Basal Ganglia Anatomy Pharmacology and Clinical Physiology and by MikeJenny

VIEWS: 21 PAGES: 18

									Basal Ganglia: Anatomy, Pharmacology, and
Clinical Physiology and Surgical Treatments
Guy M. McKhann, MD

January 19, 2008



Chapter 1: Introduction

Guy M. McKhann: Again a little bit of the basal ganglia and specifically focus
mostly on Parkinson’s disease, a little bit on dystonia and essential tremor,
conditions that we’re all able to treat with various surgical modalities right now. I’ll
talk about the treatments, the logic behind them, rationale, how they are working
as well as some degree of future directions.

Without reiterating too much of what Pietro was saying in terms of the basal
ganglia, he was explaining that the basal ganglia are a critical part of the control
circuitry that are important for preparing and adjusting for movements. Not
actually initiating them, but basically taking the amplitude and effort and scaling
that in relation to the execution of various tasks that we do.

I’m going to show a lot of some kind of loop diagram of the basal ganglia. It’s
nice, and it fits some of our models well, and a lot of it has been validated in
primate work and rodent work as well as some of the human work, but it’s very
important to realize that this is actually a slide made several years ago by Seth
Pullman, one of our movement physiologists. These are actually just some of the
known connections among these various structures. Keep in mind that it’s not
nearly as simple as we make it out to be with block diagrams, although what we
do talk about with them is actually valid for the most part in terms of what we see
in human physiology.




http://www.cumc.columbia.edu/dept/cme/neuroscience/                                    1
Chapter 2: Basic Anatomy and Pathways of the Basal Ganglia

Here’s a very schematic diagram of the basal ganglia. The basal ganglia are
functionally located strategically between the cerebral cortex and the output to
the brain stem and spinal cord down to the rest of the body. We talk about the
inputs from the cerebral cortex directly into the striatum, and the two output
nuclei of the basal ganglia are the internal segment of the globus pallidus and the
reticular segment of the substantia nigra. In between these—input striatum and
output nuclei—there’s two main pathways.

There’s the direct pathway which is mediated by or facilitated in excitatory
fashion by D1 dopamine receptors. The direct pathway is a monosynaptic
pathway directly from the striatum to the output nuclei. The indirect pathway, in
this case, is inhibited by dopamine by D2 receptors from the striatum to the
external segment of the globus pallidus to the subthalamic nucleus and then to
the output nuclei. This pathway is exciting the output nuclei, and the direct
pathway is inhibiting the output nuclei, so there’s a constant balance between
these two. In normal conditions, the direct and indirect pathways are together
exerting a slightly tonic inhibitory influence, because we don’t want to be all
moving all the time.

The system is under a slightly tonic inhibitory control, all of the time, and this
slightly tonic inhibition acts on the ventral thalamic nuclei here. The ventral
thalamic nuclei then synapse back onto the cerebral cortex, so the ventral
thalamus is always under slightly tonic inhibition. When there’s an imbalance
between these pathways, this is the main thought for what accounts for hypo- or
hyperkinetic movement disorders. Again, I’ll just reiterate the complexity of the
system, and the fact that it’s a lot more complex than this, but for model
purposes, and in terms of what we do surgically, it’s a lot clearer and easier to
just focus on the block diagrams.

What happens in Parkinson’s? In Parkinson’s disease we have a loss of
dopaminergic neurons in the pars compacta of the substantia nigra. This has
dual effects—it affects both the indirect and the direct pathways. We lose the
facilitation of the direct pathway and, at the same time, we lose the inhibition of
the indirect pathway, so we have more indirect pathway activation and less direct
pathway activation. What that means is that through the indirect pathway, the
subthalamic nucleus becomes hyperactive.

This is an excitatory drive onto the output nuclei, the internal segment of the
globus pallidus and the substantia nigra, pars reticulata, and both of these
structures become hyperactive in addition. When these are hyperactive, then
they are having an increased inhibition of the thalamus, which is then inhibiting

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                  2
the cortex. This pretty neatly explains the rigidity, the bradykinesia, the negative
symptoms, the hypokinetic symptoms of Parkinson’s. It does not clearly explain
the tremulous, the thalamic origination of the tremor that we see in Parkinson’s. It
does not clearly explain some of the midline balance features, particularly
primary imbalance of Parkinson’s that can be very medication-refractory.

If you look back at this model, the structures that are hyperactive in Parkinson’s
are the subthalamic nucleus and the output nuclei, the globus pallidus internal
segment and the substantia nigra. Either one of these are logical targets to
modulate to treat Parkinson’s. We’re not treating in a potentially curative fashion.
—this is a symptomatic treatment for Parkinson’s, trying to treat the symptoms of
the basal ganglia motor loop. We’re not treating cognitive or emotional aspects,
we’re not treating dementia, we’re treating the motor aspects of Parkinson’s. It’s
important to realize that, because if somebody’s Parkinson’s progresses over
time, even though their motor symptoms may be well-controlled, their disease
may still become more severe and more disabling over time.

