14 replies to this topic

[Sneezy12]

  My recent bcr/abl results on the Int. Scale after 2 yrs. on 400mg.of Gleevec are:

   11-2-09-0.2%

    1-18-10-PCRU

     2-1-10-0.01%

     5-10-10-0.03%

      8-17-10-0.02%

      8-26-10-0.0003%

      11-15-10-0.008%

       2-21-11-PCRU

       5-2-11-0.01%

Will someone calculate the Log increase for me and show me how to do it for future reference. Is it significant and what should be the next step? Thank you for this and all the great information on this site! Frank

[jjg]

Hi Frank,

A log (logarithm) is really just a way to re-scale numbers. Why the medical profession wants to use a log scale to describe results to non mathematical people is a complete mystery. Using more simple ratios is I guess just too simple!!!

If you are into maths then an explanation of the log scale (showing some really good reasons for using a log scale) is here:

http://en.wikipedia.org/wiki/Logarithm

A log scale can be based on anything, so it is technically wrong to talk about a log scale without saying what base you are using. In the CML context a base of 10 is being used. A natural base (e) is the one I use most but base 10 is also pretty common. If you are using a standard scientific calculator (iPhone users rotate your screen to landscape) base 10 will probably appear as "log" and base e will appear as "ln". I would like to write log followed by subscript of 10 but the formatting in here is not agreeable so I will write log.

The log scale is being used here to make the ratio of your PCR results more compact (& possibily harder to understand). Your log reduction is given by the following:

Log change from dx = log (PCR at dx / PCR now)

Log change from some other time point = log (PCR then / PCR now)

If your PCR at dx was 100 and your PCR now is 10 then, log change = log (100/10) = log (10) = 1

If your PCR at dx was 1 and your PCR now is 0.1 then, log change = log (1/0.1) = log (10) = 1

So as you can see the log is just another way of looking at the ratios of you PCR results.

If the ratio is greater than 1 (i.e. your numbers are decreasing) then the log change will be positive, or a log reduction

If the ratio is 1 (i.e. no change) then the log change is zero

If the ratio is less than 1 (i.e. your numbers are increasing) then the log change is negative, or a log increase

A shift of one decimal place, or one order of magnitude is 1 log (base 10) change. A shift of 2 decimal places, or two orders of magnitude is a 2 log change.....

If you look at your most recent PCR of 0.01 compared to your 2/09 PRC of 0.2 then the log change is log(0.2/0.01) = log (20) = 1.3 (positive number representing a log reduction)

Unfortunately (or fortunately for you) your 2nd last result is a PCRU i.e. 0 so if we did a log of zero it would be negative infinity i.e. you can't calculate a log increase using a PCRU

If you compare you 3rd last result with your last result the your log change is log(0.008/0.01) = log(0.8) = -0.097 (negative number representing a log increase)

What exactly this means I am not well placed to say, but I do note that you went from PCRU to 0.01 before and then back to PCRU again.

Best wishes.

Josie

[Sneezy12]

Josie, thank you so much for the lucid and comforting explanation. I'll sleep better tonight! Regards, Frank

[Trey]

Frank,

PCRU is roughly a 5 log decrease (depending on the lab), and .01 on the IS is -3 log.  That is up 2 logs.  You should repeat the PCR.

[NOTE: THE .01 SHOULD ACTUALLY BE .1 FOR MMR, AS NOTED IN RESPONSES BELOW -- THANKS GUYS]

[Sneezy12]

Thanks, Trey

[Sneezy12]

Josie, BTW, If I divide 0.008 by 0.01, and get 0.8, what Table do I use to obtain the number of  0.097? My Log Table only starts with a number of 1.0? Thanks again, Frank

[jjg]

I'm too young for log tables :-) i.e. too hard

You could try this online calculator:

http://www.rapidtabl..._Calculator.htm

[Sneezy12]

Josie,

Thanks again! Regards, Frank-"the old-timey log table user".

[hannibellemo]

Trey,

I'm confused now. I was told that .2% on the IS is CCyr; and .1% (not .01%) on the IS is MMR, or roughly a 2 log and 3 log reduction respectively, but perhaps it's not possible to make that comparison?  I am a .05% IS.

This makes my brain hurt!

Pat

[PhilB]

Okay, you've lured me back out of the woodwork.

International scale has 100% as the base so:

1 log = 10%

2 log = 1% (generally taked as roughly equivalent to CCR)

3 log = 0.1% (MMR)

4 log = 0.01%  (where sneezy is)

Taking Trey's figure of PCRU being 5 log that would mean a 1 log increase'  Depending to an extent on the sensitivity of the particular test used it looks as though Sneezy is possibly just bouncing around the limits of detection.  Personally I wouldn't be getting unduly worried (oh yeah? who am I kidding?) unless the next PCR shows further increase.  You also have to consider the entirely possible scenario that the PCRU was a 'false negative' - in which case you are holding steady at a very nice 4 log reduction.

Phil

[Trey]

PhilB,

Yippee, waaaahhhhoooo -- if I had only known a typo could lure you back I would have made a gazillion of them.  MMR is indeed  .1 on the IS scale.

Glad you came back into the fray.

[PhilB]

Hi Trey,

I had been having a very nice 'CML holiday' - pretend it's a vitamin pill and ignore the disease.  The NHS did its best to  remind me by making me battle with the fact that as I'm in MMR I can only see the consultant every 6 months, but she's only allowed to write a 3 month prescription, but I somehow got through that with my equanimilty largely intact.

Today I had my routine medical which put me into health mode and, later on, as my computer was largely tied up running some macros, I thought I'd log on whilst I was waiting to see what you were all up to and was caught by the trap carefully laid by Pat and your good self.

On the train home afterwards the doctor who'd done the medical phoned my mobile to let me know that she was referring me to a prostate cancer specialist based on my PSA results.  I think the holiday in Pat's realm may be over for a while.  Ho hum.

Phil

[hannibellemo]

Trey-

High 5!

Pat

[hannibellemo]

Phil,

I sent you an e-mail!

PSA numbers and what they actually mean are so under fire right now, but I'm glad you're going to get it checked out.

Pat

[Sneezy12]

Phil-I had the same problem. Keep us posted.  All will be well. Frank