Pre-IDE Submission: Clinical Protocol 
Neuroblastoma Bone Marrow Purging System 
BAXTER HEALTHCARE CORPORATION, HYLAND DIVISION 
marrows which are positive by immunocytology before purging. As indicated 
previously, it is unlikely that infiltration will be high enough to show 3-3.5 logs of tumor 
removal from any marrows, and unlikely with this sample size, that demonstration of 
removal to undetectable levels in immunocytologically positive marrows will provide a 
statistically adequate basis for analysis. There are not enough previous data from 
clonogenic assays to predict what the relationship will be between detection of 
neuroblasts by immunocytology and clonogenic assays. 
6. 2. 3. 2 Efficacy of Purging 
An aliquot of marked marrow will be purged and mixed with a second aliquot of 
unpurged marrow in which the tumor cells were marked with a different gene. Both 
aliquots will be used for autologous transplantation. If relapse occurs, the neuroblastoma 
cells will be separated from the marrow by flow cytometric cell sorting of cells that are 
negatively stained for the CD34 and CD45 antigens (detecting marrow progenitor cells 
and leukocytes respectively), and positively stained with the UJ13A monoclonal antibody 
(directed against the neural cell adhesion antigen found on neuroblastoma cells). The 
polymerase chain reaction (PCR) and restriction fragment length polymorphism assays 
(RFLP) will be used to determine whether the sorted cells contain the marker genes, and 
whether both markers are present. Samples of marrow will also be grown in clonogenic 
assays for neuroblastoma progenitors, and colonies analyzed for the presence and 
frequency of each marker gene, to determine the contribution of the purged and unpurged 
aliquots to relapse. 
There are no data available on the number of progenitor cells involved in neuroblastoma 
relapse. However, based on statistical considers ions, utility of gene marking to assess 
the contribution of transplanted marrow to relapse requires that relapses be polyclonal. 
As indicated in Table 6.2-2, if only one cell gives rise to a relapse then the liklihood of 
detecting marked cells is negligible. If however, 100 or more cells give rise to a relapse, 
then there is a high probability that marked cells would be detected. Thus the detection 
[478] 
Recombinant DNA Research, Volume 16 
