,492 --> 00:08:49,228
interventions are ultimately unsuccessful
276
00:08:49,228 --> 00:08:50,329
is important up front.
277
00:08:50,329 --> 00:08:51,063
And you have a unique
278
00:08:51,063 --> 00:08:53,399
opportunity while you're doing stereo-EEG
279
00:08:53,733 --> 00:08:55,067
to test out types of
280
00:08:55,067 --> 00:08:56,335
stimulation that you'll do,
281
00:08:56,602 --> 00:08:58,437
look at patterns of seizure propagation,
282
00:08:58,471 --> 00:08:59,472
and have a backup plan
283
00:08:59,472 --> 00:09:01,507
that DBS will provide if
284
00:09:01,807 --> 00:09:03,609
patients continue to have seizures after
285
00:09:03,609 --> 00:09:04,543
one of those resective
286
00:09:04,610 --> 00:09:05,711
or ablative procedures.
287
00:09:06,078 --> 00:09:08,314
This is from a paper that a neurologist
288
00:09:08,314 --> 00:09:08,915
that I work with,
289
00:09:08,915 --> 00:09:10,383
Anthony Fine, and I made about
290
00:09:10,383 --> 00:09:12,318
decision-making in stereotactic epilepsy
291
00:09:12,318 --> 00:09:14,420
surgery. And what you'll see
292
00:09:14,453 --> 00:09:15,821
here is a roadmap that takes
293
00:09:15,955 --> 00:09:18,491
the initial epilepsy consultation down to
294
00:09:18,824 --> 00:09:19,892
a number of different
295
00:09:20,192 --> 00:09:21,561
interventions that we do.
296
00:09:21,894 --> 00:09:24,230
And it's quite a complicated roadmap, as
297
00:09:24,230 --> 00:09:25,932
you see here, but I think
298
00:09:25,932 --> 00:09:27,600
it can be helpful to organize
299
00:09:27,600 --> 00:09:29,936
your thoughts about how you come to a
300
00:09:29,936 --> 00:09:31,370
definitive therapy from a
301
00:09:31,370 --> 00:09:33,673
patient coming typically through
302
00:09:33,673 --> 00:09:35,074
one of the neurologists that you work
303
00:09:35,074 --> 00:09:36,709
with if you're a surgeon or as a
304
00:09:36,709 --> 00:09:37,376
neurologist that will
305
00:09:37,376 --> 00:09:40,479
come to you at an academic center from a
306
00:09:40,746 --> 00:09:42,949
community setting. You start out by
307
00:09:42,949 --> 00:09:45,751
looking at the imaging
308
00:09:45,785 --> 00:09:48,421
that you have, the electrophysiology that
309
00:09:48,421 --> 00:09:49,322
you have, and the
310
00:09:49,322 --> 00:09:50,590
patient's seizure symptoms,
311
00:09:50,890 --> 00:09:52,959
and then having a conversation between
312
00:09:53,225 --> 00:09:54,026
neurology,
313
00:09:54,060 --> 00:09:56,395
neuroradiology, neuropsychiatry,
314
00:09:56,862 --> 00:09:59,031
neuropsychology, and neurosurgery in a
315
00:09:59,031 --> 00:09:59,899
multidisciplinary
316
00:09:59,899 --> 00:10:01,467
conference where all the
317
00:10:01,467 --> 00:10:02,902
different pieces of evidence are
318
00:10:02,902 --> 00:10:05,738
presented and a discussion is had. And
319
00:10:05,738 --> 00:10:06,772
for these, we look for
320
00:10:06,772 --> 00:10:08,174
concordance between the seizure
321
00:10:08,174 --> 00:10:10,109
semiology, which is the symptoms that
322
00:10:10,109 --> 00:10:11,010
patients have before,
323
00:10:11,043 --> 00:10:13,613
during, and after a seizure. The imaging
324
00:10:13,646 --> 00:10:16,248
that we get on MRI, PET
325
00:10:16,248 --> 00:10:19,485
scans, SPECT scans, and MEG,
326
00:10:19,518 --> 00:10:21,754
although MEG sort of flanks imaging and
327
00:10:21,754 --> 00:10:23,122
electrophysiology, I think.
328
00:10:23,689 --> 00:10:25,291
And for electrophysiology,
329
00:10:25,358 --> 00:10:27,760
we look at changes from the scalp from
330
00:10:27,760 --> 00:10:29,362
EEG and ask, "Do those
331
00:10:29,362 --> 00:10:30,730
agree with the imaging
332
00:10:30,896 --> 00:10:33,799
and the semiology that we have?" Okay,
333
00:10:33,799 --> 00:10:35,634
and it's the synthesis of that
334
00:10:35,634 --> 00:10:37,036
information that happens
335
00:10:37,436 --> 00:10:39,205
at our multidisciplinary committee. And
336
00:10:39,205 --> 00:10:39,872
there will be special
337
00:10:39,872 --> 00:10:41,340
cases of this. For example,
338
00:10:41,607 --> 00:10:44,210
if you have bihemispheric distributed EEG
339
00:10:44,210 --> 00:10:45,811
seizure architecture with
340
00:10:45,811 --> 00:10:47,013
semiologies that include
341
00:10:47,013 --> 00:10:48,347
frequent drop attacks, you're thinking
342
00:10:48,347 --> 00:10:50,282
corpus callusotomy. If you have
343
00:10:50,282 --> 00:10:51,584
hemispheric pathology
344
00:10:51,884 --> 00:10:53,953
that we may see for perinatal stroke,
345
00:10:54,520 --> 00:10:57,323
hemimaganscephaly, rasmuschines disease,
346
00:10:57,423 --> 00:10:59,892
and associated seizures, pars continua,
347
00:10:59,959 --> 00:11:01,327
you'll be thinking hemispherectomy.
348
00:11:01,661 --> 00:11:04,663
If you identify a concordant lesion that
349
00:11:04,930 --> 00:11:06,966
has electrophysiology
350
00:11:07,133 --> 00:11:08,601
from EEG showing the seizures
351
00:11:08,601 --> 00:11:10,970
originate from that area, and sometimes
352
00:11:11,170 --> 00:11:12,538
semiology that points to the
353
00:11:12,538 --> 00:11:14,240
same area, you may go directly
354
00:11:14,240 --> 00:11:17,276
to stimulation, or you may go directly to
355
00:11:17,276 --> 00:11:19,378
surgical resection, sometimes laser
356
00:11:19,378 --> 00:11:20,546
ablation in that setting.
357
00:11:21,247 --> 00:11:24,016
But for the remainder of patients where
358
00:11:24,183 --> 00:11:25,384
either you don't have
359
00:11:25,384 --> 00:11:26,852
full concordance or you have
360
00:11:26,852 --> 00:11:29,021
concordance but the extent is unknown,
361
00:11:29,021 --> 00:11:29,955
you're going to be looking at
362
00:11:29,955 --> 00:11:31,090
intracranial monitoring.
363
00:11:31,624 --> 00:11:33,192
Either with stereo EEG or
364
00:11:33,192 --> 00:11:35,995
electrocorticography. And from there,
365
00:11:36,262 --> 00:11:37,930
that will lead you either to
366
00:11:37,930 --> 00:11:40,633
a more specialized resection or ablation
367
00:11:40,633 --> 00:11:42,301
strategy or to
368
00:11:42,301 --> 00:11:44,870
neurostimulation. Again, these are all from
369
00:11:44,870 --> 00:11:47,073
this paper about epilepsy surgery that
370
00:11:47,073 --> 00:11:48,007
Dr. Fine and I put
371
00:11:48,007 --> 00:11:51,210
together. And when we look at how we
372
00:11:51,210 --> 00:11:53,512
treat with stimulation, and we'll come
373
00:11:53,512 --> 00:11:55,081
back to that stereo EEG in
374
00:11:55,081 --> 00:11:57,416
a moment, I think that it's
375
00:11:57,416 --> 00:11:58,918
a very important language between the
376
00:11:58,951 --> 00:12:01,087
zone of seizure onset,
377
00:12:01,220 --> 00:12:02,755
the zone of seizure spread,
378
00:12:02,955 --> 00:12:05,424
or both in tandem. And when we think
379
00:12:05,424 --> 00:12:08,360
about seizure onset, we're talking about
380
00:12:08,360 --> 00:12:09,595
the focus of the seizure.
