Hematopathology / Aberrant Immunophenotype in Burkitt Lymphoma

Immunophenotypic Variations of Burkitt Lymphoma

Katalin Kelemen, MD, PhD,1 Rita M. Braziel, MD,1 Ken Gatter, MD,1 Tony C. Bakke, PhD,2
Susan Olson, PhD,3 and Guang Fan, MD, PhD1

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
Key Words: Burkitt lymphoma; Flow cytometry; Aberrant immunophenotype

DOI: 10.1309/AJCP93LJPTRQPDKR

Abstract Burkitt lymphoma (BL), a highly aggressive B-cell lym-

Burkitt lymphoma (BL) exhibits a characteristic phoma, represents approximately 2.5% of all non-Hodgkin
immunophenotype that is positive for pan–B-cell lymphomas (NHLs). BL preferentially involves extranodal
antigens and CD10 and expresses clonal surface sites, such as the small intestine or jaw, or may manifest as
immunoglobulins (SIgs). We evaluated 35 BLs and acute leukemia. The genetic hallmark of BL is a reciprocal
identified atypical immunophenotypes in 4 including translocation of the MYC gene on chromosome 8 most com-
SIg light chain negativity in 4, negativity for B-cell monly with IGH gene or, alternatively, with the κ or λ immu-
antigens in 2, lack of CD10 in 1, and CD4 expression noglobulin light chain genes.
in 1. All 4 cases showed morphologic features The diagnosis of BL is based on a combination of mor-
characteristic of BL, and all were confirmed by phologic, immunophenotypic, and cytogenetic findings. The
cytogenetic analysis. The 4 BL cases included 1 girl, 2 distinction between BL and other high-grade B-cell lympho-
men with HIV/AIDS, and a third man with a history of mas is clinically important, not only because BL is a tumor of
heart transplantation. Two patients died shortly after extremely high proliferation rate but because patients are at
diagnosis; the other 2 completed hyperfractionated special risk for central nervous system involvement. The mor-
cyclophosphamide, vincristine, doxorubicin, and phologic features of BL—sheets of monomorphic medium-
dexamethasone with highly active antiretroviral therapy sized B cells with basophilic cytoplasm, numerous mitotic fig-
and achieved complete remission. Our study indicates ures, and admixed macrophages (“starry-sky pattern”)—are
that an unusual immunophenotype is not uncommon not specific because lymphoblastic lymphoma, diffuse large
in BL. If the immunophenotype is interpreted in B-cell lymphoma (DLBCL), plasmablastic lymphoma, and
isolation, the diagnosis could be missed. Recognizing even a high-grade T-cell lymphoma can have this picture.
the variability of immunophenotype is essential for Evaluation of the tumor immunophenotype might help
establishing an accurate diagnosis of BL. to differentiate BL from these entities. The characteristic
immunophenotype of BL is a mature B cell with germinal
center cell differentiation. The tumor cells are positive for
B cell–associated antigens such as CD19, CD20, CD22, and
CD79a and express surface immunoglobulins (SIgs) with
light chain restriction and the germinal center cell markers
CD10 and bcl-6. The tumor cells are negative for CD5, CD23,
and terminal deoxynucleotidyl transferase (TdT) and usually
negative for bcl-2; however, bcl-2 can be expressed in 10%
to 20% of cases.1,2 A deviation from the classic immunophe-
notype can cause diagnostic confusion or might even result

© American Society for Clinical Pathology Am J Clin Pathol 2010;134:127-138 127

127 DOI: 10.1309/AJCP93LJPTRQPDKR 127
Kelemen et al / Aberrant Immunophenotype in Burkitt Lymphoma

in false exclusion of the diagnosis of BL. Currently, limited PA). Cells were analyzed on a FACSCalibur flow cytometer
data are available regarding the frequency of an aberrant flow (BD Biosciences). Data were analyzed by using CellQuest
cytometric immunophenotype in BL. The aim of our study software (BD Biosciences).
was to evaluate the frequency and spectrum of aberrant flow
cytometric immunophenotypes in BL and correlate the flow Immunohistochemical Analysis
cytometric results with immunohistochemical, cytogenetic, Immunohistochemical analysis was performed on for-
and clinical features. malin-fixed, paraffin-embedded tissue samples using the

