54
controls can be instituted and these must be clearly stated so that one can be certain that the
controls are appropriate. Examples are: controls for the instrument, i.e. is the flow cytometer
functioning correctly?; controls for the procedure, i.e. were the correct stains used?; controls
for the sample, i.e. was the sample obtained correctly?; controls for the reagents, i.e. have
they gone bad since last used?, etc.
There are three types of controls that are relatively unique to flow cytometry that we might
discuss. First is that of an internal standard which may be added to each sample and which,
if everything is correct, has known characteristics. This is a very good method of controlling
for a number of variables. Such internal standards must have properties that are similar to,
but clearly distinct from, the biological particles of interest. An example of this will be given
below.
Another type of control which has not been widely used in flow cytometry but which is ideally
suited to flow cytometry data is that of testing the shape of the distribution of some parameter,
either over a population of particles or over a population of samples. The distribution shape
is often very sensitive to errors in the instrument, the procedure, or the sample. Relatively
little software is available for this type of testing but it will likely be exploited to a greater
extent in the future.
Additionally, it is frequently possible, due to the multi -parameter nature of flow cytometry
data, to find ways of testing for internal consistency in a data set. If the same particles are
counted in different ways or in different samples they should give the same results. Tests for
this type of internal consistency are now being applied to great advantage in medical flow
cytometry.
ABSOLUTE CALIBRATION
Finally we come to the most difficult aspect of quantitating flow cytometry data: expressing
the results in terms of quantities that are not intrinsically flow cytometric, i.e. in absolute
values. This is can be done in terms of quantity, e.g. the conversion of arbitrary fluorescence
to molecules of fluorochrome or moles of antigen, or the conversion of light scatter to size.
Alternatively, conversion to absolute concentration (particles per unit volume) may be
required. Both of these types of conversions are difficult with current flow cytometers.
controls can be instituted and these must be clearly stated so that one can be certain that the
controls are appropriate. Examples are: controls for the instrument, i.e. is the flow cytometer
functioning correctly?; controls for the procedure, i.e. were the correct stains used?; controls
for the sample, i.e. was the sample obtained correctly?; controls for the reagents, i.e. have
they gone bad since last used?, etc.
There are three types of controls that are relatively unique to flow cytometry that we might
discuss. First is that of an internal standard which may be added to each sample and which,
if everything is correct, has known characteristics. This is a very good method of controlling
for a number of variables. Such internal standards must have properties that are similar to,
but clearly distinct from, the biological particles of interest. An example of this will be given
below.
Another type of control which has not been widely used in flow cytometry but which is ideally
suited to flow cytometry data is that of testing the shape of the distribution of some parameter,
either over a population of particles or over a population of samples. The distribution shape
is often very sensitive to errors in the instrument, the procedure, or the sample. Relatively
little software is available for this type of testing but it will likely be exploited to a greater
extent in the future.
Additionally, it is frequently possible, due to the multi -parameter nature of flow cytometry
data, to find ways of testing for internal consistency in a data set. If the same particles are
counted in different ways or in different samples they should give the same results. Tests for
this type of internal consistency are now being applied to great advantage in medical flow
cytometry.
ABSOLUTE CALIBRATION
Finally we come to the most difficult aspect of quantitating flow cytometry data: expressing
the results in terms of quantities that are not intrinsically flow cytometric, i.e. in absolute
values. This is can be done in terms of quantity, e.g. the conversion of arbitrary fluorescence
to molecules of fluorochrome or moles of antigen, or the conversion of light scatter to size.
Alternatively, conversion to absolute concentration (particles per unit volume) may be
required. Both of these types of conversions are difficult with current flow cytometers.