Changes in scoliosis treatment due to Scoliscore

[Pooka1]

and I'm not absolutely sure now, but I believe I have read in a post, not precisely a good comment from Dr. McIntire.. but of course the best would be to ask him, maybe I'm wrong.

I think I remember that. If I am remembering correctly, I think he was making a point about PT being okay to suggest when there is no evidence for bracing efficacy during the watch and wait period. If that was his point, I don't disagree but ONLY if the kid wanted to try it and knew there was no good evidence of efficacy. Hard bracing is not ethical during the watch and wait period in my opinion due to the vast majority of those kids either not needing bracing or not being helped by bracing... some 90% perhaps. And the other 10% may only find that their progression was delayed by bracing. Who knows.

[Kevin_Mc]

I have widely replied before to all what are you saying now, but about W&W I remember that in the first or second scoliosis article I have read, researchers did a hard and of course well founded criticism to it; I'll try to find it.. and I'm not absolutely sure now, but I believe I have read in a post, not precisely a good comment from Dr. McIntire.. but of course the best would be to ask him, maybe I'm wrong.

I think I remember that. If I am remembering correctly, I think he was making a point about PT being okay to suggest when there is no evidence for bracing efficacy during the watch and wait period. If that was his point, I don't disagree but ONLY if the kid wanted to try it and knew there was no good evidence of efficacy. Hard bracing is not ethical during the watch and wait period in my opinion due to the vast majority of those kids either not needing bracing or not being helped by bracing... some 90% perhaps. And the other 10% may only find that their progression was delayed by bracing. Who knows.

If I'm understanding what I think you two are saying; Yes, I would propose PT during the W&W time period. In fact, I would predict that this would be the best time to do exercise as it would allow the best biomechanical environment for the spine/muscles/tendons/etc... There aren't huge, irreversible changes made to the structural components of the spine (assuming the heuter-volkmann principle and not some genetic growth issue), and so it would provide strength/exercise the best possible scenario for success. The larger the curve gets, I feel the less effective exercise would be. Or at least, the more complicated the protocol would need to be.

This would also be what I would propose to any surgeon that would allow me to recruit from their practice. Basically, let me work with every W&W patient, be it personal weekly training or home-ex. If, after 4 months of training, the curves have progressed, proceed with whatever standard of care the clinic usually prescribes. If they haven't progressed, then they continue training. Something like trunk rotational strength training is perfectly suited for this type of study. It's cheap, easy to train and provides the family with something concrete to do, which would improve patient satisfaction. It would probably take a couple of years to get enough numbers in each group to have any meaning since, as is noted, most of the W&W patients don't need any form of treatment anyway. But, if you use those who decline to participate in training, i.e. W&W group, then you have your control and 2 forms of exercise methods..... At least that's sort of how I see any pilot study getting off the ground.

[Pooka1]

Yes I think it is ethical to propose PT during W&W whereas it is NOT ethical to propose hard bracing during that period.

I think you have good reasoning for PT during that period though you would need huge numbers because your baseline for non-progression is going to be the large majority of your study population. You will be looking for small numbers against a huge background which is always challenging. You're essentially competing with the treatment consisting of sitting on the couch easting ice cream which will have at least a 75% efficacy rate with curves <25*.

[Kevin_Mc]

Yes I think it is ethical to propose PT during W&W whereas it is NOT ethical to propose hard bracing during that period.

I think you have good reasoning for PT during that period though you would need huge numbers because your baseline for non-progression is going to be the large majority of your study population. You will be looking for small numbers against a huge background which is always challenging. You're essentially competing with the treatment consisting of sitting on the couch easting ice cream which will have at least a 75% efficacy rate with curves <25*.

Yep... Although, the other part of the hypothesis for strength training is that it can/would stop all progression. Not just the ones who would progress to surgery. But including those patients who would never progress beyond 30° (for example). So it's not just avoiding surgery, it would be maintaining as small of a curve as possible. And since the W&W period patients generally have curves in the 11°-25°, you wouldn't have such a large range of curve sizes. So, in theory, differences in average curve size could be a bit better identified. But, I also don't really think group curve size is the most appropriate stat to run. But that's a different discussion.

continued...

Does the scoliscore test change those number requirements at all? High scores with no progression must mean more than just any old no progression.

So this gets to the Scoliscore issue. I've said before that what I fear most from the Scoliscore is that all studies going forward (if not retrospectively) would be discounted if you don't show what the Scoliscore for each patient is, i.e. if they aren't at a high risk for surgery, then you haven't proven anything. But in thinking about this (thanks to this thread), avoiding surgery is definitely the primary goal. But I think keeping the curve size small is not just possible but potentially even likely in those with a lower Scoliscore. In fact, screening patients with a high scoliscore out of studies (or at least stratifying that way) would be potentially very valuable. The goal being that you can be screened for surgical risk, and then if that's low, then there's this additional set of exercises (or insert other conservative therapy here) and we can keep the curve as small as possible. OTOH, and more to your point, yes, if you have a group of patients with a score >150 and any treatment keeps that curve <35°-40° then I think you've really proven the effectiveness.

