Vol. 10, 5587–5594, August 15, 2004
Clinical Cancer Research 5587
Degree of CD25 Expression in T-Cell Lymphoma Is Dependent on
Tissue Site: Implications for Targeted Therapy
Dan Jones,1 Sherif Ibrahim,3 Kaushali Patel,1
Rajyalakshmi Luthra,1 Madeleine Duvic,2 and
L. Jeffrey Medeiros1
Departments of 1Hematopathology and 2Dermatology, University of
Texas M. D. Anderson Cancer Center, Houston, Texas, and the
3
Department of Pathology, New York University School of Medicine,
New York, New York
ABSTRACT
Purpose: Using concurrent tumor samples from different anatomical sites, we compared expression of the therapeutic targets CD25 and CD30 in T-cell lymphoma (TCL).
Experimental Design: We examined levels of CD25 and
CD30 by flow cytometry in tumor cells from peripheral
blood and lymph node in 13 cutaneous TCL patients and by
immunohistochemistry in concurrent lymph node and skin
biopsy specimens in 17 additional TCL cases, mostly mycosis fungoides. Tumor cell expression was correlated with
patterns of expression in nonneoplastic lymphocytes in 14
reactive lymph node and 10 skin samples showing chronic
dermatitis. Expression of CD25 and CD30 in all biopsy
samples was compared with that of cutaneous lymphocyte
antigen (CLA), a mediator of skin homing.
Results: By flow cytometry, we noted significantly decreased expression of CD25 in lymph node compared with
peripheral blood in 8 of 13 TCLs, with no changes in CD30
levels in 4 cases studied. Using immunohistochemistry,
CD25 was strongly expressed in epidermotropic tumor cells
in 13 of 17 (76%) TCL skin specimens but was decreased in
the corresponding lymph node in 12 of these cases. CD30
was expressed at roughly equal intensity in tumor cells from
both sites, except in 1 case. CLA showed a similar pattern to
CD25, being expressed by tumor cells in 16 of 17 (94%) skin
specimens, but was largely absent in tumor cells in the
corresponding lymph node in 12 of these patients. In T cells
from reactive lymph node biopsy specimens, CD25 was
highly expressed only in dermatopathic lymphadenitis associated with transient skin rashes.
Conclusions: We demonstrate in vivo that decreased
levels of CD25 expression occur in TCL when it involves
lymph node, similar to what is seen with CLA. This demon-
strable variation related to anatomical localization has implications for the measurement of surface expression of
CD25 and for understanding the response of patients with
cutaneous TCL to interleukin 2 receptor-targeted immunotherapy.
INTRODUCTION
Lymphoma cells express a variety of cytokine receptors
and signaling molecules that are current or potential targets for
immunomodulatory therapy. One such agent is denileukin diftitox/ONTAK (DAB389IL-2), a recombinant diphtheria toxininterleukin 2 (IL-2) fusion protein, which targets the IL-2 receptor. This receptor has several isoforms with the high-affinity
complex comprised of three subunits, including ␣ chain (CD25),
␤ chain (CD122), and ␥-chain (CD132). DAB389IL-2, which
preferentially targets lymphocytes bearing the high-affinity IL-2
receptor (1), has been used for treatment of mycosis fungoides
(MF)/Sezary syndrome (SS) (2– 4) and is in clinical trials for
other T-cell and B-cell lymphoma types (5, 6). Selection of
suitable patients for therapy often includes pretreatment assessment of CD25 expression in tumor cells or measurement of
serum IL-2 receptor levels (7). Antibodies and chimeric ligands
targeting the tumor necrosis family receptor CD30 are also
under active investigation for lymphoma treatment (8 –10).
However, large studies comparing expression of either CD25 or
CD30 on lymphoma cells at different tissue sites in vivo have
not yet been reported.
We compare here the immunophenotype of T-cell tumors,
mostly MF/SS, simultaneously sampled at different sites for
CD25 and CD30 and compare them to expression of cutaneous
lymphocyte antigen (CLA), a labile surface adhesion molecule
that mediates skin homing. We show that expression of CD25,
as with CLA, is highly variable in most cutaneous T-cell lymphomas (CTCLs), being expressed most intensely by lymphoma
cells in the epidermis of skin and peripheral blood (PB), with
moderate intensity by tumor cells within the dermis and lymph
node sinusoids and lowest intensity by tumor cells in the parenchyma of lymph node. Modulation of CD25 expression in vivo
by microenvironmental influences has important consequences
for assessment of CD25 expression and in selecting appropriate
patients for immunomodulatory therapy.
MATERIALS AND METHODS
Received 5/12/03; revised 11/12/03; accepted 4/1/04.
The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
advertisement in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
Requests for reprints: Dan Jones, Department of Hematopathology, The
University of Texas M. D. Anderson Cancer Center, Box 72, 1515 Holcombe Boulevard, Houston, TX 77030. Phone: (713) 745-2598; Fax:
(713) 745-0736; E-mail: [email protected].