381
00:12:10,162 --> 00:12:11,697
Okay, and typically these are cortical
382
00:12:11,697 --> 00:12:12,898
sites. They may be
383
00:12:12,898 --> 00:12:14,734
neocortex, they may be limbic,
384
00:12:15,034 --> 00:12:17,703
they may be hippocampal. And here we
385
00:12:17,703 --> 00:12:19,138
really are trying to
386
00:12:19,138 --> 00:12:21,273
detect where these are from
387
00:12:21,373 --> 00:12:23,375
often stereo EEG or maybe there's an
388
00:12:23,409 --> 00:12:24,677
eloquent lesion and that may
389
00:12:24,710 --> 00:12:26,145
point us towards stimulating
390
00:12:26,445 --> 00:12:28,914
cortex. When we talk about stimulating
391
00:12:28,914 --> 00:12:31,217
the circuit or circuit stimulation,
392
00:12:31,484 --> 00:12:32,852
usually for that we mean
393
00:12:32,918 --> 00:12:35,754
stimulating nuclei and the thalamus. And
394
00:12:35,754 --> 00:12:36,755
increasingly, we'll
395
00:12:36,755 --> 00:12:38,557
consider tandem stimulation
396
00:12:38,557 --> 00:12:40,860
of multiple of these structures together.
397
00:12:41,060 --> 00:12:43,195
Stereo EEG, as we discussed,
398
00:12:43,195 --> 00:12:45,865
involves placement of depth
399
00:12:46,232 --> 00:12:49,602
throughout the cortex and subcortical
400
00:12:49,602 --> 00:12:51,070
structures using the
401
00:12:51,070 --> 00:12:52,404
stereotactically placed bolts in the
402
00:12:52,404 --> 00:12:54,907
skull through which multicontact leads
403
00:12:54,907 --> 00:12:57,109
are placed. And when we target the
404
00:12:57,109 --> 00:12:58,744
thalamus, as we often do,
405
00:12:58,778 --> 00:13:00,813
in order to help us know whether or not
406
00:13:00,813 --> 00:13:01,747
subsequent deep brain
407
00:13:01,747 --> 00:13:03,516
stimulation may help with patient
408
00:13:03,516 --> 00:13:05,384
seizures, there's a number of different
409
00:13:05,384 --> 00:13:06,452
things that we consider.
410
00:13:06,719 --> 00:13:08,120
The first of these is seizure
411
00:13:08,120 --> 00:13:10,990
timing. So recording from the thalamus
412
00:13:11,023 --> 00:13:12,091
can help you delineate
413
00:13:12,091 --> 00:13:13,592
between seizures that may start,
414
00:13:13,926 --> 00:13:16,228
let's say, from the anterior cingulate or
415
00:13:16,228 --> 00:13:17,930
from the thalamic
416
00:13:17,963 --> 00:13:19,465
structures. And if you see
417
00:13:19,832 --> 00:13:22,301
that your seizures are detected in the
418
00:13:22,301 --> 00:13:23,836
thalamus before they're
419
00:13:23,869 --> 00:13:25,905
detected in any neocortical site,
420
00:13:26,138 --> 00:13:27,973
I think our first assumption has to be
421
00:13:27,973 --> 00:13:30,409
that in some way our more
422
00:13:30,409 --> 00:13:32,111
peripheral measurements that
423
00:13:32,111 --> 00:13:34,013
we're making with stereo EEG, meaning our
424
00:13:34,013 --> 00:13:35,714
cortical sites, are not in
425
00:13:35,714 --> 00:13:37,349
the exact zone of seizure onset.
426
00:13:37,349 --> 00:13:38,817
And so you see propagation through the
427
00:13:38,817 --> 00:13:40,653
thalamus before you see subsequent
428
00:13:41,086 --> 00:13:42,521
cortical onset sites.
429
00:13:43,189 --> 00:13:45,357
And we can also determine laterality if
430
00:13:45,357 --> 00:13:47,193
we have bilateral stereo
431
00:13:47,193 --> 00:13:49,562
EEG thalamic inputs. And so
432
00:13:49,628 --> 00:13:51,430
thalamic stereo EEG can really help you
433
00:13:51,430 --> 00:13:53,432
understand seizure timing and whether
434
00:13:53,632 --> 00:13:54,900
seizures start before
435
00:13:54,900 --> 00:13:56,101
the neocortical sites that you've
436
00:13:56,101 --> 00:13:57,903
measured and whether they start in one
437
00:13:57,903 --> 00:13:58,837
hemisphere versus the
438
00:13:58,837 --> 00:14:01,340
other hemisphere. We'll often do test
439
00:14:01,340 --> 00:14:05,611
stimulation through stereo EEG for 24 to
440
00:14:05,611 --> 00:14:07,379
48 hours once we think
441
00:14:07,379 --> 00:14:08,981
we've determined where the seizure onset
442
00:14:08,981 --> 00:14:10,683
is and where the seizure
443
00:14:10,683 --> 00:14:12,084
spread through the thalamus is.
444
00:14:12,451 --> 00:14:14,386
And really with that, we test a number of
445
00:14:14,386 --> 00:14:15,588
different programming
446
00:14:15,588 --> 00:14:17,523
parameters that will mimic what we
447
00:14:17,523 --> 00:14:18,791
would do with the permanent deep brain
448
00:14:18,791 --> 00:14:20,326
stimulation electrode. One of my
449
00:14:20,326 --> 00:14:22,294
colleagues, Dr. Nick Greg,
450
00:14:22,561 --> 00:14:24,496
does this primarily. He does it in tandem
451
00:14:24,496 --> 00:14:26,098
with Dr. Brian Lundstrom,
452
00:14:26,398 --> 00:14:28,534
Dr. Greg Worrell, Dr. Gamal
453
00:14:28,534 --> 00:14:31,503
Osman, Dr. Keith Starnes. So the patterns
454
00:14:31,503 --> 00:14:33,439
that are used can be one of a number of
455
00:14:33,439 --> 00:14:34,306
different types. One
456
00:14:34,306 --> 00:14:36,375
is responsive, which is the neural pace
457
00:14:36,375 --> 00:14:37,743
device. That's where you
458
00:14:37,743 --> 00:14:39,578
continuously record activity.
459
00:14:39,778 --> 00:14:43,148
You set an algorithm that identifies what
460
00:14:43,148 --> 00:14:44,216
you believe is an onset of
461
00:14:44,216 --> 00:14:45,317
seizure that will trigger
462
00:14:45,317 --> 00:14:48,454
stimulation. Another is high frequency
463
00:14:48,487 --> 00:14:50,856
stimulation. And usually this is 130
464
00:14:50,990 --> 00:14:52,291
hertz at our institution.
465
00:14:52,558 --> 00:14:54,793
And here philosophy of the stimulation is
466
00:14:54,793 --> 00:14:55,694
that you're trying to
467
00:14:55,694 --> 00:14:57,796
suppress a region. We'll also
468
00:14:57,796 --> 00:15:00,199
stimulate sometimes at 30 hertz or lower.
469
00:15:00,799 --> 00:15:02,434
And what we're doing in this case is
470
00:15:02,434 --> 00:15:03,502
trying to modulate the
471
00:15:03,502 --> 00:15:05,671
circuit, but not through blanket
472
00:15:05,671 --> 00:15:07,773
suppression, but actually by entraining
473
00:15:07,773 --> 00:15:09,575
the circuit on that 30
474
00:15:09,608 --> 00:15:14,013
hertz stimulation. So the classic SANTE
475
00:15:14,046 --> 00:15:15,247
protocol, which is the
476
00:15:15,247 --> 00:15:16,882
initial protocol that was made
477
00:15:17,216 --> 00:15:19,318
for stimulating the thalamus, in
478
00:15:19,318 --> 00:15:20,085
particular the
479
00:15:20,085 --> 00:15:21,553
anteronucleus of the thalamus,
480
00:15:22,021 --> 00:15:23,889
that's an intermittent high frequency
481
00:15:23,889 --> 00:15:26,959
stimulation at 145 hertz with 90
482
00:15:26,959 --> 00:15:28,260
microsecond pulse width,
483
00:15:28,427 --> 00:15:30,529
basically cycling on for one minute and
484
00:15:30,529 --> 00:15:31,797
five minutes off. This
485
00:15:31,797 --> 00:15:33,299
was pioneered by Dr. Bob
486
00:15:33,299 --> 00:15:35,901
Fisher at Stanford University and was
487
00:15:35,901 --> 00:15:36,669
done in tandem with
488
00:15:36,669 --> 00:15:40,172
Medtronic. So when we do this test
489
00:15:40,172 --> 00:15:42,274
therapeutic stimulation, we may be
490
00:15:42,274 --> 00:15:44,310
measuring from multiple sites at once and
491
00:15:44,310 --> 00:15:45,611
then testing stimulation
492
00:15:45,978 --> 00:15:47,713
at those different sites while measuring
493
00:15:48,113 --> 00:15:49,214
usually interictal
494
00:15:49,214 --> 00:15:50,482
epileptiform discharges,
495
00:15:50,783 --> 00:15:52,951
also called spikes, or actually measuring
496
00:15:53,185 --> 00:15:54,320
seizure frequency if the
497
00:15:54,320 --> 00:15:55,587
seizures are frequent enough.
