Beyond the Joint: Mechanisms and Management of Central Pain in Systemic Rheumatic Diseases

Beyond the Joint: Mechanisms and Management of Central Pain in Systemic Rheumatic Diseases


>>So thank you, Dan. And so the space bar
is the way to go, huh?>>Yes.>>Alright. So, I don’t have any relevant
financial disclosures. And so first, I want to really
thank Neal for setting me up really well for
my talk and talking about rheumatoid arthritis
and the peripheral mechanisms. But I — do you want
me to turn it down? I’m short, so —
I’m old and short. So these are my disclosures. So the objectives for the
next 30 minutes or so are to describe the characteristics
of centralized pain in patients with rheumatic disease, and then
to discuss management strategies for treating centralized pain in
patients with rheumatic disease, primarily focusing on
rheumatoid arthritis which is the most prevalent. So this is an outline of
what I want to go over, first to just give a generalized
overview of centralized pain in the context of
rheumatic disease, and then to talk more
practically speaking, I think, about how one might distinguish,
say, a centralized pain disorder such as primary fibromyalgia from an early inflammatory
arthritis, and then how you may
distinguish a flare of secondary fibromyalgia
versus flares of inflammatory disease — because you treat
them differently — and then finally concluding
with a little bit on how to manage centralized
pain within the context of rheumatic disease patients. And so this was a figure taken
from a really great review by David Walsh and
Dan McWilliams in Nature Review Rheumatology
that shows the many factors that can contribute to pain
in rheumatoid arthritis, and then how they may change
during the course of the disease from premorbid to
established RA. So as Neal spoke about earlier, obviously there is
peripheral information at the joints causing pain. I think the things that we
don’t want to lose sight of is that there also may be other
things such as comorbidities, particularly fibromyalgia. So patients may come in with
primary fibromyalgia before they even get RA, and that
certainly impacts kind of their perception,
their course of disease. And then they may develop more
of a secondary fibromyalgia or more of a secondary
centralization as well later in their disease,
after they have gotten rheumatoid arthritis. And so this was a study
we did, now back in 2013, that looked at the development of fibromyalgia in
early RA patients. So we used the Canadian
Early Arthritis Cohort. And what we found was
that the incidence of fibromyalgia is highest
in the first 12 months after diagnosis of early
inflammatory arthritis. So during that first 12 months,
there seems to be this window where patients are particularly
susceptible to the development of a centralized pain state. And we found that this was
predicted by high pain levels, poor sleep, and things
like that. And then following on this, we
looked more at centralized pain in established RA patients. So these are patients with a
disease duration on average of about nine or ten years. And we used pressure
pain threshold testing which Neal spoke
about as well — basically a very
simple dolorimeter with a rubber tip placed against
the patient’s skin and you push, and then you ask the patient to
say when they first feel pain. And when that is is the
pressure pain threshold. So lower pain thresholds
equal higher pain sensitivity. And basically what we found
was that at joint sites that are commonly affected by RA
— so the wrist and the knee — the pain thresholds were lower
in RA patients than in controls. And this was statistically
significant, indicating at least a level
of peripheral sensitization. And then at non-joint sites — so the thumbnail
and the trapezius — they were also lower, though
not statistically significant in this study, suggesting
maybe a component of the centralized
process as well. And then furthermore,
we went on to look at a more specific
measure of centralization, specifically targeting the
descending inhibitory pathways. And we found, basically,
that the RA patients in blue in the graph shifted
more towards the left, so lower inhibition
of pain than controls that were age and sex matched. And so really, I
think, the goal here is to treat the inflammation. I’m not saying don’t
treat the inflammation. Definitely as Neal pointed out,
we have lots of great drugs now to treat the inflammation. So to continue doing
that, but then as well to think a little bit might
about the pain mechanisms and to try and dissect them out,
and if needed to treat the pain as well adjunctively through
many of the strategies that have been discussed
yesterday and earlier today. So now I just want
to go into some of the more practical
aspects, I think, of for if you’re a clinician and if you’re seeing
patients in front of you. How do you tell, like, what kind
of pain state that they have? Do they have more of
