OBSERVATION

Granulomatous Mycosis Fungoidesand Granulomatous Slack Skin

A Multicenter Study of the Cutaneous Lymphoma Histopathology Task Force Groupof the European Organization for Research and Treatment of Cancer (EORTC)

Werner Kempf, MD; Sonja Ostheeren-Michaelis, MD; Marco Paulli, MD; Marco Lucioni, MD; Janine Wechsler, MD;Heike Audring, MD; Chalid Assaf, MD; Thomas Rüdiger, MD; Rein Willemze, MD; Chris J. L. M. Meijer, MD;Emilio Berti, MD; Lorenzo Cerroni, MD; Marco Santucci, MD; Christian Hallermann, MD; Mark Berneburg, MD;Sergio Chimenti, MD; Alistair Robson, MBBS; Martà Marschalko, MD; Dmitry V. Kazakov, MD, PhD;Tony Petrella, MD; Sylvie Fraitag, MD; Agnes Carlotti, MD; Philippe Courville, MD; Hubert Laeng, MD;Robert Knobler, MD; Philippa Golling, MD; Reinhard Dummer, MD; Günter Burg, MD

Background: Granulomatous cutaneous T-cell lympho-mas (CTCLs) are rare and represent a diagnostic chal-lenge. Only limited data on the clinicopathological andprognostic features of granulomatous CTCLs are avail-able. We studied 19 patients with granulomatous CTCLsto further characterize the clinicopathological, therapeu-tic, and prognostic features.

Observations: The group included 15 patients withgranulomatous mycosis fungoides (GMF) and 4 withgranulomatous slack skin (GSS) defined according to theWorld Health Organization–European Organization forResearch and Treatment of Cancer classification for cu-taneous lymphomas. Patients with GMF and GSS dis-played overlapping histologic features and differed onlyclinically by the development of bulky skin folds in GSS.Histologically, epidermotropism of lymphocytes was nota prominent feature and was absent in 9 of 19 cases (47%).

Stable or progressive disease was observed in most pa-tients despite various treatment modalities. Extracuta-neous spread occurred in 5 of 19 patients (26%), sec-ond lymphoid neoplasms developed in 4 of 19 patients(21%), and 6 of 19 patients (32%) died of their disease.Disease-specific 5-year survival rate in GMF was 66%.

Conclusions: There are clinical differences between GMFand GSS, but they show overlapping histologic findingsand therefore cannot be discriminated by histologic ex-amination alone. Development of hanging skin folds isrestricted to the intertriginous body regions. Granulo-matous CTCLs show a therapy-resistant, slowly progres-sive course. The prognosis of GMF appears worse thanthat of classic nongranulomatous mycosis fungoides.

Arch Dermatol. 2008;144(12):1609-1617

T HE OCCURRENCE OF SAR-coid-like granulomas is awell-known phenomenonin malignant lymphomaand is most commonly ob-

served in patients with Hodgkin disease.1,2

In contrast, granulomatous features arerarely found in primary cutaneous lympho-mas(CLs),withapproximately2%ofallCLsdisplaying granulomatous features.3,4

Granuloma formation was reported ina broad variety of primary CLs,5,6 such asSezary syndrome,7 primary cutaneous ana-plastic large T-cell lymphoma,8 subcuta-neous panniculitis-like T-cell lymphoma,and primary cutaneous B-cell lympho-mas.4 Granulomatous mycosis fungoides(MF) is the most common form of granu-lomatous cutaneous T-cell lymphoma(CTCL). In contrast, granulomatous slackskin (GSS) is a very rare form of CTCL, andto date only approximately 50 cases have

been reported in the literature.9,10 In theWorld Health Organization–European Or-ganization for Research and Treatment ofCancer (WHO-EORTC) classification forcutaneous lymphomas, GSS is considereda distinct subtype of MF with characteris-tic clinical and histologic features.11,12

There have been only a limited numberof studies on granulomatous CTCLs, par-ticularly granulomatous MF (GMF). Theclinicopathological features and the courseof granulomatous CTCLs are still poorlycharacterized.Thegranulomaformationcanbe very extensive, so that the histologic di-agnosis of lymphoma may be delayed, andthe findings are often initially misdiag-nosed as granulomatous dermatitis.4 Thereis controversy over whether the presence ofgranulomas in CLs correlates with a betterprognosis.5,13 Thus, a multicenter study wasconducted to analyze the clinical, histo-pathological, immunophenotypic, and ge-

Author Affiliations are listed atthe end of this article.

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1609

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

notypic featuresofgranulomatous reactions inCTCLs,par-ticularly GMF and GSS, as well as the course and prognosisof these granulomatous CTCLs.

METHODS

PATIENTS AND BIOPSIES

Twenty-three skin biopsy specimens from 23 patients with well-documented disease from a total of 18 European centers were sub-mitted as “granulomatous CTCL” by the members of the EORTCCutaneous Lymphoma Histopathology Task Force. Cases to beincluded and further analyzed had to show prominent granu-loma formation or numerous histiocytic giant cells or a histiocyte-rich infiltrate defined by histiocytes accounting for more than 25%of the entire infiltrate. The following clinical data were recorded:sex, age at diagnosis, biopsy site, clinical manifestation includ-ing location and distribution of skin lesions, TNM stage at diag-nosis according to TNM classification of malignant tumors,14 ageat first symptoms (if available), age at first biopsy specimen dis-playing granulomatous features, results of staging investiga-tions, treatment, response to treatment, and outcome.

