You are here
T-Cell Receptor-γ in Gamma-Delta Phenotype Cutaneous T-Cell Lymphoma Can Be Accompanied by Atypical Expression of CD30, CD4, or TCRβF1 and an Indolent Clinical Course
Clinical Lymphoma Myeloma and Leukemia
- Diagnosis of cutaneous gamma-delta T-cell lymphoma (CGDTCL) using clinical findings alone is nearly impossible because of the resemblance of the skin lesions to other diseases on the cutaneous T-cell lymphoma (TCL) spectrum.
- The wide spectrum of clinical and histological presentations of the CGDTCL patients presented herein challenges the notion of CGDTCL as a distinct entity with universally poor prognosis, because clinical presentation and pathology can overlap with other cutaneous TCLs with less aggressive clinical courses.
- In consideration of our data in cases 1, 3, 5, 6, 7, and 9, we conclude that T-cell receptor (TCR)-γ expression alone does not condemn patients to aggressive progression of CGDTCL because the TCRγ-positive (TCRγ+) finding can be accompanied by markers that are typical of less aggressive disease, namely CD30, CD4, or βF1 expression.
- CD30 expression in particular can be used quite successfully as a therapeutic target and CD4+ cases respond better to skin-directed regimens than CD4− cases.
- Additionally, we show in cases 2 and 4 that even when a classic CGDTCL immunophenotype is present, early aggressive therapy can still induce an indolent clinical course, which highlights the importance of early CGDTCL diagnosis, often requiring multiple biopsies.
Cutaneous gamma-delta T-cell lymphoma (CGDTCL) is a rare form of T-cell lymphoma that usually exhibits an aggressive clinical phenotype. To further characterize the histopathologic and clinical features of CGDTCL we identified a retrospective case series of 9 patients from our database. As has been reported elsewhere 1 , the clinical features were heterogeneous, and the diagnosis of CGDTCL was based on extensive immunophenotypic characterization, including T-cell receptor (TCR)-γ expression. The novel finding of this study is the presence of unusual immunophenotypical profiles in most of these patients, including CD4-postive (CD4+), CD30+ or TCR-βF1+, which are typical of less aggressive cutaneous T-cell lymphoma (TCL). In the minority of cases, with the classic CGDTCL immunophenotype profile, aggressive therapy initiated early induced resolution of lesions in 1 patient. Our data suggest that CGDTCLs diagnosed according to TCRγ expression are not always associated with the aggressive phenotype. These findings call into question the diagnostic power of TCRγ alone as a marker of CGDTCL, especially in cases where certain immune markers are also present. The presence of these markers might explain the indolent disease course in these patients.
The γδ TCLs are lymphomas that express γδ TCRs 2 and are rare, aggressive cancers. Primary CGDTCL accounts for < 1% of all cutaneous TCLs. 3 γδ TCLs originate from γδ T-cells, which regulate inflammation and epidermal homeostasis. 4 Lesions of CGDTCL are typically disseminated with diffuse skin involvement 2 and reported median survival time from diagnosis ranges from 15 months to 31 months.1 and 5
Cutaneous gamma-delta T-cell lymphoma is characterized by the expression of CD3 and CD2 with variable CD7 expression. 2 Activated γδ T-cells might express CD56, associated with natural killer cells, and also might express cytotoxic markers such as T-cell intracellular antigen-1 (TIA-1) and granzyme B. 2 The classic CGDTCL tumor cells do not express αβ T-cell chains and thus do not stain for TCR-βF1 and are often double-negative for CD4 and CD8. 6 Immunophenotypic diagnostic criteria for CGDTCL are not unequivocally established, and range from absence of αβ TCR expression to positive immunohistochemistry of TCRγ.
Materials and Methods
We retrospectively reviewed the medical records of 9 patients with the γδ T-cell lymphoma phenotype, diagnosed based on clinical features and skin biopsies expressing TCRγ. Their demographic characteristics, clinical courses, histopathologic features, and response to treatment were summarized, along with polymerase chain reaction (PCR) results for TCR rearrangements.