498
00:15:56,288 --> 00:15:58,557
And so what we may do is measure a
499
00:15:58,557 --> 00:15:59,391
baseline rate of
500
00:15:59,391 --> 00:16:01,260
discharges, as you see here in gray,
501
00:16:02,027 --> 00:16:03,762
and then we'll test stimulating different
502
00:16:03,762 --> 00:16:05,831
targets. And so if we
503
00:16:05,831 --> 00:16:07,433
stimulate, in this example,
504
00:16:07,433 --> 00:16:09,835
an anteronucleus of thalamus, you see
505
00:16:09,835 --> 00:16:11,270
quite robust suppression
506
00:16:11,637 --> 00:16:13,372
of interictal discharge or
507
00:16:13,372 --> 00:16:15,574
spike discharge in the amygdala,
508
00:16:15,708 --> 00:16:17,276
hippocampus, and the hippocampal head,
509
00:16:17,276 --> 00:16:18,377
these are limbic structures,
510
00:16:18,610 --> 00:16:20,279
but less statistically significant
511
00:16:20,279 --> 00:16:22,581
suppression in the occipital lobe. If we
512
00:16:22,581 --> 00:16:23,949
instead stimulate in
513
00:16:23,949 --> 00:16:25,985
the pulvinar nucleus, we may see more
514
00:16:25,985 --> 00:16:27,753
profound suppression in the
515
00:16:27,753 --> 00:16:29,488
case for this patient in the
516
00:16:29,488 --> 00:16:31,090
occipital lobe, but without any
517
00:16:31,090 --> 00:16:34,660
improvement in the limbic circuitry. And
518
00:16:34,660 --> 00:16:36,495
so by stimulating in these
519
00:16:36,495 --> 00:16:38,497
two areas independently, we can figure
520
00:16:38,497 --> 00:16:40,699
out if we're able to modulate seizure
521
00:16:40,699 --> 00:16:42,468
activity in different
522
00:16:42,501 --> 00:16:44,103
potential neocortical targets, and if
523
00:16:44,103 --> 00:16:45,537
patients have multiple
524
00:16:45,571 --> 00:16:46,839
potential circuits involved,
525
00:16:46,839 --> 00:16:48,874
we may be identifying sites where we
526
00:16:48,874 --> 00:16:49,975
would place multiple
527
00:16:49,975 --> 00:16:53,012
different thalamic electrodes to treat
528
00:16:53,045 --> 00:16:55,748
circuits. I just want to reiterate that
529
00:16:55,814 --> 00:16:56,749
seizure freedom from
530
00:16:56,749 --> 00:16:58,517
surgery is rare outside of the
531
00:16:58,517 --> 00:17:00,686
lesional temporal lobe or focal cortical
532
00:17:00,686 --> 00:17:02,388
dysplages. And I think if we're really
533
00:17:02,421 --> 00:17:04,456
honest about this as a
534
00:17:04,456 --> 00:17:06,692
community, then candidate fallback
535
00:17:06,692 --> 00:17:08,327
stimulation should always be
536
00:17:08,427 --> 00:17:10,162
a component of our stereo EEG
537
00:17:10,195 --> 00:17:13,198
construct, and test stimulation should be
538
00:17:13,198 --> 00:17:15,534
performed if we have the time to do this
539
00:17:15,634 --> 00:17:17,036
during the inpatient
540
00:17:17,069 --> 00:17:19,772
stereo EEG monitoring. I think this
541
00:17:19,938 --> 00:17:22,374
requires very strong relationships
542
00:17:22,374 --> 00:17:24,176
between the neurosurgeon,
543
00:17:24,543 --> 00:17:26,211
the primary neurologist involved with
544
00:17:26,211 --> 00:17:27,146
this patient, and with
545
00:17:27,146 --> 00:17:29,448
monitoring. And in some cases,
546
00:17:29,515 --> 00:17:31,550
as we have where you have neurologists
547
00:17:31,550 --> 00:17:33,819
who specialize in stimulation, and
548
00:17:33,819 --> 00:17:35,120
keeping communication
549
00:17:35,387 --> 00:17:37,222
throughout the monitoring period between
550
00:17:37,222 --> 00:17:38,223
these different providers
551
00:17:38,323 --> 00:17:39,658
can really make it so that
552
00:17:39,658 --> 00:17:41,160
this test stimulation gives you a good
553
00:17:41,160 --> 00:17:43,929
backup strategy if a blade
554
00:17:43,929 --> 00:17:45,664
of a resective intervention
555
00:17:45,964 --> 00:17:47,399
does not provide the patient with seizure
556
00:17:47,399 --> 00:17:48,901
freedom. We've been
557
00:17:48,901 --> 00:17:50,936
performing research related to this
558
00:17:51,003 --> 00:17:53,972
with stereo EEG that helps guide our deep
559
00:17:53,972 --> 00:17:54,807
brain stimulation. I'd
560
00:17:54,807 --> 00:17:55,974
like to share that with you.
561
00:17:56,575 --> 00:17:58,877
One way that we can probe connectivity or
562
00:17:58,877 --> 00:17:59,745
probe the way that
563
00:17:59,745 --> 00:18:01,580
stimulating at one brain area affects
564
00:18:01,580 --> 00:18:04,083
another brain area is using something
565
00:18:04,083 --> 00:18:06,085
called brain stimulation evoked
566
00:18:06,085 --> 00:18:07,419
potentials. And that's
567
00:18:07,419 --> 00:18:08,987
where we deliver brief pulses of
568
00:18:08,987 --> 00:18:11,457
electrical current through one pair of
569
00:18:11,457 --> 00:18:13,258
electrodes, one brain area,
570
00:18:13,926 --> 00:18:16,462
and then measure evoked responses in the
571
00:18:16,462 --> 00:18:17,629
raw voltage at other
572
00:18:17,629 --> 00:18:20,098
brain areas. And by measuring
573
00:18:20,199 --> 00:18:22,968
many responses, meaning usually about 8
574
00:18:23,035 --> 00:18:25,137
to 15, we're able to get
575
00:18:25,137 --> 00:18:27,172
a very robust response.
576
00:18:27,206 --> 00:18:28,474
Classically, this had been called a
577
00:18:28,474 --> 00:18:29,641
cortical evoked
578
00:18:29,641 --> 00:18:31,276
potential. We now call it a brain
579
00:18:31,276 --> 00:18:32,744
stimulation evoked potential because
580
00:18:32,744 --> 00:18:34,680
we're often stimulating with single
581
00:18:34,680 --> 00:18:35,781
pulses enthalamus and
582
00:18:35,781 --> 00:18:39,351
finding out where in cortex we're able to
583
00:18:39,384 --> 00:18:40,519
affect with stimulation at
584
00:18:40,519 --> 00:18:41,687
that site. And so since we're
585
00:18:41,687 --> 00:18:43,255
stimulating enthalamus and measuring in
586
00:18:43,255 --> 00:18:44,323
cortex, we call it brain
587
00:18:44,323 --> 00:18:45,824
stimulation evoked potentials.