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
following primary antibodies: anti-CD3 rabbit monoclonal,
anti-CD20 mouse monoclonal, anti–Ki-67 mouse mono-
clonal (Ventana Medical Systems, Tucson, AZ); anti-CD4
Materials and Methods
mouse monoclonal, anti-CD22 mouse monoclonal (Leica,
Bannockburn, IL); anti-CD10 mouse monoclonal, anti-CD79a
Case Selection and Evaluation mouse monoclonal, anti-PAX5 mouse monoclonal, anti–bcl-2
This study was approved by the institutional review mouse monoclonal (Cell Marque, Hot Springs, AR); anti-TdT
board of Oregon Health and Science University, Portland. The rabbit monoclonal (Super Techs, Bethesda, MD); and anti–
pathology database of our institution was searched for cases bcl-6 mouse monoclonal (DAKO North America, Carpinteria,
of BL from January 1999 to December 2007 that had been CA). After incubation with the primary antibody, immuno-
evaluated by morphologic evaluation and by flow cytometry. detection was performed with a biotin-conjugated second-
A total of 35 cases of BL were identified. The flow cytometric ary antibody formulation that recognizes rabbit and mouse
results of all 35 cases were reviewed. Cases with any devia- immunoglobulins (Ventana), followed by peroxidase-labeled
tion from an immunophenotype characteristic of BL were streptavidin and with diaminobenzidine chromogen as the
reevaluated for morphologic features, cytogenetic findings, substrate (Ventana UltraView Universal DAB Detection Kit).
and correlation with clinical features. Materials for review of All immunostaining was performed using a BenchMark XT
morphologic features included all diagnostic and follow-up automated immunostaining device (Ventana).
lymph node, bone marrow aspirate and core biopsy, body
fluid, and tissue biopsy specimens, if available. Clinical infor- In Situ Hybridization
mation included laboratory data at diagnosis, clinical manifes- In situ hybridization was performed on formalin-fixed,
tations, treatment received, and response to treatment. paraffin-embedded tissue samples using an Epstein-Barr
virus–encoded RNA in situ hybridization RNA probe reagent
Histologic Studies system (Ventana) and a BenchMark XT automated immuno-
Lymph node and extranodal tissue samples were fixed in staining device.
10% buffered formalin before embedding and were stained
with H&E for histologic evaluation. Peripheral blood samples Cytogenetic Studies
and bone marrow aspirate smears were stained with Wright- Cytogenetic analysis, including conventional G-banded
Giemsa for morphologic evaluation. Formalin-fixed, decalci- karyotyping and fluorescence in situ hybridization (FISH),
fied, trephine biopsy specimens were stained with H&E for was performed at OHSU Clinical Cytogenetic Laboratories,
histologic evaluation. Portland. Briefly, chromosome preparations were obtained
from peripheral blood, bone marrow, and lymph node samples
Flow Cytometric Immunophenotyping according to conventional methods following 24- and 48-hour
Cell suspensions were prepared and stained for flow cultures at 37°C in RPMI 1640 medium (GIBCO, Invitrogen,
cytometry within 24 hours of sample collection. For the Carlsbad, CA) supplemented with 10% fetal bovine serum
study, 50-μL aliquots of cell suspensions were incubated and antibiotics. The cultures were set up without the use of
with 4 color combinations of fluorescent monoclonal antibod- mitogens. At 30 minutes to 2 hours, depending on cell type,
ies, including CD3, CD4, CD5, CD7, CD8, CD10, CD19, before the initiation of the harvest, the cells were treated with
CD20, CD22, CD23, CD45, CD79a, and κ and λ light chains 0.05 μg/mL of Colcemid (GIBCO, Invitrogen). Following
(Becton Dickinson [BD] Biosciences, San Jose, CA). After hypotonic treatment in a 0.075 mol/L potassium chloride
30 minutes, 2.5 mL of ammonium chloride was added for 10 solution, the cells were fixed in freshly prepared 3:1 metha-
minutes to lyse the RBCs. The cells were pelleted and washed nol/glacial acetic acid. The cell suspension was dropped onto
with phosphate-buffered saline containing 0.1% bovine serum clean microscope slides that were then baked for 20 minutes
albumin and 0.1% sodium azide. The second cell pellet was at 90°C. Giemsa banding using a trypsin/Wright stain method
resuspended in 0.5 mL of phosphate-buffered saline contain- was performed. When available, at least 20 metaphases were
ing 1% EM grade formaldehyde (Polysciences, Warrington, analyzed. Karyotypes of Giemsa-banded chromosomes were

128 Am J Clin Pathol 2010;134:127-138 © American Society for Clinical Pathology

128 DOI: 10.1309/AJCP93LJPTRQPDKR
 Hematopathology / Original Article

described according to the 2009 International System of equipped with appropriate filters. At least 100 nuclei were
Human Cytogenetic Nomenclature.3 Abnormal clones were examined for each probe whenever possible. Images were
defined as 2 or more cells with the same structural abnormal- captured on a CytoVision computer system (Genetix-Applied
ity or the same extra chromosomes or the presence of 3 or Imaging, San Jose, CA).
more cells with loss of the same chromosome.

FISH Analysis
Results

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
Depending on availability, FISH procedures were per-
formed on cell suspensions prepared from fresh or cryopre-
served pellets, on 4- to 5-μm, unstained, paraffin-embedded Clinical Information at Diagnosis and Morphologic
tumor tissue sections, or on air-dried touch preparations. The Findings
studies were performed on interphase cells using the LSI Among the 35 patients with BL, 29 were male and 6 were
c-MYC (8q24.12-q24.13), c-MYC break-apart (8q24), IGH female. The ages ranged from 3 to 80 years with a median
(14q23), IGH/bcl-2 fusion, bcl-6 break-apart, and CEP-8 of 14. Of the 35 patients, 22 were in the pediatric age group
probes obtained from Vysis/Abbott, Downers Grove, IL. For (younger than 18 years) and 13 patients were adults. The
paraffin-embedded tissue, before hybridization, the slides 4 patients with variant immunophenotypes included 3 men
were deparaffinized according to standard laboratory proto- (ages 38, 51, and 54 years) and one 10-year-old girl. Of the
col. FISH was performed by codenaturation on a HYBrite 3 adults with aberrant immunophenotypes, 2 had HIV/AIDS
instrument (Vysis/Abbott) at a denaturation temperature and the third patient had a history of heart transplantation. The
of 72°C for 2 minutes for freshly dropped cells, followed demographic information, specimen type, flow cytometric,
by overnight hybridization at 37°C. The slides were then immunohistochemical, in situ hybridization, and cytogenetic
washed with 2× sodium chloride–sodium phosphate–EDTA findings for the 4 cases are listed in ❚Table 1❚. Clinical stage
buffer/0.3% NP-40 at 72°C for 2 minutes. Temperature and at diagnosis, treatment received, and outcome are given in
timing varied slightly for deparaffinized sections. The cells ❚Table 2❚.
were counterstained with DAPI II (Abbott, Downers Grove, Diagnostic and, when available, follow-up tissue biopsy
IL) and viewed on a Nikon E800 fluorescence microscope specimens were reviewed. The diagnostic specimens included

❚Table 1❚
Clinicopathologic Characteristics in Four Cases of Burkitt Lymphoma With Immunophenotypic Variations

Case No.

