Measuring function in cancer patients: do we need disease- or sarcopenic-specific measures?

>>Thank you and thanks to the organizers
for inviting me here today. I also harken from the University of
Alberta and my task is to talk to you about measuring function and I want to say that
with a caveat. 20
some odd years ago when I first stood up in front of a group of endocrinologists and started
talking about function, three or four minutes and
a brave soul in the back raised their hand and said: what
organ are you talking about? And I said well, I’m not talking about an
organ, weíre talking about moving in three dimensional space. I am a physiotherapist and also an exercise
physiologist by training, so I came at this from an orthopedic and endocrine
world, and Iíll try to apply that and hopefully it will be
helpful. My other comment is I sat through many presentations
now, including Vicki’s, and it made me think: do we really need more evidence or
do we need more implementation into clinical practice? That
changes. I canít imagine data from 100 years ago on
body surface area is better or that there was data
then than there is today. So with those caveats these are my disclosures
and I will say I was involved with a couple of clinical trials with a couple
of drug companies, and as a physiotherapist Iím involved
with a couple of companies in the American Physiotherapy Association creating outcomes
registry and that is to measure physical functions across
various populations in the United States mostly and across
Canada.>>I want to take you through a whirl wind
tour of my world to sort of say: what are the benchmarks and
biomarkers that we already have in place for sarcopenia, how do we actually conceptually
measure function and talk to you a little bit about
self-report compared to performance-based measures that I
think are really important to distinguish between. Iíll then turn my comments to, and I think
others have already said this, itís not a sentinel muscle,
itís not a sentinel measure. Itís probably a battery of
measures that we need. So this is been well defined by all of my
colleagues whoíve gone before, that sarcopenia is loss of muscle tissue and it
was associated very much with aging. And then from the work
in Europe they came along and said: no we are going to describe it as a syndrome. And that syndrome
includes not only the low muscle mass but also low muscle strength and/or changes in
physical performance. But what do we mean by physical performance?>>If we look at the mechanisms of sarcopenia,
what we see is the primary reason it might be aging. But if
we think of cancer and cachexia, itís the disuse atrophy that occurs, the inadequate
or change in nutritional status of individuals, and I might
point out the hormonal regulation and other diseases all
factor in to sarcopenia as well. I think, to the comments made before about
the multiple hits that one gets, the cancer and cancer treatments affect
the endocrine and neuroendocrine system. The cancer
and cancer treatments affect our ability to take in and use nutrition. They also affect the amount of
activity that we do on a day-to-day basis. So all of those lead in, and thatís why I
think a multi-factorial disease needs a multi-factorial approach.>>Youíve seen these before and itís just
to point out that: yes, in the lower extremities, quality of muscle
matters as much as quantity of muscle and we are very concerned that way back when I
did testosterone dose response studies, we saw
these changes not just with aging but also with the change
of using a G8 or H1 acting agonist in combination with various doses of testosterone, and you
could get up to 3 kilograms or 3.5 kilograms gain in
lean body mass but also increase in fat mass for those
individuals on low doses. So it is important to think not just about
the quantity of the mass but the quality as well and the intramuscular fat. Iíve seen many MRIs and CT scans of individuals
in their 80s who continue to compete as athletes and would
look more like the top figure than the bottom figure. We also have to think about if we are designing
trials with endpoints that include muscle mass and
strength, what about physical function. What are the implications, the metabolic complications,
associated with cancer and some of the treatments and how do we evaluate those? So one might want
to address fatigue and the ability to move long distances or walk to the store to get
groceries and be independent with their living. And that is more on the aerobic side and metabolism
becomes more of an issue than just simple grip strength (or muscle
strength per se).>>What biomarkers are used, or can be used,
to identify sarcopenia? The Holy Grail that we are chasing or
trying to chase, and I harken to Vicki’s comments on not setting specific cutpoints, is biomarkers
for osteoporosis they look at bone mineral density. And they associate that with falls and fractures
and the composition of bone. And I think we are dealing with something
very different here. And that is that you
have lower muscle mass and changes in mass and changes in quality of certain muscles
that may be very different. We have androgen-responsive muscles (say the
quadriceps) and the proximal musculature which are much more responsive than say the
psoas or peripheral musculature. So I think we need to be
