- Email: [email protected]
New prognostic relevant factors in primary cutaneous diffuse large B-cell lymphomas
New prognostic relevant factors in primary cutaneous diffuse large B-cell lymphomas Christian Hallermann, MD,a Christoph Niermann, MD,a Rudolf-Josef Fischer, MD, PhD,b and Hans-Joachim Schulze, MD, PhDa Munster, Germany Background: There is a growing body of literature that has enhanced our understanding of the biology of primary cutaneous diffuse large B-cell lymphoma (PCDLBCL) including in the context of gene profiling studies. Recent studies have demonstrated an activated proliferation profile associated with leg type lymphoma including overexpression of proto-oncogenes PIM1, PIM2, and cMYC, and the transcription factors MUM1 and OCT2. Although gene profiling is very useful in understanding the molecular basis of diffuse large B-cell lymphoma (LBCL), it is not practical from a routine diagnostic perspective. In this regard, the purpose of the study was to further define an armamentarium of easily applied immunohistochemical stains to accurately prognosticate PCDLBCL. Methods: In all, 35 patients with PCDLBCL, 14 of follicle center and 21 of leg type, were analyzed using antibodies against CD5, CD138, BCL2, BCL6, OCT2, MUM1, FOXP1, and cMYC. Findings were correlated with clinical data. Results: All cases stained negative for CD5 and CD138. Both subtypes differed in distinct staining patterns for BCL6, BCL2, OCT2, MUM1, and FOXP1. Staining for BCL2, OCT2, and/or MUM1 was associated with poor, and BCL6 with a favorable prognosis. Expression of cMYC was irrespective of prognosis or subtype, whereas ulceration or primary manifestation on the leg or multiple lesions was indicative for worse prognosis. Limitations: Case number was a limitation. Conclusion: Discriminating PCDLBCL supports the validity of the World Health Organization/European Organization for Research and Treatment of Cancer classification. To identify risk factors in patients with PCDLBCL we recommend thorough evaluation of clinical presentation and exploratory staining pattern for BCL2, BCL6, MUM1 and OCT2. ( J Am Acad Dermatol 2007;56:588-97.)
T
he group of primary cutaneous diffuse large B-cell lymphoma (PCDLBCL) is heterogeneous. According to the World Health Organization (WHO)/European Organization for Research and Treatment of Cancer (EORTC) classification,
From the Department of Dermatology, Institute for Tumors of the Skin,a and Department of Medical Informatics and Biomathematics,b University of Munster. Funding sources: None. Conflicts of interest: None identified. Accepted for publication December 19, 2006. Reprint requests: Christian Hallermann, MD, Dorbaumstrasse 300, 48157 Munster, Germany. E-mail: [email protected]. Published online February 12, 2007. 0190-9622/$32.00 ª 2007 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2006.12.026
588
Abbreviations used: ABC: DSS: EORTC:
activated B cell disease-specific survival European Organization for Research and Treatment of Cancer GC: germinal center LBCL: large B-cell lymphoma LBCL-L: large B-cell lymphoma, leg type LFCL: follicle center lymphoma OS: overall survival PCDLBCL: primary cutaneous diffuse large B-cell lymphoma WHO: World Health Organization
they are divided into large B-cell lymphoma (LBCL) leg type (LBCL-L), LBCLs as a variant of primary cutaneous follicle center lymphoma (LFCL), and a third group including all LBCLs that do not fulfill the criteria of the first two groups.1,2
J AM ACAD DERMATOL
Hallermann et al 589
VOLUME 56, NUMBER 4
This grouping mirrors not only the different clinical behavior of both main entities; moreover, it also correlates differences in cytogenetic aberrations and different gene expression profiles.3-8 Although it is a well-known fact that PCDLBCL-L has a poorer outcome compared with LFCL, the reason for this is still a matter of debate. It seems likely that not the location but the differences in morphology and gene expressions are reasons for the different behavior. For this reason the WHO/EORTC classification chose the term ‘‘leg type’’ and not the former designation ‘‘large B-cell lymphoma of the leg’’ as proposed by the EORTC in 1997.9 Using microarray techniques it is possible to distinguish 3 different groups of nodal diffuse LBCLs. They are characterized by different gene expression profiles: the germinal center (GC) profile, the activated B cell (ABC)-like profile, and a third group that does not fulfill the criteria of the first two groups.10 The GC subtype is defined by the expression of a panel of genes characteristic for normal GC cells. The ABC subtype is defined by the expression of genes characteristic for activated blood lymphocytes. The ABC subtype has a poorer outcome compared with the GC subtype.10 Microarray analyses in PCDLBCLs confirm the relevance of grouping into LFCL and LBCL-L. Both show profiles similar to that of GC type and ABC type, respectively. Several marker genes have been identified that express significantly differently in both groups.8 Although gene profiling is very useful in understanding the molecular basis of diffuse LBCL, it is not practical from a routine diagnostic perspective. In this regard, the purpose of the study was to further define an armamentarium of easily applied immunohistochemical stains to accurately prognosticate diffuse LBCL of the skin.
METHODS Patients and samples For examination we used formalin-fixed skin biopsy specimens from patients with LBCL-L and LFCL. An informed consent was obtained from each patient before biopsy. All biopsy specimens were taken for diagnostic reasons. For inclusion into our study the lymphoma had to be restricted to the skin at the time of diagnosis. This was confirmed by appropriate staging procedures. Patients with immunosuppression have been excluded from the study. We only included cases composed of diffuse arranged lymphoid cells resembling large centrocyts, centroblasts, immunoblasts, or a combination of these. For inclusion into the study, the number of large cells had to be at least 50% of the B cells. As defined by Anagnostopoulos and Stein,11 a
Fig 1. Biopsy specimen of representative lesion from patient with follicle center lymphoma (case 25). Mixture of large cleaved cells resembling centrocytes and some centroblasts not arranged in sheets. (Hematoxylin-eosin stain; original magnification: 31000.)
lymphoma cell is considered large if it is at least twice as big as a normal lymphocyte (bystander cell). The diagnoses were made using the criteria published in the WHO/EORTC classification.1 LFCL was defined as a lymphoid neoplasia composed of diffuse lesions with predominantly large cleaved cells (resembling large centrocytes) with a variable infiltrate of centroblasts, which were not arranged in confluent sheets (Fig 1). LBCL-L was defined as a lymphoid neoplasia with predominance of large noncleaved round cells (resembling centroblasts and immunoblasts) arranged in confluent sheets with a diffuse growth pattern (Fig 2, A). Cases with an exclusively follicular growth pattern were excluded. Because of its heterogeneity and its ill definition, patients with the diagnosis ‘‘PCDLBCL, other’’ were also excluded from the study. Clinical parameters The following regions of the body were defined: head and neck; arm; trunk; and leg. The therapeutic response was classified into 4 categories (complete remission, partial remission, stable disease, and progression). Methods A tumor biopsy specimen was fixed in formalin, embedded in paraffin, and cut into 2- to 4-m sections. In addition to hematoxylin-eosin and Giemsa staining, immunohistochemistry was performed using standard immunoperoxidase techniques. The sections were dewaxed in xylol (Merck, Darmstadt, Germany) and rehydrated in serial dilutions of ethanol. The used antibodies are listed in Table I. An autostainer was used in combination with a detection kit (Chem-Mate LSAB-KIT, DAKO, Glostrup, Denmark). All biopsy specimens were reviewed
590 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
Fig 2. Biopsy specimen of representative lesion from patient with large B-cell lymphoma, leg type (case 7). A, There are confluent sheets of large round B cells (resembling centroblasts and immunoblasts). B to D, Neoplastic cells stain positive for OCT2, MUM1, and BCL2. (A, Hematoxylin-eosin stain; B, OCT2; C, MUM1; D, BCL2; original magnifications: A, 31000; B to D, 3400.)
Table I. Antibodies used in study Antigene
Clone
Pretreatment* Dilutiony
CD20 CD79a CD3 CD138 CD5 BCL2 BCL6 cMYC
L26 JCB117 F7.2.38 MI15 CD5/54/F6 124 PG-B6p CT14.G4
OCT2
Polyclonal Heat, pH9,0
MUM-1 MUM1p FOX-P1 JC12
e Heat, pH9 Heat, pH9 Heat, pH6,1 Heat, pH9 Heat, pH9 Heat, pH9 Heat, pH6,1
Heat, pH6,1 Heat, pH6,1
1:1 1:4 1:50 1:100 1:100 1:6 1:20 1:25 1:25 1:50 1:100
Manufacturer
DAKOz DAKOz DAKOz DAKOz DAKOz DAKOz DAKOz Acris Antibodies§ LabVision Corpk DAKOz Dr Alison Banham{
*Slides were heated in an autoclave using the target retrieval solution (DAKO). y For dilution the antibody Diluent (DAKO) was used. z Glostrup, Denmark. § Hiddenhausen, Germany. k Fremont, Calif. { Dr Alison Banham, Director of the LRF Lymphoma Antigens Programme and Oxford University Research Lecturer, Nuffield Department of Clinical Laboratory Sciences, University of Oxford, United Kingdom.
independently by two experts in dermatopathology (C. H. and H-J. S.). The cases were discussed at the multiheaded microscope. According to Grange et al,12 immunohistologic staining was considered positive if more than 50% of the lymphoma cells were stained. Histopathologic pictures were captured with a microscope (Axioscop 2 plus, Zeiss, Jena, Germany) using a camera (Axiocam HRC, Zeiss). Statistical analysis The statistical methods were chosen according the recently published recommendations by Neville et al.13 Comparisons between the two groups of patients were performed using the usual chi-square test or, when appropriate, Fisher’s exact test for categorical variables and the Student t test (after confirmation of normal distribution by Kolmogorov-Smirnov test) for categorical variables (age). Correlations between data were done by Spearman’s rank correlation. Observed survival and specific survival were estimated using the product-limited method of Kaplan and Meier. End point was the death of the patient (disease related or nonspecific) or the last follow-up. For surviving patents the study end
J AM ACAD DERMATOL
Hallermann et al 591
VOLUME 56, NUMBER 4
point was January 31, 2006. Parameters assessed by univariate analysis included sex, age (\/[70 years), location on the leg, ulcerated lesions, multiple lesions, and BCL2, BCL6, OCT2, FOXP-1, MYC, and MUM1 as determined by immunohistochemistry. Differences between survival were analyzed using the log rank chi-square test or, in cases of crossing functions, the Breslow (generalized Wilcoxon test) or Tarone test. Factors to be shown significant in above-mentioned Kaplan-Meier test and in univariate Cox analysis were analyzed by using a multivariate Cox proportional hazard regression model (method: stepwise forward Wald). All analyses were performed using a statistical software package (SPSS/PC, Version 13.0, SPSS Inc, Chicago, Ill). All P values were based on two-tailed method. Statistical analysis with P less than .05 was considered as statistically significant.