1 2 3 4

Sex/age (y) M/54 M/38 F/10 F/51

Clinical background HIV/AIDS HIV/AIDS Leukemic presentation Heart transplantation
Diagnostic tissue Right leg skeletal muscle Axillary lymph node; Blood; bone marrow Left groin lymph node
chest wall mass
Flow cytometry
SIg light chain Negative Negative Negative Negative
CIg light chain Negative Cytoplasmic κ+ Cytoplasmic λ+ Negative
B-cell lineage antigens CD19–, CD20–, CD22–, CD19+, CD20+, CD22+ CD19+, CD20+, CD22+ CD20–, CD19+, dim CD22+
cCD79a+
Germinal center marker CD10+ CD10+ CD10+ CD10–
T-cell antigen CD4+ NE NE NE
Additional antigens CD38+, dim CD45+, CD138– NE NE NE
Immunohistochemical analysis CD4+, CD10+, CD20–, CD10+, CD20+, PAX5+, bcl-2– (limited studies CD20–, CD79a+, bcl-2–
CD22–, CD79a+, bcl-2–, bcl-2–, bcl-6+ done)
bcl-6–, TdT–
Ki-67 proliferation index (%) 100 100 >90 100
EBER ISH Positive Negative Negative Positive
Cytogenetic findings t(8;14) t(8;14) plus extra Complex karyotype; all Positive for c-MYC
chromosome 18 cells have t(8;14) and rearrangement by FISH
other abnormalities of c-MYC break-apart probe;
chromosomes 1, 2, 3, negative for IGH/bcl-2
10, and 13 fusion and bcl-6 break-
apart probes

CIg, cytoplasmic immunoglobulin; EBER, Epstein-Barr virus–encoded RNA; FISH, fluorescence ISH; ISH, in situ hybridization; NE, not expressed; SIg, surface
immunoglobulin; TdT, terminal deoxynucleotidyl transferase.

© American Society for Clinical Pathology Am J Clin Pathol 2010;134:127-138 129

129 DOI: 10.1309/AJCP93LJPTRQPDKR 129
Kelemen et al / Aberrant Immunophenotype in Burkitt Lymphoma

❚Table 2❚
Clinical Stage, Treatment, and Outcome in Four Patients With Burkitt Lymphoma With Unusual Immunophenotypes

Case No./
Sex/Age (y) Stage at Diagnosis/Diagnostic Tissue Treatment Outcome

1/M/54 Stage IVE/leg abductor muscle HyperCVAD; HAART Well after 16 mo

2/M/38 Stage IIA/axillary lymph node HyperCVAD; radiation; HAART Well after 13 mo
3/F/10 Stage IV/blood and bone marrow COPADM, 2 cycles Died 4 mo after diagnosis of relapse and fungal sepsis
4/M/51* Stage IVE/bilateral adrenals, kidneys, None Died 4 d after diagnosis, cardiac failure

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
liver, and para-aortic lymph nodes

COPADM, cyclophosphamide, vincristine, prednisone, and doxorubicin plus methotrexate; HAART, highly active antiretroviral therapy; hyperCVAD, hyperfractionated
cyclophosphamide, vincristine, doxorubicin, and dexamethasone.
* Heart transplant recipient.

abductor muscle excisional biopsy of the right leg (case 1), B cell–associated antigens CD19, CD20, CD22, or CD79;
axillary lymph node and chest mass (case 2), peripheral blood lack of the germinal center cell marker CD10; and aberrant
and bone marrow biopsy (case 3), and inguinal lymph node expression of CD4.
(case 4). In case 3, a follow-up bone marrow biopsy specimen The most commonly observed abnormality was lack of
was also available for review. SIg light chain expression, detected in all 4 cases. Of the 4
H&E-stained sections and immunohistochemical stains cases, 2 showed evidence of light chain restriction on stain-
were reviewed. The morphologic findings of the abductor ing for cytoplasmic immunoglobulin light chains (case 2 and
muscle biopsy (case 1), axillary lymph node and chest mass case 3, cytoplasmic κ and cytoplasmic λ, respectively). Cases
(case 2), and inguinal lymph node (case 4) were compatible 1 and 4 were negative for surface and cytoplasmic immuno-
with BL: the lymphoma was composed of diffuse prolif- globulin light chain expression.
eration of uniform, medium-sized lymphocytes with deeply Cases 3 and 4 showed several immunophenotypic
basophilic cytoplasm, several visible nucleoli, a high rate deviations in addition to lack of SIg light chain expres-
of mitosis and apoptosis, and a starry-sky pattern. Case 3, sion. Both of these cases lacked expression of 1 or more
a case of BL in leukemic phase, showed an elevated WBC B-cell–associated antigens. Case 1 lacked expression of 3
count of 22,600/μL (22.6 × 103/μL) and many large atypical B cell–associated antigens, CD19, CD20, and CD22, by
lymphocytes with deeply basophilic cytoplasm and cyto- flow cytometry. The B-cell lineage of the tumor cells in this
plasmic vacuoles, consistent with circulating BL cells. The case was confirmed by CD79 expression by flow cytometry
bone marrow aspirate smears were extensively involved, as and immunohistochemical analysis. Case 4 was positive
75% of the cells were replaced by BL cells exhibiting mor- for CD19 but lacked expression of CD20 and displayed
phologic features similar to those in the peripheral blood. dim expression of CD22. In both of these cases, the lack
The bilateral bone marrow core biopsies showed 100% cel- of expression of B-cell antigen CD20 was reproduced by
lularity and extensive monotonous sheets of medium-sized immunohistochemical analysis.
lymphoid cells with brisk mitosis and individual cell necro- Additional immunophenotypic aberrancy in case 1
sis. After initial therapy, a repeated bone marrow biopsy included strong expression of CD4, a marker expressed by
at 4 weeks showed no evidence of BL. However, 8 weeks T lymphocytes and monocytes but usually not by B cells.
after initial diagnosis, another bone marrow biopsy showed Furthermore, the lymphoma cells in case 1 showed dim
relapsed BL with approximately 10% to 15% of bone mar- expression of CD45 and expression of CD38, an immunophe-
row cellularity represented by BL cells. notype that raises the possibility of plasmacytic differentiation,
especially considering the lack of CD19, CD20, and CD22 by
Flow Cytometric Immunophenotyping flow cytometry. CD138 expression, however, was negative in
Of the 35 BL cases, 31 showed a classic immunopheno- case 1, and immunohistochemical analysis showed positivity
type by flow cytometry, including SIg light chain restriction for PAX5 and CD79a, confirming the B-cell lineage of the
(14 κ; 17 λ); expression of B cell–associated markers CD19, neoplasm ❚Image 1❚ and ❚Image 2❚.
CD20, CD22, and CD79a; and the germinal center cell– Case 4 also demonstrated multiple immunophenotypic
associated marker CD10. Four cases were found to have aberrancies in addition to lack of SIg light chain expres-
variations compared with the typical immunophenotype of sion. Although the B-cell marker CD19 was positive by
BL. Variations were detected in the following categories: flow cytometry, CD22 expression was dim and CD20 was
lack of surface or surface and cytoplasmic immunoglobulin negative. In addition, case 4 lacked expression of the germinal
light chain expression; lack of expression of one or more center marker CD10 ❚Image 3❚.

