ORIGINAL ARTICLE

T Cell Receptor-c Gene Analysis of CD30þ Large AtypicalIndividual Cells in CD30þ Large Primary Cutaneous T CellLymphomas

Sylke Gellrich,1 AnkeWilks, Ansgar Lukowsky, MartinWernicke, Astrid Mˇller, J. Marcus Muche,Tanja Fischer,Kim Christian Jasch, Heike Audring, andWolfram SterryDepartment of Dermatology, Medical Faculty (ChariteŁ ), Humboldt-University, Berlin, Germany

The hallmark of primary cutaneous CD30þ large Tcelllymphoma are large lymphoid tumor cells, at least 75%of which, by de¢nition, must be positive for CD30. Therelatively benign clinical course of this lymphoma typehas been explained with CD30-induced apoptosis,on the assumption that expression of CD30 de¢nes thetumor clone; however, this hypothesis has not beentested on the molecular level to date. In this study weanalyzed CD30þ cells in four patients with primary cu-taneous CD30þ large T cell lymphoma by single cellpolymerase chain reaction of Tcell receptor-c genes fol-lowed by sequencing. Here, we demonstrate that mostof the large CD30þ atypical cells possessed identicalT cell receptor-c gene rearrangements, indicative ofclonal proliferation. Nevertheless, polyclonally rear-

ranged T cells were present in all CD30þ samples stu-died. In addition, one patient showed a second clone ina separate biopsy and three of four patients showedchromosomal imbalances as revealed by comparativegenomic hybridization. Taken together, our data sug-gest that the CD30þ population in primary cutaneousCD30þ large T cell lymphoma indeed contains thetumor clone, thus providing molecular support for alink between clinical course and CD30-related signal-ing. Importantly, however, CD30 expression does notde¢ne the tumor clone as bystander T cells, as well asoccasional additional clones, are also present in thispopulation. Key words: CD30þ /lymphoma/CTCL/Singlecell analysis/clonality. J Invest Dermatol 120:670 ^675, 2003

CD30 expression represents a diagnostic parameter inseveral lymphoproliferative disorders. A cytokinereceptor of the tumor necrosis factor receptor super-family, CD30 is mainly restricted to the large cellvariants of non-Hodgkin lymphoma, including the

primary cutaneous T cell lymphomas (pCTCL), as well as Reed^Sternberg cells in Hodgkin’s disease (Schwab et al, 1982; Stein et al,1985; Smith et al, 1993;Willemze and Beljaards, 1993; Falini et al,1995; Gruss et al, 1996; Willemze et al, 1997; Bekkenk et al, 2000;Willemze and Meijer, 2000). Predominant CD30 expression indi-cates a favorable prognosis in pCTCL (Willemze and Beljaards,1993, Willemze et al, 1997; Bekkenk et al, 2000), in contrastto Hodgkin’s disease and nodal CD30þ anaplastic large celllymphoma (Tilly et al, 1997; Falini et al, 1999).Primary cutaneous CD30þ large T cell lymphomas (CD30þ

large pCTCL) form a distinct entity within the European Organ-ization for Research andTreatment of Cancer classi¢cation of pri-mary cutaneous lymphomas. Although histologically pCTCL issubdivided into anaplastic (80%) and pleomorphic or immuno-

blastic (20%) subtypes, all types show an identical clinicalpresentation with nodules and tumors (Willemze and Beljaards,1993; Willemze et al, 1997; Bekkenk et al, 2000). Lymphomacells show a nodular distribution within the dermis and subcutisbut not the epidermis, typically featuring round, oval, orirregularly shaped nuclei, prominent nucleoli, and abundantcytoplasm. The large CD30þ cells are CD4þ with varyingloss of pan-T cell antigens (CD2, CD3, CD5). By de¢nitionCD30 is expressed on more than 75% of morphologicallyatypical cells. Because of these features and clonal T cellreceptor (TCR)-g rearrangement present in bulk lymphomaDNA it is generally assumed that the large atypical CD30þ

cells represent the tumor cell population (Willemze andBeljaards, 1993; Willemze et al, 1997; Bekkenk et al, 2000). Todate, however, there exist no molecular data supporting thisassumption. Therefore, we investigated TCR-g rearrangementsof single cells found within the CD30þ cell population ofpCTCL.

