Recombinant DNA Advisory Committee - 5/30-31/91 
patients with the disease and they have a very distinctive morphology. They look different 
from normal hematopoietic cells in that they are larger and have neurofibriles which extend 
from them. They can also be identified more definitively^ vitro, by use of monoclonal 
antibodies. Secondly, they are able to be marked and grown on selective media. Dr. Brenner 
said the efficiency of marking is rather lower than in AML» at 1-3%, with a wide variation of 
0-14%. This can be confirmed by PCR analysis. He said he was reasonably confident that 
neuroblastoma colonies growing in selective media could be marked pre-infusion. However, 
he noted that at the 1-3% transduction efficiency level, the question was whether these could 
be detected subsequently after relapse in the patient. He said this was contingent upon how 
many cells were causing the relapse. He said that if only one cell were causing it, there would 
only be a .3% chance of detecting a marked cell in the patient. However, if hundreds or 
thousands of cells contributed to the relapse, then there would be about a 95 percent chance 
of detecting marked cells in the patient. He noted that even with a .1% transduction efficiency 
level, if 1,000 malignant cells remained behind in the remission marrow, then the investigators 
would have a 95% chance of detecting a marked relapse within the first 3 patients who 
relapse. 
He said it was important to know how likely it would be that this number of malignant 
neuroblastoma cells would be left behind in the remission marrow of the patients. He said 
that by looking at AML in which the ability to detect residual malignant cells is more 
sophisticated and more sensitive than for neuroblastoma cytogenetic methods only allow for 
the detection of 2-5% of residual blasts. However, fluorescence microscopy allows the 
detection of from 1 in 1,000 to 1 in 10,000 malignant cells. This would mean that a 20 
kilogram child could still be receiving at least 20,000 blasts in a marrow reinfusion since 100 
million marrow cells are infused per kilogram. He said that as more sensitive methods have 
come into use for detecting minimal residual disease in marrows that are said to have been 
in remission that when they have been examined retrospectively they have turned out to be 
in florid relapse. Therefore, he said that it was likely that most of the remission marrows will 
have a substantial number of contaminating neuroblastoma cells and that a 1-3% transduction 
efficiency would be adequate to detect in subsequent relapse. 
Dr. Brenner said that since as few as 1,000 neuroblastoma cells can be detected in a remission 
marrow if there is a marked relapse, and since the most sensitive technique now available will 
detect 1 in 10,000, this technique will allow the investigators to evaluate the efficacy of purging 
and determine whether or not this risky procedure should be carried out. 
Dr. Atlas said that he would first relay the comments of Dr. Kelley to the committee. Dr. 
Kelley had said that he felt Dr. Brenner had responded to the provisions of the HGTS and, 
while he was not judging how he would vote after hours of discussion which may ensue, his 
comments were indeed favorable. 
Turning to his own review. Dr. Atlas said that he was somewhat confused as to whether or 
not there had been an adequate response to the first provision set down by the HGTS that, 
in fact, there be further evaluation of the procedures for in vitro bone marrow assays to detect 
[648] 
Recombinant DNA Research, Volume 14 
