Primary cutaneous large B-cell lymphoma of the leg: Histogenetic analysis of a controversial clinicopathologic entity

Primary cutaneous large B-cell lymphoma of the leg: Histogenetic analysis of a controversial clinicopathologic entity

Primary Cutaneous Large B-Cell Lymphoma of the Leg: Histogenetic Analysis of a Controversial Clinicopathologic Entity MARCO PAULLI, MD, PHD, ALESSANDR...

217KB Sizes 6 Downloads 95 Views

Primary Cutaneous Large B-Cell Lymphoma of the Leg: Histogenetic Analysis of a Controversial Clinicopathologic Entity MARCO PAULLI, MD, PHD, ALESSANDRA VIGLIO, MD, DANIELA VIVENZA, PHD, DANIELA CAPELLO, PHD, DAVIDE ROSSI, MD, ROBERTA RIBONI BSC, MARCO LUCIONI, MD, PAOLO INCARDONA, MD, EMANUELA BOVERI, MD, MARIO BELLOSTA, MD, ESTER ORLANDI, MD, GIOVANNI BORRONI, MD, MARIO LAZZARINO, MD, EMILIO BERTI, MD, ELVIO ALESSI, MD, UMBERTO MAGRINI, MD, AND GIANLUCA GAIDANO, MD, PHD This study analyzes the pathologic and molecular features of 5 cases of primary cutaneous large B-cell lymphoma of the leg (PCLBCL-leg), recently included in the European Organization for Research and Treatment of Cancer (EORTC) classification of primary cutaneous lymphoma. PCLBCL-leg accounts for 5% to 10% of all primary cutaneous B-cell lymphoma (PCBCL), usually affects elderly patients and carries a worse prognosis than other forms of PCBCL. It has been proposed that the malignant cells of PCLBCL-leg originate from germinal center (GC)-related cells, but their effective normal counterpart is unclear, and the rationale behind the inclusion of this lymphoma as a separate entity is based on its prognosis rather than on its proved histogenesis. All of our cases of PCLBCL-leg morphologically resembled diffuse large B-cell lymphoma (DLBCL), but to better define their histogenesis, we also analyzed various phenotypic and genotypic markers, including mutations of the Ig and of BCL-6 genes, as well as expression of the bcl-6, MUM1, and CD138/syndecan-1 proteins. Immunohistochemically, all of our cases stained for the L-26/CD20cy and CD79a antigens and expressed the bcl-2, bcl-6, and MUM-1 proteins but were negative for both the

CD10/CALLA and CD138 antigens. With respect to molecular analysis, the lymphoma population of all PCLBCL-leg carried hypermutation of Ig genes, and all but 1 case also harbored mutations of the BCL-6 gene. Our results indicate that PCLBCL-leg are similar both under the morphofunctional and molecular profiles to most DLBCL of other sites. Thus, caution seems justified before definitely considering PCLBCL of the leg as a distinct entity. HUM PATHOL 33:937-943. Copyright 2002, Elsevier Science (USA). All rights reserved. Key Words: primary cutaneous B-cell lymphoma, immunohistochemistry, molecular analysis. Abbreviations: PCBCL, primary cutaneous B-cell lymphoma; PCLBCL-leg, primary cutaneous large B-cell lymphoma of the leg; DLBCL, diffuse large B-cell lymphoma; GC, germinal center; TIL, tumor-infiltrating lymphocyte; PCR, polymerase chain reaction; FCC, follicular center cell; EORTC, European Organization for Research and Treatment of Cancer; WHO, World Health Organization; REAL, revised European-American lymphoma classification; FR, framework region.