588
00:18:46,658 --> 00:18:48,293
We do find that there's a very wide
589
00:18:48,293 --> 00:18:49,595
variety of evoked potential
590
00:18:49,595 --> 00:18:51,096
shapes. And so if we want to
591
00:18:51,096 --> 00:18:53,999
capture these, we developed a tool called
592
00:18:53,999 --> 00:18:55,033
canonical response
593
00:18:55,033 --> 00:18:56,568
parameterization. For those that are
594
00:18:56,568 --> 00:18:58,136
interested, there's a QR code where you
595
00:18:58,136 --> 00:18:59,571
can find the paper that
596
00:18:59,571 --> 00:19:01,974
describes this. And we perform a
597
00:19:01,974 --> 00:19:04,543
set of analysis steps where first we take
598
00:19:04,543 --> 00:19:05,544
semi-normalized cross
599
00:19:05,544 --> 00:19:10,415
products between single responses
600
00:19:10,449 --> 00:19:12,885
and stimulation. And we develop temporal
601
00:19:12,885 --> 00:19:14,887
profiles of projections
602
00:19:14,887 --> 00:19:16,922
between these single trials that
603
00:19:16,922 --> 00:19:18,757
allow us to discover how long the
604
00:19:18,757 --> 00:19:20,993
response to a single pulse of stimulation
605
00:19:20,993 --> 00:19:22,394
lasts. And then we can
606
00:19:22,394 --> 00:19:25,264
parameterize single trials using compact
607
00:19:25,697 --> 00:19:26,698
representations where
608
00:19:26,698 --> 00:19:28,734
voltage is basically a scaled
609
00:19:28,734 --> 00:19:30,869
canonical form, which you see in this
610
00:19:30,869 --> 00:19:32,571
example here in red that's basically
611
00:19:32,571 --> 00:19:34,173
scaled on a single trial
612
00:19:34,173 --> 00:19:36,008
basis. And then we get a residual noise
613
00:19:36,008 --> 00:19:38,010
term. And from this, we can develop
614
00:19:38,010 --> 00:19:39,745
statistical measures of
615
00:19:40,279 --> 00:19:43,081
the response that these single pulses
616
00:19:43,081 --> 00:19:44,650
have. This is a very
617
00:19:44,650 --> 00:19:47,152
powerful tool. You don't need a
618
00:19:47,619 --> 00:19:49,221
background in signal processing or
619
00:19:49,221 --> 00:19:51,423
engineering to use this tool. We have
620
00:19:51,423 --> 00:19:52,624
developed an open source
621
00:19:52,624 --> 00:19:54,560
package that you can download from the
622
00:19:54,560 --> 00:19:55,794
internet. It's attached to this
623
00:19:55,794 --> 00:19:57,329
publication at the QR code
624
00:19:57,329 --> 00:19:59,331
on the screen. And this parameterization
625
00:19:59,331 --> 00:20:00,232
is a really powerful
626
00:20:00,232 --> 00:20:02,034
tool that lets us understand
627
00:20:02,401 --> 00:20:04,269
how different brain areas are connected
628
00:20:04,269 --> 00:20:05,504
to one another for
629
00:20:05,504 --> 00:20:07,206
electrical stimulation and for
630
00:20:07,306 --> 00:20:09,908
determining the best pair of electrodes
631
00:20:09,908 --> 00:20:11,910
you might stimulate for epilepsy. An
632
00:20:11,910 --> 00:20:12,978
example of this would
633
00:20:12,978 --> 00:20:15,314
be trying to disentangle stimulation in
634
00:20:15,314 --> 00:20:17,549
the pulvinar and where
635
00:20:17,649 --> 00:20:19,618
that may affect cortical
636
00:20:19,618 --> 00:20:23,021
circuits within a single patient. And in
637
00:20:23,021 --> 00:20:24,022
this example that I'm going
638
00:20:24,022 --> 00:20:25,557
to show you, we're measuring
639
00:20:25,557 --> 00:20:27,593
from two different stimulation sites in
640
00:20:27,593 --> 00:20:28,760
the pulvinar. We've put
641
00:20:28,760 --> 00:20:30,495
a stereo EEG electrode in
642
00:20:30,529 --> 00:20:31,930
from a lateral approach through the
643
00:20:31,930 --> 00:20:33,832
superior temporal gyrus and
644
00:20:33,832 --> 00:20:34,933
made our way into the pulvinar
645
00:20:34,933 --> 00:20:38,737
nucleus. And we're then measuring in
646
00:20:38,770 --> 00:20:40,806
temporal neocortex and occipital
647
00:20:40,806 --> 00:20:43,442
neocortex. And we find
648
00:20:43,442 --> 00:20:45,177
that if we stimulate in the medial
649
00:20:45,177 --> 00:20:47,713
pulvinar for this patient, we do get a
650
00:20:47,713 --> 00:20:49,047
statistically significant
651
00:20:49,047 --> 00:20:52,217
response in a temporal neocortex, but not
652
00:20:52,217 --> 00:20:53,485
in the occipital cortex.
653
00:20:53,952 --> 00:20:55,554
Conversely, if we stimulate the
654
00:20:55,554 --> 00:20:58,090
lateral pulvinar, we get a statistically
655
00:20:58,090 --> 00:20:59,258
significant response.
656
00:20:59,258 --> 00:21:00,559
And this is very highly
657
00:21:00,592 --> 00:21:03,128
significant. So less than one in a
658
00:21:03,128 --> 00:21:04,596
thousand percent chance
659
00:21:04,596 --> 00:21:07,766
that this is due to chance. So
660
00:21:07,766 --> 00:21:11,169
p-value of less than 0.001. We can say
661
00:21:11,169 --> 00:21:12,404
with very, very strong
662
00:21:12,404 --> 00:21:14,172
confidence that these evoked
663
00:21:14,172 --> 00:21:16,208
potentials reflect connectivity that
664
00:21:16,208 --> 00:21:17,376
would be useful for electrical
665
00:21:17,376 --> 00:21:19,044
stimulation, but only
666
00:21:19,278 --> 00:21:21,346
if you target the lateral pulvinar to get
667
00:21:21,346 --> 00:21:22,481
the occipital cortex
668
00:21:22,514 --> 00:21:23,949
and the medial pulvinar if
669
00:21:23,949 --> 00:21:24,883
you're trying to get to temporal
670
00:21:24,883 --> 00:21:27,219
neocortex. This tool was developed in
671
00:21:27,219 --> 00:21:28,420
collaboration with Dr.
672
00:21:28,453 --> 00:21:30,155
Hermes and Jordan Bilderbeek here at
673
00:21:30,155 --> 00:21:32,324
Mayo. And we discovered this
674
00:21:32,324 --> 00:21:33,759
using these evoked potentials
675
00:21:33,759 --> 00:21:36,795
with stereo EEG. And this has helped us
676
00:21:36,795 --> 00:21:38,430
guide our understanding of thalamic
677
00:21:38,430 --> 00:21:39,498
targets for epilepsy.
678
00:21:40,232 --> 00:21:43,135
So when we think about where we stimulate
679
00:21:43,268 --> 00:21:44,136
in the thalamus for
680
00:21:44,136 --> 00:21:46,004
epilepsy, I really think about
681
00:21:46,004 --> 00:21:47,939
stimulation of the anteronucleus to be
682
00:21:47,939 --> 00:21:50,742
connected to limbic circuitry. So
683
00:21:50,742 --> 00:21:51,877
anything involving the
684
00:21:52,010 --> 00:21:53,445
hippocampal circuits or cingulate
685
00:21:53,445 --> 00:21:56,548
circuits and for frontal distribution
686
00:21:56,615 --> 00:21:58,183
seizures. And the reason
687
00:21:58,183 --> 00:21:59,751
for the frontal distribution seizures in
688
00:21:59,751 --> 00:22:00,986
our case is because I use
689
00:22:00,986 --> 00:22:02,354
a posterior approach to the
690
00:22:02,387 --> 00:22:04,690
anteronucleus and I'm getting some of the
691
00:22:04,723 --> 00:22:06,491
more dorsal anterior
692
00:22:06,491 --> 00:22:08,694
thalamic areas on my way to the
693
00:22:08,694 --> 00:22:11,029
anteronucleus of the thalamus. For the
694
00:22:11,029 --> 00:22:12,431
centromedia nucleus, that's
695
00:22:12,431 --> 00:22:13,432
a spot that we're targeting
696
00:22:13,965 --> 00:22:16,201
either for perirhealandic seizures, so
697
00:22:16,201 --> 00:22:17,402
seizures that seem to be motor
698
00:22:17,402 --> 00:22:19,104
in nature, or for generalized
699
00:22:19,237 --> 00:22:21,139
seizures. And the centromedia nucleus of
700
00:22:21,139 --> 00:22:22,607
the thalamus has really
701
00:22:22,641 --> 00:22:24,876
evolved as a target for
702
00:22:24,910 --> 00:22:27,245
generalized epilepsies. There are trials
703
00:22:27,245 --> 00:22:28,814
going on right now with the
704
00:22:28,814 --> 00:22:30,115
neuropace device targeting
705
00:22:30,115 --> 00:22:31,850
the centromedia nucleus of the thalamus
706
00:22:32,250 --> 00:22:34,286
as a specific site for
707
00:22:34,286 --> 00:22:35,554
generalized epilepsies.
708
00:22:36,355 --> 00:22:38,623
In the pulvinar nucleus, because of these
709
00:22:38,623 --> 00:22:40,392
findings that we've had and
710
00:22:40,392 --> 00:22:42,060
also some early results with
711
00:22:42,060 --> 00:22:44,863
different patients, we really do target
712
00:22:44,863 --> 00:22:45,997
specific sub-regions of
713
00:22:45,997 --> 00:22:47,365
the pulvinar depending upon
714
00:22:47,699 --> 00:22:50,502
where we localize patient seizures to. So
715
00:22:50,502 --> 00:22:51,703
for occipital seizures,
716
00:22:51,703 --> 00:22:52,604
whether they're medial or
717
00:22:52,671 --> 00:22:55,006
lateral, we're really thinking about the
718
00:22:55,006 --> 00:22:56,641
lateral pulvinar. If we
719
00:22:56,641 --> 00:22:58,176
have parietal neocortex, we're
720
00:22:58,176 --> 00:22:59,845
thinking about dorsal medial pulvinar,
721
00:23:00,379 --> 00:23:01,480
and for temporal cortex,
722
00:23:01,480 --> 00:23:02,347
we're thinking about ventral
723
00:23:02,381 --> 00:23:04,483
medial pulvinar. And so when we think
724
00:23:04,483 --> 00:23:05,984
about targeting the
725
00:23:05,984 --> 00:23:08,286
thalamus for epilepsy targets,
726
00:23:08,620 --> 00:23:10,922
there are a set of considerations that
727
00:23:11,089 --> 00:23:11,990
are really important to
728
00:23:11,990 --> 00:23:13,658
our team. The first is,
729
00:23:13,692 --> 00:23:15,293
is there substructure within it, as I've
730
00:23:15,293 --> 00:23:16,328
shown you for the pulvinar.