RESULTS Clinical parameters In all, 35 patients with primary cutaneous LBCL were included into the study; 16 of them were male and 19 were female. The lymphomas were classified as LFCL in 14 cases and as LBCL-L in 21 cases. The mean age was 64.6 years (SD 14.8, median 68 years, range 23-90 years) at the time of diagnosis and the mean follow-up was 60.2 months (median 59.5 months, range 4-156 months). The distribution of the lesions included legs (14 cases with exclusive location on the leg: 6 right, 6 left, 2 bilateral), head (14 cases), trunk (4 cases), and arms (2 cases). Three of 35 patients presented ulcerated nodules (cases 6, 7, and 18). Fourteen patients had multiple lesions ( $ 2 nodules). One of them had an involvement of more than one region of the body, as defined above. The most frequent primary therapies were radiation (21 cases) or excision (11 cases). In none of the cases was chemotherapy used as first-line therapy. In 7 of 21 cases treated by radiotherapy and in 4 of 11 cases treated by excision, at least one further therapeutic intervention was necessary. In 19 of 35 patients, a complete remission was attained. In 3 cases a cyclophosphamide, Adriamycin, vincristine, prednisone polychemotherapy was performed as second-line therapy (cases 8, 9, and 13). At the end of the follow-up period, 17 patients were alive without disease, 6 were alive with clinical signs of disease, in 7 patients the death could not be related to the lymphoma, and 5 patients died as a result of the lymphoma. For the patients with complete remission at the end point of the study, the mean follow-up time was 70.58 months (median 61 months, range 12-156 months).
The 5-year overall survival (OS) was 67.9% and the 5-year disease-specific survival (DSS) was 81.5%. Comparing both entities Focusing on the disease entity according to the WHO/EORTC classification, we found significant clinical differences between both groups. The mean age of the patients with LBCL-L was 69.9 years (70 years median) compared with a mean age of 65.6 years (54 years median) in the other group (P = .008). As expected, the clinical presentation differed between the diseases. Patients with LFCL presented mostly with single (93%), nonulcerated (100%) tumors on the head (64%). Patients with LBCL-L presented mostly with multiple (61%), and sometimes ulcerated (14%) tumors on the lower extremities (76%). Patients with LBCL-L have a significantly worse OS and DSS compared with LFCL (P = .012 and P = .04, respectively). Because of the grouping of the lymphomas into both entities based on morphologic and cytologic criteria in this study, it can be concluded that those morphologic criteria are of statistically significant prognostic impact and the entities. Comparing the immunophenotype of both entities we found LBCL-L significantly more often positive for BCL2, OCT2, MUM1, and FOX-P1 and negative for BCL6 than LFCL (chi-square test of Pearson: P = .01, P = .046, P \ .001, P \ .001, and P = .02, respectively). The details are listed in Table II. Histology is shown in Fig 2. Prognostic features Without focusing on the disease entities, we identified parameters associated with a worse survival (Table III and IV). Location on the leg (OS: P = .02, DSS: P = .01), ulceration (OS: P = .002, DSS: P = .07), and multiple nodules (OS: P = .033, DSS: P = .003) were significantly associated with both negative OS and negative DSS. As expected, the age at time of diagnosis (\ vs [70 years) was significantly associated with a worse OS (P = .049) but not with DSS (P = .59). Positive staining for BCL2, OCT2, and FOXP1 was clearly associated with progress under first-line therapy (Spearman correlation: P = .039, P = .016, and P = .039, respectively). Moreover, the expression of BCL2 (OS: P = .02, DSS: P = .039), MUM1 (OS: P = .005, DSS: P = .013), and OCT2 (OS: P = .015, DSS: P = .034) was significantly associated with a bad OS and DSS. On the other hand the expression of BCL6 was an indicator for a better prognosis (OS: P = .014, DSS: P = .001).
592 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
Table II. Clinical and immunohistologic data No.
Type
Age, y/sex
1 2 3 4 5 6 7 8 9
LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L
75/M 66/M 68/F 83/F 73/F 57/M 61/F 60/F 61/M
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL
76/F 89/F 61/F 64/M 75/M 71/M 54/F 69/M 90/F 52/F 75/M 76/F 89/M 66/M 43/F 49/F 58/M 23/F 59/F 54/M 48/M 69/F 45/M 45/F 72/M 73/F
Lesions
Mu/leg Mu/leg Mu/leg Si/head Si/leg Mu/leg Mu/leg Mu/leg Mu/leg and other Si/leg Mu/leg Si/arm Mu/leg Mu/leg Mu/leg Si/head Mu/leg Si/head Si/head Si/head Mu/leg Si/head Si/head Si/trunk Si/head Si/head Si/head Si/head Si/arm Si/head Si/trunk Si/head Mu/trunk Si/trunk Si/head
First therapy
Therapeutic response
Follow-up time, mo
Follow-up
PUVA Rt Ex Ex Rt Ex Ex Rt Rt
Pro Pro CR CR PR Pro CR Pro Pro
64 6 19 69 120 36 118 11 10
A1 DD A D D DD A DD D
1 1 1 1 1 1 1 1 1
1 1 1 1 1
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1
1 1 1
Ex Rt Ex Ex Ex Rt Rt If Rt Rituximab Ex Rt Rt Rt Rt Ex Rt Ex Rt Rt Rt Rt Rt Rt Rt Rt
CR Pro CR Pro Pro Pro PR PR Pro CR CR CR PR CR CR Pro CR CR CR CR CR CR CR CR CR Pro
36 10 57 12 24 47 145 72 12 35 126 68 4 50 12 150 156 71 61 88 106 71 61 43 58 72
D D A A1 DD DD A1 A1 D A A A D A A A1 A A A A A A A A A A1
1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1
BCL2 BCL6 MUM1 FOXP1 OCT2 cMYC
A1, Alive with disease; A, alive without disease; CR, complete remission; D, died unrelated or not obviously related to lymphoma; DD, lymphoma-related death; Ex, excision; F, female; IF, interferon; LBCL-L, large B-cell lymphoma, leg type; LFCL, follicle center lymphoma; M, male; Mu, multiple lesions; PR, partial remission; Pro, progression; PUVA, psoralen plus ultraviolet A; Rt, radiation; Si, single lesion.
Factors shown to be of prognostic significance in Kaplan-Meier tests and significant in univariate Cox analysis (age; location on the leg; multiple lesions; ulceration; positive staining for BCL2, MUM1, and OCT2; and negative staining for BCL6) were analyzed using a multivariate Cox proportional hazard regression model. In this test ulceration, BCL6 and OCT2 have been shown to be prognostically significant. Because of the low case number, all 12 patients who died were included into this analysis (Table V).
DISCUSSION In this study we investigated clinical and immunophenotypic characteristics in a large number
of PCDLBCLs including cases with LBCL-L and LFCL. Both groups show significant differences in clinical presentation and prognosis. In LBCL-L the patients are usually older and present more often with multiple lesions. Pooling all lymphomas together we identified diagnosis, location on the leg, ulceration, and multiple lesions at presentation as significant prognostic factors. In our study all 5 patients who died as a result of the lymphoma in the first 5 years presented with multiple nodules on the legs. Location on the leg and multiple lesions has also been reported as prognostically significant in further European and North American studies.14,15 One of the largest
Hallermann et al 593
J AM ACAD DERMATOL VOLUME 56, NUMBER 4
Table III. Comparison between the two entities Total (%)
N Clinical data Sex Male Female Age (\70 y) Multiple lesions Ulceration Immunohistologic data BCL6 BCL2 OCT2 MUM1 FOXP1 cMYC CD138 CD5 Disease-specific survival 3y 5y Overall survival 3y 5y
35
LFCL (%)
14
LBCL-L (%)
P (LFCL e LBCL-L)
21
16 19 22 14 3
(46) (54) (63) (40) (8)
7 7 11 1 0
(50) (50) (78) (7) (0)
9 12 11 13 3
(43) (57) (52) (61) (14)
25 22 22 19 19 9 0 0
(71) (63) (62) (54) (54) (26) (0) (0)
13 1 6 0 2 3 0 0
(93) (7) (42) (0) (14) (21) (0) (0)
12 21 16 19 17 6 0 0
(57) (100) (76) (90) (80) (28) (0) (0)
100 100
76.53 68.03
91.67 91.67
59.42 52.82
.008 .01
.022 .001 .046 .001 .001
Only significant P values are listed. LBCL-L, Large B-cell lymphoma, leg type; LFCL, follicle center lymphoma.
studies on this topic has recently been published by Zinzani et al.16 In that study the location on the leg and the disease entity are relevant prognostic parameters in primary cutaneous B-cell lymphomas. Our data further corroborate these studies and further emphasize the clinical significance of the WHO/ EORTC classification. To our knowledge, until now the prognostic impact of ulceration has not been clearly demonstrated. Because of the low case number in our study, larger trials are necessary to validate this finding. Using immunohistochemical antibodies against different gene products of B-cell development and oncogenes, we found a different pattern in both groups. These differences were significant for some of the markers. None of the lymphomas in our study stained positive for CD138 or CD5. Naı¨ve B cells are usually positive for CD5. It is consistently expressed in B cells of the mantle zone, B-CLL, and in mantle cell lymphoma. Moreover, it is expressed in about 10% of systemic diffuse LBCLs in which its expression is associated with a worse prognosis.17-19 All analyzed
Table IV. Prognostic features in primary cutaneous diffuse large B-cell lymphoma (irrespective of the subtype) Parameter
DSS, P
OS, P
Age [ 70 y Male Location on leg Multiple lesions Ulceration BCL21 BCL6 OCT21 MUM11 cMYC1 FOXP11
.59 .21 .016* .003* .007* .04* .001* .034* .013* .96 .35
.048* .49 .02* .03* .002* .02* .014* .015* .005* .92 .31
DSS, Disease-specific survival; OS, overall survival. *Significant.