Primary cutaneous B-cell lymphoma (PCBCL) accounts for about 20% to 25% of all primary cutaneous lymphomas.1 Based on the revised European-American lymphoma classification (REAL) and World Health

Organization (WHO) general lymphoma classification,2,3 most PCBCL are represented by extranodal marginal zone lymphoma and diffuse large cell lymphoma. Recently, the Cutaneous Lymphoma Study Group of the European Organization for Research and Treatment of Cancer (EORTC) proposed a separate classification for PCBCL,4 based on the assumption that B-cell lymphomas, when originating in the skin, have different clinicopathologic features compared to lymphomas arising at other sites. One entity recognized by the EORTC classification of PCBCL is primary cutaneous large B-cell lymphoma of the leg (PCLBCL-leg).4,5 PCLBCL-leg is a relatively rare disease, comprising approximately 1% to 3% of all cutaneous lymphomas and about 5% to 10% of PCBCL. This lymphoma usually affects elderly women (often older than 70) and has been suggested to carry a worse prognosis than other forms of PCBCL. Clinically, PCLBCL-leg frequently presents as solitary or grouped red or bluish nodules, plaques or tumors on 1 or both lower extremities. Histologically, the lymphoma population is characterized by a dense, diffuse infiltrate of large cells involving the entire der-

From the Departments of Pathology, Dermatology, and Haematology, IRCCS Policlinico S.Matteo/University of Pavia, Pavia, Italy; Hematology Unit, Division of Internal Medicine, Department of Medical Sciences, Amedeo Avogadro University of Eastern Piedmont, Novara, Italy; Department of Dermatology, Ospedale Maggiore Policlinico/IRCCS and University of Milano-Bicocca, and the Department of Dermatology, Ospedale Maggiore Policlinico/IRCCS and University of Milano, Milan, Italy. Accepted for publication May 31, 2002. This study was performed under the auspices of the Gruppo Italiano Linfomi Cutanei and supported in part by grants from the Italian Ministry for University Scientific Research and Technology, by Fondazione Piera Pietro e Giovanni Ferrero, Alba, Italy and by Fondazione CRT, Torino, Italy. R. R. is the recipient of a fellowship from IRCCS Policlinico S. Matteo, Pavia, Italy. D. C. is the recipient of a fellowship from FIRC, Milan, Italy. Address correspondence and reprint requests to Marco Paulli, MD, PhD, Pathology Department, IRCCS Policlinico S.Matteo/University of Pavia, Via Forlanini, 14-27100 Pavia, Italy. Copyright 2002, Elsevier Science (USA). All rights reserved. 0046-8177/02/3309-0014$35.00/0 doi:10.1053/hupa.2002.126881

937

HUMAN PATHOLOGY

Volume 33, No. 9 (September 2002)

mis and resembling either centroblasts or immunoblasts.5-9 It has been proposed that the malignant cells of PCLBCL-leg originate from germinal center (GC)-related cells, but their effective normal counterpart is unclear, and to date, the rationale behind the inclusion of this lymphoma as a separate entity is based on its prognosis rather than on its proved histogenesis.4 In recent times, the field of B-cell lymphoma histogenesis has progressed rapidly due to the increasing availability of well-defined histogenetic markers. Genotypic markers of B-cell lymphoma histogenesis include mutations of immunoglobulin (Ig) and BCL-6 genes, which are somatically acquired at the time of B-cell transit throughout the GC.10-12 Phenotypic markers are represented by expression of the bcl-6 protein, which is restricted to GC B-cells, and expression of multiple myeloma-1 interferon regulatory factor-4 (MUM1/ IRF4), which denotes the final step of intra-GC B cell differentiation, as well as subsequent steps of B cell maturation toward plasma cells.13-17 An additional phenotypic marker of B-cell lymphoma histogenesis is represented by the expression of CD138/syndecan-1 (syn1), a proteoglycan clustering with late stages of B cell maturation.13,17 The objective of the present study was to provide additional histogenetic data on a series of 5 cases that fulfilled the clinicopathologic criteria required by the EORTC classification4 for the diagnosis of PCLBCL-leg. Our results indicate that PCLBCL-leg reflects the genotypic and phenotypic features of late centrocytes and shares many of the clinicopathologic and biologic characteristics of diffuse large B-cell lymphoma (DLBCL) commonly arising at other anatomic sites.