In this case series of 9 patients, 4 were male and 5 were female with a median age of 57.5 (range 1.5 to approximately 88) years at the time of diagnosis. Four patients had hypertension, 3 had irritable bowel syndrome, and 2 had hypothyroidism. History of cancer was present in 2 cases. Case 3 had a history of marginal zone B-cell lymphoma of the spleen and lymph nodes, and case 4 had a history of cervical cancer. Previous skin problems were present in case 5 (history of psoriasis), case 8 (history of melanoma and actinic keratosis), and case 9 (history of mycosis fungoides [MF], lichen simplex chronicus, and alopecia).
Data are summarized in Table 1 . The presenting symptom was a papular rash in 3 cases, erythematous patches in 3 cases, a subcutaneous nodule in 1 case, and a diffuse, papulosquamous MF-like rash in 1 case. In 1 of the cases (case 4) with a papular rash, the lesions were initially small and eventually developed into larger and ulcerated tumors. Interestingly, 3 patients initially attributed their papules to insect bite-like lesions. The interval between onset of symptoms and diagnostic biopsy was generally > 1 year (range, 1-121 months, median 43.3 months). Most of the initial lesions (n = 6) presented on the extremities with the next most common presenting site being the trunk (n = 4). Only 1 case (case 9) began on the face, and no cases had initial involvement of the scalp.
|Case||Race||Sex||Initial Lesion||Initial Suspicion||Initial Location||Age, Years||Lesion at Diagnosis||Areas Affected||Ulcer|
|1||Caucasian||F||N||–||U/l calf||64||SC nodules||B/l l lower leg; U/l upper arm and buttock||No|
|2||Chinese||M||Pa||E||Unspecified||1.5||Violaceous patches/plaques||B/l legs and buttocks; U/l ankle and upper arm||No|
|3||Caucasian||F||R||MF, Ps||Diffuse (body)||88||Plaques and scaly patches||B/l extremities and trunk; U/l axilla and ear||Yes|
|4||Black||F||Pa||MF||U/l thenar eminence||51||Patches, plaques, violaceous papules||U/l breast, nostril, pubis, heel, lower leg||Yes|
|5||Black||M||P||MF||Back||39||Hyperpigmented plaques and patches||Total body||Yes|
|6||Middle-Eastern||F||P||–||U/l lower leg||74||Plaques and ulcer||U/l lower extremity, flank, arm; both buttocks||Yes|
|7||Caucasian||M||Pa||E, Ps, CD||B/l extremities||42||Patches with central clearing and plaques||Extremities only||No|
|8||Caucasian||M||P||E, MF||U/l hip||82||Thick, erythematous plaques and tumors||U/l hip, flank, lower leg, soles||Yes|
|9||Black||F||P||PA, MF, Ps||B/l buttocks, eye, abdomen||42||Hypopigmented patches||B/l diffuse body; U/l scalp||No|
Age listed is age at diagnosis, based on time of TCRγ-positive biopsy.
Abbreviations: B/l = bilateral; CD = contact dermatitis; E = eczema; F = female; M = male; MF = mycosis fungoides; N = subcutaneous nodule; P = patch; Pa = papule; PA = pityriasis alba; Ps = psoriasis; R = rash; SC = subcutaneous; TCR = T-cell receptor; U/l = unilateral.
As the disease progressed, the most common site of involvement continued to be the extremities (n = 9), followed by the trunk (n = 6). Three patients also developed lesions on the face and/or scalp. All patients eventually developed lower limb involvement and 7 had concurrent upper limb involvement.
Case 1 had subcutaneous nodular lesions, which persisted. Other patients' lesions became patches, plaques, or a combination, and 5 of the other 8 patients experienced ulceration. In general, the patches and plaques were erythematous (n = 5), with 3 patients having violaceous plaques. Two of the patients' plaques were specifically noted to have ill-defined borders (1 of these had induration, the other had central clearing) and 1 had lesions that were well-demarcated.
