Title: Monocytopenia as a diagnostic clue to pediatric B-lymphoblastic leukemia with

rare circulating blasts

Authors: Sunita I. Park, MD and Beverly B. Rogers, MD

Children's Healthcare of Atlanta

1405 Clifton Road, NE

Egleston Children's Hospital, Department of Pathology

First floor, Tower one

Atlanta, GA 30322

Corresponding author: Sunita Park, MD. Address: See above. Phone: 404-785-6499

Fax: 404-785-1370, email: sunita.park@choa.org

Running head: Monocytopenia in occult leukemia

Abstract:

Background: B-lymphoblastic leukemia/lymphoma (B-LL) is the most common childhood

cancer. Occasionally, circulating blasts in the peripheral blood are rare (≤1%) and may

be missed, even when flow cytometric immunophenotyping is performed, leading to a

false negative report.

Methods: The records from all patients with a new diagnosis of B-LL at our institution

were reviewed from Jan 2009-Dec 2011. Of 130 cases with peripheral blood flow

cytometry, 15 had a blast count of ≤1%, with 14 having electronic files for gating

monocytes. The percentage of monocytes by flow cytometry and absolute monocyte

counts (AMCs) were compared with peripheral blood samples that were negative by

flow cytometry, sent due to at least one lineage cytopenia (n=39).

Results: The monocytes from the patients with leukemia averaged 0.8%, and were

statistically lower than the negative controls, which averaged 7.1% (p<0.001). 11 of the

14 (79%) patients with leukemia had monocytes <1%, compared to only 3 (8%) of the

negative controls. The AMCs were also significantly lower (p<0.001), with 93% of the

leukemia group having an AMC of <100 cells/µL, compared to only 28% of the negative

controls.

Conclusions: In patients presenting with cytopenias, assessment of percentage

monocytes may be an important diagnostic clue in determining the presence of occult

leukemia. If flow cytometry is performed, acquisition of more than the standard 10,000

events is necessary to adequately assess for leukemia. If monocytes are <1% by flow

cytometry in the setting of cytopenias, bone marrow examination is recommended, even

with negative peripheral blood flow cytometry.

Key terms: B-lymphoblastic leukemia, flow cytometry, monocytopenia, pediatric

Introduction:

B-lymphoblastic leukemia/lymphoma (B-LL) is the most common childhood tumor, with

an estimated 3,000 new cases annually in the US.1 The majority of patients present with

cytopenias, and blasts are easily seen on peripheral blood smear. Flow cytometric

immunophenotyping of the peripheral blood is then performed to characterize the

immunophenotype of the blasts, which often expresses dimmer CD45 than mature B

cells, and may aberrantly express CD10, CD34, CD58, and TDT, with lack of surface

kappa or lambda staining.1 Given the expected absence of normal B cell progenitors

(hematogones) in the peripheral blood, which may have a similar phenotype, the

leukemic blasts are typically easy to identify by flow cytometry.

Rarely, patients present with a very low circulating blast burden, which may be ≤1%,

making the diagnosis challenging, even with peripheral blood flow cytometry. In a

paper describing rare event analysis, Allan and Keeney show that given a coefficient of

variation of 20% in detecting a positive population, 25 events will define an abnormal

cluster of cells. If a sensitivity of 0.01% is desired, 2.5 x 105 cells need to be acquired2,

which is much greater than the standard 10,000 events/tube3 used in many flow labs.

This level of sensitivity may be challenging to achieve in hypocellular samples, those

with low specimen volume, or in samples with debris and/or platelets, which can fall in

the CD45 dim region and obscure a minor abnormal population. Failing to detect the B-

lymphoblasts will lead to a false negative flow report, which may lead to a false sense of

security, and/or delay in examining the bone marrow.

Monocytes are myeloid-derived white blood cells that aid in antigen presentation. They

circulate briefly in the peripheral blood, and then migrate to the tissues where they

mature into various cells of the monocyte/histiocyte/immune accessory cell system.

