Context: Demonstration of a dominant T-cell clone in skin biopsy specimens by a molecular assay constitutes an additional diagnostic criterion to differentiate cutaneous T-cell lymphomas (CTCLs) from inflammatory dermatoses.
Objective: To determine which patients, depending on their clinical presentations, could most benefit from a cutaneous T-cell clonality analysis in addition to histopathologic analysis for the diagnosis of CTCL.
Design: Comparison of sensitivity and specificity of histopathologic analysis and a combination of this method and the detection of a T-cell receptor gamma chain gene rearrangement by polymerase chain reaction denaturing gradient gel electrophoresis performed on skin biopsy specimens obtained at initial presentation.
Patients: One hundred forty consecutive patients were classified into 4 groups, depending on their clinical presentation: (1) eczematous patches suggestive of early-stage mycosis fungoides (MF) (IA and IB of the TNM classification) (n = 42); (2) plaques, nodules, or tumors that arise on or are associated with plaques suggestive of late-stage MF (IIB and III of the TNM classification) (n = 16); (3) erythroderma (n = 50); and (4) nodules or tumors that arise in normal skin, suggestive of non-MF CTCL (n = 32).
Results: When compared with histopathologic examination, the addition of clonality analysis increased the sensitivity of CTCL diagnosis in all groups of patients except those with cutaneous lesions suggestive of late-stage MF, because the diagnosis was made based on histopathologic analysis alone in 100% of these cases. The main increase in sensitivity of CTCL diagnosis was observed in patients with erythroderma: 62% with histopathologic analysis alone to 87% with the combination of both methods (P = .04). Diagnostic specificity of molecular assays decreased from 100% to 76% (P = .01) in patients with patch lesions and from 100% to 70% (P = .04) in patients with nodules that occurred in normal skin due to the detection of a T-cell clone in 6 patients with follicular mucinosis without a histologic pattern of MF and in 5 of 20 cases of T-cell pseudolymphoma (25%), respectively. In contrast, a T-cell clone was not detected in the 34 patients with erythroderma of inflammatory origin.
Conclusion: Polymerase chain reaction analysis of cutaneous T-cell clonality could be useful for the diagnosis of CTCL in patients who present with erythroderma.