this centralized primary fibromyalgia, or do they have
an early inflammatory arthritis and you really need to
get on it and refer them to a rheumatologist right
away to institute some of these therapies that
Neal has just discussed. And so, I will say that,
you know, it’s hard. It’s hard to tell. I mean, I think many
rheumatologists will say, “I know, I can tell right
away,” but I think when it comes down to it when you’re
in the trenches, it is hard because
they are very similar. They all have pain. Most of them complain
of fatigue. And you know, morning stiffness,
which as Neal mentioned, is kind of the textbook
thing that people say to look for for inflammatory disease — you know, lots of fibromyalgia
patients say they have morning stiffness, as well. And you know, we’re told an hour
cutoff is when it starts to get to be inflammatory pain. But also a lot of my fibro
patients say they have an hour of morning stiffness
or even a morning of morning stiffness
or an all day. So how do you get down to
that, what’s distinguishing? And I think the difference is
mainly distribution of pain. So is it all over, which
would indicate more of a centralized state? Is it more just at the joints? As Neal mentioned,
joint swelling is key. He showed some great pictures
of joint swelling and how to distinguish between kind of
that boggy swelling versus more of the knobby, bony, painful
osteoarthritis swelling. And then elevated
inflammatory markers can help, though you know, this
also is, you know, hard because there are plenty
of RA patients that don’t flare with high ESRs or CRPs
when they’re flaring. And then converse there
are a lot of people with it that don’t have RA that may
have high CRPs or high ESRs just because of age or obesity
or other comorbidities. And so, how do you really tell? And so this was an interesting
study, that as we got into the literature I
found, by Bennett in 2009 that tried to get at this. So they were basically looking at the revised fibromyalgia
impact questionnaire. And as you would
expect, the patients with fibromyalgia had
the highest scores. The RA-lupus patients had
kind of middling scores. And then healthy
controls had low scores. Depression was the other group
with similar to healthy controls but maybe a little bit higher. And then — I’m not sure
how well this projects but they actually went
down to the item level, each item on this questionnaire, and looked to see
what distinguished between fibro patients
and RA-lupus patients. And then the three that I
highlighted are the top three. And so, basically the activity
of vacuuming, scrubbing, or sweeping floors,
interestingly, fibro patients were
not very able to do that whereas RA-lupus patients
were better able to do that. Sitting in a chair for 45
minutes which was surprising to me because you know RA
patients are very stiff — they’re not going to
want to be sitting in a chair for 45 minutes. But actually they could do it
better than the fibro patients. And then this concept
of tenderness — so fibro patients were much more
tender than RA-lupus patients. And then they went on to
look at distribution of pain. And so this was not on the fibromyalgia impact
questionnaire but it’s separate to that that they added to see what sites were
more likely to distinguish. And fundamentally,
it was arm pain — so not at the joints,
but in the arms. Fibro patients had
higher levels of arm pain. And then hip and thigh pain
— if you look at thigh pain, intriguingly, 55% of fibro
patients endorsed thigh pain, whereas zero percent in
this population of RA and lupus patients did. And then the back —
the back was a huge area for fibro patients and
the axial skeleton not as prevalent in RA or lupus. And then they did the
statistical method where they put in all the
variables and try to figure out which ones would contribute
most to distinguishing, and in the end they ended
up with 15 variables, with the top three being
mid-lower back pain, tenderness, and neck pain. And if you included the whole
15 variables, they were able to distinguish fibro patients. They said fibro patients
were fibro patients 99% of the time correctly. And they were able to distinguish RA-lupus patients
90% of the time correctly. And then they hone down a
little bit more on this concept of tenderness versus pain. So I think we often equate
tenderness with pain, but you know, I think
they are different. And fibro patients, you’ll
see, have high levels of both. But they have more
tenderness than pain, whereas RA-lupus patients also
have both but tend to talk more about pain than tenderness. And I think in clinic I see that
a lot where patients will say, you know, I have aching
pain in the joints but they’re not necessarily
endorsing tenderness even when I press. And so then I want to
just go through and talk about a little bit
more about, now, let’s say you have a
patient that you know has RA. How do you distinguish
in them, you know, whether their pain is due to