MATERIALS AND METHODS Histology, Immunohistochemistry, and In Situ Hybridization

itive and negative controls were run with the study cases and stained appropriately. DNA Extraction Genomic DNA was isolated by cell lysis followed by digestion with proteinase K and purification using a commercial kit (QIAamp DNA mini Kit, QIAGEN, Italy) according to the manufacturer’s instructions.18 Analysis of IgVH Genes IgVH gene rearrangements were amplified with 2 sets of 6 VH gene family–specific primer and a JH primers mix in separate reactions for each VH primer, as described previously.19,20 The VH primers used in this study hybridize to sequences in the framework region 1 (FR1) or FR2 of the respective VH families. Polymerase chain reaction (PCR) was performed for 40 cycles with an annealing temperature of 60°C. PCR products were directly sequenced using a commercially available kit (ThermoSequenase, Amersham Life Sciences, Buckinghamshire, UK) as previously reported.18 Analysis of intraclonal heterogeneity was performed by conventional strategies.21 Sequences were compared to the V-BASE sequence directory (MRC Centre for Protein Engineering, Cambridge, UK) using MacVector 6.0.1 software (Oxford Molecular Group PLC, Oxford, UK) for comparison of the rearranged IgV genes to the most homologous germline sequences. To define the occurrence of antigen selection in Ig genes used by PCLBCLleg, 2 statistical methods were used: the Chang-Casali binomial distribution model22 and the multinomial distribution model.23 Mutational Analysis of the BCL-6 Gene Mutational analysis of the BCL-6 gene was performed by direct sequencing of a PCR product encompassing nucleotides ⫹404 to ⫹1142 localized in the 5⬘ noncoding region of the gene.18 PCR fragments were directly sequenced with appropriate primers, as described earlier.18 Detection of Viral Infection

For this study, 5 cases of cutaneous large B-cell lymphoma that presented on the lower extremities were enrolled from the files of the Department of Pathology of the IRCCS Policlinico S.Matteo/University of Pavia and the Institute of Dermatology, University of Milano, both members of the Gruppo Italiano Linfomi Cutanei. All cases fulfilled the criteria of the Dutch and EORTC cutaneous lymphoma project group for the definition of primary skin lymphomas,4,5 clinical information was obtained from referring physicians. Adequate paraffin blocks were available for each individual case from which 3-␮-thick haematoxilin and eosin and Giemsa stained sections were histopathologically rereviewed independently by a pathologist (M. P.) and a dermatopathologist (E. Be.) and classified according to the REAL/WHO lymphoma classification and the EORTC classification of cutaneous lymphomas.3,4 Immunohistochemistry on paraffin sections was done using antibodies to CD3, CD4, CD8, CD10, CD20cy, CD21, CD68, CD79a, CD138, Kappa, Lambda, bcl-2, bcl-6, TIA-1, MuM1/IRF4, and Mib-1/Ki-67 by means of an automated immunostainer (Dako, Glostrup, Denmark). Antigen retrieval was performed by heat-induced epitope retrieval. Pos-

In situ hybridization of paraffin sections was performed with a fluorescein-conjugated peptide nucleic acid probe specific for Epstein-Barr-Virus (EBV)– encoded RNAs (EBERs) (Dako). Analysis of Human Herpes Virus– 8 (HHV-8) infection was performed by single-strand PCR with primers KS330233-F and KS330233-R, as reported previously.24

RESULTS Case Reports The main clinical features of our cases of PCLBCLleg are summarized in Table 1. Three of 5 patients were female of advanced age (median, 80.8 years), and presented with multiple rapidly enlarging skin lesions (nodules, plaques) on the leg (Fig 1), which ulcerated in 2 of 5 cases. None of the patients showed extracutaneous disease progression for at least 6 months, thus fulfilling the criteria of the Dutch and EORTC Cutaneous Lymphoma Project Group for the definition of primary skin lymphomas.4 Although arbitrary, the cutoff at 6 months has proved useful in retrospective studies, and it was adopted in this report.

938

CUTANEOUS LARGE B-CELL LYMPHOMA OF THE LEG (Paulli et al)

FIGURE 1. Case 1: Clinically, in most cases (3 of 5) the lymphoma presented as nodular, infiltrated, reddish skin lesions.

Two patients died because of unrelated lymphoma causes (1 patient for heart failure and the other for heart failure associated with severe chronic bronchopneumopathy); a third died with evidence of disease during chemotherapy. The remaining 2 patients received chemotherapy and/or radiotherapy achieving complete remission in 1 patient and partial remission in the other patient. At the last follow-up, these patients are alive and disease-free and alive with disease, respectively.

