marrow prior to reinfusion. 81 Despite a 2-3 log purging 
efficacy on myeloma tumor cell lines, and absence of any 
detectable B lineage cells in the treated marrows, 7/11 
patients relapsed or did not respond to high-dose 
chemotherapy and autologous transplantation. Studies are 
also ongoing with other methods of in vitro purging, such as 
4-HC or ricin-conjugated immunotoxins, but results are too 
preliminary to interpret. 82,83 
These results are not surprising given the lack of proven 
value for purged versus unpurged autologous marrow in any 
malignancy involving the bone marrow. In myeloma the 
problem is compounded by our lack of ability to define the 
myeloma "stem cell" phenotypically , making the choice of 
purging agents and assessment of purging efficacy much more 
difficult. Plasma cells, even myelomatous ones, have an 
extremely low labelling index and seem to be terminally 
differentiated cells with- little capacity for proliferation. 
Self-renewal must be explained by a clonogenic "stem cell" 
earlier in the maturation sequence. 84 It also seems clear 
that myeloma tumor cells exhibit abnormal maturation and 
lineage infidelity. In almost 50% of patients, aberrant 
expression of CALLA, myelomonocytic antigens and T-cell 
antigens on tumor cells can be detected. These 
subpopulations with an immature phenotype may represent a 
stage closer to the true myeloma "stem cell," and one 
explanation is that the initial step in myeloma pathogenesis 
may involve a cell with -pluripotent characteristics. 
Recent approaches to defining the involved cell 
genotypically instead of phenotypically are more promising 
in terms of analyzing populations of cells for otherwise 
undetectable tumor. These approaches are all based on the 
detection of clonally rearranged immunoglobulin genes. By 
using Southern blotting, which is sensitive down to 1-5% 
clonal cells, populations of malignant cells have been 
detected in the, peripheral blood of myeloma patients, even 
in the. absence, of any' phenotypically identifiable 
circulating tumor cells. 5,76 A much^more sensitive technique 
using consensus, primers to P£R-amplify the rearranged 
immunoglobulin locus has recently been applied to myeloma. 6 
Even patients in CR by standard criteria (i.e. lack of any 
detectable monoclonal paraprotein, <5% bone marrow plasma 
cells) had specific tumor rearrangements detected by this 
technique, which is sensitive down to at least one cell in 
10 5 . This technique has not been used to monitor myeloma 
patients post-transplant, and the. significance of residual 
clonally-rearranged cells is uncertain. 
It seems more likely that patients who relapse after 
autologous bone marrow transplantation do so because of 
failure of the pre-transplant conditioning regimen to 
Recombinant DNA Research, Volume 16 
[45] 
