Differential diagnosis of cutaneous infiltrates of B lymphocytes with follicular growth pattern

The differential diagnosis of cutaneous B-cell infiltrates with follicular pattern of growth is one of the most vexing problems in dermatopathology. In this study we focused on histopathologic, immunophenotypic, and molecular differential diagnostic criteria between Borrelia burgdorferi (Bb)-associated lymphocytoma cutis (LC), primary cutaneous follicle center cell lymphoma (FCCL), and primary cutaneous marginal zone lymphoma (MZL) with reactive germinal centers (GCs). A total of 47 patients were included in the study, including 12 cases of LC (M:F = 2:1; mean age: 38.0; median: 31; range: 9-75), 29 cases of FCCL (M:F = 1.2:1; mean age: 57.5; median: 57; range: 24-97), and 6 cases of MZL (M:F = 1:1; mean age: 63.8; median: 67.5; range: 38-86). In all cases complete phenotypic data were available. In addition, the IgH gene rearrangement and the t(14;18) were analyzed using the polymerase chain reaction technique (PCR) in 41 (FCCL = 27, LC = 10, MZL = 4) and 18 cases (FCCL = 15, LC = 2, MZL = 1), respectively. Histology revealed in all cases of FCCL one or more atypical feature of the follicles including the lack of or a reduced mantle zone, lack of polarization, tendency to confluence, and absence of tingible body macrophages. In most cases of Bb-associated LC, the GCs were devoid of mantle zone, lacked polarization, and revealed tendency to confluence as well, but all cases showed the presence of several tingible body macrophages. In MZL, follicles showed typical features of reactive GCs. Immunohistology revealed a reduced proliferative activity of neoplastic follicles as detected by MIB-1 antibody in 23 of 29 cases of FCCL (79.3%), but only in 1 case of LC (8.3%). Proliferation of the GCs was normal in all cases of MZL. Positivity for CD10 and/or Bcl-6 was found in small clusters outside the follicles in 19 cases of FCCL (65.5%), and in 3 cases of LC (25%), but in no case of MZL. The intensity of CD10 staining on follicular cells on average was stronger in cases of FCCL, but overlapping features could be observed. Finally, staining for Bcl-2 protein was consistently negative on GC cells in cases of LC and MZL, and was positive on a variable proportion of the cells in 8 cases of FCCL (28.6%). Molecular analyses showed no evidence of the t(14;18) in all cases tested. Analysis of the IgH gene rearrangement revealed a monoclonal pattern in 1 of 10 cases of LC (10%), 14 of 27 cases of FCCL (51.9%), and 2 of 4 cases of MZL (50%) tested. In summary, Bb-associated LC and FCCL show sometimes overlapping histopathologic, immunohistochemical, and molecular features, whereas follicles in MZL show clear-cut aspects of reactive GCs. Absence of tingible body macrophages within follicles, reduced proliferation of the follicles as detected by immunohistology, presence of positivity for Bcl-2 protein within follicular cells, and monoclonality by PCR are the main criteria suggestive of malignancy. Diagnosis of cutaneous infiltrates of B lymphocytes with follicular growth pattern should be achieved by integration of clinical data with histopathologic, immunohistochemical, and molecular features of the lesions.