FIGURE 2. Case 5: The tumor is composed predominantly of large centroblast-like B cells showing a diffuse growth pattern and no epidermotropism. (Hematoxylin and eosin, original magnification ⫻40.) Inset: the lymphoma population consisted of a variable admixture of centroblast and immunoblast-like cells. (Giemsa, original magnification ⫻100.)

Histologic and Immunophenotypic Findings Histomorphology

In all cases, the skin biopsy showed a nonepidermotropic, neoplastic dense lymphoid infiltrate involving the dermis and subcutis (Fig 2). A thin grenz zone was observed in all but 1 case, in which the overlying epidermis was focally ulcerated. The lymphoma growth pattern was diffuse in all cases; a residual, vaguely, nodular appearance was still detectable (at least at low magnification) in 3 of 5 cases. Dermal adnexial structures were effaced diffusely in 2 of 5 cases but only partially in 3 of 5 cases. Cytologically, the lymphoma cellular infiltrate consisted of large cells. These were almost exclusively centroblast-like cells with large noncleaved nuclei, with nucleoli attached to the nuclear membrane in 1 of 5 cases. In 3 of 5 cases, cells were represented by an admixture of centroblast- and immunoblast-like cells with large, oval, vesicular nuclei, distinct nuclear mem-

FIGURE 3. Case 5: The MUM1/IRF4 antibody intensely reacts with most of the large neoplastic cells (nuclear positivity and fine, diffuse cytoplasmic reactivity). (SABC, original magnification ⫻40.)

939

HUMAN PATHOLOGY

Volume 33, No. 9 (September 2002)

TABLE 1. Clinical Data Case

Age/Sex

Type of Lesion

Therapy

1 2

73/M 94/F

No therapy No therapy

3

73/M

4

71/F

Multiple ulcerated nodules, lower right leg Multiple ulcerated nodules and plaques, right leg Sarcomatous-like vegetating plaques, lower left leg and foot Single nodule, lower right leg

5

93/M

Multiple ulcerated nodules and plaques, right leg

Radiotherapy and chlorambucil

Prednisone, onchocarbide, vinchristine Radiotherapy

brane, and prominent nucleoli. In these 4 cases, a variable number of tumor cells showed evident nuclear lobations. In the remaining case, the predominant population of large cells was admixed with a heavy inflammatory cellular infiltrate consisting of numerous small lymphocytes and abundant histiocytes. Accompanying fine interstitial fibrosis was detected in 2 of 5 cases, whereas broad bands fibrosis dividing the tumor in large rounded nodules occurred in a single case. Immunohistochemistry

Immunohistochemical findings are reported in Table 2. Briefly, in all cases the lymphoma cells uniformly expressed the CD20cy (membrane-staining) and CD79a (cytoplasmic-staining) B-cell–specific antigens. The lymphoma population lacked the CD10/CALLA and CD138 antigens but intensely reacted for both the bcl-2 (diffuse cytoplasmic staining) and bcl-6 (nuclear staining) oncoproteins. In all cases, the MUM1/IRF4 antibody strongly reacted with most of the large neoplastic cells (strong nuclear positivity and fine, diffuse cytoplasmic reactivity) (Fig 3). The proliferative rate, as assessed by Mib-1/Ki-67 antibody, was high in all cases,

Follow-Up Died of disease 3 weeks after diagnosis Died of unrelated cause 1 month after diagnosis Died of disease during the third chemotherapy cycle Complete remission; alive without disease Partial remission; alive with disease

ranging from 60% to ⬎90% of the lymphoma infiltrate. The mild to abundant reactive cellular background observed in a single case (case 4) consisted of a predominant population of tumor-infiltrating lymphocytes (TILs) and numerous CD68-positive histiocytes. In all cases, the anti-CD21 antibody failed to evidentiate a tight dendritic cell meshwork, reacting only with scattered cells interspersed within the lymphoma population (CD21 negative). The TIL population exhibited a CD2, CD3, CD43, CD45RO T-cell phenotype and contained both the CD4 and CD8 subsets, the latter increased in number (CD4:CD8 ratio of 1:1). In the same case, the TIA-1 antiserum immunostained numerous scattered cells mostly resembling TILs in cell-size and cytomorphology, whereas it was uniformely negative in the lymphoma population. Molecular Studies Analysis of IgVH Genes

