post remission chemotherapy was used in AML-83. Most recently we have intensifled 
therapy with targeted plasma concentrations of ara-C and VP-16 (AML-87). The 
overall event-free survival of these three approaches is similar and suggests that the 
gains made by intensive therapy have reached their limits with our current drugs. 
The improvements made in autologous transplantation and the outcome for this 
procedure in patients in second complete remission suggests this may be a good 
treatment for children with AML. It permits dose escalation and the use of drugs 
such as busulfan which have excellent antileukemic properties but unpredictable and 
dose limiting marrow toxicity. Therefore, in this proposal we plan to use autologous 
marrow transplantation for all children with AML (without an HLA-match) in first 
complete remission. 
3.1 Gene Markers for Autologous Marrow 
Because autologous marrow is phenotypically and genetically identical to 
residual host tissues, it has not been possible to date to investigate either the 
mechanisms of marrow regeneration after ABMT or the source of malignant 
relapse in marrow derived malignancies following ABMT. The availability of 
transduction procedures able to mark harvested progenitor cells and leukemic 
cells means that both issues can now be addressed. 
3.2 Origin of Malignant Cells During Relapse 
Retroviral vectors have been successfully used to mark malignant cells and cell 
lines in man. In rodent models, marked malignant cells have re-established 
malignant disease. If residual leukemic cells in the harvested marrow could be 
marked, it would be possible to determine whether the malignant cells 
appearing in patients who relapse after ABMT are derived from the infused 
marrow. Detection of such marked tumor cells would provide justification for 
the continued exploration and development of marrow purging technologies. 
Moreover, if the leukemic progenitor cells were marked, analysis of vector 
insertion sites would show whether relapse was monoclonal or polyclonal, a 
difference which would have implications for monitoring the efficiency of 
different purging techniques. 
3.3 Investigation of Stem Cell Involvement in the Malignant Process 
The transduction of marker genes into human marrow prior to autograft would 
also provide information of another kind. Malignancies arising from marrow 
cells may be of two kinds. They may arise from a malignant pluripotent stem 
cell or they may arise from a lineage committed progenitor cell. An example 
[508] 
Recombinant DNA Research, Volume 14 
