Human Gene Therapy Subcommittee - 4/5/91 
Dr. Epstein asked how many cells are expected to cause the relapse and wanted to know 
if marking 1 in 3,000 cells would detect a relapse. Dr. Deisseroth replied that from the 
most conservative estimates, based on the observed responses to chemotherapy which 
result in a decrease in both the blast crisis and the early indolent form of the leukemia, 
it is believed that the number of blasts that will be marked in the marrow at the time of 
storage will be between 40,000 and 4 million. At the very lowest limit, there are at least 
40,000 marked blast cells in the patient's marrow at the time of collection, each one of 
which has the capability of growing up. Therefore, the probability of seeing a polyclonal 
relapse is very high, as defined by multiple retroviral integration sites, if the marrow is 
the origin of relapse. 
Dr. Epstein asked if Dr. Deisseroth thought a large number of those tumor cells would 
contribute to relapse. Dr. Deisseroth replied that leukemia is a disease which differs 
from neuroblastoma in that most of the cells are dividing all of the time, and blast crisis 
CML is composed of cells that are dividing. He agreed with Dr. Parkman that there are 
cells in the indolent, non-dividing population which may ultimately contribute to relapse. 
If one were to assess probabilities of representation in the relapsed population of 
indolent versus dividing cells, the cells that are dividing at the time may have at least 
equal or greater probability of contributing to relapse. 
Dr. Mclvor asked if CML comes from a primitive stem cell. If it comes from a primitive 
stem cell, then those populations will have to be marked before the protocol can 
succeed. Dr. Deisseroth explained that if they were studying the indolent form of the 
disease, the chronic phase, to distinguish where relapse would occur, they might have to 
isolate the progenitor cell of the leukemic population to assess marking. Because the 
very late stage of the disease was chosen for study, one is looking at cells that have 
already evolved from an early stage progenitor. Thus, the necessity of actually proving 
one can mark that very early progenitor cell really doesn't apply. 
Discussion ensued as to the possible clinical outcomes of the therapy and the ability to 
discriminate between them. Dr. Deisseroth emphasized that the most important point 
focuses on the ability to mark leukemic cells that can be detected at the time of relapse. 
This is an informative result that can permit the development of beneficial therapy. If 
there is evidence of marrow contributing to systemic relapse, one will then have the 
means to deal with that problem decisively. It may be of benefit not only for CML but 
for other forms of leukemia as well. 
Dr. Mclvor requested some information regarding the preclinical data and how the 
marrow was going to be handled large-scale. He asked how many marrows containing 
CML had been transduced to date. Dr. Deisseroth replied that five marrows had been 
transduced on the diagnostic scale and none on the preparative scale. Dr. Mclvor 
wanted to know what proportion of the 10-20% of the colonies that were drug-resistant 
Recombinant DNA Research, Volume 14 
[587] 