MATERIALS ANDMETHODS

Patient I A 36 y old male initially developed a single erythematousnodule on the left lower leg. After ruling out systemic disease (using chestX-ray and ultrasound of abdomen and lymph nodes as well as bonemarrow biopsy), CD30þ large pCTCL was diagnosed. Various treatmentmodalities, including excision, radiotherapy, acitretin, interferon a, andchemotherapy with methotrexate and CHOP (Cyclophosphamide,Hydroxydaunomycin, Oncovin, and Prednisone) did not induce completeremission (Fig 1IA). Three years after the occurrence of the ¢rst skin lesion

Reprint requests to: Sylke Gellrich, M.D., Department of Dermatology,Medical Faculty (ChariteŁ ), Schumannstr. 20/21, 10117 Berlin, Germany.Email: Sylke.Gellrich@charite.deAbbreviations: pCTCL, primary cutaneous T cell lymphoma; CD30þ

pCTCL, CD30-positive large primary cutaneous T cell lymphoma; FFA,£uorescent fragment analysis

1Joint ¢rst authors.

Manuscript received March 18, 2002; revised October 2, 2002; acceptedfor publication November 6, 2002

0022-202X/03/$15.00 � Copyright r 2003 by The Society for Investigative Dermatology, Inc.

670

the patient died of cerebral metastases and multiple skin involvement.Biopsies from the initial lesion showed a CD30þ large pCTCLin¢ltrating the entire dermis down to the subcutaneous fatty tissue.Atypical cells appeared as medium- to large-sized cells with a CD3^,CD4þ, CD30þ phenotype and a mitotic rate of about 70% (Fig1IB,IC). Molecular biologic investigation of blood and skin samplesrevealed a monoclonal TCR-g gene rearrangement via polymerase chainreaction (PCR) from bulk DNA and following £uorescent fragmentanalysis (FFA). For single cell PCR, CD30þ cells were dissected from thedermal in¢ltrate as described below.

Patient II A 75 y old male with a CD30þ large pCTCL presented witha single erythematous nodule on the left elbow without extracutaneousinvolvement as documented by chest X-ray, ultrasound of the abdomenand peripheral lymph nodes and computed tomography of the chest andabdomen. Excision and postexcisional radiotherapy induced completeremission of about 15 mo duration. Thereafter, new lesions developed onthe left hand and lower arm (Fig 1IIA), which were excised. Eightmonths later this patient is alive in partial remission. Biopsy from theinitial lesion showed a CD30þ large pCTCL in¢ltrating the wholedermis. Atypical cells were medium sized to large sized with a CD3^,

Figure1. Patients with CD30þ large primary cutaneous Tcell lymphoma. Patient I, lower leg; patient II, ¢nger; patient III, upper leg, biopsy IIIa;patient IV, hand wrist. (A) Clinical occurrence. (B) Hematoxylin and eosin staining of lesional skin, original magni¢cation� 40. (C) Immunohistochemicalanti-CD30-staining of lesional skin, original magni¢cation� 40.

TCR-g GENE ANALYSIS OF CD30þ CELLS IN CD30þ pCTCL 671VOL. 120, NO. 4 APRIL 2003

CD4þ, CD30þ phenotype and a mitotic rate of almost 100% (Fig1IIB,IIC). Monoclonal TCR-g rearrangements were detected in bloodsamples as well as skin samples by PCR from bulk DNA followed byFFA. For single cell PCR, CD30þ cells were dissected from the dermalin¢ltrate as described below.

Patient III A 62 y old female with a CD30þ large pCTCL presentedwith multiple erythematous papules on the left upper leg (Fig 1IIIA).The patient was treated with so-called mimotope vaccination therapy(two nonapeptides mimicking unknown tumor antigens) for a total of10 mo (Linnemann et al, 2000, 2001). After 2 mo of treatment completeremission was achieved. Relapse occurred after an additional 3 mo, andthe vaccination therapy was continued leading to a second completeremission 3 mo later. A second, very aggressive relapse of skin lesions, aswell as tumor involvement of the bone marrow and lymph nodes,occurred 10 mo after the beginning of therapy. For that reason,vaccination therapy was discontinued. The patient was treated withCHOP chemotherapy and died 19 mo after occurrence of the ¢rst skinlesions.Biopsy of the initial lesion showed an in¢ltrate extending through the

whole dermis up to the subcutaneous fatty tissue. Morphologically,T cellswere medium-sized cells with a CD3þ, CD4þ, CD30þ phenotype and amitotic rate of about 10%. AT cell population with a monoclonal TCR-grearrangement was detected in all blood and skin samples. A medium-sizedpleomorphic T cell lymphoma with CD30 expression was diagnosed.Subsequent lesions showed large CD30þ tumor cells with aberrantphenotypes. For single cell PCR, CD30þ cells were dissected from thedermis of two separate skin biopsies and from one blood sample:

Skin biopsy IIIa This skin biopsy was taken from a papule developing dur-ing the ¢rst relapse after mimotope vaccination. Its histologic and morpho-logic features were identical to pretreatment biopsies (Fig 1IIIB,IIIC).