130 Am J Clin Pathol 2010;134:127-138 © American Society for Clinical Pathology

130 DOI: 10.1309/AJCP93LJPTRQPDKR
 Hematopathology / Original Article

4 4
1,000 10 10
R2: 4.67% R2: 0.10%

CD45 PerCP-Cy5.5

CD20 PerCP-Cy5.5
800 3 3
10 10

FSC Height
600
2 2
10 10
400
R5: 93.27%
1 1
10 10
200 R5: 97.88%

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
0 0
0 10 10
0 200 400 600 800 1,000 0 200 400 600 800 1,000 0 200 400 600 800 1,000
SSC Height SSC Height SSC Height
4 4 4
10 10 0.93% 0.97%
10 75.97% 0.17%
CD20 PerCP-Cy5.5

3 3 3
10 10 10

CD4 PE
λ PE
2 2 2
10 10 10
R7
1 1 1
10 10 10 R7

0 0 98.10% 0.00% 0 23.78% 0.08%

10 10 10
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
CD22 APC κ FITC CD56 FITC
4 4
1,000 10 10 0.51% 0.94%
CD45 PerCP-Cy5.5

CD19 PerCP-Cy5.5
800 3 3
10 10
FSC Height

600
2 2
10 10
400 R1: 84.14%
1 1
10 10 R11
200
0 0 22.62% 75.93%
0 10 10
0 1 2 3 4 0 1 2 3 4
0 200 400 600 800 1,000 10 10 10 10 10 10 10 10 10 10
SSC Height CD38 APC CD10 APC
4 4 4
10 0.03% 3.04%
10 0.39% 16.29%
10 69.85% 3.49%

3 3 3
10 10 10
CD79a PE
CD34 PE

λ PE

2 2 2
10 10 10
R19 R7 R7
1 1 1
10 10 10

0 1.30% 95.64% 0 42.23% 41.10% 0 26.40% 0.25%

10 10 10
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
CD10 FITC κ FITC cTdT

❚Image 1❚ (Case 1) Flow cytometric immunophenotype, right leg abductor muscle biopsy. The analysis shows a predominant
population of cells (93% of total cellular events) with the following immunophenotype: positive for CD4, CD10, CD38, dim
CD45, and cytoplasmic CD79a; negative for CD19, CD20, CD22, and surface and cytoplasmic immunoglobulin light chains. APC,
allophycocyanin; cTdT, cytoplasmic terminal deoxynucleotidyl transferase; Cy5.5, cyanine 5.5; FITC, fluorescein isothiocyanate;
FSC, forward scatter; PE, phycoerythrin; PerCP, peridinin chlorophyll protein; SSC, side scatter.

Immunohistochemical Results cytometry, aberrant immunohistochemical findings were not

Immunohistochemical stains were reviewed with a spe- expected. Consistent with the flow cytometric findings, case
cial emphasis on correlation with the aberrant immunopheno- 1 was negative by immunohistochemical analysis for CD20
typic findings by flow cytometry. The immunohistochemical and aberrantly expressed CD4; otherwise, it was positive
findings in cases 2 and 3 were consistent with BL without for CD10 and CD79 (Image 2). Similarly, the aberrant flow
anomalies. Because in these 2 cases the lack of SIg light cytometric immunophenotype was reproduced by immuno-
chain expression was observed as a single anomaly by flow histochemical analysis in case 4: it was negative for CD20

© American Society for Clinical Pathology Am J Clin Pathol 2010;134:127-138 131

131 DOI: 10.1309/AJCP93LJPTRQPDKR 131
Kelemen et al / Aberrant Immunophenotype in Burkitt Lymphoma

A B

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
C D

❚Image 2❚ (Case 1) Immunohistochemical findings, right leg abductor muscle biopsy. H&E-stained section (A, ×200) shows
morphologic findings typical of Burkitt lymphoma, notably, diffuse sheets of monomorphic medium-sized cells with basophilic
cytoplasm, high nuclear/cytoplasmic ratio, and moderately dispersed chromatin with multiple peripherally placed small nucleoli.
Numerous mitotic figures and frequent apoptotic cells are present, and admixed macrophages create a “starry-sky” pattern. The
tumor cells express CD10 (B, ×200) and CD79 (C, ×200), show a very high Ki-67 proliferation index approaching 100% (D, ×200),
and are strongly positive for Epstein-Barr virus by Epstein-Barr virus–encoded RNA (EBER) in situ hybridization (ISH) (E, ×200).

