Pre-IDE Submission: Clinical Protocol 
Neuroblastoma Bone Marrow Purging System 
BAXTER HEALTHCARE CORPORATION, HYLAND DIVISION 
6.2 DATA ANALYSIS 
6.2. 1 STATISTICAL CONSIDERATIONS FOR DETECTING GENE MARKERS 
The transduction efficiency for fresh malignant neuroblastoma progenitor ceils by LNL-6 
and GINa is between 0.3% and 3% of clonogenic cells, as assessed in colony assays in 
G418 and by PCR. At present there is no way of determining if clonogenic tumor cells 
and neuroblastoma progenitor cells are identical, or have the same transduction 
efficiency. In the following examples, the transduction efficiency of clonogenic cells and 
neuroblastoma progenitor cells is assumed to be identical, and an average efficiency of 
1 % is used. 
If NC denotes the number of neuroblastoma progenitor cells giving rise to relapse in a 
specific patient, TE the transduction efficiency and MT the proportion of the marrow 
transduced, (in our protocol, 50% of the returned marrow will have been exposed to one 
vector), then under some assumptions the number of marked cells in a sample of size NC 
is binomially distributed. If each of the NC progenitor cells were examined, the 
probability of identifying at least one marked cell (pj is given by the expression: p m = 
1 -(1-Te x MT) NC . pm is the probability that the progenitor cells giving rise to a relapse 
include at least one marked cell. The following table shows values of pm as a function 
of NC. 
Recombinant DNA Research, Volume 16 
[473] 