Skin biopsy IIIb A second nodule was excised after 10 mo of mimotopevaccination, during an aggressive relapse and showed an in¢ltrate of thewhole dermis reaching the subcutaneous fatty tissue. Tcells were large cellswith a loss of CD3 and CD4 expression (CD3^, CD4^, CD30þ pheno-type) and a mitotic rate of 90^100%. Analysis of blood skin samples byPCR from bulk DNA followed by FFA revealed a T cell population witha biallelic TCR-g rearrangement in theVg2 primer set. Disease progressioninto a CD30þ anaplastic large T cell lymphoma was diagnosed.

Blood sample IIIc Because of leukemic manifestation of T cell lymphoma,we also dissected single large CD30þ blood lymphocytes from a cytospinslide to compare TCR-g genes of these cells with cells from skin biopsies(IIIa,b). The leukemic cells exhibited the same phenotype as those de-scribed in biopsy IIIb.

Patient IV A 24 y old male presented with plaques on the right lowerarm and wrist. By histology, the diagnostic criteria of CD30þ largepCTCL were met (see below); there was no evidence of systemic disease(by X-ray, ultrasound of abdomen and lymph nodes) (Fig 1IVA).Treatment with local electron beam irradiation of the right lower arminduced complete remission, lasting 3 mo at the time of writing.Skin specimen, analyzed by single cell PCR, showed a subepidermal

CD30þ atypical T cells with a proliferation rate of 50% (Fig 1IVB,IVC).Almost all CD30þ large atypical cells were CD3^ and CD4^.Routine molecular biologic investigation (FFA^TCR-g) revealed clonal

TCR rearrangements in skin biopsy and a polyclonal PCR product inblood.All human studies were performed in accordance to the Declaration of

Helsinki Principles.

Preparation of single cells from lesional skin sections viamicromanipulation The skin biopsies were snap frozen in liquidnitrogen and stored at ^801C. Frozen skin sections (10 mm thick) werestained with monoclonal mouse anti-CD30 (DAKO, Glostrup, Denmark)using biotinylated Fab anti-mouse followed by streptavidin-peroxi-dase conjugated complexes (LSAB-Kit 2: DAKO, Hamburg, Germany)and hydrogen peroxidase^chromogen (AEC Substrate System: DAKO,Hamburg, Germany) as developing reagents. Subsequently, counterstain-ing with hematoxylin was performed.CD30þ cells were dissected microscopically (Nikon, Dˇsseldorf,

Germany) by means of a hydraulic micromanipulator (Narishige,Dˇsseldorf, Germany). Thereafter, single cells were aspirated into amicropipette (Narishige), placed into 20 ml PCR bu¡er supplementedwith 1ng rRNA per ml (both from Boehringer, Mannheim, Germany),and stored at ^201C. Photographs were taken before and after micro-manipulation of each cell.

Preparation of single cells from peripheral blood In the case ofpatient III CD30þ blood cells were investigated in addition to the skinsamples. Peripheral blood mononuclear cells were prepared from 10 mlheparinized blood (sample IIIc) by density gradient centrifugation usingFicoll Hypaque (Pharmacia, Freiburg, Germany). Peripheral bloodmononuclear cells were cytospun on slides. Immunohistologic stainingand micromanipulation of single CD30-positive cells was subsequentlyperformed as 0described above.

PCR ampli¢cation of rearranged single cell TCR-c genes andsequence analysis Initially, cells were incubated with 0.25 mg per mlProteinase K (Boehringer) for 50 min at 551C and for an additional10 min at 951C (enzyme inactivation). Rearranged TCR-g genes wereampli¢ed by a seminested PCR with primers Vg1 (50CTACATCC-ACTGGTACCT 30), recognizing Vg1^Vg8 gene families (Volkenandtet al, 1993), Vg9ext (ATTGGTATCGAGAGAGAC), Vg10/11ext(CACTGGTACKKGCAGAAAC) (Muche et al, 1997), and Jg1/2(50CAACAAGTGTTGTTCCAC 30) in the ¢rst round, and with Vgseq(50AGRCCCCACAGCRTCTTC 30), recognizing Vg1^Vg8 gene families,Vg9int (GGAAAGGAATCTGGCATTCCG), Vg10 (AATCCGCAG-CTCGACGCAGCA), and Vg11 (GCTCAAGATTGCTCAGGTGGG)(Trainor et al, 1991) in the second round.The reaction mix (50 ml) for the ¢rst round of ampli¢cation consisted

