Transcriptional profiles predict disease outcome in patients with cutaneous T-cell lymphoma.

PURPOSE Average survival of cutaneous T-cell lymphoma (CTCL) is associated with clinical stage at diagnosis, where stage I has a favorable survival prognosis, whereas patients with more advanced stages succumb to their disease within 5 years. Although the majority of patients present with an early-stage CTCL, 15% to 20% of them will inevitably progress. Current state-of-the-art clinical criteria cannot identify individuals with stage I disease who are at risk of progression. The purpose of the current work is to gain novel molecular insight into the pathophysiology of CTCL to be able to identify patients with poor versus favorable prognosis. Our previous work used microarray analysis of skin biopsies from 62 CTCL patients to perform an unsupervised analysis of gene expression, which revealed three distinct transcription profile clusters. EXPERIMENTAL DESIGN In the present study, we used reverse transcription-PCR to confirm gene expression levels for a subset of representative genes in each cluster. We also performed a Kaplan-Meier analysis of survival and disease progression based on the 6 years of clinical follow-up. RESULTS Our reverse transcription-PCR results confirmed the upregulation of representative genes for each cluster, whereas clinical analysis documents that all stage I cases that progressed to stage II and beyond were in poor and intermediate prognosis clusters 1 and 3 and none were in favorable prognosis cluster 2. This analysis also identified certain genes that were preferentially expressed in favorable (e.g., WIF-1) versus poor (e.g., IL-17F) prognosis clusters. CONCLUSION This work suggests that it may be possible to stratify CTCL patients into low-risk, intermediate-risk, and high-risk groups based on gene expression.