Key Words: Primary cutaneous follicle center lymphoma; Follicular lymphoma; Translocation t(14 ... primary cutaneous B-cell lymphomas, is the most common.
AJCP / Original Article
Diagnostic and Prognostic Value of BCL2 Rearrangement in 53 Patients With Follicular Lymphoma Presenting as Primary Skin Lesions Anne Pham-Ledard, MD, PhD,1,2 Anne Cowppli-Bony, MD,3 Adélaïde Doussau, MD,3 Martina Prochazkova-Carlotti, PhD,2 Elodie Laharanne, PhD,2,4 Thomas Jouary, MD,1 Marc-Antoine Belaud-Rotureau, PhD,2 Béatrice Vergier, MD, PhD,2,4 Jean-Philippe Merlio, MD, PhD,2,4 and Marie Beylot-Barry, MD, PhD1,2 From the 1Dermatology Department, CHU de Bordeaux, Bordeaux, France; 2EA2406 Histology and Molecular Pathology of Tumours, University of Bordeaux, Bordeaux, France; 3Clinical Epidemiology Unit, CHU de Bordeaux, Bordeaux, France; and 4Pathology Department, CHU de Bordeaux, Bordeaux, France. Key Words: Primary cutaneous follicle center lymphoma; Follicular lymphoma; Translocation t(14;18); BCL2 rearrangement; FISH Am J Clin Pathol March 2015;143:362-373 DOI: 10.1309/AJCP4SUBR4NPSPTN
ABSTRACT Objectives: To study the diagnostic value of BCL2 rearrangement in follicle center lymphoma (FCL) presenting as primary skin lesions, evaluate its prevalence and the prognostic value in primary cutaneous FCL (PCFCL), and assess prognostic factors in PCFCL. Methods: Fifty-three patients with a cutaneous presentation of FCL without a history of nodal lymphoma were selected retrospectively. Clinical and histologic data were collected together with staging and follow-up data. A fluorescence in situ hybridization (FISH) test for BCL2 split probes was performed on skin biopsy specimens. Results: Initial staging procedures identified 47 PCFCLs and six cases of secondary skin involvement of FCL (SSIFCL). FISH detected seven cases carrying a BCL2 rearrangement: four (8.5%) of 47 PCFCLs and three (50%) of six SSIFCLs. These seven cases coexpressed BCL2 and CD10. In PCFCL, cutaneous relapse rate was 42.6%. A small/medium centrocytic cell population was associated with a higher probability of skin relapse in univariate (P = .008) and multivariate ( P = .028) analysis, and BCL2 rearrangement detection was associated with secondary extracutaneous spreading (P = .05). Conclusions: We observed that BCL2 rearrangement in PCFCL is rare, associated with initial positivity of staging (diagnostic value) or with secondary extracutaneous spreading (prognostic value). In selected cases with BCL2CD10 coexpression, FISH testing could detect patients with poor outcome and require closer monitoring.
362 Am J Clin Pathol 2015;143:362-373 DOI: 10.1309/AJCP4SUBR4NPSPTN
Primary cutaneous lymphomas (PCLs) of B-cell origin account for 25% of all PCLs. Three main types of primary cutaneous B-cell lymphomas are recognized: primary cutaneous follicle center lymphoma (PCFCL), primary cutaneous marginal zone lymphoma (PCMZL), and primary cutaneous diffuse large B-cell lymphoma (PCLBCL), leg type.1,2 PCFCL, representing approximately 55% of all primary cutaneous B-cell lymphomas, is the most common B-cell lymphoma to occur as a primary tumor of skin.3 PCFCL usually involves the head and neck or trunk, with solitary or grouped plaques and tumors, while multifocal skin lesions are possible. Morphologically, PCFCL exhibits a proliferation of neoplastic follicle center cells, usually a mixture of small/medium and large centrocytes (sometimes spindle-shaped type) and a variable proportion of centroblasts. Architectural pattern is variable along a continuum from follicular, nodular, sometimes focally periadnexal to diffuse growth patterns. PCFCL, even with a predominance of large cells (large centrocytes or