Pre-IDE Submission: Clinical Protocol 
Neuroblastoma Bone Marrow Purging System 
BAXTER HEALTHCARE CORPORATION, HYLAND DIVISION 
relapse. For this study to be useful, however, polyclonal relapse is required. The 
marrow relapse rate (pr) in patients receiving ABMT for advanced (Stage D) 
neuroblastoma exceeds 60% at 12 months (Anderson et al, 1991). If this rate of relapse 
is assumed, as well as the previous assumptions of an average transduction efficiency of 
1%, and an equal number of cells contributing to relapse in all patients, Table 6.2-2 
indicates that it is unlikely that marrow relapse from ^ 10 progenitors would be detected 
in this study of 12 patients. If, however, 100 progenitors give rise to relapse, then there 
is a 96% probability that 2-3 patient relapses with marked cells would be detected, and 
if 1000 progenitors, then a 100% probability that several marked relapses (7.15) would 
be detected, and so on. 
Table 6.2-2: Relationship between the number of progenitor cells giving rise to a 
relapse, the probability of detecting a marked relapse, and the number of relapses in the 
study cohort. 
p r = 0.6 and TE x MT = 0.005 
NC 
N=5 
Prob (EXP) 
N=10 
Prob (EXP) 
N=12 
Prob (EXP) 
N=15 
Prob (EXP) 
1 
0.015 (.015) 
0.030 (0.03) 
0.035 (0.04) 
0.044 (0.05) 
10 
0.138 (0.15) 
0.257 (0.29) 
0.300 (0.35) 
0.360 (0.44) 
100 
0.741 (1.18) 
0.933 (2.37) 
0.961 (2.84) 
0.983 (3.55) 
1000 
0.989 (2.98) 
1.000 (5.96) 
1.000 (7.15) 
1.000 (8.94) 
3000 
0.990 (3.00) 
1.000 (6.00) 
1.000 (7.20) 
1.000 (9.00) 
From the information in Table 6.2-1, if 100 neuroblastoma progenitor cells are involved 
in a relapse, then by plating 100,000 neuroblastoma progenitor cells, we would expect 
39,000 of the colonies to be marked (G418 resistant). The number of marked progenitor 
cells giving rise to the relapse will be estimated. Technical considerations limit further 
PCR analysis to 400 marked colonies, which will be selected at random. In each of the 
Recombinant DNA Research, Volume 16 
