FLOW CYTOMETRY IN PRECLINICAL TOXICOLOGY/SAFETY ASSESSMENT - Flow Cytometry in Drug Discovery and Development - page 139

stored refrigerated overnight. Smears and cytospins were evaluated and scored by an
experienced hematology technologist.
In order to confirm that manual cytologic characterization was in agreement with
flow cytometric immunophenotype for the major lineage populations, nucleated
erythroid, myeloid, and lymphoid lineage cells were sorted and reanalyzed to assess
purity and cytospins were prepared for cytologic evaluation. Nucleated erythroid cells
were identified as HO þ /CD45 /ter119 þ events. Myeloid cells were identified as
HO þ /CD45 þ /ter119 /CD11b þ and included both granulocytes and monocytes.
Lymphocytes were identified as HO þ /CD45 þ /ter119 /CD11b events. 7-AAD þ ,
ter119 þ /CD45 þ double-positive, and ter119 /CD45 double-negative events were
excluded as dead, aggregates, and debris, respectively. Sorting experiments revealed
that flow cytometric identification of nucleated erythroid, lymphoid, and myeloid
bone marrow subsets agreed very well with manual cytologic classification. The cell
type sorted agreed with cytologic characterization in all cases and variation in stated
purity of sorted populations by flow cytometric reanalysis versus cytologic assess-
ment of cytospins was less than 3% for all populations.
In an effort to further investigate the utility of flow cytometric bone marrow
analysis in toxicology studies, recombinant human erythropoietin (rhEpo) was used
as a tool compound and administered to mice. A single subcutaneous injection of
rhEpo at a dose of 3000 U/kg is sufficient to stimulate erythroblast production in
mouse bone marrow with maximal effect observed 48 h postdose [35]. Therefore,
bone marrow was collected approximately 48 h after a single injection. A dose of
1000 U/kg was included to investigate dose-response relationships. The effects on
mouse bone marrow populations were monitored by flow cytometric analysis and
cytologic assessment of bone marrow smears and results were compared.
6.7.2 Results and Discussion
A comparison of flow cytometry and cytology results produced similar trends in all
monitored end points. Both methods revealed dose-responsive changes in bone
marrow, although more instances of statistical significance were observed in flow
cytometry data. This can be attributed to lower intragroup variation present in
flow cytometry results versus cytology that is not surprising given that a much larger
number of cells are interrogated in flow cytometric analysis. A comparison of
methods using least square means was used to statistically confirm this assertion
(Figure 6.5). Results strongly suggest greater sensitivity of flow cytometric
methodologies to identify drug-induced bone marrow changes as demonstrated by
dose-dependent changes seen in nucleated erythroid and lymphoid compartments on
a percentage basis, nucleated erythroid on an absolute basis, and M:E (myeloid:
erythroid) ratio whereas no dose-dependent changes of statistical significance were
observed by cytology.
A trend for flow cytometric erythroid enumeration values to be below those
determined cytologically was observed (Figure 6.6). Although this discrepancy in
erythroid enumeration appeared striking at first glance, paired t-test comparison of
methods revealed statistically significant difference among methods only in absolute