Recombinant DMA Advisoiy Committee - 5/30-31/91 
Dr. Krogstad asked what was projected to be necessary to answer the question of the site of 
relapse in terms of numbers of patients and projected time and eiq)ense. Dr. Deisseroth said 
that they had undertaken an analysis of this question from a quantitative standpoint and that 
basically there are as few as 10,000 up to as many as a million marked leukemic cells in the 
marrow and as many as 4 X lO’ nucleated cells. However, there are so many variables that 
contribute to the repopulation with marked leukemic cells competing with other cells that all 
that can be said at present is that cells can be marked. The estimated numbers are sufficient 
to give an expectation that under conditions of reconstitution there is a reasonable probability 
of a polyclonal relapse occurring from the marrow, but it may also occur from the peripheral 
blood. However, the redeeming feature of this type of clinical investigation is that the tools 
are now available to formally address many of the explanations for any given outcome. He 
noted that a few years ago in this form of leukemia they were limited to light microscopy 
which lacked sensitivity and specificity, but that now there are molecular assays which are 
totally specific and possess the ultimate sensitivity, and therefore he felt the time was now 
right to try to answer many of these questions in the clinical investigative setting. 
Dr. Geiduschek said he felt the issue that had been raised regarding the fundamental 
objective of the study was the source of relapse and he said he felt this was an over-statement 
of the aims of the study. He said it seemed the problems with that stated fundamental 
objective were a matter of arithmetic, rather than anything more complicated. He said there 
was a certain inability to be able to write the equation to confirm the source of relapse and 
that he felt the fundamental objective was harmed by the possibility of coming up with no 
outcome because of the inability to determine the source of the relapse. 
Dr. Deisseroth said that the reason they want to do the study is to determine what the 
absolute probability is of there being 10 successive reconstitutions without marking a cell. As 
far as levels of cells in the marrow and the marking frequency he said that what could be said 
is that the probability of relapse occurring from the marrow was now technically capable of 
being detected. 
Dr. Mclvor said that he thought what was being sought was the statistical analysis of the 
problem. He said that if the assumption is made that a leukemic cell infused with the marrow 
has an equal probability of contributing to relapse as a cell that remains in the body, then you 
need the transduction frequency and based on this transduction frequency and the percentage 
of cells that are actually contributing to the relapse you can predict the probability of an 
informative outcome or not. He said this is what was needed to answer the question. 
Dr. Deisseroth presented a slide to summarize this. He said that there were somewhere 
between 10,000 and one million leukemic blasts in the transduced marrow that are marked, 
and the gene marking rate is 3%, therefore the result of this shows that they are at about the 
10% level. However, he stressed that there are many unknown variables to be taken into 
account and therefore this is the reason for going ahead with the investigation itself. He said 
there was no way to know what the outcome would be without doing the experiment. 
[644] 
Recombinant DNA Research, Volume 14 
