Looking at numbers…again!

Premise: in my blog, my intention has always been to report on my own experience concerning both health issues and life in general, with some funny stuff thrown in here and there. I have always intended to do this in a straightforward and humorous manner…that’s how I am in real life, after all! And if I make a mistake, or base a post on incorrect information that I found on my own or that I received from someone else, I try to set things right as soon as possible. Well, I made a mistake in my April 16 2009 post. I meant to write a “correction” post earlier, but other things (life!) got in the way…then a blog reader reminded me of my duty just this morning. OOPS!

I owe a huge debt of thanks to that same blog reader who wrote me a message back in early May (!) concerning my above-mentioned post, the one about the risk of progression from SMM to MM. After reading it, in fact, he wrote directly to one of the authors of the NEJM study. He then sent me their exchange, which I, without naming names for reasons of privacy, will post about today, with his prior permission, of course. Since numbers, as those of you who know me well, are not at all my forte, sigh, I will simply post the examples that my blog reader and the expert used:

Blog reader: I have a friend who currently has Smoldering Multiple Myeloma and has had that diagnosis for 10 years. She has been advised that according to your graph – “Figure 2. Probability of Progression to Active Multiple Myeloma…” she has a 66%  probability of developing MM sometime in the future. Further, that as time passes, since the cumulative probabilities are increasing, if she makes it to 15 years, she will then have a 73% probability of progressing to MM and eventually she will have a virtually 100% chance of progression.

I see the data as retrospective rather than prospective. That is, on average, 66% of people with smoldering MM already have progressed to MM after 10 years. As for those who have not progressed at 10 years (about 34% of the original group), an additional 7% in absolute terms or about 21% in relative terms (21% of 34% = 7%)  will progress  – thus arriving at the 73% cumulative probability at 15 years.

I realize the 21% is a theoretical and imperfect estimate since it does not take into consideration many factors such as death from unrelated causes etc. However, would you say that the 66% value in your graph refers to the proportion of people who have progressed to MM at 10 years and is not an estimate of the proportion of people who will progress after 10 years?

The expert: You are correct in your assumption that the 66% value refers to the proportion of people who have progressed to MM or AL amyloidosis at 10 years and is not an estimate of the proportion of patients who will progress after 10 years.  As a matter of fact, the risk of progression after 10 years is approximately 1% to 2% per year.

Our data indicates that the patient is at greatest risk during the first five years and then the risk decreases.  As you pointed out, we are all at risk of succumbing to something as times goes on.

So, the rather dismal statistics in the NEJM study refer only to those who HAVE PROGRESSED from a smoldering state to active myeloma after 10 years. It does not take into consideration those who HAVE NOT PROGRESSED. Indeed, if you have NOT progressed, your risk factor decreases with each passing year, which, by the way, coincides with my original interpretation of the study. In the study author’s own words, As a matter of fact, the risk of progression after 10 years is approximately 1% to 2% per year. Hah.

By the way, another myeloma list member recently left a comment on my April 16 post. Here is her take on the study (my emphasis): 

“The overall risk of progression was 10% per year for the first 5 years, approximately 3% per year for the next 5 years, and 1% per year for the last 10 years.” What this means is that in year 1, 10% of the group of smolderers progressed to full MM, same in each of years 2-5. Then, after that, of those who were left in the group, in each of years 6-10, 3% of the group progressed. Next, again out of those who were left, only 1% of them progressed each year thereafter. So basically, the rule is “the longer you have been smoldering, the more successful you will be at staying in the smoldering state”, statistically speaking and in layman’s terms. You don’t accumulate [ ie. add ] the percentages as you go along. For example, if you make it as far as year 6, then your chances of progressing during year 7 are 3%.

Okay, so, having SMM does not mean that some day you will inevitably progress to active myeloma. Some of us will, some of us won’t.

I hope this is reassuring to those of you who, like me, have been in a stable smoldering state for a while (2010 will be my fifth year as a…smolderer). And, once again, I deeply apologize for having waited so long to post about all this…ah, what can I say in my defence? I must be super allergic to numbers…in fact, here it comes…ahhh…ahhh…atchooooo! (etciù! in Italian…)

8 Comments

  1. Your post is a bit misleading. For one thing, you cannot multiply percentages since a percentage of a percentage is meaningless. 10% of 100 is 10, but 10% of 1,000,000 is 100,000.

    Secondly, the rate of progression is highest in the first five years (five percent per year), three percent for the next five years, and one percent per year for the next 10 years. I’m not clear, but it seems as though the rate of progression doesn’t magically decrease to zero, but remains at the 1-2% per year thereafter.