their RA versus whether it’s due to a more centralized
process such as a flare of fibro or something else? And so in order to do
that, I think I first need to take a step back and explain
how disease activity is measured in rheumatoid arthritis. So rheumatologists often use
what’s called composite disease activity measures. So these are things like
the disease activity score in 28 points, the simplified
disease activity index, the clinical disease
activity index. And so basically they
take different components and they somehow
put them together, whether through addition
or multiplication or whatever to get a score. And so on the right side, I show
you a graph of the components of the DAS28 which includes
a sedimentation rate, so an objective lab marker;
the tender joint count which the physician
presses on 28 joints and asks the patients
whether they’re tender or not; a swollen joint count
assessed by the physician; and a VAS global health which
the patient reports themselves. And you can also see the
kind of contribution of each to the score, with the SED
rate being the highest, followed by the tender
joint count, and then the swollen joint
count, and then the VAS global. I think the thing to
really note is that a couple of these measures, particularly
the tender joint count and the VAS global, are very
subjective and highly influenced by pain, regardless
of what that pain is. And so you can get elevations
of this disease activity marker by other causes of pain
that are not inflammation. And so this was a study done
by Tonn [phonetic] et al. that showed with increasing
fibromyalgia tender points on the X, you get overall
increases in the DAS-28 score. So, the more tender you
are, the higher your DAS, regardless of what level
of inflammation you have. And in particular, the
tender joint count — that little dotted
line — skyrockets. And so that’s probably not too
surprising because, you know, tender points, tender joints. You’re both feeling them. They’re both assessing
tenderness. So — but something to keep
in mind when you’re thinking about assessing disease
activity in an RA patient. And as well, the VAS
global health increases, whereas the more
objective measures — the SED rate and the
swollen joint count — stay steady pretty much across
all tender point classes. And so this was a study that used the fibromyalgia
survey scale which I believe was
talked about yesterday. So just to remind you
guys, on the left, there’s the fibromyalgia
survey scale that shows the widespread
pain index and the symptom severity score. And basically, they looked
at different levels of this. And it’s indicated
as PSD on the graph because it’s also called the
poly-symptomatic distress score, but same thing. And they basically graphed
what the physician’s assessment of global disease
activity was on the X and then the patient’s
report of pain on the y-axis. And you can see that
patients with low scores on the fibromyalgia
survey score or low scores of PSD had a pretty good
correlation, whereas if you go to the high scores which
are the open circles on top, you can see there’s pretty much
no correlation between the pain and MD global activity. So these are the patients
with high centralized pain and less inflammatory pain. And so how do you distinguish? You know, I think there’s a
lot of clinical gestalt here. There have been things
in research studies that have been suggested
to be useful. Two of these, the top two,
look at the discrepancy between the tender joint count
and the swollen joint count — so how much tenderness you
have versus actually swelling. And then the last is a measure
of what’s called the DAS-28P which is a measure
of the proportion of subjective components in
this composite disease activity measure compared to more
objective components. And I’ll go into a little
bit of these following. So these were some studies
that looked at the difference between the tender and
minus swollen joint count. And so they identified this
optimal threshold of seven. So if you had seven
more tender than swollen on this 28-joint count, they found that it
predicted the presence of at least 11 fibromyalgia
tender points with 72% sensitivity
and 98% specificity, which is what they
called fibromyalgic RA. And then to see how this
affected disease activity assessments of RA and these
composite measures are the graph on the bottom with
a DAS-28 on the left and the C-DIE in the middle. And you can see that
the patients in black, the fibromyalgic RA patients, had much higher scores
than the others. But when you looked at a
more objective, arguably, measure of disease
activity, which was a tender and swollen joint count
greater than or equal to three and a SED rate greater than
or equal to I think 28, there wasn’t actually
that much difference. So it really affects how you
measure the disease activity in these patients. And then on the right
you see another study, a more recent study, by
