of the former is chronic granulocytic leukemia (CGL); while most acute 
lymphoblastic leukemias (ALL) are examples of the latter. 
It is important to discriminate between these two possibilities since the strategy 
required for cure will be different in each. There is little purpose in 
attempting to cure CGL by ablation of relatively mature marrow cells by 
chemotherapy, since eradication of stem cells is required. Similarly, these 
patients would be unlikely to benefit from standard autologous bone marrow 
transplantation since the normal stem cells within the marrow would be 
admixed with the stem cells of malignant origin. In contrast, development of 
a malignancy in a cell already committed to one or another lineage may be 
much more successfully treated with chemotherapy since its biological 
characteristics would be quite distinct from that of a normal stem cell. 
Discriminating between these two possibilities is only feasible in the presence 
of a naturally occurring gene marker which can show whether a single 
malignant progenitor can produce cells in more than one lineage. Although 
investigation of monoclonality using X-linked polymorphisms can provide 
information in some contexts, this is an unsatisfactory approach where a 
monoclonal subset may be concealed among a clonally heterogeneous normal 
population. Instead gene markers unique to the malignant population must be 
used. For CGL, the marker which demonstrates the stem cell origin of the 
disorder is the Philadelphia chromosome. Conversely, in ALL 
immunoglobulin/T cell receptor gene rearrangements or occasional 
chromosomal translocations give added support to the concept that this disease 
represents an aberration in a committed progenitor cell. Less is known of the 
origin of AML, since only 30% of AML have gene translocations. Although 
a significant number of AML clones may have rearranged T cell receptor or 
Ig genes, or may have cellular proto-oncogene mutations these are much less 
rigorous markers for the multi-lineage potential of the originating malignant 
clone. Gene rearrangement and oncogene activation may occur after lineage 
commitment, by the progeity of a malignant stem cell, since they may have 
only limited impact on the malignant process. Indeed detection of activated 
oncogenes may fluctuate during the clinical course of AML. 
If it could be shown that AML generally involved an early, multilineage 
progenitor, then current approaches to therapy would need to be greatly 
modified; simple intensification of cyto-reductive therapy (the current strategy) 
would be of little or no benefit. Certainly, study of those AML which do have 
gene translocations has suggested this consideration is correct. Patients with 
monosomy 7 which is a multilineage disorder respond poorly to chemotherapy. 
Conversely, patients with 15:17 or 8:21 translocations or inv-16 which are 
clearly lineage restricted,do well with conventional therapy. Marker gene 
Recombinant DNA Research, Volume 14 
[509] 
