WEBVTT
00:00:18.130 --> 00:00:21.370
It's June 16th, and we have a lot to talk about.
00:00:22.240 --> 00:00:27.030
For years, the standard medical advice for someone diagnosed with multiple sclerosis
00:00:27.220 --> 00:00:32.130
was simple, cautious, and as we now know, incredibly limiting.
00:00:33.050 --> 00:00:37.710
Take it easy, don't overexert yourself, and get plenty of rest.
00:00:38.600 --> 00:00:42.650
Physical therapy was often treated as a reactive last resort measure,
00:00:43.020 --> 00:00:46.820
something to help you adapt to a loss of function after the fact.
00:00:47.500 --> 00:00:53.300
But today, like almost every other aspect of MS care, neurological rehabilitation
00:00:53.300 --> 00:00:55.060
is being entirely rewritten.
00:00:55.550 --> 00:01:01.060
It's no longer about just managing decline. It's about leveraging emerging technologies
00:01:01.060 --> 00:01:03.350
to proactively build resilience,
00:01:03.680 --> 00:01:08.780
retrain the brain, and adopt cutting-edge exercise science to fundamentally
00:01:08.780 --> 00:01:11.630
change the trajectory of life with MS.
00:01:12.390 --> 00:01:17.240
This week, we're diving deep into how the landscape of MS rehabilitation has
00:01:17.240 --> 00:01:23.190
evolved from a passive, cautious approach into a dynamic, science-driven,
00:01:23.190 --> 00:01:25.690
evidence-based approach to patient care.
00:01:26.240 --> 00:01:30.330
And there's no better way to illustrate this paradigm shift than talking with
00:01:30.330 --> 00:01:34.400
my three guests who are at the forefront of this research and practice,
00:01:34.950 --> 00:01:39.780
Dr. Brian Sandroff, Dr. Brad Willingham and Dr. Laura Rice.
00:01:40.440 --> 00:01:45.310
We're going to discuss how exercise serves as a powerful tool for both cognitive
00:01:45.310 --> 00:01:51.120
health and physical health, how AI is impacting MS rehabilitation by putting
00:01:51.120 --> 00:01:52.640
the patient at its center,
00:01:53.340 --> 00:01:58.360
and how modern physical rehab strategies are being designed to accommodate people
00:01:58.360 --> 00:02:02.680
living with MS at every level of ability and mobility.
00:02:03.380 --> 00:02:05.660
You're about to hear three conversations
00:02:05.660 --> 00:02:09.510
about breaking old boundaries and redefining what's possible.
00:02:10.480 --> 00:02:15.390
Dr. Brian Sandroff is the director of the Exercise Neurorehabilitation Research
00:02:15.390 --> 00:02:20.960
Laboratory at Kessler Foundation's Center for Neuropsychology and Neuroscience Research.
00:02:21.710 --> 00:02:25.850
I caught up with Dr. Sandroff at the recent Consortium of MS Center's annual
00:02:25.850 --> 00:02:32.530
meeting where his presentation focused on whether exercise can be used to treat cognitive impairment.
00:02:32.960 --> 00:02:36.400
In a moment, we'll hear my conversation with Dr. Brian Sandroff.
00:02:41.710 --> 00:02:46.940
Dr. Brian Sandroff directs the Exercise Neurorehabilitation Research Laboratory
00:02:46.940 --> 00:02:51.220
at Kessler Foundation's Center for Neuropsychology and Neuroscience Research,
00:02:51.570 --> 00:02:53.730
where he also serves as Assistant Director.
00:02:54.410 --> 00:02:59.200
The lab focuses on creating and evaluating exercise training programs to enhance
00:02:59.200 --> 00:03:03.480
the physical, cognitive, and brain health of people with multiple sclerosis.
00:03:04.010 --> 00:03:08.110
And it's always my pleasure to meet up with Dr. Sandroff at CMSC. It's good to see you.
00:03:08.210 --> 00:03:10.450
Great to see you too, Jon. Thanks for talking with me today.
00:03:11.080 --> 00:03:16.440
You know, historically, cognitive rehabilitation focused on compensatory strategies
00:03:16.780 --> 00:03:19.150
like using planners or memory aids.
