FLOW-373905.

18

Sponsored and reviewed by ICCS Quality and Standards Committee

Title: Compensation Tips for Beckman Coulter 10-Color Navios Platform
Written by:
Salima Janmohamed-Anastasakis Ph.D., Applications Scientist, Beckman Coulter Life
Sciences, Division of Beckman Coulter, Inc.
Amr Rajab BSc, MLT, SCYM (ASCP)CM, Flow Cytometry Technical Specialist, LifeLabs, Toronto,
Canada
Andrea Illingworth MS, H(ASCP)SCYM, Manager of Flow Cytometry, Clinical Trials and HPV
Testing, Dahl-Chase Diagnostic Services
Date: June 18 2018
______________________________________________________________________

OUTLINE

Compensation is an important component of assay-specific optimization of a flow cytometer. Incorrect

compensation has the potential to lead to false-positive or false-negative interpretation of antigen
expression. This module will help the reader understand the technical background of compensation,
provide guidance in optimizing instrument settings and some basic troubleshooting tips specific to the
Navios.

A previous ICCS Module entitled “Instrument optimization - Adjusting PMT voltages and compensation”1
should be read as a prerequisite to this module.

Applicable fluorochromes for the 10-color Navios flow cytometer:

Blue laser line

Navios Detector FL1 FL2 FL3 FL4 FL5
ECD PC5.5 PC7
Fluorochrome FITC PE
(PE-Texas Red) (PE-Cy5.5) (PE-Cy7)

Red laser line Violet laser line

Navios Detector FL6 FL7 FL8 FL9 FL10

Fluorochrome APC APC-A700 APC-A750 Pacific Blue Krome Orange

The recommendations in this document are based on the spectral properties of these 10 fluorochromes.
Other fluorochromes are commercially available that can be used in place of some of these
fluorochromes, however their spectral properties may be different.

1
 FLOW-373905.18

Spectral emission properties of applicable Navios fluorochromes:

Spectral emission graph obtained

from:

[Link]
n/home/life-science/cell-
analysis/labeling-
chemistry/fluorescence-
[Link]

Spectral spillover occurs most

significantly from a fluorochrome into
an adjacent detector on the same
laser line.
For example, light from the FITC
fluorochrome will also be detected by
FL2, which is used to detect PE. This is
illustrated in this uncompensated plot
of a FITC single-stained sample. The
positive signal observed in FL2 is due
to spillover from FITC.

Spectral spillover also occurs

between laser lines. For example, this
is an uncompensated plot of a PC5.5
single-stained sample. Light emitted
by PC5.5 spills over into FL7, which is
the detector used for APC-A700.

2
 FLOW-373905.18

Typical compensation matrix from a 10-color application on the Navios

To correct for spectral spillover, interfering light is subtracted from the total light entering each
detector. The relative amount of light subtracted is represented as a percentage in a compensation
matrix.

In the Navios software, the columns of the compensation matrix represent each fluorochrome, and the
rows represent each detector (schematically shown below). Therefore, when reading down a column,
these are the compensation values required to correct spillover from the fluorochrome listed (at the top
of the column) into other detectors (listed as FL1 through FL10). The size of each bubble is directly
proportional to the amount of spillover from the fluorochrome (column) into the detector (row).

(This schematic matrix was based on a compensation matrix obtained on a Navios flow cytometer following voltage setup as per the previous
module.1 Spillover values will vary depending on voltage setup.)

The percentage obtained for each position in the matrix is influenced by the voltages set for each
detector. Therefore, voltages can be adjusted to minimize compensation values (refer to the previous
module for guidance on setting voltages).1 The larger the value in the matrix, the more distorted or
spread-out the population may appear. Antibody conjugate selection and panel design will also
influence how distortion/spreading affects sensitivity of detection of dim antigens.

Details of Navios-specific panel design can be found in the “Know Your Flow” Panel Design Resource
from Beckman Coulter (Appendix A at the end of this document).

3
 FLOW-373905.18

Single-stained controls for generating a compensation matrix

Historically, lymphocytes or CYTO-COMP/CYTO-TROL Cells stained with anti-CD45 conjugated to each of

the 10 fluorochromes have been used to prepare single-stained controls, which are then used with the
AutoSetup algorithm in the Navios software. The algorithm only requires a positively stained population
to calculate a compensation matrix, and it assumes that the negative peak is located at 0.3 on the scale.

