Advanced Flow Workshop Shumway Murphey

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
M E S F
Molecules of Equivalent Soluble
Fluorochrome
FCXM Interpretation Using MESF
1) MESF units are NOT the key to obtaining the clearest
perspective on FCXM data
2) The key is differential comparison of fluorescence
intensities that are expressed as linear values
3) MESF are simply standardized linear values
4) Standardized linear values alone are not sufficient to
universally standardize the FCXM
5) In Flow cytometry, the term “linear values” does not
always describe values that are linear!
XM: DONOR + RECIPIENT SELECTION
ANTIBODY:
LOTS OF IT
N I H
Crossmatch Method
AMOS
MW EXT
AHG
FLOW
CLINICAL
IMPACT:
IMMEDIATE
XM: DONOR + RECIPIENT SELECTION
Best donor recipient pair
= least antibody detected.
How do we determine when the Flow Crossmatch is
detecting more or fewer antibodies?
NOT MUCH
Best donor recipient pair
= least antibody detected.
DELAYED
MCF
MFI
MESF
THE CHANNEL SCALE
XM: DONOR + RECIPIENT SELECTION
(50)
#1
50 CHANNEL SHIFT
(30)
#2
50 is 2/3 higher than 30
30 CHANNEL SHIFT
220 250
120 170
Best donor recipient pair
= least antibody detected
= lowest channel shift
200 400 600 800 1000
Neg Ctl
Sample
1

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
Cutoff = 40 Channel Shift
(50)
#1
(30)
#2
220 250
120 170
50 CHANNEL SHIFT
30 CHANNEL SHIFT
Best donor recipient pair
= “no significant” antibody detected
200 400 600 800 1000
Neg Ctl
= channel shift < Cutoff
Sample
Cutoff = 40 Channel Shift
(50)
#1
(30)
#2
220 250
120 170
50 CHANNEL SHIFT
30 CHANNEL SHIFT
The “25% below cutoff” result
identifies 1/3 more antibody than the
channel shift that is not only 2/3
higher – but also “25% above cutoff”
200 400 600 800 1000
Neg Ctl
Sample
Cutoff = 40 Channel Shift
Monitoring Antibody Reduction
(50)
#1
50 CHANNEL SHIFT
(30)
#2
120 170 [50] 2.94 4.61 [1.67]
30 CHANNEL SHIFT
220 250 [30] 7.23 9.47 [2.24]
Antibody Reduced by:
Ch. Shift Reduced by:
50%
15%
220 250
120 170
200 400 600 800 1000
Neg Ctl
Sample
200 400 600 800 1000
Neg Ctl
Sample
Reduced
Fluidics & Optics
Log “Mode”: 1024 data points in log relation (Log Amp)
LASER
FSC
200 400 600 800
1000
SSC
ADC
FL-1
FL-2
FL-3
Amplifier
Log / Linear
SHEATH
SHEATH
SAMPLE Injector
Pulse
Processor
Pre-Amp
200 400 600 800 1000
Linear “Mode”: 1024 data points in linear relation
2

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
W H E N I N L O G “M M O D E” E :
Log “Mode”: 1024 data points in log relation (Log Amp)
End of 1 st decade
End of 2 nd decade
End of 3 rd decade
End of 4 th decade
CHANNEL VALUES
256 512 768 1024
BD = “Log Data”
Coulter = “Linear Data”
1
RLV
200 400 600 800
10 100 1000
1000
10,000
10 0 10 1 10 2 10 3 10 4
200 400
600 800 1000
200 400 600 800 1000
Linear “Mode”: 1024 data points in linear relation
Log “Mode”: 1024 data points in log relation (Log Amp)
End of 1 st decade
End of 2 nd decade
End of 3 rd decade
End of 4 th decade
VALUES THAT WOULD HELP US BEST UNDERSTAND
THE QUANTITATIVE RELATIONSHIP BETWEEN THESE
TWO GROUPS OF PEOPLE?
256 512 768 1024
1
RLV
200 400 600 800
10 100 1000
1000
10,000
256 512
1 2 3 4 5 6 7 8 910
200 End of 1 st decade 400
Fluorescence Intensity
End of 2 nd decade
20 30 40 50 60 70 100
4
96 additional units
154 358 Channel Shift
100
512
10 0
RLV
Log “Mode”: 1024 data points in log relation (Log Amp)
End of 1 st decade
End of 2 nd decade
End of 3 rd decade
256 512 768 1024
200 400 600 800
10 1 10 2 10 3
Linear “Mode”: 1024 data points in linear relation
End of 4 th decade
1000
10 4
256 512 768 1024
CHANNELS USED FOR
“BINNING” DATA
Variations in light intensity identifiable by PMT ~infinite
Represented as a finite, linear range for practicality
Entire range = 1024 increments (Channels)
Linear Mode:
0 – 1.0 1.1 – 2.0 3.1 – 4.0
2.1 – 3.0
200 400 600 800 1000
Scales LOOK the same – BUT hidden log scale beneath in Log Mode!!
In Linear Mode the channels ARE “numbers”!!
Moving from the origin (0) to the end of the scale
(1024), each subsequent channel serves as a “bin”
for a higher range of values.
3

