in each of 10 cases of lymphomatoid papulosis.48 The. 2 large-cell lymphomas with RS-like cells were theonly. T-cell lymphomas in which the majority of tumor ...
Immunopathology
of
Cutaneous
TCell
Lymphomas
From the Departments of Laboratory Medicine, Pathology, and Dermatology, University of Washington, Seattle, Washington; Cedars-Sinai Medical Center, Los Angeles, California; and the Department of Dermatology, Temple University Health Sciences Center
KAORI NASU, MD, JONATHAN SAID, MD, ERIC VONDERHEID, MD, JOHN OLERUD, MD, DIANNE SAKO, BS, and MARSHALL KADIN, MD
In this study the authors attempted to establish immunopathologic criteria for the distinction of various Tcell lymphomas affecting the skin. We studied skin specimens from 27 patients with mycosis fungoides (MF) (n = 12), the Sezary syndrome (SS) (n = 6), adult T-cell leukemia (ATL) (n = 4), and nonepidermotropic T-cell lymphoma of large cell (n = 4) and lymphoblastic (n = 1) types. Identification of tumor cells in mixed cell populations and detection of weak expression of surface antigens by tumor cells was facilitated by immunoelectron microscopy. The mature helper T-cell phenotype (T11+ T3+ T4+) was found in 14 of 18 cases of MF/SS. One case of MF had a cytotoxic/suppressor (T4- T8+ 3A1+) phenotype; one with frequent blastic cells showed only weak expression of T4 antigen; 2 cases of SS were T11-. Tumor cells infiltrating the skin expressed 3AI antigen in 44%
and cellular activation antigens Ia and/or Tac in 78% of patients with MF/SS. No consistent phenotypic differences were found between ATL cells from ATLV (HTLV) antibody-positive patients and tumor cells of patients with MF/SS who lacked this antibody. In contrast, a group of nonepidermotropic T-cell lymphomas showed phenotypic differences from MF/SS and ATL in all but 1 case. These cases were distinguished by the frequent absence of T3, T4, and Leu 1 antigens in 3 large-cell lymphomas; frequent expression of Ki-1 antigen, a Hodgkin's diseaseassociated antigen, in 2 cases with RS-like cells; and an immature thymocyte phenotype in lymphoblastic lymphoma. These findings demonstrate that tumor-cell phenotypes can be useful in distinguishing different histologic types of cutaneous T-cell lymphoma. (AmJ Pathol 1985, 119:436-447)
WE INVESTIGATED the immunopathologic features of various T-cell lymphomas affecting the skin. We attempted to find differences that might be useful in the differential diagnosis and prognosis of these disorders. In particular, we hoped to establish whether immunopathologic criteria could be used to distinguish the skin lesions of patients with chronic T-cell lymphomas known as mycosis fungoides (MF) and the Sezary syndrome (SS) from skin lesions of patients with the subacute virus-associated adult T-cell leukemia (ATL). We previously reported that leukemic cells expressed Ia and 3A1 antigens in 2 of 3 nonendemic cases of ATL.1 Ia antigens have also been detected on leukemic cells of at least 200/o of Japanese ATL cases.2'3 However, Ia and 3A1 antigens were reported to be absent in circulating cells of patients with MF/SS.4 Therefore, we investigated whether the different expression of these antigens could facilitate the distinction of skin lesions of patients with ATL from those with MF/SS. We also compared the phenotype of these epidermotropic disorders with that of a group of nonepidermotropic cutaneous T-cell lymphomas. Our study was aided by the use of a combined
electron-microscopic-immunoperoxidase technique that was used to identify the type of cell expressing Tcell and Ta antigens. Previous studies did not distinguish whether these antigens were expressed by tumor cells with cerebriform nuclei or other (nonneoplastic) cells on a cell-to-cell basis. The excellent preservation of cell surface antigens and ultrastructure was achieved by fixation of skin specimens fixed in periodate-lysineparaformaldehyde (PLP). These specimens could be sent from great distances to our laboratory.
Materials and Methods From December 1982 to April 1984 we evaluated 27 patients with T-cell malignancies involving the skin.
Supported by Grant CH-254 from the American Cancer Society. Accepted for publication January 28, 1985. Address reprint requests to Dr. Marshall Kadin, Department of Pathology, Beth Israel Hospital, 330 Brookline Avenue, Boston, MA 02215.
436
IMMUNOPATHOLOGY OF CTCL
Vol. 119 * No. 3
Table 1-Clinical Data On Patients With Cutaneous T-Cell Lymphoma Sites Stage Age Sex Case Mycosis fungoides 1 76 2 70 52 3 4 50 80 5 48 6 7 58 37 8 9 30 62 10 11 45 12 73 Sezary syndrome 13 51 14 80 15 55 60 16 17 79 57 18 Adult T-cell leukemia 57 19 20 50 21 51 22 78 Nonepidermotropic cutaneous 23 52 24 74 25 75 26 73 27 22
437
Lesion type
IIB
Trunk, upper extremities Upper legs, buttocks Face, trunk, extremities Generalized Face Generalized Breast, buttock Breasts, trunk, buttocks, extremities Buttock Trunk, buttocks, legs Generalized Face, trunk, extremities
Papules and nodules Papules and plaques Erythematous nodules Ulcerated nodules Erythema Patches and plaques Ulcerated nodules Plaques Plaque Patches with poikiloderma Erythroderma with plaques Patches, plaques, nodules
III III III III IVA III
Generalized Generalized Generalized Generalized Generalized Abdomen
Erythroderma Erythroderma Erythroderma Erythroderma Erythroderma Recurrent patch
NA M M NA M NA NA M T-cell lymphoma M IVA M IVB F IIB M IIB M NA
Trunk Generalized Generalized Generalized
Papules Macules and patches Ulcerated plaques Plaques
Trunk, upper extremities Lower extremities, trunk Trunk, extremities Scalp, trunk, extremities Trunk, extremities
Papules, plaques, and nodules Plaques and nodules Erythematous nodules and patches Erythema, ulcerated nodules Nodules
M M M M M M F F M M M M
IVA IA IVA IVA IIA IA
M M M M M F
l1B IB IA IIA
Ill
NA, not applicable.