Criteria for Diagnosis and Inclusion. All cases were
drawn from the hematopathology service at The University of
Texas M. D. Anderson Cancer Center (Houston, TX). Thirteen
paired PB and lymph node aspirate samples from patients with
MF/SS were assessed by flow cytometry, with samples from the
two sites taken within 7 days of each other. Three of these
patients were untreated, and the rest had received a variety of
agents (median of two different therapies). No patients had
received denileukin diftitox at time of sample analyzed. In all
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
5588 Modulation of CD25
cases, the cytomorphology of the tumor at each of the sites was
assessed by Giemsa or Pap-stained smears and by H&E-stained
paraffin-embedded sections of biopsies and cell-block cytologic
preparations when available. Morphology was graded as small
cell/cerebriform, mixed small and large nucleolated forms, or
predominantly large cells.
We also studied 17 patients with TCL who had archival
paraffin tissue blocks of both skin and lymph node biopsy
specimens taken within 30 days of each other. Four patients
were untreated; the rest had received a variety of treatments
(median of three prior agents). None had received denileukin
diftitox at time of biopsy. In 16 patients, the lymph node
sampled was draining the area from which the skin biopsy was
taken. In the remaining case, the lymph node was obtained from
a distant site. All biopsy specimens that showed evidence of
histological transformation in the skin biopsy also showed transformation in the corresponding lymph node biopsy specimen.
The clonal identity of the tumor, at both sites, was determined
using previously described multicolor PCR analysis methods
(11). In 12 cases, identical monoclonal T-cell receptor ␥ rearrangements were identified by PCR analysis in skin and lymph
node, in 3 cases, different amplified products were present
between the two sites, and in 2 cases, insufficient material was
available for comparison.
Diagnoses were low-grade MF/SS in 5 patients, large-cell
transformation of MF/SS in 8 patients, human T-cell lymphotropic virus I-positive adult T-cell leukemia/lymphoma (ATLL)
in 1 patient, and CTCL unspecified type in 3 patients. A diagnosis of MF required the presence of epidermotropism and
compatible clinical findings, with SS diagnosed in those patients
with tumor involvement of PB. Transformed MF was diagnosed
in cases in which large cells comprised ⱖ25% of the tumor
cellularity, as described previously (12, 13).
For comparison of staining patterns, we studied nonneoplastic lymph nodes that showed different reactive patterns of
expansion obtained from 20 patients. These included biopsy
specimens showing marked follicular hyperplasia in 4 patients,
progressive transformation of germinal centers in 7 patients,
follicle lysis in 2 patients, paracortical immunoblastic reaction
in 3 patients, and dermatopathic lymphadenitis in 4 patients. The
latter diagnosis required the presence of an interfollicular to
diffuse lymphohistiocytic proliferation with numerous dendritic
cells containing melanin pigment. All dermatopathic lymphadenitis cases occurred in lymph nodes of patients with reactive
skin rashes and were negative for monoclonal T-cell receptor ␥
chain rearrangements by PCR (11). Seven cases of chronic
spongiotic dermatitis and 3 cases of psoriasis were also included.
Flow Cytometric Analysis. We used a standard T-cell
antigen panel to assess PB and lymph node aspirate samples
from 10 MF/SS patients, as described previously (14). Two-,
three-, or four-color flow cytometric analysis was performed on
all samples, following a standard red-cell lysis method for PB,
using mouse monoclonal antibodies directed against CD45
[clone 2D1, peridin-chlorophyll-a-protein conjugated], as well
as various combinations of CD3 [SK7, FITC conjugated, allophycocyanin conjugated, or phycoerythrin (PE) conjugated],
CD4 (SK3, PE), CD7 (4H9, FITC), CD8 (SK1, FITC or PE),
CD19 (SJ25C1, FITC), CD25 (2A3, PE), CD26 (L272, FITC),
and CD16/CD56 (B73.1/NCAM16.2, PE). All antibodies were
from BD Biosciences (San Jose, CA), and analysis was performed using FACScan or FACScaliber cytometers.
The degree of CD25 expression in the T-cell tumor population was assessed using cluster analysis to distinguish tumor
from nonneoplastic T cells. A tumor cluster was defined as a
discrete cluster of analyzed events having altered levels of
expression of one or more pan T-cell antigens relative to an
internal, immunophenotypically normal T-cell population. The
normal pattern of marker expression on nonneoplastic T cells
was previously studied in a large sampling of PB samples (14,
15). The range of immunophenotypic aberrancies encountered
in these TCL cases has been summarized previously (14). We
have previously shown a high correlation between the presence
of an abnormal T-cell population identified by this flow cytometry panel and the number of morphologically identifiable tumor
cells on examination of the PB smear (15). For each antibody,
negative staining levels were set by comparison to an isotypematched control.
