Intramuscular Electrical Stimulation for Post-Stroke…

dr. J associate professor of physical medicine and rehabilitation and biomedical engineering at the Case Western Reserve University he received his bachelor’s and master’s degree in BME biomedical engineering from Duke University and Dartmouth College respectively he received his MD and his clinical training in PMR from New Jersey Medical School he completed the rehabilitation medicine scientist training program fellowship in 1998 at Case Western Reserve University currently he’s the director of research for the department of physical medicine and rehabilitation and the associate director of clinical affairs for the Cleveland functional electrical stimulation Center his research is funded by NIH and focuses on the application of Fes for neural prosthesis neural plasticity and shoulder dysfunction in hemiplegia he has disseminated his information in the form of multiple publications and he commands an international reputation and a national reputation he serves on the editorial boards of physical medicine and rehabilitation archives and neural rehabilitation and neural repair he recently completed service as a member and chair of the NIH ni CHD review panel called function integration and rehabilitation sciences and he currently serves on the advisory board of the National Center for medical rehabilitation research at the NIH he has also been invited to give numerous presentations worldwide and and he has been asked and appointed to review on scientific panels worldwide he has received an award for almost every year after his graduation which is a really extraordinary when you when you think about it his most recent award and I should say that one of the toughest audiences that we have to face as professionals is our own peer group and from this group the Association of academic physiatrist dr Chae received the 2007 Branum research award which recognizes an individual who over the previous decade has had the most significant impact on the science and practice of rehabilitation medicine I think a really singular award so with that dr. Shane Thank You Janice that’s a lot more than I gave you I don’t know where you got all that information from so good morning off I think this end P might be a little bit different than traditional I really would like some of your feedback and kind of share a little bit of my soul if you will on things that’s worked well and things that haven’t so work work well so today’s topic is percutaneous intramuscular electrical stimulation for having pleaded shoulder pain as usuai do you have some disclosures I will describe a device that’s been investigated under an FDA IDE I also serve as a consultant to NDI medical who has a commercial interest in the device or the approach that we will be talking about today so where’d it all start well fall of 1989 just started my senior year in medical school and I was doing a rotation at Esalen Institute for 4 p.m. and our and during that rotation senior resident told me John if you can figure out a treatment for shoulder pain and stroke patients you’ll be a very rich man well I’m still not not rich but this is a an area that’s a significant interest to me I finished my residency in 94 at Kessler and New Jersey Medical School and then I came to case in 94 where I started a fellowship and rehabilitation medicine scientist training program with Hunter as my mentor I finished out in 98 so that brief background was my objective today we’re gonna describe the epidemiology and postulated ideologies of hanging through your shoulder pain we will describe the development of intramuscular left Racal stimulation as a treatment modality for this particular

condition well that’s the effectiveness and safety to muskie electrical stimulation and then we’ll describe necessary steps for clinical implementation of this approach so let’s talk about prevalence prevalence is defined as the amount number of individuals that have a particular condition and a particular time and point as versus incidence which is a number of new cases prevalence in stroke survivors for shoulder pain has to be really presented from two perspective one is what they call unselected population that is all comers and then selected those individuals who are severely impaired enough to require inpatient or some sort of rehabilitation and prevalence in terms of stroke is always linked to the onset or time onset of the actual event and really the most important statement here is really this the most recent article by Lindgren and Association Journal Stroke the prevalence of stroke painti hemiplegic shoulder pain is 22% at four months but at 12 months is still 24 percent and there are some data that goes beyond this to approximate four years and but those are only really for individuals who are more severely impaired now having said that if you look at a selected group of stroke survivors the prevalence is substantially higher okay so prevalence is almost 60 percent at four months post stroke if you have a if you yes if you’re really impaired enough to require inpatient rehabilitation in this particular group if you go out four years approximately 25 to 30% still have significant shoulder pain so it’s quite a prevalent condition and a very serious one now why is it relevant well hey you put your shoulder pain is very strongly associated motor impairment associate reduce activity limitation as well as reduce quality of life and therefore is a major rehabilitation problem impact is quite significant if you look at the incidence a number of new stroke survivors per year at 70 50 thousand per year approximately a 17 percent will die within 30 days that transit to each year approximately 150,000 and new individuals at one year post stroke exhibiting shoulder pain now you add that to the prevalent population which is number of individuals who actually have stroke right now in the US as 5.8 million in that 5.8 million if I estimate approximately 10% having shoulder pain that’s five hundred eighty thousand individuals certainly this has clinical implications but also commercial entities love to hear these numbers as well clearly shoulder pain has a significant impact in the public in public health now this is a very busy slide and we asked a question what’s the cause of shoulder pain and hanging pleasure the answer is we don’t know and so like any good research we have to come up with a conceptual framework to start your process so here’s my conceptual framework well there’s true or not I don’t know but that’s what we’ll be testing so I’m a background a little bit before I do this you know unlike most joints in the human body the shoulder is extremely mobile joint and as a result it’s very unstable by nature the joint is stabilized primarily by muscles in particular the rotator cuff muscles the supraspinatus infraspinatus subscapularis and teres minor there’s very little ligamentous mechanical stability for the shoulder in comparison to other joints of the body so then you have a stroke and what a stroke do well cause spasticity well reactivity when you don’t want it and weakness decreased strength when you do need it and i postulate this these two entities have a significant impact in the mechanical instability of the shoulder with a mechanical instabilities manifested by a number of things joint mount alignment we do this range of motion scapula now rotation impingement syndrome as well subluxation but these I believe put our joint shoulder joint a significant risk from micro/macro trauma leading to inflammation and degeneration of the capsule as well as extra capsular tissue including tendons muscles and Bursa leading the capsulitis in contractures of the capsule tendinitis tendonosis tears as well as per situs and of course leading the pain and this pain then further reduces the immobility of the increase in mobility of the shoulder and this immobility is a significant impact and struggle course as you all know stroke recovery is activity dependent so if you have immobility this further accentuates our hemiparesis and from the neuroplasticity perspective it’s got a very negative impact which then may have a vicious cycle of worsening mechanical

