RE: HLA TYPING B27 BY FLOW
Snider Denis (sniderd@fhs.csu.mcmaster.ca)
Mon, 21 Apr 1997 16:31:17 -0400
Vivian's experience is basically the same as ours with the B-D kits. We
are a regional testing center for B27 and use the kit to screen the
200-300 samples per month. From initial pilot data we developed a
screening scheme for "gray-area" flow results representing B27
fluorescence intensities near the B-D "cut-off" channel number (+/- 6
channel numbers). Those within this range represent about 10% of our
samples, and we immediately send these to cytotoxicity or DNA typing for
confirmation. Of those only 50% turn out to be false positive or
negative (cytotoxicity as "gold" standard), with about equal
distribution of negatives and positives. Of false positive, 75% are B7+
(cytotoxicity or DNA). Since this represents only 2% of our work load,
it does not make sense to further test for B7 in flow. The false
negatives are not easily explained although some relate to overall loss
of class I expression, tested in flow using the W6/32 mAb. In all, the
B27 results are >95% accurate, and fairly inexpensive (even by B-D
standards of inordinate cost!!!), compared to cytotoxicity alone or
testing in conjunction with SSP DNA.
As I see it the cytotoxicity tests have never been examined to a
reliability of 95% so some of the "discrepancies" could be related to
variability in the cytotoxicity accuracy in a given laboratory.
So for large volume, the B-D reagents are economical and a very reliable
screen. In the tough cutback times for Health Care in southern Ontario,
that has a "real" meaning. Of course, if you expect to have a large
volume such as ours, you should try to "extract" a little compensation
in reagent cost from B-D.
Good luck,
Denis Snider
>