Immunohistochemistry. Biopsy specimens were fixed
in 10% neutral-buffered formalin, embedded in paraffin, and
processed to produce 5-␮m sections. Immunostaining was performed with mouse monoclonal antibodies directed against
CLA (CLA-HECA452; PharMingen, San Diego, CA), CD25
(4C9; Novocastra), and CD30 (BerH2; Dako, Carpinteria, CA),
as described previously (12). CD4 (1F6; Novocastra, Peterborough, United Kingdom) and CD7 (BC-272) immunostaining
was also done in most cases to aid in identification of tumor
cells. Immunostaining for CD1a (235; Novocastra) and S-100
(rabbit anticow polyclonal; Dako) was done in dermatopathic
lymphadenitis to confirm the identity of dendritic cells. In a
subset of cases, the lymphatic endothelium marker, D2-40 (Signet, Dedham, MA), was used to highlight lymphocytes within
intranodal lymphatics. Antigen retrieval was performed in a
microwave with 10 mM citrate buffer (for CLA) or Dako target
retrieval buffer (for CD25 and CD30). Staining was detected
using biotin-avidin-peroxidase-conjugated reagents (LSAB⫹
kit; Dako) and diaminobenzidine or 3-amino-9-ethylcarbazole
as the chromogenic substrate.
Percentages of immunoreactivity in cytologically identified
tumor cells were assessed by counting three ⫻40 high-power
fields and averaging the results. Staining pattern was correlated
with tumor cell localization in lymph node (sinusoidal, subcapsular, interfollicular, clustered, and diffuse) and skin (epidermal
and dermal). The Kruskal-Wallis test and Spearman’s rank
correlation test were used to correlate expression of markers.
Statistical significance was defined as ␣ ⫽ 0.05.
RESULTS
Comparison of MF/SS in PB and Lymph Node
By flow cytometry, we examined PB and lymph node
aspirate samples obtained within 7 days of each other from 13
patients with MF/SS. Eleven of 13 (85%) of the analyzed
MF/SS patients showed ⱖ5% CD25⫹ tumor cells in PB (Table
1). In four of these cases, there were reduced numbers of
CD25⫹ tumor cells in the lymph node aspirate, with four cases
showing a ⱖ20% decrease in the number of CD25⫹ tumor
cells. In the four cases with ⱖ5% CD30⫹ tumor cells in PB
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
Clinical Cancer Research 5589
Table 1 Comparison of CD25 expression levels in Sezary syndrome
tumor cells from simultaneously sampled PB and lymph sitesa
Sample no.
PB
% positive tumor
(tumor cell size)
LN
% positive tumor
(tumor cell size)
Change
PB to LN
1
2
3
4
5
6
7
8
9
10
11
12
13
97 (small)
87 (mixed)
57 (large)
20 (mixed)
21 (mixed)
27 (mixed)
18 (small)
15 (mixed)
2 (mixed)
29 (large)
4 (small)
39 (mixed)
21 (mixed)
0 (small)
7 (mixed)
29 (large)
0 (large)
3 (large)
11 (mixed)
9 (mixed)
8 (large)
4 (mixed)
32 (large)
9 (small)
47 (mixed)
32 (mixed)
-97
-80
-28
-20
-18
-16
-9
-7
⫹2
⫹3
⫹5
⫹8
⫹11
a
PB, peripheral blood; LN, lymph node.
(range, 5–9%), similar numbers of CD30⫹ tumor cells were
also present in lymph node aspirate specimens.
Fig. 1 shows a comparison of tumor cells sampled from PB
and lymph node within a 24-h period. Tumor cells at both sites
had a CD3-dim- and CD26-negative immunophenotype that
allowed separation from the (likely) nonneoplastic T cells (Fig.
1, left and middle panels). Nearly all of the tumor cells in PB
were uniformly positive for CD25 (Fig. 1, top left panel),
whereas tumor cells in a sampled inguinal lymph node were
completely negative for CD25 (Fig. 1, bottom left panel). Comparison of a PB smear and the lymph node aspirate smears
revealed similarly small to intermediate-sized cerebriform lymphocytes (Fig. 1, right panels). No other immunophenotypic
variations were seen between the two populations with the
tested antigens (i.e., CD3, CD4, CD5, CD7, CD8, CD19, CD26,
CD56/CD16, and T-cell receptor).
Comparison of TCL Immunophenotype in Skin and
Lymph Node
We also studied 17 additional cases of TCL simultaneously
involving lymph node and skin, including 8 cases of transformed MF/SS, 5 cases of lower grade MF, 1 case of ATLL, and
3 cases of CTCL, unspecified type. In two of the transformed
MF cases, there was a greater percentage of large tumor cells in
the lymph node as compared with the skin with a greater
percentage of large tumor cells seen in skin in a third case.
Otherwise, the morphological spectrum of tumor cells was similar between the two sites. Patients included 10 women and 7
men whose mean age at time of biopsy was 55 years (range,
26 – 80 years). Lymph node biopsies were done, on average, 15
days from the skin biopsy (range, 0 –28 days). Table 2 summarizes the patterns of CLA, CD25, and CD30 immunostaining of
tumor cells in skin and lymph node, and Fig. 2 illustrates a
representative case.
Fig. 1 Comparison of CD25 expression in mycosis fungoides/Sezary syndrome cells sampled from lymph node and peripheral blood (PB). Upper
panels, analysis of PB from a patient with mycosis fungoides/Sezary syndrome reveals a discrete population of CD3-dim/CD25⫹ tumor cells (boxed)
that can be distinguished from variable CD25 positivity of nonneoplastic T cells. This CD3-dim population is also uniformly CD26-negative consistent
with a neoplastic clone. PB smear demonstrates a prominent enlarged cerebriform tumor population. Lower panels, analysis of a fine needle aspirate
from an enlarged lymph node done the same day revealed a uniform population of CD3-dim tumor cells that were completely negative for CD25.