instability there’s some other issues as well for example spasticity may direct directly cause pain but causing periosteum and so on so again mechanical instability is the major factor I believe in contributing to hemiplegia shoulder pain now this doesn’t mean that neuropathic pain is doesn’t have a role certainly does have a role but in this particular model we were focusing on mechanical causes of shoulder pain there lot of treatments in fact if you were to survey the literature and-and-and a surveys have been done in natural clinical practice there are more than hundred fifty different treatments and generally there’s no consensus as to which is effective which means that we don’t really know what we’re doing however with the exception of surface electrical stimulation to reduce subluxation I’m sorry I’ll just say that only exception to this is surface electrical stimulation to reduce subluxation all other treatment options have been shown to be ineffective this statement is supported by three practice guidelines or evidence-based reviews that came out since 2004 first one is by t-cell came out in 2004 I believe and statement is quite strong there’s strong evidence that FES reduces pain and subluxation and improves muscle function and shoulder range of motion Pam Duncan and her group with VA Department of Defense practice guidelines says that surface electrical stimulation may be effective for reducing subluxation and pain based on good evidence and the most recent the Ottawa panel from Canada also recommend FES for acute and chronic phases so clearly literature supports to use surface electrical stimulation to reduce shoulder pain well let’s go back and I’m back to you now in 1994-1995 sure strong evidence may be effective work clinicians use ok no these are these are these are what they believe the evidence shows at this point so this is the best evidence that exists as compared to let’s say shoulder positioning moderate evidence with no effect may be effective based on fair evidence inaudible panel recommend I consensus we don’t believe his evidence their shoulder slings moderate of evidence for no effect the v8 didn’t even choose to comment audible panel says recommend but though no evidence so that these statements have to be compared to these statements you know which is uh and then for many it’s kind of like looking at the Cochran review and if you guys have seen the Cochran review in medicine I don’t ever remember them ever saying anything works so I think in medicine where we don’t have the ability to control every variable our ability say something is far more limited and so it’s more of a probability statement if you will so so I come to case in 94 and at that time the surface stimulation studies have been whirring merging and of course all the pain exists so I said well let’s try it and I tried it multiple patients and of course they loved it initially and by the way one Deaton initial surface simulation studies were done they were all done in inpatient rehab when people were able to stay for 6 months 4 months that’s not real anymore you know our patients take a couple of weeks and so all this has to happen now in the outpatient setting well they loved it I showed them how to do it then they call me he’s doctor said this isn’t working anymore a week or two afterwards and of course they Ivan come in and the electrodes are nowhere near where I put them and I asked them so are you doing six hours a day Wow six hours a day that’s s that’s a lot you know maybe an hour so if you were to survey the therapists the rehab physicians this is not being used most of our stroke survivors can not tolerate six hours a day for six weeks of surface stimulation the electrodes reliability is poor stimulation is poor because you need Co to find these motor points and as a result poor compliance and of course doesn’t help to this right now lack of reimbursement for this intervention so not surprising in 1986 one Baker and partner who to poop out the way to the first randomized clinical trial of surface electrical stimulation with shoulder pain said this until implanted electrode systems become available long-term use of surface electrical stimulation can be managed by only a few patients and their families by the way this is the one that actually lets the development of ions I didn’t know that at the time because I was just naive first years that can you’re attending so I said well why not try something different death yes Center has

this percutaneous system if I implant this thing there shouldn’t be any pain was pain with stimulation and of course the electrodes don’t have to be donned and off each time so let’s try it and so I apply for a pepper pilot project grant Bob do you remember reviewing this so Bob review this particular grant and he said okay so I got a little money to do some studies and of course I had no idea how to implant these things so guess who I want to hi Ron Ron Hart taught me how to do percutaneous implants the nice job and that we had this n ps4 which is a pretty heavy box but this is all we had at the time and of course my patients walk and you know this was a problem but nevertheless this from a perspective of feasibility we thought let’s give it a try so our first patient we implanted it what the implantation was remarkably smooth the IRB the IRB environment has changed a lot since that time so I would never do it the way I did it then but clearly here I took an x-ray this is the unaffected shoulder well approximated glowing humeral joint this is the affected side without the simulation with a subluxation and when stimulation is on you reduce the supplication so from a biological perspective it looks like oh this is working well and this patient had no pain I mean that stimulation went on I didn’t this feels great in fact this guy fell asleep with it so our initial suspicion was it was appear to be correct is well tolerated in plantations were all to be easy and it’s appears to be quite effective in reducing subluxation well like in 1996 and I was asked to present at Tom’s applied neural control research day question lis thing I had was this at the time I was the only been case for two years and then I presented this and then just a little short guy from neural control named Zhi ping Feng comes up and he says so John what do you need to make this thing work well you know right now it’s not really a clinically viable system because my stroke survivors walk and they can’t lug around this huge expense this heavy device so I need a small system a little small system like a beeper so I can do that for you and so we’ve got SP they got SBIR funding with met with myself as a co-investigator and develop this little stimulator pages eye stimulator and with that SBIR funding and the addition of the pepper grant would did a small study a kick series it’s before and after trial eight chronic stroke survivors with pain and subluxation and they were stimulated for six hours a day for six weeks without any difficulty we stimulate supraspinatus and post to adulthood why because that’s what surface stimulation study said and David you who I was trained at the time also did some specific muscle selection studies indicating that the post adult oil was actually quite good turns out supraspinatus was not but we did it this way only because that’s how surface simulation studies did it they follow the patients for three months and then of course we test a hypothesis the pain level as follow-up will be lower compared to baseline and what do we find well first of all from outcome measure BPI 15 a brief pain inventory question 15 is simply a zero ten numeric rating scale zero means no pain ten is the worst pain ever but it really simply asked what’s your pain right now and pre treatment was four point eight and post treatment one point nine and stay there at three months and these are statistically significant the other common outcome measure was pain-free external rotation range of motion and this improves significantly from eighteen point five to thirty six point nine but after just after the treatment was done that that value actually declined a little bit which is okay because suggest some system specificity of treatment from a mechanistic perspective we looked at shoulder subluxation using radiographs and there was a ten point three millimeter subluxation at baseline that went down to five and that won’t even further to three point three well this is real or not I don’t know but certainly compared to baseline these are significant we were hypothesizing or speculating that electrical stimulation has an impact and motor impairment or motor recovery and so we de fuga Meyer and that went up in five to six then to fifteen point five which is kind of odd because you still well gee if this is what the stimulation is why is there only such a small change here and a big change there and I don’t know the answer to that question specifically but I can speculate perhaps what this did really not provide any mechanical increase but I simply reduce the pain and as they reduce the pain they use to limb more and that is they use a little more they actually had more function of this improvement in arm function yes these are beyond six months so you wouldn’t necessarily expect recovery but you know what I haven’t seen a single case series that’s been