centroblasts), has an indolent behavior characterized by a 5-year survival rate of 95% as opposed to the 41% rate of PCLBCL, leg type.1,4,5 Multifocal lesions do not have a poorer outcome than solitary lesions.3 PCFCL on the leg displays a poorer prognosis, with 5-year disease-specific survival reported at 41%.4 Skin relapse is observed in 30% to 45% of patients. Recurrence does not affect prognosis and is usually confined to the skin. Secondary extracutaneous dissemination may occur in 5% to 10% of cases.4,5 Furthermore, secondary skin involvement by nodal follicular lymphoma (SSIFCL) may share clinicopathologic similarities with PCFCL, especially if skin lesions are the initial manifestation of a systemic disease. Therefore, a negative staging is required to assert PCFCL
© American Society for Clinical Pathology
AJCP / Original Article
❚Table 1❚ BCL2 Rearrangement (or t(14;18)) Detection by PCR and/or FISH in PCFCL
Author (Year)
No. of Patients
t(14;18) Detection by PCR, No./ Total No. (%)
Geelen et al (1998)8 Child et al (2001)9 Vergier et al (2004)7 Aguilera et al (2001)11 Lawnicki et al (2002)12 Mirza et al (2002)13 Kim et al (2005)14 Streubel et al (2006)10
8 5 30 17 20 32 13 27
0/8 0/5 9/30 (30) 3/17 (18) 4/20 (20) 11/32 (34) 4/13 (31) 0/27
t(14;18) Detection by FISH, No./ Total No. (%)
Materials and Methods 0/17
11/27 (41)
FISH, fluorescence in situ hybridization; PCFCL, primary cutaneous follicle center lymphoma; PCR, polymerase chain reaction.
and to rule out SSIFCL since both diseases require a different treatment.6 The hallmark of nodal follicular lymphoma is the t(14;18)(q32;q21) translocation, which consists of a reciprocal translocation between the BCL2 proto-oncogene and the immunoglobulin heavy chain (IgH) gene, leading to the overexpression of BCL2 protein. Therefore, the t(14;18) should be detected either by polymerase chain reaction (PCR) for BCL2-IgH breakpoint amplification or by fluorescence in situ hybridization (FISH) for either BCL2-IgH fusion or BCL2 separation or split. In nodal follicular lymphoma, interphase FISH has detected a higher rate of positive cases for BCL2 rearrangement than the BIOMED-2 PCR protocol.7 While the BCL2 rearrangement is observed in 80% to 90% of cases of follicular lymphoma, this genetic alteration was not detected in several series of PCFCL, including the one from our group using both FISH and PCR with BIOMED-2 techniques.1,7-9 However, Streubel et al10 found a BCL2-IgH fusion by FISH in 41% of PCFCL cases, although they were not able to amplify the BCL2-IgH breakpoint by the BIOMED2 PCR protocol in FISH-positive cases. Alternatively, PCR amplification of the BCL2-IgH breakpoint provided positive results in some series of PCFCL cases, but we clearly demonstrated that molecular detection alone is able to pick up B cells carrying BCL2-JH rearrangement as described in healthy carriers ❚Table 1❚.7-14 Besides the technical aspects, most of the above studies did not provide extensive staging and follow-up data that may account for such discordance in patients with different inclusion criteria. This prompted us to analyze by FISH the prevalence of the BCL2 rearrangement in follicle center lymphoma (FCL) with skin presentation before staging procedures. The main goal of this study was to estimate the prevalence of BCL2 rearrangement in the largest series of PCFCL cases studied so far and to determine if the detection of a BCL2
© American Society for Clinical Pathology
rearrangement could represent a biomarker that would help patient management at diagnosis. Thanks to an extended follow-up, we also investigated prognostic factors that would predict relapse or extracutaneous spreading in PCFCL.