00:03:19.730 --> 00:03:25.660
Your work is helping to prove that exercise can actually target the underlying neuropathways.
00:03:26.080 --> 00:03:32.180
Your presentation today directly asked, can exercise be used to treat cognitive impairment?
00:03:32.840 --> 00:03:38.980
You know, for a long time, exercise was viewed as a way to improve general health or mobility in MS.
00:03:39.450 --> 00:03:44.910
How does framing exercise as a direct treatment for cognitive impairment change
00:03:44.910 --> 00:03:47.540
how a neurologist might prescribe it in the clinic?
00:03:48.110 --> 00:03:52.520
So that is a wonderful question, because that's the question that really drives
00:03:52.520 --> 00:03:59.800
our research, is if exercise can demonstrably change the brain,
00:04:00.130 --> 00:04:02.630
we know that cognition comes largely from the brain.
00:04:03.050 --> 00:04:06.240
And if we can develop our programs in such a way that target the brain,
00:04:06.240 --> 00:04:10.570
that can make changes in networks in the brain that are so important for cognitive
00:04:10.570 --> 00:04:12.110
performance and thinking performance,
00:04:12.570 --> 00:04:17.540
we've essentially created a target for improving cognition and everyday life
00:04:17.540 --> 00:04:19.180
outcomes in people with MS.
00:04:19.800 --> 00:04:25.030
So if we can provide the biological evidence to support exercise benefits on
00:04:25.030 --> 00:04:26.610
the brain and central nervous system,
00:04:27.190 --> 00:04:31.960
and then we can develop evidence that shows that that extends into cognitive
00:04:31.960 --> 00:04:35.820
tests, and we have evidence where it helps people with MS in their daily life,
00:04:36.210 --> 00:04:39.230
then hopefully that evidence can get integrated into clinics where neurologists
00:04:39.230 --> 00:04:44.880
and primary care providers can recommend exercise as an approach to help cognition
00:04:44.880 --> 00:04:46.740
in people who are living with the disease.
00:04:47.760 --> 00:04:52.730
When we look at standard disease-modifying therapies, they're highly effective
00:04:52.730 --> 00:04:59.530
at reducing relapses, but have a limited direct impact on repairing pre-existing cognitive damage.
00:05:00.290 --> 00:05:06.630
Do you see exercise as a necessary synergistic therapy alongside DMTs to protect
00:05:06.630 --> 00:05:08.290
and restore that cognitive reserve?
00:05:08.990 --> 00:05:12.480
That is a trial I would absolutely love to do and one that's on our radar.
00:05:13.610 --> 00:05:17.910
We've argued in the past that exercise could have potential disease-modifying
00:05:17.910 --> 00:05:22.710
properties associated with its biological benefits, but we never actually directly tested it
00:05:23.440 --> 00:05:26.300
either against disease-modifying therapies or, more importantly,
00:05:26.300 --> 00:05:28.640
in concert with disease-modifying therapies.
00:05:29.070 --> 00:05:33.040
As you're right, the DMTs are not designed to improve cognition,
00:05:33.280 --> 00:05:37.280
but they've been wonderfully effective at improving many other aspects.
00:05:37.700 --> 00:05:41.900
So it is something that we wonder often and something we would love to test.
00:05:42.820 --> 00:05:47.600
We'll keep our eye on that then, for sure. In your targeted randomized control
00:05:47.600 --> 00:05:52.510
trials, you've looked specifically at supervised treadmill walking and its impact
00:05:52.510 --> 00:05:53.940
on verbal learning and memory.
00:05:54.720 --> 00:05:59.790
Why does aerobic walking specifically seem to have such a profound impact on
00:05:59.790 --> 00:06:04.370
learning and memory compared to low-intensity resistance or stretching?
00:06:04.870 --> 00:06:08.670
So our research not only is focused on learning and memory, but also on cognitive
00:06:08.670 --> 00:06:12.350
processing speed as the primary cognitive problem in MS.
00:06:12.760 --> 00:06:16.840
And what we think is going on is that exercise involves all these different
00:06:18.130 --> 00:06:22.860
demands that cause the brain to work really hard to be able to continue to exercise.