and CD10 and positive for CD79 ❚Image 4 ❚. In all 4 cases, translocation by FISH. In case 2, the axillary lymph node
bcl-2 was negative. Ki-67 proliferation indices were greater sample showed evidence of a t(8;14) translocation in addi-
than 95% in all 4 cases. Epstein-Barr virus–encoded RNA in tion to an extra copy of chromosome 18. Case 3 showed a
situ hybridization was positive in cases 1 and 4 and negative complex karyotype in which all analyzed cells were abnor-
in cases 2 and 3. mal and carried a t(8;14); however, additional structural
abnormalities of chromosomes 1, 2, 3, 10, and 13 were
Cytogenetic Findings also observed as follows: 46,XX,t(1;13)(q25;q12),t(2;3)
Cytogenetic findings were reviewed in all 4 BL cases (p3?2;p21),t(8;14)(q24;q32),inv(10)(p11.2;q32)
with atypical immunophenotypes. Representative results [1]/46,idem,del(1)(q21),der(13)t(1;13)(q25;q12)add(13)
are shown in ❚Image 5❚. In case 1, the right leg abduc- (p11)[6]/46,XX,t(2;3),t(8;14),inv(10)[2]/46,idem,add(1)
tor muscle biopsy sample showed evidence of a t(8;14) der(1)t(1;13)(q2?5)[16]. In case 4, touch preparation of the

132 Am J Clin Pathol 2010;134:127-138 © American Society for Clinical Pathology

132 DOI: 10.1309/AJCP93LJPTRQPDKR
 Hematopathology / Original Article

E F

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
❚Image 2❚ The tumor cells are negative for expression of CD20 (F, ×200).

4 4 4
10 65.5% 3.3%
10 16.0% 0.7%
10 3.8% 0.5%
CD19 PE FL2-H

CD20 PE FL2-H

CD23 PE FL2-H
3 3 3
10 10 10
R3
2 2 2
10 10 10
R3 R3
1 1 1
10 10 10

0 3.4% 27.7% 0 81.2% 2.0% 0 93.7% 2.0%

10 10 10
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
CD5 FITC FL1-H CD103 FITC FL1-H CD10 FITC FL1-H
4 4 4
10 1.8% 1.9%
10 2.7% 1.0%
10
2.7% 6.7%
CD11c PE FL2-H

3 3 3
10 10 R3 10
λ PE FL2-H

R3
λ PE

2 2 2
10 10 10
R3

1 1 1
10 10 10

0 42.5% 53.7% 0 95.2% 1.1% 0 85.3% 5.3%

10 10 10
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
CD22 FITC FL1-H κ FITC FL1-H κ FITC

❚Image 3❚ (Case 4) Flow cytometric immunophenotype, left inguinal lymph node. The analysis reveals an abnormal
B-lymphocyte population (65% of cellular events) with the following immunophenotype: positive for CD19, CD22, and CD45;
negative for CD5, CD10, CD23, CD11c, CD103, and surface and cytoplasmic immunoglobulin light chains. The remaining
lymphocytes are a mixture of CD5+ T cells and a few polyclonal B cells. FITC, fluorescein isothiocyanate; FL, fluorochrome
label; H, height; PE, phycoerythrin.

inguinal lymph node showed evidence, by a c-MYC break- Therapy Received and Clinical Follow-up
apart probe, of a c-MYC translocation. In addition, FISH Patient 1, a 54-year-old man with HIV/AIDS was
study of case 4 was negative using an IGH/bcl-2 fusion treated with hyperfractionated cyclophosphamide, vincristine,
probe set and a bcl-6 break-apart probe, excluding the pres- doxorubicin, and dexamethasone (hyperCVAD) and highly
ence of t(14;18) and translocation of bcl-6. These findings active antiretroviral therapy (HAART) and achieved complete
support the diagnosis of a BL vs a “double-hit” high-grade remission and remained in remission at the time of latest fol-
B cell lymphoma. low-up of 16 months. Patient 2, a 38-year-old man who was

© American Society for Clinical Pathology Am J Clin Pathol 2010;134:127-138 133

133 DOI: 10.1309/AJCP93LJPTRQPDKR 133
Kelemen et al / Aberrant Immunophenotype in Burkitt Lymphoma

A B

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
C D

❚Image 4❚ (Case 4) Immunohistochemical findings, left inguinal lymph node. A, H&E-stained section shows classic
morphologic features of Burkitt lymphoma with a “starry-sky” pattern (×200). The lymphoma cells, however, are negative
for expression of CD20 (B, ×200). B-cell lineage differentiation is evidenced by expression of CD79 (C, ×400). Epstein-Barr
virus–encoded RNA (EBER) in situ hybridization (ISH) is positive (D, ×400). Ki-67 proliferation index (not shown) approaches
100% in the lymphoma cells.

diagnosed with BL and HIV/AIDS at the same time received before the diagnosis of BL, died 4 days after the diagnostic
hyperCVAD combined with local radiation to the axillary and inguinal lymph node excisional biopsy of cardiac failure. He
chest area and HAART. He achieved complete remission and did not receive therapy for BL.
was well at 13 months of follow-up. Patient 3, a 10-year-old
girl, received 2 cycles of COPADM chemotherapy (cyclo-
phosphamide, vincristine, prednisone, and doxorubicin plus Discussion
methotrexate), and she achieved a short remission at 4 weeks; Accurate diagnosis of BL is critical because of the
however, the disease relapsed in the bone marrow 8 weeks extremely short doubling time of this lymphoma and because
after initiation of therapy, and the patient died of fungal sepsis of the high risk of central nervous system involvement. The
and multiorgan failure. Patient 4, a 51-year-old man with a diagnosis of BL is made by examining the morphologic fea-
history of heart transplantation for Chagas disease 6 years tures of tissue biopsy specimens in conjunction with finding a

