Objective: To provide evidence that lymphomatoid keratosis should be categorized as an epidermotropic subtype of cutaneous lymphoid hyperplasia.
Design: Clinicopathological, immunohistochemical, and molecular biological studies of epidermotropic and dermal bandlike infiltrates of lymphocytes without necrotic keratinocytes, Civatte bodies, or Max-Joseph spaces and solar lentigo or seborrheic keratosis adjacent to the lesion, but with epidermal hyperplastic change (clinically scaly plaque) in cases of lymphomatoid keratosis. Conventional histopathologic study as well as immunohistochemical examinations for CD1a, CD3, CD4, CD8, CD20, and CD79a and S100 protein and genotypic examinations were performed.
Setting: University departments comprising 2 sections of dermatology and 1 section of pathology.
Main outcome measures: Ratio of T to B cells and of CD4(+) to CD8(+) cells, and the phenotype of epidermotropic cells were evaluated. Gene rearrangement of the immunoglobulin heavy chain gene and T-cell receptor (TCR)-beta and TCRgamma genes was also investigated by the polymerase chain reaction method.
Results: Immunohistochemically, epidermotropic CD20(+) and/or CD79a(+) cells were present. In the upper dermal lymphocytic infiltrates, the CD3(+)/CD79a(+) cell ratio ranged from 5:5 to 8:2. The CD4(+)/CD8(+) cell ratio was within normal limits. Rearrangements of the TCRgamma gene were demonstrated in 2 cases and of the TCRbeta gene in 1 case.
Conclusions: Our results indicate that lymphomatoid keratosis is a clinically benign keratotic lesion but histologically malignant, simulating mycosis fungoides. Immunohistochemical findings showed a reaction pattern in all cases, but genotypical examination showed some clonal dermatoses. Therefore, "lymphomatoid" keratosis should be classed as a pseudolymphoma, namely, a subtype of cutaneous lymphoid hyperplasia with epidermotropism.