marrow involvement by breast cancer ABMT is unknown. Two 
studies have suggested that the rate of relapse for patients with 
early stage (Stages I, II) or locally advanced breast cancer (Stage 
III) was higher for patients with occult metastases (demonstrated 
by immunohistochemistry) versus those without occult metastases 
(29, 30). Several methods of bone marrow purging have been shown 
to reduce the burden of metastatic breast cancer in stem cell 
harvests by 3-5 logs (31, 32). Whether this degree of purging will 
be clinically beneficial remains to be demonstrated. 
The major concern regarding reinfusion of stem cell harvests 
containing occult metastases is whether these reinfused breast 
cancer cells will cause early relapse. Several studies have reported 
that breast cancer relapse post-ABMT occurs primarily at sites of 
original disease (9-11). However, there is currently no method of 
distinguishing breast cancer cells that may have been reinfused 
from those that represent relapse of original disease. In this pilot 
protocol, the CD34+ population hematopoietic of stem cells will be 
purified using the anti-CD34 antibody 12.8 and the immuno- 
absorption column Separate-SC. This column has been approved by 
the FDA for experimental use. This technique has been shown to 
yield a population of CD34+ cells that is 60-70% pure (15). These 
cells will then be incubated in 3 - 6 day cultures with the retroviral 
vector as described below. We expect that few breast cancer cells 
will contaminate the CD34+ population after the positive selection 
process and that any that did would be unlikely to survive the 3 - 6 
day cell culture process. However, no data currently exist on 
whether this is the case. It is possible that some breast cancer 
cells will also be transduced with the neo marker gene. We will 
therefore obtain bone marrow biopsies, and if safely possible, tumor 
biopsies on patients with relapsed disease post-ABMT to test for 
the presence of the neo gene by PCR. To do this we will need to 
culture the patients' breast cancer cells prior to PCR analysis to 
separate the cancer cells from contaminating blood cells. Finding 
the marker gene in sites of relapsed disease in this study would be 
an unexpected finding given the above considerations. Should this 
occur, however, it would force revision of our current thinking about 
the source of relapse. 
[96] 
Recombinant DNA Research, Volume 16 