    So, after five years, 50% of patients will have progress. By 10 years, a total of 65% will have progressed. After 20 years, 75% will have progressed.

    I suppose you could make the case that, living long enough, every MM patient will have progressed, but I’m willing to bet my buttons that that’s not true. There are some patients who will never progress. It may only be 5%, but that is not zero.

    You also need to recognize that the treatment picture is not static (far from it). Newer and better treatments are being developed every day. Soon, the percentages will be meaningless since the disease will be curable.

    Look at your curcumin work and work with other ‘alternative’ treatments. I personally believe that you are delaying the onset of the full-blown disease, and perhaps these people will start having the risk of progression fall into that lower percentage.

  2. Hi Barry,
    You wrote, “Soon, the percentages will be meaningless since the disease will be curable.”
    How “soon” is “soon” do yo think?
    Paul

  3. Wonderful insight (as always).

    My hubby was dx’ed 4/06 MGUS (not SMM but I am hoping a similar five-year highest-risk rule applies…).

    So, it does make sense to do all we can (curcumin) during these inital years… (One argument for early dx & putting up with watchful waiting…)

    One extreme end of that effort to do all we can during the 1st 5 years is that we decided to live on one income (at least for awhile). I love my job (and have the insurance coverage). However, Hubby had been through two very stressful jobs and we believe that job stress played a big role in the shingles/myelitis event that lead to dx.

    So, we cut back & now live on less than half of prior income and are doing fine. Hubby is a stay-at-home Dad for teenager (which for some might be more stressful…) Not saying everyone can do it, but IF it prevents hubby from progressing – it’s worth it!

    Better to pause work for a few years and live longer than never have the chance to use all your retirement savings….

    Another example may be eating vegatarian, organic (or in extreme try Garison) for the 1st five years…

    Just an idea.
    Sandy

  4. Margaret,
    The following is an explanation of why the more dire interpretation of the NEJM article is the correct one. First a little background on statistics: The NEJM article is a study of the risk of progression of smoldering and MGUS patients that have had no chemo or other treatments. So if we ask the question “Do these statistics apply to me a smoldering or MGUS patient with no chemo or treatment?” then the answer is statistically, yes. If this group does not represent you because you are taking cyclopamine like Nick did or you are taking circumin like Margaret is or are on the Gearson diet or some other protocol, then these statistics may not apply to you. If you are on some protocol you should hope that these statistics do not apply to you because you are trying the protocol to better your odds over those that may be doing nothing but watching and waiting which is what we have to assume that most of the patients in the NEJM study were doing (although we don’t know for sure). One of the uses of statistical studies is to try to predict what will happen in one group based on statistics taken from a random sample from a similar group. Exit polls at election time are and example of a random sample of a very small percentage of the electorate to predict what the entire universe of electorate will decide. If the sample is truly a random sample of voters then it does a very good job of predicting the outcome of the entire group of voters based on a relatively small sample. So if we think we belong to a similar group to the patients in the NEJM study then we can reasonably assume that our results will turn out in a similar manner. After all, this is all we have so we have to use this information to the best of our ability.
    Now to the interpretation of the NEJM article. Figure 2 of the study is a graph of the “Probability of Progression”(also referred to as “Risk of Progression” in other parts of the article) vs. “Years since Diagnosis”. This means you can determine the risk of progression of the people in this study if you know the time since diagnosis. So if you are interested in a smoldering patient and that patient has been smoldering for 15 years the graph says that the risk of progression for that patient is 73%. This is a very high risk that the disease will progress to active myeloma. If the patient is a MGUS patient the probability of progression after 15years is 16%. This is low risk of progression to active myeloma. So let’s assume that you have been smoldering for 15 years. You ask the question, “Do these statistics apply to me”? So based on the above discussion on statistics, if you are similar to the people in the study, then statistically the answer is yes and your risk of progression is very high at 73%. If you don’t belong to a similar group then the answer is “I don’t know”.
    Now to all the confusion about the NEJM article. Some like me say the risk of progression of smoldering to active myeloma after 15 years is 73% and some say it is 1 to 2 %. So what is going on? Well the graph of Figure 2 unambiguously says the answer is 73%. Also the following is a direct quote from Dr. Kyle, probably the most authoritative author of the 12 authors of the NEJM article, to a question about the interpretation of the NEJM article data. Dr. Kyle states “You are correct in that 34% of the patients with SMM have not progressed at 10 years and 27% have not progressed at 15 years of followup”. Since we are interested in the patients that have progressed we have to subtract these numbers from 100 and get the result that 66% of the patients with SMM have progressed at 10 years and 73% have progressed at 15 years of follow-up. Please note that his answers of 66% and 73% agree exactly with the 66% and 73% from the graph of Figure 2. So I agree with Dr. Kyle that the correct answer is 73% (a very high risk of progression at 15 years). I think that the other group of people are getting confused because of the language used in the report to describe in words the results of the graph of Figure 2. As so often happens, when we try to put information that can clearly be shown in a graph into words we get in trouble. If words were adequate to describe graphical concepts we wouldn’t need graphs (or pictures for that matter – a picture is worth a 1000 words!). So here is what happens in the report. The report states in the results section “The overall risk of progression was 10% per year for the first 5 years, approximately 3% per year for the next 5 years and 1% per year for the last 10 years: the cumulative probability of progression was 73% at 15 years”. Note that at 15 years we get the same correct answer of 73%! These numbers are handy for your onc. to put into his palm pilot so if you ask him, so Doc, I have been smoldering for 15 years, what are my odds of progressing to active myeloma? He can do a quick calculation of 10% times 5 plus 3% times 5 plus 1% times 5 and get the approximate answer of 70%. If he had the graph of Figure 2 on his palm pilot he could get the exact answer of 73%. So how do some people get and answer of 1 or 2 % for the probability of progression after 10 years? It is because the slope of the graph is 1 to 2 % per year from year 10 to 20. By the way the slope of the MGUS data is also about 1% per year from year 0 to 20. So when the quoted expert says “As a matter of fact, the risk of progression after 10 years is approximately 1% to 2% per year” he/she is in stark disagreement with Dr. Kyle’s answer of 66% at 10 years and the unambiguous data in Figure 2 which is also 66% at 10 years. By the way, if you use the 1% slope interpretation, after 10 years you will get the same probability of progression for smoldering and MGUS, namely 1%. This is clearly wrong as evidenced by the statement in the NEJM article (first paragraph of the discussion) “Smoldering multiple myeloma resembles monoclonal gammopathy of undetermined significance (MGUS) in that end-organ damage is absent, but clinically it is far more likely to progress to active myeloma or amyloidosis at 20 years (a 78% probability for smoldering multiple myeloma vs. 21% for MGUS”. Note that neither answer is anywhere near 1%!
    I rest my case. If you are still in doubt ask your doctor (good luck).