Clinically, the diagnosis was suspected to be MF in 5 cases. In 5 cases, the clinical presentation was initially suspicious for eczema or psoriasis. Data are summarized in Table 1 .
All 9 patients had 1 or more lesional skin biopsies taken from characteristic skin lesions. In all but 3 cases, the biopsy results initially supported other dermatologic conditions, making it clear that multiple biopsies were needed to obtain the correct diagnosis. The final diagnosis of γδ T-cell lymphoma was established according to the histopathological and immunohistochemical findings of the biopsy tissue—specifically, TCRγ+ and TCRβF1−. TCRγ+ is important to note because TCR-silent lymphomas exist, where expression of αβ and γδ are both negative.7 and 8
All cases exhibited an atypical lymphoid infiltrate involving the dermis (n = 3), epidermis (n = 1), or both (n = 3), with 2 cases involving the dermis and subcutaneous tissue. In 7 cases there was epidermotropism, which is a diagnostic criteria for MF. Case 1 was the only case with a subcutaneous granulomatous infiltrate with admixed tumor cells characterized by hyperchromatic nuclei and scant eosinophilic cytoplasm. The tumor cells ranged from small (n = 6) to medium (n = 6) to large (n = 2), with many cases showing a range of cell sizes. Two cases showed extravasation of red blood cells. Of note, histopathologically, case 7 lacked some of the typical features of CGDTCL, such as conspicuous dyskeratotic keratinocytes. Complete pathological descriptions are noted in Table 2 .
|Case||Infiltrate||Depth||EDT||Cell Size||Previous Biopsies|
|1||Granulomatous with lymphocytes||SC, D||No||S/M||αB-panniculitic TCL; granulomatous MF|
|2||Lymphocytes with irregular nuclear contours||D, E||Yes||S/M||None|
|3||Atypical lymphocytes||SC, D||Yes||S/M, L||None|
|4||Atypical lymphocytes with cytologic atypia||D, E||Yes||M/L||Pagetoid reticulosis type TCL; CD8+ lymphoproliferative TCL|
|5||Atypical lymphocytes with syringotropism||D||NA||NA||MF|
|6||Lymphocytes, rare germinative centers||D||Yes||S, M, L||Granuloma annulare|
|7||Lymphocytes with mild cytologic atypia||NA||Yes||S||Contact dermatitis, MF|
|8||Atypical lymphocytes with hyperchromatic nuclei||D, E||Yes||NA||MF in 3 biopsies|
|9||Atypical lymphocytes with mild cytologic atypia||D||Yes||S/M||MF|
Abbreviations: D = dermis; E = epidermis; EDT = epidermotropism; L = large; M = medium; MF = mycosis fungoides; S = small; SC = subcutaneous; TCL = T-cell lymphoma.
Immunophenotyping of the infiltrates was available for all of the 9 cases. The atypical lymphocytes were positive for CD3. All 9 cases had positive staining for TCR-gamma chain but 3 of these had concurrent expression of βF1, with the remaining 6 patients characterized as βF1−. It is believed that expression of TCRαβ or TCRγδ heterodimers is mutually exclusive, 9 so these 3 patients did not fit the generally accepted clinical standards and are puzzling. Of the 8 samples tested for CD30, 6 were positive to varying degrees. Interestingly, the infiltrates were CD8+ in 4 patients and CD4+ in 3 patients. Of the 5 patient samples tested, none were Epstein-Barr virus-positive, which is consistent with previous reports. 3 Granzyme B and TIA were coexpressed in 6 of the patients tested. Case 7 again lacked definitive expression of cytotoxic markers. CD5 expression was positive in 3 of 6 patients and CD7 expression was positive in 3 of 8 patients.