These patients included 12 individuals with MF, 6 with SS, 4 with ATL, and 5 with nonepidermotropic lymphoma, including 4 large-cell and 1 lymphoblastic lymphoma. The diagnosis was established by the usual clinical and histopathologic criteria without inclusion of immunologic data. Patients with MF and the SS had tumor cells with cerebriform nuclei infiltrating the epidermis (epidermotropism). Patients with clinical diagnosis of ATL were from Hawaii (Patient 19) or Japan (Patients 20-22). They had leukemic cells with more coarsely lobated nuclei and usually less epidermotropism than patients with MF or the SS. Patients with large-cell lymphoma had a predominance of large tumor cells, lacking cerebriform nuclei, but with pleomorphic, multilobated, or, in 2 cases (25 and 26) ReedSternberg-like cells. One patient had a lymphoblastic lymphoma in which cells with fine nuclear convolutions, delicate chromatin, small nucleoli, and frequent mitoses were found. Serum antibodies against human T-cell leukemia virus (HTLV) or adult T-cell leukemia virus (ATLV) were assayed by methods previously published.5 6 Antibodies against HTLV (ATLV) were detected only in the ATL group (Patients 19-22). NonATL patients with negative antibody screens were 3, 5, 11, 13, 14, 16-18, 24, and 26. Clinical information is
summarized in Table 1. Clinical staging is based on the TNM classification of cutaneous T-cell lymphoma.7 Four- or 6-mm punch biopsies of skin were obtained from all patients. Specimens from 11 patients were directly snap-frozen, and specimens from 6 patients (16, 19, and 20-23) were placed in Zeus fixative (an ammonium-sulfate-based fixative) (Zeus Scientific, Raritan, NJ) and sent to us before freezing. Specimens from the remaining 10 patients (2, 6, 8, 10, 11, 14, 15, 17, 18, and 24) were fixed in PLP fixative according to the method of McLean and Nakane8 for 4-6 hours at 4 C. PLP-fixed specimens were transferred through increasing concentrations of sucrose (7%To, 15%To, and 25%/o plus 10%o glycerol) in 0.05 M phosphate buffer (pH 7.4) and then snap-frozen. Each specimen was cut at approximately 6 , and placed in acetone at -20 C (for snapfrozen or Zeus-fixed material) or in PLP at 4 C (for PLP-fixed specimens) for 5 minutes. Cryostat sections were then washed in PBS for 15 minutes. Horse serum was used to block nonspecific binding of primary antibodies. Sections were then allowed to react with various murine monoclonal antibodies, listed in Table 2. Previous studies have shown that the antigens of interest are well preserved in both Zeus22 and PLP8 23 fixatives. In our experience material fixed in PLP gave the
438
NASU ET AL
AJP * June 1985
Table 2-Monoclonal Antibodies and Their Reactivity Antigen(s) Antibody 9.6 (OKT11)
E-rosette receptor
Leu 1 (10.2)
p45 p65
Leu 4 (OKT3) Leu 3a (OKT4) Leu 2a (OKT8)
9
T10 HB1Oa (7.2) HLB-3 OKT6 Ki-1
All E-rosetting T cells Thymocytes, all T cells except some cytotoxic/ suppressor T cells Mature thymocytes, peripheral T cells Thymocyte subpopulations, helper T cells Thymocyte subpopulations, cytotoxic/suppressor T cells Thymocytes, all T cells except some helper T cells Activated and functionally mature T cells Transferrin receptor on proliferating cells
10
p19 p62 p32-33 p70 (unreduced) p40 Interleukin-2 receptor p94 reduced p190 (unreduced) line p45 Framework region of la p29-34 p1 68-34 p49 unknown
3A1 Tac T9
Reactive cells
Reference
11
12 13 14 15 16
Thymocytes and activated T cells B cells, dendritic cells, macrophages, activated T cells Pan B (except plasma cell) Cortical thymocytes, Langerhans cells Hodgkin's (Reed-Sternberg) cells and unidentified smaller cells in lymph node, tonsil
17 18
19 20 21
best preservation of ultrastructure and surface antigen staining. For convenience, when duplicated antibodies were used for detection of T-cell antigens, we will refer only to the OKT series of antibodies in the text. The 10.2 antibody detects the so-called pan-T-cell antigen (Leu 1).