So the reasons that we target the subthalamic nucleus and the globus pallidus,
as I just explained, because of the dopamine depletion there’s this series of
changes in the basal ganglia circuits that result in hyperactivity of the
subthalamic nucleus, increased excitatory drive onto the globus pallidus internal
segment and the substantia nigra pars reticulata which results in excess
inhibition of motor projections to cortex. So if we can release that
hyperexcitability in either the STN or the GPi, then we can release some of that
hyperinhibition of the cortex.

Chapter 3: Ablative Surgery and Brain Lesioning

This is the logic for why the subthalamic nucleus or the globus pallidus internal
segment are the targets that we use primarily for treating Parkinson’s disease.
There’s a number of ways in which you can treat those structures. Ablative
surgery, was the original thing that was done for the majority of several decades,
really, from basically the 40s, 50s, into the 60s. Thalamotomies, in particular, but
also pallidotomies were very common surgical procedures, particularly to control
tremor but also some of the other features.

Any place where you could explore this network, people went and tried to lesion
the brain symptomatically. Obviously, as you can imagine, there’s a couple of
problems. One is that if you have a side effect, it’s permanent, if you make a
lesion in the brain. Secondly the circuitry was not nearly as well understood then
as it is now. What happened was that dopamine came along, and dopamine was
developed as a drug. As that was developed, the movement disorder surgery
basically went away completely. Then that people realized over time, as I’ll talk

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                 3
about in a minute, that dopamine has problems in these patients that lead to
fluctuations in patients between their on and their off state and that over time,
patients need more dopamine, have a shorter period where the dopamine
actually works. Those are the patients that we’re looking at primarily as surgical
candidates.

Chapter 4: Transplantation

We almost never do ablative procedures anymore—ninety-nine percent of
movement disorder surgery in this country now is electrical stimulation based,
either in the thalamus, the internal segment of the globus pallidus, or the
subthalamic nucleus. Transplantation right now is basically at a research level.
There is one trial going on downtown at Cornell, there’s other trials with various
growth factors that are going on in various places around the country and around
the world. So far, the fetal transplant trials are two highly publicized trials. One
had the neurology aspects done here, and the transplants were done in
Colorado, about 5, 6 years ago and that trial showed 2 things. One, it showed
that the efficacy was not nearly as good as what we currently get with deep brain
stimulation. Two, that there were a lot of patients, particularly younger patients,
who ended up with uncontrollable side effects of the transplants, presumably due
to aberrant rewiring of the cell populations.

So there’s a lot of work that needs to be done in the transplantation field. It is
being done, but right now, none of it has equaled deep brain stimulation in its
ability to potentially improve quality of life. That’s the reason why we use
stimulation versus lesioning. Again, lesioning has decades of proven efficacy. It’s
lower cost because you don’t have to implant a system, there’s less risk of
infection because you don’t have to implant a system. You don’t have to manage
in terms of programming and fine-tuning an electrical system, and there’s no
batteries or mechanical breakdown. So some places in the world where for
instance it’s very difficult to follow up patients and monitor them still do a lot more
lesioning. But in the US, it’s very, very rare now.

The reasons why we stimulate is that: One, it’s less invasive. There’s less
morbidity, because we’re not actually having to make a lesion in the brain, so the
risk of hemorrhage and permanent morbidity is less. If you’re in a place where
you have a side effect that you don’t like, you just adjust it—change electrodes,
move the electrodes slightly—it’s not a permanent side effect. It’s easier and
safer to use bilaterally. It’s removable. The question of neuroprotection still
remains completely unsettled. In addition, if somebody comes up with a direct
transplant technology that’s efficacious, there’s no reason that it can’t be
combined into a patient who already has stimulators in place because the targets
will be different and the cell populations will be different.

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                   4
Chapter 5: Basics of Deep Brain Stimulation

Deep brain stimulation is really a misnomer in the way that it’s used, because
while we’re stimulating at a high frequency, in fact what we’re really doing is we
are most of the time inhibiting deep neural somatic structures. We use high
frequency stimulation, so in this case, this a graph of log of frequency here. Our
usual frequency that we stimulate at for tremors is 185 hertz, and that is actually
an inhibitory frequency to these neuronal structures. There’s all sorts of
hypotheses on why this is. The exact mechanisms of deep brain stimulation have
not been completely unraveled, although there’s a lot of experimental
hypotheses.

What are the ideal targets for deep brain stimulation, anywhere in the brain?
One, we have to be able to find a target. We have to be able to localise
stereotactically with a combination of MR imaging as well as merging with brain
atlases. Ideally, it’s nice if that target has some kind of physiological signature, so
when we get to the target, we can identify the target, either with a microelectrode
that allows us to identify the neuronal populations or a macroelectrode that
allows us to stimulate the target and find some kind of reproducible response.

Obviously, there has to be an acceptable regional morbidity. There’s some
evidence that possibly targets that have a higher basal firing frequency may be
more amenable to stimulation. There’s a number of targets that we currently use
for movement disorders. The VIM—this is the European nomenclature of the
thalamus which is the ventralis intermedialis, which is a nucleus of the thalamus
which kind of in the nomenclature that we learn in the US in medical school is
basically just in front of sensory thalamus, just in front of VPL and VPM.