134 Am J Clin Pathol 2010;134:127-138 © American Society for Clinical Pathology

134 DOI: 10.1309/AJCP93LJPTRQPDKR
 Hematopathology / Original Article

A B

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
C D

❚Image 5❚ Cytogenetic findings in 4 Burkitt lymphoma cases with unusual immunophenotypes. A (Case 1), Right leg abductor
muscle biopsy. Interphase fluorescence in situ hybridization (FISH) shows evidence of t(8;14) translocation. B (Case 2), Axillary
lymph node sample shows evidence of a t(8;14) translocation in addition to an extra copy of chromosome 18 (A and B, red,
MYC; green, IGH; aqua, chromosome 8; yellow, MYC/IGH fusion). C (Case 3), A complex karyotype, in which all analyzed cells
carry a t(8;14); however, additional structural abnormalities of chromosomes 1, 2, 3, 10, and 13 were also observed. D (Case
4), Interphase FISH shows evidence of a t(8;14) translocation (red, MYC; green, IGH). Yellow, normal c-MYC; 1 red/1 green/1
yellow, c-MYC rearrangement.

characteristic immunophenotype. Flow cytometry has gained markers CD19, CD20, CD22, and CD79; and clonally
increasing importance in the diagnostic workup of B-cell restricted SIg light chains. In our study, 31 (89%) of 35 BL
lymphomas because, unlike immunohistochemical analysis, cases were found to express a classic immunophenotype by
immunoglobulin light restriction is readily detected and the flow cytometry. The remaining 4 cases (11%), however,
results are often available within a few hours after the biopsy showed deviations from the classic phenotype.
is performed. Despite the growing importance of flow cytom- The most commonly observed deviation was lack of
etry in the diagnostic workup, there are few data available on SIg light chain expression, observed in all 4 cases. Mature
variations of the flow cytometric immunophenotype in BL. B cells express κ or λ immunoglobulin light chains on the
Immunophenotypically, BL is defined as a mature germinal cell surface. The demonstration of SIg light chain restric-
center cell–type B cell expressing CD10; B cell–associated tion indicates monoclonality of the proliferating mature

© American Society for Clinical Pathology Am J Clin Pathol 2010;134:127-138 135

135 DOI: 10.1309/AJCP93LJPTRQPDKR 135
Kelemen et al / Aberrant Immunophenotype in Burkitt Lymphoma

B cells, a feature useful to support the diagnosis of lym- Lack of B-cell lineage–associated markers may represent
phoma. Early studies with the relatively insensitive method a major challenge in the diagnosis of B-cell malignancies. In
of immunohistochemical analysis demonstrated lack of κ 2 of our BL cases with aberrant immunophenotypes, the lack
or λ SIg light chain proteins in up to one third of B-cell of SIg light chains was further complicated by the absence
NHLs.4 Subsequent data obtained with the more sensitive of 1 or more B cell–associated markers. Both of these cases
flow cytometry indicated that SIg– malignant B-cell NHLs lacked expression of CD20 by flow cytometry, and this find-
were rare; the reported frequency ranged from 3.4% to ing was reproduced in both cases by immunohistochemical

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
12.2%. In these studies, the most frequently observed B-cell analysis. CD20, a membrane-embedded nonglycosylated
lymphoma that lacks SIg expression was DLBCL, followed phosphoprotein is expressed on the surface of all mature B
by follicular lymphoma; only rare cases of BL and low lymphocytes and considered one of the best markers of B-cell
grade B-cell lymphomas were described.5-11 Kaleem et al11 lymphomas. CD20 is not expressed in immature B-cell pre-
described 10 cases of SIg– NHLs, including 2 cases of BL. cursors or in plasma cells. In a study of CD20– B-cell neo-
Li et al12 also reported 2 BLs among 36 B-cell NHLs with- plasms, Chu et al13 concluded that mature B-cell lymphomas
out SIg light chain expression. These reports are intriguing; with absent CD20 expression are rare outside the setting of
however, the overall incidence of SIg– BL is not clear from rituximab therapy and usually are seen in HIV-associated
these studies. high-grade DLBCLs, such as primary effusion lymphoma and
B-cell NHLs without SIg light chain expression need to plasmablastic lymphoma, both of which are commonly nega-
be distinguished from precursor B-cell acute lymphoblastic tive for CD19, CD20, and CD79a. B-lineage determination
leukemia (pre-B ALL), plasma cell neoplasms, and some of B-cell malignancies not expressing CD20 may depend on
cases from benign B-cell populations, such as hematogones demonstration of CD19, CD22, or CD79 expression by flow
in a bone marrow sample or early germinal center cell B cytometry; however, in plasmablastic lymphoma and primary
lymphocytes in lymph node samples with florid follicular effusion lymphoma, OCT.2 or BOB.1 immunohistochemi-
hyperplasia. Leukemic manifestation of BL with absent SIg cal studies or documentation of clonal immunoglobulin gene
expression might represent a diagnostic challenge against rearrangement might be necessary for diagnosis.14-16
precursor B-ALL because pre-B ALL and BL usually have The immunophenotype in case 1 is remarkable for lack
a CD10+ immunophenotype. The almost universal expres- of 3 distinct B-cell lineage markers, CD20, CD19, and CD22,
sion of TdT by ALL and its lack in BL is the only reliable by flow cytometry, in addition to the strong uniform expres-
criterion in the differential diagnosis. Case 3 represents an sion of CD4. In fact, this case in our study showed several
example of this diagnostic dilemma: it involves blood and similarities to plasmablastic lymphoma: the patient is HIV+
bone marrow samples from a 10-year-old girl with a leuke- and the lymphoma is Epstein-Barr virus (EBV)-associated,
mic-type picture of BL with absent SIg light chain expres- extranodal, and, in addition to the lack of CD19, CD20,
sion. TdT was negative on the bone marrow core biopsy and CD22 by flow cytometry, the lymphoma cells strongly
specimen, and cytogenetic analysis confirmed the presence express CD38. This case also expresses CD4, a finding highly
of a t(8;14) chromosomal translocation embedded in a com- unexpected in B-cell lymphomas because its expression is so
plex karyotype. Similarly, a diagnostic dilemma may arise rare. Expression of CD4 has been described in at least 1 case
in fine-needle aspiration biopsies of lymph nodes involved of oral plasmablastic lymphoma.17 On the other hand, our
by SIg– BL; diagnostic pitfalls may include lymphoblastic case did not express CD138, and the B-cell lineage was con-
lymphoma, SIg– DLBCL, and, less likely, florid follicular firmed by expression of CD79 and PAX-5. Most important,
hyperplasia. In a review of almost 500 reactive lymphoid the diagnosis of BL in case 1 was confirmed by the presence
populations, Li et al12 found that in reactive follicular hyper- of t(8;14) by cytogenetic studies.
plasia, SIg– B cells represented less than 17% of total B In our study, 2 of the cases with deviant immunophe-
cells. They concluded that the complete absence of SIg light notypes (cases 1 and 2) were HIV-associated. In both of the
chain expression in a mature B-cell proliferation can be used patients, the lymphoma represented the first clinical mani-
as a surrogate marker of malignant B-cell lymphoma.12 festation of AIDS, and in case 2, HIV testing was prompted
In many laboratories, including ours, the finding of by the diagnosis of BL. Overall, BL is the most common
absent SIg light chain expression is followed by staining HIV-associated B-cell NHL and accounts for 30% of all HIV-
for cytoplasmic immunoglobulin light chains. In 2 of the 4 associated lymphomas. Of note, two thirds of cases show plas-
deviant cases (cases 2 and 3), cytoplasmic immunoglobulin macytoid differentiation, a relatively unique feature of HIV-
light chain staining demonstrated clonally restricted cytoplas- associated BL and a feature well illustrated by our case 1.18
mic immunoglobulin light chains. Cases 1 and 4, however, Case 4 occurred in a heart transplant recipient and, as
failed to show both surface and cytoplasmic immunoglobulin such, represents a monomorphic posttransplantation lym-
expression, a feature that has not been reported before in BL. phoproliferative disorder (PTLD), BL type. By definition,