00:06:23.220 --> 00:06:28.670
And by continuing to exercise those brain areas, the brain becomes more efficient
00:06:28.670 --> 00:06:33.430
at communicating, uh, within itself in terms of long-term communication from
00:06:33.430 --> 00:06:35.720
the front to the back, from the right to the left hemisphere.
00:06:36.150 --> 00:06:40.950
And by becoming more efficient, the brain then, uh,
00:06:41.890 --> 00:06:46.160
the brain becomes more efficient. And that increased efficiency results in downstream
00:06:46.160 --> 00:06:50.190
improvements in cognition and other outcomes, including learning and memory,
00:06:50.520 --> 00:06:55.430
or cognitive processing speed that rely upon those cognitive, the brain outcomes,
00:06:55.820 --> 00:07:00.230
where that increased efficiency not just supports exercise, it also supports
00:07:00.460 --> 00:07:01.970
cognition or thinking performance.
00:07:03.120 --> 00:07:08.810
Your pilot neuroimaging data showed that 12 weeks of treadmill training actually
00:07:08.810 --> 00:07:13.980
helped preserve hippocampal volume in patients with relapsing-remitting MS.
00:07:14.720 --> 00:07:19.340
Can you explain the mechanisms at play there? Is it driven by increased blood
00:07:19.340 --> 00:07:24.910
flow, neuroplasticity, or perhaps neuroprotective factors like BDNF?
00:07:25.310 --> 00:07:31.400
So, our research suggests that it has to do with increased connectivity in the
00:07:31.400 --> 00:07:35.570
brain as a primary way to potentially preserve hippocampal volume.
00:07:36.310 --> 00:07:40.270
And one of the interesting findings of that particular study is that the results
00:07:40.270 --> 00:07:43.330
are consistent with those from animal research done in
00:07:43.670 --> 00:07:50.330
1992, for example, where environmental enrichment in mice was associated with
00:07:50.330 --> 00:07:52.640
increased neurogenesis in the hippocampus.
00:07:52.960 --> 00:07:56.520
And exercise is certainly an enriched environment where when people are walking
00:07:56.520 --> 00:07:59.610
on a treadmill, that's not something you do on your everyday,
00:07:59.880 --> 00:08:04.830
you know, life and your everyday existence where that's really making your brain work hard.
00:08:05.200 --> 00:08:09.630
And the hippocampus is a region that really responds to increased environment.
00:08:09.630 --> 00:08:15.110
So environmental enrichment. So one of our going hypotheses was that by continuing
00:08:15.110 --> 00:08:20.340
to exercise three days a week for three months, we really had that stimulation going on.
00:08:21.450 --> 00:08:25.190
We're not quite sure if it's blood flow or we're not quite sure if it's molecular
00:08:25.190 --> 00:08:29.360
cellular mechanisms like BDNF or IGF-1, for example.
00:08:29.490 --> 00:08:35.800
But we think that those are more support mechanisms with the primary mechanism
00:08:35.800 --> 00:08:39.260
being all that work that your brain has to do to be able to continue to exercise.
00:08:39.620 --> 00:08:41.170
And then that causes adaptations.
00:08:42.000 --> 00:08:47.800
Thalamic and hippocampal atrophy can occur very early in the MSD's disease process, right?
00:08:48.230 --> 00:08:53.240
Does your research suggest that exercise is most effective as a preventative
00:08:53.240 --> 00:08:58.290
strategy early in the diagnosis, or can it still restore volume and functional
00:08:58.290 --> 00:09:01.980
connectivity in patients with longstanding progressiveness?
00:09:02.280 --> 00:09:05.340
So that's been one of the million-dollar questions in our lab as well.
00:09:06.100 --> 00:09:10.650
Our research largely supports these effects in people who have had some symptoms
00:09:10.650 --> 00:09:14.900
and have had some advancement of the disease, but not quite to the point of
00:09:15.280 --> 00:09:16.780
advanced progressive MS.