Monocytes are increased in physiologic conditions, such as in the neonatal period and

in marrow recovery from agranulocytosis. They are also increased in reactive

conditions, such as infection, autoimmune diseases, and certain neoplasms.4

Decreased monocytes is uncommon, and the differential diagnosis includes marrow

failure states, such as aplastic anemia, and glucocorticoid administration, hemodialysis,

sepsis, and certain hematologic malignancies, such as hairy cell leukemia.5

This study examines the role of using decreased monocytes, as determined either by

flow cytometry or absolute monocyte count, as a diagnostic clue to aid in the diagnosis

of B-LL with very low circulating blasts.

Methods:

Study Groups:

With appropriate IRB approval, the records from all patients with a new diagnosis of B-

lymphoblastic leukemia at Children's Healthcare of Atlanta (CHOA) were retrospectively

reviewed over a 3 year period (Jan 2009-Dec 2011). Of 171 cases, 130 had peripheral

blood flow cytometry performed at our institution.

B-lymphoblastic leukemia with ≤1% blasts: Fifteen of the above 130 patients had a

blast count of ranging from 0.08 – 1.0% by flow cytometry. In all cases, a bone marrow

biopsy confirmed the diagnosis of B-LL, by both morphology and flow cytometry. The

percentage of monocytes by flow cytometry and absolute monocyte count were

compared to the negative control group. Fourteen of the 15 cases had electronic files

that allowed for re-gating of the monocytes by CD33 and CD64 (see below).

Controls: This group is comprised of all patients seen in 2011 whose peripheral blood

was sent for flow cytometry because of least one lineage cytopenia stated in the clinical

history. Of 43 total cases, 39 were included in the study, as they had electronic files

that allowed gating of the monocytes. Patients with a prior diagnosis of a hematologic

malignancy, solid tumor, and Down syndrome were excluded. Clinical follow up, and

bone marrow examination, when performed, were negative for leukemia.

B-lymphoblastic leukemia with >25% blasts: Flow cytometry reports were reviewed

from 52 patients with "typical" B-LL who presented with blast counts >25% in the

peripheral blood. Percent monocytes from the flow cytometry report was recorded,

which was generated by gating CD14+ cells with low to intermediate side scatter

properties.

Comparison with T-lymphoblastic leukemia/lymphoma (T-LL): Records from patients

with T-LL were also reviewed over the same time period. Twenty-five new diagnoses of

T-LL were reviewed, 22 of which had peripheral blood flow cytometry at CHOA.

However, none of these patients presented with a peripheral blood blast count of ≤1%.

Percent monocytes from the flow cytometry report was recorded, similar to above.

Multiparameter flow cytometry: Four color flow cytometric immunophenotyping was

performed with a broad panel of antibodies, evaluating for B-lymphoid, T-lymphoid, and

myeloid disease. The pertinent antibodies used for diagnosis of B-lymphoblastic

leukemia and for monocytes are detailed in table 3.

Peripheral blood samples collected in EDTA were adjusted to a cell concentration of

between 5-10 x 103 cells/µL. The peripheral blood sample was then washed three times

to remove plasma and platelets, and resuspended in cell wash (PBS, Fetal bovine

serum, sodium azide). Fifty microliters of this sample was added to each tube

containing the specific cocktail of antibodies, and the cells were incubated in the dark

for 15 minutes. Ammonium chloride was then added to lyse the erythrocytes, and the

tubes were incubated in the dark again for 15 minutes. The tubes were then

centrifuged, decanted and washed twice in cell wash, and 500 microliters of 1% reagent

grade formalin added to fix the cells. Acquisition of 10,000 - 200,000 cells/tube was

performed on a BD Canto II flow cytometer, and analyzed using Diva software. For

each antibody, negative staining levels were set by the use of an isotype-matched

control.

Gating strategy: To assess percentage monocytes, a gate was placed on FSC vs SSC

to include all mononuclear cells and a portion of the granulocytes. The monocytes were

then gated using a combination of CD64 vs CD33, which allows separation from the

normal granulocytes, which express dim CD64 and CD33 (see Figure 1). The percent

monocytes were obtained using all events, to include all WBCs, as the denominator.