more selective about what we are trying to measure and why. The quantity in many can be done by
various imaging analyses, the strength -and I see much of the literature looking towards
not lower extremity strength but the quality of the
muscle strength as far as the upper extremities. And Iíll not
show some of the work that, Peggy you have done and have in the past, and that is that
muscle strength of the handgrip is probably an appropriate
measure. But in my world, what if the patient walks
in with sarcopenia but is also sarcopenic obesity
with osteoarthritis of the hand, or rheumatoid arthritis and
now I need to go to a completely separate measure. Most of us use gait speed for mobility (thatís
again walking and functioning in the real world).>>The European group said we need to define
sarcopenia. So they came up with (and I think it holds
up today) is low muscle mass in combination with
low muscle strength and/or low physical performance. If
you have all three, then we will call that severe sarcopenia. But in terms of the clinical applications
of this, they reversed it and I think quite rightly
to say: let’s measure activity levels with gait speed first. If
an individual has adequate gait speed, which is defined as greater than 0.8 meters per
second, then we would move to evaluate your muscle strength. If theyíre okay there, they don’t have sarcopenia
and if not, then we go on to evaluate muscle mass. So for clinical operations they reversed the
order of that. They also gave us lots of different ways,
and are we are all familiar with how we measure things in
research and, in the red boxes, how we typically see things measured in clinical practice.>>FNIH then looked at the same type of thing
and asked if we could look at the same cutpoints for that. And develop cutpoints for grip strength of
less than 26 kilograms for men and 16 kilograms for women
for grip strength. Low muscle mass is appendicular lean muscle
mass for men and for women. So at least
we have some idea, albeit not perfect, of the cutpoints. If I think back to my world of how we measure
physical function and physical therapy and people moving about in the real world, we
used the ICF model. It used to be Impairment, Disability And Handicap. Weíre much more friendly in terms of our
terminology now, and it is an International Classification of Functioning, Disability
and Health. The two
areas that weíre really focused on then are body structure and function is all that we
are measuring when we think of mass and measures of strength
like grip strength. To get an activity-based measure we
move to gait speed. So the relationship between these two in many
populations (orthopedic, cancer, rheumatology, infectious disease, inflammatory
disease processes) is really weak correlations between
the two at best. Which is why we need to measure the measures
of from each of those categories and not try to get one measure to be the be-all
and end-all and the sentinel measure for everything we are
doing.>>So why do functional and physical performance
measures? I got asked this a lot way back when I had
a colleague who sat down and said: Linda you
could do a quality of life measure, something that a patient
fills out. And the domains are so much broader for what
we evaluate, why do we have to walk them up and down the hall to get a gait measure. It is difficult to do an in a physicianís
office, etc.. So let me walk
you through the arguments that we had and the data we came out with, and itís important
to see why we actually had those results. We use measures of function to discriminate
functional status and that’s what you have seen with the sarcopenia criteria. We use them to predict subsequent function
and, in this case, survival for cancer patients who
have survived cancer and treatments and are going on now to
what we would consider to have a chronic disease. As well as, and I will take some time to show
you how we use them to evaluate change in function
over time, where the measure has to be not only valid
but reliable and sensitive to the change that you are trying to measure, be it an intervention
or the change in the individualís functional status. So you have seen lots of these graphs today
and this is another one where non-small cell cancer where
they actually looked at a cutpoint of 400 meters to
predict cancer survival and how an individual would do with chemotherapy and other treatments. And
for those of you who arenít familiar with the six minute walk test, you walk as far
as you can in six minutes and typically individuals who are
healthy would be around 650 to 700 meters in that six
minutes.>>We also know from the NHANES study about
cancer survival that if we looked at a shorter physical
performance battery, so these are tests of balance (standing balance). They are tests where you rise and
sit down again in a chair five times within a minutes. And walking short distances for meters of
gait speed, as well as the fast walking pace. In this case it was more than six meters per
second, and what you see is that those physical function measures
relate to survival and can be used to predict subsequent events. So how do we evaluate, then, change in function
which I think is of more interest to you in terms of interventions you use.>>We have to think of the appropriate mechanism
