Need help reading deep, long, reaearch paper.

[LL60]

Many of you know that I do my own research, and am always looking for the answers to my questions. I hit brick walls all the time, even with trying to deal directly with people who could give me the answers if they wanted to. (Biogen)

There is a wonderful paper I downloaded this morning on figuring a more reliable why to predict PML than what is currently out there. It is 45 pages long and involves a lot of math. A lot!! But it covers many topics such as drug "holidays" and the effects of going off of Tysabri.

One frustration is that this paper, just published in April, STILL stops at the level of 1.5 when talking about JCV+ patients. I just have got to accept the fact that I will never get my question answered: does 4.1 mean your risk is higher, or does it mean that after 1.5 everyone's risk is the same?

Anyway, would anyone like to download and read this with me and help me pull out useful information, and see if we can condense it down to post to help others? Not even sure if that is allowed. I don't see why not, we should all be helping each other. I have learned so much from these boards, but a lot of people with helpful information no longer post.

The paper is titled " Carefully estimating the incidence of natalizumab-associated PML" and it by Julian Borchardt and was published April 19th 2015

[jreagan70]

Hi LL60:

Your interest in the deep science is remarkable!

If your main purpose is just to get this question answered:

does 4.1 mean your risk is higher, or does it mean that after 1.5 everyone's risk is the same?

...the easiest thing to do may be to just email the author of the article and ask him. If your request is compelling enough, and it will only take him a few minutes to reply, he just might do that.

As for the rest of the paper, it isn't just math. It's hardcore statistics , and it will take a hardcore statistician to be able to explain everything.

If you'd like the statistics explained, one approach could be to look for an online forum where statisticians hang out and ask if someone would be willing to go through the paper and explain how the key points were arrived at.

Another possibility is to contact an MS expert such as Dr. Vollmer at the Rocky Mountain MS Center and ask for an explanation. His explanation of the paper could actually turn into an article for him, so he might be willing to answer your questions.

Another possibility is to search for experts' comments on the paper. It isn't unusual for a journal to publish an article along with a commentary on the pro side and a commentary on the con side, which shows that answers aren't always clear cut. Yet another is to search for forums for neurologists or neuroimmunologists and see if there are any comments on the paper there.

The nature of the paper is so complex that only people with experience with statistics should interpret it. (I've taken courses in probability and statistics and I'm not even going to try it.) And for this reason, I think that a group effort to try to summarize it is not a good idea.

I've been reading forums for 10+ years, and I've found that it's the nature of well-meaning but unqualified efforts by individuals or groups who don't have the background to interpret information to result in misinterpretation and postings that are so completely wrong that they're harmful rather than helpful. Just the other day, there was a post on this forum (about a news release) that didn't even get the body parts correct even though they were clearly identified in the short article. That doesn't bode well for interpretation of 40 pages of statistics. I think that's one reason why you've found that people with helpful information no longer post.

You're onto something important here and I commend your interest. It would be awesome if the author of the paper, or another expert, were to answer your questions and you would share them. THAT would be valuable for everyone interested in Tysabri. I know that it would get my attention.

[LL60]

jregan70,

I have emailed the author of the paper, have not heard anything yet.

I have driven to hear Dr. Vollmer speak. It was a very informative talk. He stated that he felt Tysabri and an excellent exercise program was the best chance MS patients had for the best outcome when dealing with MS. Having said that, later in the talk he also shared that he took his JCV+ patients off of Tysabri because he felt the risk was high when there where now other options available with less risk.

The closet "MS specialist" to my area, also follows the same practice and takes his JCV+ patients off.
Both of these doctors, I know, do in fact have patients still on Tysabri being JCV+; so I'm guessing they do take their patients wishes into consideration even though they may feel otherwise. I am my own doctor's only JCV+ patient on Tysabri.

I started rereading the paper last night, line by line. Statistics are there to prove your point and even though they are black and white, they can and are manipulated to prove that point. The math I see now, is to further prove that point and can be skipped over. (at this point in my reading) I am pulling out some useful easy to understand points and will share them when I am finished. Since he does talk about the new white blood cell testing that I had done, I am very interested in going further into that part. Right now, I am starting at the beginning and not skipping around as I did the first two go around. It was only after I had this other testing done at the Mayo Clinic that my doctor stopped trying to talk me into going off of Tysabri.

Dr. Vollmer made a good point in his talk to us; don't expect your doctor to be an expert on your care - things are changing so fast now that we need to be our own expert and then share information with our physician, and they should willing to have that type of relationship with us.

I started Tysabri in the summer of 2009. My risk of PML at that time with what was known was 1:10,000. It is now 1:59. I continue to make my decision on a month by month bases.

I will let you know if I get my question answered.

[jreagan70]

Hi LL60:

Unless and until Julian Borchardt tells you something different, I'm interpreting his paper exactly as written.

In this paper, the upper JCV antibody index interval still stops at ">1.5" because this paper is only recalculating data and risk based on intervals used in a 2014 study that were established using the risk-stratification algorithm from the STRATIFY studies back in 2011-2012. The purpose of the paper was not to formulate a new algorithm or establish new intervals, so those are still the same as in previous calculations.

You'll have to go back to the original studies to find out how the index intervals were decided on. But by convention, the major factor that determines the upper and lower cut-off values for intervals is that -- for whatever reason -- the differences between the data points above and below the cut-off values aren't statistically significant.

Because all of the data above the cut-off point of 1.5 is statistically the same (based on available data), you can consider all of the index values above 1.5 to be statistically the same. That means that an index of 4.1 has the same statistical risk as any other value above 1.5. You might want to believe that an index of 4.1 carries a higher risk than an index of 1.6, but the existing data don't and can't support that belief.

Another factor that contributes to the choice of a cut-off value is how relevant the data/calculations are to the question at hand. In the case of the JCV interval cut-offs, lack of relevance may certainly have contributed to the lower interval. After all, patients who are seronegative for the JCV antibody aren't relevant to a calculation of risk when no one who was seronegative developed PML. Relevance wouldn't logically be a factor in choosing the upper value because a greater risk based on a higher JCV index is absolutely relevant.

In like manner, I'm interpreting everything else in the paper as written.

I'm curious to know if Julian Borchardt says anything different about his paper.

[Teekybird]

very interested in reading the research article

LL60:
Thanks for sharing this with us. I have not yet gone to look for the research paper, but I want to read it for sure. I am interested in seeing the condensed points you make (and depending on when I do read it) I can share my feedback/things that strike me as important as well.

If you do hear anything from the author, please do share.

I've had Tysabri infusion #2. JC negative (at least at initial testing). I get re-tested in August.

Also, thanks for mentioning the importance of coupling an exercise program with Tysabri treatment. This is the area where I am currently slacking off in my battle with MS. It is good for me to hear (again and again) the importance of this. Just this evening, I was thinking "I could go and work out tonight..." I need to make this happen. Thanks.