As in PB, this CD3-dim tumor cell population was also uniformly negative for CD26. Examination of the Pap-stained smears from lymph node
revealed a homogeneous population of cerebriform tumor cells identical to those in the PB. This case is sample no. 1 in Table 1.
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
5590 Modulation of CD25
Table 2
Summary of immunohistochemical comparison of CTCL in skin and lymph node biopsy specimensa
CD25 (diagnosis)
CLA
CD30
Staining pattern in skin
13/17 cases (76%)b
Increased in epidermis
16/17 (94%) casesc
(Strong/uniform in 9)
5/17 cases (29%)d
Mostly in dermal component
Immunostaining decreased in
lymph node
12 cases (92%)e
(5 MFt, 4 MF, 2 TCL, 1 ATLL)
12 cases (75%)f
(5 MFt, 3 MF, 3 TCL 1 ATLL)
1 (MFt)
Expressed, no changer: 4 (MFt)
No change in immunostaining
level
Not expressed, either site: 4
(1 MF, 2 MFt, 1 TCL)
Expressed, no change: 4
(2 MFt, 2 MF)
Not expressed, 1 (MFt)
Expressed, no change: 4 (MFt)
Not expressed: 12
(3 MFt, 5 MF, 3 TCL, 1 ATLL)
Immunostaining increased in
lymph node
1 case (MFt)
None
None
a
CTCL, cutaneous T-cell lymphoma; CLA, cutaneous lymphocytes antigen; MFt, transformed mycosis fungoides; TCL, cutaneous TCL,
unspecified type; ATLL, adult T-cell leukemia/lymphoma.
b
CD25 positivity defined as ⱖ5% tumor cells positive for CD25.
c
Defined as 20% or more CLA⫹ tumor cells.
d
CD30 positivity defined as ⱖ5% CD30⫹ tumor cells.
e
CD25 staining was largely restricted to sinuses in 4 and to sinus and directly subcapsular areas in another 4.
f
CLA staining was largely restricted to sinuses in 6 and to sinus and directly subcapsular areas in another 4.
CD25 Expression
CD25 was expressed in ⱖ5% of cytologically atypical
lymphocytes in skin in 13 of 17 (76%) cases, with a median of
20% CD25⫹ tumor cells across all cases. In seven of these
cases, CD25 was more highly expressed in the intraepidermal
tumor component than in the tumor cells in the dermis. In 12 of
these CD25⫹ tumors (92%), the number of CD25⫹ tumor cells
was decreased in lymph node as compared with skin (Fig. 2, B
and E). There was often a staining intensity gradient in lymph
node with the strongest CD25⫹ staining located within the
sinusoidal and subcapsular components. Using the antibody
D2– 40 to highlight intranodal lymphatics, we noted that in four
cases (2 transformed MF, 1 ATLL, and 1 CTCL), the only
strongly CD25⫹ tumor cells were present in lymph node sinusoids (Fig. 2H and data not shown). In another four cases, the
CD25⫹ tumor cells were located primarily in the subcapsular
region in association with CLA⫹/S100⫹ dendritic cells. In 4 of
17 cases (24%), CD25 was expressed by ⬍5% of tumor cells in
skin. In three of these cases, tumor cells in lymph node were
also negative for CD25. However, in one case, intrasinusoidal
tumor cells were strongly CD25⫹.
CD30 Expression
CD30 was expressed in ⬎5% of tumor cells in skin of 5 of
17 (29%) cases (range, 5–50% of CD30⫹ tumor cells; median,
15%). In 12 (71%) skin cases, CD30 was not expressed or was
positive in only rare tumor cells. CD30 expression was similar
between skin and lymph node samples, with the number of
positive cells within 5% of each other in 16 of 17 cases (94%).
The one exception was a transformed MF/SS case, which
showed 50% CD30⫹ tumor cells in skin and only rare CD30⫹
cells in lymph node.
CLA Expression
Tumor cells in skin strongly expressed CLA in 9 cases,
variably expressed CLA in 7 cases (Fig. 2, C and I), and were
completely negative for CLA in a single transformed MF case.
CLA was more strongly expressed in low-grade MF and expressed at lower levels in transformed MF. Similar to the results
with CD25, CLA staining of tumor cells in lymph node was
decreased in 12 of 16 cases as compared with staining levels in
skin. As with CD25, the strongest CLA⫹ tumor cells were
frequently in the sinusoids or subcapsular areas of lymph node
(Fig. 2F and data not shown). In cases with numerous CLA⫹
dendritic cells, it was often difficult to distinguish their strong
positivity from the weak positivity of adjacent tumor cells.
By pairwise rankings, expression of CLA, CD25, and
CD30 were not significantly correlated. There was, however, a
statistically significant association between those cases with the
largest decreases in CLA staining in lymph node as compared
with skin and those tumors that had the most variable CD25
expression (r ⫽ 0.61). CLA expression was also inversely
correlated with CD30 staining levels (r ⫽ ⫺0.40). Besides the
association of CD30 and decreased CLA expression with transformed MF, there was no significant association of CD25 or
CLA modulation with specific diagnostic categories, although
the limited numbers of cases in some categories precluded
meaningful subset analysis.