Reaction of primary antibodies was localized with the avidin-biotin peroxidase complex according to the method of Hsu et al.24 The reaction product was darkened with 1%7o osmium. For light microscopy, the immunoperoxidase-labeled frozen sections were covered with mounting medium and a coverslip. Thirteen
(2, 3, 6, 8-10, 13-15, 17-19, and 24) were studied by immunoelectron microscopy for Ki-l, Leu Ml, Tac, Ia, and T-cell-specific antigens. Frozen sections prepared and antibody-labeled were fixed in 3 7o gluteraldehyde before the diaminobenzidine (DAB) reaction and then dehydrated in a graded series of alcohols and embedded in Epon 812 (Ted Pella, Inc., Tustin, Calif) by rapidly inverting gelatin embedding capsules filled with the epoxy resin. The Epon was allowed to harden overnight at 60 C. Detachment of the capsule from the glass slide was accomplished by gentle heating. Ultrathin sections were cut from the face of the Epon block with cases
Table 3-Phenotype of Tumor Cells in Skin of Patients With Mycosis Fungoides
Immunologic markers T-cell-specific 9.6 (OKT11) Leu 1 (10.2) Leu 4 (OKT3) Leu 3a (OKT4) Leu 2a (OKT8) 3A1
Activated T-cell Tac OKT9 OKT1 0 HB10a (anti-la)
B-cell-specific HLB-3
Case 1
A
B
Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Case 9 Case 10 Case 11 Case 12
+
+
+
+
+
+
+
+
+
+
+
0
+
+
+
+
+
+
+c 0 +
+
+
+
+
+
+
+
+
+
+
+
+
+
+c
+
+
+
+
+
+
0
0
0
0
0
0
0
0
0
+
0
+a
+a
0
0
0
0
+
+
+
+c
+a
0
0
+a +a
+b
+a
0
+b
+b +b
0
+b +b 0
0
0
+a
0
0
+a
+
0 +
+b ND ND ND
ND
+
0
+
0
+b +b
0
0
0
0
+a
+a
0
+a
+a
ND
ND
0
ND
0
ND
0
ND
0
ND
0
0
slg
0
0
ND
0
ND
0
ND
0
ND
0
ND
ND
Other OKT6 Ki-1
O
O
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
ND
0
ND
ND
0
+b
A, skin; B, lymph node; 0, negative; ND, not done. +, more than 50% of infiltrating cells are positive; + a, more than 20% and fewer than 50% of infiltrating cells of infiltrating cells are positive; +c, most infiltrating cells are weakly positive.
are
positive;
ND 0
ND ND
ND + 0b
+
b, fewer than 20%
IMMUNOPA'THOLOGY OF CTCL
Vol. 119 * No.3
Table 4-Phenotype of Tumor Cells in Skin (S) and Blood (B) of Patients With Sezary Syndrome Case 14 Case 15 Case 16 Case 13 Immunologic B S S S B S markers T-cell-specific 9.6 (OKT11) Leu 1 (10.2) Leu 4 (OKT3) Leu 3a (OKT4) Leu 2a (OKT8) 3A1 Activated T-cell Tac OKT9 OKT1 0 HB10a (anti-la) B-cell-specific HLB-3 slg
Other OKT6 Ki-1
o
0
+
+
0 0
0
+b +a
0 + +
+
+
+
+a
+
+ +
+ +
Case 18
Case 17 S
B
+
+
o
0
0
+0
+C
O
ND
+ +
0 +
+a +b
+b +b
0
0
0
ND ND
+
+
0
0 0
ND ND ND 0
ND O
ND ND
0 ND
0 ND
ND O
o o
o 0
0 0
0 0
O
O
+, more than 50% of cells are positive; + a, more than 20% and fewer than 50% of infiltrating cells cells are positive; + c, weakly positive, detected only by immunoelectron microscopy. ND, not done; 0, negative.
a diamond knife on an LKB ultramicrotome and examined either unstained or stained with uranyl acetate under a JEOL electron microscope. In cells that reacted with monoclonal antibodies, a continuous layer of granular electron-dense reaction product outlined the cell membrane and cell processes between adjacent cells. No reaction product was observed within the cytoplasm, overlying the nucleus, or in the interstitium. Control sections stained with an irrelevant antibody were examined ultrastructurally in all cases and revealed no staining. Phenotypes of circulating tumor cells were studied by immunofluorescence microscopy as previously described.25
Results As shown in Tables 3 and 4, immunopathologic studies in MF/SS demonstrated that the infiltrating tumor cells usually expressed the pan-T-cell antigens Leu 1, TI1, and T3. However, the tumor cells from Patients 13 and 14 with SS lacked the Tll antigen corresponding to the E-rosette receptor. In 11 of 12 MF cases and 6 of 6 SS cases, tumor cells expressed the T4 antigen of the helper/inducer phenotype. However, the T4 antigen was only weakly expressed by a majority of infiltrating cells of Patient 3, who had many blastic cells in his skin lesion (Figure 1). The tumor cells of Patient 9 also lacked the T4 antigen but expressed the T8 antigen of the suppressor/cytotoxic phenotype (Figure 2). On tumor cells of 5 of 12 patients with MF (Cases 1, 6, 7, 8, 10; Table 3) and only 1 of 6 patients with the
0
+ +
ND O
ND ND
ND 0
ND ND
0
ND
+b
O
positive;
+ b,
S
0 0 +b
+a
+
are
439
+b
0
0 ND 0 +b
fewer than 20% of infiltrating
SS (Case 17; Table 4) 3A1 antigen was detected. Our immunoelectron-microscopic study confirmed that 3A1 antigen was expressed by tumor cells with cerebriform nuclei in these 3AM-positive cases (Figure 3A). Moreover, in SS Case 15 (Table 4) immunoelectron microscopy revealed 3Al staining of cerebriform cells that appeared negative by light microscopy (Figure 3B). Ia (HLA-DR) antigen was detected in skin-infiltrating tumor cells in 9 of'12 cases of MF and 5 of 6 cases of SS (Figure 4). Both skin and blood specimens were studied for Ia antigen in 3 patients (13, 16, and 17) with SS. Sezary cells in the skin, but not in the blood, of 1 patient (Case 17) expressed Ia antigen. In the other 2 patients the phenotype of Sezary cells was the same in both skin and blood, Ia' in Case 16 and la- in Case 13. Anti-Tac antibody, which recognizes the 11-2 receptor,26 reacted with only a minor population of cerebriform cells. T9 antigen, corresponding to the transferrin receptor, was expressed on a minimum of 1-2/o of tumor cells in Cases 5 and 9 to a maximum of 80% of tumor cells in Case 3. In 7 patients with MF/SS, only a few T8+ residual normal T cells were found in the epidermis, whereas 9 other patients had numerous T8+ cells in the epidermis. T6+ Langerhans' cells (LCs) were usually confined to the epidermis, but some aggregates of T6+ LCs were found in reticular dermis of 9 patients with MF/SS. An involved lymph node from Patient 1 also contained a large number of T6+ LCs coexistent with many cerebriform cells that were of the same phenotype as tumor cells in the skin. Ki-1 antigen, a Hodgkin's-disease-associated antigen that is also ex-
440
AJP * June 1985
NASU ET AL
- ...X.;ilr ~~~~~~~~~~~~~~~~~~~~~~~~~~. ...