Inclusion criteria included hematoxylin-eosin and immu-nohistochemical stainings of diagnostic quality, written de-tailed data or photographs of clinical presentation, and infor-

mation on therapeutic interventions, as well as follow-up oncourse and outcome.

All biopsy specimens were formalin-fixed and paraffin-embedded.Hematoxylin-eosin stainingaswell as staining forelas-tic fibers was performed. Immunohistochemical staining for lym-phocytic (CD3, CD4, CD8, CD30) and histiocytic (CD68)antigens was visualized by the streptavidin-biotin or alkaline phos-phatase–anti–alkaline phosphatase method according to stan-dard protocols. Rearrangement of T-cell receptor � genes was as-sessed by polymerase chain reaction as previously described.15,16

Statistical analysis was performed with SPSS version 15.0software (SPSS Inc, Chicago, Illinois).

RESULTS

According to the WHO-EORTC classification for cutane-ous lymphomas,11,12 19 cases of CTCL could be identifiedand classified as GMF (n=15) or GSS (n=4). Two casesoriginally submitted as GSS were reclassified as GMF be-cause the skin lesions did not evolve to hanging skin foldsduring the follow-up period. Four additional cases wereclassified as primary cutaneous peripheral T-cell lym-phoma, unspecified, and were excluded from further analy-sis because this study focused on GMF and GSS. In addi-

Table 1. Clinical and Therapeutic Data From a Series of 19 Patients With Granulomatous Cutaneous T-Cell Lymphomas

Patient No./Sex/Age, y Location Lesion Typea TNM Stage Therapy

TreatmentResponse

ExtracutaneousSpread

2ndNeoplasia

Outcome/Follow-up, y

Granulomatous Mycosis Fungoides1/F/51 Buttocks,

extremitiesPapules, patches IB, T2N0M0 IFN, PUVA PR None None AWD/8

2/F/20 Face, arm Patches III, T4N0M0 IFN, chemo SD LN, spleen, ocular None DOD/13/M/35 Groin Plaque (unilesional) IA, T1N0M0 CS, MC, RT, IFN CR None None ACR/204/F/81 Generalized Plaques (pigm) IB, T2N0M0 CS, MC PR None None AWD/65/M/39 Generalized Plaques, nodules

(pigm)IIB, T3N0M0 RT, chemo PR None HL DOD/2

6/M/52 Generalized Patches, plaques,nodules

IB, T2N0M0 PUVA, RT PD BM, retroperitonealLN, spleen

None DOD/1

7/F/64 Trunk Patches IA, T1N0M0 Imiq, IFN, PUVA, RT CR LN None ACR/68/F/61 Generalized Patches, plaques

(pigm)IA, T1N0M0 Chemo SD None None AWD/1

9/M/57 Generalized Patches, plaques IVA, T3N2M0 Chemo CR LN None AWD/410/M/47 Trunk, legs Plaques IIA, T1N1M0 Pred PD None HL AWD/511/M/30 Trunk Plaques

(poikilodermatous)IA, T1N0M0 PUVA, IFN PR None None AWD/4

12/F/28 Trunk Patches, plaques IA, T1N0M0 PUVA, RT, IFN PD LN, BM, liver Myeloidleukemia

DOD/9

13/M/27 Trunk Plaques, nodules IA, T1N0M0 CS, PUVA, IFN, ret SD None None AWD/714/M/72 Generalized Plaques, nodules IB, T2N0M0 PUVA, IFN, RT, chemo PD None None DOD/115/M/47 Legs Plaques, nodules

(pigm)IA, T1N0M0 CS, IFN, ret, RT PD None Nodal CD30�

ALCLDOD/5

Granulomatous Slack Skin16/M/55 Groin Patches, plaques,

hangingIA, T1N0M0 PUVA, ret, IFN PR None Cutaneous

CD30� LPDAWD/10

17/M/71 Axilla, groin Plaques, hanging IA, T1N0M0 NA NA None None AWD/1618/F/35 Trunk, groin,

legPatches, plaques,

hangingIA, T1N0M0 Excision, CS, MC, PUVA,

IFNPD None None AWD/15

19/F/22 Trunk includingaxilla/groin

Patches, plaques,hanging

IA, T1N0M0 Excision, carmustine PR None None AWD/28

Abbreviations: ACR, alive with complete remission; ALCL, anaplastic large cell lymphoma; AWD, alive with disease; BM, bone marrow; chemo, chemotherapy;CR, complete tumor regression; CS, topical corticosteroids; DOD, died of disease; hanging, hanging skin folds; HL, Hodgkin lymphoma; IFN, interferon alfa;imiq, imiquimod; LN, lymph nodes; LPD, lymphoproliferative disorder; MC, mechlorethamine hydrochloride; NA, not available; pigm, hyperpigmented; PD, progressivedisease; PR, partial tumor regression; pred, oral prednisone; PUVA, psoralen–UV-A light therapy; ret, retinoids; RT, radiotherapy; SD, stable disease.

aLesions were multiple in all except patient 3.

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1610

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

tion to the23casesofCTCL,3casesof secondarycutaneousinvolvement by systemic T-cell non-Hodgkin lym-phoma— including 1 case of nodal Lennert lymphoma (asa variant of nodal peripheral T-cell lymphoma), 1 case ofangioimmunoblastic T-cell lymphoma, and 1 case of nodalCD4� T-cell non-Hodgkin lymphoma not otherwise speci-fied—had been submitted but were excluded from thisstudy because of their primary extracutaneous origin.

The clinical data, therapy, and outcome are pre-sented in Table 1. Table 2 displays the histopatho-logical, immunophenotypical, and genotypic features.

