10% of the marrow) . The recent literature concerning 
residual leukemia suggest that the transplanted leukemia 
burden is actually much greater (NEJM 323:448, 1990; 
Blood 74:1762, 1989). Therefore, we have calculated the 
likelihood of detecting marked cells over a range of 1 to 
10,000 leukemic cells contributing to relapse to estimate 
our likelihood of success. 
Using our minimal estimates of cells in the transplanted 
marrow at 100 cells, and marking 1% of the total 
transplanted leukemic cells, the probability of detecting 
marked cell is 63.4% in a single patient. If larger 
number of cells are contributing, for example 1000, the 
likelihood of detecting marked cells will be >99% in the 
first patient. 
Columns = # of transplanted leukemic cell contributing to 
relapse 
Rows = Percentage of cell contributing to relapse which 
contain the LNL6 vector 
Data = Probability of detecting marked cells in a single 
patient 
1 
10 
100 
300 
1000 
3000 
10000 
30000 
.1% 
0.1 
1.0 
9.5 
25.9 
63 . 2 
95.0 
>99.99 
>99.99 
. 3% 
0.3 
3 . 0 
26.0 
59.4 
95.0 
99.99 
>99.99 
>99.99 
1% 
1.0 
9.6 
63.4 
95.1 
>99.99 
>99.99 
>99.99 
>99.99 
3% 
3.0 
26.3 
95.3 
99.99 
>99.99 
>99.99 
>99.99 
>99.99 
The likelihood of detection increases as the number of 
patients studied increases. The estimates given below 
indicate a 95% probability of detecting marked leukemic 
cells in one of the first three patients if cell marking 
is at the expected 1% transduction level and only 100 
cells contribute to relapse. Even if the transduction 
efficiency was ten times lower than estimated by pre- 
clinical data, the likelihood of success is still 95% if 
ten patients are studied. 
[494] 
Recombinant DNA Research, Volume 15 