Table V. Multivariate Cox proportional hazard regression model of prognostic significant parameters (method: stepwise forward Wald) Coefficient (CD) SE
Ulceration
Wald
P
Exposition (B) HR
3.18
1.08 8.72 .003
24.15
BCL6
1.46
0.64 5.24 .022
0.23
OCT2
2.09
1.06 3.93 .04
8.11
95% CI
2.91199.8 0.070.81 1.0264.39
CD, Coefficient of determination; CI, confidence interval; HR, hazard ratio.
lymphomas in this study failed to stain for this marker. This is in line with former studies arguing for a GC origin of both groups.20 CD138 (syndecan-1) is expressed in the post-GC phase of B-cell development. Its expression is typically seen in plasma cells and in plasmocytoid lymphocytes and their neoplasms.21 As stated by Slater,22 the data about CD138 expression in primary cutaneous B-cell lymphoma are limited. Our results are concordant with previous studies, showing that this antigene is not expressed in primary cutaneous B-cell lymphoma.23,24 In summary, it is very likely that both entities of PCDLBCL derive neither from naı¨ve nor from postGC B cells. Focusing on markers expressed in GCs such as BCL6, OCT2, and MUM1, we found significant differences in both entities. BCL6 is expressed by GC B cells of secondary follicles in the dark and the light zone. Interestingly, MUM1 is also expressed by B cell of the light zone of the GC but it is associated with a later stage of B-cell maturation. It is involved in the
594 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
progression from GC cells to plasma cells. MUM1positive GC cells usually fail to express BCL6.25 OCT2 is of relevance for the terminal B-cell differentiation. Interestingly, in gene expression profiles of diffuse LBCL, BCL6 belongs to the GC profile and MUM-1 and OCT2 belong to the ABC profile.8,26 This was mirrored in our study on protein level. We found BCL6 significantly more often expressed in LFCL without being restricted to this group (P = .02). This is supported by data from Kodama et al23 showing a strong trend in the same direction. It also corresponds with the data by Sundram et al.27 In their study, LBCL-L stained significantly less often positive for BCL6. On the other hand, Hoefnagel et al28,29 demonstrated BCL6 expression in most cases with LBCL-L and found no difference to primary cutaneous follicular lymphoma for this marker. BCL6 is not only known to be a marker for the GC type of diffuse LBCL. Furthermore, its expression has been linked to a favorable survival.8,27,30-32 Pooling all 35 cases of our study, our data demonstrate a positive prognostic impact of BCL6 expression on the DSS and OS for PCDLBCL. MUM1 is expressed significantly more frequently in LBCL-L than in LFCL. MUM1 is a marker for the ABC type of diffuse LBCL and it is associated with an inferior survival in systemic diffuse LBCL.33,34 The WHO/EORTC classification describes that MUM1 expression is characteristic for LBCL-L.1 At that time, the statement was based on limited data from Hoefnagel et al28 (16 cases) and Paulli et al24 (5 cases). In the meantime, the significance of this hypothesis was confirmed by larger immunohistochemical studies.23,27,29 Our data further validate these results showing MUM1 expression in 19 of 21 LBCL-L cases but in none of the LFCL cases (only 5/14 cases stained positive in \30% of tumor cells; as mentioned, above this was considered negative using a threshold of at least 50% tumor cells). The difference between both groups is highly significant (P \ .001). Moreover, we demonstrate that MUM1 is of prognostic impact also in primary cutaneous lymphomas. Its expression was associated with a significantly worse DSS and OS. OCT2 belongs to the POU family of transcription factors. It is expressed not only in a subtype of B cells but also in activated T cells, macrophages, and neuronal cells.35-37 The physiologic function of OCT2 is a matter of debate but it seems to be necessary for terminal B-cell differentiation. Its overexpression has been demonstrated in different types of lymphoma including PCDLBCL. Hoefnagel et al8 demonstrated an overexpression of OCT2
significantly more often in LBCL-L than in LFCL using real-time polymerase chain reaction and microarray techniques. OCT2 was associated with the ABC profile. Using immunohistochemistry, the same group recently published data about OCT2 expression that confirmed the gene-expression findings.29 In our study we found a positive staining with the OCT2 antibody significantly more often in LBCL-L than in LFCL. Its expression was significantly associated with a worse DSS and OS. The production of the antiapoptotic protein BCL2 is physiologically suppressed in the GC phase of B-cell development.38 In almost all other phases of development, B cells stain positive with antibodies against this protein. The same applies for B-cell lymphomas that stain positive for BCL2 if they derive from the pre-GC or post-GC phase of B-cell development. Nodal diffuse LBCLs frequently express BCL2 regardless of whether they belong to the ABC or the GC subtype.13,33,34 For LFCLs the matter is more complicated. Although nodal LFCLs usually stain positive for BCL2, the cutaneous variant of LFCLs is usually negative for this marker. In primary cutaneous B-cell lymphomas, a positive staining for BCL2 is almost restricted to marginal zone lymphoma and LBCL-L.1,3,6,11,23,39 For PCDLBCL its expression has been shown to be the strongest independent prognostic factor of survival.11 Although, in systemic diffuse LBCL, BCL2 is an important negative prognostic marker especially in the ABC subgroup.32,40,41 Our study further confirms the different frequency of BCL2 expression in both investigated groups (P \ .001). A positive staining was associated with an inferior OS and DSS (P = .024 and .039, respectively). In our study all cases with predominance of confluent sheets of round noncleaved lymphoid cells resembling large centroblasts and or immunoblasts stained positive for BCL2. This is in contrast to the study of Kodama et al,23 who identified rare cases with predominance of large round cells and absence of BCL2 expression. They propose that those cases may represent a distinct subtype. Although nearly all LBCL-Ls overexpress BCL2, the reason for this remains unclear. In nodal lymphoma the translocation t(14;18) is known to be one of the most important mechanism leading to BCL2 overexpression.42 This mechanism seems to be of minor relevance in PCDLBCL.7,43 An amplification of the BCL2 gene is demonstrable in the minority of primary cutaneous B-cell lymphomas.6 A recent study has detected BCL2 amplification in 6 of 12 LBCL-Ls.44 In a subgroup of nodal follicular lymphoma, OCT2 transcription factors regulate BCL2 expression by inducing BCL2 promoter activity.45
J AM ACAD DERMATOL
Hallermann et al 595
VOLUME 56, NUMBER 4
In this context it is of interest that BCL2 and OCT2 staining is favorably seen in the same group of PCDLBCL. Further studies have to clarify whether this is more than a coincidence and whether OCT2 overexpression interacts with the BCL2 expression also in primary cutaneous B-cell lymphomas. FOXP1, a winged helix transcription factor, is expressed by various tissues including B cells in different stages of development within and outside the GC but not by plasma cells.46 In systemic diffuse LBCL, FOXP1 is linked to the ABC type.33,47 In PCDLBCL the expression of FOXP1 is associated with the LBCL-L type.8,23 Concordantly in our study, LBCL-L stains more often positive for FOXP1 than does LFCL (17/21 vs 2/14, P \ .001). In systemic diffuse LBCL it is matter of debate whether FOXP1 is associated with a bad prognosis or not.33,48 In our study on PCDLBCL, FOXP1 had no effect on DSS and OS. Rearrangements of the cMYC oncogene are the hallmark of Burkitt’s lymphoma and are also seen in about 10% of nodal diffuse LBCL.49,50 Recently our group demonstrated its significance for PCDLBCL by demonstrating the translocation t(8,14) in 5 of 14 LBCL-Ls.7 By gene expression profiling, Hoefnagel et al8 demonstrated an increased expression of cMYC in LBCL-L compared with LFCL. To our knowledge, immunohistochemical data about cMYC in primary cutaneous B-cell lymphoma are missing. In this study we showed a positive staining for cMYC in 9 of 35 cases of PCDLBCL, underlining its significance also in cutaneous lymphomas, but there were no significant differences in frequency between both groups and no prognostic impact (P = .63 and P = .96, respectively). In conclusion, the clinical and immunohistologic data presented in this study further confirm the significance and reliability of stratifying PCDLBCL according to the EORTC/WHO classification. Moreover, the attribution of GC profile to LFCL and of ABClike profile to LBCL-L was further corroborated by the determination of characteristic staining patterns with antibodies for GC or ABC profile. We identified clinical (location, multiple lesions, ulceration) and histologic (BCL2, BCL6, OCT2, MUM1) parameters of prognostic significance. All 5 patients in our study who died as a result of the disease presented with multiple tumors on the leg with an OCT21 and BCL21 phenotype. Because of the relatively low case number, further multicenter studies are necessary to validate these results. Using commonly available antibodies, it is possible to earn similar prognostic and diagnostic information compared to much more complex microarray techniques. In addition, in times when microarray techniques revolute lymphomas classifications,
the use of immunohistochemistry for diagnostic and prognostic routine seems to be still of great value. We recommend the use of BCL6, BCL2, MUM1, and OCT2 to estimate the prognosis in patients with PCDLBCL. This may be a significant clue for the choice of the first treatment. The authors thank Margret Leygraf and coworkers for their skilled technical assistance. We sincerely thank Alison Banham for the supply of the FOXP1 antibody. REFERENCES 1. Willemze R, Jaffe ES, Burg G, Cerroni L, Berti E, Swerdlow SH, et al. WHO-EORTC classification for cutaneous lymphomas. Blood 2005;105:3768-85. 2. Burg G, Kempf W, Cozzio A, Feit J, Willemze R, Jaffe ES, et al. WHO/EORTC classification of cutaneous lymphomas 2005: histological and molecular aspects. J Cutan Pathol 2005; 32:647-74. 3. Vermeer MH, Geelen FA, van Haselen CW, van Voorst Vader PC, Geerts ML, van Vloten WA, et al. Primary cutaneous large B-cell lymphomas of the legs: a distinct type of cutaneous B-cell lymphoma with an intermediate prognosis; Dutch cutaneous lymphoma working group. Arch Dermatol 1996;132:1304-8. 4. Geelen FA, Vermeer MH, Meijer CJ, Van der Putte SC, Kerkhof E, Kluin PM, et al. Bcl-2 protein expression in primary cutaneous large B-cell lymphoma is site-related. J Clin Oncol 1998; 16:2080-5. 5. Mao X, Lillington D, Child F, Russell-Jones R, Young B, Whittaker S. Comparative genomic hybridization analysis of primary cutaneous B-cell lymphomas: identification of common genomic alterations in disease pathogenesis. Genes Chromosomes Cancer 2002;35:144-55. 6. Hallermann C, Kaune KM, Siebert R, Vermeer MH, Tensen CP, Willemze R, et al. Chromosomal aberration patterns differ in subtypes of primary cutaneous B cell lymphomas. J Invest Dermatol 2004;122:1495-502. 7. Hallermann C, Kaune KM, Gesk S, Martin-Subero JI, Gunawan B, Griesinger F, et al. Molecular cytogenetic analysis of chromosomal breakpoints in the IGH, MYC, BCL6, and MALT1 gene loci in primary cutaneous B-cell lymphomas. J Invest Dermatol 2004;123:213-9. 8. Hoefnagel JJ, Dijkman R, Basso K, Jansen PM, Hallermann C, Willemze R, et al. Distinct types of primary cutaneous large B-cell lymphoma identified by gene expression profiling. Blood 2005;105:3671-8. 9. Willemze R, Kerl H, Sterry W, Berti E, Cerroni L, Chimenti S, 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 1997;90:354-71. 10. Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, Rosenwald A, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000;403:503-11. 11. Anagnostopoulos I, Stein H. Large B-cell lymphomas: variants and entities. Pathologe 2000;21:178-89. 12. Grange F, Petrella T, Beylot-Barry M, Joly P, D’Incan M, Delaunay M, et al. Bcl-2 protein expression is the strongest independent prognostic factor of survival in primary cutaneous large B-cell lymphomas. Blood 2004;103:3662-8. 13. Neville JA, Lang W, Fleischer AB Jr. Errors in the Archives of Dermatology and the Journal of the American Academy of Dermatology from January through December 2003. Arch Dermatol 2006;142:737-40.