GROUP 1: GMF

Clinical Features

This group consisted of 15 patients, 9 men and 6 women,resulting in a male to female ratio of 1.5:1. Median age at

diagnosis was 48 years, with a broad range (20-72 years).In 2 patients, the disease had started in childhood beforeage 10 years. All patients in this group exhibited patchesand plaques (Figure 1), some of them with atrophy ofthe skin but without cutis laxa–like features. In 1 patient,the disease was restricted to a solitary plaque represent-ing unilesional MF. In 5 of 15 patients (33%), skin le-sions were hyperpigmented (Figure 2). At the time ofdiagnosis, 13 of 15 (87%) were in stage I or II accordingto the TNM staging system.14 First symptoms of the dis-ease had been reported to be present years or decades (me-dian, 11 years; range, 1-15 years) before the diagnosis ofGMF was established. Four patients had developed othertypes of lymphoid or myeloid neoplasms before or afterthe occurrence of GMF: Two patients had had nodal Hodg-kin lymphoma, nodular sclerosing type, 20 years beforeand 4 years after the occurrence of GMF. In the third pa-tient, nodal CD30� anaplastic large-cell lymphoma devel-

Table 2. Histologic, Immunophenotypic, and Genotypic Data From a Series of 19 Patients With Granulomatous CutaneousT-Cell Lymphomas

PatientNo.

GrowthPattern Epidermotropism Cell Size Granuloma

GiantCells

ElLoss Elastophagocytosis Eos Plasm Angio Phenotype Genotype

Granulomatous Mycosis Fungoidesa

1 Diffuse � S-M pleo � � � � � − � CD3�, 4−, 8�,30−; TIA-1�

TCR�

2 Diffuse − S-M pleo � � NA NA − − − CD3�, 4� NA3 Perivascular �/− Lining up S � � (Few) NA NA − − − CD3�, 4� TCR�

4 Perivascular − S − � (Few) NA NA − − − NA TCR�

5 Perivascular/periadnexal

� S-M pleo � (Few) � � − � − − CD3�, 4�, 8−,30−/�;TIA-1�/−

TCR�

6 Perifollicular/periadnexal

� S � � � − � − − CD3�, 4�, 8−,30−; TIA-1−

TCR�

7 Perivascular/nodular

− S-M-L pleo � − � − � (Few) � − CD3�, 4−,8�, 30−;TIA-1�/−

TCR�

8 Nodular − S-M pleo � − � − − − − CD3�, 4�, 30−;TIA-1�

TCR�

9 Nodular −/� S � − � − � − − CD3�, 4�, 8−,30−; TIA-1−

TCR�

10 Nodular − S � − � − � − − CD3�, 4−, 8� TCR−11 Diffuse − S − � � − − − � CD3�, 4�, 8−,

30− ;TIA-1−TCR�

12 Diffuse � S-M pleo � − � − � − � CD3�, 4�, 8−,30−; TIA-1−

TCR�

13 Diffuse −/� S � − � − � − � CD3�, 4�, 8−,30−; TIA-1−

TCR�

14 Nodular −/� S � − � − � � − CD3�, 4�, 8−,30−; TIA-1−

TCR�

15 Diffuse − S � � � − − − − CD3�, 4�, 8−,30−; TIA-1−

TCR�

Granulomatous Slack Skin16 Diffuse − S − � � − � (Few) � � CD3�, 4�, 8−,

30−; TIA-1−TCR�

17 Diffuse − S � � � � � (Few) � − CD3�, 4−,8�, 30−;TIA-1−;granzyme B−

TCR�

18 Diffuse − S-M pleo − � � � � − − CD3�, 4�, 8− TCR�

19 Diffuse � S-M pleo − � � − � � � CD3�, 4�, 8− TCR�

Abbreviations: Angio, angiocentric growth; El Loss, loss of elastic fibers on elastica staining; Eos, eosinophilic granulocytes; L, large; M, medium-sized; NA, notavailable; Plasm, plasma cells; pleo, pleomorphic; S, small; TIA-1, T-cell intracellular antigen 1; TCR, T-cell receptor rearrangement (�, monoclonal; −, polyclonal);�, present/positive; −, absent/negative; −/�, very few cells positive (�5%); �/−, few positive cells (10%-20%).

aAll except patient 3 had multiple lesions. Patient 6 had folliculotropic disease.

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1611

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

oped 4 years before the diagnosis of GMF, and the fourthpatient had had myeloid leukemia in childhood 21 yearsbefore the diagnosis of GMF. In all 4 patients, completeremission from nodal non-Hodgkin lymphomas or my-eloid leukemia was observed.

Treatment of GMF was heterogeneous, involving com-bined treatment with psoralen–UV-A and interferon alfain 7 patients. Three patients received chemotherapy withthe CHOP regimen (cyclophosphamide, doxorubicin, vin-cristine sulfate, and prednisolone acetate), whereas 2 pa-tients were treated with single-agent chemotherapy. Ra-diation was applied in 7 of 15 patients. Other treatmentmodalities included topical corticosteroids, imiquimod, andsystemic retinoids. Complete tumor regression was ob-served in only 3 of 15 patients (20%), but recurrence de-veloped within 2 years in 1 patient. In both patients withcomplete remission, the remission followed treatment withinterferon alfa. Progression of the disease was observed in6 of 15 patients (40%), and extracutaneous spread was ob-

served in 5 (33%) with involvement of lymph nodes, liver,and bone marrow (Table 1). In 3 of 15 patients (20%), trans-formation into CD30� large-cell phenotype was observed.Six of 15 patients (40%), including the 3 patients with trans-formation into a CD30� large-cell phenotype, died of lym-phoma after a median follow-up of 5.3 years (range, 1-20years) after diagnosis and 16 years (range, 2-54 years) af-ter onset of the disease, ie, the appearance of first symp-toms. Disease-specific 5-year survival rate in GMF was 66%.