In designing a study, I doubt the study could pay for the testing unless there were a deep research based discount. Perhaps some type of collaboration. The other alternative is to have that be purely a clinical decision and present whatever scores are available.

[Pooka1]

Kevin_Mc,

I think you are roughing out a pretty good study. It would be very good if it could get off the ground. Maybe once you finish your post doc... :-)

And I agree group curve size is not going to be a fine enough point... you will need to sort out curve types... like Katz et al. should have done. :-)

You are also going to have to beat the numbers from the Soucacus (sp?) Greek study showing a surprising percentage of kids with small curves who spontaneously improved and even resolved completely compared to the number that progressed. That is also going to raise the baseline in terms of claims about regression of curves and complete regressions due to TRS.

I agree that if 100% of the kids with a Scoliscore >150 will reach surgical range in adolescence then that is a good test population for conservative treatments. It might actually be the only legitimate test population. I am trying to remember... what percentage of kids >150 reached surgical range? Was it really close to 100%?

[LindaRacine]

Yep... Although, the other part of the hypothesis for strength training is that it can/would stop all progression. Not just the ones who would progress to surgery. But including those patients who would never progress beyond 30° (for example). So it's not just avoiding surgery, it would be maintaining as small of a curve as possible. And since the W&W period patients generally have curves in the 11°-25°, you wouldn't have such a large range of curve sizes. So, in theory, differences in average curve size could be a bit better identified. But, I also don't really think group curve size is the most appropriate stat to run. But that's a different discussion.

continued...




So this gets to the Scoliscore issue. I've said before that what I fear most from the Scoliscore is that all studies going forward (if not retrospectively) would be discounted if you don't show what the Scoliscore for each patient is, i.e. if they aren't at a high risk for surgery, then you haven't proven anything. But in thinking about this (thanks to this thread), avoiding surgery is definitely the primary goal. But I think keeping the curve size small is not just possible but potentially even likely in those with a lower Scoliscore. In fact, screening patients with a high scoliscore out of studies (or at least stratifying that way) would be potentially very valuable. The goal being that you can be screened for surgical risk, and then if that's low, then there's this additional set of exercises (or insert other conservative therapy here) and we can keep the curve as small as possible. OTOH, and more to your point, yes, if you have a group of patients with a score >150 and any treatment keeps that curve <35°-40° then I think you've really proven the effectiveness.

In designing a study, I doubt the study could pay for the testing unless there were a deep research based discount. Perhaps some type of collaboration. The other alternative is to have that be purely a clinical decision and present whatever scores are available.

Kevin...

The ScoliScore folks are interested in some collaborative studies. They've approached the BRAIST people, but haven't been able to work anything out. There is actually little to no cost for the ScoliScore test for the vast majority of patients. Pricing is based on a sliding scale. Everyone who has a household income <$450K per year gets some discount. Most patients pay nothing. If you're interested, send me a PM, and I'll send an introductory email to Ken Ward.

--Linda

[Kevin_Mc]

Kevin_Mc,

I think you are roughing out a pretty good study. It would be very good if it could get off the ground. Maybe once you finish your post doc... :-)

And I agree group curve size is not going to be a fine enough point... you will need to sort out curve types... like Katz et al. should have done. :-)

You are also going to have to beat the numbers from the Soucacus (sp?) Greek study showing a surprising percentage of kids with small curves who spontaneously improved and even resolved completely compared to the number that progressed. That is also going to raise the baseline in terms of claims about regression of curves and complete regressions due to TRS.

I agree that if 100% of the kids with a Scoliscore >150 will reach surgical range in adolescence then that is a good test population for conservative treatments. It might actually be the only legitimate test population. I am trying to remember... what percentage of kids >150 reached surgical range? Was it really close to 100%?

Yeah, after my postdoc...:> I'm working on my next grant/project now and am hoping it will lead to a move back east (Raleigh or Atlanta).

Basically, the only way I would ever really see something like this gaining enough patients to have any statistical power would be to change the way a particular practice operates. And I feel this way for most clinical studies. Clinical practices that want to 'do' research need to look at every patient as data. And I don't mean that to sound insensitive. But there are things that could be measured in every patient population that clinics don't measure but could be potentially useful. To use AIS as the example. On top of x-ray and surface measurements, a panel of strength tests or proprioception or quality of life or cytokine or homonal levels etc..., could all be collected. This is extremely idealistic of course and there's always the issue of cost. But had the practice where I did my grad work allowed me the access to at least strength test their patients, the amount of data would have been huge. The MD was frequently at least an hour or so behind and my testing took ~20 minutes. There are certainly ethical and analysis considerations. Sometimes I think that since it's nearly impossible to collect the perfect patient population for a study, then we just need to stop trying and get as many people into a study as possible. Present the data as a particular practice's outcome instead of a particular group of patient's outcome.... Lot's of problems with that I know. And sort of thinking out loud and expressing the reality that something in clinical research has to change.