Patient Selection Cases were retrieved retrospectively from the Aquitaine database of cutaneous lymphoma and from dermatologic departments of the University Hospital of Bordeaux from 1994 to 2009 (regional referral center for cutaneous lymphomas). The study was performed after patient information and consent, according to the guidelines of the French bioethical law and the Declaration of Helsinki principles. Patients were consecutively included by the following criteria: 1. Cutaneous lesions of FCL at presentation in patients without a history of B-cell lymphoma. A systematic review of clinical, staging, and histopathologic features was made by an expert panel of dermatologists and pathologists, using the criteria of the World Health Organization classification. 2. Formalin-fixed, paraffin-embedded (FFPE) tissue blocks or frozen material available for molecular studies. Histologic and Clinical Data Collected For all cases, H&E-safranin–stained sections and routine immunostaining were reviewed retrospectively. BCL2, BCL6, and CD10 immunostainings were considered positive if 50% or more of tumor cells expressed the protein. Comparison with CD3 staining on consecutive sections allowed us to discard the strong expression of BCL2 by reactive T cells in such PCFCL cases considered BCL2 positive since the staining was moderate and restricted to B cells. In cases predominantly composed of small/medium centrocytic cells, architectural, cytologic, and immunohistochemical features excluded PCMZL. For three cases, small biopsy specimens limited the interpretation of architectural features, but no specific feature of PCMZL was observed (typical colonization of follicular dendritic cell meshwork, immunophenotype, immunoglobulin light chain–restricted plasma cell) on the primary lesion or relapse (n = 2). In cases predominantly composed of a monotonous population of large centrocytes or centroblasts, a “nonactivated B-cell” (germinal center) phenotype permitted us to rule out PCLBCL, leg type. Staging procedures were performed according to international recommendations,6,7 including detailed history and complete physical examination, routine laboratory tests, computed tomography of the thorax and abdomen, and bone marrow biopsy for most patients (n = 37 [70%], including
Am J Clin Pathol 2015;143:362-373 363 DOI: 10.1309/AJCP4SUBR4NPSPTN
Pham-Ledard et al / t(14;18) in Cutaneous Follicular Lymphoma
all patients with t(14;18)-positive detection). This allowed separating the patients into two groups: PCFCL (n = 47) and SSIFCL with concomitant skin involvement (n = 6). Patients with PCFCL were classified according to the adapted TNM classification.15 Sex, age at diagnosis, site and size of the skin lesion, initial therapy, cutaneous relapses, secondary extracutaneous spreading, status at last follow-up, and duration of follow-up were recorded. Follow-up was updated in March 2013. PCR Amplification for B-Cell Clonality Analysis DNA was extracted from frozen or FFPE skin biopsy specimens (n = 45) and from peripheral blood lymphocytes (n = 44). FR3-JH primers (Ca1/Ca2) were used to amplify the hypervariable third complementarity-determining region (CDR3) of the IgH gene.16 FISH Analysis of the BCL2 Locus The translocation t(14;18)(q32;q21) at both the major breakpoint region or the minor cluster region was detected by interphase FISH with separation probes (BCL2 FISH DNA probe, Split Signal; DAKO, Trappes, France). The 4-µm tissue sections were obtained from FFPE skin tumoral biopsy specimens (n = 50), and FISH analysis was conducted with the histology FISH accessory kit (DAKO). In three cases, frozen skin touch imprints were hybridized as described before.17 FISH patterns were determined by analyzing 200 nonoverlapped nuclei as reported.7 Cases with a BCL2 breakpoint were then referred to as t(14;18)-positive FCL, because translocation partners other than IgH are exceedingly rarely affected in FCL.2 PCR Amplification of BCL2-JH Rearrangement The IGH-BCL2 fusion was sought by PCR amplification targeting the JH joining region of the IgH gene and distinct regions of the BCL2 gene. DNA was extracted from peripheral blood lymphocytes and frozen skin biopsy specimens of positive FISH cases. According to the BIOMED-2 protocol,18 three master mixes were used to identify BCL2 breakpoints MBR (mix A), 3′ MBR (mix B), mcr, and 5′ MCR (mix C). Response Criteria and Statistical Analysis Patients with PCFCL were included for statistical analysis of prognosis factors, while patients with a positive extracutaneous staging were excluded. The primary end point was disease-free survival (DFS). In Kaplan-Meier analysis and Cox proportional hazards model of DFS, time was defined as the time elapsed from date of diagnosis to the date of first disease recurrence or death. Time was censored for patients who had not experienced disease recurrence or had not died at the time of last follow-up. Potential prognosis
364 Am J Clin Pathol 2015;143:362-373 DOI: 10.1309/AJCP4SUBR4NPSPTN
factors examined in our study were patient age (dichotomized in two classes: >60 years vs 5 cm vs