00:09:17.170 --> 00:09:20.520
But we do have other evidence, specifically from the COG-X trial,
00:09:20.800 --> 00:09:26.210
that exercise plus cognitive rehabilitation did induce some brain changes in
00:09:26.210 --> 00:09:31.560
people with progressive MS over a 12-week period. it. So again, there's signal there.
00:09:31.870 --> 00:09:35.350
We just have to really do a good job of isolating it to figure out what exact
00:09:35.350 --> 00:09:41.070
prescription might be best for individuals who present differently with the disease.
00:09:41.630 --> 00:09:46.160
Well, speaking of prescriptions, clinicians often struggle with prescribing
00:09:46.160 --> 00:09:48.900
a specific dose of exercise for cognition.
00:09:49.930 --> 00:09:54.360
Based on your current data, is there an ideal frequency, intensity,
00:09:54.740 --> 00:09:58.860
duration that's required before a patient starts showing measurable improvement?
00:09:59.220 --> 00:10:05.070
So our data right now are not quite in the advanced stages of having a definitive
00:10:05.070 --> 00:10:07.170
100% proven prescription.
00:10:07.170 --> 00:10:12.010
But the data from our lab that has resulted in cognitive and brain improvements
00:10:12.010 --> 00:10:14.880
largely involved walking at a moderate
00:10:15.400 --> 00:10:19.280
to vigorous intensity, starting very small at 10 minutes per session,
00:10:19.280 --> 00:10:22.590
a couple days a week, moving up to three days a week, and all the way up to
00:10:22.590 --> 00:10:25.750
about 30 minutes of moderate to vigorous exercise.
00:10:26.160 --> 00:10:29.600
Now, the nice thing about that is that aligns really well with the Canadian
00:10:29.600 --> 00:10:34.050
Physical Activity Guidelines for Adults with MS, where our walking intervention
00:10:34.230 --> 00:10:38.470
really mirrors the aerobic prescription that is appropriate for people with MS.
00:10:38.470 --> 00:10:42.130
So those guidelines are publicly available, and they give tips and tricks for
00:10:42.130 --> 00:10:46.170
how to do the exercise, where to do the exercise, and what people need to do
00:10:46.530 --> 00:10:50.470
for general health benefits. But maybe that could extend to cognition and brain health, too.
00:10:52.160 --> 00:10:56.150
At CMSC, you're speaking to an audience of neurologists, nurses,
00:10:56.150 --> 00:10:58.380
physical therapists, and psychologists.
00:10:59.290 --> 00:11:03.220
So if a clinic wants to implement your findings next Monday,
00:11:03.540 --> 00:11:05.780
what's the biggest barrier they need to overcome?
00:11:05.780 --> 00:11:10.530
Is it patient adherence, or is it a lack of standardized cognitive screening,
00:11:10.530 --> 00:11:14.550
something like SDMT, to track progress, or is it something else?
00:11:14.550 --> 00:11:18.660
I think the biggest barrier would be personnel and time, because our interventions
00:11:18.660 --> 00:11:22.460
involve a lot of behavioral support with participants, where participants interact
00:11:22.460 --> 00:11:24.600
with behavior coaches who help teach
00:11:25.000 --> 00:11:29.220
participants with MS the skills and knowledge to change and maintain their exercise
00:11:29.220 --> 00:11:30.980
behavior so that they can reap the benefits.
00:11:31.450 --> 00:11:36.210
Now, if a clinician wants to deliver that on Monday, I'm not sure how they would
00:11:36.210 --> 00:11:37.870
allocate time and personnel
00:11:38.800 --> 00:11:42.340
needed to be able to deliver that behavioral support, especially considering
00:11:42.340 --> 00:11:47.020
that our behaviorists in our trials are really thoroughly trained to
00:11:48.040 --> 00:11:52.340
think quickly on their feet, use supportive accountability mechanisms to help
00:11:53.260 --> 00:11:57.620
when barriers arise and help each individual participant with their specific situation.
00:11:58.080 --> 00:12:01.700
Now, I don't know if those personnel are readily available in clinic. I'm guessing not.
00:12:02.250 --> 00:12:07.270
But the next phase of research for us, one of the next phases involves looking
00:12:07.270 --> 00:12:10.270
at the potential implementation of our programs in clinic, because at the end
00:12:10.270 --> 00:12:11.700
of the day, that's where the magic happens.