T-cell receptor γ chain gene rearrangement analysis was performed using PCR and detected in all 7 tested patient biopsies. TCR-γ amplicons used the V-gamma 1 family only in 2 cases, the V1,2 family in 1 case, the V1,3 family in 1 case, and the V1,2,3 family in another. Of these, 3 were also positive for rearrangements in TCR-β. Data are summarized in Table 3 .
|Case||CD3||CD5||CD7||CD4||CD8||CD56||CD30||EBER||GZB||TIA-1||βF1||TCR-γ||TCR-β PCR a||TCR-γ PCR a|
a No indicates no rearrangement detected, yes indicates rearrangement detected.
Abbreviations: EBER = EBV-encoded RNA; GZB = Granzyme B; PCR = polymerase chain reaction; R = rare; S = significant; TCR = T-cell receptor; TIA1 = T-cell intracellular antigen-1.
Treatment and Response
Average survival from onset of symptoms was 87.4 months (7.3 years) and average survival from date of diagnosis was 9.8 months. The 9 CGDTCL patients are still living, and 1 is in complete remission.
In 8 of the patients, initial treatment was topical steroids and was ineffective in all as monotherapy. Oral prednisone was effective in treating the subcutaneous nodules when combined with bexarotene in case 1. The histone deacetylase (HDAC) inhibitor vorinostat was used in cases 5 and 8. Although there was an excellent response to vorinostat in case 5, the medication was unfortunately discontinued because of an increase in creatinine level. Another HDAC inhibitor, romidepsin, provided some clinical benefit for this patient although not as dramatic as with vorinostat.
Cutaneous chemotherapy with topical nitrogen mustard was prescribed in 2 cases as an adjuvant therapy with limited response. In general, nitrogen mustard served to stabilize or improve existing lesions but did not prevent formation of new lesions. Cases 1, 5, 8, and 10 underwent therapy with the oral retinoid X receptor, bexarotene, but only half of the patients experienced the clinical benefit of either flattened lesions or reduced erythema.
Multi-agent chemotherapy is often used to treat CGDTCL because of its aggressive behavior. 10 When oral chemotherapy failed in case 5, intravenous chemotherapy with gemcitabine provided no response. Aggressive treatment was initiated in case 2 because of the age of the patient. He had no response to cyclophosphamide, doxorubicine, vincristine and prednisone (CHOP) with bleomycin but a complete response to Berlin-Frankfurt-Munster (BFM) therapy with non-Hodgkin lymphoma Berlin-Frankfurt-Münster-95 (NHL-BFM-95) that was partially stabilized with gemcitabine. Case 4 was also treated aggressively because of her definitive CGDTCL immunophenotype; she underwent chemotherapy with rescue stem-cell transplant and completed only 2 cycles of rituximab, cyclophosphamide, vincristine, doxorubicin, dexamethasone, methotrexate, cytarabine (R-hyper-CVAD) with alternating methotrexate and cytarabine before having complete resolution of skin lesions. In all patients who received targeted radiation therapy, palliation of the lesions was achieved.
Targeted anti-CD30 therapy with brentuximab vedotin (BV) was initiated for the MF-like lesions in case 3 after radiation of a tumor. There was a partial response but after 7 treatments the drug was stopped because of Grade 2 neuropathy. Because BV might have a synergistic effect with chemotherapy, 11 this could be a useful therapy for CD30+ CGDTCLs. Case 10 also had diffuse CD30 staining but has not received BV.