CD25 and CD30 Expression in Nonneoplastic
Lymphocytes in Skin and Lymph Node
Nonneoplastic Lymph Nodes. We correlated the levels
of CD25 and CD30 expression in T cells in nonneoplastic
tissues. In cases showing follicular hyperplasia or immunoblastic paracortical expansion, CD25 was highly expressed
only in scattered interfollicular clusters of histiocytes and
lymphocytes and in rare germinal center cells, overall comprising much less than 1% of lymphocytes (Fig. 3A). In cases
with progressively transformed germinal centers, focal increases in CD25⫹ cells were present in association with
transformed follicles, marking up to 5% of lymphocytes in
these areas (data not shown). In four cases of dermatopathic
lymphadenitis, there was marked expression of CD25 by
interfollicular T cells (up to 30% of lymphocytes) and weaker
expression by dendritic cells at various stages of maturation
(Fig. 3B). These cases all occurred in lymph nodes of patients’ draining transient severe skin rashes and, in contrast to
lymphomatous involvement, showed expansion of only the
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
Clinical Cancer Research 5591
Fig. 2 Modulation of CD25
and cutaneous lymphocyte antigen (CLA) expression in human T-cell lymphotropic virus
I-positive adult T-cell lymphoma/leukemia. A, adult T-cell
lymphoma/leukemia in skin
shows a perivascular and epidermotropic infiltrate of tumor
cells. B, CD25 is strongly expressed in 100% of tumor cells
in skin. C, tumor cells in skin
show variable strong reactivity
for CLA. D, an enlarged lymph
node biopsied 1 day later
shows an interfollicular infiltration of tumor cells surrounding and encroaching on a germinal center (GC). E, CD25 is
strongly expressed only in sinusoidal tumor and is expressed weakly in the subcapsular location, consistent with
progressive antigen loss upon
migration into the nodal parenchyma. F, CLA is strongly positive only in tumor cells within
the subcapsular sinuses. G,
higher magnification of indicated areas in (E) shows pleomorphic cerebriform tumor
cells identical to those in skin.
H, higher magnification of F
shows CD25⫹ intrasinusoidal
tumor cells adjacent to CD25negative tumor cells within
lymph node. I, higher magnification of F shows CLA-negative tumor in deep paracortex
admixed with CLA⫹ high endothelial venules.
interfollicular nodal regions with no encroachment on the
follicular compartment (Fig. 3C). The density of CD25⫹ T
cells in dermatopathic lymphadenitis was related to the number of CD1a⫹/S100⫹ dendritic cells, with cases showing
only focal melanin deposition having the lowest numbers of
CD25⫹ lymphocytes (Fig. 3D).
CD30 in all nonneoplastic lymph nodes was expressed in
a small number (less than 1%) of intrafollicular lymphocytes,
most with immunoblastic morphology. CD30 was also
weakly expressed in some plasma cells. In all nonneoplastic
lymph nodes, CLA was expressed in less than 1% of interfollicular lymphocytes, CLA was highly expressed in highendothelial venules as well as a small subset of migrating
dendritic cells and clusters of plasmacytoid monocytes, located primarily around HEV (not shown). In 4 lymph nodes
with benign dermatopathic lymphadenitis, CLA was variably
and weakly expressed in numerous dendritic cells, particularly in subcapsular areas.
Skin Biopsies with Inflammatory Infiltrates. We studied seven skin biopsy specimens from patients with chronic
dermatitis. CD25 positivity in lymphocytes ranged from 5 to
20%, with the scant lymphocytes in the epidermis more
strongly positive for CD25 than the dermal lymphocytes.
CD25 was highly expressed in 30 –50% of the lymphocytes in
three psoriasis biopsy specimens in both dermis and epidermis. In both types of lesions, CD25 expression was characteristic of larger, activated-appearing lymphocytes (data not
shown). CD30 expression was not detected in lymphocytes
in any of the biopsy specimens. In all chronic dermatitis and psoriasis samples, the majority of infiltrating lymphocytes in the dermis and epidermis expressed CLA (data not shown).
DISCUSSION
We demonstrate here that IL-2 receptor-␣/CD25 is expressed in TCL cells at varying levels in simultaneous samples of different tissue sites. Among the 13 CD25⫹ TCLs in
skin studied by immunohistochemistry, there was decreased
expression by the tumor cells in lymph nodes in 12 (92%). In
13 additional patients with TCL studied by flow cytometry, 8
(61%) had decreased CD25 expression in the tumor cells
from lymph node samples, as compared with those in PB. We
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
5592 Modulation of CD25
Fig. 3 CD25 expression in T cells in benign dermatopathic lymphadenitis. A, CD25 expression in a lymph node with follicular hyperplasia is limited
to rare scattered interfollicular lymphocytes (arrows). B, strong CD25 expression in T cells is seen in benign dermatopathic lymphadenitis in a patient
with a severe skin rash. Lymphocytes are strongly positive for CD25, whereas dendritic cells more weakly express CD25 (inset). C, unlike cutaneous
T-cell lymphoma, the benign dermatopathic infiltrate expands the interfollicular regions, leaving the follicular compartment intact. D, a case of milder
dermatopathic lymphadenitis shows fewer CD25⫹ T cells.
observed lower expression of CD25 in tumor cells from
lymph node samples even in a human T-cell lymphotropic
virus I-positive ATLL patient who had uniformly strongly
CD25⫹ tumor cells in skin. Among cases of reactive lymphadenitis, high CD25 expression by T cells was associated
with reactive dermatopathic lymphadenitis, occurring in
lymph nodes that are draining sites of transient skin rashes,
and suggesting a transient pattern of CD25 expression by
lymphocytes recently entering the lymph node.