00:12:12.040 --> 00:12:15.630
And that is something that we very much need to do if our program truly does work.
00:12:16.810 --> 00:12:21.870
People living with MS who are experiencing cognitive fatigue often find it incredibly
00:12:21.870 --> 00:12:25.470
difficult to motivate themselves to begin an exercise regimen.
00:12:26.290 --> 00:12:30.700
What strategies or behavioral hooks have you found to be effective in helping
00:12:30.700 --> 00:12:34.840
patients overcome that initial barrier so they can stick with the program long
00:12:34.840 --> 00:12:36.350
enough to see those benefits?
00:12:36.640 --> 00:12:40.950
I think the initial part is the hardest, for sure, especially when you're feeling fatigued.
00:12:41.720 --> 00:12:43.790
One of the things that you can keep in the back of your mind,
00:12:43.790 --> 00:12:48.940
though, is that of all the outcomes studied in response to exercise in terms
00:12:48.940 --> 00:12:52.840
of the benefits, the single most well-established one is that exercise reduces
00:12:52.840 --> 00:12:54.300
fatigue in people with MS.
00:12:54.660 --> 00:12:58.080
And it's to the effect of about a half a standard deviation in magnitude,
00:12:58.420 --> 00:13:02.120
which has been shown in the literature to be universally remarkable.
00:13:02.640 --> 00:13:07.870
Half standard deviation is a big deal. So that is one thing you can keep in
00:13:07.870 --> 00:13:10.450
the back of your mind. But in the front of your mind, it starts small.
00:13:10.890 --> 00:13:14.100
It's every little bit is like saving a little bit of money in the bank.
00:13:14.100 --> 00:13:17.890
And over time it grows and you have to listen to your body on some off days.
00:13:18.140 --> 00:13:21.700
You don't need to go to the gym and exercise vigorously for an hour,
00:13:22.000 --> 00:13:26.930
starting small, sitting less, moving more and slowly accumulating more and more
00:13:26.930 --> 00:13:31.490
exercise over the long haul. It's, it's a marathon, not a sprint. Pardon the pun.
00:13:32.170 --> 00:13:35.750
Well, Dr. Brian Sandroff, thank you for all you do to improve the lives of people
00:13:35.750 --> 00:13:38.690
who are Living with MS. Thanks so much for talking with me today.
00:13:38.690 --> 00:13:40.500
Of course, and thank you for all you do as well.
00:13:41.910 --> 00:13:46.160
Dr. Brad Willingham is a director of the Multiple Sclerosis Research Program
00:13:46.160 --> 00:13:47.920
at the Shepherd Center in Atlanta.
00:13:48.540 --> 00:13:53.230
I began my conversation with Dr. Willingham by asking him how that continuous
00:13:53.230 --> 00:13:56.610
stream of data that a wearable device produces
00:13:56.990 --> 00:14:02.250
can be synthesized into actionable information that a clinician can put to use
00:14:02.350 --> 00:14:03.900
when their patient is in front of them.
00:14:04.510 --> 00:14:07.760
In a moment, will hear my conversation with Dr. Willingham.
00:14:14.330 --> 00:14:18.230
Every day, there's a new device generating new streams of data,
00:14:18.230 --> 00:14:20.620
and the resolution is getting greater and greater.
00:14:20.890 --> 00:14:23.850
Continuous data streams, millions of data points coming in.
00:14:24.330 --> 00:14:27.750
But how do we make sense of it? How do we extract data from that?
00:14:27.750 --> 00:14:32.080
Well, fortunately, that's exactly what AI is good at. And specifically,
00:14:32.080 --> 00:14:38.070
generative AI, large language model technology, it excels at ingesting large
00:14:38.070 --> 00:14:41.160
amounts of information from very different sources,
00:14:41.770 --> 00:14:45.620
And assimilating it, synthesizing it, and summarizing it.
00:14:46.080 --> 00:14:50.610
And so by applying that to monitoring data sources in the clinic,
00:14:51.050 --> 00:14:54.730
we can now realize the full potential of these technologies. Because they are.
