Autologous Bone Marrow Transplantation - Blog Science Connections

86 Mafosfamide Purging in Leukemia
dose-response curve was too steep to select appropriate drug doses to work
with. Others have made similar observations, and the inverse relation between
mafosfamide cytotoxicity to progenitor cells and the RBC contamination
responsible for nonspecific binding and diversion of the drug from the primary
target is now well established.
Our preliminary studies also showed a wide range of sensitivity among
patients; further, a few patients tested sequentially (data not shown) revealed an
increasing resistance to the drug that developed in parallel with disease
progression. When we started the clinical trial, these observations prompted us
to study each patient's sensitivity to mafosfamide immediately before marrow
collection to select the incubation dosage rather than using a constant
amount—hence, we used the PIT. As in our preliminary studies, the PIT showed
considerable variations among patients in response to identical drug concentrations,
which suggested that adjusting the dose for individuals might be
appropriate. Because we could not directly evaluate the tumor log-cell kill and
because the biological importance of cells detected in the CFCI-L assay is
unknown, we decided to define the incubation dose of mafosfamide for marrow
as the highest possible dose that would spare enough normal stem cells to
ensure consistent engraftment. A residual amount of 5% ± 5% CFCl-GM in the
cleansed marrow was considered a safe margin beyond which further
cytotoxicity would not be measurable, at least by conventional laboratory
means. With this definition, doses of mafosfamide ranging from 50 ug to 140
ug/2 x 10 7
buffy coat cells were necessary.
Since we initiated our program, other investigators have reported similar
individual variations in sensitivity to mafosfamide both in animal progenitor cell
populations (8) and in human leukemic and lymphoid cells (9). Our long-term
culture studies indicated that even marrows totally depleted of CFCl-GM still
retained a self-regenerating ability. Although they confirmed the higher
resistance of early progenitors to CY derivatives, as suggested by initial
preliminary reports of successful autologous engraftment in humans with
marrow deprived of CFCl-GM (10), these studies also showed that the selfrenewal
capacity was proportional to the surviving fraction of CFCl-GM in the
sample initially placed in culture, which reassured us in our intention not to go
below the threshold of 5% residual CFCl-GM in the cleansed marrow. Fifty of 52
patients in our study experienced engraftment, although they received, as
predicted, very low numbers of CFCl-GM.
We can draw some conclusions. First, mafosfamide can be used to cleanse
the marrow of patients with acute leukemia. Because patients show a wide
range of sensitivity to the drug, we believe that individual dosage adjustment is
an option, the clinical relevance of which must be tested further. Second, the
differences in kinetics of hematopoietic recovery suggest that the engraftment
potential has been altered more severely in ANLL than in ALL, which may reflect
both the intensity of the in vitro treatment and the intrinsic fragility of the
stem cell pool in ANLL.