Histologic Features

The infiltrate was diffuse in 6 of 15 cases (40%), nodularin 4 (27%) (Figure 3), and perivascular or periadnexalin 5 (33%), and it extended throughout the entire dermisin 8 cases (53%) and into the subcutis in 5 (33%). Epi-dermotropism of lymphocytes was a prominent feature inonly 4 cases (27%) and was subtle with only a few lym-phocytes in another 4. In the remaining 7 biopsy speci-mens (47%), epidermotropism of lymphocytes could notbe detected. The lymphocytic component of the infiltrateconsisted of small lymphocytes without significant nuclearatypia in 4 cases (27%), whereas small lymphocytes withcerebriformnucleiwere found in5(33%). In6cases (40%),tumor cells were small to medium-sized with pleomor-phic nuclei, and in 1 of these cases large pleomorphic lym-phocytes were intermingled with the predominant smalltomedium-sized tumorcells. Eosinophils werepresent andreadily identifiable in 9 of 15 cases (60%) (Figure4). Clus-ters of plasma cells, which were not related to overlyingulceration, were observed in 2 (13%) of the biopsy speci-mens. Granuloma formation with aggregations of histio-cytes was found in 13 of 15 cases (87%), and multinucle-ated histiocytic giant cells were present in 8 biopsyspecimens (53%) (Figure 4 and Figure 5). In all caseswith granuloma formation, there was a sarcoid-like pat-tern of granulomas (Figure 4), whereas a granuloma an-nulare–like pattern could not be found in any of the speci-mens. Granulomas were absent in 2 specimens, butnumerous multinucleated giant cells were scattered in adiffuse lymphocytic infiltrate in those 2 cases (Figure 6and Figure 7). In 4 cases (27%), infiltration of dermal or

Figure 1. Granulomatous mycosis fungoides showing erythematous patchesand plaques. Note that clinical features are not suggestive of granulomatoushistologic findings.

Figure 2. Granulomatous mycosis fungoides showing hyperpigmentedpatches.

Figure 3. Granulomatous mycosis fungoides showing dense nodularlymphocytic infiltrates throughout the entire dermis. Note the grenz zone and theabsence of epidermotropism (hematoxylin-eosin, original magnification �25).

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1612

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

subcutaneous vessels by lymphocytes was found and, in 2of these 4 cases, numerous multinucleated giant cells wereobserved around and within the walls of large veins in thesubcutis (Figure 8). Elastica staining was available in 12cases. Loss of elastic fibers throughout the infiltrated areaswas found in all 12 biopsy specimens, but elastophago-cytosis was found in only 1 of 12 specimens (8%).

Immunophenotype and Genotype

The lymphocytes expressed a CD3�, CD4�, CD8− phe-notype in 12 of 15 cases (80%). One of those cases showedexpression of TIA-1 by CD4� lymphocytes. Three cases(20%) exhibited a CD3�, CD4−, CD8� cytotoxic pheno-type. Clonal rearrangement of T-cell receptor � genes wasdetected by polymerase chain reaction in 13 (87%) of thebiopsy specimens.

GROUP 2: GSS

Clinical Features

The GSS group included 2 men and 2 women. Median ageat diagnosis was 46 years (range, 22-71 years). All 4 pa-

B

A

Figure 4. Granulomatous mycosis fungoides. A, Infiltrate of smalllymphocytes and granuloma formation (hematoxylin-eosin, originalmagnification �40). B, The lymphocytes exhibit nuclear atypia. Noteadmixed eosinophils (hematoxylin-eosin, original magnification �200).

Figure 5. Granulomatous mycosis fungoides showing infiltrate withhistiocytic multinucleated giant cells (hematoxylin-eosin, originalmagnification �100).

Figure 6. Granulomatous mycosis fungoides showing diffuse lymphocyticinfiltrate covering all dermal layers (hematoxylin-eosin, original magnification�25).

Figure 7. Granulomatous mycosis fungoides showing scatteredmultinucleated giant cells and absence of granuloma formation(hematoxylin-eosin, original magnification �20). Inset, Note the scatteredmultinucleated giant cells with phagocytosis of small lymphocytes(hematoxylin-eosin, original magnification �100).

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1613

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

tients showed poikilodermatous patches and plaques in theintertriginous areas (axillae and groins) with the develop-ment of characteristic bulky skin folds (Figure 9). In 1patient who had additional skin lesions on nonintertrigi-nous areas of the trunk, only the lesions in the axillae andgroin underwent cutis laxa–like changes, whereas skin le-sions at other sites did not evolve in a similar way. All pa-tients experienced an indolent, slowly progressive coursewithout extracutaneous spread and were alive with dis-ease after a median follow-up of 17 years. One patient de-veloped a second lymphoid neoplasia (CD30� lympho-proliferative disorder of the skin) after the occurrence ofGSS. Partial remission was achieved in 2 patients by pso-ralen–UV-A or topical carmustine. However, none of theother therapies, including surgical excision, topical corti-costeroid, and mechlorethamine hydrochloride, or sys-temic therapies such as interferon alfa, in combination withretinoids, was effective, and in none of the patients was com-plete tumor regression observed. All patients were alive af-ter a median follow-up of 17 years (range, 10-28 years),resulting in a 5-year survival rate of 100%.