Complete Blood Count with Differential (CBCD): The CBCD was performed on a

Siemen's Advia 120 or 2120 instrument (Malvern, PA). The absolute monocyte count

was calculated by multiplying the WBC by the percentage of monocytes, and expressed

as cells/microliter. In most cases, for both the leukemic and negative control groups, the

monocyte % was generated by manual differential.

Statistical analysis: The Wilcoxon test was used to compare percent monocytes and

AMCs from the leukemic group to the negative control group. Significance of the

reported p values was defined as p<0.05.

Results:

Study Group: Clinical characteristics of patients presenting with ≤1% circulating blasts

are detailed in Table 1, and include 9 girls and 6 boys, ranging from 10 months-15 years

of age. No recurring cytogenetic abnormalities were noted. The clinical characteristics,

types of cytopenias, and final diagnoses of the negative control group are detailed in

Table 2. There is no statistically significant difference in the ages (P=0.663, Wilcoxon

test) or sexes (P=0.94, Chi Square test) between the leukemia group and the negative

controls.

Flow cytometry: The monocytes from the patients with B-LL with ≤1% blasts averaged

0.8%, and were statistically lower than the control group, which averaged 7.1%

(Wilcoxon test (p<0.001), Figure 2a). Eleven of the 14 (79%) patients with B-LL with

≤1% had monocyte counts that were less than 1%, compared to only 3 (8%) of the

control group. Upon follow up, each of these 3 control patients with monocytes <1% had

a marrow production defect, as 2 had aplastic anemia, and one had severe vitamin B12

deficiency.

Peripheral blood flow cytometry showed monocytes <1% in 31 of 52 (60%) of patients

with "typical" newly diagnosed B-LL, who had >25% circulating blasts. Fourteen (27%)

had monocytes of 1%, and 6 (12%) showed monocytes of 2%. One case (2%) had

monocytes >3%. Similarly, patients newly diagnosed with T-LL showed monocytes of

<1% in 8 of 22 (36%), =1% in 7 (32%), 2% in 3 (14%), and ≥3% in 4 (18%).

Absolute monocyte count: The absolute monocyte count (AMC) was calculated for each

patient in the leukemia and control group. The AMC from the patients with leukemia

was statistically lower than the control group, (Wilcoxon test (p<0.001), Figure 2b). The

AMC was <100 cells/µl in 14 of 15 (93%) of the leukemia group, whereas only 11 of 39

(28%) of the control patients had AMCs at this level.

Discussion:

Peripheral blood flow cytometry is performed routinely at our institution, in cases ranging

from unexplained pancytopenia to overt leukemia. While this practice varies from

institution to institution, we have found that establishing the type of leukemia prior to

performing the marrow allows for several advantages. These include obtaining proper

COG consent, drawing appropriate study tubes, ordering appropriate cytogenetics and

FISH panels, and placement of the correct port for therapy at the time of BMA. This

obviates a second bone marrow biopsy, or repeat sedation for port placement. It is

therefore essential that the Pathologist interpreting the flow cytometry be thorough in

the evaluation, as to minimize false negatives.

Monocytes should comprise at least 4% of total white blood cells, with an absolute

monocyte count (AMC) of at least 180 cells/μl (0.18 cells x 109/L) in patients older than

2 months. 6 Circulating monocytes provide a window into bone marrow production, as

they are often increased first in marrow recovery, with neutrophils following. 4

Monocytes only circulate for 12-24 hours before entering tissue7, compared to the

relatively longer 5.4 day circulation of neutrophils8, making monocytes a more sensitive

indicator of marrow production.

Although both neutrophils and monocytes are typically decreased in leukemia,

decreased monocytes are more informative for occult leukemia, as many of the non-

neoplastic conditions that lead to neutropenia in children, such as infections, cyclic

neutropenia, and autoimmune neutropenia, often cause a relative increase in circulating

monocytes.4 Decreased monocytes are not often seen in children, and most often

indicate a bone marrow failure syndrome, such as aplastic anemia. Monocytopenia

may also be seen in the setting of glucocorticoid administration, sepsis, or

hemodialysis.5 To assess whether these factors were involved in the monocytopenia

observed in the leukemic patients, medical records were reviewed at presentation.