731
00:23:17,028 --> 00:23:18,497
The other is, is if there's a
732
00:23:18,530 --> 00:23:20,098
structure that we want to target, we want
733
00:23:20,098 --> 00:23:22,167
to get along the long axis of that
734
00:23:22,167 --> 00:23:24,369
structure. And we're
735
00:23:24,369 --> 00:23:25,704
still trying to figure out what the
736
00:23:25,704 --> 00:23:26,938
optimal targeting
737
00:23:26,938 --> 00:23:28,673
approach may be. I tend to use
738
00:23:28,673 --> 00:23:30,842
posterolateral approaches and pure
739
00:23:30,842 --> 00:23:32,677
lateral approaches, more
740
00:23:32,677 --> 00:23:33,812
than many in the community.
741
00:23:34,613 --> 00:23:36,014
But for me, that's because I'm getting
742
00:23:36,014 --> 00:23:37,349
along the long axis of
743
00:23:37,349 --> 00:23:38,517
the centromedian nucleus.
744
00:23:38,984 --> 00:23:41,286
If I take a parietal approach or the long
745
00:23:41,286 --> 00:23:42,387
axis of the pulvinar
746
00:23:42,387 --> 00:23:43,955
nucleus from a purely lateral
747
00:23:43,955 --> 00:23:45,657
approach, and then for the anterior
748
00:23:45,657 --> 00:23:46,792
nucleus, it's the same
749
00:23:46,792 --> 00:23:47,926
thing where I get along the long
750
00:23:47,926 --> 00:23:50,162
axis of it. And I also get some of the
751
00:23:50,362 --> 00:23:52,898
adjacent thalamic nuclei. I
752
00:23:52,898 --> 00:23:54,232
also think about feasibility
753
00:23:54,232 --> 00:23:56,635
and safety. So it's more difficult for me
754
00:23:56,635 --> 00:23:58,637
to have confidence in my
755
00:23:58,637 --> 00:24:00,705
ability to be within a structure
756
00:24:00,739 --> 00:24:03,308
if I'm more confident if I've stayed out
757
00:24:03,308 --> 00:24:04,176
of the ventricles. And the
758
00:24:04,176 --> 00:24:05,644
reason for that is not that
759
00:24:06,077 --> 00:24:08,046
I think that my accuracy is from
760
00:24:08,046 --> 00:24:09,381
deflection of a cannula or
761
00:24:09,381 --> 00:24:10,482
something like that, but actually
762
00:24:10,515 --> 00:24:12,818
because as you pass through the ventricle
763
00:24:12,851 --> 00:24:14,252
and enter into a
764
00:24:14,252 --> 00:24:15,620
structure, for example, the
765
00:24:15,921 --> 00:24:17,989
anteronucleosophthalamus, the thalamus
766
00:24:17,989 --> 00:24:19,658
itself deforms in
767
00:24:19,658 --> 00:24:21,426
response to the incident cannula.
768
00:24:21,626 --> 00:24:22,828
And that will actually change your
769
00:24:22,828 --> 00:24:23,795
trajectory as you're in the
770
00:24:23,795 --> 00:24:26,665
structure. And also, I'm less
771
00:24:26,665 --> 00:24:29,000
confident about things like vascular
772
00:24:29,000 --> 00:24:30,936
safety if I pass through a
773
00:24:30,936 --> 00:24:33,305
ventricle. I feel more confident
774
00:24:33,305 --> 00:24:35,674
about being away from vascular anatomy if
775
00:24:35,674 --> 00:24:37,175
I have not gone through
776
00:24:37,175 --> 00:24:39,811
sulci, although oftentimes I do
777
00:24:40,078 --> 00:24:42,280
across sulci. So avoiding going through
778
00:24:42,280 --> 00:24:43,682
sulci is sort of a soft
779
00:24:43,882 --> 00:24:45,116
contraindication for me, and I
780
00:24:45,116 --> 00:24:47,219
will often have to do that in my DBS
781
00:24:47,219 --> 00:24:48,753
trajectories for epilepsy.
782
00:24:49,221 --> 00:24:50,689
When we think about targeting in
783
00:24:50,689 --> 00:24:53,124
general for deep brain stimulation, there
784
00:24:53,124 --> 00:24:53,892
are a few different
785
00:24:53,892 --> 00:24:55,193
considerations that we have.
786
00:24:55,493 --> 00:24:56,995
The first is what's called indirect
787
00:24:56,995 --> 00:24:58,363
targeting, and that's
788
00:24:58,363 --> 00:24:59,664
where we have a proxy landmark.
789
00:24:59,698 --> 00:25:01,366
Classically, this was the anterior
790
00:25:01,366 --> 00:25:02,868
commissure and the posterior commissure,
791
00:25:02,868 --> 00:25:03,768
and I'll show you examples
792
00:25:03,768 --> 00:25:06,738
of that. Another is by using an atlas,
793
00:25:06,972 --> 00:25:07,973
which is determined
794
00:25:08,039 --> 00:25:09,841
either from knowledge of prior
795
00:25:09,875 --> 00:25:12,344
efficacy and stimulation or from
796
00:25:12,677 --> 00:25:15,380
histology, for example. And we often are
797
00:25:15,380 --> 00:25:16,882
able to merge in atlases
798
00:25:16,882 --> 00:25:19,284
that can be custom deformed to individual
799
00:25:19,284 --> 00:25:21,486
patient's anatomy. And then the last
800
00:25:21,486 --> 00:25:22,153
thing that we use for
801
00:25:22,153 --> 00:25:23,855
targeting is direct visualization of the
802
00:25:23,855 --> 00:25:24,523
structure that we're
803
00:25:24,523 --> 00:25:26,124
interested in, and I'll show you some
804
00:25:26,124 --> 00:25:28,693
examples of each. So indirect targeting,
805
00:25:28,727 --> 00:25:29,761
again, is when we use
806
00:25:29,761 --> 00:25:31,830
landmarks in order to determine
807
00:25:31,863 --> 00:25:34,366
where we believe our target should be.
808
00:25:34,432 --> 00:25:35,800
Classically, this was
809
00:25:35,834 --> 00:25:37,369
the anterior and posterior
810
00:25:37,369 --> 00:25:39,504
commissures, and the reason for that was
811
00:25:39,504 --> 00:25:40,238
that even on a
812
00:25:40,238 --> 00:25:41,706
ventricular gram, where you flooded
813
00:25:41,706 --> 00:25:43,241
the ventricles with contrast and then
814
00:25:43,241 --> 00:25:44,809
took an X-ray purely
815
00:25:44,809 --> 00:25:46,611
laterally, you could see these
816
00:25:46,611 --> 00:25:49,247
landmarks and target accordingly. And so
817
00:25:49,247 --> 00:25:49,981
for targeting, for
818
00:25:49,981 --> 00:25:51,850
example, in movement, the VIM
819
00:25:51,850 --> 00:25:54,619
nucleus of thalamus, you will know
820
00:25:54,619 --> 00:25:56,821
classically that this is a
821
00:25:56,821 --> 00:25:58,189
quarter of the way from the
822
00:25:58,223 --> 00:26:00,258
posterior commissure going anteriorly,
823
00:26:00,525 --> 00:26:03,194
usually at the level of ACPC from a
824
00:26:03,194 --> 00:26:04,729
dorsoventral perspective,
825
00:26:05,096 --> 00:26:07,465
and then usually something like 10
826
00:26:07,465 --> 00:26:09,668
millimeters off of the ventricular wall.