Histologic Features

Five biopsy specimens of the 4 patients were availablefor histologic evaluation. In 4 of the 5 specimens therewas a diffuse lymphocytic infiltrate throughout the en-

tire dermis and the upper parts of the subcutis with nu-merous scattered multinucleated giant cells displayingmore than 10 nuclei per cell (Figure10 and Figure11).In addition, granuloma formation was identified in 1 of4 cases. One biopsy specimen exhibited a lichenoid in-filtrate of small to medium-sized lymphocytes mostly inthe upper and mid-dermis with sarcoid-like granulomaand a few giant cells. This pattern was not related to ini-tial disease manifestation because this biopsy specimenwas obtained from established lesions with hanging skinfolds. Epidermotropism of lymphocytes was present in

Figure 8. Infiltration of a large subcutaneous vein by the infiltrate containingmultinucleated giant cells (hematoxylin-eosin, original magnification �20).

Figure 9. Granulomatous slack skin showing bulky skin folds in the right axilla.

Figure 10. Granulomatous slack skin showing dense lymphocytic infiltratethroughout the entire dermis (hematoxylin-eosin, original magnification �25).

Figure 11. Granulomatous slack skin showing numerous multinucleatedgiant cells but lack of granuloma formation (hematoxylin-eosin, originalmagnification �100).

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1614

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

only 1 case, but absent or very subtle with only a few scat-tered intraepidermal lymphocytes in the remaining 3 cases.The lymphocytes were small without nuclear atypia in2 cases and small to medium-sized pleomorphic cells inthe remaining 2 cases. Eosinophils could be found in allspecimens, and mostly scattered plasma cells were pres-ent in 3 of 4 cases. In 2 cases, infiltration of large veinsin the subcutis by multinucleated giant cells and lym-phocytes was observed. Elastica staining demonstratedloss of elastic fibers in the infiltrated area in all cases. Elas-tophagocytosis was subtle and found in only 2 of 4 cases.

Immunophenotype and Genotype

In 3 of 4 patients, lymphocytes displayed a CD3� CD4�

CD8− phenotype. In contrast, 1 case showed a CD3� CD4−

CD8� phenotype. Monoclonal rearrangement of T-cellreceptor � genes could be demonstrated by polymerasechain reaction in all cases. In addition, monoclonal re-arrangement of T-cell receptor � genes was demon-strated by Southern blot technique in 1 of the cases.

COMMENT

In our series, GMF was the most common disorder, ac-counting for 79% of the analyzed cases. Patients affectedby GMF and GSS were on average diagnosed in their fifthdecade of life, with a male predominance in GMF. In bothdisorders, the disease extent and distribution of skin le-sions at diagnosis corresponded to TNM stage Ia in themajority of patients (11 of 19 [58%]), with the trunk rep-resenting the predilection site (Figure 1). The clinical pre-sentation of skin lesions in GMF was not indicative of thehistologically detected granulomatous features (Figure 1).One-third of the patients with GMF manifested hyperpig-mented skin lesions (Figure 2). Hyperpigmented MF hasbeen reported as a common feature in CD8� MF,17 but noneof the cases with hyperpigmented skin lesions in our se-ries displayed a CD8� cytotoxic T-cell phenotype.

Psoralen–UV-A and/or interferon alfa in addition toradiotherapy were the most commonly used treatmentmodalities. Complete tumor regression could be achievedin only 3 of 15 patients (20%) with GMF. In half of thepatients, the disease showed a slowly progressive course.

Extracutaneous spread was observed in a third of pa-tients with GMF and was associated with transforma-tion into CD30� anaplastic large-cell lymphoma in 20%of the patients. Six of 15 patients (40%) with GMF, in-cluding the 3 patients with large-cell transformation, diedof the disease. The percentage of patients with unfavor-able outcome is identical to that in the study by Chen etal,13 who reported death due to the disease in 40% of pa-tients with GMF. Whereas the occurrence of granulo-mas in MF has been considered to be associated with afavorable prognosis by some authors,6,18 other investiga-tors could not confirm this observation.19-21 These andour findings demonstrate that GMF is not associated witha better prognosis than classic, nongranulomatous MF.In fact, the 5-year survival rate of GMF (66%) is worsethan that of classic MF and similar to that of folliculo-tropic MF.22,23

In contrast, all patients with GSS were still alive aftera median follow-up of 17 years despite the fact that GSSwas therapy resistant and complete remission could notbe achieved in any of the patients. Recently, response ofGSS to topical mechlorethamine was observed in 2 pa-tients,24 but definitive therapy for GSS has yet to be es-tablished. Extracutaneous spread is exceedingly rare inGSS and did not occur in our series of patients with GSS.

Patients with GMF and GSS are known to be at risk forthe development of second lymphoid neoplasias. In ourseries, 4 of 19 patients (21%) with GMF or GSS had ex-perienced a second lymphoma before or after occurrenceof granulomatous CTCLs, and an additional patient hadhad myeloid leukemia. This prevalence is lower than thatreported in the literature, with 48% of the patients withGSS having second lymphoid neoplasias.10,25,26 These pa-tients may be overrepresented in the literature because ofthe development of a second lymphoma and, eventually,misinterpretation of large-cell transformation as develop-ment of a second anaplastic large-cell lymphoma unre-lated to GSS. Hodgkin lymphoma is the most common sec-ond neoplasia in patients with granulomatous CTCLs inthe literature as well as in our series.10,25,26 The interval be-tween lymphoid neoplasias and GMF or GSS may be yearsor even decades, as seen in 1 patient in our series who de-veloped nodal Hodgkin lymphoma 20 years before onsetof GMF. Thus, lifelong observation of patients with GMFand GSS is required.