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
Electronics
Amplifier
Log / Linear
Linear Scaling
(“normal numbers”: : movement
along scale according to amount
of increase)
Log “Mode”: 1024 data points in log relation (Log Amp)
Channels
256 512
Background
2x
4x
200 End of 1 st decade 400
End of 2 nd decade
1 2 3 4 5 6 7 8 910
Fluorescence Intensity
20 30 40 50 60 70 100
200 400 600 800 1000
Fluorescence Intensity
Integers that directly represent
quantities of fluorescence
CHANNELS USED FOR
“BINNING” DATA
Variations in light intensity identifiable by PMT ~infinite
Represented as a finite, 4 log range for practicality
Entire range = 1024 increments (Channels)
Log Mode:
1 X 1.009 X1.009 X1.009 X1.009
Electronics
Amplifier
Log / Linear
Background
2x
4x
8x
16x
Log Scaling
(movement along scale according
to proportion of increase)
32x
128x
64x 256x 512x 1024x 2048x
Moving from the origin (10 0 ) to the end of the scale
(10 4 ), each subsequent channel serves as a “bin”
for a higher range of values.
200 400 600 800 1000
Fluorescence Intensity
CHANNEL SHIFTS AND THE MULTIPLES OF
THE “STARTING VALUE” (BACKGROUND
FLUORESCENCE) THAT THEY REPRESENT
MULTIPLE
2
3
4
5
6
7
8
9
10
1024
SCALE
77
122
154
179
199
216
231
244
256
256
SCALE
19
31
39
45
50
54
58
61
64
** Bottom Line – channel shifts represent a
PROPORTIONAL relationship! **
** Bottom Line – channel shifts represent a
PROPORTIONAL relationship! **
THEREFORE: “Channel Shift” and “MFI Ratio”
are EXACTLY the SAME method of analysis!
MFI
Ratio
2
3
4
5
6
7
8
9
10
MCF
Shift
(1024)
77
122
154
179
199
216
231
244
256
(256)
19
31
39
45
50
54
58
61
64
4

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
XM: DONOR + RECIPIENT SELECTION
Best donor choice = least
fluorescence from
secondary antibody
3
detected.
FCXM #1
3
The quantity of signal
XM:
above background is of
- Bkgnd:
36
200% of background
significance (because it
Aby:
3
(77 ch. . Shift)
directly reflects the
Y
number of patient
FCXM #2
1
antibodies bound), the
XM:
1
proportion of background
- Bkgnd:
12
200% of background
is IRRELEVANT*!
Aby:
1 (77 ch. . Shift)
*with regard to the number of antibodies present
Y
Y
Y
Y
Y
Y
Y
Channel Shifts and the Illusion of Increased
Sensitivity When Using Pronase
Y Y
The same number of
Untreated B Cells
antibody molecules
XM:
binding against a lower
- Bkgnd:
background produce a
36
greater channel shift – Aby:
3
but remain the same
77 ch. . Shift
number of antibody
molecules!
Pronase Treated B Cells
Linear data reflects the
XM:
true quantity above
- Bkgnd:
14
background – a value
Aby:
3
that is independent of
154 ch. . Shift
the background.
Y Y
Y
Y
Y
3
Y
Y
1
Y
Y
Y
3
background
1
background
Relative Linear Values (RLV) are frequently referred to as MFI (Median
or Mean Fluorescence Intensity) values. It is these RLV / MFI values
that are “numbers” in the conventional sense (NOT the channel
values!), they have simply been plotted on a log scale. Events with an
MFI value of 600 are twice as bright as those with an MFI of 300 (the
expected, conventional, “linear” relationship). HOWEVER, if an event
reads at channel 600, an event half as bright will read at channel 523.
So how do we convert the channels to
the “numbers”?
CHANNEL VALUE
Calculate “R” = the
channel value divided
by channels per decade
512
512/256 = 2
End of 1 st decade
End of 2 nd decade
End of 3 rd decade
256 512 300 768 1024
200 400 600 600 800
End of 4 th decade
1000
10 R = RELATIVE
LINEAR VALUE
(RLV – aka MFI)
10 2 = 100
1
RLV
10 100 1000
Fluorescence Intensity
10,000
W H E N I N L O G “M M O D E” E :
CHANNEL VALUES
BD = “Log Data”
Coulter = “Linear Data”
RELATIVE LINEAR VALUES
BD = “Linear Data”
Coulter = “Log Data”
End of 1 st decade End of 2 nd decade End of 3 rd decade
End of 4 th decade
256 512 768 1024
RELATIVE LINEAR VALUE
is a much better term than
Median Fluorescence Intensity
1
RLV
200 400
600 800 1000
10 100 1000 10,000
5