Amplification of the rearranged VH genes used by PCLBCL-leg yielded a single PCR product in 3 of 5 tumor samples, whereas a biallelic rearrangement was

TABLE 2. Immunohistochemical Findings Antibody and Source B-cell markers L26 (D) CD79a (D) CD10/CALLA (N) BB4 (IQP) MUM1/IRF4 (SC) Anti-k (D) Anti-␭ (D) T-cell markers CD3 (D) T4 (D) T8 (D) Proliferation marker MIB-1 (I) Miscellaneous CD21 (D) PGM1 (D) TIA-1 (I) Bcl-2 (D) Bcl-6 (Prof. Falini)

CD; Specificity

Lymphoma Cells

CD20cy; pan B CD79a; pan B CD10; germinal centre B cells CD138; post-germinal centre B cells NC; Interferon Regulatory Factor-1; B lymphocytes following germinal centre selection k light chains ␭ light chains

⫹ ⫹ ⫺ ⫺ ⫹ ⫺ ⫺

CD3; pan T CD4; T helper cells CD8; T suppressor cells

⫺ ⫺ ⫺

NC; nuclear antigen of cellular proliferation CD21; dendritic cells CD68; histiocytic cells NC; cytotoxic granule-associated protein NC; bcl-2 protein NC; bcl-6 protein

60%–90% ⫺ ⫺ ⫹ ⫹

Abbreviations: D, Dako AS (Glostrup, Denmark); N, Novocastra Laboratories (Newcastle Upon Tyne, UK); IQP, Immuno Quality Products (Groningen, The Netherlands); SC, Santa Cruz Biotechnology. (Santa Cruz, CA); I, Immunotech (Marseille, France); NC, not clustered.

940

CUTANEOUS LARGE B-CELL LYMPHOMA OF THE LEG (Paulli et al)

TABLE 3. Analysis of Ig VH Genes Used by PCLBCL-leg

Case

VH Family

VH Gene

Percentage of VH Difference

1 3 4 5

V H4 VH3 VH3 VH3

DP-64/3d2 DP-51⫹ DP-35/V3⫺ DP-58/hv3

16% 13% 4.2% 18%

Analysis by Multinomial Distribution Model†

Analysis by Binomial Distribution Model* P CDR

P FR

P CDR

P FR

0.0915 0.100 0.012 0.037

0.090 0.025 0.075 0.042

0.173 0.107 0.007 0.048

0.205 0.019 0.052 0.063

*P values are calculated according to Chang and Casali (binomial distribution model). †P values are calculated according to Lossos et al. (multinomial distribution model).

detected in 1 case (Table 3). In 1 case, no VH gene rearrangement could be amplified with the approach used. All rearranged VH genes used by PCLBCL-leg carried somatic Ig hypermutation. In particular, sequencing analysis demonstrated that the similarity between the VH sequences used by PCLBCL-leg and the closest germ line genes ranged from 95.8% to 82.0%. In the sample with biallelic Ig rearrangement, 1 VH-D-JH gene segment was functional, whereas the other segment was out of frame. Functional VH genes were derived from VH3 (3 cases) and VH4 (1 case) families. Analysis of intraclonal heterogeneity scored negative in all tested cases. Statistical analysis of mutations was performed for all productively rearranged Ig genes by the binomial (Chang-Casali) and the multinomial distribution methods (see Table 3). The results of the binomial and the multinomial statistical methods were superimposable. Evidence of antigen selection was observed in 2 cases, in which the number of R mutations in complementarity determining regions (CDR) was significantly higher than would be expected to arise solely by chance (P ⬍ 0.05). Evidence for a significant counterselection of replacement mutations within FRs was observed in an additional PCLBCL-leg (P ⬍ 0.05). Mutational Analysis of the BCL-6 Gene