Complete treatment and response summaries are listed in Table 4 . Although not included in this table, the benefits of wound management in these patients cannot be understated. In several patients, Burrow's solution soaks and whirlpool therapy provided improvement, and treatment of secondary infection is assumed to have prevented further complications.
|Bexarotene and prednisone||PR|
|2||CHOP and bleomycin (2 cycles)||NR|
|NHL-BFM-95 (4 cycles)||CR, then relapse|
|Gemcitabine and clobetasol||PR|
|3||Triamcinolone and hydrocortisone||NR|
|Radiation therapy: axilla||LR|
|Brentuximab and bleach baths||PR|
|Brentubimab discontinued midcourse (peripheral neuropathy)||Relapse|
|Radiation therapy: scalp||LR|
|Radiation therapy: neck, perineum, inner thigh||LR|
|R-hyper-CVAD alternating with methotrexate, cytarabine||CR and PIH|
|5||Clobetasol and triamcinolone||NR|
|Vorinostat and narrow band UVB||PR|
|Vorinostat decreased (increased creatinine)||Relapse|
|Gemcitabine (3-4 cycles)||NR|
|Total skin electron beam therapy and whirlpool||PR|
|6||Triamcinolone||PR, then relapse|
|Radiation therapy and nitrogen mustard||LR, then relapse|
|Burrrow's solution soaks||NR|
|EPOCH||CR, then relapse|
|Triamcinolone (discontinued, allergic reaction)||NR|
|Nitrogen mustard and bleach baths||PR|
|Bexarotene decreased (hyperlipidemia)||Relapse|
|Bexarotene and vorinostat||PR|
|Low dose skin electron beam||PR|
|Bexarotene only||PR, Relapse|
|Narrow band UVB||PR|
|Narrow band UVB discontinued (photosensitivity)||Relapse|
Abbreviations: CHOP = cyclophosphamode, doxorubicin, vincristine and prednisone; EPOCH = etoposide, prednisone, vincristine, cyclophosphamide and doxorubicin; LR = local response; NHL-BFM-95 = non-Hodgkin lymphoma Berlin-Frankfurt-Münster-95; NR = no response; PIH = Post inflammatory hyperpigmentation; R-hyper-CVAD = rituximab, cyclophosphamide, vincristine, doxorubicin, dexamethasone, methotrexate, cytarabine; UVB = ultraviolet B phototherapy.
Cutaneous γδ TCL is a rare malignancy consisting of a clonal population of mature, activated γδ T cells with a cytotoxic phenotype. 12 Most cases of CGDTCL in the literature appear at a median age of 42 years with a female predominance. 13 Five-year survival is not reached in classic CGDTCL patients. 10 Our report shows 9 cases of relatively indolent courses of CGDTCL evidenced by longer than expected survival from date of presentation.
The classically described immunophenotype of CGDTCL is CD3+, CD4−, CD5−, and CD8− with variable expression of CD7 and CD56. Cytotoxic markers, granzyme B and TIA-1, are usually positive, and TCR-βF1 is negative. In the larger study from Guitart et al, all cases expressed at least 1 cytotoxic marker, either TIA-1 or granzyme B. 1 Interestingly, they also noted that TIA-1+ and CD5− were associated with overall mortality and granzyme B+ and CD5− was associated with shorter time to death. All 9 of our patients had the unifying finding of TCRγ positivity. However, other immunohistochemical findings varied greatly. Despite that the immune phenotypes of some cases exhibit features of non-CGDTCL cutaneous TCLs, we believe that these are indolent CGDTCLs rather than mimickers of less aggressive TCL. Not only are our cases TCRγ+ but the clinical presentations of our cases make other cutaneous TCLs less likely.
CD30 expression is rarely reported in cases of CGDTCL but was diffusely present in 2 of our patients' tumor cells and present in 8 of 9 patients' lesions. The classic primary CD30+ lymphoproliferative disorders are lymphomatoid papulosis (LyP) and cutaneous CD30+ anaplastic large-cell lymphoma (C-ALCL), both of which have excellent prognosis.14 and 15 It is unlikely that our 2 CD30+ patients have either of these diseases, however, because they do not fit the generally accepted clinical standards. The most common presentation of C-ALCL is ulcerated tumors or nodules and the most common presentation of LyP is multiple self-regressioning pink, red, or brown papules and nodules of < 1.5 to 2.0 cm. 14 Neither of these presentations was typical of any of our cases. It is possible that the diffuse CD30+ in case 3 and 9, plus the rare, scattered CD30+ atypical lymphocytes in 5 other cases, is contributing to their more indolent-course CGDTCL, although our small sample size is far too small to draw prognostic conclusions. To our knowledge, CD30+ cutaneous TCL with γδ-phenotype is rare. Again, drawing comparison to Guitart et al, CD30+ was observed in only 1 of 45 patients. 5 Concerning their patient with atypical γδ T cells and CD30 expression, they observed that the lesions also presented with an indolent clinical course but concluded that the lesions represented a variant of LyP rather than CGDTCL. Regardless, in our 2 CD30+ cases, therapy with the anti-CD30 agent BV was beneficial in resolving lesions for 1 patient and has just been started in the other patient.