.1e
A.~
.W
~
..
Vol. 119
*
. ;,^.- ;
IMMUNOPATHOLOGY OF CTCL
No. 3
441
s--/ #*
,,,, ,.
-.
X
,.= ,,,,;
..
X...;.
% e.'d'
|
: w i-. {jf
r
A f
.;Si
;"il s
i: ..^
..1,.
s3WF
rr
^
,.S
.F A *;'
s
ti ,,,,,
IX'
J ,,
Figure 2-Patient 9, with a plaque lesion of mycosis fungoides. A-Immunoperoxidase staining of skin biopsy shows aggregates of T8-positive tumor cells (arrow) in the basal layer of the epidermis. B-Tumor cells (arrow) appear (x 480) to be negative for 3A1 antigen. (x 480) However, immunoelectron microscopy revealed weak staining of tumor cells for 3A1 antigen (similar to Figure 3B).
.loi
-F j;
^
'; rf ?S
4-s_
w
41
;,.
7-,
W."
a.
Lil . ;
~~~~A.
pressed on a subset of activated helper T cells,2" was found infrequently in skin biopsies of patients with MF and SS. It was detected on a minority of cells, generally with a transformed appearance, in 2 patients with MF (Patients 10 and 11) and 2 with SS (Patients 17 and 18). As shown in Table 5, immunopathologic studies of skin-infiltrating ATL cells revealed a uniform helper Tcell phenotype similar to the phenotype of ATL cells in the peripheral blood; 3Al antigen was not detected. Tumor cells from 3 of 4 ATL patients expressed la antigens. Tac antigen was expressed on ATL cells in the skin in all cases but at low frequency, compared with
.i.;.
..
ATL cells in peripheral blood. T9 was positive in only one ATL case. Numerous T6' LCs were found in the epidermis, but not in the dermis. Ki-l antigen was not detected, although large RS-like cells were not found in the skin biopsies of these patients with ATL. Nonepidermotropic T-cell lymphomas (Table 6) included a heterogenous group of 4 large-cell and 1 lymphoblastic lymphomas. Case 23 had immunologic characteristics similar to those of MF/SS and ATL. Case 24 was different from MF/SS and ATL in that both Leu 1 and T4 antigens were absent. Among large-cell lymphomas, there were 2 cases (25 and 26) that contained
Figure 1-Patient 3, with rapidly growing tumors of mycosis fungoides. A-Epidermotropism of convoluted cells. (H&E, x 800) B-Blastic appearance and mitoses in dermal infiltrate. (x 1250) C-Immunoperoxidase staining of frozen sections shows many cells are positive for Leu 1. (x800) D-Most cells appear to be negative for T4. (x800)
442
NASU ET AL
AJP * June 1985
Figure 3A-An electron micrograph shows immunoperoxidase staining of 3A1 antigen as
a dark line on the membrane of a cerebriform tumor cell and a
small normal lymphocyte in Patient 8, who had mycosis fungoides. (x 13,999) B-Similar 3A1 antigen staining in a skin biopsy specimen of Patient 15, who had the Sdzary syndrome. (x 19,000)
0r
3'
,
,.
:1
fa' wt
frequent multinucleated Reed-Sternberg-like cells expressing Hodgkin's-disease-associated antigen Ki-I (Figure 5). These Reed-Sternberg-like cells also expressed T-cell-specific antigens Til and T4, and in Case 26, T3 (Figure 5B), but no Leu 1 or T8 antigens. In all large-cell lymphomas, la and Tac antigens were ex-
pressed by the infiltrating tumor cells, and 3A1 antigen was absent. Case 27, a 22-year-old man who presented with numerous cutaneous nodules, had tumor cells that were negative for Tll, T3, T4, T8, T6, and la but positive for Leu 1, TIO, 3A1, indicating an immature thymo-
IMMUNOPATHOLOGY OF CTCL
Vol. 119 * No. 3
443
Figure 4-An electron micrograph showing staining of la antigen on the membrane of a cerebriform tumor cell of Patient 15, with the Sezary syndrome (x 13,333)
cyte phenotype. This patient did not have a mediastinal mass, but developed central nervous system involvement consistent with the usual clinical course of lymphoblastic lymphoma.
