Our past experience and the above estimates suggest that there will be 14 
relapses in which to evaluate the efficacy of the two purging techniques. For 
each marked relapse the number of progenitor cells transfected with each 
vector and giving rise to the relapse will be estimated. The Wilcoxon signed 
rank test will be used to compare the two techniques in terms of the paired 
differences between the numbers of progenitor cells marked with each vector. 
No information exist on which reasonable power calculations can be based. 
This experiment is considered the first step in investigating these two 
techniques and based on the observations a definitive trial will be proposed 
to the Pediatric Oncology Group. 
8.4 Historical Control Comparison of Event-Free Survival 
The experience of St. Jude Protocol AML-87 will be used as the historical 
control for evaluating the event-free survival and duration of CCR achieved 
by ABMT. Kaplan-Meier estimates of duration of CCR for AML-87 are 
shown in the following table. 
YEARS 
PROB 
SE 
1 
0.55 
0.07 
2 
0.35 
0.06 
3 
0.32 
0.07 
Using the technique of Simon and Dixon, we estimate that the study will have 
approximately 75% power to detect and improvement in duration of CCR 
from the observes .35 to .55 for ABMT. The test will have a significance level 
of 0.05 and will be a one sided log-rank comparison. 
Kaplan-Meier estimates of both event-free survival and duration of CCR will 
be calculated. 
9.0 DRUG INFORMATION 
9.1 Busulfan (Myleran®, BU) NDC# 0081-0713-25 
Busulfan is available as Myleran® (Burroughs-Wellcome and Co.) in 2 mg 
scored tablets. It is impossible to prepare a parenteral form of the drug 
because it is insoluble in water. Oral preparations are well tolerated. 
Busulfan for this study will be given four times a day (q 6 hr) at a dose of 4 
mg/kg/day PO; four consecutive days prior to CTX administration. The dose 
is per kg of actual body weight or ideal body weight, whichever is less . Dose 
[246] 
Recombinant DNA Research, Volume 17 