negative so by definition you know case series are always suspect so this could be our reviewers wishful thinking what could be a placebo effect now after we finished our study our colleagues in Enschede Martinizing they said you know this is pretty interesting stuff you know I think we want to try this I said great so we sent David you over there to train them how to use the implantation but we weren’t involved at all in the study that this is your study we want this to be done independently of us so they did their study with 15 stroke survivors chronic again with pain and subluxation same exact stimulation paradigm six hours a day for six weeks same muscles but instead of three months they chose to go six months and instead of using BPI 15 they chose to use BPI twelve which is actually a better measure BPI 12 says what was the worst pain in the last seven days not just simply a random one time but rather in the last seven days all sure worse pain and furthermore they said we won’t measure quality of life we want to know what the real-life impact of this intervention is in quality of life so their study was at their study was very similar in results I think their results actually a little more dramatic their pain level was higher because they were using BPI 12 which again is at worst pain in last 7 days well sue market was after the explant 12 weeks and 24 weeks their worst pain school was actually better than ours now I can’t explain why that’s the case but there was also quite quite remarkable a little bit different but still consistent with what we found sf-36 is what the user quality of life and as expected the bodily pain component was substantially better compared to baseline other measures were not terribly significant actually this is a baseline I’m sorry right here and so all these measures seem to have improved some but not reach statistical significance but bodily pain and not only reach statistical significance was also maintained at six months when they looked at pain free range of motion that also improves substantially in contrast to our study their pain free range of motion actually was maintained and even got better up to six months their mode of function well their response is more like what I would have expected that is more of an improvement during the stimulation period and then just a gradual well this really isn’t a change you know this is very small changes small changes in fuga Myers so what is real clinically or not it’s not all that clear but statistically they’re significant now there were other preliminary studies by the way I didn’t in the interest of time I chose not to review those so what do we learn from these Colima nary studies well it seems like this is clinically fusable it may reduce shoulder pain it may improve quality of life and it may improve motor function and subluxation and I say these are all made because I alluded to or stated earlier I have never seen a case series that’s been published that’s negative so this may all be false these strokes of eyes preferred the implanted system over a sling it was clearly no discomfort during stimulation it was easy to apply and for those patients we’ve seen a pretty very safe so by this time neural controls is very very excited and so they obtained additional funding SBIR phase two an additional venture capital funding and they go forward toward a pivotal trial the the the meet with FDA and I can’t tell you all the details of what happened but they choose to do a 510 K route and and for you guys who may not be familiar 510 K says this is a fast-track approach where we try to streamline approval devices but the company needs to prove that it is essentially equivalent to a predicate device a predicate device that already exists has already been approved and so again company decision that’s what they choose to do and what I’m going to do now is just simply present that data present a 6-month data and then a 12-month follow-up data as well multiple places of course with a lead Center I think we hit miss Charlotte RIA Charlotte interested rehabilitation I wasn’t was one of our sites as well as Cleveland Clinic Kessler Institute for rehabilitation now Sahni Medical Center New York or IC as well as University Pennsylvania so I’m not gonna give you all the details but here the key inclusion criteria we took chronic stroke survivors some would argue we’re in three months it’s not chronic but I think I’ll show you later as I’ll show you later on most of these patients were quite beyond this time period they have to have clinical subluxation they have to have shoulder pain at least a two on BPI question 12 because BPI question 12 requires them to remember last seven days they also have

to exhibit adequate memory also because many stroke survivors have visual spatial and normalities they also have to demonstrate they can use it easy return numeric rating scale in addition to superspinatus and the poster delta we elected to also stimulate Milla deltoids in the trapezius six hours a day for six weeks this group also received a sling in the control group also control group all day we see it was a sling a primary hypothesis was the participants treated would intramuscular electrical stimulation will exhibit significant greater reduction in shoulder pain compared to controls secondary outcomes or the degree to which paint interferes with daily activity that’s really a misnomer this is really not daily activities actually a quality-of-life measure which I’ll show you a little later we want to look from mechanistic perspective subluxation motor recovery pain free range of motion hypertonia and disability computer controls so we will refer 562 patients 157 were found to be eligible however only 61 chose to enroll primarily because of anxiety of the minimum minimally invasive procedure we randomized 32 the treatment group and 29 to the control group however to study was terminated before full accrual of 66 patients and the reason we was terminated is because neural control was a little antsy about the slow recruitment rate they elected to do a interim analysis on the 61 and realize that even if five additional were in and they were all negative it wouldn’t change the results so they chose to stop the study at 61 instead of 66 having said that I’ll show you the results of course in the end where randomized clinical trial the first thing you do is show us show the audience that the randomization was good and for the most part it was the age demographics gender distribution stroke onset stroke type here in stroke on 623 weeks that’s the average considering there’s 52 weeks in a year that’s quite these are chronic stroke survivors stroke characters are also quite similar as well as used specific medications when we looked at the outcome measures there was a little bit of difference the treatment group actually appear to have worse pain there BPI 12 score seven point six computer six point five and this and this is a trend toward significance BPI twenty three which is the pain interference a party of life was also a little bit higher but not as not as significant as a control group but otherwise all the other measures looked actually pretty similar the pain free external rotation range of motion subluxation if you go my motor assessment spasticity disability and so on this is our longitudinal analysis as you can see here the blue is the intramuscular electrical stimulation the red is the control as indicated the intramuscular stimulation group had a higher pain at baseline seven point five but wouldn’t by end of treatment this came down all the way to three and it was maintained as you can see here the control group just experienced a gradual decline in pain which is what we would expect from a non treated interventional group this graph also shows the results the solid line represents the intent to treat protocol and and the dashed line is the it for protocol so the protocol and tend to treat had very similar results so overall the results look quite good and this was it was significant but in order to address this clinical relevance of an intervention you have to say well what is clinically important and so we came up with a metric where we use what we knew was the minimum clinically important difference in pain studies and that is in a zero to ten numeric grading scale individuals must reduce their pain by two points to be considered clinically important or thirty percent and so we said well if we have a stroke patient reduces their pain by two points and it’s maintained throughout follow-up then we see that’s clinically important and that’s that’s valuable if the patient experiences two point reduction but obviously goes up to loses that then we say there appears to be a short-term effect but no long-term effect and if we’re not clinically important if a stroke patient has does not have to point reduction at end of treatment but does so at twelve months we say well that’s real is top probably not specific to the actual intervention so having said that what is a success rate if you use a two-point criteria or 30% criteria well turns out success rate even for the treatment group is only 63% not a great difference not not a great number not compared to the control board is 21%