though large numbers of LCs were observed in both diseases. The role of the LCs in the pathogenesis of these diseases is not yet clarified. In a separate report we describe a difference in the stability of T4 antigen expres-
Discussion
Table 5-Phenotype of Tumor Cells in Skin of Patients With Adult T-Cell Leukemia Immunologic markers Case 19 Case 20 Case 21 Case 22
Although MF/SS is more common in Western countries and ATL is more frequent in the Orient, especially in Japan, recent studies have shown that both groups of diseases have a worldwide distribution.' 28-3' Because of the expected poor prognosis in ATLI 3.132 and its frequent association with a T-cell leukemia virus,33-36 it is important to distinguish MF/SS from ATL. ATL and MF/SS often have overlapping clinical and histologic features.3',37-40 Therefore, we sought to establish possible immunopathologic criteria to distinguish MF/SS from ATL. However, this was not possible, because tumor cells from most MF/SS and all ATL patients were both of the same helper T-cell phenotype (Tll+, T3+, T4+). Unexpectedly, we found many Tac-positive cells in the skin of the patients with MF/SS; Tac expression was previously reported as the distinguishing feature of ATL cells.3" 5 Furthermore, no consistent differences in tumor-cell expression of Ia, T9, and 3A1 antigens were found in skin lesions of patients with MF/SS and ATL. The presence of numerous T6+ LCs in the dermis was more common in MF/SS than ATL in this series, al-
T-cell-specific 9.6 (OKT11) Leu 1 (10.2) Leu 4 (OKT3) Leu 3a (OKT4) Leu 2a (OKT8) 3A1 Activated T-cells Tac OKT9 OKT1 0 HB10a (anti-la) B-cell-specific HLB-3 slg Other OKT6 Ki-1
+b + +
+ +
0
+
+
0
0 0
0 +a +a
+b
0 0 0
0
0 +
0
0+b 0
0 ND
0 ND
0 ND
0 ND
0 0
0 0
0 0
0 0
+, more than 50% of infiltrating cells are positive; + a, more than 20% and fewer than 50% of infiltrating cells are positive; + b, fewer than 20% of infiltrating cells are positive. ND, not done; 0, negative.
444
NASU ET AL
AJP * June 1985
Table 6-Phenotype of Tumor Cells in Skin of Patients With Cutaneous T-Cell Lymphoma Without Epidermotropism Immunologic markers Case 23* Case 24* Case 25* Case 26* T-cell-specific 9.6 (OKT11) Leu 1 (10.2) Leu 4 (OKT3) Leu 3a (OKT4) Leu 2a (OKT8) 3A1 Activated T-cell Tac OKT9 OKT1 0 HB10a (anti-la) B-cell-specific HLB-3
+ + +
0+ 0
ND
+
+
0 +
0 0
0 00
+a ND
Case 27t +b
0
+
+ +
0 0
0 0
0 +C 0
0
ND ND 0
0 0
+
sig
00 0
0 ND
0 0
0 0
ND 0
Other OKT6 Ki-1
0 0
0 0
0
0
+
0 0
+, more than 50% of infiltrating cells are positive; +a, more than 20% and fewer than 50% of infiltrating cells are positive; of infiltrating cells are positive; +c, most infiltrating cells are weakly positive. * Large-cell lymphoma. t Lymphoblastic lymphoma. 0, negative; ND, not done.
sion by MF/SS and ATL cells treated by phorbol ester in vitro as a possible means of distinguishing between these two groups of patients.?5 In this study we confirmed the heterogeneity of tumor-cell phenotypes in MF/SS. Haynes described one patient with MF whose tumor cells had an unusual nonhelper phenotype, T4-, T8+, 3Al+.4' We found a similar case of MF with the suppressor/cytotoxic phenotype T4-, T8+, 3A1+, that of a 30-year-old male who had atrophic slightly scaling patches in a bathing trunk distribution for more than 20 years without other symptoms. The skin biopsy in this case showed a peculiar pattern of epidermotropism, with cerebriform cells confined mainly to the basal layer, where they sometimes formed Pautrier's microabscesses (Figure 2). Scheen et al recently reported 2 young patients who had long-standing red scaling lesions similar to those of our patient; MF and then Hodgkin's disease developed.42 Further studies may show whether the expression of T8 antigen by infiltrating tumor cells identifies a group of patients with a different prognosis among those with MF. We studied another atypical patient (Patient 3) with advanced MF who had rapidly progressive disease with formation of large tumor nodules on his face. These nodules showed marked infiltration of blastic tumor cells with a T3+, T4+, 3A1-, T9+ phenotype. However, the T4 antigen was only weakly expressed and not readily detected by the usual light-microscopic immunoperoxidase technique. The weak expression of T4
+ b,
fewer than 20%
antigen accompanied by blastic transformation of tumor cells has previously been described in some cases of advanced MF.43 We have shown that the loss of T4 antigen induced by phorbol ester in vitro is also accompanied by blastic transformation of cerebriform cells.25 These findings suggest that the usual immunopathologic criteria applied for the diagnosis of MF may be too rigid for recognition of all cases. Despite the initial reports of a uniform 3A1 phenotype of MF/SS cells in peripheral blood,4 subsequent studies of skin lesions,4" 14 including the present study, show a variable expression of 3A1 antigens by cerebriform cells (Figures 3 and 4). Our study reveals that the absence of 3A1 antigen is correlated with a high frequency of leukemic dissemination in patients with MF/SS. We do not know whether 3A1 antigen expression determines the homing of cells into the skin or is acquired after skin infiltration because of the influence of the skin microenvironment. In 1 patient with SS, la was expressed by infiltrating cells in the skin but not in the peripheral blood. Haynes et al also reported a disparity of antigen expression in the skin and peripheral blood45 but did not distinguish on a cell-to-cell basis whether these antigens were expressed by cells with cerebriform nuclei or nonneoplastic cells. In this study it was possible to confirm the expression of 3A1 and Ia antigens by tumor cells with cerebriform nuclei in some cases by immunoelectron microscopy. We found 2 cases of SS negative for sheep erythrocyte rosette (ER) receptor. This may cause diagnostic