of 1�PCR bu¡er (Boehringer), 2.5 mM MgCl2 (Boehringer), 200 mM ofeach nucleotide (deoxyadenosine triphosphate, deoxythymidine triphos-phate, deoxycytidine triphosphate, deoxyguanosine triphosphate) (AGS,Heidelberg, Germany), 7 nM of each primer, and 3.5 U Expand high¢delity Taq-Polymerase (Boehringer). PCR ampli¢cation was carried outon a Personal Cycler (Biometra, G˛ttingen, Germany): one cycle at 951Cfor 2 min, 651C for 1 min (addition of Taq-Polymerase), and 721C for 1min, followed by 35 cycles at 951C for 1 min, 581C for 1 min and 721Cfor 1 min, and a ¢nal extension at 721C for 5 min. The second round ofPCRwas carried out using 1 ml of the PCR product from the ¢rst round.Reaction mixture: 1�PCR bu¡er (Boehringer), 2.5 mM MgCl2(Boehringer), 200 ml of each nucleotide (deoxyadenosine triphosphate,deoxythymidine triphosphate, deoxycytidine triphosphate, deoxy-guanosine triphosphate) (AGS), 7 nM of each primer and 1 U Taq-polymerase (Boehringer). The second round cycle program was carriedout on aTrio-Thermoblock (Biometra) and consisted of one cycle at 951Cfor 2 min, 681C for 5min (addition of Taq-Polymerase), and at 721C for 1min, 45 cycles at 951C for 1 min, at 581C for 1 min, 721C for 1 min, and a¢nal extension at 721C for 5 min. A 5 ml aliquot of the reaction mixturewas analyzed on a 2% agarose gel for successful ampli¢cation.In order to avoid contamination with DNA, micromanipulation, ¢rst

and second round of PCR were carried out separately in three di¡erentrooms. Every 10 tubes one tube without cells was inserted as negativecontrol, failing to show any products. PCR products were puri¢ed usingthe Gel DNA Extraction Kit (Boehringer) followed by direct sequencingwith 5 ml Dye Terminator Cycle Sequencing Ready Reaction Kit (PerkinElmer,Weiterstadt, Germany) on the automated DNA Sequencing SystemABI 373A (Perkin Elmer Applied Biosystems,Weiterstadt, Germany) using30 ng puri¢ed PCR product and 5 mM Vgseq,Vg9,Vg10, and Vg11 primer.Sequence evaluation was done using Sequence Navigators software(Perkin Elmer Applied Biosystems). DNA sequences were analyzed onlineusing DNA BLOTs software (created by H.H. Althaus) and comparedwith the EMBL gene bank data (Heidelberg, Germany) for known TCR-g genes. In total, 257 single CD30þ cells were analyzed. Two hundred andthirty were localized in the dermis or in the subcutaneous fatty tissue.Twenty-seven CD30þ cells were isolated from a cytospin sample ofperipheral blood lymphocytes. DNA could be ampli¢ed and sequenced in137 of 257 isolated single CD30þ cells, what represents a method relevante⁄cacy (Kˇppers et al, 1995).

FFA^TCR-c Primers Vg1 and Vg2 (V9, V10/11) or Jg1/2 were 50 -labeledwith 5-carboxy-£uorescein (FAM). Labeled PCR products or productmixtures were subjected to FFA on the ABI 310 PRISM (capillary; PEApplied Biosystems) or, as indicated, on the ABI 373A (polyacrylamidegel electrophoresis) sequencers (PE Applied Biosystems). For FFA on theABI 310 PRISM, 12 ml deionized formamide and 0,5 ml Genescan 500TM

ROX internal lane standard (PE Applied Biosystems) were added to 1 ml ofthe PCR ampli¢cation, denatured at 901C for 2 min and chilled on ice toproduce single stranded DNA. Each run was performed at 601C and 15 kVwith a 5 s injection and 36 min separation time in a 47 cm POP 6-¢lledcapillary (PE Applied Biosystems).For analysis on the ABI 373A, 5 ml deionized formamide, 0.5 ml

ethylenediamine tetraacetic acid, 0.5 ml GeneScan 2500TM ROX internallane standard (PE Applied Biosystems) and 0.5 ml dextran blue wereadded to 2 ml of the labeled PCR product, denatured as described aboveand loaded on to a 4.75% polyacrylamide gel (LC 6 premixed gel, FMC

672 GELLRICH ETAL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Bioproducts, Rockland MD). The runs of both sequencers were analyzedusing the ‘‘GeneScan 372’’ software (PE Applied Biosystems).

Comparative genomic hybridization (CGH) Lesional skin samplesof all four patients were investigated for chromosomal aberrations(Table I) as described byTonnies et al (2001).