this is very very good but from a clinical perspective boy I really wish I could have more than 63 percent success rate if you use the 30% criterion that that number comes down to 56% so statistically significant but from a clinical perspective ours to g4 that invasive procedure I would have loved to have a better success rate secondary measures BPI 23 let me just describe that for a second it’s the degree to which the pain interferes with the following domains general activity mood walking normal work relationships sleep and enjoyment of life now if you can agree with me these domains and not this is not activities daily living this is quality of life and there was a significant reduction in BPI 23 as well however there was no effect in fuga Myer our motor ability test subluxation external rotation range of motion spasticity therapy utilization or medication use which is very perplexing so now all these both groups declined so which explains why the k-series seem to have improved because that’s what they’re gonna do anyway it’s about the control group and a treatment group showed improvements in these measures but when you compare the two together they were not significant in terms of adverse events it was actually quite good there was no infection at a lecture site no broking electro stirring treatment however during removal 656 electrodes broke and it wasn’t a tip actually actually left a fair amount of length behind contrary to what we had thought that maybe only the distal tip would break but it was actually several centimeters that would be left behind infections or granuloma broken electrodes we didn’t see any now at the electrode exit site we did see granulomas but after the lecture to remove these things disappeared and we weren’t concerned about it as expected there was some bruising and swelling at the needle insertion sites and that again resolved without treatment skin irritation from the adhesives was common but again not very significant so overall the adverse event rate was quite good so from my perspective the risks associated with the use of intramuscular electrical stimulation as demonstrated by the randomized clinical trial were minimal adverse events in particular were minor and all resolved without treatment we expect the effectiveness and safety the primary endpoint was met the electrical stimulation reduces hemiplegia shoulder pain and the effect is enduring for 12 months post treatment also improves pain related quality of life appears to be safe and and risk to minimal so our conclusion based on this study was that intramuscular stimulation is effective and safe in reducing hemiplegic shoulder pain among chronic stroke survivors so you would think that maybe this would going well because I don’t know if you all know it’s 510 K in most cases there’s you don’t present data it’s basically a paper argument this is what we have it’s substantially equivalent to what you’ve already approved and so our perspective the time was wow this is overkill we did a randomized clinical trial unfortunately things don’t always work out that way and it was it disapproved and it was just a proof of the following reasons inadequate predicate a sling is not a an acceptable predicate the fact of the matter is what was the predicate what could you use as a predicate and so because of that they said that’s a major reason second you didn’t tell us you’re gonna do an interim analysis and you didn’t tell us you’re gonna do early stop it to the study so you termina study prior to accrual participants so you violated the protocol that next they said there’s inadequate control and accounting for concomitant therapies including pain I agree with them though it’s inadequate accounting for this this is very difficult to account for it’s very difficult to control our patients and what medication they took or for let’s even tell us what medications they took and didn’t take and how much well some do be back to the drawing board so I left there was no control for them to deal with but for me I said you know I gotta move on I gotta ask some more questions why is it such a low success rate and so I took the data that we had in this and logistic regression we looked at only treatment subjects and ask the question what predicts treatment success so a dependent variable was treatment success based on a two point and 30% criterion our explanatory variables will be with independent variables who were the patient characteristics demographics stroke characteristics and when we did this

analysis the only thing that came out significant was time from a stroke if they had this intervention with an 18 month of their stroke that was significant but at beyond 18 months it was different well how so so we do this analysis it’s a little busy slide this represents individuals who receive treatment within 18 months of their stroke what a dramatic drop from 8 all the way down to 1 and they stay there yeah it does doesn’t it in contrast this little dash line are the people who will treat it beyond 18 months and by the way this is the control group right here so this is interesting individuals who were 18 beyond 18 months day they had some sheep in effect at in the treatment but by 3 months oops sorry about that 3 months they were approaching the controls again so it appears that if you’re beyond 18 months and this post hoc analysis you go back to baseline so it’s not enduring but if you less than 18 months the effect appears to be enduring and if you look at from the criterion perspective you see the two-point criterion is a 90% success rate compared to those who are beyond 77 which is 30% and third and 30% reduction was very similar again 30% reduction rate reduction criteria is much more rigorous but essentially the results are very similar well let’s go back to neural control it’s time to regroup so go back to FDA said well how do we do this differently now because you need a predicate I said sir for simulation no because surface stimulation is not an approved modality shoulder pain in spite of all the data by the way they said because they’re looking at what FDA has already approved and FDA has not approved surface stimulation for shoulder pain subluxation what they approved was tense for pain in general so they said tense in fact when you use tense you can’t just simply use a six hours a day they can use it whenever because that’s how it was approved so whereas electrical stimulation you only limit to six hours a day ten you can use forever you can use it for even beyond the study period boy something was not sounding right but this is what they chose to do we chose same study population treat for six weeks follow for six months mainly because we felt like the 12-month data was really superfluous and they they accepted that much tighter control concomitant their appease and looked at weekly outcomes assessment via telephone as verses every three months well we did this for a little while and then another pothole maybe an abyss when a patient died but the 50th patient by the way this patient had electrode related infection this patient failed to return for the first follow-up visit after implantation he was a diabetic he was self treating but did not inform us and this was a different site actually but but this patient did not show up he finally showed up several days after the scheduled appointment in this whole arm B few red purulent discharge the lectures were removed started antibiotics and he was responsive to the antibiotics however hospital day number 3 he collapsed after going to the bathroom and died there was no next of kin and the medical examiner declined to do an autopsy you know actually we actually pursued us quite vigorously and they refused to do it and of course there was no next to Kim so we couldn’t ask a family member and so cause of death is oh no my guess is that he had the infection cause hypercoagulable state you had in my party infarction at a clot in his leg and through a pulmonary embolus just as my speculation at this point so we don’t know why this guy died and maybe nothing I mean that all be related to the infection at all shortly thereafter Invacare who owned neuro control had a internal audit by their accounting and they were told that whatever money you spend in neural control has to be part of Invacare finances until that time it was not and based on that they said we can’t go on but it was really we had