If somebody really has tremor—so essential tremor, MS tremor, Parkinson’s
disease that is truly tremor-predominant, almost exclusively tremor with very little
bradykinesia or rigidity—then the VIM thalamus is the best target to use. For
other features of Parkinson’s, the subthalamic nucleus and the globus pallidus
are our targets. If somebody has significant tremor with it, the subthalamic
nucleus is better. We use the subthalamic nucleus for probably 85 to 90% of our
Parkinson’s patients. If somebody has a significant amount of parkinsonian
dystonia, then we use the globus pallidus and we also use the globus pallidus for
idiopathic dystonia.

Where these targets are in the brain relatively speaking? This is a sagittal
schematic and this is a coronal schematic, and basically, each of these targets,
what we do is we look at them relative to the anterior posterior commissure line.
All of these are standardly targeted once we identify the anterior commissure and

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                   5
the posterior commissure in somebody’s brain. We can then morph their brain to
standard stereotactic atlases and then we say, okay, this is where the
subthalamic nucleus should be. This is where the VIM thalamus should be. Then
we can either put in a microelectrode and record the neuronal populations and
make sure we’re in the right place or we can put in a macroelectrode, particularly
in the case of the thalamus, stimulate and make sure that we get the response
that we want and that we get the side effects that we would expect to get.

The indications for deep brain stimulation in Parkinson’s disease are: severe
symptoms with best medical management. So tremor, bradykinesia, rigidity,
medication-induced dyskinesias, dystonia. Again, as I said before, we’re treating
the motor circuits, so somebody who has significant dementia, significant bulbar
symptoms or autonomic dysfunction, these are not things that are going to
respond—first off, you have to make sure somebody actually has Parkinson’s
and not a Parkinson-plus syndrome. These are also things that do not generally
respond. Imbalance, if it’s due to the off-medication rigidity response, if it’s
primary imbalance, it’s present even when on Sinemet does not respond
particularly well.

Somebody has to have Parkinson’s in that they have to be dopa responsive, and
they need to have uncontrollable fluctuations between their on and off medication
state. So again, an ideal candidate: someone with typical Parkinson’s disease
with or without tremor, but tremor responds very well to stimulation. L-dopa
responsive. Somebody who has medication-induced dyskinesias, wearing-off
spells, so fluctuations between on-medication dyskinesia and off-medication
wearing-off periods with rigidity, bradykinesia and/or tremor, and good general
overall health.

Somebody who is significantly demented or has a significant psychological
history, postural instability, poor response to L-dopa, atypical Parkinson’s or
severe medical problems is someone who generally is going to be a much poorer
candidate to consider for stimulation. Again, just looking at the on and off
fluctuations. Most patients who come to us for surgical consideration are having
to take their medications every 2 to 4 hours and at least half of that cycle, half of
that 2 hour cycle or 4 hour cycle they’re either off medications and bradykinetic or
rigid or tremulous, or on medications but with so much medication-induced
dyskinesias that they are disabled on either side of the curve. They may have
half of the day where they have some kind of nondyskinetic state where they are
not so rigid that they can actually function and move. These are the patients who
are generally seen most of the time for surgery.




http://www.cumc.columbia.edu/dept/cme/neuroscience/                                 6
Chapter 6: Stimulation of the Sub-thalamic Nucleus

Going back to kind of a block diagram, in Parkinson’s disease, the subthalamic
nucleus is hyperactive, we’re putting an electrode in, we’re stimulating at high
frequency to decrease the output from the subthalamic nucleus. To turn the
output nuclei from basically increased neuronal firing to decreased neuronal
firing, to remove the inhibition off of the cortex. The same thing, you can draw the
same block diagram for the globus pallidus where you’re just, now in this case,
directly stimulating the globus pallidus, again to slow down the neuronal firing to
convert it from a structure with increased inhibition of the thalamus to decreased
inhibition of the thalamus.

This is just an example actually from some human neuronal recordings carried
out by Professor Benabid in Grenoble, who was actually the first person to
develop deep brain stimulation in humans just showing that when you stimulate
the subthalamic nucleus, you get a decreased firing frequency in the subthalamic
nucleus, which results in decreased firing in the output nuclei of the basal ganglia
and results in releasing the thalamus.

It’s really a team effort and this is just a small component of our team. Blair Ford
runs our Movement Disorder Surgery Center from the neurology aspect. We
have a whole host of skilled neurologists here who evaluate these patients. Seth
Pullman runs our electrophysiology in the operating room and probably our most
important person is Linda Winfield who actually tirelessly programs and
reprograms people over time and really is kind of the point person for a lot of the
program. Bob Goodman, one of my colleagues, and I do the surgery on these
patients.

This is kind of something that—and I can say this in all fairness because my
dad’s a neurologist—I think that in general, most neurologists generally think on
the phylogenetic curve that neurosurgeons are kind of somewhere back here,
somewhere between almost upright and not quite carrying the spear yet. I think
that's one of the things that’s nice in fields like epilepsy and movement disorder
and the things that I work, that that relationship has become a lot closer to a
partnership. I don’t think we’re quite, as neurosurgeons, allowed to sit at the
computer yet but we’re allowed to give up the jackhammers and kind of talk
about these concepts in partnership.