RESULTS

Large, atypical CD30þ cells of four patients with CD30þ largepCTCL were examined by single cell PCR analysis followed bysequencing of the TCR-g chain genes (Table I). The majority ofdermal CD30þ cells displayed identical individual TCR-g genesin all patients (patient I: 81.8%; patient II: 81.5%; patient III:biopsy IIIb 74.0%, blood IIIc 78.6%; patient IV: 78.2%). The onlyexception was skin sample IIIa of patient III containing only aminority of cells with an identical TCR-g rearrangement (patientIII: biopsy IIIa 18.2%). For details seeTable I.Two di¡erent clonal TCR-g chain genes could be ampli¢ed

from several cells of patients I^IV. Therefore, both TCR-g genescan be assigned to a biallelic, clonal (malignant) Tcell population(Table II); however, the remainder of CD30þ Tcells analyzed inthis investigation was polyclonal (sequences not shown). In 18 of38 of these polyclonal cells in all investigated specimens, bothTCR-g chain genes could be ampli¢ed and sequenced success-fully. Morphologically, there was no di¡erence between CD30þ

cells belonging to the predominant clone and their polyclonalcounterparts.A remarkable additional observation was made in patient III,

where three samples were investigated by single cell analysis. Inbiopsy IIIa, two di¡erent clonal Tcell populations were identi¢ed(lengths of PCR products in clone 1: 241 bp, 250 bp and in clone2: 251 bp). Clone 1 (241 bp, 250 bp) persisted during disease pro-gression. Four of 22 CD30þ cells (18.2%) (Table I) in biopsy IIIa

showed a biallelic TCR-g chain gene rearrangement (clone 1:241 bp, 250 bp); in subsequent samples (IIIb, IIIc), clone 1 had ex-panded, accounting for 70^80% of all CD30þ cells in biopsyIIIb and blood sample IIIc. Clone 2 (251 bp;Table II) was foundin ¢ve of 18 CD30þ cells taken from biopsy IIIa, showing onlyone rearranged TCR-g chain gene. This clone was no longer de-tectable in subsequent samples (IIIb, IIIc). Most CD30þ cells insample IIIa were of a polyclonal nature (59.1%) (Table I). Inbiopsy IIIb and blood sample IIIc, the majority but not all (simi-lar to patient I, II, and IV) of CD30þ T cells belonged to theclonal population (74% and 78.6%, respectively). In order to con-¢rm these results, FFA^TCR-g was performed from bulk DNAof samples IIIa, IIIb, and IIIc. Two clonal PCR products of iden-tical sequence lengths were demonstrated as previously detectedin single cell PCR (clone 1: 241 bp, 250 bp) (Fig 2D,E,Table II). The second, transient, clone 2 (251 bp) most likely cor-responds to a dominant PCR product in FFA^TCR-g (length251 bp) in one biopsy prior to treatment (Fig 2A) and in speci-men IIIa (Fig 2B) under vaccination therapy.Finally, an analysis of skin samples by CGH detected the fol-

lowing chromosomal imbalances: trisomy 7 in patients I and II,deletion of 7q and trisomy 14 in patient III and no aberration atall in patient IV (Table I).

DISCUSSION

The CD30 antigen, a member of the tumor necrosis factor recep-tor superfamily, has been shown to be involved in the process oflymphoid development and di¡erentiation (Falini et al, 1995;Gruss et al, 1996; Horie andWatanabe, 1998). Apart from its phy-siologic function, CD30 expression may play a pivotal role intumorigenesis and tumor survival, as CD30 expression representsan important diagnostic and prognostic marker in patients with

Table I. Results of micromanipulation and single cell PCR for TCR-g chain gene rearrangement in CD30þ largepCTCL and CGH

Patient Compartment CGH results Cell size ofCD30þ cells

No. ofisolated cells

Clonal/total sequences Polyclonal/total sequences

I Dermis Trisomy 7 Medium-sized 50 27/33 (81.8%) 6/33 (18.2%)subcutaneous to largefat

II Dermis Trisomy 7 Medium-sized 45 22/27 (81.5%) 5/27 (18.5%)to large

IIIa Dermis No material Medium-sized 36 4/22 (18.2%) 13/22 (59.1%)(5/22a) (22.7%)

IIIb Dermis, Partial deletion 7q large 54 17/23 (74%) 6/23 (26%)subcutaneous trisomy 14fat

IIIc Blood No material large 27 11/14 (78.6%) 3/14 (21.4%)IV Dermis No aberration large 45 13/18 (72.2%) 5/18 (27.8%)

aIn skin biopsy IIIa identical rearrangements were found in ¢ve of 22 cells that do not belong to the tumor cell population

Table II. Clonal TCR-g rearrangements of the patients, detected by single cell PCR