actually had a grant-funded you know 1.2 million dollar grant from NIH who was ready to give to us but this company was no longer and who could not accept we could not accept to accept the grant the company closed in June of 2007 with 56 participants enrolled all lectures removed so yeah I don’t know if this is maybe it’s just real life in science I don’t know but I’m just being honest with you all this is what happened but you know we just can’t die we gotta persevere there’s more to science than having a company fault I had still ongoing questions even if the FDA approved the device with a clinical world except to intervention I wasn’t sure why well the clinical trial shows superiority over a sling there was neither best practice you know was it usual care I think it’s better than what I normally experience but I wasn’t sure if the clinic if the clinical world would accept that and furthermore what about those who are beyond 18 months so with these new set of questions I decided to go forward and so I investigated of course as I indicated before world’s best practice and clearly all best practice parameters suggest that surface electrical stimulation needs to be included but not only that almost a third party exercise it’s not an evidence for therapeutic exercise is not as strong a surface electrical stimulation but in general there was appear to be a consensus the best practice for shoulder pain is hemiplegia was electrical stimulation and therapeutic exercises so in evidence-based review t-cell said is moderate evidence that overhead police worsen pain and moderate evidence that gentle range of motion reduces pain the VA practice guidelines again says avoid overhead police but recommended general gentle lateral range of motion stretching mobilization techniques ice heat soft tissue massage and shoulder girdle strengthening based on fair evidence and this may be effective in reducing shoulder pain the audible panels again also that avoid overhead police recommend gentle range of motion exercise by qualified clinicians all these other approaches were inconsistent throughout three approaches and thus I excluded them from so-called best practice so I said well maybe I should compare to muskie electrical stimulation best practice now so I submit an ro1 application with that in mind chose to enroll chronic stroke survivors who has shoulder pain and subluxation for only for two to 18 months okay just be consistent with our post-op data so the randomized clinical trial one group obviously receiving intramuscular lectric of stimulation the best practice group using it’s receiving both surface electrical stimulation as well as outpatient therapies and by the way this makes a lot of sense because surface electrical stimulation in isolation is not going to work in real practice when it was originally developed it was developing the contents of inpatient rehabilitation which we no longer have now in that inpatient rehabilitation is only three weeks post-stroke our patient therapy however may work very well because these individuals are come in two or three times a week and they will be able to work with a therapist in positioning these electrodes and perhaps changing parameters so I thought you know if surface electrical simulation is gonna work this is the only way it’s gonna work of course this cause an imbalance now how could you give outpatient therapy to best practice and not to mention muskie electrical stimulation so in anticipation of review panels questions well let’s go and use outpatient therapy so that group also received outpatient therapy they were treated for six weeks as per usual and they were followed up for six months and our aim one was looking at BPI twelve EMT was looking at quality of life a name three which I was less concerned about was mechanistic looking at subluxation motor impairment and spasticity respect to expected result we can power this study based on the certainty of respect to the intramuscular stimulation and power to based on our post hoc analysis but what about best practice well I had no idea what best practice would do so I said well in order for this to be clinically implemented it has to be better than best practice it has to be better beyond the so-called minimum clinically important difference which fortunately for pain has been well defined as at least two point or 30% improvement so using that data using the variance that I have from my preliminary data we came up with sample size using alpha 0.05 power of ninety percent and forty five per group and this doesn’t account for the 18% drop I think with a with a drop out I think we said 110 total and so this is what we submitted to the NIH and what did they say so I want your feedback by the way they were

okay with it but no cigar so squirts a priority score two hundred and thirty six percent percentile what were their issues not enough clinical impact what happens the two groups are similar you need to use you a care group in other words you in other words you said well my definition best practice is not real right he does it yes study Care Group yeah and so the end of the doing the historical control it’s funny that one when Steve walked at his study he was criticized for doing a usual care group you know I think it’s maybe this is how we act in study panels whatever you didn’t do you’re gonna be thanked for so anyway that’s that was that then they said well there may be other factors that affect clinical acceptance you know you need a survey assessing the interference of intervention and daily activity and by the way you need to do a cost-effectiveness analysis so one side of the study section was into this clinical impact and then the other side chimed in comparison this is really a comparison of compliance rates it’s not particularly interesting in an ideal environment with ideal parameters to be the same so what you have to do you got to maximize the compliance rate for the surface stimulation group okay then somebody else at too many differences between groups because intramuscular stimulation we’re able to stimulate 12 Hertz constant duty cycle because I didn’t have to change it because patients tolerate that but for the surface stimulation they were going at 25 to 35 Hertz because they can’t tolerate a lower stimulation frequency and either change a duty cycle from a real marginal to high later on so that was a those are the differences so these are the difference between surface stimulation and chomsky simulation is that no no you can’t do that all parameters must be exactly the same by the way you got evaluate mechanisms yeah nothing about mechanisms here so this is a yeah I’m just kind of surprised that it was priority scores only 200 not 300 based on these because these are hard to address and I’ll tell you why the tension the tension that I saw here was different to an explanatory clinical trials and a pragmatic clinical trial explanatory clinical trial focus on efficacy treatment effect in an idealized laboratory environment whereas pragmatic looks at effectiveness treatment affecting anticipated clinical practice explanatory clinical trials folk was an internal validity I control the variables focus on mechanism ie biologic effect pragmatic trial says extra nobility so what often have looser control variables focus on generalizability utility and clinical practice so at the end I asked a question can a face to clinical trial be completely explanatory pragmatic and demonstrate cost-effectiveness well there’s no way you can do this in a face to clinical trial it’s tough on a phase 3 clinical trial now I happen to know all the reviewers I know that because this this particular grant goes to my study section and because I’m a member of that study section they had to put an ad hoc study section that hoc study section had only four members it was only one application so I can figure how who was and and based on their background it was really an interesting experience there’s one basic scientist actually there were three basic scientists and one clinician a very young clinician and so I don’t know how you would address this but let me tell you how I address didn’t let me know how I did so here’s an introduction to my resubmitted application we thank the reviewers for their thorough and thoughtful review of our application by the way you always have to say that well you believe it or not the reviewers raise important concerns are threatened internal and external validity of the proposed study while we endeavor to address both types of concerns these are often inherently in conflict with one another this is well understood we made a concerted effort to reduce threats to Internet validity nevertheless given previous demonstration of the superiority of intramuscular and surface electrical simulation compared to controls and the review in the reviewers emphasis on impact in clinical practice the clinical trials revised to be more pragmatic in nature with higher priority place and effectiveness and external validity I recognize that I can’t address all of them so I had to choose which one am I gonna go and I chose based the the individuals who are pushing clinical effectiveness I thought was stronger in the review and so I elected to go with them it was also easy for me to address