Histologically, a diffuse infiltrate of lymphocytes ex-tending throughout the entire dermis and the subcutis wasthe most common growth pattern (Figure 4). Granulo-mas were sarcoid-like in all biopsy specimens (Figure 6).Other patterns of granulomatous reactions, such as granu-loma annulare–like, palisaded, or necrobiotic granuloma,as reported in the literature,27-30 were not found in our se-ries. In 70% of the CTCL cases, the infiltrates containedhistiocytic giant cells displaying 10 or more nuclei(Figure 11). Remarkably, infiltration of the vessel walls oflarge veins by small lymphocytes and even by large mul-tinucleated histiocytic giant cells was observed in a thirdof the cases (Figure 8). This is more common than re-ported in the literature.4,5 Epidermotropism of lympho-cytes was previously reported as a common finding in granu-lomatous CTCLs and considered to be a useful histologicfeature for discrimination of granulomatous CTCLs fromreactive granulomatous disorders.5 In our series, how-ever, epidermotropism was absent in almost half of the cases,limiting its value as a diagnostic marker in GMF and GSS.Although classic MF commonly displays epidermotro-pism, this histologic feature is not a prerequisite for MFaccording to the WHO-EORTC classification. Thus, lackof epidermotropism does not exclude the diagnosis of MF.In those cases, diagnosis of MF relies on the characteristicclinical presentation with patches and plaques.

Diagnosis in granulomatous CTCLs is often delayed,with a latency of years to decades after the onset of initialsymptoms. Diagnosis is particularly challenging in caseswith predominant granuloma formation in the absence ofnuclear atypia or epidermotropism of tumor cells. Detec-tion of a neoplastic T-cell clone, which was present in nearlyall cases, may thus be a useful diagnostic adjunct in granu-lomatous CTCLs.

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1615

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

The histologic features of GSS have been reported aspathognomonic, with a diffuse lymphocytic infiltrate har-boring numerous scattered multinucleated giant cells5,31

(Figures 6 and 7), but identical histologic features canalso be observed in GMF.32 Remarkably, this pattern wasalso found in 2 of the 15 GMF cases in our series (Figures 6and 7) and is therefore not pathognomonic for GSS. Theseobservations demonstrate that GMF and GSS differ clini-cally but show overlapping histologic findings and there-fore cannot be discriminated by histologic examinationalone. It should be recalled that both diseases are con-sidered variants of a single disease,33 which implies thatGSS may be listed in future classifications for cutaneouslymphomas as another variant and not a subtype of MF.As emphasized by Scarabello and coworkers,4 diagnosisof GSS should rest on clinical grounds and be restrictedto patients presenting clinically with characteristic bulkyskin lesions.

The classic pathogenetic concept links the develop-ment of hanging skin folds in GSS to destruction of elas-tic fibers due to elastophagocytosis by histiocytic giantcells. However, loss of elastic fibers was found in all ex-amined cases. The extent of loss of elastic fibers corre-lated with the extent of the granulomatous infiltrate butwas not restricted to GSS. Moreover, only skin lesionsin skin folds such as the axilla and the groin underwentcutis laxa–like changes, whereas skin lesions present atother body sites did not evolve in a similar way. Theseobservations suggest that development of hanging skinfolds is a location-related phenomenon and not solely theresult of the destruction of elastic fibers. Hypotheti-cally, the continuous stretching of elastic fibers in the in-tertriginous body areas during physiologic movementsmay facilitate the loss of their function when these re-gions become affected by lymphomatous infiltrates.

The pathogenetic mechanisms of granuloma forma-tion in lymphoid neoplasms are poorly understood. Granu-lomatous reaction has been regarded by some authors asa local tissue response to the infiltrating malignant cellsor their antigens,34 but this hypothesis has been criticizedby others on the basis of the occurrence of granulomas inhistologically lymphoma-free tissues.35 In addition, treat-ment with interferon alfa or bexarotene36 may induce sar-coid-like reactions. However, granuloma formation in ourseries was not related to previous therapy.

Recently, genetic alterations with t(3;9)(q12;p24) havebeen reported in a case of GSS,37 which may indicate ge-netic predisposition to granuloma formation. The patho-genetic processes underlying granuloma formation ingranulomatous CTCLs remain to be elucidated.

Accepted for Publication: February 19, 2008.Author Affiliations: Departments of Dermatology, Uni-versity Hospital, Zürich, Switzerland (Drs Kempf, Os-theeren-Michaelis, Golling, Dummer, and Burg), HopitalHenriMondor,Creteil,France (DrWechsler),Charite, Ber-lin, Germany (Drs Audring and Assaf), Leiden Univer-sity, Medical Center, Leiden, the Netherlands (DrWillemze),UniversityMilan-Bicocca,Milan, Italy(DrBerti),Medical University of Graz, Graz, Austria (Dr Cerroni),University Hospital Tübingen, Tübingen, Germany (DrBerneburg), University of Rome, Rome, Italy (Dr