Vol. 119 * No. 3
IMMUNOPATHOLOGY OF CTCL 0
Figure 5A-Reed-Sternberg-like cell in a skin tumor of Patient 26, who had nonepidermotropic T-cell lymphoma. B-T3 staining of (H&E, x 1200) a Reed-Sternberg-like cell in a skin Ctumor of Patient 26. (x 1200) Immunoperoxidase staining shows nearly all cells are positive for Ki-1 antigen. (x 480)
difficulties, because SS is commonly known as a mature helper T-cell malignancy. In these ER-receptornegative cases, the diagnosis of SS was confirmed by the expression of the T-cell antigens Leu 1, T3, and T4. Moreover, we have previously shown that the ER receptor can be induced by phorbol ester on ER-negative SS cells.25
Five of our patients with T-cell malignancies had nonepidermotropic skin infiltrates. As in the study of Wood et al,46 the lesions of these patients were heterogeneous in both morphologic features and phenotype. Four patients had large-cell and 1 a lymphoblastic lymphoma. In each case there was a lack of one or more mature T-cell antigens that were found in the MF/SS and ATL groups. Among the large-cell lymphomas, 2
445
:Pqww:
B
had invasive characteristics and contained frequent RSlike cells. At least one of these cases could represent progression of lymphomatoid papulosis to an immunoblastic T-cell lymphoma.4' We have found frequent expression of Ki-1 antigen by large atypical RS-like cells in each of 10 cases of lymphomatoid papulosis.48 The 2 large-cell lymphomas with RS-like cells were the only T-cell lymphomas in which the majority of tumor cells expressed the Hodgkin's-disease-associated antigen Ki1. These lymphomas also had a high frequency of Ta-, Tac-, and T9-positive cells. A recent report27 revealed that helper T cells activated in vitro by autologous or allogeneic lymphocytes could express Ki-l antigen, consistent with our findings of co-expression of Ki-1, Tl1, and T4 antigens by malignant helper T cells.
446
NASU ET AL
One 22-year-old patient who presented with many cutaneous nodules was shown to have a nonepidermotropic small-cell lymphoma with an immature thymocyte phenotype (3A1+, Leu 1+, T10+, T6-, T1l-, T3-, T4-, T8-). Immunopathologic studies helped to establish the correct diagnosis of T-cell lymphoblastic lymphoma and to predict the subsequent clinical course of central nervous system disease. Previous reports have indicated marked difficulty in distinguishing cutaneous T-cell malignancies from benign T-cell infiltrates by immunopathologic techniques alone.4" 49 We did not include benign T-cell disorders in this series but suggest that their distinction from cutaneous T-cell lymphomas requires correlation of clinical, morphologic, and immunologic data. A diagnosis of T-cell malignancy can be supported on immunologic grounds in those cases in which an immature phenotype (Case 27) or an anomalous phenotype (Cases 13, 14, and 24-26) is discovered on the majority of cells.
References 1. Kadin ME, Kamoun M: Nonendemic adult T cell leukemia/lymphoma. Hum Pathol 1982, 13:691-693 2. Hattori T, Uchiyama T, Takatsuki K, Uchino H: Presence of human B-lymphocyte antigens on adult T cell leukemia cells. Clin Immunol Immunopathol 1980, 17:287-295 3. Takatsuki K, Uchiyama T, Ueshima Y, Hattori T, Toibana T, Tsudo M, Wano R, Yodoi J: Adult T cell leukemia: Proposal as a new disease and cytogenetic, phenotypic, and functional studies of leukemic cells. Gann Monograph on Cancer Research. Edited by M Haneoka, T Takatsuki, M Shimoyama. Tokyo, Japan Scientific Societies Press; New York, Plenum Press, 1982, 28: 13-22 4. Haynes BF, Metzgar RS, Minna JD, Bunn PA: Phenotypic characterization of cutaneous T-cell lymphoma: Use of monoclonal antibodies to compare with other malignant T cells. N Engl J Med 1981, 304:1319-1323 5. Hinuma Y, Nagata K, Hanaoka M, Nakai M, Matsumoto T, Kinoshita K, Shirakawa S, Miyoshi I: Adult T-cell leukemia: Antigen in an ATL cell line and detection of antibodies to the antigen in human sera. Proc Natl Acad Sci USA 1981, 78:6476-6480 6. Kalyanaraman VS, Sarngadharan MG, Bunn PA, Minna JD, Gallo RC: Antibodies in human sera reactive against an internal structural protein of human T-cell lymphoma virus. Nature 1981, 294:271-273 7. Bunn Jr PA, Lamberg SI: Report of the committee on staging and classification of cutaneous T-cell lymphomas. Cancer Treat Rep 1979, 63:725-728 8. McLean IW, Nakane PK: Periodate-lysine-paraformaldehyde fixative: A new fixative for immunoelectronmicroscopy. J Histochem Cytochem 1974, 22:1077-1083 9. Kamoun M, Martin PJ, Hansen JA, Brown MA, Siadak AW, Nowinski RC: Identification of a human T lymphocyte surface protein associated with the E-rosette receptor. J Exp Med 1981, 153:207-212 10. Martin PJ, Hansen JA, Nowinski RC, Brown MA: A new human T cell differentiation antigen: Unexpected expression on chronic lymphocytic leukemia cells. Immunogenetics 1980, 11:429-439 11. Reinherz EL, Kung PC, Goldstein G, Schlossman SF: A monoclonal antibody with selective reactivity with func-
AJP * June 1985
12.