so I revised the application certainly we had intramuscular electrical stimulation with outpatient therapy best practice as per their request I chose to put usual care the problem is what is usual care we have no idea because when you do these surveys they’re all over the board there’s no consensus what you should care is so I said okay while the reviewers suggested how about just getting therapies okay so now I can’t get faulted you know okay you recommend a therapist as well I’ll give you outpatient therapy but the mechanist in there so you’re going to say gee but look at the other groups they have electrical stimulation this group doesn’t not so I said alright I’ll give you electrical stimulation I’ll give you sensory stimulation without motor activation it probably has some therapeutic effect by the way but this is one way to kind of balance the higher emphasis of external validity but also maintain some intern ability as well my expected results what I chose to do is I chose to provide this is the usual care group I chose to give them a clinically important improvement throughout the baseline which is fine which is good because the difference between this and this is so big that the actual power analysis is not significantly affected besides the addition of a third group and because the challenge of additional challenge of additional group I elected to go back to the more traditional power of 80% 70% this came up with 34 per group after you adjust for dropouts it comes out to total hundred twenty-six patients over a five year period so this is one way to respond for clinical impact I chose to add the modified BPI 23 which says I want you to rate the degree to which a treatment protocol interfere with all these things and then there are other factors that they asked for which can’t someone can address some cannot average the event some medical risks are easy that’s part of clinical trial anyway they want us to address the impact of a need for clinician training in clinical acceptance in the West kind of ridiculous you don’t do that at this face you do that later on cost-effectiveness similarly you can’t do this in this face so I gently said maybe in later phase three trials we can do this treatment compliance it’s something that needs to be addressed and this is a quote that I think comes in very handy differentiating efficacy explanatory trials and pragmatic there was a group in this review panel who thought that this was not an important factor in fact this was not interesting at all and so I let’s share this with you it says efficacy studies assess differences in effect between two or more conditions under ideal highly controlled conditions while effectiveness studies assess differences in effect between two or more conditions when used in normal real-world clinical circumstances explanatory trials deal with efficacy whereas pragmatic trials are more closely associate with effectiveness a particular treatment approach might be shown to be efficacious but may prove not to be clinically effective an example this could be an extremely complicated time consuming physiotherapy exercise regime in a university research environment with a participant conveniently selected from a keen student population a financially compensated for the time the treatment works very well however upon transfer to the clinical environment where patients may not have the same time energy users participants may not perform the exercise regime at all the results the treatment was demonstrated efficacy from a highly controlled trial proving ineffective when utilized in a standard clinical situation so this in the surface electrical stimulation in a highly structured idealized environment works well but in real life it doesn’t work at all a pragmatic trial tries to differentiate these effects and I realize that in this particular audience this is probably more of a foreign concept because I think most of the work you guys do is efficacy but when in order to get to the real clinical world you got demonstrate effectiveness and and pragmatic trials are very hard to do as a result so non adherence is actually in major part of clinical practice even if the reviewers have failed to recognize that pragmatic trials ask does the treatment work in real clinical situation or the anticipated real clinical situation we’re for a variety of reasons treatment application a deterrence may not be ideal any effort of maximizing that humans in that clinical tribe must be clinically realistic for example I could have a therapist go to the patient’s home every day to put those electrodes on but what world this does this belong in we can’t do that the only assurance company in the world will be able to pay for that well it’s gonna be willing so that intervention is an efficacy trial