596 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
14. Grange F, Bekkenk MW, Wechsler J, Meijer CJ, Cerroni L, Bernengo M, et al. Prognostic factors in primary cutaneous large B-cell lymphomas: a European multicenter study. J Clin Oncol 2001;19:3602-10. 15. Smith BD, Smith GL, Cooper DL, Wilson LD. The cutaneous B-cell lymphoma prognostic index: a novel prognostic index derived from a population-based registry. J Clin Oncol 2005; 23:3390-5. 16. Zinzani PL, Quaglino P, Pimpinelli N, Berti E, Baliva G, Rupoli S, et al, for the Italian Study Group for Cutaneous Lymphomas. Prognostic factors in primary cutaneous B-cell lymphoma: the Italian study group for cutaneous lymphomas. J Clin Oncol 2006;24:1376-82. 17. Harada S, Suzuki R, Uehira K, Yatabe Y, Kagami Y, Ogura M, et al. Molecular and immunological dissection of diffuse large B cell lymphoma: CD51, and CD5 with CD101 groups may constitute clinically relevant subtypes. Leukemia 1999;13:1441-7. 18. Yamaguchi M, Seto M, Okamoto M, Ichinohasama R, Nakamura N, Yoshino T, et al. De novo CD51 diffuse large B-cell lymphoma: a clinicopathologic study of 109 patients. Blood 2002;99:815-21. 19. Tagawa H, Tsuzuki S, Suzuki R, Karnan S, Ota A, Kameoka Y, et al. Genome-wide array-based comparative genomic hybridization of diffuse large B-cell lymphoma: comparison between CD5-positive and CD5-negative cases. Cancer Res 2004;64: 5948-55. 20. Gellrich S, Rutz S, Golembowski S, Jacobs C, von Zimmermann M, Lorenz P, et al. Primary cutaneous follicle center cell lymphomas and large B cell lymphomas of the leg descend from germinal center cells: a single cell polymerase chain reaction analysis. J Invest Dermatol 2001;117:1512-20. 21. Sebestyen A, Berczi L, Mihalik R, Paku S, Matolcsy A, Kopper L. Syndecan-1 (CD138) expression in human non-Hodgkin lymphomas. Br J Haematol 1999;104:412-9. 22. Slater DN. The new World Health Organization-European Organization for Research and Treatment of Cancer classification for cutaneous lymphomas: a practical marriage of two giants. Br J Dermatol 2005;153:874-80. 23. Kodama K, Massone C, Chott A, Metze D, Kerl H, Cerroni L. Primary cutaneous large B-cell lymphomas: clinicopathologic features, classification, and prognostic factors in a large series of patients. Blood 2005;106:2491-7. 24. Paulli M, Viglio A, Vivenza D, Capello D, Rossi D, Riboni R, et al. Primary cutaneous large B-cell lymphoma of the leg: histogenetic analysis of a controversial clinicopathologic entity. Hum Pathol 2002;33:937-43. 25. Falini B, Fizzotti M, Pucciarini A, Bigerna B, Marafioti T, Gambacorta M, 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 2000; 95:2084-92. 26. Wright G, Tan B, Rosenwald A, Hurt EH, Wiestner A, Staudt LM. A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lymphoma. Proc Natl Acad Sci U S A 2003;100:9991-6. 27. Sundram U, Kim Y, Mraz-Gernhard S, Hoppe R, Natkunam Y, Kohler S. Expression of the bcl-6 and MUM1/IRF4 proteins correlate with overall and disease-specific survival in patients with primary cutaneous large B-cell lymphoma: a tissue microarray study. J Cutan Pathol 2005;32:227-34. 28. Hoefnagel JJ, Vermeer MH, Jansen PM, Fleuren GJ, Meijer CJ, Willemze R. Bcl-2, Bcl-6 and CD10 expression in cutaneous B-cell lymphoma: further support for a follicle center cell origin and differential diagnostic significance. Br J Dermatol 2003; 149:1183-91.
29. Hoefnagel JJ, Mulder MM, Dreef E, Jansen PM, Pals ST, Meijer CJ, et al. Expression of B-cell transcription factors in primary cutaneous B-cell lymphoma. Mod Pathol 2006;19:1270-6. 30. Lossos IS, Jones CD, Warnke R, Natkunam Y, Kaizer H, Zehnder JL, et al. Expression of a single gene, BCL-6, strongly predicts survival in patients with diffuse large B-cell lymphoma. Blood 2001;98:945-51. 31. Rosenwald A, Wright G, Chan WC, Connors JM, Campo E, Fisher RI, et al. Lymphoma/leukemia molecular profiling project: the use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma. N Engl J Med 2002;346:1937-47. 32. Berglund M, Thunberg U, Amini RM, Book M, Roos G, Erlanson M, et al. Evaluation of immunophenotype in diffuse large B-cell lymphoma and its impact on prognosis. Mod Pathol 2005; 18:1113-20. 33. Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie J, Ott G, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood 2004;103:275-82. 34. Lossos IS, Czerwinski DK, Alizadeh AA, Wechser MA, Tibshirani R, Botstein D, et al. Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes. N Engl J Med 2004;350:1828-37. 35. Staudt LM, Clerc RG, Singh H, LeBowitz JH, Sharp PA, Baltimore D. Cloning of a lymphoid-specific cDNA encoding a protein binding the regulatory octamer DNA motif. Science 1988; 241:577-80. 36. Kang SM, Tsang W, Doll S, Scherle P, Ko HS, Tran AC, et al. Induction of the POU domain transcription factor Oct-2 during T-cell activation by cognate antigen. Mol Cell Biol 1992; 12:3149-54. 37. Dunn TL, Ross IL, Hume DA. Transcription factor Oct-2 is expressed in primary murine macrophages. Blood 1996; 88:4072. 38. Hockenbery DM, Zutter M, Hickey W, Nahm M, Korsmeyer SJ. BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc Natl Acad Sci U S A 1991;88:6961-5. 39. Cho-Vega JH, Vega F, Rassidakis G, Medeiros LJ. Primary cutaneous marginal zone B-cell lymphoma. Am J Clin Pathol 2006;125(Suppl):S38-49. 40. Iqbal J, Neppalli VT, Wright G, Dave BJ, Horsman DE, Rosenwald A, et al. BCL2 expression is a prognostic marker for the activated B-cell-like type of diffuse large B-cell lymphoma. J Clin Oncol 2006;24:961-8. 41. Muris JJ, Meijer CJ, Vos W, van Krieken JH, Jiwa NM, Ossenkoppele GJ, et al. Immunohistochemical profiling based on Bcl-2, CD10 and MUM1 expression improves risk stratification in patients with primary nodal diffuse large B cell lymphoma. J Pathol 2006;208:714-23. 42. Willis TG, Dyer MJ. The role of immunoglobulin translocations in the pathogenesis of B-cell malignancies. Blood 2000;96:808-22. 43. Child FJ, Russell-Jones R, Woolford AJ, Calonje E, Photiou A, Orchard G, et al. Absence of the t(14;18) chromosomal translocation in primary cutaneous B-cell lymphoma. Br J Dermatol 2001;144:735-44. 44. Dijkman R, Tensen CP, Jordanova ES, Knijnenburg J, Hoefnagel JJ, Mulder AA, et al. Array-based comparative genomic hybridization analysis reveals recurrent chromosomal alterations and prognostic parameters in primary cutaneous large B-cell lymphoma. J Clin Oncol 2006;24:296-305. 45. Heckman CA, Duan H, Garcia PB, Boxer LM. Oct transcription factors mediate t(14;18) lymphoma cell survival by directly regulating bcl-2 expression. Oncogene 2006;25:888-98.