The typical profile of CGDTCL is CD4−CD8− and this phenotype, when seen with TCRγ, is associated with more aggressive disease than CD4+ TCLs. 16 CD4+ cells in TCL are associated with nonaggressive behavior in MF and under the provisional category of primary cutaneous CD4+ small/medium-sized pleomorphic T-cell lymphomas, 17 although pleomorphic TCL would not be on the differential for a clinical presentation suspected to be CGDTCL. We were able to exclude MF and pleomorphic TCL based on the constellation of clinical presentation and histopathologic and immunophenotypic findings. Clinically, the predilection of lesions on the extremities compared with on the trunk and buttocks points to CGDTCL over MF. Histologically, a major difference is presence of CD4+CD8− infiltrate along with a lack of TCRγ in MF. 14 Taken together, these features support a diagnosis of CGDTCL in our patients despite the apparently indolent course involving MF-like plaques. CD4 positivity in CGDTCL is rare in the literature; in a reported study of 53 CGDTCL patients, only 4 were CD4+, 1 making our group very unique. As with CD30 expression, we believe that CD4 expression in TCRγ+ CGDTCL patients is likely a contributor to the more indolent course. Alternatively, it is possible that these patients represent MF with aberrant expression of TCRγ. Notably, in our patients, lesions that presented like MF tended to respond to skin-directed skin regimens faster than the other lesions.
Despite the belief that TCRβ and TCRγ expression are mutually exclusive, the double-positive phenomenon we see in 3 of our patients is not unique. TCRβF1 and TCRγ double-positive neoplastic cells were reported in lesional biopsies in another study in 1 of 5 CGDTCL patients and 3 of 5 LyP patients. 18 Among the double-positive cases presented by Rodríguez-Pinilla et al, 18 all 3 LyP patients had diffuse CD30 expression and the CGDTCL patient had low CD30 expression. In our double-positive cases, CD30 expression was low. This further supports the diagnosis of CGDTCL in our 3 patients with TCRγ expression alongside βF1 expression, but possibly explains why the progression of disease is less aggressive, similar to LyP.
Taken together, the novel finding of single markers displaying atypical expression in CGDTCL cases might be the most important contributing factor to the less-than-expected aggressiveness of the TCLs reported herein. Specifically, expression of CD30, CD4, or βF1, despite TCRγ staining, might explain the indolent course of cases 1, 3, 5, 6, 7, and 9. The remaining cases (2, 4, and 8) are more puzzling because they display the expected immunophenotype of aggressive CGDTCL yet the patients were treated and achieved at least a partial response. Their indolent courses might be explained by the aggressive therapy they received, initiated sooner after diagnosis. In those 2 cases, treatment with NHL-BFM-95 and R-hyper-CVAD, respectively, induced complete resolution of lesions.
The authors have stated that they have no conflicts of interest.