Mutational analysis of the BCL-6 gene was performed on a PCR fragment spanning a 739-bp region located downstream of the first noncoding exon of BCL-6 and containing ⬎95% of BCL-6 mutations occurring in human neoplasms.18 BCL-6 mutations were detected in 4 of 5 cases (Table 4). In total, 24 mutational events were observed among the 4 mutated samples. Table 4 summarizes the characteristics of the BCL-6

mutations detected in PCLBCL-leg. In all mutated cases, mutations occurred in the heterozygous state and were represented exclusively by single nucleotide substitutions. The average frequency of mutation was 0.428 ⫻ 10⫺2 bp (range, 0.07 to 1 ⫻ 10⫺2 bp, corresponding to 1 to 14 mutations per allele) (see Table 3). Viral infection

Infection by EBV and by HHV-8 was consistently absent in all PCLBCL-leg (see Table 4). DISCUSSION This study has analyzed the pathologic and molecular features of 5 cases of PCLBCL-leg, as originally described by Vermeer et al.5 and subsequently by other authors.7-9 In the EORTC cutaneous lymphoma classification,4 PCLBCL-leg is retained as a separate category, but this distinction is based mainly on its prognosis rather than on its proved histogenesis, and the existence of such entity is still questioned.25 Apart from immunocytoma and rare provisional entitities (e.g., cutaneous plasmocytoma and intravascular B-cell lymphoma), the EORTC classification considers most PCBCLs to be follicular center cell (FCC) derived. However, the term FCC lymphoma encompasses various histologic subtypes (small cell, large cell, centrocytic, centroblastic, and mixed), with nodular and/or diffuse growth patterns. Such a broad spectrum of cytologic features of FCC lymphoma probably led to the inclusion within this category of cases otherwise classified as DLBCL. In particular, a matter of debate is the relationship of PCBCL composed mainly of large cells, including the so-called PCLBCL-leg, to other histologic

TABLE 4. Molecular Characterization of BCL-6 Mutations in PCLBCL-leg Case 1 2 3 4 5

BCL-6* C 3 G (475), A G 3 A (754), T 3 C (781) C 3 T (704), G — G 3 A (819), A

3 T (477), G 3 T (506), C 3 T (522), T 3 C (565), T 3 C (626), T 3 G (723), G 3 C (751), G 3 A (756), G 3 C (804), A 3 C (1069), A 3 C (1071) 3 A (787), C 3 T (930), T 3 C (931), A 3 C (1071) 3 G (970), G 3 A (1065), T 3 A (1098)

*The positions of nucleotide substitutions are indicated in parentheses.

941

HUMAN PATHOLOGY

Volume 33, No. 9 (September 2002)

categories of B-cutaneous lymphomas, because a distinction between large FCC and DLBCL may be sometimes difficult by morphology alone. In the present series, all cases were easily morphologically classified as DLBCL, because in all cases the lymphoma population consisted of large cells (exclusively centroblasts in 2 of 5 cases and an admixture of centroblasts and immunoblasts in 3 of 5 cases). To corroborate morphologic findings and define the histogenesis of PCLBCL-leg, we used a model that has proven useful in improving understanding the derivation of other nodal and primary extranodal B-cell lymphoma subtypes. According to this model,12,13,17 the different stages of mature B cells can be identified by the analysis of various histogenetic markers, including mutations of the Ig and of BCL-6 genes, as well as expression of the bcl-6, MUM1, and CD138/syn-1 proteins. Because Ig and BCL-6 mutations are acquired by B cells at the time of their transit through the GC, their occurrence denotes that the lymphoma derives from GC-related B cells.10-12 Ig and BCL-6 mutations are retained by B cells on GC exit and further differentiation, hampering the discrimination between GC and post-GC B cells. Such distinction can be aided by the expression analysis of the bcl-6 protein (restricted to GC B cells) and of the MUM-1 protein, which is expressed by the final step of intra-GC maturation (i.e., late centrocytes) and by post-GC B cells.14-16 Expression of CD138 clusters with late stages of B cell maturation.17 Immunohistochemically, all of our cases positively immunostained for the L-26/CD20cy and CD79a antigens and expressed the bcl-2, bcl-6, and MUM-1 proteins, but were negative for both the CD10/CALLA and CD138 antigens. The foregoing phenotype is consistent with histogenesis from GC-related B-cells. In particular, PCLBCL-leg express the bcl-6⫹/MUM-1⫹ phenotype that is also found in a large proportion of both nodal and extranodal DLBCL. Because the expression of bcl-6 and MUM-1 is mutually exclusive in normal GC B cells, their simultaneous expression may be considered a tumor-specific phenotype caused by molecular deregulation of the BCL-6 gene.15 Notably, the bcl-6⫹/ MUM-1⫹ phenotype clearly distinguishes PCLBCL-leg from grade III follicular lymphoma, usually bcl-6⫹ but MUM-1⫺.15,16 Furthermore, in accordance with previous reports,5,9 all PCLBCL-leg included in this study also expressed the bcl-2 protein. With respect to molecular analysis, we found that, in all tested cases, the lymphoma population carried hypermutation of Ig genes and all but 1 case also harbored mutations of the BCL-6 gene. The fact that a single case of PCLBCL-leg failed to carry BCL-6 mutations despite the presence of Ig mutations is not surprising, because the frequency of BCL-6 mutations is physiologically lower than that of Ig mutations.10,11 Overall, these findings confirm that PCLBCL-leg definitely represent DLBCL both under the morphofunctional and molecular profile.10,12,18 A reduced disease-free survival and a worse prognosis as compared with the morphologically similar