This is how we actually do the surgery. This is a woman who is in the holding
area in the operating room and what happens is a patient comes down. They’re
awake, we give them four shots of numbing medication in the head and we put
this stereotactic ring into the skull. The skull, essentially, has almost no pain
fibers, so once the scalp is numb, it doesn’t hurt. You give four shots of numbing

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                    7
medication—this is the worst part of the day. Sometimes we sedate patients if
they’re really anxious. When you ask patients the two things they don’t like about
surgery: getting the frame put on because for about 5 or 10 minutes, it really,
really hurts because there is so much pressure to it. It’s about 60 pounds of
pressure force. The other thing is that Parkinson’s patients hate being off, and for
surgery, they have to be off their medications the entire day, so we try to sedate
them as much as possible in the operating room, so they are only awake for the
few minutes that we need to test them.

Once we put on the frame, we get an MRI scan and we build out our 3-
dimensional work station, and we’re looking for the subthalamic nucleus. The
subthalamic nucleus is located about 7 cm from the cortex. This is the blow up of
a coronal scan. This is an axial plane scan and the subthalamic nucleus is about
the size of an almond. In the dorsal-ventral direction, it’s about 5 to 6 mm thick,
so it’s like an almond that we’re aiming for, we’re aiming for just basically
posterior to the midline of that. We expect to get 4.5 to 5 mm of subthalamic
nucleus.

The whole reason that we go through all the electronics in the OR is that, say we
get 2 mm, we have to be able to figure out, are we too far in the front, the back,
inside or outside on the subthalamic nucleus? We make an educated guess on
which direction we’re off, move the electrode slightly and retest it. That’s called
microelectrode recording, which is when we pass a microelectrode down. We
listen to all the target structures along the way and basically identify each
structure’s receptive field and firing patterns. This is used to optimize the
electrode placement. It increases the OR time. It results in a higher number of
brain penetrations and the longer that the head is actually open, the more the
brain can shift from getting air in the head and things like that. So for the thalamic
stimulation, because the thalamus has a very, very readily and reproducibly
identifiable sensory response, because we’re right in front of the sensory
thalamus, then we actually do not do microelectrode recording. Some places do.
We do not.

With the single-unit recording, as I said, we pass a microelectrode down through
the gray matter and white matter structures and what we do is we look at the
neuronal firing pattern on each one. This is a patient in the operating room, this is
the electronic set-up which we actually have been using, I think, for 14 years and
now a couple of turnkey systems and hopefully we’re going to be switching to
one of those rapidly—it’s sort of jerry-rigging this together with duct tape week to
week, but it’s actually a highly accurate system, with basically sub millimeter
spatial precision.




http://www.cumc.columbia.edu/dept/cme/neuroscience/                                  8
This is a patient who is heavily sedated but not intubated, so the frame is
attached to the head. Everything is sterile above them, not sterile down below, so
they can be tested by the physiology team. Again, just another view based on the
whole electronic set up. Sterile above the head. The anesthesiologists are over
here. The patient’s got just a small basically a little cannula giving them extra
oxygen. As soon as we turn off the sedation, in 10 minutes they are able to wake
up, they are able to follow commands, we are able to test their motor function.

So as we go down through the basal ganglia, we come through the caudate. We
come through the thalamus. So each structure that we go through has a
particular neuronal firing frequency and pattern that you can see on the
oscilloscope and you can hear in the operating room. So as we go through the
caudate, the caudate has a very low frequency of firing. We actually most of the
time don’t start recording until we get into the thalamus. When we get to the
thalamus, it’s an irregular activity, slightly higher frequency, anywhere from 10 to
80 hertz and oftentimes, in the thalamus if somebody has tremor, there can be
direct tremor-related cells that we can pick up. We then go through the zona
inserta, white matter between the thalamus and the subthalamic nucleus. Again,
it’s going to be without neuronal activity in white matter.

We get to the subthalamic nucleus. The subthalamic nucleus has a very
oscillatory pattern, 20 to 50 hertz, sometimes with burst discharges and very
rhythmic discharges that we hear and it’s very characteristic. You can actually
hear it very clearly on the oscilloscope in the operating room. Also, in the
subthalamic nucleus when somebody has tremor, there are actually tremor-
responsive cells. If we let somebody lighten up the anesthesia a little bit and they
start shaking a little bit or there are limb-responsive cells just in terms of the limb
proprioception.

When we get beyond the subthalamic nucleus, then we end up in the substantia
nigra and again, a characteristic pattern there, so that when we are all the way
through, we basically construct a map of where we think we are, based on what’s
the size of the zona inserta, how much subthalamic nucleus are we getting,
what’s the gap here between the subthalamic nucleus and the substantia nigra,
so we can say exactly where we think we are. Usually if it’s the first site in the
OR, we do a second pass to confirm that. If it’s the second site we already have
one electrode in successfully, we will place it at the mirror location to the first site.