827
00:26:10,101 --> 00:26:10,902
And these rules will
828
00:26:10,902 --> 00:26:12,470
change a little bit depending upon the
829
00:26:12,470 --> 00:26:13,872
width of the thalamus from the
830
00:26:13,872 --> 00:26:15,140
ventricular wall to the
831
00:26:15,273 --> 00:26:17,042
capsule, which you can now see on white
832
00:26:17,042 --> 00:26:18,944
patterned hole imaging. You
833
00:26:18,944 --> 00:26:20,745
may also use atlases that are
834
00:26:20,745 --> 00:26:23,148
built into software. So this is showing
835
00:26:23,148 --> 00:26:24,282
you software from
836
00:26:24,282 --> 00:26:25,817
BrainLab, but they're equivalent
837
00:26:26,084 --> 00:26:29,054
software programs in other targeting
838
00:26:29,087 --> 00:26:30,922
softwares. You can use these
839
00:26:30,922 --> 00:26:31,923
atlases. The one thing I would
840
00:26:31,923 --> 00:26:33,725
say is, is anytime you use an atlas
841
00:26:33,725 --> 00:26:34,893
that's built in, make sure you
842
00:26:34,893 --> 00:26:36,194
look at it yourself and review
843
00:26:36,194 --> 00:26:38,663
it carefully. We can also use customized
844
00:26:38,663 --> 00:26:39,364
segmentation
845
00:26:39,364 --> 00:26:41,366
programs. We will use these to
846
00:26:41,366 --> 00:26:43,768
help us in an off-label fashion. These
847
00:26:43,802 --> 00:26:45,403
things are designed for research, but
848
00:26:45,804 --> 00:26:47,005
those allow us to do
849
00:26:47,038 --> 00:26:48,840
perform custom deformation of patient
850
00:26:48,873 --> 00:26:50,542
anatomy. It allows us to
851
00:26:50,542 --> 00:26:52,844
take our own results and use
852
00:26:52,844 --> 00:26:54,546
aggregated statistical measures from our
853
00:26:54,546 --> 00:26:55,847
results in a standardized
854
00:26:55,847 --> 00:26:57,816
space and then warp those into
855
00:26:58,183 --> 00:26:59,818
an individual patient space. And we do
856
00:26:59,818 --> 00:27:01,987
this for many of our epilepsy patients,
857
00:27:02,287 --> 00:27:04,122
particularly when we're targeting, for
858
00:27:04,122 --> 00:27:05,156
example, the hotspot of
859
00:27:05,156 --> 00:27:07,092
the centromedian nucleus for
860
00:27:07,092 --> 00:27:08,860
generalized or perirhealandic epilepsies.
861
00:27:10,595 --> 00:27:13,131
So for direct targeting, I think the
862
00:27:13,131 --> 00:27:15,700
antinucleus thalamus is a great example
863
00:27:15,700 --> 00:27:17,502
because for me, the mammal
864
00:27:17,502 --> 00:27:19,304
thalamic tract and the output
865
00:27:19,304 --> 00:27:20,805
of the mammal thalamic tract is really
866
00:27:20,805 --> 00:27:22,607
what we're trying to perform stimulation
867
00:27:22,607 --> 00:27:25,143
on. And the way that I
868
00:27:25,243 --> 00:27:27,312
get the outflow of the mammal thalamic
869
00:27:27,312 --> 00:27:30,515
tract is to look at it on imaging. And
870
00:27:30,515 --> 00:27:31,383
you can see it pretty
871
00:27:31,383 --> 00:27:33,284
well here on this image. If you look at
872
00:27:33,284 --> 00:27:34,953
the bottom left, you can see the mammal
873
00:27:34,953 --> 00:27:35,987
thalamic tract coming
874
00:27:35,987 --> 00:27:38,923
up. And what I'll do is I'll put, and you
875
00:27:38,923 --> 00:27:39,724
can see it actually in the
876
00:27:39,758 --> 00:27:41,893
top left as well, I'll put a
877
00:27:41,893 --> 00:27:45,096
target sitting on the top fanning out
878
00:27:45,096 --> 00:27:46,664
portion of the mammal
879
00:27:46,664 --> 00:27:48,600
thalamic tract. I think it looks like
880
00:27:48,633 --> 00:27:50,235
something called a baobab tree, which is
881
00:27:50,235 --> 00:27:51,436
a really, really cool tree
882
00:27:51,436 --> 00:27:52,904
that grows in Africa. And one of
883
00:27:52,904 --> 00:27:54,472
the neurologists I work with, Nick Gregg,
884
00:27:54,472 --> 00:27:55,373
really thinks it looks more
885
00:27:55,373 --> 00:27:56,474
like the top of the mushroom,
886
00:27:56,741 --> 00:27:58,243
but really we get at the top of the
887
00:27:58,243 --> 00:28:00,445
mammal thalamic tract. And here you can
888
00:28:00,445 --> 00:28:01,813
see it. And you know,
889
00:28:01,813 --> 00:28:03,882
I identify the fanning out of the mammal
890
00:28:03,882 --> 00:28:05,650
thalamic tract and I put a target there.
891
00:28:06,618 --> 00:28:08,386
And then I'll extend my trajectory
892
00:28:08,386 --> 00:28:10,989
posteriorly. I will stay close to the
893
00:28:10,989 --> 00:28:12,390
ventricle. My reason for
894
00:28:12,390 --> 00:28:15,927
doing that is I really, if I can, want to
895
00:28:15,927 --> 00:28:17,028
stay out of the sensory
896
00:28:17,028 --> 00:28:18,430
motor portions of the internal
897
00:28:18,463 --> 00:28:20,799
capsule for where my cannula is passing
898
00:28:20,799 --> 00:28:22,333
through. I do also want to
899
00:28:22,333 --> 00:28:23,435
stay out of the ventricle and a
900
00:28:23,435 --> 00:28:25,070
little, a few millimeters away from the
901
00:28:25,070 --> 00:28:27,439
ventricular wall because I do worry that
902
00:28:27,439 --> 00:28:28,840
the appendema and the
903
00:28:28,873 --> 00:28:30,809
immediate sub-appendema is more likely to
904
00:28:30,809 --> 00:28:32,744
bleed. I obviously completely avoid
905
00:28:32,744 --> 00:28:33,545
vessels, but I try to
906
00:28:33,545 --> 00:28:35,713
avoid sulci as well. I'll look at this on
907
00:28:35,713 --> 00:28:37,282
different sequences. So you saw white
908
00:28:37,282 --> 00:28:37,882
matter and all in the
909
00:28:37,882 --> 00:28:40,752
last slide, you see T2 here. And then I
910
00:28:40,752 --> 00:28:41,653
look at it again in my
911
00:28:41,653 --> 00:28:43,922
CTV-Venogram and I run through both my
912
00:28:43,922 --> 00:28:48,193
CTV-Venogram and my MRI T1 post contrast
913
00:28:48,493 --> 00:28:49,360
the whole way along the
914
00:28:49,360 --> 00:28:50,695
span to make sure vessels
915
00:28:51,096 --> 00:28:53,431
are avoided. And then once I have that
916
00:28:53,465 --> 00:28:55,600
optimal trajectory, I then extend
917
00:28:56,167 --> 00:28:57,402
anteriorly all the way
918
00:28:57,402 --> 00:29:00,371
to the sub-appendemal zone. So I know
919
00:29:00,371 --> 00:29:02,740
that my electrodes at the tip, my
920
00:29:02,740 --> 00:29:03,908
contacts at the tip of
921
00:29:03,908 --> 00:29:06,878
the TBS lead are sitting in the anterior
922
00:29:06,878 --> 00:29:08,179
nucleus itself. And the
923
00:29:08,313 --> 00:29:10,148
second and third levels of contacts
924
00:29:10,482 --> 00:29:11,850
are sitting in the fanning out of the
925
00:29:11,850 --> 00:29:12,584
mantle of the laminar
926
00:29:12,617 --> 00:29:13,885
tract so that my colleagues
927
00:29:14,185 --> 00:29:16,087
in neurology can directly stimulate the
928
00:29:16,087 --> 00:29:16,788
mammal of the laminar
929
00:29:16,821 --> 00:29:18,523
tract for optimal therapy.
930
00:29:19,624 --> 00:29:20,358
When I'm targeting the centromedian
931
00:29:20,358 --> 00:29:23,862
nucleus, I do often use an atlas burn-in
932
00:29:23,895 --> 00:29:25,063
as you see here. And
933
00:29:25,063 --> 00:29:27,165
for us, this means burning the atlas
934
00:29:27,198 --> 00:29:29,100
using a research software
935
00:29:29,134 --> 00:29:30,635
called lead DBS and then
936
00:29:30,635 --> 00:29:32,203
burning that back into an MRI and then
937
00:29:32,203 --> 00:29:33,037
bringing it into our
938
00:29:33,037 --> 00:29:34,739
planning software. You have to be
939
00:29:34,739 --> 00:29:36,307
really careful in the centromedian
940
00:29:36,374 --> 00:29:37,075
because anterolaterally
941
00:29:37,075 --> 00:29:39,811
you'll have the VC nucleus. So the
942
00:29:39,911 --> 00:29:42,947
sensory area of the thalamus and then
943
00:29:42,947 --> 00:29:44,315
infralaterally and
944
00:29:44,315 --> 00:29:45,483
directly inferiorly you have
945
00:29:45,483 --> 00:29:47,418
laminar scar fibers and those can produce
946
00:29:47,418 --> 00:29:48,520
sensory side effects as
947
00:29:48,520 --> 00:29:50,054
well. And so we take particular
948
00:29:50,054 --> 00:29:52,724
care to avoid these. I think that this
949
00:29:52,724 --> 00:29:54,392
manuscript here by Warren et
950
00:29:54,392 --> 00:29:56,694
al is really nice in showing
951
00:29:56,694 --> 00:29:58,096
you the anatomy and how to think about
952
00:29:58,096 --> 00:29:58,763
the centromedian
953
00:29:58,763 --> 00:30:00,532
nucleus. I've included a QR code
954
00:30:00,532 --> 00:30:02,500
if you want to get it. In our imaging,
955
00:30:02,534 --> 00:30:04,002
I'm really looking below the medial
956
00:30:04,035 --> 00:30:05,170
dorsal nucleus because
957
00:30:05,170 --> 00:30:07,205
it's easy to see that. And if you look
958
00:30:07,205 --> 00:30:08,173
here on the right, you can
959
00:30:08,173 --> 00:30:09,674
see the medial dorsal nucleus
960
00:30:09,674 --> 00:30:11,743
and then I target directly below that.
