re quality controls for CD34 enumeration (ISHAGE) on BD FACSCALIBUR

D. Robert Sutherland rob.sutherland at
Fri Oct 26 15:50:08 EDT 2007

Hi Adamari,
we have used BD's high and low controls for QC'ing CD34 enumeration  
on a Calibur running the single platform ISHAGE protocols.
The target range for the 'high' control (usually around 22.5 - 27.5  
CD34+ cells per microlitre) is pretty difficult to miss. Furthermore,  
you will acquire well in excess of the minimum number of 100 CD34+  
cells required to ensure a CV of 10% or less, if you are collecting a  
minimum number of 75,000 CD45+ cells.
The low range control in this 'kit' is in fact made from a normal  
blood sample and generally comes with an assayed CD34+ cell count  
between 0.75 to about 1.5 CD34+ cells per microlitre (although the  
range as supplied is usually 'zero to about 2.5'). (How this assayed  
CD34+ cell concentration is accurately and precisely determined at  
this low level boggles the mind!!).
At that concentration, you will be lucky to find up to 50 CD34+ cells  
in your listmode file even with a 15 minute acquisition time (the  
effective maximum time that the Trucount beads will remain in  
adequate suspension for this assay).
If you are using Flowcount beads as part of the Beckman Coulter Stem- 
kit on your Calibur, you will be lucky to get 20 CD34+ cells in the 5  
minute time limit we set ourselves for retaining the larger Flowcount  
beads in adequate suspension.
When these QC samples are analysed on Beckman-Coulter instruments,  
further issues arise due to the different angles of the forward light  
scatter detectors utilised by BD and Beckman Coulter instruments.  
This causes an apparent change in the light scatter characteristics  
of clustered CD34+ cells compared to normal samples. The degree of  
light scatter change varies between instrument manufacturers. When  
combined with about 20 cells or less, one can never be very confident  
of what to include and what to exclude from the final light scatter  
gate when such low range stabilised cells are used with this  
protocol. In other words, controls at this level are of limited use,  
and I am not aware of any Bone Marrow Transplant program that makes  
'clinical decisions' at the level of 1 CD34+ cell per microlitre of  

We have discussed some of these issues with the manufacturers of  
stabilised QC samples and urged them to produce 'low range' controls  
in the 10 - 12 CD34+ cells per microlitre range. At this level, one  
should be able to acquire approximately 100 CD34+ cells per listmode  
file on either instrument platform, regardless of whether Flowcount,  
Trucount, (and perhaps other) beads are in use. Even though the  
clusters of CD34+ cells in the final light scatter gate will still  
move around a bit compared to fresh patient material, the number of  
events in the cluster should give confidence in the reported result.

If you would like to send me a listmode file or two of your  
problematic analyses, I would be happy to look at them for you.

Rob Sutherland
University Health Network Toronto Ontario
Mike Keeney
London Health Sciences Centre London Ontario

More information about the Cytometry mailing list