The pattern of immunostaining observed for CLA was
similar to CD25. In the 16 cases of TCL with CLA⫹ tumor cells
in skin, 12 cases (75%) showed decreased expression of CLA in
the lymph node tumor cells. This could not be readily attributed
to morphological changes because the cytomorphology of
lymph node tumor cells was similar in all but three cases. This
pattern of CLA staining in tumor cells was paralleled by the
observed pattern of CLA expression in lymph node-dendritic
cells. These cells were frequently strongly CLA⫹ in the sinuses
and in areas directly adjacent but were much more weakly
positive or negative in deeper paracortical areas. We observed a
reproducible microanatomical pattern of CD25 and CLA expression by tumor cells within lymph node, namely high in
sinusoidal and subcapsular locations with loss in the deeper
paracortex. This gradient of decreased CD25 and CLA expression in lymph node suggests both proteins can be modulated on
tumor lymphocytes during their passage through the lymph
node, although the presence of multiple immunophenotypically
stable or genetically distinct tumor populations cannot be completely excluded.
In nonneoplastic T cells, regulation of the CD25/IL-2
receptor-␣ gene is complex, with established roles for CD25 in
early thymic T-cell maturation and as an activation marker in
mature T cells. CD25 is an essential component of the threechain high-affinity IL-2-receptor; however, IL-2 signaling can
still occur through the medium-affinity CD122/CD132 complex
(16). We show here that wide variations in the levels of CD25
occur in different benign lymph node conditions and inflamma-
tory dermatoses (Fig. 3). Recently, there has been great emphasis on the role of bright CD25⫹ CD4⫹ T cells as mediators of
the regulatory/suppressor T-cell response, with CD25 loss leading to overexpansion of normal T-cell populations (16, 17). Our
findings suggest that bright CD25⫹ CD4⫹ tumor cells isolated
from PB or seen in the lymphatic sinuses may have the capacity
to become negative for CD25 once they enter the lymph node
parenchyma. This suggests that CD25 levels on normal T cells
may also be rapidly regulated because they appear to be on
CTCL cells and might not be a stable feature of regulatory T
cells. Indeed, bright CD25⫹ T cells can be isolated in great
numbers from lymph nodes that are draining sites of inflammation (18, 19) but were seen in much lower numbers in most of
the reactive lymph node conditions analyzed here. Strong CD25
expression on CD4⫹ T cells within the lymph node itself was
common only in dermatopathic lymphadenitis.
We demonstrate here, for the first time, differential expression of CD25 in vivo between TCL tumor cells in skin, lymph
node, and PB. This differential expression occurred even in a
case of human T-cell lymphotropic virus I-positive ATLL,
which is commonly regarded as uniformly, strongly positive for
CD25 (20). This is of interest because CD25 is currently a
therapeutic target in T-cell tumors, using agents such as
DAB389IL-2 (21). Our results demonstrate that the expression
level of CD25 by tumor cells measured as a possible indicator of
therapeutic response will depend on the site sampled. Thus,
overall systemic measures of IL-2 receptor levels, such as by
soluble serum assay, may give a more accurate median expression level and better predictor of therapeutic response (7). Our
results also suggest that drugs targeting CD25 may be more
effective in treating skin disease than tumors with predominantly lymph node involvement. Previous studies measuring
IL-2 receptor levels in serum have also suggested lower systemic levels of IL-2 receptor in patients with nodal disease (22).
However, given the multiple topical and systemic treatments that
most of the patients studied here had undergone, it is possible that
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
Clinical Cancer Research 5593
modulations in CD25 expression may be affected by certain treatments more than others.
We also demonstrated parallel patterns of expression of
CD25 and CLA in most CTCL cases. CLA is expressed by the
vast majority of skin-homing T cells as well as by a small subset
of circulating benign lymphocytes, especially in patients with
chronic dermatitis (23–26). CLA is a modified selectin ligand
that mediates binding of a subset of circulating lymphocytes to
dermal endothelium (27, 28). We have previously demonstrated
that CLA is highly expressed in tumor cells in skin at all stages
of CTCL (12). CLA can also be expressed in MF/SS in PB, but
analysis of sorted cell populations has demonstrated that circulating
tumor cells are present in both the CLA⫹ and CLA-negative T-cell
populations (29). This suggests that CTCL may shed or rapidly
modulate CLA once outside of the skin microenvironment as has
been demonstrated for nonneoplastic T cells (30, 31).