- 1 J. Guitart, D.D. Weisenburger, A. Subtil, et al. Cutaneous γδ T-cell lymphomas: a spectrum of presentations with overlap with other cytotoxic lymphomas. Am J Surg Pathol. 2012;36:1656-1665 Crossref
- 2 C. Tripodo, E. Iannitto, A.M. Florena, et al. Gamma-delta T-cell lymphomas. Nat Rev Clin Oncol. 2009;6:707-717 Crossref
- 3 J.R. Toro, M. Beaty, L. Sorbara, et al. Gamma delta T-cell lymphoma of the skin: a clinical, microscopic, and molecular study. Arch Dermatol. 2000;136:1024-1032
- 4 A.S. MacLeod, W.L. Havran. Functions of skin-resident γδ T cells. Cell Mol Life Sci. 2011;68:2399-2408 Crossref
- 5 J.R. Toro. Gamma-delta T-cell phenotype is associated with significantly decreased survival in cutaneous T-cell lymphoma. Blood. 2003;101:3407-3412 Crossref
- 6 C. De Wolf-Peeters, R. Achten. Gamma delta T-cell lymphomas: a homogeneous entity?. Histopathology. 2000;36:294-305 Crossref
- 7 S.S. Chuang. Primary cutaneous γδ T-cell lymphoma should be limited to cases with TCRγδ expression. Am J Clin Pathol. 2013;140:135 Crossref
- 8 A. Garcia-Herrera, J.Y. Song, S.S. Chuang, et al. Nonhepatosplenic γδ T-cell lymphomas represent a spectrum of aggressive cytotoxic t-cell lymphomas with a mainly extranodal presentation. Am J Surg Pathol. 2011;35:1214-1225 Crossref
- 9 M. Ciofani, J.C. Zúñiga-Pflücker. Determining γδ versus αβ T cell development. Nat Rev Immunol. 2010;10:657-663
- 10 R. Willemze. WHO-EORTC classification for cutaneous lymphomas. Blood. 2005;105:3768-3785 Crossref
- 11 L.C. Pinter-Brown. SGN-30: a basis for the effective treatment of CD30 positive hematopoietic malignancies. Expert Opin Investig Drugs. 2008;17:1883-1887 Crossref
- 12 D.C. Endly, R.H. Weenig, M.S. Peters, D.S. Viswanatha, N.I. Comfere. Indolent course of cutaneous gamma-delta T-cell lymphoma: gamma-delta T-cell lymphoma. J Cutan Pathol. 2013;40:896-902
- 13 P. Aguilera, J.M. Mascaró Jr., A. Martinez, et al. Cutaneous gamma/delta T-cell lymphoma: a histopathologic mimicker of lupus erythematosus profundus (lupus panniculitis). J Am Acad Dermatol. 2007;56:643-647 Crossref
- 14 R. Willemze. Prognostic factors in cutaneous T-cell lymphoma. Haematol Rep. 2006;2:89-92
- 15 C. Querfeld, T.M. Kuzel, J. Guitart, S.T. Rosen. Primary cutaneous CD30+ lymphoproliferative disorders: new insights into biology and therapy. Oncology (Williston Park). 2007;21:689-696 discussion 699-700
- 16 D. Jones, F. Vega, A.H. Sarris, L.J. Medeiros. CD4-CD8-‘double-negative’ cutaneous T-cell lymphomas share common histologic features and an aggressive clinical course. Am J Surg Pathol. 2002;26:225-231 Crossref
- 17 S.H. Swerdlow. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. (International Agency for Research on Cancer, 2008)
- 18 S.M. Rodríguez-Pinilla, P.L. Ortiz-Romero, V. Monsalvez, et al. TCR-γ expression in primary cutaneous T-cell lymphomas. Am J Surg Pathol. 2013;37:375-384
1 Department of Dermatology, The University of Texas M.D. Anderson Cancer Center, Houston, TX
2 Department of Pathology, The University of Texas M.D. Anderson Cancer Center, Houston, TX
∗ Address for correspondence: Madeleine Duvic, MD, Department of Dermatology, M.D. Anderson Cancer Center, Pickens Tower, FCT11.6078, 1515 Holcombe Blvd, Unit 1453, Houston, TX 77030 Fax: 713-745-3597
© 2014 Elsevier Inc., All rights reserved.