(IBC) and the Recombinant DNA Advisory Committee (RAC). These protocols have been 
implemented and we are actively enrolling patients. As in these protocols, the target for 
retroviral transduction will be CD34+ immunoselected hematopoietic cells. CD34+ cells have 
been shown to reconstitute patients undergoing ABMT (19-21). We have shown that these 
cells can be transduced with recombinant retroviral vectors in vitro (22), and preliminary data 
indicates that these retrovirally transduced cells contribute to engraftment in patients (C. 
Dunbar, unpublished observations, Appendix I). 
An important difference between the neomycin marking protocols and the MDR-1 protocol is 
the use of a marker gene intended to change cell phenotype. While this will allow us to 
address new, biologically relevant questions, it also incorporates potential risks. It is 
theoretically possible that breast cancer cells may be transduced with the MDR- 1 retroviral 
vector in this study. In order to reduce this risk, patients with evidence of bone disease 
evident on bone scan after induction therapy, or with histologic evidence of bone marrow 
involvement on bone marrow biopsy prior to or at harvest, will be excluded. Even in patients 
whose bone marrow is histologically negative for breast cancer, more sensitive 
immunohistochemical techniques show breast cancer cells in the marrow in a high proportion 
of patients with metastatic breast cancer (23). However, we believe that transduction of small 
numbers of breast cancer cells with the MDR- 1 vector is very unlikely to adversely affect 
patient outcome. Breast cancer patients relapse primarily in sites of their original metastatic 
disease following ABMT, implying that contaminating tumor cells within unfractionated bone 
marrow do not significantly contribute to post-transplant relapse (2-4). Furthermore, we 
propose to use CD34 immunoselection to purify and transduce only a fraction of harvested 
bone marrow and mobilized peripheral blood cells. CD34 purification with the CellPro 
column will result in an additional 2-5 log reduction in any tumor cell contamination of 
harvested hematopoietic cells (Dr. E J. Shpall, personal communication). Primary breast 
cancer cells grow poorly in culture (K. Cowan, unpublished observations), and it is not 
known whether remaining tumor cells would survive during the transduction period, or what 
proportion of surviving tumor cells would be transduced with the vector. Even in the unlikely 
event that viable, transduced tumor cells contributed to taxol and vinblastine resistant relapse, 
other treatment options remain. Antineoplastic drugs that are not substrates for Pgp, but 
which have activity in breast cancer, including antimetabolites (methotrexate, fluorouracil), 
and alkylating agents (thiotepa, cytoxan, melphalan, and cis-platinum), could be used for 
palliation. We plan to evaluate patients for evidence of relapse with transduced tumor cells. If 
patients relapse following transplant, biopsies of accessible tumor will be assayed for evidence 
of proviral integration, proviral mRNA expression, and immunologic evidence of Pgp 
expression. 
3.0 Patient Eligibility 
3 . 1 Histologically-documented Stage IV breast cancer. 
3.2 Patients must have achieved at least a partial response to induction chemotherapy. 
Patients achieving a complete response are eligible for study. 
3.3 Age: 18-60. 
3.4 No history or clinical evidence of metastatic CNS disease. 
3.5 Patients who have received treatment with more than two prior chemotherapy regimens 
or who have received mitomycin C are not eligible. Patients who have undergone a 
prior ABMT are excluded. 
3.6 Normal cardiac function: No history of angina, myocardial infarction, CHF. LVEF 
must be >40%. Exercise treadmill test must be normal. 
3.7 Creatinine clearance > 90cc/min, bilirubin <1.5., SGOT <2x normal, normal PT, 
PTT, calcium. 
3.8 Negative HIV serology and hepatitis B surface antigen. 
Recombinant DNA Research, Volume 18 
[279] 
