of these leukemic cells in CLL. The M. D. Anderson 
program will consist of taking 30% of a population of 
purged autologous cells, which are to be used to 
reconstitute hematopoiesis following TBI and intensive 
preparative therapy, exposing it in vitro to the 
replication-incompetent GINa and LNL-6 neo positive 
retroviral vectors, thus marking the cells in vitro, and 
then returning those marked cells along with the rest of 
the autologous stem cells to restore hematopoiesis after 
intensive therapy. Since the two safety modified 
retroviruses are used for separately marking the 
autologous marrow at different stages of the in vitro 
fractionation are marked with viruses which differ only 
in a PCR detectable difference in sequences in the virus, 
the relative contribution of the marrow cells at 
different stages of the f ractionatidn to hematopoietic 
reconstitution and to relapse can be measured. 
The patients will then be followed at three weekly 
intervals during and after hematopoietic reconstitution 
and at six monthly intervals thereafter with PCR's for 
lymphoid markers and the two neo positive safety modified 
retroviral sequences in the marrow. This analyses will 
clarify the relative contribution of the cells left in 
the marrow to relapse. This analysis will be done at the 
time of relapse with colony growth, PCR, FISH, and 
cytogenetic assays of the colonies to determine if the 
blast cells that arise after relapse are marked with 
either of the retroviral vectors and, therefore, arising 
from the infused autologous cells or whether they arise 
from systemic disease which is left after the delivery of 
preparative therapy (in this case, no marker will be 
detectable) . There are four potential results of this 
experiment: (1) A rare patient will have virus positive 
autologous leukemia cells, but these will be clonal 
(i.e., the vector integrating site will be clonal). The 
same normal cells may have a polyclonal vector 
integration site. This will indicate that the infused 
autologous cells are the origin of relapse but that 
relapse arises from a single cell; (2) Leukemia cells are 
positive for vector and the vector integration site is 
polyclonal. Normal cells may also have polyclonal 
integration sites. This will indicate that multiple 
leukemia cells contained within the infused autologous 
stem cells give rise to the relapse; (3) Both normal and 
leukemia cells are marked with a clonal vector 
integration site. This will indicate that the disease 
arises from a pluripotent stem cell which is undergoing 
leukemic conversion spontaneously; and (4) No marker is 
seen suggesting relapse arises from systemic disease. In 
cases 1 and 2, further development of purging procedures 
for the autologous cells needed for transplant would be 
indicated. Retroviral marking will permit us to select 
either the purification of normal progenitor cells or 
Recombinant DNA Research, Volume 16 
[395] 