961
00:30:11,843 --> 00:30:13,511
And I tend to take a poster lateral
962
00:30:13,511 --> 00:30:14,746
approach into that target.
963
00:30:15,113 --> 00:30:16,881
And here with the dash white line, I've
964
00:30:16,881 --> 00:30:17,916
circled the medial dorsal
965
00:30:17,916 --> 00:30:19,651
nucleus and I pass below it to
966
00:30:19,651 --> 00:30:21,753
get to centromedian. For pulvinar, I'm
967
00:30:21,753 --> 00:30:23,354
coming pretty purely
968
00:30:23,354 --> 00:30:25,423
laterally. I'll usually be coming
969
00:30:25,423 --> 00:30:27,192
in at the superior temporal gyrus.
970
00:30:27,392 --> 00:30:29,694
Sometimes I'll be coming in as the
971
00:30:29,694 --> 00:30:30,895
superior temporal gyrus
972
00:30:30,895 --> 00:30:33,231
transitions around to become the super
973
00:30:33,231 --> 00:30:35,133
marginal gyrus. You have to be very
974
00:30:35,133 --> 00:30:36,000
careful here because
975
00:30:36,000 --> 00:30:38,670
you're in the Wernicke's vicinity. And so
976
00:30:38,670 --> 00:30:40,038
I try to avoid coterie of
977
00:30:40,038 --> 00:30:41,539
the cortex there or being too
978
00:30:41,539 --> 00:30:43,107
aggressive when cutting the arachnoid
979
00:30:43,107 --> 00:30:45,543
open to pass my cannula in. And I try to
980
00:30:45,543 --> 00:30:46,544
avoid the thick white
981
00:30:46,544 --> 00:30:48,379
matter of the arc of the fasciculus.
982
00:30:48,446 --> 00:30:50,381
Obviously on the dominant side, we do
983
00:30:50,381 --> 00:30:52,083
worry about interrupting
984
00:30:52,083 --> 00:30:54,219
language if there's a problem. And the
985
00:30:54,219 --> 00:30:55,053
last thing I want to talk
986
00:30:55,053 --> 00:30:56,354
to you about today is tandem
987
00:30:56,354 --> 00:30:58,523
targeting. And by tandem targeting, we're
988
00:30:58,523 --> 00:30:59,991
really thinking about when
989
00:30:59,991 --> 00:31:02,794
we're using multiple DBS leads,
990
00:31:03,094 --> 00:31:05,296
whether we're targeting two sites within
991
00:31:05,296 --> 00:31:06,264
a connected circuit. And
992
00:31:06,264 --> 00:31:07,699
I'll show you some examples
993
00:31:07,699 --> 00:31:09,467
of that. We're targeting separate
994
00:31:09,467 --> 00:31:10,902
circuits, as I showed you before,
995
00:31:10,902 --> 00:31:12,737
targeting the different
996
00:31:12,737 --> 00:31:14,706
regions maybe within the pulvinar or
997
00:31:14,706 --> 00:31:17,342
targeting pulvinar and anteronucleus
998
00:31:17,408 --> 00:31:19,010
together to get limbic
999
00:31:19,010 --> 00:31:20,678
circuitry from the anteronucleus and
1000
00:31:20,678 --> 00:31:22,447
temporal neocortex from the
1001
00:31:22,447 --> 00:31:24,382
pulvinar. Here, I just have
1002
00:31:24,382 --> 00:31:26,517
a cartoon where two different circuits
1003
00:31:26,517 --> 00:31:27,852
are represented in blue
1004
00:31:27,852 --> 00:31:29,687
and orange. And let's say
1005
00:31:29,687 --> 00:31:31,256
we're targeting two different circuits
1006
00:31:31,256 --> 00:31:33,024
that would be targeting one node of the
1007
00:31:33,024 --> 00:31:34,325
blue circuit and one
1008
00:31:34,325 --> 00:31:36,294
node of the orange circuit. But we may
1009
00:31:36,294 --> 00:31:37,629
also be targeting connected
1010
00:31:37,629 --> 00:31:39,197
circuits where we have one
1011
00:31:39,197 --> 00:31:41,566
node of the blue and often that would be
1012
00:31:41,633 --> 00:31:43,001
a target in the thalamus
1013
00:31:43,401 --> 00:31:44,602
and another node of the blue,
1014
00:31:44,602 --> 00:31:46,871
which might be a neocortical site that we
1015
00:31:46,871 --> 00:31:48,273
believe is involved with C4
1016
00:31:48,273 --> 00:31:51,376
onset. I use a Boston Scientific
1017
00:31:51,376 --> 00:31:54,112
device, which is off label, which is a
1018
00:31:54,112 --> 00:31:56,314
four lead device that has four eight
1019
00:31:56,314 --> 00:31:57,415
channel leads that you
1020
00:31:57,448 --> 00:31:59,851
can use. And so you can allow for
1021
00:31:59,851 --> 00:32:02,053
targeting neocortex as well
1022
00:32:02,053 --> 00:32:04,422
as targeting thalamus together.
1023
00:32:06,424 --> 00:32:07,859
I use that also for
1024
00:32:07,859 --> 00:32:09,494
targeting multiple thalamic circuits.
1025
00:32:09,494 --> 00:32:11,129
So here you see an example of targeting
1026
00:32:11,129 --> 00:32:12,130
anteronucleus and
1027
00:32:12,130 --> 00:32:13,464
centromedian nucleus together.
1028
00:32:13,932 --> 00:32:15,166
For this construct, I'm often
1029
00:32:15,166 --> 00:32:16,834
able to use the same burr hole
1030
00:32:17,235 --> 00:32:17,802
to target both.
1031
00:32:18,469 --> 00:32:20,738
Targeting different nodes in tandem
1032
00:32:21,239 --> 00:32:21,873
is something that I
1033
00:32:21,873 --> 00:32:23,107
also find very useful.
1034
00:32:23,107 --> 00:32:24,409
So this is an example of targeting
1035
00:32:24,409 --> 00:32:26,277
the centromedian nucleus of the thalamus
1036
00:32:26,644 --> 00:32:28,413
for stereo-EEG-identified
1037
00:32:28,413 --> 00:32:29,647
perirhealandic epilepsy.
1038
00:32:30,048 --> 00:32:32,617
So here we had initial DBS leads
1039
00:32:32,650 --> 00:32:34,686
in the superior frontal gyrus,
1040
00:32:35,086 --> 00:32:36,187
the pre-central gyrus
1041
00:32:36,187 --> 00:32:37,622
and the post-central gyrus.
1042
00:32:37,622 --> 00:32:39,023
And then I put DBS
1043
00:32:39,090 --> 00:32:40,992
leads in cortical targets
1044
00:32:41,292 --> 00:32:42,360
with the eight contact
1045
00:32:42,360 --> 00:32:44,262
linear lead in cortex
1046
00:32:44,262 --> 00:32:45,997
and then the eight contact segmented lead
1047
00:32:45,997 --> 00:32:47,265
in centromedian nucleus.
1048
00:32:47,832 --> 00:32:49,233
We've also used the
1049
00:32:49,300 --> 00:32:51,102
Medtronic RC-Plus-S device
1050
00:32:51,369 --> 00:32:53,471
to target the hippocampus
1051
00:32:53,671 --> 00:32:55,707
alongside anteronucleus of the thalamus.
1052
00:32:55,707 --> 00:32:56,607
This was our portion
1053
00:32:56,607 --> 00:32:57,875
of the RC-Plus-S trial
1054
00:32:58,009 --> 00:32:59,610
run by Dr. Worrell here at Mayo.