Then what we do is we put the permanent electrode in and we stimulate. We
hook it up to an external battery pack and we stimulate and make sure that the
patient has a response that we expect to see, so that we’re able to suppress their
tremor, decrease their rigidity or bradykinesia. Usually this is something that
happens basically immediately in the operating room and it’s a great thing, you

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                     9
know, when we have medical students every week, we always call them in for
that part because it’s, you know, a dramatic thing. It’s good for the patients
because if they are struggling through the OR and they can see that the
stimulation stops their tremor or makes them less rigid, then they kind of get the
gumption up to make it through the second side of the surgery. Postoperatively,
everybody gets an MRI scan to make sure that the electrode is targeted exactly
where we think it is. Again, we can build it on a 3-dimensional work station and
make sure that’s it in exactly where we think the STN should be, that there are no
surprises.

Chapter 7: Medtronic's Stimulation System and Implantation Surgery

This is the actual implantable system. This is a battery pack. The Medtronic
system is currently the only FDA-approved system. There’s another company
that has a system that is currently in trial called ANS, advanced neuromodulation
out of Dallas. This technology came directly out of the pacemaker technology.
This is implanted like a pacemaker below the clavicle. There’s a lead extension
that goes up behind the ear and then electrode itself goes into the brain.

The electrode that goes in has four contacts, so any of these contacts can be
used. They can be used either in a monopolar configuration where a single
contact is used relative to the pulse generator. That gives you a sphere of
stimulation where you can stimulate between two contacts, which results in a
cylindrical stimulation. It’s going to have a greater dorsal ventral extent but a
smaller width to the area of stimulation. It’s trial and error—patients come back
several times in the first six months for fine tuning of their stimulation to try to
maximize their benefit and minimize any sort of side effects. This is actually the
second phase of the surgery. They come back a week to two after the first
surgery. This is a patient who is actually asleep. The electrode gets put in on top
of the head. They get an incision behind the ear and this is just the connection.
We just drill a trough in the bones so the connection is recessed so it doesn’t
bother people, and then the battery pack goes underneath the clavicle. That’s the
full extent of gore that anyone’s going to have to witness this morning.

The overall stimulation results are that all of the cardinal features of Parkinson’s
disease are improved. What we’re really doing here is that you’re going to make
a patient significantly better than they are in their best medication-on state. If
somebody does well most of the day on medication and doesn’t have a lot of
fluctuations, we’re not going to make them better, and so that’s why we wouldn’t
put stimulators in. But this is somebody who has a lot of dyskinesias as I said, or
spends a significant amount of time off medication. The off-medication state, in
patients who are properly selected, improves by generally 60 to 75%. Their on-
medication state may improve by 10% but really, very modestly. In addition, most

http://www.cumc.columbia.edu/dept/cme/neuroscience/                               10
patients are able to decrease their medications. On average, it’s about 50% Very
rarely, somebody comes off medication. Sometimes it’s only 20 or 30% Their
dyskinesias tend to improve both from a combination of a direct stimulation
effect, but also more importantly, from being able to decrease their medication
dose over time.

I’ll skip the fine details of the programming. Bilateral placement is what we
normally do. In general, probably 80% of our patients that we see are fairly
symmetric or asymmetrically involved, severe enough on both sides for bilateral
stimulation and 80% of the time we’re able to put both electrodes in at a single
operative setting. If somebody develops confusion in the operating room, we
actually stop and come back in a week or two because that’s a sign that if we
proceeded on, that confusion would last longer, and I’ll touch on that in a minute.

That’s Jack Nicholson from A Few Good Men, that scene with Tom Cruise where
he talks about, you know, being able to be truthful. It’s very important. These
patients are mostly 50, 60, 70s, a lot of them are fairly medically frail. You need
to be very realistic, and tell people about what to expect. This is not a totally
benign procedure. If you look here, the average age is about 60. Parkinson
duration—most patients have had Parkinson’s for a mean of 12 years. As I said,
most patients end up having a bilateral procedure. You’re 4 hours in the OR with
a frame on, sedated. It’s a significant period of time in the OR, and then coming
in for a second procedure under general anesthesia 1 to 2 weeks later. We end
up making on average, 1.8 microelectrode passes per side, but up to 7 electrode
passes have been done to adequately identify the subthalamic nucleus.

Chapter 8: Evidence-Based Changes to Stimulation Practices

This is from 100 patients that were pooled together about 4 or 5 years ago, and
it's actually resulted in somewhat how we changed our practice. You can see, in
this group, this is actually our first 100 patients, 50 of which were part of the
original FDA-approval pilot study. When the FDA first approved this, patients in
general were even more severe than mostly what we are seeing now, and if you
have patients who are more severely or are more on the borderline cognitively,
you’re much more at risk at developing significant confusion. As part of this,
everybody had a bilateral stimulation.

Now we in general are down to less than 5% confusion, because as I said, if
somebody starts to get confused, we stop the surgery. We don’t open up the
other side of the head. This is probably a combination of anesthesia, air in the
head as well as making multiple trajectory passes—even though the cannula we
put into the brain is less than 2 millimeter in size, we’re still making a 2 millimeter
cannula pass through the frontal lobes.