An antibody against an abundantly expressed antigen (e.g., CD3, CD4, CD8), or the same fluorochrome-
antibody conjugate used in a lab’s fully stained panel can also be used, as long as a bright positive
population is generated. An advantage of using the actual conjugate present in the fully stained panel is
that the spillover properties of tandem dyes will be accurately calculated.

VersaComp antibody-capture beads (For Laboratory Use Only, and not for use in diagnostic procedures)
may also be used. Further details on this topic, as well as how to use Kaluza (For Research Use Only. Not
for use in Diagnostic Procedures) as an alternative method for generating a compensation matrix
(instead of AutoSetup), can be found in a previous ICCS module.1

Adjusting/troubleshooting compensation after AutoSetup

For a 10-color application, the compensation matrix automatically generated by AutoSetup must be
manually fine-tuned and validated. After the single-stained samples have been acquired during
AutoSetup, there is an opportunity for a verify tube to be acquired. It is possible to adjust compensation
while the verify tube is acquiring. However, due to the more complicated nature of 10-color
compensation, alternative analysis programs such as Kaluza can be used to adjust compensation on the
acquired LMD file. A major advantage to this is that logicle scales can be used to properly visualize the
data and examine potential compensation issues. Also, the lab staff can take their time to adjust
compensation, instead of being rushed to complete it as the tube is actively acquiring.

The following examples demonstrate common fluorochrome combinations that typically require manual
adjustment. The advantages of using logicle scaling in Kaluza are also shown.

4
 FLOW-373905.18
Example #1: PE & ECD

Initial gating plots:

Correct compensation can be visually

confirmed by the appearance of
Correct compensation: symmetrical populations. Logicle
Log scale Logicle scaling is required for such visual
judgment. A more accurate way of
determining correct compensation
can be accomplished by observing the
means or medians of the populations.

For instance, correct compensation of

PE spillover into FL3 (ECD) can be
visually estimated by whether the
population in quadrant A +- appears
symmetrical (e.g., around the dashed
green line provided for reference),
when logicle scales are used. Note
that this is more difficult to confirm
when log scales are used.

In this example, correct

Over-compensation: compensation can be further
Log scale Logicle confirmed by looking for similar Y-
medians between quadrants A - - and
A +- (green arrows).

Laboratories should develop their

own criteria for acceptability for
compensation settings, e.g.,
1. Visual: Populations should
appear symmetrical;
populations not be "crushed"
against axis or "bleed" into
the A++
2. Numerical: develop
acceptable range for medians

5
 FLOW-373905.18

Example #2: PE & ECD

Initial gating plots:

Over-compensation of
PE spillover into FL3
(ECD) leads to loss of
CD19+ Lambda+ cells
and leads to a
Kappa/Lambda ratio of
3/1

Corrected
compensation:
Kappa/Lambda ratio is
now 1.4/1

6
 FLOW-373905.18

Example #3: PE & ECD

In this example, a bright lambda-PE positive population causes compensation issues with CD10-ECD. If
this spillover is under-compensated, a CD10+Lambda+ population could be erroneously reported:

Under-Compensation Corrected Compensation

FL3-FL2 = 50% FL3-FL2 = 53%

7
 FLOW-373905.18

Example #4: PC5.5 & PC7

Gating plots:

Excess (under-compensated) spillover

signal from PC5.5 into FL5 (the detector
used for PC7) (left plot) leads to CD11b
monos (green) and grans (brown) to
appear artificially CD34-positive (right
plot).

Optimal compensation between

PC5.5 & PC7

8
 FLOW-373905.18

Example #5: APC-A700 & APC-A750

The following example illustrates the “spread-out” nature of populations stained with APC-A700 and
APC-A750. These two fluorochromes exhibit spillover into each other’s detectors, leading to a post-
compensation spreading effect. In addition, these fluorochromes emit light at relatively long
wavelengths, which is less efficiently detected by photomultiplier tube (PMT) detectors. This leads to a
greater error in the measurement of fluorescent light, and therefore a more spread-out appearance of
the population. Logicle scaling enables the full population to be viewed onscale, which helps determine
if compensation appears correct.

Log scale Logicle

After editing the compensation matrix in Kaluza:

Once the compensation has been adjusted in Kaluza, this edited matrix can be entered into the
appropriate cytosettings protocol in the Navios software, so that future LMD files will be recorded with
this edited matrix (see detailed instructions in a previous ICCS module).1 If the fluorescence emitted by
all 10 fluorochromes is stable day to day, this edited compensation matrix should not need to be
adjusted. This can be confirmed by running a verify tube routinely.