PCBCL on the head and trunk are considered among the hallmarks of PCLBCL-leg. Proposed unfavorable prognostic markers for PCLBCL-leg include the proportion of large round-lymphoma cells and the expression of bcl-2 protein, with the latter suggested to be site-related.26 Despite their elderly median age (80.8 years), only 1 of 5 patients in our series died of lymphoma; however, among our cases, no significant differences were found with respect to the morphologic features as well as the bcl-2 expression status of the lymphoma cells. These findings are in keeping with the results of a recent study9 on a series of patients with PCLBCL-leg, indicating that the primary (leg) lesional site is not associated with different survival probabilities, bcl-2 expressional status, and response to treatment. In all likelihood, similar to other B-cell malignancies, various recognized clinical prognostic factors (e.g., number and extent of lesions, tumor burden) other than site of presentation may be responsible for the clinical outcome of PCLBCL-leg. In summary, the results of our pathologic and molecular analyses indicate that the genetic and immunophenotypic status of the lymphoma cells in PCLBCLleg largely overlap with that of most large B-cell lymphomas of other sites. Thus caution seems to be justified before definitely considering PCLBCL-leg a distinct clinicopathologic entity.

REFERENCES 1. Isaacson PG, Norton AJ: Cutaneous lymphomas. In: Isaacson PG, Norton AJ, eds. Extranodal Lymphomas. London, Churchill Livingstone, 1994:131-91. 2. Harris NL, Jaffe ES, Stein H, et al: A revised EuropeanAmerican classification of lymphoid neoplasms: A proposal from the International Lymphoma Study Group. Blood 84:1361-1392, 1994 3. Harris NL, Jaffe ES, Diebold J, et al: The World Health Organization Classification of neoplastic diseases of the hematopoietic and lymphoid tissues. Report of the Clinical Advisory Committee meeting, Airlie House, Virginia, November 1997. Ann Oncol 10:14191432, 1999 4. Willemze R, Kerl H, Sterry W, et al: EORTC Classification for primary cutaneous lymphomas: A proposal from the Cutaneous Lymphoma Study Group of the European Organization for Research and Treatment of Cancer. Blood 90:354-371, 1997 5. Vermeer MH, Geelen FAMJ, van Haselen CW, et al: Primary cutaneous large B-cell lymphomas of the leg. A distinct type of cutaneous B-cell lymphoma with an intermediate prognosis. Arch Dermatol 132:1304-1308, 1996 6. Kerl H, Cerroni L: Primary B-cell lymphomas of the skin. Ann Oncol 8:S29-S32, 1997 7. Starz H, Balda BR, Bachter D, et al: Secondary lymph node involvement from primary cutaneous large B-cell lymphoma of the leg. Sentinel lymph nodectomy as a new strategy for staging circumscribed cutaneous lymphomas. Cancer 85:199-207, 1999 8. Torres-Paoli D, Sanchez JL: Primary cutaneous B-cell lymphoma of the leg in a chronic lymphedematous extremity. Am J Dermatopathol 22:257-260, 2000 9. Fernandez-Vazquez A, Rodriguez-Peralto JL, Martinez MA, et al: Primary cutaneous large B-cell lymphoma. The relation between morphology, clinical presentation, immunohistochemical markers, and survival. Am J Surg Pathol 25:307-315, 2001 10. Pasqualucci L, Migliazza A, Fracchiolla N, et al: BCL-6 mutations in normal germinal center B cells: evidence of somatic hyper-