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                  11
In addition, about 5% of patients end up having to have a revision for one reason
or the other. We’ve had a couple of patients have seizures, a couple of patients
with wound hematomas. Again, they are mechanical devices—they can erode
through the skin over time. We’ve had one subdural hematoma, one intracerebral
hematoma. We’ve had one patient in our last, I guess, 250 now who had major,
major, major—not irreversible, the patient got a lot better—but a significant
morbidity when the patient never got back to close to their baseline.

Generally speaking, I tell every patient it’s a 1 to 2% risk. Our numbers are less
than 1% right now, but it is a 1 to 2% risk statistically. For thalamic DBS in
general, we get about an 80% reduction in contralateral tremor in tremor-
predominant Parkinson’s or essential tremor. There is no beneficial effect on gait,
sometimes a very mild improvement in parkinsonian bradykinesia or rigidity. If
you do bilateral thalamic stimulation, there’s a significant effect of potentially
softening the speech, or potentially some impact on imbalance. The globus
pallidus, as I said, we use for dystonic Parkinson’s or for primary generalized
dystonia.

What’s happening is there has really been somewhat of an evolution of
indications over time. Initially, surgery was considered when medical treatment
completely failed and the patient had to be severely disabled. The problem is, if
someone is really severely disabled and you improve them by 60 or 70%, that
may not make that much difference in quality of life than if they are moderately
disabled and you improve them by 50 or 60 or 70%. We are actually seeing a lot
more patients now that are moderately disabled, not severely disabled, and we’re
actually having a lot more younger patients with Parkinson’s who actually come
in now, saying "Okay. Maybe I’m not moderately disabled. But maybe for me it’s
moderately affecting my quality of life and I don’t want to wait until I’m moderately
disabled." Sometimes, those patients, if you have educated discussions with
them and you see them 2 or 3 times and go through it, they elect that they want
to have deep brain stimulation at an earlier phase.

Chapter 9: DBS as a Neuroprotective Process

As I said, the mechanism of action is largely experimental based on rodent work
and still being sorted through, but there’s all sorts of things that deep brain
stimulation can do. It can cause a depolarization block, there can be ion channel
inactivation, there could be activation of white matter en passage GABAergic
axon terminals. There’s a number of mechanisms that all, or multiple of which,
may be at play in deep brain stimulation.




http://www.cumc.columbia.edu/dept/cme/neuroscience/                               12
There’s still a big question about whether or not DBS could be neuroprotective.
The logic here is that if you have a subthalamic nucleus that’s a glutamatergic
nucleus that’s basically hyperactive, it is essentially causing glutamate toxicity.
There’s one interesting rodent study from out of France. Here, what they did was
they used a 6-Hydroxydopamine model of Parkinson’s and when you use the 6-
Hydroxydopamine model and you look overall at dopaminergic terminal staining,
you can see that there’s a significant loss of the dopaminergic neurons in this
model with a loss of their striatal terminals.

In this case they use a lesion, not stimulation—they lesion the subthalamic
nucleus first, so you’re making it so the subthalamic nucleus cannot become
hyperactive. You see that despite administering 6-Hydroxydopa you have
preservation of the dopaminergic neurons in the pars compacta and their
terminals. There’s a number of different primate experiments going on now in a
couple of labs around the world as well as rodent experiments, trying to look at
the hypothesis, because obviously, there would a lot more rationale for
something if it turns out it could be potentially neuroprotective.

There’s a whole lot we don’t know about this. Even simple things like "What’s the
ideal electrode target?" There is some evidence, in fact, that some of the most
effective stimulation may not be even be the subthalamic nucleus. It may be
white matter pathways, the axon pathways just above the subthalamic nucleus.
Again, which patients will respond the best is still being worked through, as are
optimal stimulation parameters. Are there better targets? When a Parkinson’s
patient has axial postural instability it does not respond to subthalamic
stimulation. There is some evidence now, there’s a couple of papers in the last 6
months, mostly out of Toronto, showing that potentially targeting the peduncular
pontine nucleus in conjunction with the STN may improve the other cardinal
features as well as the balance features. Again, there are other issues in terms of
how all this works into the model in terms of the impact on tremor.

Just one concept that again is being looked at at a rodent level right now is
combining electrical stimulation for symptomatic relief, as well as seeing whether
or not that electrical stimulation can directly essentially electroporate genes and
increase gene transfer at the same time. The model would be in the future to
potentially think about a combined therapy where you use stimulation for
symptomatic benefit and at the same time, electroporate some kind of genetic
situation to allow you to improve the overall condition and not be so stimulation-
dependent over time.

One thing about the current stimulating systems is that the battery life is about 3
to 5 years, so every 3 to 5 years patients have to come back for new batteries.
That’s about a 30-minute outpatient procedure. Hopefully with a little luck through

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                13
the companies in the FDA within 3 to 5 years from now there will be rechargeable
batteries and we won’t have to keep changing people’s batteries all the time.