but has no relevance in real life surface electrical stimulation with motor activation can only be done in a context of outpatient therapies versus just simply home alone in the present healthcare environment and so the present paradigm I presented was really try to yes we’re gonna maximize compliance in an environment as clinically realistic now to address that question all this all this all the subjects will have datalogger and they’re stimulated to see what they actually do to see how much an explanatory factor that is stimulation parameters what a hassle so I feel like I had to address it and yet it had to be realistic the frequency frequency was left at 12 Hertz for intramuscular and for usual practice because there we worked elevators just fine but for best practice what I chose to do they’ll start at 12 Hertz which they won’t tolerate by the way and they will be asked to adjust tolerance and to keep the lowest frequency possible the duty cycle for the intramuscular and usual practice will be 20 seconds on 10 seconds off the best practice will start at 10 seconds on texas at 10 seconds off but slowly increased to the point where it becomes 20 seconds on 10 seconds off duration of stimulation with 6 hours for the intramuscular and usual practice for the best practice they’ll start at 2 hours a day at the end of week 1 it will be up to 6 hours but by mid middle of week 3 they have to be seven hours so that everybody gets the exact same amount after a simulation after accounting for the duty cycle so whether they’ll be satisfied by this I have no idea but this is my best effort to make the parameters as equivalent as possible mechanisms and treatment components again pragmatic studies identify utility of a treatment for clinical practice what a sponsor looks at biological effects and the emphasis being pragmatic I basically said this is not the focus of the study pragmatic trials answer questions about the overall effectiveness of an intervention and cannot study the contribution of its different components you would use a pragmatic trial to test an overall package of care including the contribution of therapeutic relationships patient expectations and any specific therapy that is used you would generally compare the effect of this package of care with another treatment not with a placebo nevertheless we did it luck to maintain those three parameters de fuga Meyer subluxation as well as hyper hyper tonyia just to look at what is the general mechanism of how this might work well all this work respect to the NIH that is you know uncertain so I would appreciate your feedback there are cultural issues here and this is not typically what NIH funds but I thought NIH likes novelty so why not cannot contribute after all and I aged a lot of healthy people isn’t it I think we need to be clinically relevant I plan on resubmitting in November fourth based on your feedback I may change a lot of things real quickly what about those beyond 18 months post stroke my recommendation at this point is to pursue a permanent implant muscle based or nerve based and I believe there is a clinical value as well as a market for that as well future directions a lot of questions could remain biomechanical instability of the cost needs to be investigated I’m hoping that Bob Hirsch and his crew would go to that area subacromial inflammation as well as capsule inflammation I believe are important parameters and there are specific treatments for that using intra cortical intra articular subacromial corticosteroids as well as oral steroids subscapularis hypertonia there’s the emerging evidence that neural lysis of a subscapularis may indeed reduced role the pain in certain individuals so there in fact we don’t show it here but we actually done a fair amount of additional work using time to our agency stroke as well for shoulder pain I want to acknowledge these individuals David you who I trained he’s not here with us anymore by the way David is working on the Mayans in Seattle dr. frosty flying in from our Sinai LEL away from Kessler Ellie is probably gonna move on to Utah Martin Eisenman Andrew Seng he’s moved on to University at Toronto which is Roberts was at University of Pennsylvania he’s now at Hopkins and also Zhi ping Fang Maria Bennett Maria by the way that the first case series Maria used as her master’s thesis here as well as Julie grill like to acknowledge the funding this especially this first $10,000 that Bob reviewed that was really where it all started and Oz was to SBIR phase 1 and 2 from ni CHD as well as David used k12 and also the National Center for research resource to G CRC the VA who by the way the VA pay for half my salary when I

first came here to start my career and then neural control corporation I want to formally thank dr. Schaefer bringing this as this some very interesting presentation and this is just an indication early intervention data yes muscles their strength deteriorates very very quickly if they’re not used so this doesn’t surprise me at all then the second question is I guess the experience I’ve had over my career with pmn R is that early intervention is really hard a concept that’s I found very very hard to get over I’m much much more enjoyed working with surgeons that that crowd you can move fast and so I just wonder you know how this kind of thing plays out that seems like there is a cultural issue which you brought up in the end of getting pm’n are to think you in terms of early innervation and it’s it’s kind of perplexes me why they don’t see the muscle deteriorating so quickly within days and why why this would be us not be a surprise that get this thing going earlier and and it should work yeah well you shouldn’t get me started on p.m. and are because I’m probably one of their harsh harshest critics innovation is not their strong point it is a naturally conservative group certain group of people going to p.m and are and and you could and and how they look at science is reflected by that the some would say I don’t fit that mold but I think there’s some hope and there’s especially of late of late there’s a lot of young p.m. in our physicians who are not thinking that way you know for example when I first started residency I had some background in osseous bone formation during my masters and so my first rotation was amputee service and everybody’s worrying about this interfaces between the prosthesis and neski and I said I just been a very naive statement to my attending physician and said why don’t you simply attach these things skeletally so that’s a good idea but we don’t do that we’ll just leave it’s a plastic surgeons and at that point I said boy this is really frustrating and so there is a culture like that that exists now having said that there is an emerging culture where the young physiatrist are becoming much more procedure oriented they they’re not afraid of doing needles doing cuts and and this has been really buttress by the musculoskeletal component of rehabilitation I think the neuro rehabilitation component still probably resides in that area but the musculoskeletal guys are doing spine injections a lot more invasive and so because of that that won’t it jump in a lot earlier that’s one number two the nih truly hasn’t helped on this either because when we try to do something on the acute you know what do they say ah too much variability doing the chronic first okay and so in fact my first clinical trial in foot drop and I proposed a randomized clinical trial in the acute stroke population because I think if there’s going to be neurological recovery using Peroni nurse simulation it’s gonna happen in the acute population not later on Wow they did it because now cutes not too many variability go do chronic in my mind it’s not gonna work but I wasn’t stupid so okay and that’s what I did and eventually got funded on chronic the the effects are small but that’s exactly what I would expected no matter how big your numbers are and so the NIH has not helped us any either the the individuals at the review panels are also conserved and they’re not TM in our Docs you know they’re basic scientists and clinicians and so so there are two things going on I think the NIH doesn’t help in any culture of p.m. it hasn’t helped but I think hopefully both are both are moving and and we were actually able to do some additional studies get additional acute studies funded by making some very good arguments and I think that’s what you need to do one question and a point the question is could you argue that a study could not be effective if there wasn’t intrinsic efficacy so that jumping to a pragmatic study which is what you want to do that if you get a result it would imply that you’re doing something beneficial yes so can you jump to an efficacy study without demonstrating