136 Am J Clin Pathol 2010;134:127-138 © American Society for Clinical Pathology

136 DOI: 10.1309/AJCP93LJPTRQPDKR
 Hematopathology / Original Article

monomorphic PTLDs fulfill diagnostic criteria of one of the with EBV positivity but might be related to a partial plas-
well-defined B- or T-cell neoplasms that is recognized in an macytic differentiation in HIV-associated BLs. Recognition
immunocompetent host. Flow cytometric immunophenotyp- of immunophenotypic variations of BL is important, and the
ing is important in PTLDs because detection of clonality is atypical immunophenotype should not delay correlation with
one of the essential pieces of information in the classification, morphologic findings and with cytogenetic studies for an
and detection of the expression of CD20 enables the inclu- accurate diagnosis of BL.
sion of anti-CD20 monoclonal antibody therapy. Yet, there

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
are few data available on the immunophenotypic spectrum From the 1Department of Pathology, 2Clinical Flow Cytometric
of PTLDs by flow cytometry. Several studies reported a Laboratory, and 3Clinical Cytogenetic Laboratory, Oregon Health
& Science University, Portland.
higher incidence of absent CD20 expression and a higher
incidence of lack of SIg expression in different PTLDs of Address reprint requests to Dr Kelemen: Dept of Pathology,
a mature B-cell type compared with non-PTLD mature Oregon Health & Science University, Mail Code L-471, 3181 SW
Sam Jackson Park Rd, Portland, OR 97239-3098.
B-cell NHLs.19,20 Kaleem et al20 reported flow cytometric
immunophenotypes in 3 BL-type monomorphic PTLDs, 2
of which showed immunophenotypic aberrancies, including
1 with lack of SIg light chain expression and 1 with absent
CD10 expression. Kowal-Vern et al21 reviewed features of References
21 B-cell PTLDs occurring in cardiac transplant recipients.
1. Rodig SJ, Vergilio JA, Shahsafaei A, et al. Characteristic
They concluded that PTLDs in cardiac transplant recipients expression patterns of TCL1, CD38, and CD44 identify
are usually highly aggressive EBV+ lymphomas. Their study aggressive lymphomas harboring a MYC translocation. Am J
included 1 BL-type PTLD, and this case displayed a classic Surg Pathol. 2008;32:113-122.
flow cytometric immunophenotype (CD10+/CD19+/CD20+). 2. Chuang SS, Ye H, Du MQ, et al. Histopathology and
immunohistochemistry in distinguishing Burkitt lymphoma
Case 4 in our study is in line with their observation in that from diffuse large B-cell lymphoma with very high
it is an EBV+ lymphoma; however, the immunophenotypic proliferation index and with or without a starry-sky pattern:
deviation of case 4 is more complex than any of the previously a comparative study with EBER and FISH. Am J Clin Pathol.
2007;128:558-564.
reported BL PTLD cases: It combines the lack of surface and
3. Shaffer LG, Slovak ML, Campbell LJ, eds. An International
cytoplasmic immunoglobulin expression with absent CD10 System for Human Cytogenetic Nomenclature. Basel,
and CD20 by flow cytometry. Switzerland: Karger; 2009.
From a prognostic viewpoint, it is difficult to determine 4. Picker LJ, Weiss LM, Medeiros LJ, et al. Immunophenotypic
whether BL with an aberrant flow cytometric immunopheno- criteria for the diagnosis of non-Hodgkin’s lymphoma. Am J
Pathol. 1987;128:181-201.
type might behave differently compared with BL with a clas-
5. Liendo C, Danieu L, Al-Katib A, et al. Phenotypic analysis
sic immunophenotype. Of the 4 patients, 2 died shortly after
by flow cytometry of surface immunoglobulin light chains
diagnosis; however, both had multiple factors to contribute and B and T cell antigens in lymph nodes involved with
to the poor outcome, such as advanced stage, leukemic pre- non-Hodgkin’s lymphoma. Am J Med. 1985;79:445-454.
sentation, and complex cytogenetic findings in case 3 and the 6. de Martini RM, Turner RR, Boone DC, et al. Lymphocyte
history of heart transplantation and cardiac failure in case 4. immunophenotyping of B-cell lymphomas: a flow cytometric
analysis of neoplastic and nonneoplastic cells in 271 cases.
The 2 other patients responded to standard chemotherapy as Clin Immunol Immunopathol. 1988;49:365-379.