Chapter 10: Dystonia

Just moving on quickly to dystonia in the last 5 minutes or so. I know that Pietro
covered dystonia a little bit. This is the original description of a childhood
syndrome of twisted posture, muscle spasms, bizarre gait, rapid rhythmic jerking
movements and progression to sustain postures and deformities. Dystonia can
be classified based on age of onset, based on the overall distribution of which
muscle groups are involved, or based on the etiology of the dystonia itself.

Primary dystonia is genetic or idiopathic dystonia, and then a number of
secondary dystonias occur. We’re much better at treating primary dystonias with
stimulation than we are at treating secondary dystonias. Primary dystonias get
tested for the DYT1 mutation, which is a GAG deletion in the torsion-A ATP-
binding protein on chromosome 9. There’s usually a childhood onset of the
condition, and these are the patients that actually have the best response to
deep brain stimulation. There’s a number of potential treatments for dystonia, a
number of medical treatments and we don’t generally consider people for surgery
again unless they fail standard medical therapy. And again, there’s a number of
different targets that have been tried traditionally, pallidotomy or thalamotomy but
now the most common site that is being stimulation is the internal segment
globus pallidus.

The dystonia model is not nearly as clean as the Parkinson’s model in terms of
understanding what happens. Lesioning the GPi in some patients can relieve
symptoms, but also there are situations where it can actually induce dystonia.
One of the things that is unusual about dystonia in Parkinson’s is that you put in
an electrode, you turn on the electrode—if somebody has tremor, they stop
immediately. If they have parkinsonian rigidity, they loosen very, very rapidly.
You can bring somebody in the office, turn off their stimulator and their symptoms
will come back in 30 seconds or a minute, usually very rapidly.

Dystonia, when we put the deep brain stimulating system in, oftentimes it takes
many months, sometimes 3 to 6 months, for somebody to have maximal benefit.
So any model of how basal ganglia manipulation is impacting dystonia has to
somehow incorporate the fact that there are manipulations taking 3 to 6 months
to have full efficacy. DBS for dystonia is now FDA-approved, it has been for
several years now. The series are growing larger now and in DYT1, there can be
up to as high as 90% improvement. In generalized dystonia, however, it can be
as low as 20%, depending on whether it’s secondary dystonia or even in some
primary dystonia patients.

http://www.cumc.columbia.edu/dept/cme/neuroscience/                               14
It’s also being used for cervical dystonia and tardive dystonia to some degree,
but as I said, it’s much more variable and less successful in results in secondary
dystonia. This is one paper from a few years ago, just looking at different types of
dystonia. The white here is DYT1 genetic dystonia, primary dystonia. The gray is
primary dystonia, presumed primary dystonia but negative DYT1 mutation. This
last is secondary dystonia. If you look at the overall degree of improvement, you
can see that the majority of studies with primary dystonia show a significant
improvement in these patients, and these are the patients that do best.

Here is a patient with primary dystonia that was implanted now probably 7 years
ago by my partner, Bob Goodman. This is pre-implantation—the dystonia is so
bad on his dominant right hand that he has to write with his left hand. He has to
support his trunk as he walks, has to support his head because he has got so
much lateral collis in the neck with the dystonia he ends up essentially having to
hold his neck up when he walks. You can see that obviously this person is
cognitively in very good shape but you very motorically disabled from this primary
DYT1 genetic dystonia.

You will see in a second the patient afterward. He’s still growing back some hair
from his intraoperative cosmetic hair treatment and you can see here though he
still has straightening of the right arm, he’s back to being able to write with his
dominant right hand, and he’s actually able to hold a much straighter posture
now in terms of just sitting up. You can see now when he walks—again it’s not
perfect. He still has a little bit of pulling to the right side, but again, a fairly
dramatic quality of life improvement for him with the stimulator system on.

Essential tremor is another condition that can respond very well to deep brain
stimulation. Again, similar situation—medications first. If the patient is
medication-refractory and significantly disabled by their tremor, then we place
electrodes into the VIM thalamus. Electrode singular or plural depending on
whether the patient has bilateral or unilateral predominant tremor. Here’s a
normal spiral. Here’s somebody with moderate ET and severe ET. Everyone gets
a computerized spiral analysis before and after so we can more objectively
quantify their response to therapy.

This is just a study from several years ago, The New England Journal of
Medicine, looking at thalamic stimulation for tremor, either tremor-predominant
Parkinson’s or essential tremor. This is exactly comparing thalamotomy versus
thalamic stimulation. There was a very high degree of tremor suppression in both
groups, less adverse effects with stimulation and more overall functional
improvement in somebody’s quality of life with stimulation than with thalamotomy.



http://www.cumc.columbia.edu/dept/cme/neuroscience/                                15
That’s because you can modify the stimulation and have less overall side effects
with stimulation than with bilateral thalamotomy.

Here’s a patient, that we implanted about 6 years ago, with essential tremor.
Here he is at his baseline—you can see he’s got a significant action tremor with
his rest tremor. Here he is trying to draw a spiral, so fairly severe ET. He is left-
hand dominant here but he has it on both sides. Here he is trying to drink a cup
of water. This could be incredibly disabling—people just don’t even want to go
out to a restaurant because they can’t pick up a glass at the table. You will see
here that he says, "Oh, boy!" because when the electrical stimulation first goes
on he gets a quick electric sensory jolt that rapidly weans off. Now you can see,
he’s not perfect. If you watch closely, he’s got a slight degree of tremulousness
here that will come out in a second when he’s pouring. It’s not complete
resolution, but yet, in terms of his quality of life he’s able to do things like just eat,
brush his teeth, sit at the table and drink something. His quality of life is really
markedly improved.