Patient 6 presented in sepsis, and later grew Pseudomonas in her blood culture. This

patient died during induction chemotherapy. All the other patients in the occult leukemia

group were previously well, with no signs of sepsis, no history of steroids or

hemodialysis.

It is not only the cases with low circulating blasts that had monocytopenia. In fact, the

typical presentation of "typical" B-LL, with circulating blasts >25%, also showed a

decrease in monocytes. A review of 52 peripheral blood flow cytometry reports

revealed that 98% of newly diagnosed B-LL with circulating blasts of >25% shows a

monocytopenia of ≤ 2%, well below the 4% lower limit of the normal range as defined by

the American Association of Clinical Chemistry (AACC). 6 The reason circulating

monocytes are decreased in B-LL is unclear. One hypothesis is that the increased

blasts in the marrow space are myelophthisic, leading to an overall decrease in

hematopoiesis by replacing the bone marrow, and excluding the normal marrow

elements. However, review of the bone marrow aspirates in the 15 patients in our study

with low (≤1%) circulating blasts revealed the average blast count to be 60%, and half of

the cases had easily identifiable background hematopoiesis. In keeping with this

finding, the CBCs of these patients occasionally only showed mildly decreased counts

in only 1-2 lineages. Patient 5 with B-LL even had a normal CBC with an ANC of 7,000

cells/μl; however, the monocytes were still mildly decreased at 3%.

Other possibilities for monocytopenia associated with B-LL include dysregulation of

endogenous GM-CSF or increased endogenous glucocorticoids, both of which may lead

to decreased monocyte production. The mechanism may be similar to that seen in

hairy cell leukemia, which is also poorly understood, but may involve decreased levels

of stimulatory cytokines or secretion of inhibitory factors by the malignant hairy cells. 9

Clearly, more study is necessary to elucidate the cause of this relationship.

To assess whether a similar finding is observed with T-lymphoblastic

leukemia/lymphoma (T-LL), all new diagnosis T-LL cases were reviewed over the same

time 3 year period. Twenty five new cases were reviewed, of which 22 had peripheral

blood flow cytometry at CHOA. However, none of these cases had circulating blasts

≤1%. By flow report, monocytes were ≤ 2%, in 82% of newly diagnosed T-LL, which

suggests a similar biological mechanism to the monocyte suppression seen in B-LL.In

over 85% of newly diagnosed B-LL, establishing the diagnosis on peripheral blood flow

cytometry is straightforward, with circulating blasts >1%. However, occasionally

circulating blasts are <1%, and even <0.1%, making the diagnosis challenging, even

with flow cytometry. Decreased monocytes provide a clue to a defect in marrow

production, and may be useful in diagnosing occult leukemia. The results of our study

suggest that if monocytes are <1% on peripheral blood flow cytometry in the setting of

cytopenias or clinical suspicion of leukemia, it is optimal to acquire as many cells as

possible in the B cell tubes to look carefully for rare blasts. Practically, we try to acquire

at least 100,000 cells in our lab before calling the result negative.

Even if the flow cytometry is still negative, a bone marrow examination should be

considered in these cases to evaluate for a marrow production defect. Peripheral blood

flow cytometry may be bypassed if the AMC is <100 cells/μl with high clinical suspicion

of leukemia, and consideration given to proceed with bone marrow examination.

Acknowledgements: The authors would like to sincerely thank Traci Leong for her

expertise in the statistical analysis of our data.

References: 1. Swerdlow SH, Campo E, Harris NL, et al. WHO classification of Tumours of

Haematopoietic and Lymphoid Tissues, 4th ed. Lyon: International Agency for

Research on Cancer, 2008

2. Allan, A and Keeney, M. Circulating Tumor Cell Analysis: Technical and Statistical

Considerations for Application to the Clinic. Journal of Oncology, Volume 2010, Article

ID 426218

.3. Stetler-Stevenson M, Ahmad E, Barnett D, et al. Clinical Flow Cytometric Analysis of

Neoplastic Hematolymphoid Cells; Approved Guideline. Clinical and Laboratory

Standards Institute (CLSI) document H43-A2, 2nd ed. 2007

4.. Foucar K. Monocytosis. In: Kjeldsberg CR, ed, Practical Diagnosis of Hematologic

Disorders, Volume 1. 4th ed. Singapore, American Society for Clinical Pathology, 2006;

219-226

5. Reichard, K. Non-neoplastic granulocytic and monocytic disorders, excluding

neutropenia. In: Foucar K, Reichard K, Czuchlewski D. Bone Marrow Pathology. 3rd

ed. Singapore, American Society for Clinical Pathology, 2010;181-205.