J AM ACAD DERMATOL
Hallermann et al 597
VOLUME 56, NUMBER 4
46. Banham AH, Beasley N, Campo E, Fernandez PL, Fidler C, Gatter K, et al. The FOXP1 winged helix transcription factor is a novel candidate tumor suppressor gene on chromosome 3p. Cancer Res 2001;61:8820-9. 47. Shaffer AL, Rosenwald A, Staudt LM. Lymphoid malignancies: the dark side of B-cell differentiation. Nat Rev Immunol 2002; 2:920-32. 48. Banham AH, Connors JM, Brown PJ, Cordell JL, Ott G, Sreenivasan G, et al. Expression of the FOXP1 transcription factor is strongly associated with inferior survival in patients
with diffuse large B-cell lymphoma. Clin Cancer Res 2005; 11:1065-72. 49. Akasaka T, Akasaka H, Ueda C, Yonetani N, Maesako Y, Shimizu A, et al. Molecular and clinical features of non-Burkitt’s, diffuse large-cell lymphoma of B-cell type associated with the c-MYC/ immunoglobulin heavy-chain fusion gene. J Clin Oncol 2000; 18:510-8. 50. Kawasaki C, Ohshim K, Suzumiya J, Kanda M, Tsuchiya T, Tamura K, et al. Rearrangements of bcl-1, bcl-2, bcl-6, and c-myc in diffuse large B-cell lymphomas. Leuk Lymphoma 2001;42:1099-106.
T
he group of primary cutaneous diffuse large B-cell lymphoma (PCDLBCL) is heterogeneous. According to the World Health Organization (WHO)/European Organization for Research and Treatment of Cancer (EORTC) classification,
From the Department of Dermatology, Institute for Tumors of the Skin,a and Department of Medical Informatics and Biomathematics,b University of Munster. Funding sources: None. Conflicts of interest: None identified. Accepted for publication December 19, 2006. Reprint requests: Christian Hallermann, MD, Dorbaumstrasse 300, 48157 Munster, Germany. E-mail: [email protected]. Published online February 12, 2007. 0190-9622/$32.00 ª 2007 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2006.12.026
588
Abbreviations used: ABC: DSS: EORTC:
activated B cell disease-specific survival European Organization for Research and Treatment of Cancer GC: germinal center LBCL: large B-cell lymphoma LBCL-L: large B-cell lymphoma, leg type LFCL: follicle center lymphoma OS: overall survival PCDLBCL: primary cutaneous diffuse large B-cell lymphoma WHO: World Health Organization
they are divided into large B-cell lymphoma (LBCL) leg type (LBCL-L), LBCLs as a variant of primary cutaneous follicle center lymphoma (LFCL), and a third group including all LBCLs that do not fulfill the criteria of the first two groups.1,2
J AM ACAD DERMATOL
Hallermann et al 589
VOLUME 56, NUMBER 4
This grouping mirrors not only the different clinical behavior of both main entities; moreover, it also correlates differences in cytogenetic aberrations and different gene expression profiles.3-8 Although it is a well-known fact that PCDLBCL-L has a poorer outcome compared with LFCL, the reason for this is still a matter of debate. It seems likely that not the location but the differences in morphology and gene expressions are reasons for the different behavior. For this reason the WHO/EORTC classification chose the term ‘‘leg type’’ and not the former designation ‘‘large B-cell lymphoma of the leg’’ as proposed by the EORTC in 1997.9 Using microarray techniques it is possible to distinguish 3 different groups of nodal diffuse LBCLs. They are characterized by different gene expression profiles: the germinal center (GC) profile, the activated B cell (ABC)-like profile, and a third group that does not fulfill the criteria of the first two groups.10 The GC subtype is defined by the expression of a panel of genes characteristic for normal GC cells. The ABC subtype is defined by the expression of genes characteristic for activated blood lymphocytes. The ABC subtype has a poorer outcome compared with the GC subtype.10 Microarray analyses in PCDLBCLs confirm the relevance of grouping into LFCL and LBCL-L. Both show profiles similar to that of GC type and ABC type, respectively. Several marker genes have been identified that express significantly differently in both groups.8 Although gene profiling is very useful in understanding the molecular basis of diffuse LBCL, it is not practical from a routine diagnostic perspective. In this regard, the purpose of the study was to further define an armamentarium of easily applied immunohistochemical stains to accurately prognosticate diffuse LBCL of the skin.
METHODS Patients and samples For examination we used formalin-fixed skin biopsy specimens from patients with LBCL-L and LFCL. An informed consent was obtained from each patient before biopsy. All biopsy specimens were taken for diagnostic reasons. For inclusion into our study the lymphoma had to be restricted to the skin at the time of diagnosis. This was confirmed by appropriate staging procedures. Patients with immunosuppression have been excluded from the study. We only included cases composed of diffuse arranged lymphoid cells resembling large centrocyts, centroblasts, immunoblasts, or a combination of these. For inclusion into the study, the number of large cells had to be at least 50% of the B cells. As defined by Anagnostopoulos and Stein,11 a
Fig 1. Biopsy specimen of representative lesion from patient with follicle center lymphoma (case 25). Mixture of large cleaved cells resembling centrocytes and some centroblasts not arranged in sheets. (Hematoxylin-eosin stain; original magnification: 31000.)
lymphoma cell is considered large if it is at least twice as big as a normal lymphocyte (bystander cell). The diagnoses were made using the criteria published in the WHO/EORTC classification.1 LFCL was defined as a lymphoid neoplasia composed of diffuse lesions with predominantly large cleaved cells (resembling large centrocytes) with a variable infiltrate of centroblasts, which were not arranged in confluent sheets (Fig 1). LBCL-L was defined as a lymphoid neoplasia with predominance of large noncleaved round cells (resembling centroblasts and immunoblasts) arranged in confluent sheets with a diffuse growth pattern (Fig 2, A). Cases with an exclusively follicular growth pattern were excluded. Because of its heterogeneity and its ill definition, patients with the diagnosis ‘‘PCDLBCL, other’’ were also excluded from the study. Clinical parameters The following regions of the body were defined: head and neck; arm; trunk; and leg. The therapeutic response was classified into 4 categories (complete remission, partial remission, stable disease, and progression). Methods A tumor biopsy specimen was fixed in formalin, embedded in paraffin, and cut into 2- to 4-m sections. In addition to hematoxylin-eosin and Giemsa staining, immunohistochemistry was performed using standard immunoperoxidase techniques. The sections were dewaxed in xylol (Merck, Darmstadt, Germany) and rehydrated in serial dilutions of ethanol. The used antibodies are listed in Table I. An autostainer was used in combination with a detection kit (Chem-Mate LSAB-KIT, DAKO, Glostrup, Denmark). All biopsy specimens were reviewed
590 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
Fig 2. Biopsy specimen of representative lesion from patient with large B-cell lymphoma, leg type (case 7). A, There are confluent sheets of large round B cells (resembling centroblasts and immunoblasts). B to D, Neoplastic cells stain positive for OCT2, MUM1, and BCL2. (A, Hematoxylin-eosin stain; B, OCT2; C, MUM1; D, BCL2; original magnifications: A, 31000; B to D, 3400.)
Table I. Antibodies used in study Antigene
Clone
Pretreatment* Dilutiony
CD20 CD79a CD3 CD138 CD5 BCL2 BCL6 cMYC
L26 JCB117 F7.2.38 MI15 CD5/54/F6 124 PG-B6p CT14.G4
OCT2
Polyclonal Heat, pH9,0
MUM-1 MUM1p FOX-P1 JC12
e Heat, pH9 Heat, pH9 Heat, pH6,1 Heat, pH9 Heat, pH9 Heat, pH9 Heat, pH6,1
Heat, pH6,1 Heat, pH6,1
1:1 1:4 1:50 1:100 1:100 1:6 1:20 1:25 1:25 1:50 1:100
Manufacturer
DAKOz DAKOz DAKOz DAKOz DAKOz DAKOz DAKOz Acris Antibodies§ LabVision Corpk DAKOz Dr Alison Banham{
*Slides were heated in an autoclave using the target retrieval solution (DAKO). y For dilution the antibody Diluent (DAKO) was used. z Glostrup, Denmark. § Hiddenhausen, Germany. k Fremont, Calif. { Dr Alison Banham, Director of the LRF Lymphoma Antigens Programme and Oxford University Research Lecturer, Nuffield Department of Clinical Laboratory Sciences, University of Oxford, United Kingdom.
independently by two experts in dermatopathology (C. H. and H-J. S.). The cases were discussed at the multiheaded microscope. According to Grange et al,12 immunohistologic staining was considered positive if more than 50% of the lymphoma cells were stained. Histopathologic pictures were captured with a microscope (Axioscop 2 plus, Zeiss, Jena, Germany) using a camera (Axiocam HRC, Zeiss). Statistical analysis The statistical methods were chosen according the recently published recommendations by Neville et al.13 Comparisons between the two groups of patients were performed using the usual chi-square test or, when appropriate, Fisher’s exact test for categorical variables and the Student t test (after confirmation of normal distribution by Kolmogorov-Smirnov test) for categorical variables (age). Correlations between data were done by Spearman’s rank correlation. Observed survival and specific survival were estimated using the product-limited method of Kaplan and Meier. End point was the death of the patient (disease related or nonspecific) or the last follow-up. For surviving patents the study end
J AM ACAD DERMATOL
Hallermann et al 591
VOLUME 56, NUMBER 4
point was January 31, 2006. Parameters assessed by univariate analysis included sex, age (\/[70 years), location on the leg, ulcerated lesions, multiple lesions, and BCL2, BCL6, OCT2, FOXP-1, MYC, and MUM1 as determined by immunohistochemistry. Differences between survival were analyzed using the log rank chi-square test or, in cases of crossing functions, the Breslow (generalized Wilcoxon test) or Tarone test. Factors to be shown significant in above-mentioned Kaplan-Meier test and in univariate Cox analysis were analyzed by using a multivariate Cox proportional hazard regression model (method: stepwise forward Wald). All analyses were performed using a statistical software package (SPSS/PC, Version 13.0, SPSS Inc, Chicago, Ill). All P values were based on two-tailed method. Statistical analysis with P less than .05 was considered as statistically significant.