  5. I have to agree that the NEJM study statistics should be read as backward looking statistics rather than foreward looking statistics and thus the odds of progressing do decrease the longer the person remains in the smoldering state. Lets take my case for example. I have been smoldering for 11 years. Yesterday I visited my onocologist and we discussed the data presented in the NEJM article Figure 2 (available at http//:tinyurl.com c9f8lb ). I asked him if the following interpretation was correct. Lets say that 11 years ago myself and 9 other patients that were also just diagnosed with smoldering MM were in a room. We agreed that in 11 years those of us that are still smoldering would meet in this room. So according to Figure 2 of the study, the probability of progression at 11 years is 67% . So in round numbers this means that about 7 out of the original 10 would have progressed to MM after 11 years. So in 11 years, 3 people would walk into the room and I would be one of them. My onocologist said that this is correct. So now we have three people in the room. We then ask the question, how many of us will still be smoldering at 20 years from the original diagnosis and what is the probability that I will be one of them? Again going to Figure 2 the probability of progression at 20 years is 78%. So again in round numbers this means that at 20 years about 8 of the original 10 would have progressed to MM, leaving only 2 still smoldering. So between year 11 and 20 we lost one. So what are the odds that it will be me who progresses to MM between year 11 and 20? Well since there are three people in the room with equal odds, then there is a 1 in 3 chance it will be me. So there is a 33% chance I will progress to MM by year 20 and a 67% chance that I will not and thus be one of the 2 remaining. Again my onocologist said this is correct. He also said that I am the longest smoldering patient that he has ever seen. He also said somewhat jokingly, “if I were ever diagnosed with smoldering myeolma I would call you and ask you what you are taking”. I then asked him if he thinks that the supplements that I have been taking have increased my odds of not progressing. He said he has no way of knowing but I should continue doing what I have been doing. This bring up a question. Would it be beneficial to study smoldering MM patients who have been smoldering for 5, 10, 15 years or longer (those who have beat the odds) and see if they are merely watching and waiting or are on some self directed or other protocol? I think that some good could come from such a study and it could possibly be done on the IMF list serve. Maybe such a study has already been done but I am not aware of any.

Leave a Reply

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