9
 FLOW-373905.18

Verify tube

Compensation issues can be identified and corrected by observing populations in the single-positive
quadrants of a bivariate plot. Therefore, it is ideal to use a combination of antibodies that yield as many
single-positive populations as possible on plots that will have significant compensation applied. Ideally, a
lab would use antibodies that are already in their inventory that yield clearly defined and bright
populations.

A Kaluza template can be used to facilitate compensation verification on a routine basis. This template
should consist of all possible bivariate plots representing spillover between all channels. The plots most
likely to require adjustment can be placed at the top of the protocol (refer to the Compensation Bubble
Matrix above). Another advantage of using Kaluza software for routine verification of compensation is
that the data can be exported in an electronic PDF format as documentation of the compensation
verification and to meet accreditation and regulatory requirements.

Also, it is recommended to incorporate all possible dot plots in specimen analysis templates to evaluate
compensation per specimen, as the applied compensation may not be appropriate due to specimen-to-
specimen variation. For example, in cases of bright over-expression of a marker, care should be taken to
observe dot plots displaying the over-expressed marker versus parameters that receive spillover from
that marker. These plots may require compensation adjustments per specimen.

A reference verify tube data file (e.g., one that was acquired immediately after the assay and
compensation were optimized) can be used to compare to future verify tube samples for routine
monitoring purposes. Verify tubes should be acquired and evaluated on a routine basis.

Re-compensation should be considered according to manufacturer’s instructions, potentially including

any changes in voltages, preventative maintenance visits, or major instrument repairs. If the verify tube
sample appears significantly different from the reference verify tube data file, and no major hardware
issues can be identified as the cause, the lab should not jump into re-compensation until the underlying
cause is rectified. For example, reagents (e.g., lot changes) and sample preparation procedures (e.g.,
automated sample preparation instruments) should be investigated first.

10
 FLOW-373905.18

There are many options for types of samples that can be used for a verify tube. Examples are listed
below with advantages and disadvantages:

Sample Type Advantages Disadvantages

Normal Fresh Peripheral Cellular autofluorescence closely Normal samples may be difficult
Blood matches fresh samples; acts as a to obtain day to day; specimen-to-
process control for RBC lysis specimen variability is expected
reagents as well as antibody
reagents
CYTO-COMP/CYTO-TROL Day-to-day appearance of cells is Cells may not express all markers
Cells consistent within a lot present in a peripheral blood
(lyophilized lymphocytes) sample
IMMUNO-TROL Cells Day-to-day appearance of cells is Background autofluorescence is
(stabilized whole blood consistent within a lot; acts as a higher than fresh blood samples
product) process control for RBC lysis due to the solution used to
reagents as well as antibody stabilize the product
reagents

This is not a complete list of possible sample types. Other commercial control samples are also available,
such as CD-CHEX and CRISP control cells. In addition to the advice above, when considering which
control sample to choose for a verify tube, labs should investigate whether the sample possesses an
adequate percentage of positive cells for the antigen of interest (e.g., rare antigens).

11
 FLOW-373905.18

Summary

Achieving optimal compensation is crucial for the correct interpretation of flow cytometry results,
summarized in the following workflow:

Optimize
Run AutoSetup Refine this Enter the final
panel design Perform
to calculate an matrix using compensation
and voltages routine
initial normal and matrix into the
(refer to monitoring of
compensation abnormal settings file in the
previous compensation
matrix samples Navios software
module)1

Further resources for laboratories running flow cytometry lab-developed tests can be found at
[Link]

For any questions on this module or any other suggestions, please email info@[Link]

References:
1. “Instrument optimization - Adjusting PMT voltages and compensation” [Link]

Reviewed and approved by: Ben Hedley and Ahmad Al Samman

The documents posted on the ICCS website may contain product or vendor names which are provided for platform specific guidance. Any
reference within the ICCS Quality and Standards modules to any vendor, product or educational material by trade name, trademark or
manufacturer does not constitute or imply the endorsement or recommendation by ICCS.

12
 FLOW-373905.18

Appendix A: Know Your Flow Panel Design Resource,

Beckman Coulter

*Gallios is For Research Use Only.

Not for Diagnostic Procedures

13