Patient Allele V-gene V-segment N-sequence J-segment Length of TCR

I 1 V9 TACTACTGTGCCTTGGTGG AGAAACTTGACAGAGAAAC GAATTATTATAAGAAA 2572 V10 TACTACTGTGCTGCGTGGG C TATTATAAGAAA 240

II 1 V4 TATTACTGTGCCACCTGGGATGGG GGCGG ATAAGAAA 2492 V5 TATTACTGTGCCACCTGGGACAGG CCCAGCGG GAAA 248

IIIa,b,c 1 V9 TATTACTGTGCCACCTGGGA GACC TATTATAAGAAA 2412 V10 TACTACTGTGCTGCGTGGGA AGGCCTATGTTGA TATAAAAA 250

IIIa 1 V8 TACTACTGTGCCTTGTGGGAG CCC GAATTATTATAAGAAA 251IV 1 V10 TACTACTGTGCTGCGTGGGAT CC GAAA 235

2 V4 TACTACTGTGCCTTGTGGGATGGG TTGGG ATTATAAGAAA 252

TCR-g GENE ANALYSIS OF CD30þ CELLS IN CD30þ pCTCL 673VOL. 120, NO. 4 APRIL 2003

cutaneous lymphoma. In order to de¢ne the extent of overlapbetween CD30þ large atypical cells and the malignant T cellclone, we studied TCR-g gene sequences in single CD30þ cellsby micromanipulation and single cell PCR. Using the same ap-proach, it has previously been demonstrated that CD30þ Reed^Sternberg and Hodgkin cells in Hodgkin’s disease represent themalignant cells, 90% of which are of B cell origin and 10% ofT cell origin (Kˇppers et al, 1994; Maeuschen et al, 2000).We detected a population of T cells with clonal TCR-g gene

rearrangements in all four cases studied. All of these tumor-re-lated cells showed a biallelic TCR-g rearrangement (Tables Iand II). Unexpectedly, we also found polyclonal T cells besidesthe expected clonal population with relatively high frequency.Previously, we have shown that T cells expressing one expandedVb family comprise both reactive cells and the tumor cell popu-lation in mycosis fungoides (Gellrich et al, 2000).The role of these

bystanding T cells present within the tumor cell populationof both disease entities is currently unknown. Possibly, theyrepresent cytotoxic T cells.Several studies show that cytotoxic T cells exist in cutaneous T

cell lymphoma expressing various kinds of inhibitory receptorsinvolved in anti-tumor responses. In this study all patientsshowed a loss of CD3 within the CD30þ cell in¢ltrate. Theabsence of this cell surface receptor could represent a tumor es-cape mechanism as described by Bagot et al (2000). In the earlierstage of disease in patient III (sample IIIa) we observed expressionof CD3 and CD4 on CD30þ cells and a relatively high numberof polyclonal T cells possibly induced by mimotope vaccination.Paralleling clinical tumor progression a loss of CD3 and CD4was noted (samples IIIb and IIIc), concomitant with increasednumbers of clonal cells. Thus, continued expression of CD3 andCD4 might be relevant for an intact anti-tumor response.

Figure 2. FFA-TCR-c of Patient III. (A) Lesional skin prior to treatment:TCR-g1. (B) Lesional skin under vaccination therapy:TCR-g1. (C) Peripheralblood lymphocytes, leukemic stage: TCR-g1. (D) Lesional skin under vaccination therapy: TCR-g2. (E) Peripheral blood lymphocytes, leukemic stage:TCR-g2. Abscissa: sequence length in bp.Vertical axis: £uorescent intensity.

674 GELLRICH ETAL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Our ¢nding of a second expanded clonal T cell population insample IIIa (revealed by the same length of PCR products(251 bp) in single cell analysis and in FFA^TCR-g) suggests theexistence of a reactive cytotoxic anti-tumor cell line (Linnemannet al, 2000, 2001). In the peripheral blood of patient III an in-creased number of mimotope-activated cytotoxic T cells couldbe found by speci¢c interferon-g production of these cells (un-published data).As noted by Kadin (Mori et al, 1999), signaling pathways seem

to be important in the tumor biology of CD30þ cutaneousTcelllymphoma. Of these, the CD30/CD30 ligand interaction may beinvolved in spontaneous self-regression of tumor lesions probablydue to tumor cell apoptosis. Nevertheless, recent observations in-dicate that aberrant cell activation by CD30 may cause the oppo-site e¡ect: in anaplastic large cell lymphoma (ALCL) cell linestransformed from progressed lymphomatoid papulosis lesions,CD30/CD30 ligand signaling leads to cell proliferation insteadof cell death. Therefore, defective CD30/CD30 ligand interactionin CD30þ tumor cells could promote tumor progression by re-sistance to CD30-mediated growth inhibition (Mori et al, 1999;Levi et al, 2000). Furthermore, the interaction between transform-ing growth factor b and its receptor is compromised by muta-tions in the transforming growth factor b receptor gene (Knauset al, 1996; Schiemann et al, 1999). Interestingly, tumor cells in cu-taneous T cell lymphoma were shown to produce high amountsof transforming growth factor b leading to the inhibition ofspeci¢c anti-tumor responses (Bagot et al, 2001) and a lack of Fasexpression (Mori et al, 1999).In contrast to the favorable clinical course generally noted in