RESULTS Clinical parameters In all, 35 patients with primary cutaneous LBCL were included into the study; 16 of them were male and 19 were female. The lymphomas were classified as LFCL in 14 cases and as LBCL-L in 21 cases. The mean age was 64.6 years (SD 14.8, median 68 years, range 23-90 years) at the time of diagnosis and the mean follow-up was 60.2 months (median 59.5 months, range 4-156 months). The distribution of the lesions included legs (14 cases with exclusive location on the leg: 6 right, 6 left, 2 bilateral), head (14 cases), trunk (4 cases), and arms (2 cases). Three of 35 patients presented ulcerated nodules (cases 6, 7, and 18). Fourteen patients had multiple lesions ( $ 2 nodules). One of them had an involvement of more than one region of the body, as defined above. The most frequent primary therapies were radiation (21 cases) or excision (11 cases). In none of the cases was chemotherapy used as first-line therapy. In 7 of 21 cases treated by radiotherapy and in 4 of 11 cases treated by excision, at least one further therapeutic intervention was necessary. In 19 of 35 patients, a complete remission was attained. In 3 cases a cyclophosphamide, Adriamycin, vincristine, prednisone polychemotherapy was performed as second-line therapy (cases 8, 9, and 13). At the end of the follow-up period, 17 patients were alive without disease, 6 were alive with clinical signs of disease, in 7 patients the death could not be related to the lymphoma, and 5 patients died as a result of the lymphoma. For the patients with complete remission at the end point of the study, the mean follow-up time was 70.58 months (median 61 months, range 12-156 months).
The 5-year overall survival (OS) was 67.9% and the 5-year disease-specific survival (DSS) was 81.5%. Comparing both entities Focusing on the disease entity according to the WHO/EORTC classification, we found significant clinical differences between both groups. The mean age of the patients with LBCL-L was 69.9 years (70 years median) compared with a mean age of 65.6 years (54 years median) in the other group (P = .008). As expected, the clinical presentation differed between the diseases. Patients with LFCL presented mostly with single (93%), nonulcerated (100%) tumors on the head (64%). Patients with LBCL-L presented mostly with multiple (61%), and sometimes ulcerated (14%) tumors on the lower extremities (76%). Patients with LBCL-L have a significantly worse OS and DSS compared with LFCL (P = .012 and P = .04, respectively). Because of the grouping of the lymphomas into both entities based on morphologic and cytologic criteria in this study, it can be concluded that those morphologic criteria are of statistically significant prognostic impact and the entities. Comparing the immunophenotype of both entities we found LBCL-L significantly more often positive for BCL2, OCT2, MUM1, and FOX-P1 and negative for BCL6 than LFCL (chi-square test of Pearson: P = .01, P = .046, P \ .001, P \ .001, and P = .02, respectively). The details are listed in Table II. Histology is shown in Fig 2. Prognostic features Without focusing on the disease entities, we identified parameters associated with a worse survival (Table III and IV). Location on the leg (OS: P = .02, DSS: P = .01), ulceration (OS: P = .002, DSS: P = .07), and multiple nodules (OS: P = .033, DSS: P = .003) were significantly associated with both negative OS and negative DSS. As expected, the age at time of diagnosis (\ vs [70 years) was significantly associated with a worse OS (P = .049) but not with DSS (P = .59). Positive staining for BCL2, OCT2, and FOXP1 was clearly associated with progress under first-line therapy (Spearman correlation: P = .039, P = .016, and P = .039, respectively). Moreover, the expression of BCL2 (OS: P = .02, DSS: P = .039), MUM1 (OS: P = .005, DSS: P = .013), and OCT2 (OS: P = .015, DSS: P = .034) was significantly associated with a bad OS and DSS. On the other hand the expression of BCL6 was an indicator for a better prognosis (OS: P = .014, DSS: P = .001).
592 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
Table II. Clinical and immunohistologic data No.
Type
Age, y/sex
1 2 3 4 5 6 7 8 9
LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L
75/M 66/M 68/F 83/F 73/F 57/M 61/F 60/F 61/M
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LBCL-L LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL LFCL
76/F 89/F 61/F 64/M 75/M 71/M 54/F 69/M 90/F 52/F 75/M 76/F 89/M 66/M 43/F 49/F 58/M 23/F 59/F 54/M 48/M 69/F 45/M 45/F 72/M 73/F
Lesions
Mu/leg Mu/leg Mu/leg Si/head Si/leg Mu/leg Mu/leg Mu/leg Mu/leg and other Si/leg Mu/leg Si/arm Mu/leg Mu/leg Mu/leg Si/head Mu/leg Si/head Si/head Si/head Mu/leg Si/head Si/head Si/trunk Si/head Si/head Si/head Si/head Si/arm Si/head Si/trunk Si/head Mu/trunk Si/trunk Si/head
First therapy
Therapeutic response
Follow-up time, mo
Follow-up
PUVA Rt Ex Ex Rt Ex Ex Rt Rt
Pro Pro CR CR PR Pro CR Pro Pro
64 6 19 69 120 36 118 11 10
A1 DD A D D DD A DD D
1 1 1 1 1 1 1 1 1
1 1 1 1 1
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1
1 1 1
Ex Rt Ex Ex Ex Rt Rt If Rt Rituximab Ex Rt Rt Rt Rt Ex Rt Ex Rt Rt Rt Rt Rt Rt Rt Rt
CR Pro CR Pro Pro Pro PR PR Pro CR CR CR PR CR CR Pro CR CR CR CR CR CR CR CR CR Pro
36 10 57 12 24 47 145 72 12 35 126 68 4 50 12 150 156 71 61 88 106 71 61 43 58 72
D D A A1 DD DD A1 A1 D A A A D A A A1 A A A A A A A A A A1
1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1
BCL2 BCL6 MUM1 FOXP1 OCT2 cMYC
A1, Alive with disease; A, alive without disease; CR, complete remission; D, died unrelated or not obviously related to lymphoma; DD, lymphoma-related death; Ex, excision; F, female; IF, interferon; LBCL-L, large B-cell lymphoma, leg type; LFCL, follicle center lymphoma; M, male; Mu, multiple lesions; PR, partial remission; Pro, progression; PUVA, psoralen plus ultraviolet A; Rt, radiation; Si, single lesion.
Factors shown to be of prognostic significance in Kaplan-Meier tests and significant in univariate Cox analysis (age; location on the leg; multiple lesions; ulceration; positive staining for BCL2, MUM1, and OCT2; and negative staining for BCL6) were analyzed using a multivariate Cox proportional hazard regression model. In this test ulceration, BCL6 and OCT2 have been shown to be prognostically significant. Because of the low case number, all 12 patients who died were included into this analysis (Table V).
DISCUSSION In this study we investigated clinical and immunophenotypic characteristics in a large number
of PCDLBCLs including cases with LBCL-L and LFCL. Both groups show significant differences in clinical presentation and prognosis. In LBCL-L the patients are usually older and present more often with multiple lesions. Pooling all lymphomas together we identified diagnosis, location on the leg, ulceration, and multiple lesions at presentation as significant prognostic factors. In our study all 5 patients who died as a result of the lymphoma in the first 5 years presented with multiple nodules on the legs. Location on the leg and multiple lesions has also been reported as prognostically significant in further European and North American studies.14,15 One of the largest
Hallermann et al 593
J AM ACAD DERMATOL VOLUME 56, NUMBER 4
Table III. Comparison between the two entities Total (%)
N Clinical data Sex Male Female Age (\70 y) Multiple lesions Ulceration Immunohistologic data BCL6 BCL2 OCT2 MUM1 FOXP1 cMYC CD138 CD5 Disease-specific survival 3y 5y Overall survival 3y 5y
35
LFCL (%)
14
LBCL-L (%)
P (LFCL e LBCL-L)
21
16 19 22 14 3
(46) (54) (63) (40) (8)
7 7 11 1 0
(50) (50) (78) (7) (0)
9 12 11 13 3
(43) (57) (52) (61) (14)
25 22 22 19 19 9 0 0
(71) (63) (62) (54) (54) (26) (0) (0)
13 1 6 0 2 3 0 0
(93) (7) (42) (0) (14) (21) (0) (0)
12 21 16 19 17 6 0 0
(57) (100) (76) (90) (80) (28) (0) (0)
100 100
76.53 68.03
91.67 91.67
59.42 52.82
.008 .01
.022 .001 .046 .001 .001
Only significant P values are listed. LBCL-L, Large B-cell lymphoma, leg type; LFCL, follicle center lymphoma.
studies on this topic has recently been published by Zinzani et al.16 In that study the location on the leg and the disease entity are relevant prognostic parameters in primary cutaneous B-cell lymphomas. Our data further corroborate these studies and further emphasize the clinical significance of the WHO/ EORTC classification. To our knowledge, until now the prognostic impact of ulceration has not been clearly demonstrated. Because of the low case number in our study, larger trials are necessary to validate this finding. Using immunohistochemical antibodies against different gene products of B-cell development and oncogenes, we found a different pattern in both groups. These differences were significant for some of the markers. None of the lymphomas in our study stained positive for CD138 or CD5. Naı¨ve B cells are usually positive for CD5. It is consistently expressed in B cells of the mantle zone, B-CLL, and in mantle cell lymphoma. Moreover, it is expressed in about 10% of systemic diffuse LBCLs in which its expression is associated with a worse prognosis.17-19 All analyzed
Table IV. Prognostic features in primary cutaneous diffuse large B-cell lymphoma (irrespective of the subtype) Parameter
DSS, P
OS, P
Age [ 70 y Male Location on leg Multiple lesions Ulceration BCL21 BCL6 OCT21 MUM11 cMYC1 FOXP11
.59 .21 .016* .003* .007* .04* .001* .034* .013* .96 .35
.048* .49 .02* .03* .002* .02* .014* .015* .005* .92 .31
DSS, Disease-specific survival; OS, overall survival. *Significant.