efficacy and first of all the line between efficacy and effectiveness is not as clear-cut as people like to think the question of efficacy always usually deals with biologic so is there a sub reduction of stop localization well yes and this is kind of trivial I mean if you stimulate it does we do subluxation and in fact when in a saw and people say I have less pain I think we can justify a pragmatic study because there’s already a clinical trial with intramuscular stimulation compared to control us more unlevel efficacy surface electrical stimulators also plenty of studies already showing thats effect there compared to a control group so to some degree efficacy has already been demonstrated before the intramuscular and surface electrical stimulates question now is does it make sense in real life so I think we can make the argument but again because the line between efficacy and effectiveness is not always clear I think I could claim that efficacy stacks have already been done just second question I can’t remember actually a point in in addition to what dr. Mortimer suggests in terms of muscle weakness comparing the newer versus more chronic patients in terms of their response do you think don’t you think there’s also a component of osteoarthritis that’s developed I mean and mutton soft tissue fibrosis mmm-hmm that would be harder to overcome and so they may start to get better or improve somewhat when you stop the stimulation they regress back yeah I believe that’s going to be true and that’s why I don’t recommend this for individuals for 18 months because you have a short term benefit so that’s why I’m recommending actually a permanent implant for individuals with greater than 18 months yes Janice I had two thoughts on they probably aren’t gonna be much help for your application but they could be of some theoretical help or maybe a help for a mechanism study a clinical study in your conceptual diagram I was thinking that you could add something in there and you’ve got what are you already got it’s a very pretty complete conceptual diagram already and you’ve got what happens within the subject or the patient but there’s also a set of events that happens to the subject so the my point put on that topic would be that patients when they’re transferred by a and other helpers right where it says muscle atrophy and weakness and mechanical instability there’s an impact on that joint in that tissue by virtue of the the the patient being lifted by the humerus on both sides and there they’re hanging from their ligaments and they there could be an injury to nerves or our muscles or Bursa at that point and and it could be such a traumatic injury and repeated every day many times a day that it could be very difficult to to address with an in or any intervention no matter how good it is you probably already thought of all this and so one intervention for for shoulder pain and subluxation would really be prevention in clinical practice which might ultimately be easier to address then a treatment after the fact but of course there are other reasons for for shoulder pain ii thought i had that it was all along the lines of what bob was just just talking about that dotted line where the the the effect is is good the pain is going down and then after the treatment stops it goes back up again well it seems to me that what might be happening at least one of the things that might be happening would be that when the electrical stimulation is being offered it is improving the balance forces at the joint and so there’s no impingement on the nerve or no trauma to the to the joint tissue and and but then we we’re not stretching the the tight the tight tissue we’re only stimulating the the weak tissue on one side of the joint so as soon as we stop the treatment if we haven’t stretched the tight tissue so that it doesn’t spring back into its old position as soon as we stop the muscle activation on one side the joint it’s gonna just return to it’s

painful position so that was what I was no I think those points are excellent the in fact one of my first patients a shoulder pain it was an inpatient service and I knew he was subluxed I knew that shoulder was really at risk and he had no pain for a while and then one day I come in he said I got to change this shoulder is killing me and then he said well what happened you know when I woke up this morning my arm was stuck under the under the bar and so the question of trauma is real and that’s why I feel this micro risk for micro-macro trauma it’s very important that multiple ways this happens and one of them is how we manage our patients and so almost all practice guidelines will say proper positioning and prevention unfortunate that’s a very hard intervention to study and how do you do that and how do you look at outcomes and some of the studies that has been done it’s been negative but that doesn’t mean it’s not effective it just means that we just don’t know how to measure that the second comment you made is you’re speaking just like a clinician which is great and what I mean by that is that you know in science as you go more toward efficacy we want to change keep everything the same and change one variable at a time well in clinical practice you can’t do that you know we got yeah you could do multiple things and you’re right so I I think you know shoulder pain can develop into a huge program we’re right now for example I’m investigating not only electrical stimulation but I’m also looking at inflammation in the joint as well as subacromial bursa but not only that I am seeking suspicion the spasticity with peri osteo traction is an important part of the pain generation and so at the end I think what’s gonna happen is that you know I would love to do synergy studies but I need to make sure one one intervention works and then as I then get developed diagnostic diagnostic abilities say well which part is contributing to this is it just mechanical is it spasticity or is it XY and Z and then be able to say well we have treatment for X well cheaper and for Z so I my hope is that as we do this that we have a rational diet not diagnosis specific approach as versus well I should still one thing you know one big shebang we have another grant going in where we’re actually looking at MRIs longitudinally from stroke onset where there’s no shoulder pain to six months to a year where I know that 50% of my patients are gonna develop a little shoulder pain and look at the MRI compared to baseline and compared to the opposite shoulder right now that data does not exist we have data and cross-sectional studies where I take a stroke patient and just look at them as there are and there are a lot of abnormalities but what I don’t know is or any of those abnormality actually causing the pain okay so this is a huge area with very little data and at the end I have to confess why I treat my shoulder patients and I’ve tried everything I’m throwing the kitchen sink at it I don’t care if there’s no data and that’s what I think as a clinician I feel you know the scientists can be out there and and criticize me for doing this but they’re not calling you at night they’re calling me anything it hurts and I gotta do something whether it’s clinically justifiable scientific justifiable or not yes sir good question in your selection criteria for your patients did you allow people who had complex regional pain syndrome yeah I think they’re fundamentally different although I believe they probably start the data suggests that really in complex regional pain that it actually starts with trauma and it’s the the significant subluxation and that probably triggers this cascade of maybe autonomic maybe sympathetic you know nobody really knows anymore people are only safe sympathetic anymore any other questions yes did you hit a 63 percent success rate and you were disappointed because you wanted a a solution for your patients but and yet you you had like seven or eight reasons for a shoulder pain I mean do you and all these questions and your answers are relating to you know you can’t cure them all with this one treatment so right I guess maybe where your expectations really that good or you just were hopeful or or how are you gonna present this you always have to reason to write more grass sorry about that your point is well-taken they majors had a pain syndrome that’s not amenable to that intervention because it’s caused by something different so that’s why we go back and look at it product retro actively a post-hoc because that may still be the case if that’s the case in my my post type analysis not show me anything that is consistent with that

but in this particular case i was really surprised by the post hoc analysis and most post hoc analysis do not this good by the way and so you gotta be always be suspicious always is this real is it not but it was quite dramatic and and there’s some released theoretical reasons why it might work because this is muscle based you know if you if you do it early it makes sense and now they’re also meta-analysis of shoulder subluxation studies and they’re saying it appears to be more effective early than late it appears to be not effective and crime in true chronic business alright thank you very much for great audience