6. Soldin, S, Wong, EC, Brugnara C, Soldin, O. Pediatric reference intervals, 7th ed,

Washington DC, American Association of Clinical Chemistry Press, 2011

7. Glassy, E ed. Color Atlas of Hematology; An Illustrated Field Guide Based on

Proficiency Testing. Northfield, Il. College of American Pathologists, 1998

8. Pillay J, den Braber I, Vrisekoop N, Kwast LM, de Boer RJ, Borghans JA, Tesselaar

K, Koenderman L. In vivo labeling with 2H2O reveals a human neutrophil lifespan of 5.4

days. Blood. 2010; 4:625-7.

9. Burthem J, Cawley JC. Hairy Cell Leukaemia. London: Springer-Verlag, 1996

Table 1 Clinical characteristics of the B-lymphoblastic leukemia group with ≤1% circulating blasts by flow cytometry Patient Age (yrs) Sex Cytopenias % Monos AMC %Blasts Hgb Plt ANC

1 9 F 7.1 90 300 0.4 31 0.8

2 10 mos. F 8.6 23 1150 1.3 46 1

3 15 F 7.0 108 240 0.2 0 0.4

4 3 M 2.6 23 880 0.2 0 0.5

5 8 M Nl Nl Nl 3.5 361 0.5

6 10 F 7.1 87 360 0.5 10 0.3

7 5 F 10.3 131 Nl 0.7 0 0.7

8 1 M 2.6 37 0 0.5 63 0.3

9 2 F 3.4 32 0 0.8 47 0.2

10 7 M 8.5 Nl 490 0.9 87 0.8

11 9 M Nl 60 0 N/A 12 0.7

12 5 F 9.5 Nl 750 0.1 0 0.1

13 8 M 9.3 93 250 1.0 33 0.7

14 2 F 4.9 <10 20 0.7 21 0.9

15 12 F 7.0 Nl 30 0.6 0 0.08

Hemoglobin (Hb) expressed in g/dL, platelets (Plts) x 103/µL, absolute monocyte counts (AMC) and absolute neutrophil count (ANC) in cells/µL. Nl=value within the normal range. N/A= not able to be performed.

Table 2 Demographics and final diagnoses of negative control group, which had negative peripheral blood flow cytometry Patient Age (yrs) Sex Cytopenias % Monos AMC Diagnosis Hgb Plt ANC

1 3 M 10.5 <10 Nl 4.9 205 ITP

2 1 F 6.0 Nl 580 3.3 82 possible TEC

3 13 M Nl <10 40 33.9 383 ITP, possible virus

4 15 F 5.5 17 Nl 7.0 1135 TTP

5 9 months M Nl 12 910 5.9 588 ITP

6 2 F 5.1 11 270 0.5 0 SAA

7 6 months M Nl Nl 0 13.8 806 Autoimmune neutropenia with anti-granulocyte antibodies