expected and were well after 16 and 13 months of follow-up, 7. Batata A, Shen B. Diagnostic value of clonality of
respectively. The atypical immunophenotype did not seem to surface immunoglobulin light and heavy chains in
correlate with EBV positivity of the lymphoma (cases 1 and malignant lymphoproliferative disorders. Am J Hematol.
1993;43:265-270.
4, positive; cases 2 and 3, negative).
8. Geary WA, Frierson HF, Innes DJ, et al. Quantitative
We report flow cytometric immunophenotypic varia- criteria for clonality in the diagnosis of B-cell non-Hodgkin’s
tions in 11% of BL cases at diagnosis, including lack of SIg lymphoma by flow cytometry. Mod Pathol. 1993;6:155-161.
expression; absent surface and cytoplasmic immunoglobu- 9. Fukushima PI, Nguyen PK, O’Grady P, et al. Flow cytometric
lin expression; lack of expression of one or several B-cell analysis of kappa and lambda light chain expression in
evaluation of specimens for B-cell neoplasia. Cytometry.
lineage–associated markers CD19, CD20, or CD22; lack of 1996;26:243-252.
expression of germinal center marker CD10; and aberrant 10. Zardawi IM, Jain S, Bennett G. Flow-cytometric algorithm on
expression of CD4. If the immunophenotype is interpreted in fine-needle aspirates for the clinical workup of patients with
isolation, the diagnosis of BL could be missed. Variant flow lymphadenopathy. Diagn Cytopathol. 1998;19:274-278.
cytometric immunophenotypes are observed in adult and 11. Kaleem Z, Zehnbauer BA, White G, et al. Lack of
expression of surface immunoglobulin light chains in
pediatric age groups and in the setting of BL-type PTLDs. The B-cell non-Hodgkin lymphomas. Am J Clin Pathol.
aberrant immunophenotype in BL does not seem to correlate 2000;113:399-405.

© American Society for Clinical Pathology Am J Clin Pathol 2010;134:127-138 137

137 DOI: 10.1309/AJCP93LJPTRQPDKR 137
Kelemen et al / Aberrant Immunophenotype in Burkitt Lymphoma

12. Li S, Eshleman JR, Borowitz MJ. Lack of surface 17. Tzankov A, Brunhuber T, Gschwendtner A, et al. Incidental
immunoglobulin light chain expression by flow cytometric oral plasmablastic lymphoma with aberrant expression of
immunophenotyping can help diagnose peripheral B-cell CD4 in an elderly HIV-negative patient: how a gingival
lymphoma. Am J Clin Pathol. 2002;118:229-234. polyp can cause confusion. Histopathology. 2005;46:348-350.
13. Chu PG, Loera S, Huang Q, et al. Lineage determination of 18. Davi F, Delecluse HJ, Guiet P, et al, for the Burkitt’s
CD20– B-cell neoplasms: an immunohistochemical study. Lymphoma Study Group. Burkitt-like lymphomas in
Am J Clin Pathol. 2006;126:534-544. AIDS patients: characterization within a series of 103
14. Dong HY, Scadden DT, de Leval L, et al. Plasmablastic human immunodeficiency virus–associated non-Hodgkin’s
lymphomas. J Clin Oncol. 1998;16:3788-3795.

Downloaded from https://academic.oup.com/ajcp/article/134/1/127/1766527 by University of Iowa Libraries/Serials Acquisitions user on 01 August 2023
lymphoma in HIV-positive patients: an aggressive Epstein-
Barr virus–associated extramedullary plasmacytic neoplasm. 19. Dunphy CH, Gardner LJ, Grosso LE, et al. Flow cytometric
Am J Surg Pathol. 2005;29:1633-1641. immunophenotyping in posttransplant lymphoproliferative
15. Knowles DM, Inghirami G, Ubriaco A, et al. Molecular disorders. Am J Clin Pathol. 2002;117:24-28.
genetic analysis of three AIDS-associated neoplasms of 20. Kaleem Z, Hassan A, Pathan MH, et al. Flow cytometric
uncertain lineage demonstrates their B-cell derivation and evaluation of posttransplant B-cell lymphoproliferative
the possible pathogenetic role of the Epstein-Barr virus. disorders. Arch Pathol Lab Med. 2004;128:181-186.
Blood. 1989;73:792-799. 21. Kowal-Vern A, Swinnen L, Pyle J, et al. Characterization
16. Nador RG, Cesarman E, Chadburn A, et al. Primary effusion of postcardiac transplant lymphomas: histology,
lymphoma: a distinct clinicopathologic entity associated immunophenotyping, immunohistochemistry, and gene
with the Kaposi’s sarcoma–associated herpes virus. Blood. rearrangement. Arch Pathol Lab Med. 1996;120:41-48.
1996;88:645-656.

138 Am J Clin Pathol 2010;134:127-138 © American Society for Clinical Pathology

138 DOI: 10.1309/AJCP93LJPTRQPDKR