CD30þ pCTCL two of the investigated patients (patients I andIII) underwent a highly aggressive clinical course, both dyingwithin 3 y. Mechanisms such as chromosomal imbalances (Bey-lot-Barry et al, 1998; Zoi Toli et al, 2000) are considered importantin tumor progression. In line with this notion, we detected a tris-omy 7 in patients I and II as well as a partial deletion of 7q and atrisomy 14 in patient III by CGH analysis; however, whetherthese aberrations are present in the clonal, the polyclonal, or bothpopulations remains an open question at present.In conclusion, our results demonstrate that CD30þ large aty-

pical T lymphocytes in CD30þ large pCTCL indeed includeclonally rearranged cells, and thus the tumor clone, but do notrepresent a uniform cell population, but rather a mixture of clo-nal and polyclonal T lymphocytes. With regard to the role ofpolyclonal CD30þ Tcells within the tumor lesions, we speculatethat strongT cell activating factors, i.e.,T helper 2 type cytokinesare active within the lesion, inducing common morphologic, andlikely functional, features both in malignant in bystander T cells(Li et al, 1997). As activated Tcells are probably more vulnerable tothe genetic alterations detected in our patients, such local stimulimay be directly involved in the initial phases of lymphomagenesis.

The Deutsche Forschungsgemeinschaft supported this work, grant no. GE 1004/1-1and 1-2.We thank H.Tonnies for supporting the CGH experiments.

REFERENCES

Bagot M, Martinvallet D, Echchakir H, et al: Functional inhibitory receptors ex-pressed by a cutaneous T cell lymphoma-speci¢c cytolytic clonal T cell popula-tion. J Invest Dermatol 115:994^999, 2000

Bagot M, Nikolova M, Schirm-Chabanette F, et al: Crosstalk between tumor T lym-phocytes and reactive T lymphocytes in cutaneous T cell lymphomas. Ann N YAcad Sci 941:31^38, 2001

Bekkenk MW, Geelen FA, van-Voorst-Vader PC, et al: Primary and secondary cuta-neous CD30(þ ) lymphoproliferative disorders: A report from the Dutch Cu-taneous Lymphoma Group on the long-term follow-up data of 219 patientsand guidelines for diagnosis and treatment. Blood 95:3653^3661, 2000

Beylot-Barry M, Groppi A,Vergier B, et al: Characterization of t(2;5) reciprocal tran-scripts and genomic breakpoints in CD30þ cutaneous lymphoproliferations.Blood 91:4668^4676, 1998

Falini B, Pileri S, Pizzolo G, et al: CD30 (Ki-1) molecule: A new cytokine receptor ofthe tumor necrosis factor receptor superfamily as a tool for diagnosis and im-munotherapy. Blood 85:1^14, 1995

Falini B, Pileri S, Zinzani PL, et al: ALKþ lymphoma. Clinico-pathological ¢ndingsand outcome. Blood 93:2697^2706, 1999

Gellrich S, Lukowsky A, Schilling T, et al: Microanatomical compartment of clonaland reactive cells in mycosis fungoides: Molecular demonstration by single cellPCR of T cell receptor gene rearrangement. J Invest Dermatol 115:620^624, 2000

Gruss HJ, Duyster J, Herrmann F: Structural and biological features of the TNF re-ceptor and TNF ligand superfamilies: Interactive signals in the pathobiology ofHodgkin’s disease. Ann Oncol 7:19^26, 1996

Horie R,Watanabe T: CD30. Expression and function in health and disease. SeminImmunol 10:457^470, 1998

Knaus PI, Lindemann D, DeCoteau JF, et al: A dominant inhibitory mutant of thetype II transforming growth factor beta receptor in the malignant progressionof a cutaneous T-cell lymphoma. Mol Cell Biol 16:3480^3489, 1996

Kˇppers R, Rajewsky K, Zhao M, et al: Hodgkin disease: Hodgkin and Reed-Stern-berg cells picked from histological sections show clonal immunoglobulin generearrangements and appear to be derived from B cells at various stages of de-velopment. Proc Natl Acad Sci USA 91:10962^10966, 1994

Kˇppers R, Hansmann ML, Rajewsky K: Micromanipulation and PCR analysis ofsingle cell from tissue sections. In:Weir DM, Blackwell C, Herzberg LA (eds).Handbook of Experimental Immunology, 5th edn. Cambridge, MA: BlackwellScience, 1995