I just picked 150 out of the blue. But the article uses 180 as the cutoff point for high-risk. Also Ward says that the score can be used as a continuous variable. So, the higher the number the higher the risk. 99% with >190 progressed to a severe curve. Based on their chart, the score and risk are exponentially related, not really linearly related. The risk level hovers <20% until the score reaches ~160. Then it takes off pretty steep.

Kevin...

The ScoliScore folks are interested in some collaborative studies. They've approached the BRAIST people, but haven't been able to work anything out. There is actually little to no cost for the ScoliScore test for the vast majority of patients. Pricing is based on a sliding scale. Everyone who has a household income <$450K per year gets some discount. Most patients pay nothing. If you're interested, send me a PM, and I'll send an introductory email to Ken Ward.

--Linda

That's good information to know and is actually quite impressive of them.

[rohrer01]

This may sound really dumb, so ignore my ignorance if it is. I realize that the scoliscore is aimed at adolescents. However, our genetics don't change as we age. Could you possibly get a "control" group from adults that were in the watch and wait category since their outcomes are already known? Give them a scoliscore, say, "after the fact"? Then you don't have to find a group not willing to do PT or bracing or any other intervention to get your control. I know it is difficult to find people out there that have all of their films from adolescence to adulthood, but they are out there. That would possibley remove any ethical violation of witholding treatment from the younger generation of kids.

[Pooka1]

That would possibley remove any ethical violation of witholding treatment from the younger generation of kids.

With bracing, it is already known that the largest ethical violation occurring NOW is overtreatment. Vast amounts of overtreatment. We know this from several independent lines of evidence.

There is no good evidence that anyone would be harmed by withholding brace treatment, especially in the long run.

[LindaRacine]

This may sound really dumb, so ignore my ignorance if it is. I realize that the scoliscore is aimed at adolescents. However, our genetics don't change as we age. Could you possibly get a "control" group from adults that were in the watch and wait category since their outcomes are already known? Give them a scoliscore, say, "after the fact"? Then you don't have to find a group not willing to do PT or bracing or any other intervention to get your control. I know it is difficult to find people out there that have all of their films from adolescence to adulthood, but they are out there. That would possibley remove any ethical violation of witholding treatment from the younger generation of kids.

Hi...

Not sure I follow your point, but the ScoliScore database was built from adults whose outcome was already known.

--Linda

[Kevin_Mc]

Hi...

Not sure I follow your point, but the ScoliScore database was built from adults whose outcome was already known.

--Linda

Now I'm not sure I follow you. Is the database different than the data presented in the article? If so, then it was a retrospective study? The methods say they recruited 9-13 year olds and followed them through until an end-point was reached.

[Pooka1]

I remember some confusion about some dataset, at least on my end!

It was that discussion about dependent and independent variables and having trouble understanding their curve. I thought there might be three "things" going on. I offered that it was a calibration curve of some sort and not a more typical data curve. (In bench chemistry, you typically don't need to show the calibration curve.)

I'm not saying that was a calibration curve but I am saying that if there was really an adult population used to develop the "calibration" that was later used with a population of kids, then that might make some sense in light of the foregoing discussion. Or not! I am not clear on what exactly they did calibration and testing wise.

[Pooka1]

The other thing they were doing with adults is determining the genetic distance or something. They showed that people with IS were much more closely related than random people, strongly implicating a genetic trait.

[LindaRacine]

My understanding is that they used a very large database of adults on whom they had medical history. Then, they asked adults with scoliosis, whose outcomes were known, to add their genes to the database. I don't have a copy of the study being discussed, but I've heard the Axial Biotech pitch several times, so I'm relatively familiar with the data. As usual, before they started marketing the service, they took half the database, and scored it. Then they used the other half of the database to validate it. The validation charts are very impressive. For the tests that have been done to date, the only patients whose predictions have failed (I.e., the patient had a low score, but their curve(s) progressed significantly), have all had an underlying, unreported issue (e.g., undiagnosed congenital vertebra).

I'm pretty skeptical by nature, and almost always resistant to sales pitches, and have a low bullshit threshold. My bullshit meter never went off with the AB guys. I don't know what their motivation was in the beginning, but it appears that they'll never be profitable. But it appears that they're going to stick with it, even though they suspect there won't be any more money coming from investors, because they feel there's a huge benefit to society.

[sjmcphee]

I don't know much about all this scoliscore topic.
My personal opinion is that the genetic work in necessary, but lets not forget the fact that it's a diagnosis tool and not a treatment.
I'm not going to argue with 99.99999 % whatever the hell DNA accuracy is.

I have however owned shares in companies before.
And Linda's statement saying that AB will never be profitable is enough to set alarm bells ringing.
Why won't AB ever be profitable?

Your whole health system in the USA is different to ours, in our country the government pays for peoples health and medical costs - they provide the hospitals.
So in our country if the government can save money of the cost of health it will.
If there was a benefit to patients and it saved money overall theres no reason why they wouldn't use it.
They would pay for it.

So why are investors not interested in AB and Scoliscore if it is all its supposed to be?
If investors aren't interested in Scoliscore, then that's enough to set off my bullshit meter.

- Scott