A staging system based on histopathologic evaluation of skin, lymph nodes, blood, and visceral sites provides more comprehensive prognostic information than clinical evaluation of skin disease and adenopathy. Patients may be divided at initial presentation into three prognostic groups: good-risk patients, who have plaque-only skin disease without lymph node, blood, or visceral involvement (median survival, greater than 12 years); intermediate-risk patients, who have cutaneous tumors, erythroderma, or plaque disease with node or blood involvement but no visceral disease or node effacement (median survival, 5 years); and poor-risk patients, who have visceral involvement or node effacement (median survival, 2.5 years).

The clinical course of cutaneous T cell lymphoma (mycosis fungoides and Sezary syndrome) is generally indolent, but in occasional patients becomes fulminant. We found that biopsies from patients with accelerating disease can reveal cytologic transformation from previously observed small, convoluted lymphocytes to large cells that are similar to cells seen in large-cell lymphoma. The cerebriform nuclei characteristic of malignant T cells can only rarely be identified. Of 150 cutaneous T cell lymphoma patients we treated from 1976 to 1984, cytologic transformation was identified in 12 after review of peripheral blood smears and biopsies from skin, lymph nodes, and visceral sites. Patients who developed cytologic transformation were initially characterized by advanced stage (11 of 12), with lymph node effacement (seven of 11) and erythroderma (five of 12). The tumor cell DNA content after transformation was aneuploid (four of four), and the ability to form rosettes with sheep erythrocytes was retained in transformed cells (three of three). The median time from diagnosis of cutaneous T cell lymphoma to cytologic transformation was 21.5 months (range, 4 to 64), and the median survival from transformation was only 2 months (range, 0 to 19+). We conclude that cytologic transformation in cutaneous T cell lymphoma represents a distinct clinicopathologic entity, characterized by an aggressive clinical course.

No abstract

Peripheral blood lymphocyte morphology was evaluated prospectively by light microscopy of blood smears and E rosette preparations in 160 patients with cutaneous T cell lymphoma (CTCL). Blood involvement was related to the type of cutaneous T-stage, being present in 90% of patients with erythroderma (T4), 27% of those with cutaneous tumors (T3), 9% of those with generalized (T2), and 0% of those with limited skin plaques (T1). Untreated patients with blood involvement (38 of 105) had a higher frequency of CTCL in lymph nodes and viscera and survival inferior to that of patients with normal or nondiagnostic lymphocyte morphology (P less than .001). Multivariate analysis showed skin stage and age to be the most important pretreatment risk factors for survival. Although blood involvement was not an independent risk factor for the entire group, it appeared to have some adverse influence in the T2/T3 subsets (P = .051). Both lymphocytosis and size distribution of the circulating CTCL cells at initial diagnosis influenced survival. Patients with “mixed cell” cytology (greater than 20% large [greater than 11 microns] CTCL cells), had a worse survival than those with predominantly small circulating CTCL cells (P = .009). The former were more likely to have aggressive features, including lymph node effacement by tumor (P less than .001) and visceral disease (P = .074), than were “small cell” patients. Our data indicate that detailed review of the blood lymphocyte morphology in patients with diagnosed or suspected CTCL is helpful in predicting extent of disease and prognosis.