and the time course of the adaptation to the intervention. Time and time again I see people picking a
measure because it is easy to do in an office visit and itís quick to do but it doesn’t
necessarily relate to the physical function that the patient actually
wants or struggles with, and it doesn’t necessarily relate to the intervention. If your intervention is going
to improve, say, oxygen flow to muscle, you want to measure something on the aerobic side. And if their
interest is to reduce fatigue (to make it through a day), you want to intervene with
aerobics and measure that. If however theyíre interested in getting
out of a chair and they don’t have the strength to
do that or to transfer independently, then that becomes the strength issue and we need
to be clear on what we are measuring and why.>>I would argue that measures need to not
only be disease-specific but also condition-specific and I will
use an example from total joint arthroplasty to show you that data. As said before, we need both self-
report (so thatís the patient-oriented outcome measures that you see now, a pen and paper
task) that are specific to the questions about function,
not necessarily the same as quality of life. And they are
different from physical performance measures. This is a sample from about 2000 patients
that we originally did, and in a series of trials
that weíve done since, where we look at the lower extremity
functional scale, it takes less than two minutes to fill out, widely used across ambulatory
care centers and with patient centers will all different
types of disease (stroke, cancer, you name it). We looked at
performance measures that we did: a self-paced walk ever 200 meters, timed up-and-go (where
they rise from a chair, walk three meters and come
back and sit down again) thatís timed, and a stair test. And what you see is that they are moderate
correlations at best. The same is true if you look at cancer. So that what does that mean? If it is a moderate correlation between actual
performance-based measures and self-report measures, then theyíre
telling us different things and we need to measure
both.>>More importantly is probably these series
of studies. There has been a number since. And that is that
the two bars on the left-hand side are the WOMAC. Itís a disease-specific questionnaire for
physical function related to osteoarthritis. And the SF-36. Compared to the gait speed and 6-minute walk
test. So
while the patient’s perception is that their performance had improved and that’s what they
reported on their questionnaire, in actual fact if we
measured gait speed and 6-minute walk, theyíve gone in the
opposite direction. And thatís why it is crucial to measure both
since theyíre telling us different things. The next question I often get is how do I
know that the change is important and real? There are two
things that youíll see in the literature, often used interchangeably. So I just want to clarify them for you. One is a minimally clinically important difference
(MCID). They are anchor-based. So I might do a 6-
minute walk test and I measure how far the person walks, and I do that pre-and post-intervention. But I
will also ask them: do you perceive that youíre walking better, no change, or worse? I might also ask a
spouse or a family member to rate the individual. So there is a global rating of that activity
and thatís what you use to calculate your MCID and thatís
why we call them anchor-based measures. These are
very subject to something we call ìresponse shiftî. So as a person’s function declines, their
internal value for understanding how they are performing
in the real world and walking, functionally getting
along, changes over time. And it changes, at least in orthopedics where
weíve measured for joint arthroplasty, about every six months they
reset their expectations. So is it that we want the patientís
perception of function or do we want the actual performance-based measures? And I would argue for
doing both.>>You will also see MDC, and that is minimal
detectable change. So how do I know that the patient actually
improved from pre- to post-intervention versus simple measurement error? This is a mathematical
calculation thatís distribution-based thatís based on reliability of the measure and the
standard deviation or standard error of the mean. It is a calculation and I would argue that
performance based measures, where many go back to the anchor-based
and ask the patientís perception, which I just told
you with response shift changes over time, but we really should consider, for cancer
patients and patients living with cancer and surviving
cancer, that we should be looking at MDC and what the actual
distributive measures are.>>We looked at growth curves for 6-minute
walk tests and itís similar to what you saw this morning: males
and females on almost all performance measures are different. And thereís an expected trajectory,
which for us is recovery, that we know what we would see. If we then look, and this is just males alone,
what was the pre-op functional status, or the pre-intervention in your case functional
status? That often
dictates the trajectory of the recovery for an individual. So it is important to measure them before
and also after their intervention and treatment. The other thing to note is the MDC for us
in total joint arthroplasty -almost everyone, even orthopedic