942

CUTANEOUS LARGE B-CELL LYMPHOMA OF THE LEG (Paulli et al) mutation acting outside Ig loci. Proc Natl Acad Sci U S A 95:1181611821, 1998 11. Shen HM, Peters A, Baron B, et al: Mutations of BCL-6 in normal B cells by the process of somatic hypermutation of Ig genes. Science 280:1750-1752, 1998 12. Kuppers R, Klein U, Hansmann ML, et al: Cellular origin of human B-cell lymphomas. N Engl J Med 341:1520-1529, 1999 13. Gaidano G, Carbone A: MUM-1: A step ahead toward the understanding of lymphoma histogenesis. Leukemia 4:1209-1215, 2000 14. Cattoretti G, Chang CC, Cechova K, et al: BCL-6 protein is expressed in germinal-center B cells. Blood 86:45-53, 1995 15. Falini B, Fizzotti M, Pucciarini A, et al: A monoclonal antibody (MUM1p) detects expression of the MUM1/IRF4 protein in a subset of germinal center B cells, plasma cells, and activated T cells. Blood 95:2084-2092, 2000 16. Tsuboi K, Iida S, Inagaki H, et al: MUM1/IRF4 expression as a frequent event in mature lymphoid malignancies. Leukemia 14:449456, 2000 17. Carbone A, Gloghini A, Larocca LM, et al: Expression profile of MUM1/IRF4, BCL-6, and CD138/syndecan-1 defines novel histogenetic subsets of human immunodeficiency virus-related lymphomas. Blood 97:744-751, 2001 18. Capello D, Vitolo U, Pasqualucci L, et al: Distribution and pattern of BCL-6 mutations throughout the spectrum of B-cell neoplasia. Blood 95:651-659, 2000

19. Kuppers R, Rajewsky K, Hansmann ML: Diffuse large cell lymphomas are derived from mature B cells carrying V region genes with a high load of somatic mutation and evidence of selection for antibody expression. Eur J Immunol 27:1398-1405, 1997 20. Fais F, Gaidano G, Capello D, et al: Immunoglobulin V region gene use and structure suggest antigen selection in AIDSrelated primary effusion lymphoma. Leukemia 13:1093-1099, 1999 21. Pasqualucci L, Neumelster P, Goossens T, et al: Hypermutation of multiple proto-oncogenes in B-cell diffuse large-cell lymphomas. Nature 412:341-346, 2001 22. Chang B, Casali P: The CDR1 sequences of a major proportion of human germline Ig VH genes are inherently susceptible to amino acid replacement. Immunol Today 15:367-373, 1994 23. Lossos I, Tibshirani R, Narasimhan B, et al: The inference of antigen selection on Ig genes. J Immunol 165:5122-5126, 2000 24. Gaidano G, Pastore C, Gloghini A, et al: Distribution of human herpesvirus-8 sequences throughout the spectrum of AIDSrelated neoplasia. AIDS 10:941-949, 1996 25. Jaffe ES, Sander CA, Flaig MJ: Cutaneous lymphomas: A proposal for a unified approach to classification using the REAL/ WHO Classification. Ann Oncol 11:S17-S21, 2000 26. Geelen FA, Vermeer MH, Meijer CJ, et al: bcl-2 protein expression in primary cutaneous large B-cell lymphoma is site-related. J Clin Oncol 16:2080-2085, 1998

943