Table V. Multivariate Cox proportional hazard regression model of prognostic significant parameters (method: stepwise forward Wald) Coefficient (CD) SE
Ulceration
Wald
P
Exposition (B) HR
3.18
1.08 8.72 .003
24.15
BCL6
1.46
0.64 5.24 .022
0.23
OCT2
2.09
1.06 3.93 .04
8.11
95% CI
2.91199.8 0.070.81 1.0264.39
CD, Coefficient of determination; CI, confidence interval; HR, hazard ratio.
lymphomas in this study failed to stain for this marker. This is in line with former studies arguing for a GC origin of both groups.20 CD138 (syndecan-1) is expressed in the post-GC phase of B-cell development. Its expression is typically seen in plasma cells and in plasmocytoid lymphocytes and their neoplasms.21 As stated by Slater,22 the data about CD138 expression in primary cutaneous B-cell lymphoma are limited. Our results are concordant with previous studies, showing that this antigene is not expressed in primary cutaneous B-cell lymphoma.23,24 In summary, it is very likely that both entities of PCDLBCL derive neither from naı¨ve nor from postGC B cells. Focusing on markers expressed in GCs such as BCL6, OCT2, and MUM1, we found significant differences in both entities. BCL6 is expressed by GC B cells of secondary follicles in the dark and the light zone. Interestingly, MUM1 is also expressed by B cell of the light zone of the GC but it is associated with a later stage of B-cell maturation. It is involved in the
594 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
progression from GC cells to plasma cells. MUM1positive GC cells usually fail to express BCL6.25 OCT2 is of relevance for the terminal B-cell differentiation. Interestingly, in gene expression profiles of diffuse LBCL, BCL6 belongs to the GC profile and MUM-1 and OCT2 belong to the ABC profile.8,26 This was mirrored in our study on protein level. We found BCL6 significantly more often expressed in LFCL without being restricted to this group (P = .02). This is supported by data from Kodama et al23 showing a strong trend in the same direction. It also corresponds with the data by Sundram et al.27 In their study, LBCL-L stained significantly less often positive for BCL6. On the other hand, Hoefnagel et al28,29 demonstrated BCL6 expression in most cases with LBCL-L and found no difference to primary cutaneous follicular lymphoma for this marker. BCL6 is not only known to be a marker for the GC type of diffuse LBCL. Furthermore, its expression has been linked to a favorable survival.8,27,30-32 Pooling all 35 cases of our study, our data demonstrate a positive prognostic impact of BCL6 expression on the DSS and OS for PCDLBCL. MUM1 is expressed significantly more frequently in LBCL-L than in LFCL. MUM1 is a marker for the ABC type of diffuse LBCL and it is associated with an inferior survival in systemic diffuse LBCL.33,34 The WHO/EORTC classification describes that MUM1 expression is characteristic for LBCL-L.1 At that time, the statement was based on limited data from Hoefnagel et al28 (16 cases) and Paulli et al24 (5 cases). In the meantime, the significance of this hypothesis was confirmed by larger immunohistochemical studies.23,27,29 Our data further validate these results showing MUM1 expression in 19 of 21 LBCL-L cases but in none of the LFCL cases (only 5/14 cases stained positive in \30% of tumor cells; as mentioned, above this was considered negative using a threshold of at least 50% tumor cells). The difference between both groups is highly significant (P \ .001). Moreover, we demonstrate that MUM1 is of prognostic impact also in primary cutaneous lymphomas. Its expression was associated with a significantly worse DSS and OS. OCT2 belongs to the POU family of transcription factors. It is expressed not only in a subtype of B cells but also in activated T cells, macrophages, and neuronal cells.35-37 The physiologic function of OCT2 is a matter of debate but it seems to be necessary for terminal B-cell differentiation. Its overexpression has been demonstrated in different types of lymphoma including PCDLBCL. Hoefnagel et al8 demonstrated an overexpression of OCT2
significantly more often in LBCL-L than in LFCL using real-time polymerase chain reaction and microarray techniques. OCT2 was associated with the ABC profile. Using immunohistochemistry, the same group recently published data about OCT2 expression that confirmed the gene-expression findings.29 In our study we found a positive staining with the OCT2 antibody significantly more often in LBCL-L than in LFCL. Its expression was significantly associated with a worse DSS and OS. The production of the antiapoptotic protein BCL2 is physiologically suppressed in the GC phase of B-cell development.38 In almost all other phases of development, B cells stain positive with antibodies against this protein. The same applies for B-cell lymphomas that stain positive for BCL2 if they derive from the pre-GC or post-GC phase of B-cell development. Nodal diffuse LBCLs frequently express BCL2 regardless of whether they belong to the ABC or the GC subtype.13,33,34 For LFCLs the matter is more complicated. Although nodal LFCLs usually stain positive for BCL2, the cutaneous variant of LFCLs is usually negative for this marker. In primary cutaneous B-cell lymphomas, a positive staining for BCL2 is almost restricted to marginal zone lymphoma and LBCL-L.1,3,6,11,23,39 For PCDLBCL its expression has been shown to be the strongest independent prognostic factor of survival.11 Although, in systemic diffuse LBCL, BCL2 is an important negative prognostic marker especially in the ABC subgroup.32,40,41 Our study further confirms the different frequency of BCL2 expression in both investigated groups (P \ .001). A positive staining was associated with an inferior OS and DSS (P = .024 and .039, respectively). In our study all cases with predominance of confluent sheets of round noncleaved lymphoid cells resembling large centroblasts and or immunoblasts stained positive for BCL2. This is in contrast to the study of Kodama et al,23 who identified rare cases with predominance of large round cells and absence of BCL2 expression. They propose that those cases may represent a distinct subtype. Although nearly all LBCL-Ls overexpress BCL2, the reason for this remains unclear. In nodal lymphoma the translocation t(14;18) is known to be one of the most important mechanism leading to BCL2 overexpression.42 This mechanism seems to be of minor relevance in PCDLBCL.7,43 An amplification of the BCL2 gene is demonstrable in the minority of primary cutaneous B-cell lymphomas.6 A recent study has detected BCL2 amplification in 6 of 12 LBCL-Ls.44 In a subgroup of nodal follicular lymphoma, OCT2 transcription factors regulate BCL2 expression by inducing BCL2 promoter activity.45
J AM ACAD DERMATOL
Hallermann et al 595
VOLUME 56, NUMBER 4
In this context it is of interest that BCL2 and OCT2 staining is favorably seen in the same group of PCDLBCL. Further studies have to clarify whether this is more than a coincidence and whether OCT2 overexpression interacts with the BCL2 expression also in primary cutaneous B-cell lymphomas. FOXP1, a winged helix transcription factor, is expressed by various tissues including B cells in different stages of development within and outside the GC but not by plasma cells.46 In systemic diffuse LBCL, FOXP1 is linked to the ABC type.33,47 In PCDLBCL the expression of FOXP1 is associated with the LBCL-L type.8,23 Concordantly in our study, LBCL-L stains more often positive for FOXP1 than does LFCL (17/21 vs 2/14, P \ .001). In systemic diffuse LBCL it is matter of debate whether FOXP1 is associated with a bad prognosis or not.33,48 In our study on PCDLBCL, FOXP1 had no effect on DSS and OS. Rearrangements of the cMYC oncogene are the hallmark of Burkitt’s lymphoma and are also seen in about 10% of nodal diffuse LBCL.49,50 Recently our group demonstrated its significance for PCDLBCL by demonstrating the translocation t(8,14) in 5 of 14 LBCL-Ls.7 By gene expression profiling, Hoefnagel et al8 demonstrated an increased expression of cMYC in LBCL-L compared with LFCL. To our knowledge, immunohistochemical data about cMYC in primary cutaneous B-cell lymphoma are missing. In this study we showed a positive staining for cMYC in 9 of 35 cases of PCDLBCL, underlining its significance also in cutaneous lymphomas, but there were no significant differences in frequency between both groups and no prognostic impact (P = .63 and P = .96, respectively). In conclusion, the clinical and immunohistologic data presented in this study further confirm the significance and reliability of stratifying PCDLBCL according to the EORTC/WHO classification. Moreover, the attribution of GC profile to LFCL and of ABClike profile to LBCL-L was further corroborated by the determination of characteristic staining patterns with antibodies for GC or ABC profile. We identified clinical (location, multiple lesions, ulceration) and histologic (BCL2, BCL6, OCT2, MUM1) parameters of prognostic significance. All 5 patients in our study who died as a result of the disease presented with multiple tumors on the leg with an OCT21 and BCL21 phenotype. Because of the relatively low case number, further multicenter studies are necessary to validate these results. Using commonly available antibodies, it is possible to earn similar prognostic and diagnostic information compared to much more complex microarray techniques. In addition, in times when microarray techniques revolute lymphomas classifications,
the use of immunohistochemistry for diagnostic and prognostic routine seems to be still of great value. We recommend the use of BCL6, BCL2, MUM1, and OCT2 to estimate the prognosis in patients with PCDLBCL. This may be a significant clue for the choice of the first treatment. The authors thank Margret Leygraf and coworkers for their skilled technical assistance. We sincerely thank Alison Banham for the supply of the FOXP1 antibody. REFERENCES 1. Willemze R, Jaffe ES, Burg G, Cerroni L, Berti E, Swerdlow SH, et al. WHO-EORTC classification for cutaneous lymphomas. Blood 2005;105:3768-85. 2. Burg G, Kempf W, Cozzio A, Feit J, Willemze R, Jaffe ES, et al. WHO/EORTC classification of cutaneous lymphomas 2005: histological and molecular aspects. J Cutan Pathol 2005; 32:647-74. 3. Vermeer MH, Geelen FA, van Haselen CW, van Voorst Vader PC, Geerts ML, van Vloten WA, et al. Primary cutaneous large B-cell lymphomas of the legs: a distinct type of cutaneous B-cell lymphoma with an intermediate prognosis; Dutch cutaneous lymphoma working group. Arch Dermatol 1996;132:1304-8. 4. Geelen FA, Vermeer MH, Meijer CJ, Van der Putte SC, Kerkhof E, Kluin PM, et al. Bcl-2 protein expression in primary cutaneous large B-cell lymphoma is site-related. J Clin Oncol 1998; 16:2080-5. 5. Mao X, Lillington D, Child F, Russell-Jones R, Young B, Whittaker S. Comparative genomic hybridization analysis of primary cutaneous B-cell lymphomas: identification of common genomic alterations in disease pathogenesis. Genes Chromosomes Cancer 2002;35:144-55. 6. Hallermann C, Kaune KM, Siebert R, Vermeer MH, Tensen CP, Willemze R, et al. Chromosomal aberration patterns differ in subtypes of primary cutaneous B cell lymphomas. J Invest Dermatol 2004;122:1495-502. 7. Hallermann C, Kaune KM, Gesk S, Martin-Subero JI, Gunawan B, Griesinger F, et al. Molecular cytogenetic analysis of chromosomal breakpoints in the IGH, MYC, BCL6, and MALT1 gene loci in primary cutaneous B-cell lymphomas. J Invest Dermatol 2004;123:213-9. 8. Hoefnagel JJ, Dijkman R, Basso K, Jansen PM, Hallermann C, Willemze R, et al. Distinct types of primary cutaneous large B-cell lymphoma identified by gene expression profiling. Blood 2005;105:3671-8. 9. Willemze R, Kerl H, Sterry W, Berti E, Cerroni L, Chimenti S, 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 1997;90:354-71. 10. Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, Rosenwald A, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000;403:503-11. 11. Anagnostopoulos I, Stein H. Large B-cell lymphomas: variants and entities. Pathologe 2000;21:178-89. 12. Grange F, Petrella T, Beylot-Barry M, Joly P, D’Incan M, Delaunay M, et al. Bcl-2 protein expression is the strongest independent prognostic factor of survival in primary cutaneous large B-cell lymphomas. Blood 2004;103:3662-8. 13. Neville JA, Lang W, Fleischer AB Jr. Errors in the Archives of Dermatology and the Journal of the American Academy of Dermatology from January through December 2003. Arch Dermatol 2006;142:737-40.