Chimenti), Semmelweis Medical School, Budapest, Hun-gary (Dr Marschalko), Necker-Enfants Malades Hopital,Paris, France (Dr Fraitag), and Hopital Tarnier Cochin,Paris, France (Dr Carlotti); Departments of Pathology, Uni-versity of Pavia, Pavia, Italy (Drs Paulli and Lucioni), Ho-pital Henri Mondor, Creteil, France (Dr Wechsler), Luit-pold Hospital, Würzburg, Germany (Dr Rüdiger), VrijeUniversiteit Medical Center, Amsterdam, the Nether-lands (Dr Meijer), University Hospital, Djion, France (DrsPetrellaandCourville), andCantonalHospital,Aarau,Swit-zerland (Dr Laeng); Department of Human Pathology andOncology, University of Florence Medical School, Flo-rence, Italy (Dr Santucci); Hautklinik Linden, Hannover,Germany (Dr Hallermann); Skin Tumour Unit, St John’sInstitute of Dermatology, St Thomas’ Hospital, London,England (Dr Robson); Sikl’s Department of Pathology,CharlesUniversity,MedicalFacultyHospital,Pilsen,CzechRepublic (Dr Kazakov); and Division of Special and En-vironmental Dermatology, Department of Dermatology,Vienna, Austria (Dr Knobler).Correspondence: Werner Kempf, MD, Department ofDermatology, University Hospital Zürich, Gloriastrasse 31,CH-8091 Zürich, Switzerland (kempf@derm.unizh.ch).Author Contributions: All authors had full access toall of the data in the study and take responsibility forthe integrity of the data and the accuracy of the dataanalysis. Study concept and design: Kempf, Burg, andWillemze. Acquisition of data: All authors. Analysis andinterpretation of data: Kempf, Ostheeren-Michaelis,and Burg. Drafting of the manuscript: Kempf. Criticalrevision of the manuscript for important intellectual con-tent: All authors. Statistical analysis: Kempf. Study super-vision: Kempf, Burg, and Willemze.Financial Disclosure: None reported.Additional Contributions: We thank Beatrix Mueller forthe excellent technical assistance.

REFERENCES

1. Sacks EL, Donaldson SS, Gordon J, Dorfman RF. Epithelioid granulomas asso-ciated with Hodgkin’s disease: clinical correlations in 55 previously untreatedpatients. Cancer. 1978;41(2):562-567.

2. Kadin ME, Donaldson SS, Dorfman RF. Isolated granulomas in Hodgkin’s disease.N Engl J Med. 1970;283(16):859-861.

3. Shapiro PE, Pinto FJ. The histologic spectrum of mycosis fungoides/Sezary syn-drome (cutaneous T-cell lymphoma): a review of 222 biopsies, including newlydescribed patterns and the earliest pathologic changes. Am J Surg Pathol. 1994;18(7):645-667.

4. Scarabello A, Leinweber B, Ardigo M, et al. Cutaneous lymphomas with promi-nent granulomatous reaction: a potential pitfall in the histopathologic diagnosisof cutaneous T- and B-cell lymphomas. Am J Surg Pathol. 2002;26(10):1259-1268.

5. LeBoit PE, Zackheim HS, White CJ. Granulomatous variants of cutaneous T-celllymphoma: the histopathology of granulomatous mycosis fungoides and granu-lomatous slack skin. Am J Surg Pathol. 1988;12(2):83-95.

6. Ackerman AB, Flaxman BA. Granulomatous mycosis fungoides. Br J Dermatol.1970;82(4):397-401.

7. Carrozza PM, Kempf W, Kazakov DV, Dummer R, Burg G. A case of Sezary’s syn-drome associated with granulomatous lesions, myelodysplastic syndrome andtransformation into CD30-positive large-cell pleomorphic lymphoma. Br J Dermatol.2002;147(3):582-586.

8. Boulier I, Cohen J, Kerdel FA. Cutaneous large-cell lymphoma histologically re-sembling sarcoidosis. J Am Acad Dermatol. 1993;28(2, pt 2):327-330.

9. Balus L, Manente L, Remotti D, Grammatico P, Bellocci M. Granulomatous slackskin: report of a case and review of the literature. Am J Dermatopathol. 1996;18(2):199-206.

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1616

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from

10. Clarijs M, Poot F, Laka A, Pirard C, Bourlond A. Granulomatous slack skin: treat-ment with extensive surgery and review of the literature. Dermatology. 2003;206(4):393-397.

11. Willemze R, Jaffe ES, Burg G, et al. WHO-EORTC classification for cutaneouslymphomas. Blood. 2005;105(10):3768-3785.

12. Burg G, Jaffe ES, Kempf W, et al. WHO/EORTC classification of cutaneouslymphomas. In: LeBoit P, Burg G, Weedon D, Sarasin A, eds. WHO Books: Tu-mors of the Skin. Lyon, France: WHO IARC; 2005.

13. Chen KR, Tanaka M, Miyakawa S. Granulomatous mycosis fungoides with smallintestinal involvement and a fatal outcome. Br J Dermatol. 1998;138(3):522-525.

14. International Union Against Cancer. TNM Classification of Malignant Tumors. 6thed. New York, NY: Wiley-Liss; 2002.

15. Whittaker SJ, Smith NP, Jones RR, Luzzatto L. Analysis of beta, gamma, anddelta T-cell receptor genes in mycosis fungoides and Sezary syndrome. Cancer.1991;68(7):1572-1582.

16. Wood GS, Tung RM, Haeffner AC, et al. Detection of clonal T-cell receptor gammagene rearrangements in early mycosis fungoides/Sezary syndrome by polymer-ase chain reaction and denaturing gradient gel electrophoresis (PCR/DGGE).J Invest Dermatol. 1994;103(1):34-41.

17. Dummer R, Kamarashev J, Kempf W, Haffner AC, Hess-Schmid M, Burg G.Junctional CD8� cutaneous lymphomas with nonaggressive clinical behavior: aCD8� variant of mycosis fungoides? Arch Dermatol. 2002;138(2):199-203.

