Diagnostic difficulties in 'non-mycotic' cutaneous lymphoproliferative diseases

The diagnosis of cutaneous lymphoproliferative diseases is an area of bewildering complexity to many histopathologists. This article concentrates on 'non-mycotic' cutaneous diseases. The 'current state of the diagnostic art' is critically assessed. Cutaneous 'pseudolymphoma' is relegated to the position of an aid-memoire and is not a diagnosis. Inadequacies in the classification of cutaneous lymphoma are discussed and the non-specificity of many histopathological features is highlighted. The status of specific entities is analysed and the contribution of modern investigative techniques in diagnosis is evaluated. This includes cutaneous T-cell lymphomas with a detailed consideration of large cell lymphoma heterogeneity. Cutaneous B-cell diseases are shown to be an unresolved diagnostic maze and the necessity for new clearly defined diagnostic criteria is emphasized. Evidence is presented to show that many cutaneous lymphoproliferative diseases lie on continuous spectra that, initially, are probably antigenically driven, and that a diagnosis is best achieved by a multifaceted approach. This is exemplified by cutaneous diseases that have origins from both B- (cutaneous lymphoid hyperplasia/lymphoma) and T-cells (lymphomatoid papulosis, lymphomatoid granulomatosis and mycosis fungoides). The future diagnostic role of the polymerase chain reaction and cytogenetic analysis is discussed. Intriguingly, recent molecular evidence has shown that lymphomatoid papulosis, cutaneous T-cell lymphoma, CD30 positive large cell lymphoma and Hodgkin's disease can originate from a single T-cell clone and display an identical chromosomal translocation and T-cell receptor rearrangement. Careful clinico-pathological correlation combined with prolonged patient follow-up remains the gold standard for diagnosis.