surgeons if you ask them, they know the value of 61
meters. So if I measure a patientís 6-minute walk
test and they can do 250 meters and they come back
after the intervention and they are at 300 meters, I know that thatís a 50 meter change
but not enough change to say that it is beyond the measurement
error of the test.>>I also know that the MCID, so for cancer
and COPD patients we ask spouses, family members and the
individual whether they perceive an important change and that is somewhere around 54 meters. I think
sometimes the intervention trials are actually too short of a window and maybe not intense
enough or large enough doses of particular (if weíre
looking at myostatins and other anabolic agents or other
treatment agents) they are not large enough to actually get some of these changes and
I think in shorter duration trials, and specifically in patients
with cancer, we need to go back and calculate these. These
are just some of the measures. So we anchor any of the performance measures
that we do to say: individually how is the patient improved or
not, will I continue with this treatment? We anchor them
based on the MCD-90 evaluations.>>I will walk through a couple minutes quickly
the functional capacity of individuals. There is tremendous
variability based on gender, based on age, based on physical function, how much people
exercise and workout over time. We also know that something happens, as Vicki
said, and the individuals may lose their physical capacity. We use this threshold of functional independence,
the ability to get up and move around in the real world. With this large span and different trajectories
of recovery from this. The more
reserved capacity you have, thatís the maximum compared to the threshold for functional
independence, the greater that difference, the more insults you can take before you tip
over to have functional decline. We need to consider the longitudinal measures
that are done and avoid floor and ceiling effects. And by that I mean, if I am testing an athlete
I cannot simply do a timed up-and-go to walk three meters. The ceiling effects would be so large that
I canít measure that person prior to, after an intervention or treatment for cancer or
surgery or other, and then recovery. So I need to have
measures that are going to be over a long period of time pre-, post-intervention and
out 5, 10 and hopefully 15 and 20 years.>>The other thing I want to mention quickly
is if we look at the change, either with surgery or an
intervention, there is a functional decline and the responsiveness of the measure, if
I measured a 6- minute walk test here, I wouldn’t see any
change very operatively or in close concert to a treatment or
cancer therapy. The measure I need here is going to be very
condition-specific. For us, even with the
same disease (say osteoarthritis), a stair climb measure is much more sensitive to change
for the knees, and for hips we use the TUG. But for the broader continuum, we use the
6-minute walk test. And this is just the data to show you that:
why do we anchor these measures in TUG? We have 30
seconds to do a TUG. It means the patient is dependent on others
for care and transit. 10 seconds or
less, they are completely independent with their living. But as I said previously, for the hip patients
this is quite a responsive measure perioperatively
for us, but it is not if I measure pre-intervention and, say,
six months to a year post-intervention. I would get ceiling and floor effects and
I would not be able to measure anything. That is why you need a battery of measures.>>My last comments relate to Peggyís work,
and thatís the international group who flipped the question
around and said: do the diagnostic cut points that weíre currently using for low muscle
mass and muscle strength, actually predict -they call it mobility/disability,
but Iíd probably call it mobility limitation? What
theyíve done is theyíve analyzed, and Peggy has already gone through all that so I donít
want to repeat but the bottom line is that the muscle mass
doesnít really correlate very well with mobility disability in
all those trials. The grip strength, absolute or scaled to body
mass index, actually did predict mobility disability. And we need more of that and we need much
larger pragmatic trials to be able to measure the disability with the interventions that
you are going to use.>>So to summarize, we need to do measurements
of activity and physical function, not just structure and
function that weíre doing now. We need to have self-report, thatís both
familial support and performance-based and we need to be condition-specific. Muscle strength is probably closer to mobility
in that continuum than is simply mass. But Iím not saying that we do it instead
of. We are saying we
need to do these measures in addition to changes in muscle mass. And we need a competent
evaluation. I know it is nice to have just one or two
sentinel measures in muscle but I don’t think it is
enough. I think we need a battery of measures that,
if we consider an aging population with other comorbidities, hormonal dysregulation thatís
going to occur with cancer and cancer treatment, we need
a battery of measures that tells us the direction we need to go to be able to clearly evaluate
our intervention. Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *