CELL-DYN Ruby

Casebook

Principles
and Practice


 Introduction

Automated instruments such as the CD-Ruby are used in clinical laboratories to provide a series of analytical measure-
ments collectively referred to as the Full or Complete Blood Count (FBC/CBC). This often represents the first stage of
haematological investigation, with the data being typically used to make decisions about whether or not additional follow
up is required.

In a laboratory, the processing and subsequent review of patient sample results is ideally designed to ensure maximal
detection of clinically relevant abnormalities while at the same time minimising unnecessary supplementary testing. As the
efficiency of CBC data review consequently has a significant impact on overall performance efficiency, a good understand-
ing of numerical and graphical outputs provided by an analyser is important. Similarly, haematology instrument reports are
often accompanied by warning indicators (flags) for certain abnormalities and it is further necessary that reviewers have
access to descriptions of flagging logic and significance so that they can be best applied to laboratory decision processes.

As part of Abbott’s commitment to customer support, we have produced this CD-Ruby instrument guide to illustrate
representative haematological abnormalities with associated graphical displays and flagging patterns. Some comparative
evaluation data obtained from two independent performance evaluations of the CD-Ruby is also included where a particu-
lar understanding of instrument efficiency is required.

Dr Mathie Leers
Department of Clinical Chemistry and Hematology, Atrium Medical Center Parkstad, Heerlen, The Netherlands
([email protected])

Dr Isabel Garcia
Departement de Biologie Medicale, CH de Versailles, Paris, France
([email protected])

Dr Fadila Souni
Laboratoires Abbott, Rungis, Paris, France
([email protected])

Dr Colin Stephen Scott

Abbott Diagnostika GmbH & Co. KG, Wiesbaden-Delkenheim, Germany
([email protected])


Contents

Page
Primary CD-Ruby Measurements and Graphics.....................................................................4

1 a) Optical White Blood Cell (WBC) Differential...................................................................................................... 4

b) WBC Differential Graphical Displays................................................................................................................. 4
c) Platelet/Red Blood Cell (RBC) Graphical Displays............................................................................................ 5
Processing Modes.................................................................................................................6

2 a) Routine CBC Mode.......................................................................................................................................... 6

b) NOC (Nuclear Optical Count) Mode................................................................................................................. 6
c) RRBC (Resistant Red Blood Cell) Mode........................................................................................................... 6
Sample Alerts and Flagging...................................................................................................6
a) General Overview............................................................................................................................................. 6

3
b) WBC Count Alerts........................................................................................................................................... 7
c) WBC Differential Alerts..................................................................................................................................... 9
d) RBC Morphology Flag.................................................................................................................................... 11
e) Platelet Alerts................................................................................................................................................. 11
f) ATYPDEP Flag............................................................................................................................................... 12
Processing Modes and Flagging Flow Charts......................................................................12
a) Normal Operational (CBC/Patient) Mode........................................................................................................ 12
b) Normal Operational (CBC/Patient) Mode........................................................................................................ 13
c) Normal Operational (CBC/Patient) Mode........................................................................................................ 13

4 d)
e)
f)
Nucleated Optical Count (NOC/FWBC) Mode................................................................................................ 13
Resistant Red Blood Cell (RRBC) Mode......................................................................................................... 14
Resistant Red Blood Cell (RRBC) Mode......................................................................................................... 14
g) Resistant Red Blood Cell (RRBC) Mode......................................................................................................... 14
h) Resistant Red Blood Cell (RRBC) Mode......................................................................................................... 15
i) Resistant Red Blood Cell (RRBC) Mode......................................................................................................... 15

5 CD-Ruby Performance Studies............................................................................................16

Case studies.......................................................................................................................18
Case 1: Normal Haematological Profiles.......................................................................................................... 18
Case 2: RBC Microcytosis............................................................................................................................... 19
Case 3: RBC Macrocytosis............................................................................................................................. 21
Case 4: RBC Dimorphism............................................................................................................................... 23
Case 5: Granulocytosis................................................................................................................................... 24
Case 6: Immature Granulocytes (IG)................................................................................................................ 26
Case 7: Eosinophilia........................................................................................................................................ 27
Case 8: Basophilia.......................................................................................................................................... 28
Case 9: Lymphocytosis................................................................................................................................... 29

6 Case 10:
Case 11:
Case 12:
Variant Lymphocytes.......................................................................................................................... 31
Fragile Lymphocytes........................................................................................................................... 32
Monocytosis....................................................................................................................................... 33
Case 13: Acute Leukaemia............................................................................................................................... 34
Case 14: Neutropenia....................................................................................................................................... 36
Case 15: Pancytopenia..................................................................................................................................... 37
Case 16: NWBC Flagging Alert......................................................................................................................... 38
Case 17: Resistant Red Blood Cells (RRBC)...................................................................................................... 39
Case 18: Thrombocytopenia............................................................................................................................. 40
Case 19: Thrombocytosis................................................................................................................................. 42
Case 20: Malaria............................................................................................................................................... 43
Case 21: Reticulocyte Analysis.......................................................................................................................... 46


1 Primary CD-Ruby Measurements and Graphics
1a) Optical White Blood Cell (WBC) Differential
The WBC differential is a central part of the routine CBC, and an awareness of instrument processing
mechanisms together with associated graphical outputs is important for clinical and laboratory interpreta-
tion. The CD-Ruby CBC comprises 22 different parameters with many of the primary measurements being
made by laser light flow cytometry. For the WBC differential in particular, four individual optical scatter char-
acteristics are simultaneously recorded for each cellular event as follows:
• SIZE (0° to 3°) – this low-angle scatter measurement is mainly used to assess cell size although other
factors such as the shape of the cell and its refractive index may also be relevant.
• COMPLEXITY (10°) – provides information (refractive and reflective) about internal and external cellular
structures.
• LOBULARITY (90°-Polarised) – primarily a measurement that reflects the type and density of cytoplasmic
granulation.
• GRANULARITY (90°-Depolarised) – used to differentiate the two main granulocyte fractions in blood. In
contrast to neutrophils, eosinophils specifically depolarise laser light at 90°.
Information from these four channels is then used to construct the five-part WBC differential by a process
known as Multi-Angle-Polarised-Scatter-Separation (MAPSS).

1b) WBC Differential Graphical Displays

There are six selectable graphical displays available for visualisation of WBC population distributions and in
each of these, the five main fractions are colour coded (neutrophils – orange, eosinophils – green,
basophils – black, monocytes – purple, and lymphocytes – blue) for easier operator review of population
distributions and frequencies.

Figure 1

The most widely used CD-Ruby plot is SIZE versus COMPLEXITY (0°/10°) [Figure 1a] as this provides:
(i) a check on the general appearance of the white blood cell differential
(ii) a means of validating the separation of lymphocytes from monocytes
(iii) information about possible interferences in the white blood cell count/differential

Individual features of particular note in normal SIZE/COMPLEXITY plots include the location of eosinophils
overlapping the lower region of the neutrophil cluster, the clear separation of granulocytes from mononu-
clear cells, and the occurrence of two populations clusters corresponding to lymphocytes and monocytes.

In the GRANULARITY versus LOBULARITY (90°-D/90°-P) plot [Figure 1b], the separation of neutrophil
(orange) and eosinophil (green) granulocytes can be seen. While neutrophils and eosinophils both diffract
polarised laser light at 90°, because they are multi-lobed, eosinophil granules have the additional property
of being able to depolarise laser light (y axis). In normal blood samples, eosinophils are the only WBC that
can depolarise laser light but in some abnormal conditions other atypical depolarising events may be seen.
Of perhaps the most importance is the association between the appearance of atypical depolarising
monocytes and malaria (see case 20).

Of the other WBC differential graphics, the COMPLEXITY versus GRANULARITY (10°/90°) [Figure 1c] and
SIZE versus LOBULARITY (0°/90°) [Figure 1d] plots are perhaps the most informative. In the first of these,
light scatter at 10° and 90° is used to separate polymorphonuclear (neutrophils and eosinophils) from
mononuclear cells (lymphocytes, monocytes and basophils) using the displayed discriminator line. Interest-
ingly, basophils are nominally classified as mononuclear cells by the CD-Ruby MAPSS process because
exposure to the leukocyte diluent causes degranulation. In contrast, the 0°/90° plot gives indications about
the presence of large-sized cells of the blast type which are typically seen as an extension of the monocyte
population cluster (purple) along the 0° axis.

1c) Platelet/Red Blood Cell (RBC) Graphical Displays

The CD-Ruby provides a number of selectable graphical displays comprising either histograms or two-
dimensional plots [Figure 2].

Figure 2

Of these, the combined optical display of 0° versus 10° [Figure 2a] is perhaps the most important as it
provides an immediate impression of population separation between platelets (orange events) and RBC
(red events). When an overlap of these populations is seen, the possible existence of RBC microcytes
or schistocytes should be considered. In addition, this plot can also reveal abnormalities associated with
platelet aggregates and fragile WBC (FWBC). Of the other display options, the histograms of red cell
[Figure 2c] and platelet volume [Figure 2f] distributions are also useful for data review.

Figure 3

• Platelet Aggregates: The occurrence of events above the platelet population cluster on the 0°/10° plot
[Figure 3a) suggests the possible existence of platelet aggregates, or giant platelets, especially when
accompanied by URI/NWBC (Upper Region Interference/Non-White Blood Cell) warning flags or an
invalidated MPV (Mean Platelet Volume).
• Fragile White Blood Cells [Figure 3b]: On the 0°/10° plot, the fragile lymphocytes of conditions such
as CLL can often be seen as a second blue cluster to the left of the red blood cells (arrow). This atypical
location of WBCs suggests the presence of bare nuclei as opposed to intact cells.

2 Processing Modes
2a) Routine CBC Mode
EDTA-anticoagulated blood samples are normally processed in the first instance using the CD-Ruby CBC
operational mode. If the resulting data do not show either a FWBC (Fragile White Blood Cell) or RRBC
(Resistant Red Blood Cell) flag, then there is no need for further analyser processing. However, if either of
these flags is triggered then one of the following two processing options needs to be used.

2b) NOC (Nuclear Optical Count) Mode

The NOC mode, sometimes referred to as the FWBC (fragile white blood cell) mode, provides an optical
count of WBC nuclei after the cytoplasm of white cells has been selectively stripped. This method of sam-
ple analysis is necessary when the FWBC warning flag is triggered by a sample run in the normal opera-
tional patient CBC mode because fragile or age-deteriorated WBC may not be included in the optical
WBC count estimate. WBC counts (including bare nuclei in cases of lymphoproliferative disease) can be
obtained by the NOC WBC method with linearity to 250 x 109/L.

2c) RRBC (Resistant Red Blood Cell) Mode

The presence of lysis-resistant red blood cells generates WBC and NRBC/RRBC warning indicators which
invalidate white blood cell results. Red blood cells containing foetal haemoglobin (from premature, new-
born or haemoglobinopathy patients), or those with ‘target cell’ morphology, are relatively resistant to lysis
with CD-Ruby WBC reagents and this may result in an overestimated WBC count. The CD-Ruby RRBC
mode of analysis overcomes the influence of lysis-resistant red cells and provides accurate white blood cell
counts and differentials.

3 Sample Alerts and Flagging

3a) General Overview
As part of the CD-Ruby WBC count and differential analyses, a number of instrument algorithms are used
for internal data validation. Among these are included:

(a) The amount of stroma/debris overlapping the lower CD-Ruby size threshold. If events in this region cor-
respond to 2.9% or more of the WBC count, then there is an increased likelihood of lysis-resistant
(RRBC) or nucleated (NRBC) red cells.
(b) The degree of separation between events in the stroma region and the lymphocyte population. If this
threshold is indistinct, then the accuracy of the WBC count and differential may be affected.
(c) The rate of white cell optical counts (WOC) during the sample measurement period. A declining WBC
count rate is important because it can indicate the presence of either RRBC or FWBC.
(d) The degree of lymphocytosis. This is incorporated as a predictive factor in relation to the possible exist-
ence of fragile lymphocytes (FWBC) or variant lymphoid cells (VAR LYM).
(e) The presence of a second population of WBC events next to the RBC fraction in the 0°/RBC 10° plot.
When representing 15% or more of the total WBC count, WBC events in this location suggest the pres-
ence of FWBC.


CD-Ruby instrument warning alerts (flags) on the CD-Ruby are designed to call attention to one or more
aspects of the sample or haematological picture. As the analyser uses a dedicated optical approach for
WBC differential analysis, then the morphological characteristics of cells seen in a stained film will often
bear some direct correspondence to the MAPSS differential. This is particularly relevant to the detection of
abnormal cells, which are atypically located in graphical plots compared to their normal counterparts.

Flags should be considered as part of the decision processes regarding whether or not to examine a
stained film or to undertake supplementary analyses. As with all algorithmic procedures, the triggering of
any given flag is highly dependent on detecting cellular features that deviate outside of pre-defined limits.
However, because atypical changes may simultaneously affect more than one cellular characteristic, flexi-
bility should be used when interpreting flagging information. For example, while a flag may specifically indi-
cate the presence of abnormal cells the presence of the same flag may also be triggered less specifically
by a related morphological abnormality. The most common situations where this occurs are BAND/IG and
VAR LYM/BLAST. In broad terms, the following guide should be used as an initial basis for interpreting the
significance of the CD-Ruby WBC differential flags:

CD-Ruby Flag Possible Reasons for Flag

BAND (a) Band (non-segmented) neutrophils
(b) Occasionally when immature granulocytes are present
(c) Morphological changes associated with sample ageing/deterioration
IG (a) Immature granulocytes (above a frequency of 3%)
(b) Occasionally when non-segmented neutrophils are increased
(c) Morphological changes associated with sample ageing/deterioration
VAR LYM (a) Variant lymphoid cells (reactive or malignant)
(b) Small lymphoid appearing blasts
(c) Morphological changes associated with sample ageing/deterioration
BLAST (a) Blast cells (1% or more)
(b) Large atypical lymphoid cells (reactive or malignant)
(c) High mononuclear cell counts

3b) WBC Count Alerts

The following flag descriptions are applicable to samples that are analysed in the normal operational
patient (CBC) mode. Note that other flagging strategies are used when samples are analysed in either the
CD-Ruby NOC or RRBC modes.

The WBC warning flag is seen in two situations that can be distinguished by the patterns of accompanying
flags. The other flags that are related to the WBC count determination (and of course the associated differ-
ential) are NWBC, NRBC/RRBC and FWBC.


 WBC Flag
(a) When flag occurs simultaneously with NRBC/RRBC Flags
Cause Action Required
When both of the following conditions are met: Lysis-resistant RBCs or nucleated red blood cells
(NRBC) interfering with the WBC measurement
1. A declining optical WBC kinetic count rate is
typically causes this combination of flags to be trig-
detected
gered.
AND
Repeat analysis using the CD-Ruby RRBC mode to
2. More than 2.9% of the WBC count is located in eliminate interference caused by lyse-resistant red
the stroma region below the threshold in the blood cells.
SIZE/COMPLEXITY plot
If the WBC flag persists in the RRBC mode, review
a stained smear for the presence of NRBC.
When analysed in the RRBC mode, the reported
WBC count is accompanied by either a (NOC),
(WOC) or WBC flag depending on an internal deci-
sion matrix.

WBC Flag
(b) When flag occurs simultaneously with FWBC, VAR LYM and DFLT(NLMEB) Flags
Cause Action Required
When one or more of the following conditions is This situation is typically caused by the presence of
met: fragile white blood cells (typically lymphocytes in
most cases) which may not be included in the opti-
1. A declining optical WBC kinetic count rate is
cal WBC count (WOC).
detected but there is < 2.9% stroma interference
Repeat analysis using the CD-Ruby NOC mode to
2. Lymphocytes comprise > 80% of the differential
obtain correct WBC count and differential.
and the WBC count is > 4.1 x 109/L but there is
< 2.9% stroma interference When analysed in the NOC mode, the reported
WBC count is accompanied by a (NOC) flag next
3. > 15% Fragile WBCs are detected in the RBC/Plt
to the WBC count.
plot

NWBC Flag
Cause Action Required
The count in the region below the WBC differential Correction of the WBC count is not required.
threshold on the SIZE/COMPLEXITY 0°/10° scatter
If no other suspect parameter flags are present, the
plot is > 2.9% of the total WBC
WBC and Differential may be reported.
BUT
The NWBC flag indicates the possible presence of:
there is no declining optical WBC count rate • Low levels of NRBCs
• Unlysed RBCs
• Platelet clumps
• Giant platelets
A stained film should be reviewed if considered
necessary.


NRBC and RRBC Flags
Displayed next to the MONO %
Cause Action Required
When both of the following conditions are met: Review a stained smear for the presence of NRBC
and follow your laboratory’s review criteria.
1. A
 declining optical WBC kinetic count rate is
detected In the normal CD-Ruby operational mode, the
NRBC flag always appears simultaneously with the
AND
WBC and RRBC flags.
2. M
 ore than 2.9% of the WBC count is located in
The sample should be reanalysed in the CD-Ruby
the stroma region below the threshold in the
RRBC mode.
SIZE/COMPLEXITY plot

FWBC Flag
Displayed next to the MONO %
Cause Action Required
When one or more of the following conditions is The FWBC flag always appears in the normal CD-
met: Ruby operational mode in conjunction with WBC,
VAR LYM and DFLT(NLMEB) flags.
1. A declining optical WBC kinetic count rate is
detected but there is < 2.9% stroma interference The sample should be reanalysed in the CD-Ruby
NOC mode in order to provide the correct WBC
2. Lymphocytes comprise > 80% of the differential
count and differential.
and the WBC count is > 4.1 x 109/L but there is
< 2.9% stroma interference Review a stained smear for the presence of abnor-
mal lymphoid cells and follow your laboratory’s
3. > 15% Fragile WBCs are detected in the RBC/Plt
review criteria.
plot

3c) WBC Differential Alerts

BAND Flag
Displayed next to the NEU %
Cause Action Required
The BAND flag is triggered if any of the following Review a stained smear for the presence of bands
conditions is met: and follow your laboratory’s review criteria.
1. C
 ells in the region of scatter (on the 0°/10° plot) When bands are present, they are included in the
where bands are typically located corresponds total neutrophil count.
to >12.5% of the WBC differential
2. The ratio of suspected bands to mature neu-
trophils is > 50%
3. The CV of the neutrophil cluster on the 0° axis
exceeds pre-defined distribution limit


 IG (Immature Granulocyte) Flag
Displayed next to the NEU %
Cause Action Required
Cells in the IG the region of scatter (on the 0°/10° Review a stained smear for the presence of imma-
plot) correspond to > 3% of the WBC differential ture granulocytes and follow your laboratory’s
review criteria.
When immature granulocytes are present, they are
included in the total neutrophil count.

BLAST Flag
Displayed next to the LYM %
Cause Action Required
Cells in the region of scatter (on the 90°/0° plot) Review a stained smear for the presence of blasts
where blasts are typically located corresponds to and follow your laboratory’s review criteria.
1% or more of the WBC differential
Note that when blasts are present, they are typical-
ly included in the CD-Ruby monocyte count.

VAR LYM Flag

Displayed next to the Lymph %
Cause Action Required
1. The position, density or CV of the lymphocyte Review a stained smear for the presence of variant
cluster on the SIZE/COMPLEXITY (0°/10°) scat- lymphoid cells and follow your laboratory’s review
ter plot exceeds pre-defined limits criteria.
2. The absolute lymph count exceeds a predefined When Variant Lymphoid cells are present, they are
limit and the ratio of neutrophils to lymphocytes included in the lymphocyte count.
is < 0.2:1
Note that this flag may be displayed singly or in
3. Whenever the FWBC flag is triggered combination with the BLAST flag. If the flag is dis-
played with the blast flag it is displayed as VARL/
BLAST.

DFLT(NLMEB) Flag
Cause Action Required
When one of the following conditions is met: If the DFLT(NLMEB) flag is seen with the FWBC
flag, then reanalyse in the NOC mode.
1. If the FWBC flag is triggered
If no suspect parameter flags are present, the WBC
2. The numbers of WBCs are too low to discrimi-
and differential may be reported.
nate between individual leukocyte populations
3. The threshold between mononuclear and poly-
morphonuclear cells has too much interference

DFLT(NE), DFLT(LM) and DFLT(B) Flags

Cause Action Required
When one of the following conditions is met: The letters in parenthesis indicates which leukocyte
population(s) is suspect.
1. The numbers of WBCs in any given leukocyte
population are too low to allow floating thresh- Visual examination of the relevant scatter plots
olds to be used usually provides explanation for the flag.
2. The discrimination between individual popula- Blood film review may be required.
tions is indistinct
10
3d) RBC Morphology Flag
RBC MORPH Flag
Cause Action Required
One or more of the following parameters is outside Review a stained smear for abnormal RBC mor-
of the following limits: phology and follow your laboratory’s review criteria.
MCV <80 fL or >100 fL
MCH <25 pg or >34 pg
MCHC <29 g/dL or >37 g/dL
RDW >18.5 %

3e) Platelet Alerts

Platelet LRI (Lower Region Interference) Flag
Displayed next to the MPV
Cause Action Required
Interference in the lower threshold region (1 – 3 fL) Check the background count and if it exceeds
is greater than pre-defined limits recommended limits, troubleshoot accordingly.
If the background count is acceptable, repeat the
analysis or review a stained smear to verify the CD-
Ruby platelet count and determine the cause of the
interference.
Interferences may be biological or non-biological:
• Debris (dirty aperture)
• Contaminated reagent
• Electronic noise
• Microbubbles
• Cell fragments
• Protein aggregates

Platelet URI (Upper Region Interference) Flag

Displayed next to the MPV
Cause Action Required
Interference in the upper threshold region Review the red cell and platelet histograms. Note
(15 – 35 fL) is greater than pre-set limits that a “bumpy” platelet histogram may indicate the
presence of platelet clumps.
If the MCV is low and/or the histogram indicates an
overlap (poor separation at the upper discriminator)
in the RBC and platelet populations, review a
stained smear to determine the cause and confirm
the platelet count.
Possible URI interferences include:
• Microcytic RBCs
• Schistocytes
• Leukocyte fragments
• Sickle cells
• Platelet clumps or Giant Platelets

11
 No MPV result displayed (data suppressed)
Cause Action Required
The PLT histogram did not meet expected criteria Review a stained smear for abnormal platelet mor-
(non-log normal distribution) phology, or the presence of platelet aggregates,
and follow your laboratory’s review criteria.
Verify the platelet count.

3f) ATYPDEP Flag

ATYPDEP Flag
Cause Action Required
MAPSS 90°-P analysis indicates the presence of Visual review of the GRANULARITY/LOBULARITY
atypical depolarising mononuclear cell events plot should examine the nature and frequency of
atypical depolarising events.
Atypical depolarising monocyte (purple) events
suggest the possibility of malaria infection and
appropriate investigations should be undertaken for
further clarification.
Other causes for atypical depolarising events are
varied and include lipids and intracellular (RBC)
haemosiderin.

4 Processing Modes and Flagging Flow Charts

4a) Normal Operational (CBC/Patient) Mode
Samples with Insignificant (< 2.9%) Stroma

12
4b) Normal Operational (CBC/Patient) Mode
Samples with Significant (≥ 2.9%) Stroma

4c) Normal Operational (CBC/Patient) Mode

Supplementary Condition Irrespective of Stroma Level

4d) Nucleated Optical Count (NOC/FWBC) Mode

13
 4e) Resistant Red Blood Cell (RRBC) Mode
Decision Outcomes

4f) Resistant Red Blood Cell (RRBC) Mode

Decision Outcomes: NOC = WOC for Samples with < 2.9% Stroma

4g) Resistant Red Blood Cell (RRBC) Mode

Decision Outcomes: NOC = WOC for Samples with ≥ 2.9% Stroma

14
4h) Resistant Red Blood Cell (RRBC) Mode
Decision Outcomes: NOC > WOC

4i) Resistant Red Blood Cell (RRBC) Mode

Decision Outcomes: NOC < WOC

15
5 CD-Ruby Performance Studies
Practical evaluations are commonly undertaken by instrument manufacturers in order to validate inter-sys-
tem equivalence as well as to identify areas that may be relevant to user interpretation of data. These eval-
uations tend to provide large amounts of data which are typically converted to visual displays (plots or
graphics) that allow easier assimilation. In this user guide, the following series of plots were obtained from
two independent evaluations of the CD-Ruby in the Department of Clinical Chemistry and Hematology,
Atrium Medical Center Parkstad, Heerlen, The Netherlands, and the Departement de Biologie Medicale,
CH de Versailles, Paris, France. The comparative analysers used in these studies were the Abbott CELL-DYN
CD4000 and the Sysmex XT-2000, while reference microscopy was used to assess the accuracy and reli-
ability of the automated CD-Ruby WBC differential.

(a) (b)

(c) (d)

The four plots above show the comparative estimates of (a) granulocyte, (b) eosinophil, (c) lymphocyte and
(d) monocyte absolute counts obtained with the CD-Ruby compared to reference microscopy. Data is
shown on log scales, with the continuous black trendlines and equations being obtained by Passing-Bablok
analysis of method agreement. The 655 different samples represent the combined contributions of the two
evaluation sites, and the results indicate generally very good equivalence. The only significant outliers in the
lymphocyte and monocyte plots correspond to samples where blasts seen by microscopy were categorised
as lymphocytes and/or monocytes in the automated CD-Ruby 5-part differential.

16
The above plot shows the comparative determinations of absolute platelet counts obtained with the CD-Ruby
and CELL-DYN 4000 optical methods (n = 293). There is essentially no significant difference between the
instruments for the sample ranges tested although it needs to be noted that relatively few samples with
extreme low (< 20 x 109/L) counts were included in this sample series.

This plot shows the comparative determinations of absolute reticulocyte counts between the CD-Ruby and
the CELL-DYN 4000 (n = 293). Although there is some variability between the CD-Ruby dye scatter and
CD4000 fluorescent methods, and a general trend for higher counts with the CD-Ruby, there are no major
outliers.

17
6 Case studies
Case 1: Normal haematological profiles
Case Images

1a

1b

18
CaseText
The graphical displays shown with these two haematologically normal samples are those that are most
commonly used for routine laboratory review. For the SIZE/COMPLEXITY plot of WBC distributions, the
following points should be noted:
• clear separation between granulocytes (neutrophils plus eosinophils) and mononuclear cells
(lymphocytes plus monocytes)
• discrete clustering of the basophil population
• eosinophil localisation over the lower size region of the neutrophils
• relatively few events below the size threshold line

In the second WBC plot (GRANULARITY versus LOBULARITY), the separation of eosinophils from neu-
trophils by means of laser light depolarisation can be seen. Note also that the neutrophil population barely
extends to the upper limit of the LOBULARITY axis. This is consistent with a normal degree of segmenta-
tion or lobulation.

The other two plots show the normal Gaussian distribution of RBC volumes (lower left) and the separation
between platelets and RBC (lower right) that would be expected with samples containing normocytic RBC,
normal-sized platelets and an absence of interferences. With the latter plot, the events coded as blue
(WBC) form a single cluster only. A second blue cluster to the left of this would indicate the presence of
fragile WBC.

Morphological Images

Normal segmented neutrophil. Normal lymphocyte. Normal segmented monocyte. Note folded
nucleus and opaque cytoplasm with occa-
sional small cytoplasmic vacuoles.

Normal segmented eosinophil. Normal segmented basophil.

19
 Case 2: RBC Microcytosis
Case Images

2a

2b

Case Text
These samples both show RBC microcytosis. Case 2a is further characterised by moderate anaemia, a
decreased MCHC (< 30 g/dL) and a slightly increased (>15%) RDW. The RBC count is relatively normal, as
are the WBC and platelet results. Review of the graphical displays does not reveal anything of major note
apart from the reduced degree of degree of separation between the platelets and RBC (lower right plot).
This profile is consistent with iron deficiency anaemia (IDA) but supplementary investigations would be
required to further confirm this.

20
Case 2b differs in that the patient is not particularly anaemic and the RBC count is increased. While this
pattern of microcytic erythrocytosis could be suggestive of b-thalassaemia minor, this 51 year old patient in
fact had polycythaemia rubra vera (PRV). Of further note with this case is the presence in the SIZE versus
COMPLEXITY WBC plot of an additional population of events (colour coded pink) below the size threshold
line. These may well correspond to lyse-resistant ‘target RBC’. Note that there is no RRBC flag because
the population of lyse-resistant RBC does not exceed 2.9% of WBC events.

Morphological Images

Red blood cell microcytes on a background Red blood cell hypochromasia in a patient Red blood cell hypochromasia with many
of normocytosis. with iron deficiency anaemia (IDA). Note target cell forms in a patient with b-thalas-
occasional target cells. saemia minor.

Supplementary Notes
Most microcytic disorders arise from decreased haemoglobin production during red cell development
because of either (a) inadequate synthesis of heme – iron deficiency (IDA), (b) globin production defects
(haemoglobin gene abnormalities), (c) decreased synthesis of normal globin chains (thalassaemias), or (d)
decreased synthesis of abnormal globin chains (haemoglobinopathies). RBC microcytosis can also be
seen in some chronic disorders such as Rheumatoid Arthritis (Anaemia of Chronic Disease – ACD), where
defective re-utilisation of iron is thought to play a key role, and the rare hereditary form of sideroblastic
anaemia.

Routinely, a serum ferritin estimation represents the most useful single diagnostic test in recognising simple
IDA although it needs to be interpreted with caution when there is a coexisting inflammatory state, liver
damage or malignancy. Levels of <10 mg/L are typical for iron deficiency while a value of >50 mg/L is
inconsistent with IDA (even in the presence of inflammation or liver disease).

With regards to CBC parameters, various discriminant functions have been proposed to aid the differential
distinction of thalassaemia minor and IDA. These include the ratio of MCV to RBC count, the ratio of MCH
to RBC count, the ratio of RDW to red cell count, a weighted formula comprising both the MCV and RBC
count, and a combined MCV and RDW algorithm. The central role of the RDW in many of these diagnostic
algorithms is based on observations suggesting that microcytic pictures with an increased RDW are more
indicative of IDA while microcytic pictures with a normal RDW are more likely to be due to ACD or thalas-
saemia minor. However, while there is some agreement that it is unusual for a patient with iron deficiency
to have a normal RDW, an abnormal RDW alone has a low specificity for IDA (almost one half of thalassae-
mia minor cases show similar MCV and RDW patterns to IDA).

Another distinction, even though its overall sensitivity is quite poor, is that thalassaemia is more likely to
show both microcytosis and erythrocytosis (68% of cases) compared to IDA (3% of cases). The presence
of target cells in patients with microcytic red cells is, however, less helpful as a discriminator (30% of thalas­
saemia minor versus 18% of IDA). The reticulocyte number may also have some value since it tends to be
low or normal in IDA compared to thalassaemia. Thalassaemia intermedia and major tend to have higher
reticulocyte counts than b-thalassaemia minor, while a-thalassaemia patients have higher absolute reticu-
locyte counts compared to b-thalassaemia.

21
 Case 3: RBC Macrocytosis
Case Images

3a

3b

Case Text
The first sample (Case 3a) with marked macrocytosis (130 fL) shows a non-specific leukocytosis, moder-
ate thrombocytosis and moderate anaemia. The SIZE versus COMPLEXITY WBC plot is unremarkable
apart from the occurrence of relatively poor lymphocyte/monocyte population separation (suggesting the
possibility of some atypical lymphoid cells despite the absence of a VAR LYM flag) and a mild eosinophilia.
By comparison, Case 3b shows a less severe macrocytosis (114 fL) without anaemia, a mild neutropenia

22
with an otherwise normal leukocyte differential, and a normal platelet count. This second case also shows
an extension of the neutrophil cluster along the LOBULARITY axis suggesting the possibility of a right shift
(hypersegmentation). The patient was a 18-year old boy, who went to a general practitioner because of
complaints of weakness and fever.

Morphological Image

Red blood cell macrocytosis.

Supplementary Notes
It is important to note that macrocytosis is frequently observed in the absence of anaemia. Data from early
studies (1964 to 1981) indicated that the commonest single cause of macrocytosis was alcoholism and/or
liver disease (26% of patients), followed by B12 or folate deficiency (21%), and then malignancy/chemo-
therapy (15%). The remaining 38% of cases had a diverse range of clinical pictures and the cause of mac-
rocytosis was actually unexplained in 16% of patients. During this survey period, myelodysplastic syn-
dromes (MDS) were not well defined entities and thus not recognised as a specific cause of macrocytosis.
In a more recent review of almost 40,000 routine outpatient blood counts where MDS was recorded as a
separate category, the commonest causes of macrocytosis were attributed to chronic alcohol abuse
(27%), chemotherapy (23%) and MDS (18%). Only 11% of patients had B12 or folate deficiency.

In children, the causes of macrocytosis are quite different. One US study of 60,000 paediatric blood
counts recorded 146 cases of macrocytosis, one third of which were associated with the use of drugs
such as anticonvulsants, anti-retroviral agents, and immunosuppressants. Other aetiologies included con-
genital heart disease (14%), Downs Syndrome (8%), reticulocytosis (7%) and marrow failure/myelodyspla-
sia (4%). A significant minority of children had no obvious cause for a macrocytosis, although such cases
typically showed only a marginal increase in the MCV, and no cases of B12 and/or folate deficiency were
observed.

When the quantitative degree of MCV measurements in differential diagnosis is considered, it is seen to
be lower than 115 fL in 98% of patients with alcohol-associated macrocytosis, 92% of patients with
macro­cytosis due to chemotherapy, 95% of myelodysplasia cases and 96% of patients with liver disease.
Conversely, an MCV exceeding 115 fL in adults is more likely to be associated with B12 or folate deficien-
cy (although a lower MCV does not exclude this possibility).

The commonest cause of B12 deficiency in the Western Hemisphere is a failure of the stomach to secrete
intrinsic (B12 binding) factor. This disorder, which is known as Pernicious Anaemia, is generally controlled
with maintenance B12 therapy. Other causes of B12 deficiency include gastrectomy, tapeworm (diphyllo-
bothrium latum) infection, ileal resection, serum transcobalamin (I or II) deficiency, and some drug treat-
ments. Dietary insufficiency is rarely a cause of B12 deficiency. Among the elderly, the relatively high preva-
lence of B12 deficiency appears to be mainly due to a decline in gastric function, with at least 30% of
elderly individuals having varying degrees of atrophic gastritis. In these patients, decreased stomach acid
leads to reduced food B12 absorption.

In contrast to vitamin B12 deficiency, there are many different causes of folate deficiency with the com-
monest being dietary insufficiency. Poor dietary intake is often seen in individuals with high alcohol depend-
ency, and may also exist when there is increased body demand for folate such as during pregnancy. Other
causes of folate deficiency include malabsorption, often as a consequence of disorders affecting the upper
small intestine, and treatments with folate antagonists such as methotrexate.
23
 Case 4: RBC Dimorphism
Case Images

4a

4b

Case Text
Both cases show bimodal RBC populations. These profiles are most commonly seen following blood
transfusion for anaemia. Dimorphism may also be seen in iron deficiency patients during therapy, patients
with combined B12/folate and iron deficiency and, rarely the hereditary form of sideroblastic anaemia.
Case 4a was a 82-year old lady known with a colorectal carcinoma. One month before the analysis she
underwent a successful response to a blood transfusion because of a prolonged microcytic anaemia.
Case 4b additionally shows neutrophilia and eosinophilia, as well as the marked microcytosis.

24
Morphological Image

Red blood cell anisocytosis with significant

variation in size. Note occasional sphero-
cyte forms.

Case 5: Granulocytosis
Case Images

5a

25
 5b

Case Text
Both cases show a significant granulocytosis (neutrophilia). For Case 5a, the neutrophil cluster retains a
relatively normal shape but its position is slightly higher on the SIZE axis compared to normal. The sample
also shows monocytosis, moderate normocytic anaemia and a high normal platelet count. The BAND flag
suggests the possibility of a left shift.

For comparison, Case 5b (51 year old male following nephrectomy for renal cell carcinoma) shows a less
marked neutrophilia, with relatively normal RBC and platelet results. With regards to the graphical displays,
the neutrophil population cluster (upper left plot) is also relatively normal but the extended location of the
cluster along the LOBULARITY axis in the other WBC plot (upper right) indicates that the neutrophils are
hypersegmented.

Morphological Images

Non-segmented neutrophil (Band Cell). Neutrophil hypersegmentation.

26
Supplementary Notes
Band Cells: The normal mean percentage of band cells in healthy adults is approximately 6% but there is
wide inter-individual variation and the upper normal limit may extend to 15%. Up to 6 months of age, the
band cell count is approximately 50% higher than adult levels. An increased proportion of band cells is
referred to as a ‘left shift’ and although this is often accompanied by an increase in the absolute neutrophil
count, this may not always be the case.

Conventional teaching suggests that acute inflammation is haematologically indicated by an increased pro-
portion of band cells in the WBC differential but this requires an assumption that morphologists can reliably
and consistently distinguish them from other neutrophils. Studies by the College of American Pathologist
(CAP), however, suggests that inter-laboratory consensus agreement for band cell identification rarely
exceeds 80% and even within individual laboratories the consistent delineation of band cells is problemat-
ic. In addition, poor inter-observer consistency is compounded by the statistical influence of cell counts
which means that (for example) a true band cell count of 10% in a 100 cell differential has a statistical con-
fidence range of 2% to 20%.

In addition to morphological limitations, most clinical investigations have also concluded that the band cell
count is relatively inefficient as a laboratory test for detecting infection. Even when band cells comprise
20% or more of the WBC differential, there is still only a 53% sensitivity (with 79% specificity) in identifying
infection. Consequently, neutrophilia with a left shift should be regarded not as an indicator of infection but
rather as a non-specific response to emotional or physical stress. Conditions in which this can occur
include acute and chronic inflammation, benign and malignant tumours, surgical trauma, haemorrhage or
haemolysis, splenectomy, myeloproliferative disorders, asthma, seizures or drugs. In assessing patients
for possible infection, the most useful haematological information is the absolute neutrophil count and the
presence of immature granulocytes (i.e. myelocytes and metamyelocytes) although these remain relatively
uninformative compared to CRP, PCT and cellular markers such as neutrophil CD64. Moreover, it is
important to note that some infections induce neutropenia rather than neutrophilia, and many infections
actually develop because of a pre-existing neutropenia.

Neutrophil Hypersegmentation: The normal mean neutrophil lobe count is 2.7 and a ‘right shift’ is indicat­
ed when the average lobe count (obtained by examining 100 neutrophils) exceeds 3.5. Formal determina-
tion of a lobe count in this way is, however, rarely performed nowadays and a simpler way of defining hyper­­­
segmentation is to count the proportion of neutrophils with five lobes or more. As normal samples only
show occasional neutrophils with five lobes, hypersegmentation can be alternatively defined as any sample
where >5% of neutrophils have at least five lobes, or where 1% or more of the neutrophils have six lobes.

Neutrophil hypersegmentation is mainly associated with macrocytic anaemia in general and megaloblastic
anaemia in particular. The finding of hypersegmentation indeed appears to be more discriminative than
macrocytosis in the context of B12 and/or folate deficiency. For example, in a review of 25,000 blood
smears, 200 patients were found with hypersegmentation in the absence of macrocytic red blood cells.
On investigation, it was found that most patients with hypersegmentation had folate or vitamin B12 defi-
ciencies or were uraemic. A further series of patients with normal red cell parameters (including haemo-
globin concentration and red cell MCV) where neutrophil hypersegmentation was noted had significantly
reduced serum folate levels. However, although hypersegmentation may be more sensitive (91%) than
either red cell macrocytosis (62%) or increased RDW (54%) values in detecting B12 or folate deficiency, it
is not specific. In various surveys, hypersegmented neutrophils have also been documented in patients
with macrocytosis secondary to chronic alcohol abuse, chemotherapy, myelodysplasia, and liver disease.

27
 Case 6: Immature Granulocytes (IG)
Case Images

6a

6b

Case Text
In contrast to the previous two cases (5a and 5b), the granulocytosis in these two patients is characterised
by distinctively abnormal neutrophil population cluster shapes. This cluster pattern is suggestive of imma-
ture granulocytes (IG), and this possibility is further indicated by the presence of IG/BAND flagging alerts.
There are no other major abnormalities with regards to the WBC differentials although the increased mono-
cyte and basophil counts in Case 6b may have relevance in the context of possible myeloproliferative dis-
ease. Both patients also show a moderate normocytic anaemia and mild to moderate reductions in the
absolute platelet count. Reference microscopic review of these two samples revealed the presence of 19%
IG in Case 6a and 16% IG in Case 6b.
28
Morphological Images

Immature Granulocyte (neutrophilic) showing a Immature Granulocyte (neutrophilic) with a

single nucleus, nucleolation, and cytoplasmic single, slightly indented, nucleus (Myelocyte/
granulation (Promyelocyte/Myelocyte). Metamyelocyte).

Supplementary Notes
Neutrophilia (>7.5 x 109/L) is seen in many clinical states and can be transient or persistent in nature (Table 1).
Neutrophilia can also be further characterised according to the absence or presence of immature granulo-
cytes (IG, promyelocytes, myelocytes and metamyelocytes), where presence is simply defined as the
existence of 2% or more IG on two separate occasions (transient occurrences of low IG numbers have
no significance). On this basis, increased IG on a background of neutrophilia are typical findings in myelo-
proliferative disorders (especially CML but also polycythaemia and thrombocythaemia), some variants of
myelodysplasia with proliferative elements, leukaemoid blood pictures, severe acute infections, and after
growth factor (G-CSF) treatment.

IG can also be seen in the absence of neutrophilia during bone marrow recovery (drug-induced or post-
transplantation) and with some severe bacterial infections. If IG are seen together with erythroblasts, the
term ‘leukoerythroblastic picture’ is used. Interestingly, despite the perceived link between leukoerythrob-
lastic pictures and marrow infiltration by non-haemopoietic tumour, these in fact only represent a minority
of cases. In one survey, only 40% of leukoerythroblastic pictures in adults were directly associated with
marrow infiltration with one half of these being due to leukaemia. In addition, it is relatively rare for patients
to present with leukoerythroblastic pictures secondary to tumour infiltration of bone marrow without a pre-
existing history of malignancy. Of the other clinical conditions in which leukoerythroblastosis is seen are
included infections, haemolytic anaemias, megaloblastic anaemia, inflammatory disorders, haemorrhage,
severe trauma and anorexia. Consequently, it is more appropriate to regard leukoerythroblastosis as a non-
specific phenomenon.

Table 1: Neutrophilia and Clinical Associations

Clinical Association Additional Comments
Acute Infection Commonest cause of neutrophilia. May be associated
with the presence of immature granulocyte forms, and
with morphological changes such as toxic granulation or
Dohle bodies. Counts rarely exceed 30 x 109/L.
Chronic Inflammation Neutrophilia is often accompanied by monocytosis and is
usually less marked than seen in acute infections.
Post-operative, after acute blood loss, Neutrophilia in these situations is usually transient and
following myocardial infarction and following normalises rapidly.
intensive exercise
Drug-induced Neutrophilia Steroids and growth factor treatment in particular may
cause neutrophilia.
Marrow Hyperplasia Neutrophilia may be seen in patients with increased mar-
row haemopoiesis (e.g. chronic or acute haemolytic pic-
tures). Immature granulocytes are often seen.
Primary Malignant Disorders These essentially comprise the myeloproliferative disorders
and some of the myelodysplastic syndromes.
Leukaemoid Reaction A marked neutrophilia, predominantly comprising mature
forms, is occasionally seen in patients with solid tumours. 29
 Case 7: Eosinophilia
Case Images

7a

7b

30
Case Text
These samples show varying degrees of eosinophilia. With Case 7a, the eosinophilia is seen on a back-
ground of an otherwise normal WBC count (and differential), red blood cell and platelet profiles. Note the
clear separation of eosinophils from other leukocytes by virtue of higher depolarisation in the LOBULARITY
versus GRANULARITY plot. In Case 7b, the eosinophilia is more marked, and the WBC differential analysis
also reveals a mild neutrophilia with a higher location for the population cluster on the size axis (upper left
plot). There is also a mild monocytosis and when this is seen it is important to review the degree popula-
tion separation between monocytes and lymphocytes. If, as is seen in this case, the separation is not dis-
tinct, then the possibility of some atypical mononuclear cells should be suspected. The patient is also mild-
ly anaemic and the platelet count is slightly increased. Supplementary CD-Ruby analysis further indicated a
normal reticulocyte count.

Morphological Images

Normal bilobulated eosinophil. Immature eosinophil (metamyelocyte).

Supplementary Notes
Eosinophil counts exceeding 0.42 x 109/L are defined as eosinophilia. Increased eosinophil counts are
generally less than 1.0 x 109/L and are most commonly associated with drug allergies (e.g. gold, sulphona-
mides and penicillin), allergic syndromes such as asthma and urticaria, and parasitic infections (especially
hookworm, tapeworm, filariasis and schistosomiasis). Less frequently, eosinophilia is seen in patients with
some skin disorders (e.g. Pemphigus) and malignant diseases. Profound persistent eosinophilia corre-
sponds to eosinophilic leukaemia or hypereosinophilic syndrome (HES). Idiopathic HES is an important but
relatively rare haematological disorder characterised by three main criteria; (a) persistent eosinophilic gran-
ulocytosis exceeding 1.5 x 109/L for a period of at least 6 months, (b) no detectable underlying cause such
as parasitic infection or allergic reaction, and (c) the presence of signs and symptoms of tissue damage
(cardiac and/or pulmonary infiltration neurological manifestations etc.).

31
 Case 8: Basophilia
Case Images

8a

8b

Case Text
These two examples of basophilia are seen on relatively complex haematological pictures. Case 8a shows
a basophilia of 0.5 x 109/L (4.3%) together with marked monocytosis, mild normocytic anaemia and mild
thrombocytopenia. The basophil cluster is clearly seen in the SIZE versus COMPLEXITY plot, and this also
shows an abnormal neutrophil cluster with a subpopulation extending into the upper region of the size
axis. This pattern indicates the possible presence of immature granulocytes and is consistent with the IG/
BAND flag. This plot also shows a minor number of pink-coded events below the size threshold but these
represent < 2.9% of WBC as there is no RRBC, NRBC or NWBC flag.
32
The second example (Case 8b) has a basophilia of 0.62 x 109/L and this is accompanied by a neutrophil
leukocytosis in particular and a non-specific increase of all myeloid cells in general. The neutrophil cluster is
not as irregular as the previous case but it is sufficiently abnormal to generate the BAND flag. The patient
also has a macrocytic anaemia and a moderate thrombocytopenia.

Morphological Image

Granulocytosis, with immature forms and a

basophil, in a case of chronic myeloid leu-
kaemia (CML).

Supplementary Notes
Basophilia is poorly defined because of the statistical limitations of manual differentials and analytical limita-
tions of automated haematology analysers. Nominally, basophilia can be defined as counts in excess of
0.1 x 109/L although significant basophilia is only reached above 0.25 x 109/L. Basophilia is a common
finding in chronic myeloid leukaemia (CML) and (to a lesser extent) the other myeloproliferative disorders.
Basophils in these conditions may have atypical morphological features, and an increasing basophil count
is an indicator of disease progression/acceleration. Non-malignant conditions associated with mild to
moderate basophilia include hypothyroidism, IgE-mediated hypersensitivity reactions, inflammatory disor-
ders such as rheumatoid arthritis and ulcerative colitis, and some viral infections.

Case 9: Lymphocytosis
Case Images

9a

33
 9b

Case Text
These CBC profiles correspond to relatively uncomplicated cases of lymphocytosis. Both show a modest
increase in the absolute lymphocyte count, there is no significant anaemia and the platelet counts are
either normal or slightly increased. Case 9a was a 38-year old female patient with long-standing chronic
rheumatoid arthritis, while Case 9b was a 3 year old boy with Type 1diabetes mellitus. Together with the
lymphocytosis, the WBC differentials show slightly increased monocyte counts with good population sepa-
ration from lymphocytes, normal neutrophil counts with normal-appearing population clusters in the SIZE
versus COMPLEXITY plots, and generally normal eosinophil counts.

Morphological Images

Atypical lymphocyte with clumped chroma- Atypical lymphocyte with clumped chroma-
tin and a clear blue cytoplasm. tin, opaque blue cytoplasm and nuclear
nucleolation.

Supplementary Notes
Conditions associated with a lymphocytosis (> 4.5 x 109/L) are diverse (Table 2). With regards to the ex­­
pected patterns of instrument flagging, the CD-Ruby VAR LYM flag should be viewed simply as a generic
descriptor for the presence of atypical lymphoid cells. However, to maximise the utility of this flag in the
laboratory review process it is necessary to understand the morphological variability and conditions in
which lymphoid morphological changes can be seen. At its broadest level, a VAR LYM flag indicates signifi-
cant deviation from normality without inferring either (a) the degree of morphological atypicality, (b) the type
of morphological changes that may be present (i.e. nuclear or cytoplasmic), or (c) whether the cells are
reactive or malignant in nature. This vagueness largely reflects the fact that ‘variant’ lymphoid morphology
is interpreted differently by individual observers. For these reasons, it is also reasonable to use the term
‘blast’ for lymphoid cells that may not be malignant. This is the case for B-cell maturation where antigenic
34
stimulation of ‘resting’ B-lymphocytes induces development to an immunoblast stage as part of the matu-
ration to antibody-secreting plasma cells. B-cell immunoblasts are most commonly seen in the peripheral
blood in reactive states and share many of the features (large size, nucleoli, and deep cytoplasmic
basophilia) typically associated with immature blasts. Similarly, in samples with activated suppressor T-cells
(e.g. infectious mononucleosis) a BLAST flag may also be triggered. Recognition of these interrelationships
is necessary in order to avoid a common misconception in flagging diagnostics that peripheral blood blasts
are both immature and malignant. It is also important to understand that inter-related flagging alerts such
as VAR LYM and BLAST need to be interpreted in a flexible rather than specific manner.

Table 2: Lymphocytosis and Clinical Associations

Clinical Association Additional Comments
Acute Infections Infections, such as cytomegalovirus (CMV), infectious hepatitis
and toxoplasmosis, can cause a modest lymphocytosis of up
to 15.0 x 109/L. In infectious mononucleosis and pertussis, the
lymphocyte count is often higher.
Chronic Infections Brucellosis, typhoid fever, tuberculosis and rickettsia may be
associated with a modest lymphocytosis.
Chronic Lymphoproliferative Disorders Most cases of lymphoproliferative disease, as well as a propor-
tion of lymphomas, show an increased lymphoid cell count.
These can be highly variable and comprise lymphocytes with
morphology ranging from a typical normal lymphocyte appear-
ance to highly abnormal cells with blastoid features.
Lymphoblastic Leukaemia Most (but not all) patients with acute lymphoblastic leukaemia
present with an absolute increase in the lymphoid cell count.
The morphology and counts may vary significantly between
different patients.
Other Increased lymphocyte counts may be found in conditions asso-
ciated with stress (e.g. myocardial infarction) or trauma, follow-
ing splenectomy and after vigorous exercise.

35
 Case 10: Variant Lymphocytes
Case Images

10 a

10b

36
Case Text
The two samples show relatively complex profiles with the main abnormalities being associated with the
presence of atypical lymphoid cells (VAR LYM flag). Case 10a has a reasonably normal absolute lym-
phocyte count, together with neutropenia, normocytic anaemia and moderate thrombocytopenia, but the
cluster profile (SIZE versus COMPLEXITY plot) is clearly abnormal. The lack of separation between lym-
phocytes (blue) and the events categorised as monocytes (purple) usually indicates the presence of atypi-
cal lymphocytes although small blasts could also show a similar cluster pattern. As a general rule, when
cluster patterns such as this are seen the lymphocytes tend to be underestimated and cells in the mono-
cyte fraction overestimated. The observation that the patient has pancytopenia of the main haemopoeitic
cell elements is suspicious and indicates a need for further follow up of a potential high-grade malignancy.

Case 10b shows a similar overlap of lymphocyte and monocyte populations although the size range is
wider than seen with the previous case. In further contrast, while this sample also has a mild neutropenia
the red cell and platelet results are normal. The FWBC flag indicates that there could be degenerate or
fragile cells in the sample, and the WBC flag is triggered as an operator alert that that WBC may be erro-
neous because of the FWBC. In such cases, the sample is simply reprocessed in the CBC+NOC mode.
The FWBC flag does not discriminate between reactive or malignant mononuclear cell abnormalities but it
does indicate the need for specific follow up.

Morphological Images

Atypical lymphocyte with clumped Pleomorphic atypical lymphocyte with dark

chromatin and a clear blue cytoplasm. clear cytoplasm.

37
 Case 11: Fragile Lymphocytes
Case Images

11a

11b

Case Text
When the proportions of fragile cells (usually lymphocytes) are high, then the primary optical WBC count
could well underestimate the true WBC count. Consequently, samples processed in the normal CBC
mode with a FWBC flag should be reprocessed in the CBC+NOC mode to confirm the correct WBC
count.

The nature of fragile cells in individual samples show wide variability and it is not always possible to predict
the effect on the WBC count determination. The case shown here is a good example. With the primary
CBC analysis (Case 11a), a sample from a patient with chronic lymphoproliferative disorder (probably CLL)
38
showed a WBC count of 53.4 x 109/L and a FWBC flag. This alert could have been triggered by any of the
main algorithms, including the existence of >15% fragile WBC in the RBC/Plt plot (lower right) that is seen
as a second WBC population (blue) cluster. As the FWBC flag requires sample reprocessing in the
CBC+NOC mode, this was undertaken (Case 11b). It can be seen that the WBC of 50.4 x 109/L is slightly
lower than obtained with the CBC mode and suggests that although FWBC are present they were not
excluded from the initial CBC analysis (the CBC+NOC analysis of course has little or no effect on the
measurement of any other parameter). Although this sample showed only a minimal WBC difference, it is
possible for significant discrepancies to occur with some abnormal samples. Microscopic review of this
particular case revealed the presence of many smear (Gumprecht) cells and moderate numbers of atypical
lymphocyte forms.

Morphological Images

Chronic lymphocytic leukaemia (CLL) with Chronic lymphocytic leukaemia (CLL).

predominance of small lymphocytes and
occasional ‘smear’ cells.

Case 12: Monocytosis

Case Images

12a

39
 12b

Case Text
These patient samples both show a monocytosis. With Case 12a the monocytosis of 1.8 x 109/L is seen
as part of a general non-specific increase in all WBC elements and is therefore of lesser diagnostic rele-
vance although the eosinophilia of 1.3 x 109/L is notable. The sample also shows a high haemoglobin con-
centration, macrocytic red cells and a high normal platelet count. This patient was a 1 day old female who
developed bacterial infection after initial presentation. By contrast, the second patient sample (Case 12b)
shows a monocytosis of greater magnitude (7.2 x 109/L) that is accompanied by a specific neutrophilia and
an apparent left shift (BAND flag and a reduced level of neutrophil lobulation in the GRANULARITY versus
LOBULARITY plot). The patient is also anaemic (normocytic) and thrombocytopenic. The monocyte cluster
(purple) in this sample is reasonably homogeneous and this is more in keeping with typical rather than
atypical morphology.

Morphological Images

Typical mature monocyte with convoluted Atypical mature monocyte.

nucleus, opaque cytoplasm and small cyto-
plasmic vacuoles.

Supplementary Notes
A monocytosis (> 0.8 x 109/L) is most commonly associated with chronic bacterial infections (e.g. tubercu-
losis, brucellosis and typhoid fever) and some parasitic diseases (e.g. malaria and trypanosomiasis). Less
commonly, monocytosis can be observed in bacterial endocarditis and Hodgkin’s disease. A monocytosis
(including immature forms) is also the hallmark of malignancies such as chronic myelomonocytic leukaemia
(CMML) and acute myeloid leukaemias of M4 and M5 subtype. For CMML, the monocyte count exceeds
1.0 x 109/L and the monocytosis is typically seen on a background of dysplasia.

40
Case 13: Acute Leukaemia
Case Image

Case Text
The sample shows all the features of a high-grade malignancy such as lymphoblastic lymphoma or acute
leukaemia. There is an increased WBC count, a marked neutropenia and severe thrombocytopenia, and
an anaemia that is slightly macrocytic. The graphical WBC displays essentially show a single population
cluster with moderate size variability and evidence of WBC fragility (FWBC flag). Although the predominant
population is nominally classified as lymphoid, it is possible for blast cells in the more immature (agranular)
acute myeloid leukaemias to show similar optical characteristics. It is also not uncommon for the abnormal
cells in acute leukaemias to show abnormally increased osmotic fragility (FWBC). Finally, the LURI flag indi-
cates the presence of both lower and upper range interferences in the platelet count. This is a relatively
common feature in some high-grade proliferations that have circulating cell fragments or abnormal plasma
components.

Morphological Images

Acute myeloid leukaemia (AML) showing Acute lymphoblastic leukaemia (ALL) with
predominance of immature blast cells. Note immature blasts showing nuclear clefts.
the occasional cytoplasmic Auer rods.

41
 Supplementary Notes
The haematological and morphological features of acute leukaemia are varied in relation to the cellular ori-
gin and level of maturation. The classification and differential diagnosis of acute leukaemias is extensively
reviewed in the literature and will not be covered in detail here. However, when interpreting the results of
blood count analysis there are a number of important and relatively consistent patterns. As a general rule,
acute lymphoblastic leukaemias (ALL) are mainly seen in young children while acute myeloid leukaemias
(AML) are more commonly seen in adults over the age of 40. Haematologically, most patients present with
a normocytic anaemia. Total WBC counts may be reduced, normal or increased but the differential usually
includes a variable blast cell component on a background of neutropenia. Thrombocytopenia is commonly
present, particularly in AML, and thrombocytopenia in acute promyelocytic leukaemia (AML-M3) is often
accompanied by evidence (including schistocytes) of disseminated intravascular coagulation (DIC). In
approximately 10% of patients with AML, blast cells may be absent and a modest thrombocytopenia and
a low-grade anaemia may the only haematological abnormalities. The finding of morphologically abnormal
(dysplastic) granulocytes, and a leukoerythroblastic picture with immature granulocyte forms is not uncom-
mon. Dysplastic features are particularly evident in cases of acute leukaemia that have evolved from prima-
ry myelodysplasia. Bone marrow examination is typically characterised by hypercellularity and an increase
in blast cells.

Case 14: Neutropenia

Case Image

Case Text
This patient sample shows a specific reduction in the neutrophil count (neutropenia) with all other haemo-
poietic elements being within normal limits. The graphical displays are also unremarkable. This pattern of
results is relatively unusual, as most patients with neutropenia have co-existing decreases in other cell
types.

42
Supplementary Notes
Defined as a neutrophil count below 2.0 x 109/L, reduced neutrophil counts are associated with suscepti-
bility to infection with a risk that is related to the severity of neutropenia. Patients with neutrophil counts of
< 0.5 x 109/L are at particular risk for severe and potentially life-threatening infection. Neutropenia can be
variously regarded as congenital or acquired, specific or non-specific (i.e. in conjunction with other cytope-
nias), chronic, cyclical or transient, and mild or severe.

Interpretation of neutropenia is best achieved by considering the clinical picture and patient history, togeth-
er with other haematological features. Thus for asymptomatic patients with no history, an isolated neutro-
penia can often be managed conservatively. In one study of incidentally observed adult neutropenia during
routine CBC analysis, 34% of cases were categorised as chronic idiopathic neutropenia. Of the remainder,
the neutropenia was variously attributed to exposure to chemical agents, infections, autoimmune disor-
ders, haematological diseases, thyroid disorders, ethnic neutropenia, cyclic neutropenia and iron deficien-
cy. By comparison, most specific neutropenias seen in childhood are associated with viral infections.

Case 15: Pancytopenia

Case Images

15a

43
 15b

Case Text
These sample results both show pancytopenia but the haematological pictures are markedly different.
Case 15a is a typical profile for a patient receiving myelosuppressive therapy. There is a severe reduction in
the neutrophil count, and the patient has moderate normocytic anaemia and thrombocytopenia. The SIZE
versus COMPLEXITY plot shows a solitary population of lymphocytes, with only occasional monocytes,
and the LOBULARITY versus GRANULARITY plot confirms the almost complete absence of depolarising
eosinophils.

With the second case (15b), the neutropenia is only mild but there is a marked anaemia (with RBC macro-
cytosis) and a moderate thrombocytopenia.

Supplementary Notes
Pancytopenia is any condition where there is a reduction in the production of all three main haemopoietic
elements. Classifications of pancytopenia are not well established but there is general acceptance that
they may be categorised as (a) aplastic anaemia, (b) pancytopenia secondary to drugs or radiation,
(c) pancytopenia secondary to marrow infiltration, (d) transient pancytopenia secondary to viral infection,
(e) myelodysplastic syndromes, (f) hypersplenism and (g) megaloblastic anaemia. In children, pancytopenia
is only rarely caused by infectious agents and is more often due to leukaemia, aplastic anaemia, or neuro­
blastoma.

The most common cause for pancytopenia is drug-mediated myelosuppression associated with modern
therapeutic regimens in cancer treatment. These typically have a non-specific effect on normal marrow cell
production, with the kinetics of the pancytopenia and its subsequent recovery being reasonably predicta-
ble. Pancytopenia following marrow transplantation typically lasts from 2 – 4 weeks, with successful
engraftment being indicated by recovery of normal cell production.

Aplastic anaemia represents an important type of pancytopenia that describes cases where recovery is
neither spontaneous nor predictable. Aplastic anaemia typically results from physical or chemical damage
to the marrow, and some cases seem to be due to autoimmune attacks. While the cause in most patients
is unknown, some have been shown to result from exposure to chemicals such as benzene, insecticides,
antibiotics, and anti-inflammatory agents. Viral infections have also been implicated in the pathogenesis of
aplastic anaemia, and the 0.2% of Hepatitis (mainly type C) infections that develop aplastic anaemia as a
secondary complication represents about 5% of all cases.
44
Case 16: NWBC Flagging Alert
Case Image

Case Text
This sample has a reasonably normal WBC count and differential but the results are accompanied by a
NWBC (non-white blood cell) alert. The SIZE versus COMPLEXITY plot is also abnormal and shows a clear
population of events adjacent to the lymphocyte population and below the size threshold.

The NWBC flag can be triggered by a number of different factors including low numbers of nucleated red
blood cells (NRBC), unlysed RBC (RRBC), platelet clumps, or giant platelets. Examination of graphical
plots, together with consideration of the other haematological parameters can often provide clues as to the
reason for an NWBC flag. In this example, the red cell parameters are normal but there is an apparent
moderate thrombocytopenia of 70 x 109/L, a URI flag and a suppressed MPV. All of these are suggestive
of platelet clumps (or giant platelets) and this was confirmed by reference microscopy. The low platelet
count was consequently interpreted as pseudothrombocytopenia.

Morphological Images

Mature (haemoglobised) nucleated red Platelet clumps.

blood cell (erythroblast).

45
 Case 17: Resistant Red Blood Cells (RRBC)
Case Images

17a

17b

Case Text
The upper report (Case 17a) shows a sample that was processed with the CD-Ruby CBC mode. The
WBC count (12.6 x 109/L) and the absolute lymphocyte count (6.0 x 109/L) both appear to be increased
but these results are accompanied by WBC and NRBC/RRBC flags, and data-invalidating * alerts. Exami-
nation of the SIZE versus COMPLEXITY plot shows a distinct population of events (pink) below the size
threshold. The patient also has anaemia, with slight microcytosis, but the platelet count is normal. The
RRBC flag indicates the need for sample re-processing with the CBC + RRBC mode.

46
When reanalysed with the CBC + RRBC mode, a lower WBC count was obtained (the WOC flag indicates
that it is the optical count that is being shown) but there was still a persistence of residual events suggest-
ing the presence of RRBC. Consequently, the NRBC/RRBC flag combination remains and the WBC count
and differential data retains the * alerts. However, even though these alerts indicate the need to independ-
ent validation, visual review of the SIZE versus COMPLEXITY plot suggests that these results are unlikely to
be affected to any great extent by the residual RRBC. Moreover, reference microscopy confirmed the pres-
ence of 8 NRBC/100 WBC.

Morphological Images

Target cells. Sickle cells.

Supplementary Notes
Instrument WBC counts and differentials are invariably obtained after sample dilutions are exposed to lys-
ing agents. These are designed to remove RBC while at the same time preserving WBC morphology.
Resistant RBC are red cells that have a relative resistance to lysing agents and remain as a ‘contaminant’
in the WBC analysis. If a sample contains a significant number of lyse-resistant red cells, the CD-Ruby
analytical algorithms will trigger a RRBC flag and place an * next to the potentially affected parameters (i.e.
the WBC count and differential). To deal with such samples, the CD-Ruby provides a supplementary oper-
ational mode that prolongs the exposure time of red cells to the lysing agent. Samples with a RRBC flag
should therefore be reanalysed with this mode.

The occurrence of lyse-resistance is most often associated with target cells and any condition in which
they are present can therefore occasionally present analytical difficulties. These include liver disease, thal­
assaemia and some haemoglobinopathies. RRBC are also sometimes seen in samples from neonates
(cord bloods). Early studies suggested that cord blood red cells had a generalised property of increased
osmotic-resistance, similar in nature to that associated with target cells, but more recent work indicates
that this resistance may relate only to a subpopulation of red cells. With the exception of occasional
homozygotic haemoglobinopathies, in which the red cells can be hyper-resistant, the CD-Ruby RRBC
mode is generally able to overcome the problem of lyse-resistance.

47
 Case 18: Thrombocytopenia
Case Images

18a

18b

Case Text
These case examples show severe thrombocytopenia in two quite different haematological situations. The
first (18a) is from a young child with idiopathic thrombocytopenic purpura (ITP). Apart from the markedly
reduced platelet count there are no other abnormalities of significance. A suppressed MPV result in ITP
cases is not unexpected as the platelet size distribution is often abnormal. By contrast, the thrombocyto-
penia in the second case (18b) is seen on a background of general pancytopenia.

48
Supplementary Notes
Thrombocytopenia is defined as a platelet count of < 150 x 109/L. In stable patients, a susceptibility to
bleeding is not usual until the platelet count falls below 50 x 109/L. Below 20 x 109/L, there is a major risk
of bleeding and patients may need platelet transfusion support. Apart from some rare hereditary disorders
(Fanconi’s syndrome, Wiskott-Aldrich syndrome, May-Hegglin’s anomaly and Bernard-Soulier syndrome),
thrombocytopenia is mainly due to either (a) decreased marrow production, (b) increased destruction or (c)
increased sequestration (Table 3). Conditions associated with increased destruction can be further subdi-
vided into immune and non-immune. Samples containing platelet clumps or satellitosis are often charac-
terised by pseudothrombocytopenia and it is important that this possibility is fully excluded when investi-
gating any patient with apparent thrombocytopenia.

Table 3: Causes of thrombocytopenia.

Decreased Production Increased Destruction Increased Sequestration
• Aplastic anaemia Non-immune Splenomegaly and hypersplenism
• Non-specific myelosuppression • Microangiopathies (thrombocytopenia may be
by chemotherapy/radiotherapy accompanied by neutropenia).
• Specific suppression of mega­ Immune
karyopoiesis by drugs or virus • Idiopathic thrombocytopenic
infection (especially HIV and purpura (ITP)
parvovirus) • In association with disorders
• Bone marrow infiltration by such as SLE, CLL and NHL
tumour cells (leukaemia or • Drug-associated (especially
carcinoma) or fibrosis (myelo­ heparin, gold, penicillin and
fibrosis) quinine); drug-platelet complex
• Ineffective thrombopoiesis (vita- autoantibodies
min B12 or folate deficiency, • Post-transfusion purpura
paroxysmal nocturnal haemo- (resulting from transfusion of
globinuria) plasma isoantibody)
• Neonatal purpura

Determination of accurate platelet counts in severe thrombocytopenia can be problematic. Impedance

methods in particular are prone to erroneous overestimation in patient groups where plasma or cellular
interferences are commonly seen (e.g. leukaemia and myeloma). Optical methods may also be affected by
interferences and it is therefore recommended that independent validation of platelet counts is undertaken
whenever a patient shows clinical evidence of increased bleeding or the instrument platelet count is at a
level close to or below the platelet transfusion trigger point. In such circumstances, the method of choice
for validation is immunoplatelet analysis (CELL-DYN Sapphire or flow cytometry).