Michelson that also looked at this tender minus
swollen joint count. They use different cutoffs
here and they looked at remission assessed
by a whole slew of different composite
disease activity measures in both rheumatoid
arthritis on the top and psoriatic arthritis
on the bottom. And the bottom line
to these graphs is that the more discrepant you are in the tender swollen
joint counts — so the further out
to the right — the less likely you are to be
considered in remission based on these disease
activity measures. So this is just kind of a
cautionary that, you know, when you’re looking at
disease activity measures, you need to also consider
what other causes of pain that this patient may have
because you may not want to go escalate therapy
to the next DMAR, to the stronger
immunosuppressant agent, because maybe in these patients, even though their disease
activity composite measures are saying they’re high, maybe they’re not actually
as high as they seem. And then just another
illustration of this similar concept but
now looking at ratios instead of differences — and so this
group specified the low ratio of less than 0.5 of swollen
to tender ratio as to be kind of the fibromyalgic
patient, so basically twice as many tender joints
as swollen. And once again, they looked at different disease
activity measures. Here is the ACR-20,
another composite measure where if you have low
STR, you’re less likely to get a 20% response. And similarly, you’re less
likely to get a 50% response — an ACR-50 — if you
have low STR. And then when you put this in a multivariable model
including things such as sex, physical function, steroid
use, baseline disease activity, the thing that really predicted
poor treatment response was the low STR. And then just one
slide on the DAS-28P because I think it’s a little
bit more complicated and harder to do directly in clinic. But this group here
looked at the fraction of the total DAS-28
score reported by patient reported components,
so the tender joint count and patient global assessment,
and divided by the whole, and saw an association between
that and fibromyalgia-ness as assessed by the
fibromyalgia survey scale. So in the last five
minutes or so, I just want to discuss
managing centralized pain in rheumatic disease patients. And you know, I think
largely it’s similar to managing centralized
pain in other patients, as has been discussed, you
know, by other colleagues in more depth earlier
today and yesterday. These are the ULAR
revised recommendations for the management
of fibromyalgia by Dr. McFarland
and his colleagues. And you know, they
focused on starting with the non-pharmacologic
therapies and then adding the
pharmacologic therapies. I will say that as with
what others have discussed previously, patient
education is key. I think you’ve seen the
diagram here before yesterday, but really talking to them,
spending that time to validate that their pain is real. And then I think a lot of the
studies that other colleagues such as Neal have done
in neuroimaging — so I just stole a picture from
one of his publications here — are helpful in this, in really
showing patients, yes, you know, kind of similar to some
of the ultrasound images that Neal showed earlier which
showed the red, fiery images in the joints to convince people that they have joint
inflammation. Also, saying that, hey,
there’s something going on in the head and it is real. There is a lot of value to
that because it gets buy-in and it gets them to get with
you to do the next step, to do some of those
non-pharmacologic therapies and to put the work in. So I think that is
probably the key point here. Then, I will go a little
bit into discussing some of the non-pharmacologic
therapy. I’m going to just look
at the data in patients with systemic rheumatic
diseases, as in to contrast what
other speakers have spoken up about beautifully in
fibromyalgia and other states, but in patients with
systemic rheumatic diseases. There is some data on
aerobic exercise as well as psychological
interventions, specifically CBT, stress management, and
education which they lump into psychological
interventions. There was a systematic
review in Arthritis Care and Research looking at pain
outcomes for aerobic exercise and they found that there was
a small effect size favoring aerobic exercise for
pain in these patients. So a lot of patients
will come in and say, well is it safe to exercise? Will it aggravate my pain? You know, no. It actually will
improve, if anything. Interestingly, they also looked
at joint count which was more of a measure, I think,
of inflammation. They didn’t actually see
any significant effect for inflammation, so arguing
that some of this effect on pain may be through more of the centralized
pain pathways rather than through the
peripheral pathways. And then there was also
another review in Arthritis Care and Research that looked at 13
studies with followup ranging from two to 14 months
