Basic HTML Version
35
october 7-9, 2012 • NEW ORLEANS, LA
CD95, CD127 CD152, CD154, CD178, FOXP3, TCRab,
TCRgd. This study may lead to the need for new clinical and
therapeutic approaches based on the results obtained. The
use of biomarkers as a tool to monitor this group of patients
may predict response to treatment and predict the risk of
chronic graft-versus-host and the same diagnosis.
P9
RESULTS FROM FLOWCAP2 - CRITICAL ASSESSMENT
OF AUTOMATED FLOW CYTOMETRY DATA ANALYSIS
TECHNIQUES ON CLINICAL DATA FOR DIAGNOSIS
Ryan Brinkman
1
, Nima Aghaeepour
1
, Greg Finak
2
, Holger
Hoos
3
, Tim Mosmann
4
, Raphael Gottardo
2
, Richard
Scheuermann
5
1
BC Cancer Agency,
2
Fred Hutchingson
Cancer Research Center,
3
University of British Columbia,
4
University of Rochester,
5
J Craig Venter Institute
An important use case for FCM analysis is the discovery
of biomarker patterns in FCM data for the purposes of
sample classification (i.e., diagnosis). We assembled a
benchmark of three datasets in which the subjects/samples
were associated with an external variable that could be
used as an independent measure of truth for sample
classification. The benchmark consisted datasets for: (1)
studying the effect of HIV exposure on 44 African infants
using 6 tubes of 8 color assays (HIV-exposed emph in utero
but uninfected (HEU) vs. unexposed (UE)); (2) diagnosis
of acute myeloid leukemia (AML) using 8 tubes of 5 color
assays on 359 subjects provided by a reference laboratory
(AML vs. non-AML); (3) discriminating between two antigen
stimulation groups of post-HIV vaccine T-cells using two
tubes of 8 color assays on 48 subjects (Gag-stimulated vs.
Env-stimulated). We received a total of 43 submissions of
computational approaches for the analysis of flow cytometry
data and we compared the sensitivity, specificity, accuracy
and F-measure of these approaches to correctly predict
patient outcome. For two of the datasets (AML and HIV
Vaccine Trials Network (HVTN)) many algorithms were able
to perfectly predict the external variables even under very
conservative conditions (i.e., using an independent test set
as large as the training set).
P10
EXPRESSION OF CD2 AND CD25 ON MAST CELL
POPULATIONS OUTSIDE THE SETTING OF SYSTEMIC
MASTOCYTOSIS
Sindhu Cherian, Vivian McCulloch, Katy Dougherty, Valerie
Miller, Jonathan Fromm, Brent Wood
University of
Washington
Systemic mastocytosis (SM) is a diagnosis made using a
variety of clinical, laboratory, and histologic parameters. The
demonstration of aberrant CD2 and/or CD25 expression on
mast cells provides one minor criterion for a diagnosis of
SM. In order to validate a tube (CD45/CD117/CD2/CD25)
for mast cell evaluation in our laboratory, we performed
flow cytometry (FC) using this tube on marrow samples
submitted for routine analysis by FC. Cases included in the
study (n=44) had either an unknown diagnosis or a lymphoid
neoplasm while cases with a myeloid stem cell neoplasm,
cases in which SM was a clinical consideration, and cases
with too few mast cells for evaluation (<0.01% mast cells)
were excluded. The threshold for defining a case as positive
was determined using a fluorescence minus one control. 36
cases showed mast cell populations with no expression of
CD2 or CD25. 8 cases showed expression of CD2 and/or
CD25 on ≥10% of the mast cell population (CD25 in 5 cases,
CD2 in 2 cases, and both in 1 case). Review of available
clinical data revealed no history of SM in the positive cases.
The percentage of mast cells showing aberrant expression
of CD2 and CD25 ranged from 12.1% to 98.8% and 22.2%
POSTER ABSTRACTS