These are the patients who sometimes will get to close to a heavy refrigerator
magnet or get hand-checked by security by a magnetic system at the airport and
their system goes off and all of a sudden, they think that their batteries have worn
out and you get this frantic call. All these patients now, we actually give them an
access review device so they can check their system. It tells them if they are on
or off and allows them to turn themselves on or off. They can’t fine tune it but
they can turn themselves on or off.

Chapter 11: Other Uses of DBS

In the last 2 minutes, a number of other disorders. I don’t have time to cover
these today, but we kind of focused on movement disorders. Again, I’ve touched
on Parkinson’s, essential tremor, dystonia—I haven’t talked about Tourette’s at
all, but Tourette’s is also being treated with deep brain stimulation. The concept
of deep brain stimulation can be applied anywhere where there is a focal target
that you want to modulate the neuronal activity of to change a structure function
relationship in the brain.

There’s actually a deep brain stimulation trial out of Medtronic for medically
refractory epilepsy. There’s actually two trials. There’s a trial using the anterior
nucleus of the thalamus that’s just been finished and the patients are in follow up.
We’ll be getting reports on that efficacy in the next year. Also the subthalamic
nucleus is also being tried in Europe as a trial in epilepsy.

Psychosurgery is a rapidly evolving field in deep brain stimulation. Pain has been
an area of interest for a long time. Potentially hypothalamic stimulation for

http://www.cumc.columbia.edu/dept/cme/neuroscience/                                     16
obesity—right now it is at the primate and rodent level. For psychosurgery, both
depression and OCD are conditions that are being targeted. The really important
thing about any one of these sorts of endeavors is that it really needs to be an
effort that’s really driven out of a collaboration between the primary field caring
for these patients and the people who are delivering the therapy, which is us.

I mean, the real danger—and I can say this easily being a neurosurgeon—the
real danger in a field like depression or OCD is that somebody is going to break
apart from doing this in a responsible, structured way and just say, "Okay, you
know, we can put electrodes in depressed people or people with OCD and
potentially help them." But these are incredibly complicated, difficult patients and
just like in movement disorder, the surgical therapy of movement disorder has
been driven from the movement disorder neurologist to surgeons. Psychosurgery
really has to work the same way—these are patients, that for depression or for
obsessive compulsive disorder, need to be primarily managed and primarily this
whole academic interest has to evolve out of the psychiatry world if it’s going to
have a chance of actually both just medically as well as socially being accepted.

I don’t have time for more than one minute to touch on depression. But this is a
fascinating area where basically what happened is MRI and PET scanning
identify a number of regions in the brain, and in particular the subgenual part of
the cingulate gyrus which is Brodman area 25, was shown by Helen Mayberg
initially in PET scanning and then in functional MRI to be hyperactive in primary
generalized depression and primary major depression. So the thought was,
"Okay. Here’s a hyperactive structure we can readily target in the brain. So let’s
try it as a pilot.”

This is also an area when you look at successful treatment of depression with
SSRIs or electroconvulsive therapy, and now with other sorts of electrical
stimulations, that these areas normalize with treatment. So the thought was, if
you had somebody who was refractory to ECT, refractory to medications, been
through everything you have, who is a young enough and healthy enough
patient, offer them deep brain stimulation in the subgenual cingulate bilaterally for
depression. Helen Mayberg who was at Toronto and is now at Emory did a pilot
study together with the neurosurgeons in Toronto in six patients. They did not
use microelectrode guidance, just anatomically-based electrodes. These are the
descriptions from the patients themselves: kind of a disappearance of void, a
connectedness, increased sharpness.

One thought was, well, maybe you are replacing depression with basically
stimulation like a nucleus-accumbens type phemonenon. They found it’s really
not that—that you’re not replacing depression with mania. You are actually
reversing the depression without these patients becoming hypomanic. This is

http://www.cumc.columbia.edu/dept/cme/neuroscience/                               17
their initial result from the 6 patients published in Neuron. They are now up to 13
patients in a pilot and they’ve got a 50% sustained response rate with about a
third of patients who actually were in remission. For medically refractory
depression that has failed ECT, this is actually dramatically better than anything
else that has come along. So now a multicenter trial that’s just getting started.
Actually, we’re going to a planning meeting for that in about two weeks down in
Texas. Again, I think that this is being done very slowly and hopefully will remain
very responsibly so that a lot of the bad things that happened in the past history
of psychosurgery don’t have an opportunity to be repeated going forward. And I
think that’s it. Thank you. Yeah?

Question #1: Does deep brain stimulation work with stuttering?

Guy M. McKhann: Stuttering. I don’t know anybody that has tried it for
stuttering. I actually had a patient who . . .




http://www.cumc.columbia.edu/dept/cme/neuroscience/                               18

								
To top