post-treatment of psychological interventions. And they found a small but statistically significant
post-treatment effect size with a Hedge’s G of 0.18. They didn’t find any
statistically significant difference between
types of interventions. So they were only able to look
at CBT, patient education, and stress management. And there was no
difference between the three. Though if you look at
each of them individually, the stress management seemed to
have the greatest effect size. But once again, not a lot of
studies, a lot of heterogeneity, a little hard to tell. And then just a brief nod
to the pharmacologic therapy in systemic rheumatic diseases. There has been some data on
tricyclic antidepressants. There’s been a little
bit of data on SNRIs. And I put anti-convulsants here
though I could not find actually great data on anti-convulsants. But it is something that
I do use in my own arsenal so I thought I would
put that there. And so this was a Cochrane
review on tricyclics for pain in patients with systemic
rheumatic diseases. They were only able to look at
a very small number of trials. They looked at pain intensity
improvement in the short term, which was one to six
weeks, versus placebo, as well as in the longer term
which is greater than six weeks. In the short term, they were
only able to find three trials with extractable data on
a consistent pain outcome which was the VAS, the
visual analog scale. And two out of three trials
favored the tricyclic. They could not do a formal
metanalysis, however, because there was just so much
difference between the studies that it was impossible to do. When they looked at
more longterm results, unfortunately, it
was less obvious. Only one out of three
trials favored the TCA. Once again, metanalysis
couldn’t be done due to heterogeneity in the studies. They did look at side effect. As would be excepted, there are
a higher number of side effects in patients taking
TCAs than placebo. Interestingly, there were no
difference in GI side effects but the side effects were
more CNS side effects and anti-cholinergic side
effects with the TCAs. So in conclusion, I think TCAs
may be effective for management of pain and RA in
the short term, though the effect
sizes are small. And then one does need to
be concerned about looking out for side effects,
particularly the CNS and anti-cholinergic effects. And then we did do a study
— this was a study that Dan and I collaborated on. Dan helped mentor me through
this which was a pilot trial of milnacipran, which is
a serotonin-norepinephrine reuptake inhibitor, on
pain in patients with RA. This was a randomized,
blinded cross-over trial. So patients crossed over. They received both
interventions. Sadly, the primary
analysis was null. So our primary outcome
was the BTI pain intensity and there was no difference
for when patients were on placebo versus milnacipran. The thumbnail pressure
pain threshold, though, did seem to be increased more — so less pain sensitive
when they were on the milnacipran than not. And then the interesting
part was that in the subgroup
analysis where we really hone down to those patients with only
one or fewer swollen joints — so very well controlled
inflammatory disease — there was a significant
difference. And those on the
milnacipran group at that time period
did improve more than when they were
not on milnacipran. And similar effect on
the pain threshold, that the higher increase
in pain threshold while on milnacipran than placebo. And then we also looked
at adverse effects and they were very similar
to what’s been reported in the RCTs for fibromyalgia. The greatest was nausea at 27%
of patients reporting that, loss of appetite in
10%, vomiting in 7%, and sleep problems also in 7%. So, I think conclusions here
are it’s really important to carefully phenotype patients,
determine the cause of pain, and we’re going to work on more
studies to be able to do that. SNRIs may be efficacious
for RA patients with very well controlled
inflammation but side effects were common. So really, in closing,
I think I want to hone in on these take
home points: one, the characteristics
distinguishing a centralized pain condition and an
inflammatory arthritis — tenderness to touch;
difficulty sitting for 45 minutes intriguingly; pain in the axial
skeleton, arms, thighs. And then when you have a patient
with an inflammatory condition, how do you distinguish between
a flare of that versus a flare of more of their fibromyalgia? Maybe look at discrepancies
between the tender and swollen joint count
or subjective components of the DAS versus objective. And then management — really very similar to
primary fibromyalgia, and a lot of what the
previous speakers have spoken about today are really important
take home points there as well. So I just want to close
and acknowledge my mentors, collaborators, and funders. Thank you. [ Applause ]

Leave a Reply

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