8 14 M 7.3 Nl Nl 1.7 32 HIV, disseminated MAI

9 3 F 7.9 <10 160 0.7 0 SAA

10 10 months F 9.1 <10 Nl 6.3 585 ITP

11 14 F 4.9 70 1360 4.8 167 Drug induced marrow suppression

12 1 F 3.8 Nl 1080 1.9 64 Pearson's syndrome

13 1 F 3.3 Nl Nl 1.9 1792 Autoimmune hemolytic anemia

14 15 M Nl 31 620 6.4 64 Ehrlichia chaffeensis infection

15 10 M 5.9 Nl Nl 4.5 1342 Kawasaki disease

16 7 F 94 1,070 8.4 86 Probably viral infection

17 12 M 6.1 Nl Nl 9.6 1563 New dx sickle cell disease with pain crisis

18 16 M 4.5 118 Nl 0.1 0 Severe vitamin B12 deficiency

19 1 M 7.7 Nl 890 4.2 343 TEC

20 11 M 11.2 Nl 1110 10.4 394 Benign ethnic neutropenia with viral

illness

21 6 M 9.9 <10 Nl 8.2 369 ITP

22 1 F 10.1 101 620 3.2 164 Bacteremia

23 16 M 4.3 93 1,600 8.1 151 PNH

24 16 F 11.3 126 450 4.0 286 Proprionic acidemia and viral infection

25 5 F Nl 54 Nl 12.5 225 Viral illness

26 13 M 12.7 <10 370 26 346 ITP with possible viral syndrome

27 17 F Nl <10 Nl 10.2 374 ITP

28 16 F 6.3 Nl 320 3.1 290 Sickle cell disease with viral illness

29 15 F 9.8 33 Nl 4.7 0 ITP and acute appendicitis

30 1 F 8.3 Nl 450 24 1202 Pneumonia, multiple infections, immunodeficiency

31 2 F 7.3 <10 780 1.7 120 SAA

32 9 F 10.4 19 540 3.4 54 SAA

33 4 F Nl 32 Nl 7.2 955 ITP

34 9 M Nl 11 Nl 3.3 598 ITP secondary to EBV infection

35 14 F 10.2 Nl 610 2.9 97 Anti-granulocyte antibodies, likely SLE

36 7 F 10.3 Nl Nl 3.6 120 Possible immunodeficiency

37 7 M Nl 75 Nl 8.9 345 Autoimmune hepatitis with hypersplenism and sequestration

38 1 F Nl 15 Nl 5.1 879 ITP

39 2 months M 7.7 Nl Nl 7.7 199 Neutropenia, resolved without therapy

Hemoglobin (Hb) expressed in g/dL, platelets (plts) x 103/μl, absolute monocyte count (AMC) and absolute neutrophil count (ANC) in cells/μl. Nl=value within the normal range. ITP= idiopathic thrombocytopenic purpura; TEC= Transient erythroblastopenia of childhood; TTP= Thrombotic thrombocytopenic purpura; SAA= Severe aplastic anemia; MAI= Mycobacterium avium intracellularae; PNH=paroxysmal nocturnal hemoglobinuria; EBV= Epstein Barr Virus; SLE= Systemic lupus erythematosus

Table 3

Monoclonal antibodies used for immunophenotypic characterization

 

Monoclonal Antibody

Clone In this study , used for

Source

CD19 SJ25C1 B-LL BD

CD10 HI10a B-LL BD

CD34 8G12 B-LL BD

CD20 L27 B-LL BD

CD58 1C3 (AICD58.6) B-LL BD Pharmingen

TDT HT1, HT4, HT8, HT9

B-LL Beckman Coulter

Kappa/Lambda TB28-2/1-155-2 B-LL Dako

CD33 P67.6 Monocytes BD

CD64 22 Monocytes Beckman Coulter

       

B-LL: B-lymphoblastic leukemia; BD: Becton Dickenson

Figure 1: Sample gating of monocytes. A large gate is drawn on FSC vs. SSC to include all

mononuclear cells as well as some granulocytes. Monocytes are then gated from CD64 vs.

CD33, which separates them from the granulocytes, which show dimmer CD64 and CD33

expression. Percent monocytes are expressed using all events (including all white cells) as the

denominator.

Figure 2a: Boxplots of the percentage monocytes as determined by peripheral blood flow

cytometry for patients with B-lymphoblastic leukemia vs. the negative control group. The top,

middle, and bottom bars of the box represent the 75th, 50th (median), and 25th percentiles,

respectively. The lines drawn from the box represent 1.5* intra quartile range (75th percentile -

25th percentile). Using the Wilcoxon test, the B-lymphoblastic leukemia group has statistically

lower monocytes than the negative controls (p<0.0001).

Figure 2b: Boxplots comparing the absolute monocyte counts for patients with B-lymphoblastic

leukemia vs. the negative control group. These plots are derived in a similar way as figure 2a.

Using the Wilcoxon test, the B-lymphoblastic leukemia group has statistically lower absolute

monocyte counts than the negative control group (p<0.0001).