596 Hallermann et al
J AM ACAD DERMATOL APRIL 2007
14. Grange F, Bekkenk MW, Wechsler J, Meijer CJ, Cerroni L, Bernengo M, et al. Prognostic factors in primary cutaneous large B-cell lymphomas: a European multicenter study. J Clin Oncol 2001;19:3602-10. 15. Smith BD, Smith GL, Cooper DL, Wilson LD. The cutaneous B-cell lymphoma prognostic index: a novel prognostic index derived from a population-based registry. J Clin Oncol 2005; 23:3390-5. 16. Zinzani PL, Quaglino P, Pimpinelli N, Berti E, Baliva G, Rupoli S, et al, for the Italian Study Group for Cutaneous Lymphomas. Prognostic factors in primary cutaneous B-cell lymphoma: the Italian study group for cutaneous lymphomas. J Clin Oncol 2006;24:1376-82. 17. Harada S, Suzuki R, Uehira K, Yatabe Y, Kagami Y, Ogura M, et al. Molecular and immunological dissection of diffuse large B cell lymphoma: CD51, and CD5 with CD101 groups may constitute clinically relevant subtypes. Leukemia 1999;13:1441-7. 18. Yamaguchi M, Seto M, Okamoto M, Ichinohasama R, Nakamura N, Yoshino T, et al. De novo CD51 diffuse large B-cell lymphoma: a clinicopathologic study of 109 patients. Blood 2002;99:815-21. 19. Tagawa H, Tsuzuki S, Suzuki R, Karnan S, Ota A, Kameoka Y, et al. Genome-wide array-based comparative genomic hybridization of diffuse large B-cell lymphoma: comparison between CD5-positive and CD5-negative cases. Cancer Res 2004;64: 5948-55. 20. Gellrich S, Rutz S, Golembowski S, Jacobs C, von Zimmermann M, Lorenz P, et al. Primary cutaneous follicle center cell lymphomas and large B cell lymphomas of the leg descend from germinal center cells: a single cell polymerase chain reaction analysis. J Invest Dermatol 2001;117:1512-20. 21. Sebestyen A, Berczi L, Mihalik R, Paku S, Matolcsy A, Kopper L. Syndecan-1 (CD138) expression in human non-Hodgkin lymphomas. Br J Haematol 1999;104:412-9. 22. Slater DN. The new World Health Organization-European Organization for Research and Treatment of Cancer classification for cutaneous lymphomas: a practical marriage of two giants. Br J Dermatol 2005;153:874-80. 23. Kodama K, Massone C, Chott A, Metze D, Kerl H, Cerroni L. Primary cutaneous large B-cell lymphomas: clinicopathologic features, classification, and prognostic factors in a large series of patients. Blood 2005;106:2491-7. 24. Paulli M, Viglio A, Vivenza D, Capello D, Rossi D, Riboni R, et al. Primary cutaneous large B-cell lymphoma of the leg: histogenetic analysis of a controversial clinicopathologic entity. Hum Pathol 2002;33:937-43. 25. Falini B, Fizzotti M, Pucciarini A, Bigerna B, Marafioti T, Gambacorta M, 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 2000; 95:2084-92. 26. Wright G, Tan B, Rosenwald A, Hurt EH, Wiestner A, Staudt LM. A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lymphoma. Proc Natl Acad Sci U S A 2003;100:9991-6. 27. Sundram U, Kim Y, Mraz-Gernhard S, Hoppe R, Natkunam Y, Kohler S. Expression of the bcl-6 and MUM1/IRF4 proteins correlate with overall and disease-specific survival in patients with primary cutaneous large B-cell lymphoma: a tissue microarray study. J Cutan Pathol 2005;32:227-34. 28. Hoefnagel JJ, Vermeer MH, Jansen PM, Fleuren GJ, Meijer CJ, Willemze R. Bcl-2, Bcl-6 and CD10 expression in cutaneous B-cell lymphoma: further support for a follicle center cell origin and differential diagnostic significance. Br J Dermatol 2003; 149:1183-91.
29. Hoefnagel JJ, Mulder MM, Dreef E, Jansen PM, Pals ST, Meijer CJ, et al. Expression of B-cell transcription factors in primary cutaneous B-cell lymphoma. Mod Pathol 2006;19:1270-6. 30. Lossos IS, Jones CD, Warnke R, Natkunam Y, Kaizer H, Zehnder JL, et al. Expression of a single gene, BCL-6, strongly predicts survival in patients with diffuse large B-cell lymphoma. Blood 2001;98:945-51. 31. Rosenwald A, Wright G, Chan WC, Connors JM, Campo E, Fisher RI, et al. Lymphoma/leukemia molecular profiling project: the use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma. N Engl J Med 2002;346:1937-47. 32. Berglund M, Thunberg U, Amini RM, Book M, Roos G, Erlanson M, et al. Evaluation of immunophenotype in diffuse large B-cell lymphoma and its impact on prognosis. Mod Pathol 2005; 18:1113-20. 33. Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie J, Ott G, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood 2004;103:275-82. 34. Lossos IS, Czerwinski DK, Alizadeh AA, Wechser MA, Tibshirani R, Botstein D, et al. Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes. N Engl J Med 2004;350:1828-37. 35. Staudt LM, Clerc RG, Singh H, LeBowitz JH, Sharp PA, Baltimore D. Cloning of a lymphoid-specific cDNA encoding a protein binding the regulatory octamer DNA motif. Science 1988; 241:577-80. 36. Kang SM, Tsang W, Doll S, Scherle P, Ko HS, Tran AC, et al. Induction of the POU domain transcription factor Oct-2 during T-cell activation by cognate antigen. Mol Cell Biol 1992; 12:3149-54. 37. Dunn TL, Ross IL, Hume DA. Transcription factor Oct-2 is expressed in primary murine macrophages. Blood 1996; 88:4072. 38. Hockenbery DM, Zutter M, Hickey W, Nahm M, Korsmeyer SJ. BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc Natl Acad Sci U S A 1991;88:6961-5. 39. Cho-Vega JH, Vega F, Rassidakis G, Medeiros LJ. Primary cutaneous marginal zone B-cell lymphoma. Am J Clin Pathol 2006;125(Suppl):S38-49. 40. Iqbal J, Neppalli VT, Wright G, Dave BJ, Horsman DE, Rosenwald A, et al. BCL2 expression is a prognostic marker for the activated B-cell-like type of diffuse large B-cell lymphoma. J Clin Oncol 2006;24:961-8. 41. Muris JJ, Meijer CJ, Vos W, van Krieken JH, Jiwa NM, Ossenkoppele GJ, et al. Immunohistochemical profiling based on Bcl-2, CD10 and MUM1 expression improves risk stratification in patients with primary nodal diffuse large B cell lymphoma. J Pathol 2006;208:714-23. 42. Willis TG, Dyer MJ. The role of immunoglobulin translocations in the pathogenesis of B-cell malignancies. Blood 2000;96:808-22. 43. Child FJ, Russell-Jones R, Woolford AJ, Calonje E, Photiou A, Orchard G, et al. Absence of the t(14;18) chromosomal translocation in primary cutaneous B-cell lymphoma. Br J Dermatol 2001;144:735-44. 44. Dijkman R, Tensen CP, Jordanova ES, Knijnenburg J, Hoefnagel JJ, Mulder AA, et al. Array-based comparative genomic hybridization analysis reveals recurrent chromosomal alterations and prognostic parameters in primary cutaneous large B-cell lymphoma. J Clin Oncol 2006;24:296-305. 45. Heckman CA, Duan H, Garcia PB, Boxer LM. Oct transcription factors mediate t(14;18) lymphoma cell survival by directly regulating bcl-2 expression. Oncogene 2006;25:888-98.
J AM ACAD DERMATOL
Hallermann et al 597
VOLUME 56, NUMBER 4
46. Banham AH, Beasley N, Campo E, Fernandez PL, Fidler C, Gatter K, et al. The FOXP1 winged helix transcription factor is a novel candidate tumor suppressor gene on chromosome 3p. Cancer Res 2001;61:8820-9. 47. Shaffer AL, Rosenwald A, Staudt LM. Lymphoid malignancies: the dark side of B-cell differentiation. Nat Rev Immunol 2002; 2:920-32. 48. Banham AH, Connors JM, Brown PJ, Cordell JL, Ott G, Sreenivasan G, et al. Expression of the FOXP1 transcription factor is strongly associated with inferior survival in patients
with diffuse large B-cell lymphoma. Clin Cancer Res 2005; 11:1065-72. 49. Akasaka T, Akasaka H, Ueda C, Yonetani N, Maesako Y, Shimizu A, et al. Molecular and clinical features of non-Burkitt’s, diffuse large-cell lymphoma of B-cell type associated with the c-MYC/ immunoglobulin heavy-chain fusion gene. J Clin Oncol 2000; 18:510-8. 50. Kawasaki C, Ohshim K, Suzumiya J, Kanda M, Tsuchiya T, Tamura K, et al. Rearrangements of bcl-1, bcl-2, bcl-6, and c-myc in diffuse large B-cell lymphomas. Leuk Lymphoma 2001;42:1099-106.