The differential expression of CLA and CD25 is contrasted
with the relatively uniform pattern of CD30 expression by tumor
cells at different sites observed here. Only 1 of 17 cases showed
substantial differences in the expression of CD30 between
lymph node and skin. This suggests that induction of CD30
expression may be a more stable, likely transformation-related
property of the tumor cells, rather than being more transiently
influenced by tissue microenvironment. We have previously
observed that CD30 up-regulation in transformation of MF is
associated with many other phenotypic changes, including expression of cytotoxic proteins (e.g., perforin) and chemokine
receptors (e.g., loss of CXCR3) that would be consistent with a
stable change in tumor phenotype (12).
However, the majority of T-cell tumors do show tissuespecific variations in CD25 and CLA levels in a manner resembling patterns of antigen modulation in nonneoplastic T cells.
Therefore, it may be useful to define those markers used for
targeted therapies as being either relatively labile (as with
CD25) or as being relatively stably expressed by tumors (such
as CD30). Because the pattern of antigen expression in normal
cells may not always directly mimic the pattern observed in
tumors, microenvironmental influences need to be directly confirmed on involved tumor samples from different sites. For these
reasons, we believe that for lymphoid tumors, in vivo monitoring of the microenvironmental regulation of a therapeutic
marker is an important component of understanding the consequences of targeting that antigen.
ACKNOWLEDGMENTS
We thank Wenhua Lang and Thomas Brooks for assistance with
immunostaining and Linda Powers and the staff of the M. D. Anderson
Immunology Laboratory for flow cytometric studies.
REFERENCES
1. Shao RH, Tian X, Gorgun G, Urbano AG, Foss FM. Arginine
butyrate increases the cytotoxicity of DAB(389)IL-2 in leukemia and
lymphoma cells by up-regulation of IL-2Rbeta gene. Leuk Res 2002;
26:1077– 83.
2. Foss FM, Borkowski TA, Gilliom M, et al. Chimeric fusion protein
toxin DAB486IL-2 in advanced mycosis fungoides and the Sezary
syndrome: correlation of activity and interleukin-2 receptor expression
in a Phase II study. Blood 1994;84:1765–74.
3. Olsen E, Duvic M, Frankel A, et al. Pivotal Phase III trial of two dose
levels of denileukin diftitox for the treatment of cutaneous T-cell lymphoma. J Clin Oncol 2001;19:376 – 88.
4. Duvic M. Bexarotene and DAB(389)IL-2 (denileukin diftitox, ONTAK) in treatment of cutaneous T-cell lymphomas: algorithms. Clin
Lymphoma 2000;1(Suppl 1): S51–5.
5. Talpur R, Apisarnthanarax N, Ward S, Duvic M. Treatment of
refractory peripheral T-cell lymphoma with denileukin diftitox (ONTAK). Leuk Lymphoma 2002;43:121– 6.
6. Zhang M, Zhang Z, Garmestani K, et al. Pretarget radiotherapy with
an anti-CD25 antibody-streptavidin fusion protein was effective in therapy of leukemia/lymphoma xenografts. Proc Natl Acad Sci USA 2003;
100:1891–5.
7. Wasik MA, Vonderheid EC, Bigler RD, et al. Increased serum concentration of the soluble interleukin-2 receptor in cutaneous T-cell lymphoma.
Clinical and prognostic implications. Arch Dermatol 1996;132:42–7.
8. Hombach A, Muche JM, Gerken M, et al. T cells engrafted with a
recombinant anti-CD30 receptor target autologous CD30(⫹) cutaneous
lymphoma cells. Gene Ther 2001;8:891–5.
9. Huhn M, Sasse S, Tur MK, et al. Human angiogenin fused to human
CD30 ligand (Ang-CD30L) exhibits specific cytotoxicity against CD30positive lymphoma. Cancer Res 2001;61:8737– 42.
10. Borchmann P, Schnell R, Fuss I, et al. Phase I trial of the novel
bispecific molecule H22xKi-4 in patients with refractory Hodgkin lymphoma. Blood 2002;100:3101–7.
11. Vega F, Medeiros LJ, Jones D, et al. A novel four-color PCR assay
to assess T-cell receptor gamma gene rearrangements in lymphoproliferative lesions. Am J Clin Pathol 2001;116:17–24.
12. Lu D, Duvic M, Medeiros LJ, Luthra R, Dorfman DM, Jones D. The
T-cell chemokine receptor CXCR3 is expressed highly in low-grade
mycosis fungoides. Am J Clin Pathol 2001;115:413–21.
13. Diamandidou E, Colome-Grimmer M, Fayad L, Duvic M, Kurzrock
R. Transformation of mycosis fungoides/Sezary syndrome: clinical
characteristics and prognosis. Blood 1998;92:1150 –9.
14. Washington LT, Huh YO, Powers LC, Duvic M, Jones D. A stable
aberrant immunophenotype characterizes nearly all cases of cutaneous
T-cell lymphoma in blood and can be used to monitor response to
therapy. BMC Clin Pathol 2002;2:5.
15. Jones D, Dang NH, Duvic M, Washington LT, Huh YO. Absence of
CD26 expression is a useful marker for diagnosis of T-cell lymphoma in
peripheral blood. Am J Clin Pathol 2001;115:885–92.