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
1
“RELATIVE” LINEAR VALUES
RLV-BD
0.1 RLV-BC
10 100 1000 10,000
1 10 100 1,000
64
128 192
50 100 150 200 250
End of 1 st decade
End of 2 nd decade
End of 3 rd decade
256 512 768
256
End of 4 th decade
1024
“RELATIVE” LINEAR VALUES
Current XM:
Last XM PMT settings
X 4 =
Last XM:
Last XM PMT settings
MCF = 256 MCF = 410
MFI Ratio = 4
MCF = 77 MCF = 231
MFI Ratio = 4
154 MCF shift 154 MCF shift
8 – 2 = 6 MFI Shift 40 – 10 = 30 MFI Shift
400%
OF BACKGROUND
400%
OF BACKGROUND
200 400 600 800 1000
1
2 3 4 5 6 7 8910
200 400
20 30 40 50 60 70 100
“RELATIVE” LINEAR VALUES
X 4 =
Current XM:
NEW XM PMT settings
MCF = 256 MCF = 410
MFI Ratio = 4
154 MCF shift
40 – 10 = 30 MFI Shift
400%
OF BACKGROUND
“RELATIVE” LINEAR VALUES
X 4 =
Current XM:
CONSTANT PMT settings
Last XM:
CONSTANT PMT settings
MCF = 256 MCF = 410
MFI Ratio = 4
MCF = 77 MCF = 231
MFI Ratio = 4
154 MCF shift 154 MCF shift
8 – 2 = 6 MFI Shift 40 – 10 = 30 MFI Shift
400%
OF BACKGROUND
400%
OF BACKGROUND
1
2 3 4 5 6 7 8910
200 400
20 30 40 50 60 70 100
1
2 3 4 5 6 7 8910
200 400
20 30 40 50 60 70 100
MESF Values **
**** MESF Values are Independent of Instrument and Settings ****
200,000
50,000
12,500
MESF line at initial PMT settings
MESF line at increased PMT settings
** Relative
positions are not
to scale. For
illustration only.
MESF Acquisition
2 8 10 40
Cytometer Scale (RLV) **
Reference
Bead MESF
2546
6393
19019
47474
6

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
Universal Standardization of the FCXM
Will Require More Than MESF
Conversion due to the Variability of:
1) FCXM protocols
2) Secondary Ab characteristics
3) Negative control sera reactivity
4) HLA expression
FCXM Interpretation Using MESF
1) The key to obtaining the clearest perspective on FCXM
data is the differential comparison of fluorescence
intensities that are expressed as linear values
2) MESF are simply standardized linear values
3) If a lab FCXMs on multiple instruments, MESF
conversion provides a convenient way to standardize
data between machines - BUT
4) MESF conversion alone is not sufficient to universally
standardize the FCXM between laboratories
5) MESF reference lines can be used to verify the proper
functioning of fluorescence detection systems
Auto XM
Negative
Positive Donor XM
IS NOT
Due to auto-ab
ab
Patient Cells +
Patient Serum
Positive
Positive Donor XM
MAY BE
Due to auto-ab
ab
Negative
Positive Donor XM
IS NOT
Due to auto-ab
ab
Patient Cells +
Patient Serum
Positive
Positive Donor XM
MAY BE
Due to auto-ab
ab
Cell Washer
IF Positive Donor XM
Donor Cells + Donor Adsorbed patient serum
Negative
HLA-ab
likely
Positive
HLA-ab
NOT likely
7

Workshop 2: Advanced Flow Workshop-
Wayne Shumway, CHT, CHS and
Cathi Murphey, MT(ASCP), CHS, BLM
Advantages of an Automated Cell Washer:
1) Uniform and consistent washes – therefore
2) No wash technique effects on results
3) Wash processing time reduced 85-95%
Advantages of an Automated Cell Washer:
1) Uniform and consistent washes – therefore
2) No wash technique effects on results
3) Wash processing time reduced 85-95%
Disadvantages of an Automated Cell Washer:
1) .
2) .
3) .
Highly Recommended:
Helmer Ultra CW
www.helmerinc.com
1.800.743.5637
8