49
 Case 19: Thrombocytosis
Case Images

19a

19b

Case Text
The first patient example (19a) shows a markedly increased platelet count of 1009 x 109/L together with a
mild neutrophilia and moderate normocytic anaemia. The platelet size distribution appears to be narrow
(lower right plot) and there is good separation between platelets and RBC. The normal MPV further sug-
gests that there are relatively few large platelet forms. While the second patient (19b) has a similarly high
platelet count of 1044 x 109/L, there are a number of additional features of note. For example, there is a
moderate neutrophilia, and the anaemia is associated with moderate microcytosis. It is the platelet/RBC

50
optical plot (lower right) that is, however, of particular interest. Because of the RBC microcytosis it can be
seen that there is relatively little separation between the microcytes and platelets. In such situations, it is
important for the review process to consider whether or not the separation is acceptable for the purposes
of data reporting. In this case, the separation appears adequate but there may be some occasions when it
is not. This overlap between platelets and RBC is common to all analysers irrespective of whether the
methods are impedance or optical. Resolution of suspect results can only be achieved by alternative pro-
cedures such as the immunoplatelet count.

Morphological Image

Platelet thrombocytosis with occasional

‘giant’ forms.

Supplementary Notes
A thrombocytosis exceeding 450 x 109/L may be primary or secondary in nature. Cases of primary throm-
bocytosis are clonal and correspond to the myeloproliferative disorders in general and essential thrombo-
cythaemia (ET) in particular. In these primary conditions, a thrombocytosis is typically persistent in nature
and it is often accompanied by other haematological abnormalities. These include giant and morphologi-
cally atypical platelet forms. This is in contrast to secondary thrombocytosis where the increased platelet
count is usually modest (<1000 x 109/L) and the platelets tend to morphologically normal even though
some may be large.

Secondary thrombocytosis is associated with various clinical conditions including haemorrhage, trauma,
surgery (especially splenectomy), iron deficiency anaemia, inflammatory disorders and infections, and
malignancy. Note also that in children younger than 12 months of age, the median platelet count is higher
(425 x 109/L) than adults, and platelet counts up to 750 x 109/L are not uncommon. For haematologically
normal children of 1– 3 years and 3 – 7 years, the observed frequencies of platelet counts exceeding
500 x 109/L have been reported as 12% and 6%, respectively.

51
 Case 20: Malaria
Case Image

Case Text
These four representative images of GRANULARITY versus LOBULARITY show the most important
changes associated with abnormal CELL-DYN depolarisation patterns. The first of these (upper left) is a
normal profile in which the area circumscribed by the continuous yellow oval boundary represents the
position normally associated with light-depolarising eosinophils (green), while all other leukocytes do not
depolarise light. Note in particular that the area within the red boundary does not contain any leukocyte
events. The second plot (upper right) shows the presence of occasional depolarising purple events (mono-
cytes) located within the red boundary, while the third plot (lower left) is characterised by a higher frequency
of abnormal depolarising purple events together with a small number of depolarising green events that fall
outside of the normal eosinophil region. The final plot (lower right) shows the same admixture of atypical
depolarising purple and green events but differs in that the abnormal green events are particularly prominent.

Morphological Images

Intracellular parasites (ring forms) from a Intracellular malarial pigment Intracellular malarial pigment
patient with P. falciparum malaria. (haemozoin) in a mature neu- (haemozoin) in a mature mono-
trophil. cyte.

52
Supplementary Notes
Between 300 and 500 million cases of malaria are reported each year with more than 90% of these occur-
ring in tropical Africa. Infection with Plasmodium falciparum is associated with the highest levels of blood
parasitaemia, clinical complications and mortality. Three other plasmodium species (vivax, malariae and
ovale) also cause malaria in humans but these rarely have fatal outcomes.

In countries where there are comprehensive community healthcare systems, patients with fever and
malaise usually consult a clinician within a short time of developing symptoms. While clinicians may not ini-
tially consider malaria as part of the differential diagnosis of pyrexia (especially if the clinical or travel history
is inadequate), it is nevertheless likely that ‘screening’ tests such as the CBC will be undertaken. However,
the routine use of thin films in haematology for morphological examination and the limited experience
of most morphologists in non-endemic areas means that the likelihood of a ‘chance’ finding of malaria is
small. This is a particular concern because malaria mortality in non-endemic parts of the world, such as
Europe and the US where malaria is a relatively rare disease, is almost always due to a missed or late diag-
nosis. This critical point is well illustrated by one study showing that a diagnosis of malaria was missed in
59% of returning travelers with infection and that the average delay before correct diagnosis and initiation
of treatment was 7.6 days. This delay in diagnosis was almost certainly a major factor in the development
of severe malaria by a significant minority of these patients.

Haematological changes in patients with malaria in endemic areas are usually non-specific and complicat-
ed by other disturbances related to abnormal haemoglobins, poor nutrition, other parasitic infections and
chronic infections. Patients from non-endemic areas are less likely to have these complications but there is
nevertheless no distinctive haematological pattern that can alert a clinician to the possible existence of
malaria infection. Thrombocytopenia has been regarded as an indicative feature but this is not always
seen.

Following the introduction of CELL-DYN haematology analysers in the 1980s, a number of reports were
received about atypical depolarising events in the automated WBC differential plots of patients with malar-
ia. These were associated in particular with the monocyte fraction although other abnormal depolarising
populations were also observed. In confirming this, a subsequent South African study further suggested
that by providing an indication of malaria as part of the routine CBC analysis, MAPSS depolarisation analy-
sis could be of particular value if malaria was not clinically suspected.

The explanation for the abnormal depolarising WBC events is interesting. During the intra-erythrocytic
stage, a malaria parasite digests haemoglobin and detoxifies the heme by converting it to haemozoin.
Haemoglobin does not depolarise light but haemozoin does. When the red cells rupture, free haemozoin is
released and this is subsequently ingested by monocytes. Monocytes are normally non-depolarising but
when they contain ingested haemozoin they behave like ‘pseudo-eosinophils’. Because of the way in
which CELL-DYN systems determine the WBC differential, haemozoin-positive mononuclear cells remain
colour-coded as blue or purple. As these are clearly abnormal, the CD-Ruby is able to provide an operator
alert in the form of an ATYPDEP flag. In some cases, there may also be haemozoin ingestion by neu-
trophils. As these events are depolarising cells with cytoplasmic granulation, the MAPSS analysis colour-
codes them green (for eosinophils) but their location on the NEU-EOS plot is different. Importantly, it should
be noted that while the numbers of abnormal depolarising WBC events are not directly related to the
degree of red cell parasitaemia there is evidence that the number of haemozoin-containing WBCs may be
an independent prognostic indicator.

53
 Case 21: Reticulocyte Analysis
Case Images

54
Case Text
The CD-Ruby reticulocyte method is based on supravital dye staining with New Methylene Blue. After initial
off-line exposure to the dye, the blood is processed using a dedicated reticulocyte ‘mode’ and RBC are
analysed in the optical flow cell by three angles of laser light (0°, 10° and 90°). The two examples shown
here correspond to samples with a range of different reticulocyte counts. By means of three-dimensional
optical analysis and subsequent automated algorithmic gating procedures, the various cell populations are
colour-coded in the 90°/0° and 90°/10° scatter plots as follows:
• Pink-red events – Red blood cells (non-reticulocytes)
• Dark blue events – Mature reticulocyte fraction
• Light blue events – Immature reticulocyte fraction (IRF)
• Black events – White blood cells
• Yellow/Orange events – Platelets (at the lower limits of the 0° axis and 10° axis)

The first example shown here (21a), from a patient with a normocytic anaemia of 9.3 g/dL, has a normal
reticulocyte percentage of 1.2% and an absolute count of 66 x 109/L (derived through automatic reference
to the red cell count). The other example (21b) has a higher reticulocyte count of 159 x 109/L (4.0%) and
was from a patient with a more severe normocytic anaemia of 7.6 g/dL.

Morphological Image

Red blood cell polychromasia. Polychromatic

red cells typically correspond to reticulocytes
seen with supravital staining methods.

Supplementary Notes
Reticulocytes require 44 hours within the bone marrow and a further 29 hours within the peripheral blood
to mature. Reticulocytes synthesise up to one third of the final erythrocyte haemoglobin content although
much of this probably takes place in the bone marrow before release into the peripheral blood. During the
transition from reticulocyte to mature red cell, the mean cell diameter and MCV both decrease by about
30% and the haemoglobin concentration increases by almost 20%. Determinations of reticulocyte counts
are therefore important for understanding the kinetic status of red cell production and for the monitoring of
treatments that suppress or enhance erythropoiesis.

The normal reticulocyte reference range for adults is between 0.5% and 1.6%, with absolute counts in the
order of 25 to 75 x 109/L. In a 3-month foetus, 90% of the red cells are reticulocytes and this falls to
between 15 and 30% at the 6th month of foetal life. Cord blood samples from full-term infants also show
increased (compared to adult) reticulocyte numbers, with a reported mean of 3.11± 0.75% (SD) and an
absolute count of 137.3 ± 33.3 x 109/L. These fall to adult levels within 5 days. Premature infants tend to
have higher reticulocyte counts than full-term infants.

In interpreting reticulocyte information, it is necessary to be aware of analytical interferences. Howell-Jolly

bodies (DNA) and basophilic stippling (RNA) are potentially important sources of error because they can
theoretically be included in the reticulocyte count. In practice, however, many investigators have concluded
that Howell-Jolly bodies rarely interfere with reticulocyte estimates because their relative number in most
patients is low. Interferences affect all automated methods to some extent and so it is important that when
samples show unexpectedly high reticulocyte counts that they should be validated by alternative procedures.

55
Abbott GmbH & Co. KG
Abbott Diagnostics Europe
Max-Planck-Ring 2
65205 Wiesbaden
Tel. (+49) 61 22 58 0
Fax (+49) 61 22 58 12 44
www.abbottdiagnostics.com

Axxxxx/uk CELL-DYN Ruby Casebook 10/07/x