16. Papiernik M. Natural CD4⫹ CD25⫹ regulatory T cells. Their
role in the control of superantigen responses. Immunol Rev 2001;
182:180 –9.
17. Almeida AR, Legrand N, Papiernik M, Freitas AA. Homeostasis of
peripheral CD4⫹ T cells: IL-2R alpha and IL-2 shape a population of
regulatory cells that controls CD4⫹ T cell numbers. J Immunol 2002;
169:4850 – 60.
18. Green EA, Choi Y, Flavell RA. Pancreatic lymph node-derived
CD4(⫹)CD25(⫹) Treg cells: highly potent regulators of diabetes that
require TRANCE-RANK signals. Immunity 2002;16:183–91.
19. McNally JM, Dempsey D, Wolcott RM, Chervenak R, Jennings
SR. Phenotypic identification of antigen-dependent and antigenindependent CD8 CTL precursors in the draining lymph node during
acute cutaneous herpes simplex virus type 1 infection. J Immunol
1999;163:675– 81.
20. Dahmoush L, Hijazi Y, Barnes E, Stetler-Stevenson M, Abati A.
Adult T-cell leukemia/lymphoma: a cytopathologic, immunocytochemical, and flow cytometric study. Cancer (Phila.) 2002;96:110 – 6.
21. Foss FM. DAB(389)IL-2 (ONTAK): a novel fusion toxin therapy
for lymphoma. Clin Lymphoma 2000;1:110 – 6; discussion 117.
22. Vonderheid EC, Zhang Q, Lessin SR, et al. Use of serum soluble
interleukin-2 receptor levels to monitor the progression of cutaneous
T-cell lymphoma. J Am Acad Dermatol 1998;38:207–20.
23. Dworzak MN, Froschl G, Printz D, et al. Skin-associated lymphocytes
in the peripheral blood of patients with atopic dermatitis: signs of subset
expansion and stimulation. J Allergy Clin Immunol 1999;103:901– 6.
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
5594 Modulation of CD25
24. Picker LJ, Michie SA, Rott LS, Butcher EC. A unique phenotype of
skin-associated lymphocytes in humans. Preferential expression of the
HECA-452 epitope by benign and malignant T cells at cutaneous sites.
Am J Pathol 1990;136:1053– 68.
25. Picker LJ, Treer JR, Ferguson-Darnell B, Collins PA, Bergstresser
PR, Terstappen LW. Control of lymphocyte recirculation in man. II.
Differential regulation of the cutaneous lymphocyte-associated antigen,
a tissue-selective homing receptor for skin-homing T cells J Immunol
1993;150:1122–36.
26. Santamaria Babi LF, Picker LJ, Perez-Soler MT, et al. Circulating allergen-reactive T cells from patients with atopic dermatitis and
allergic contact dermatitis express the skin-selective homing receptor, the cutaneous lymphocyte-associated antigen. J Exp Med 1995;
181:1935– 40.
27. Santamaria Babi LF, Moser R, Perez-Soler MT, Picker LJ, Blaser
K, Hauser C. Migration of skin-homing T cells across cytokine-activated human endothelial cell layers involves interaction of the cutaneous
lymphocyte-associated antigen (CLA), the very late antigen-4 (VLA-4),
and the lymphocyte function-associated antigen-1 (LFA-1). J Immunol
1995;154:1543–50.
28. Fuhlbrigge RC, Kieffer JD, Armerding D, Kupper TS. Cutaneous
lymphocyte antigen is a specialized form of PSGL-1 expressed on
skin-homing T cells. Nature (Lond.) 1997;389:978 – 81.
29. Borowitz MJ, Weidner A, Olsen EA, Picker LJ. Abnormalities of
circulating T-cell subpopulations in patients with cutaneous T-cell
lymphoma: cutaneous lymphocyte-associated antigen expression on
T cells correlates with extent of disease. Leukemia (Baltimore)
1993;7:859 – 63.
30. Campbell DJ, Butcher EC. Rapid acquisition of tissue-specific
homing phenotypes by CD4(⫹) T cells activated in cutaneous or mucosal lymphoid tissues. J Exp Med 2002;195:135– 41.
31. Armerding D, Kupper TS. Functional cutaneous lymphocyte antigen can be induced in essentially all peripheral blood T lymphocytes. Int
Arch Allergy Immunol 1999;119:212–22.
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.
Degree of CD25 Expression in T-Cell Lymphoma Is
Dependent on Tissue Site: Implications for Targeted Therapy
Dan Jones, Sherif Ibrahim, Kaushali Patel, et al.
Clin Cancer Res 2004;10:5587-5594.
Updated version
Cited articles
Citing articles
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://clincancerres.aacrjournals.org/content/10/16/5587
This article cites 29 articles, 12 of which you can access for free at:
http://clincancerres.aacrjournals.org/content/10/16/5587.full#ref-list-1
This article has been cited by 5 HighWire-hosted articles. Access the articles at:
http://clincancerres.aacrjournals.org/content/10/16/5587.full#related-urls
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from clincancerres.aacrjournals.org on June 16, 2017. © 2004 American Association for Cancer
Research.