1055
00:32:59,877 --> 00:33:01,079
Sometimes we will
1056
00:33:01,079 --> 00:33:03,214
have insular epilepsies,
1057
00:33:03,214 --> 00:33:04,015
and I think those can
1058
00:33:04,015 --> 00:33:04,916
be quite complicated
1059
00:33:04,916 --> 00:33:05,516
because I think
1060
00:33:05,516 --> 00:33:07,952
different circuits are involved
1061
00:33:07,986 --> 00:33:09,253
with insular epilepsy.
1062
00:33:09,253 --> 00:33:09,821
So you can have
1063
00:33:09,821 --> 00:33:11,556
different, both limbic circuitry
1064
00:33:11,923 --> 00:33:13,658
as well as perirhealandic,
1065
00:33:13,691 --> 00:33:14,692
what I think of as
1066
00:33:14,692 --> 00:33:17,228
centromedian involved circuits.
1067
00:33:17,228 --> 00:33:17,862
And so this is an
1068
00:33:17,862 --> 00:33:19,097
example where we targeted
1069
00:33:19,397 --> 00:33:22,633
two insular sites and two thalamic sites
1070
00:33:22,800 --> 00:33:23,835
with the same construct
1071
00:33:23,835 --> 00:33:25,136
all in one hemisphere.
1072
00:33:25,703 --> 00:33:28,072
The take home messages I'd like to convey
1073
00:33:28,306 --> 00:33:29,607
from this lecture is that
1074
00:33:29,707 --> 00:33:30,908
having multidisciplinary
1075
00:33:30,942 --> 00:33:31,943
interactions between
1076
00:33:31,943 --> 00:33:32,744
all the different types
1077
00:33:32,744 --> 00:33:33,478
of clinicians,
1078
00:33:33,711 --> 00:33:35,847
neurologists, neuroradiologists,
1079
00:33:35,847 --> 00:33:36,814
neuropsychologists,
1080
00:33:36,814 --> 00:33:39,250
neuropsychiatrists, neurosurgeons
1081
00:33:39,751 --> 00:33:42,520
can produce optimal outcomes in workup
1082
00:33:42,520 --> 00:33:43,287
and decision making.
1083
00:33:43,755 --> 00:33:45,289
Multi-stage workup for
1084
00:33:45,289 --> 00:33:46,657
epilepsy should involve
1085
00:33:47,191 --> 00:33:48,092
consideration and
1086
00:33:48,092 --> 00:33:49,827
discussion about concordance
1087
00:33:49,827 --> 00:33:51,295
between electrophysiology,
1088
00:33:51,863 --> 00:33:53,831
brain imaging, and semiology,
1089
00:33:53,831 --> 00:33:54,432
the symptoms the
1090
00:33:54,432 --> 00:33:55,767
patients have during seizures,
1091
00:33:56,234 --> 00:33:57,702
and working potential
1092
00:33:57,702 --> 00:33:59,637
downstream deep brain stimulation
1093
00:33:59,637 --> 00:34:01,572
to your stereo EEG construct
1094
00:34:01,873 --> 00:34:03,374
is a very important piece of this.
1095
00:34:03,374 --> 00:34:04,909
And the reason is not because you think
1096
00:34:04,909 --> 00:34:05,843
that you're gonna be going
1097
00:34:05,843 --> 00:34:08,346
to DBS upfront necessarily,
1098
00:34:08,546 --> 00:34:11,516
but because if they fail a resective or
1099
00:34:11,516 --> 00:34:13,084
ablative intervention,
1100
00:34:13,117 --> 00:34:14,118
deep brain stimulation
1101
00:34:14,118 --> 00:34:15,386
remains a viable option
1102
00:34:15,420 --> 00:34:16,788
you've already tested and
1103
00:34:17,121 --> 00:34:18,823
determined the optimal approach
1104
00:34:18,823 --> 00:34:20,591
for within a stimulation paradigm
1105
00:34:20,858 --> 00:34:22,527
during your stereo EEG monitoring.
1106
00:34:22,860 --> 00:34:24,595
Something that I didn't talk a lot about
1107
00:34:24,762 --> 00:34:25,963
during this lecture,
1108
00:34:25,963 --> 00:34:26,898
but is very important,
1109
00:34:26,931 --> 00:34:28,399
is that advances in
1110
00:34:28,433 --> 00:34:30,201
imaging and improvements
1111
00:34:30,234 --> 00:34:30,635
in stereotactic
1112
00:34:30,635 --> 00:34:32,336
targeting as more and more people
1113
00:34:32,336 --> 00:34:34,272
use robotics is reducing
1114
00:34:34,272 --> 00:34:35,473
complications associated
1115
00:34:35,473 --> 00:34:37,141
with stereo EEG, the
1116
00:34:37,141 --> 00:34:39,010
reduction in the caliber
1117
00:34:39,310 --> 00:34:41,779
of stereo EEG electrodes from
1118
00:34:41,779 --> 00:34:44,482
1.2, 1.3 millimeter diameter
1119
00:34:44,682 --> 00:34:46,017
electrodes down to 0.8
1120
00:34:46,050 --> 00:34:47,452
millimeter diameter electrodes
1121
00:34:47,452 --> 00:34:48,553
has made it much safer.
1122
00:34:48,886 --> 00:34:51,756
So that including thalamic stereo EEG,
1123
00:34:51,756 --> 00:34:52,857
I think is much safer
1124
00:34:52,857 --> 00:34:54,559
today than it would have been
1125
00:34:54,959 --> 00:34:56,160
15 to 20 years ago.
1126
00:34:56,727 --> 00:34:58,229
Overall, I do think that
1127
00:34:58,229 --> 00:35:00,198
stimulation for epilepsy
1128
00:35:00,231 --> 00:35:02,266
is an evolving and growing field.
1129
00:35:02,667 --> 00:35:03,534
And we're really
1130
00:35:03,534 --> 00:35:04,802
seeing remarkable results
1131
00:35:04,802 --> 00:35:06,304
across many institutions,
1132
00:35:06,504 --> 00:35:07,472
not just here in America,
1133
00:35:07,505 --> 00:35:09,674
but also our colleagues around the world.
1134
00:35:09,740 --> 00:35:11,342
This is a paper that we put out
1135
00:35:11,342 --> 00:35:12,944
that talks about thalamic targeting.
1136
00:35:13,211 --> 00:35:14,946
And these are some general
1137
00:35:14,979 --> 00:35:16,447
papers related to stereotaxy
1138
00:35:16,447 --> 00:35:17,815
that I think are useful to review.
1139
00:35:17,982 --> 00:35:19,150
I'm very grateful to be able
1140
00:35:19,150 --> 00:35:20,485
to share all of these concepts
1141
00:35:20,485 --> 00:35:21,385
with you on the STEIN
1142
00:35:21,385 --> 00:35:22,653
Fireside Podcast.
1143
00:35:22,753 --> 00:35:24,188
Well, I hope you enjoyed that lecture.
1144
00:35:24,222 --> 00:35:24,755
Kai, it was really
1145
00:35:24,755 --> 00:35:26,057
interesting to see the variety
1146
00:35:26,057 --> 00:35:27,191
of approaches because
1147
00:35:27,191 --> 00:35:28,793
in movement disorders,
1148
00:35:28,793 --> 00:35:30,561
we have a fixed set of targets
1149
00:35:30,561 --> 00:35:31,362
and are usually applying
1150
00:35:31,362 --> 00:35:33,097
those to a fixed set of disorders.
1151
00:35:33,965 --> 00:35:35,399
And epilepsy, well, obviously they all,
1152
00:35:35,433 --> 00:35:36,501
all these patients have seizures.
1153
00:35:36,701 --> 00:35:39,036
There's a very, very rich variety of
1154
00:35:39,036 --> 00:35:40,872
underlying pathology
1155
00:35:40,872 --> 00:35:41,706
and pathophysiology.
1156
00:35:41,706 --> 00:35:43,241
And so it takes a
1157
00:35:43,541 --> 00:35:44,575
different set of approaches.
1158
00:35:44,742 --> 00:35:46,777
Yeah, and it's really an evolving toolkit
1159
00:35:46,777 --> 00:35:47,879
that we have to treat
1160
00:35:47,879 --> 00:35:48,880
patients with epilepsy
1161
00:35:48,880 --> 00:35:50,748
using stimulation, really
1162
00:35:50,748 --> 00:35:51,849
identifying the difference
1163
00:35:51,849 --> 00:35:53,351
between where the seizures start from
1164
00:35:53,351 --> 00:35:54,485
and where they spread through
1165
00:35:54,886 --> 00:35:56,153
and trying to use the
1166
00:35:56,153 --> 00:35:57,622
best of current technology
1167
00:35:57,622 --> 00:35:59,724
to have a toolkit to
1168
00:35:59,757 --> 00:36:00,758
stop patient seizures.
1169
00:36:01,526 --> 00:36:02,527
Well, we hope you enjoyed the talk
1170
00:36:02,560 --> 00:36:03,227
and we'll see you next time.