13.
14.
15.
16. 17. 18.
19.
20.
21.
22.
tionally mature human thymocytes and all peripheral mature T cells. J Immunol 1979, 123:1312-1317 Ledbetter JA, Evans RL, Lipinski M, CunninghamRundles C, Good RA, Herzenberg LA: Evolutionary conservation of surface molecules that distinguish Tlymphocyte helper/inducer and cytotoxic/suppressor subpopulations in mouse and man. J Exp Med 1981, 153: 310-323 Evans RL, Wall DW, Platsoucas CD, Siegal FP, Fikrig SM, Testa CM, Good RA: Thymus-dependent membrane antigens in man: Inhibition of cell-mediated lympholysis by monoclonal antibodies to the TH2 antigen. Proc Natl Acad Sci USA 1981, 78:544-548 Haynes BF, Eisenbarth GS, Fauci AS: Human lymphocyte antigens: Production of a monoclonal antibody that defines functional thymus-derived lymphocyte subsets. Proc Natl Acad Sci USA 1979, 76:5829-5833 Uchiyama T, Broder S, Waldmann TA: A monoclonal antibody (anti-Tac) reactive with activated and functionally mature T cells: I. Production of anti-Tac monoclonal antibody and distribution of Tac (+) cells. J Immunol 1981, 126:1393-1397 Trowbridge I, Omary B: Human cell surface glycoprotein related to cell proliferation is the receptor for transferrin. Proc Natl Acad Sci USA 1981, 78:3039-3043 Reinherz E, Schlossman S: Biochemical studies of the human thymocyte cell-surface antigens T6, T9 and TIO. Cell 1981, 23:771-780 Hansen JA, Martin PJ, Nowinski RC: Monoclonal antibodies identifying a novel T cell antigen and Ia antigens of human lymphocytes. Immunogenetics 1980, 10:247260 Kikuchi K, Koshiba H, Kasai K, Hiratsuka H, Fujimoto N, Vede T: B-cell malignancies and differentiation antigens defined with monoclonal antibodies. Jpn J Clin Oncol 1983, 13:517-532 McMichael AJ, Pilch JR, Galfre G, Mason DY, Fabre JW, Milstein C: A human thymocyte antigen defined by a hybrid myeloma monoclonal antibody. Eur J Immunol 1979, 9:205-210 Schwab U, Stein H, Gerdes J, Lemke H, Kirchner H, Schaadt M, Diehl V: Production of a monoclonal antibody specific for Hodgkin and Sternberg-Reed cells of Hodgkin's disease and a subset of normal lymphoid cells. Nature 1982, 299:65-67 Borowitz MJ, Newby S, Brynes RK, Cousar JB, Whitcomb CC, Crissman JD, Byrne GE, Collins RD: Multi-
institution study of non-Hodgkin's lymphomas using frozen section immunoperoxidase: The Southeastern Cancer Study Group experience. Blood 1984, 63:1147-1152 23. Said JW, Shintaku P, Chien K, Sassoon AF: Peripheral T-cell lymphoma (Immunoblastic sarcoma of T-cell type): An immuno-ultrastructural study. Hum Pathol 1984, 15:324-329 24. Hsu SM, Raine L, Fanger H: Use of Avidin-BiotinPeroxidase Complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 1981, 29:577-580 25. Kadin ME, Nasu K, Sako D, Su IJ: Distinctive phorbol ester-induced morphological and surface antigen changes in mycosis fungoides, Sezary syndrome, and adult T-cell leukemia. Cancer Res 1984, 44:3383-3387 26. Leonard WJ, Depper JM, Uchiyama T, Smith KA, Waldmann TA, Greene WC: A monoclonal antibody that appears to recognize the receptor for human T-cell growth factor: Partial characterization of the receptor. Nature 1982, 300:267-269 27. Andreesen R, Osterholz J, Lohr GW, Bross KJ: A Hodgkin cell-specific antigen is expressed on a subset of autoand alloactivated T (helper) lymphocytes. Blood 1984, 63:1299-1302
Vol. 119 * No. 3
28. Uchiyama T, Yodoi J, Sagawa K, Takatsuki K, Uchino H: Adult T-cell leukemia: Clinical and hematologic features of 16 cases. Blood 1977, 50:481-492 29. Catovsky D, Greaves MF, Rose M, Galton DAG, Goolden AWG, McCluskey DR, White JM, Lampert I, Bourikas G, Ireland R, Brownell Al, Bridges JM, Blattner WA, Gallo RC: Adult T-cell lymphoma-leukemia in blacks from the West Indies. Lancet 1982, 1:639-643 30. Tricot GJK, Broeckaert-Van Orshoven A, den Ottolander GJ, de Wolf-Peeters C, Meyer CJLM, Verwilghen RL, Jansen J: Adult T-cell leukemia: A report on two white patients. Leukemia Res 1983, 7:31-42 31. Jaffe ES, Cossman J, Blattner WA, Robert-Guroff M, Blayney DW, Gallo RC, Bunn Jr PA: The pathologic spectrum of adult T-cell