18. Schwartz RA, Burgess GH, Holtermann OA. Mycosis fungoides associated withflorid sarcoid reactions. J Surg Oncol. 1980;14(4):347-357.

19. Schewach-Millet M, Azizi E, Semah D, Bubis JJ, Chaitchuk S. Cutaneous T-celllymphoma first presenting at the tumor stage: a clinicopathologic study. Int JDermatol. 1984;23(4):275-278.

20. Dabski K, Stoll HJ. Granulomatous reactions in mycosis fungoides. J Surg Oncol.1987;34(4):217-229.

21. Papadavid E, Yu RC, Bunker C, Scoones D, Chu AC. Tumour progression in apatient with granulomatous mycosis fungoides. Br J Dermatol. 1996;134(4):740-743.

22. Willemze R, Kerl H, Sterry W, et al. EORTC classification for primary cutaneouslymphomas: a proposal from the Cutaneous Lymphoma Study Group of the Eu-ropean Organization for Research and Treatment of Cancer. Blood. 1997;90(1):354-371.

23. Kim YH, Liu HL, Mraz-Gernhard S, Varghese A, Hoppe RT. Long-term outcome

of 525 patients with mycosis fungoides and Sezary syndrome: clinical prognos-tic factors and risk for disease progression. Arch Dermatol. 2003;139(7):857-866.

24. Hultgren TL, Jones D, Duvic M. Topical nitrogen mustard for the treatment ofgranulomatous slack skin. Am J Clin Dermatol. 2007;8(1):51-54.

25. Noto G, Pravata G, Miceli S, Arico M. Granulomatous slack skin: report of a caseassociated with Hodgkin’s disease and a review of the literature. Br J Dermatol.1994;131(2):275-279.

26. van Haselen CW, Toonstra J, van der Putte SJ, van Dongen JJ, van Hees CL, vanVloten WA. Granulomatous slack skin: report of three patients with an updatedreview of the literature. Dermatology. 1998;196(4):382-391.

27. Barksdale SK, Perniciaro C, Halling KC, Strickler JG. Granuloma annulare in pa-tients with malignant lymphoma: clinicopathologic study of thirteen new cases.J Am Acad Dermatol. 1994;31(1):42-48.

28. Garrie SA, Hirsch P, Levan N. Granuloma annulare–like pattern in mycosis fungoides.Arch Dermatol. 1972;105(5):717-719.

29. Miyamoto T, Mihara M. Subcutaneous granuloma annulare with Hodgkin’s disease.J Dermatol. 1996;23(6):405-407.

30. Rongioletti F, Cerroni L, Massone C, Basso M, Ciambellotti A, Rebora A. Differ-ent histologic patterns of cutaneous granulomas in systemic lymphoma. J AmAcad Dermatol. 2004;51(4):600-605.

31. LeBoit PE. Granulomatous slack skin. Dermatol Clin. 1994;12(2):375-389.32. Metzler G, Schlagenhauff B, Krober SM, Kaiserling E, Schaumburg-Lever G, Lischka

G. Granulomatous mycosis fungoides: report of a case with some histopatho-logic features of granulomatous slack skin. Am J Dermatopathol. 1999;21(2):156-160.

33. Topar G, Zelger B, Schmuth M, Romani N, Thaler J, Sepp N. Granulomatous slackskin: a distinct disorder or a variant of mycosis fungoides? Acta Derm Venereol.2001;81(1):42-44.

34. Kahn LB, Gordon W, Camp R. Florid sarcoid reaction associated with lymphomaof the skin. Cancer. 1974;33(4):1117-1122.

35. Anderson R, Geraint James D, Peters PM, Thomson AD. Local sarcoid-tissuereactions. Lancet. 1962;1(7241):1211-1213.

36. Ruiz-de-Casas A, Carrizosa-Esquivel A, Herrera-Saval A, Rios-Martın JJ, Cama-cho F. Sezary syndrome associated with granulomatous lesions during treat-ment with bexarotene. Br J Dermatol. 2006;154(2):372-374.

37. Ikonomou IM, Aamot HV, Heim S, Fossa A, Delabie J. Granulomatous slack skinwith a translocation t(3;9)(q12;p24). Am J Surg Pathol. 2007;31(5):803-806.

Archives Web Quiz Winner

C ongratulations to the winner of our Septemberquiz, Rajesh Chhetia, MD, Dermatology Consul-

tant, Sanchaiti Hospital, Kandivali, Mumbai, India. Thecorrect answer to our September challenge was lipoid pro-teinosis. For a complete discussion of this case, see theOff-Center Fold section in the October Archives (ColeJA, Novosel TA, Williams JV. Pocklike scarring and sub-lingual papules in a child. Arch Dermatol. 2008;144[10]:1383-1388).

Be sure to visit the Archives of Dermatology Web site(http://www.archdermatol.com) to try your hand at theinteractive quiz. We invite visitors to make a diagnosisbased on selected information from a case report or otherfeature scheduled to be published in the following month’sprint edition of the Archives. The first visitor to e-mailour Web editors with the correct answer will be recog-nized in the print journal and on our Web site and willalso receive a free copy of The Art of JAMA II.

(REPRINTED) ARCH DERMATOL/ VOL 144 (NO. 12), DEC 2008 WWW.ARCHDERMATOL.COM1617

©2008 American Medical Association. All rights reserved. at Bibliosan, on March 7, 2012 www.archdermatol.comDownloaded from