00:14:55.000 --> 00:15:01.380
They're giving us unprecedented insight into the needs and progress and abilities of patients.
00:15:01.840 --> 00:15:06.110
But how do we realize that potential? Well, we need some help synthesizing it.
00:15:06.110 --> 00:15:07.450
And that's what AI can do.
00:15:07.790 --> 00:15:13.370
So if you think about converting millions of data points into a brief,
00:15:13.370 --> 00:15:19.410
succinct paragraph of what a clinician needs to do, then we have a way to integrate it into workflow.
00:15:19.970 --> 00:15:26.080
So that AI tool, if you will, is sort of serving as middleware between the millions
00:15:26.080 --> 00:15:28.970
of data points and that one succinct paragraph.
00:15:29.340 --> 00:15:35.570
Absolutely. But a key to that is informing the middleware with clinical perspective
00:15:35.570 --> 00:15:36.790
and patient perspective.
00:15:37.090 --> 00:15:40.350
And so when we think about translating innovation into impact,
00:15:40.860 --> 00:15:46.270
it's critical to engage the patients and the clinicians early and often so that
00:15:46.270 --> 00:15:49.480
we can design system prompts or tailored AI
00:15:50.110 --> 00:15:54.900
so that the summaries that it's generating and the synthesis that it's performing
00:15:55.120 --> 00:15:57.160
is relevant to the objective, right?
00:15:57.160 --> 00:16:01.400
So interviewing patients, interviewing clinicians, what's important to you in
00:16:01.400 --> 00:16:06.030
this data, what can inform your decisions and empower you, and then making sure
00:16:06.030 --> 00:16:10.150
the AI is focused on that part of the information and providing
00:16:10.930 --> 00:16:13.190
and prioritizing that piece of information.
00:16:13.440 --> 00:16:17.110
I'm going to circle back to that in just a moment. I wanted to ask you,
00:16:17.110 --> 00:16:21.980
when you're designing an AI-enabled decision support tool for MS rehabilitation,
00:16:22.600 --> 00:16:27.810
Given how wildly unpredictable and individualized MS symptoms can be,
00:16:28.190 --> 00:16:33.040
how do you train the system to replicate or enhance a human clinician's reasoning?
00:16:33.260 --> 00:16:38.470
That is where the clinician in the loop and the patient in the loop strategies become key.
00:16:38.930 --> 00:16:45.630
So the AI, it excels again at ingesting large amounts of information and extracting
00:16:45.630 --> 00:16:47.450
key points. Is there an anomaly?
00:16:47.780 --> 00:16:51.000
Was there a change two weeks ago, a week ago? Is there a trend?
00:16:51.030 --> 00:16:55.730
And, but then to your point, to actually make that clinically actionable,
00:16:56.050 --> 00:16:59.730
to personalize it within the context of that individual patient's journey,
00:17:00.210 --> 00:17:03.420
that's where it becomes critical to have the patient in the loop,
00:17:03.420 --> 00:17:08.030
clinician in the loop models, where the AI is summarizing and synthesizing the
00:17:08.030 --> 00:17:13.080
data, but then it's clinical decision support, not clinical decision making, right?
00:17:13.340 --> 00:17:16.560
And so I think the patient and clinician in the loop model is really the key
00:17:16.560 --> 00:17:17.750
to that personalization.
00:17:18.690 --> 00:17:22.490
I've heard you talk about the six domains of clinical decision support.
00:17:23.050 --> 00:17:26.790
Which of these domains do you think is the most important or perhaps the most
00:17:26.790 --> 00:17:30.710
difficult to get right when managing progressive forms of MS?
00:17:31.020 --> 00:17:38.360
So when you think about progressive MS and the large timescales at which data must be considered,
00:17:39.690 --> 00:17:44.430
one of the things that's going to become critical is creating care models that
00:17:44.430 --> 00:17:46.820
are predictive and proactive.
00:17:47.170 --> 00:17:52.670
And so by getting data streams over long periods of time and then creating predictive
00:17:52.670 --> 00:17:56.740
models that can then cue a, you know, there might be a change in function,
00:17:57.040 --> 00:17:58.660
there might be a disease progression.