leukemia/lymphoma in the United States: Human T-cell leukemia/lymphoma virus-associated lymphoid malignancies. Am J Surg Pathol 1984, 8:263-275 32. The T- and B-cell malignancy study group: Statistical analysis of immunologic, clinical and histopathologic data on lymphoid malignancies in Japan. Jpn J Clin Oncol 1981, 11:15-38 33. Hinuma Y, Komoda H, Chosa T, Kondo T, Kohakura M, Takenaka T, Kikuchi M, Ichimaru M, Yunoki K, Sato I, Takiuchi Y, Uchino H, Hanaoka M: Antibodies to adult T-cell leukemia-virus-associated antigen (ATLA) in sera from patients with ATL and controls in Japan: A nationwide sero-epidemiologic study. Int J Cancer 1982, 29: 631-635 34. Yoshida M, Miyoshi I, Hinuma Y: Isolation and characterization of retrovirus from cell lines of human adult T-cell leukemia and its implication in the disease. Proc Natl Acad Sci USA 1982, 79:2031-2035 35. Robert-Guloff M, Nakao Y, Notake K, Ito Y, Sliski A, Gallo RC: Natural antibodies to human retrovirus HTLV in a cluster of Japanese patients with adult T cell leukemia. Science 1982, 215:975-978 36. Blayney DW, Jaffe ES, Blattner WA, Cossman J, RobertGuroff M, Longo DL, Bunn PA, Gallo RC: The human T-cell leukemia/lymphoma virus (HTLV) associated with American adult T-cell leukemia/lymphoma (ATL). Blood 1983, 62:401-405 37. Watanabe S, Nakajima T, Shimosato Y, Shimoyama M, Minato K: T-cell malignancies: Subclassification and interrelationship. Jpn J Clin Oncol 1979, 9 (Suppl):423-442 38. Hanaoka M: Clinical pathology of adult T-cell leukemia. Acta Hematol Jpn 1981, 44:1420-1430 39. Shimoyama M, Minato K, Tobinai K, Nagai M, Setoya T, Takenaka T, Ishihara K, Watanabe S, Hoshino H, Miwa M, Kinoshita M, Okabe S, Fukushima N, Inada N: Atypical adult T-cell leukemia-lymphoma: Diverse clinical manifestations of adult T-cell leukemia-lymphoma. Jpn J Clin Oncol 1983, 13 (Suppl 2):165-188
IMMUNOPATHOLOGY OF CTCL
447
40. Yamaguchi K, Nishimura H, Kohragi H, Jono M, Miyamato Y, Takatsuki K: A proposal for smoldering adult T-cell leukemia: A clinicopathologic study of five cases. Blood 1983, 62:758-766 41. Haynes B, Hensley LL, Jegasothy BV: Phenotypic characterizations of skin-infiltrating T cells in cutaneous T-cell lymphoma: Comparison with benign cutaneous T-cell infiltrates. Blood 1982, 60:463-473 42. Scheen III SR, Banks PM, Winkelmann RK: Morphologic heterogeneity of malignant lymphomas developing in mycosis fungoides. Mayo Clin Proc 1984, 59:95-106 43. Willemze R, de Graaff-Reitsma CB, Cnossen J, Van Vloten WA, Meijer CJLM: Characterization of T-cell subpopulations in skin and peripheral blood of patients with cutaneous T-cell lymphomas and benign inflammatory dermatoses. J Invest Dermatol 1983, 80:60-66 44. Wood GS, Deneau DG, Miller RA, Levy R, Hoppe RT, Warnke RA: Subtypes of cutaneous T-cell lymphoma defined by expression of Leu-l and Ia. Blood 1982, 59: 876-882 45. Haynes BF, Hensley LL, Jegasothy BV: Differentiation of human T lymphocytes: II Phenotypic difference in skin and blood malignant T-cells in cutaneous T-cell lymphoma. J Invest Dermatol 1982, 78:323-326 46. Wood GS, Burke JS, Horning S, Doggett RS, Levy R, Warnke RA: The immunologic and clinicopathologic heterogeneity of cutaneous lymphomas other than mycosis fungoides. Blood 1983, 62:464-472 47. Tucker WFG, Leonard JN, Smith N, Coulter C, Woods B, Clayton RJ: Lymphomatoid papulosis progressing to immunoblastic lymphoma. Clin Exp Dermatol 1984, 9:190-195 48. Kadin ME, Nasu K, Said J, Vonderheid E, Sako D: Lymphomatoid papulosis: A cutaneous proliferation of activated helper T cells expressing Hodgkin's disease associated antigens. Am J Pathol 1985, 119:315-325 49. Chu A, Patterson J, Berger C, Vonderheid E, Edelson R: In situ study of T-cell subpopulations in cutaneous T cell lymphoma: Diagnostic criteria. Cancer 1984, 54:2414-2422
Acknowledgments We thank Dr. Yorio Hinuma for assistance in detecting serum antibodies of ATLV (HTLV), Dr. Kokichi Kikuchi, Dr. Takashi Uchiyama, and Dr. Harald Stein for providing monoconal antibodies, and Ms. Gwen Mallet for secretarial
assistance.