Levi E,Wang Z, Petrogiannis-Haliotis T, et al: Distinct e¡ects of CD30 and Fas sig-naling in cutaneous anaplastic lymphomas: A possible mechanism for diseaseprogression. J Invest Dermatol 115:1034^1040, 2000

Li G, Salhany KE, Rook AH, et al: The pathogenesis of large cell transformation incutaneous T-cell lymphoma is not associated with t(2;5)(p23;q35) chromoso-mal translocation. J Cutan Pathol 24:403^408, 1997

Linnemann T,Wiesmuller KH, Gellrich S, et al: A T-cell epitope determined withrandom peptide libraries and combinatorial peptide chemistry stimulates Tcells speci¢c for cutaneous T-cell lymphoma. Ann Oncol 11:95^99, 2000

LinnemannT,Tumenjargal S, Gellrich S, et al: Mimotopes for tumor-speci¢c T lym-phocytes in human cancer determined with combinatorial peptide libraries.Eur J Immunol 31:156^165, 2001

Maeuschen M, Rajewsky K, Brauninger A, et al: Rare occurrence of classicalHodgkin’s disease as a T cell lymphoma. J-Exp Med 191:387^394, 2000

Mori M, Manuelli C, Pimpinelli N, et al: CD30^CD30 ligand interaction in primarycutaneous CD30(þ ) T-cell lymphomas. A clue to the pathophysiology ofclinical regression. Blood 94:3077^3083, 1999

Muche JM, Lukowsky A, Asadullah K, et al: Demonstration of frequent occurrenceof clonal T cells in the peripheral blood of patients with cutaneous T-cell lym-phoma. Blood 90:1636^1642, 1997

SchiemannWP, Pfeifer WM, Levi E, et al: A deletion in the gene for transforminggrowth factor beta type I receptor abolishes growth regulation by transform-ing growth factor beta in a cutaneous T-cell lymphoma. Blood 15:2854^2861,1999

Schwab U, Stein H, Gerdes J, et al: Production of a monoclonal antibody speci¢c forHodgkin and Sternberg-Reed cells of Hodgkin’s disease and a subset of nor-mal lymphoid cells. Nature 299:65^67, 1982

Smith CA, Gruss HJ, Davis T, et al: CD30 antigen, a marker for Hodgkin’s lympho-ma, is a receptor whose ligand de¢nes an emerging family of cytokines withhomology to TNF. Cell 73:1349^1360, 1993

Stein H, Mason DY, Gerdes J, et al:The expression of the Hodgkin’s disease associatedantigen Ki-1 in reactive and neoplastic lymphoid tissue: Evidence that Reed-Sternberg cells and histiocytic malignancies are derived from activated lym-phoid cells. Blood 66:848^858, 1985

Tilly H, Gaulard P, Lepage E, et al: Primary anaplastic large-cell lymphoma in adults:Clinical presentation, immunophenotype, and outcome. Blood 90:3727^3734,1997

Tonnies H, Stumm M,Wegner RD, et al: Comparative genomic hybridization basedstrategy for the analysis of di¡erent chromosome imbalances detected in con-ventional cytogenetic diagnostics. Cytogenet Cell Genet 93:188^194, 2001

Trainor KJ, Brisco MJ,Wan JH, et al: Gene rearrangement in B- and T-lymphoproli-ferative disease detected by the polymerase chain reaction. Blood 78:192^196,1991

Volkenandt M, Burmer GC, Schadendorf D, et al: The polymerase chain reaction.Method and applications in dermatopathology. Am J Dermatopathol 15:118^126,1993

Willemze R, Beljaards RC: Spectrum of primary cutaneous CD30 (Ki-1)-positivelymphoproliferative disorders. A proposal for classi¢cation and guidelines formanagement and treatment. J Am Acad Dermatol 28:973^980, 1993

Willemze R, Meijer CJ: EORTC classi¢cation for primary cutaneous lymphomas: acomparison with the R.E.A.L. Classi¢cation and the proposedWHO Classi¢-cation. Ann Oncol 11:11^15, 2000

Willemze R, Kerl H, Sterry W, et al: EORTC classi¢cation for primary cutaneouslymphomas: A proposal from the Cutaneous Lymphoma Study Group ofthe European Organization for Research and Treatment of Cancer. Blood90:354^371, 1997

Zoi Toli O, Vermeer MH, De-Vries E, et al: Expression of Fas and Fas-ligand inprimary cutaneous T-cell lymphoma (CTCL): Association between lack ofFas expression and aggressive types of CTCL. Br J Dermatol 143:313^319, 2000

TCR-g GENE ANALYSIS OF CD30þ CELLS IN CD30þ pCTCL 675VOL. 120, NO. 4 APRIL 2003