SAN FRANCISCO Dermatologists and oncologists have long fretted over when to pull out the therapeutic big guns for patients with early-stage cutaneous T-cell lymphoma.
Gene expression profiling of lesional skin may provide the answer.
"We want to know who's going to progress in a high-risk way and who's not. And we're looking to see which patients will respond to which therapies. We think we can begin to do so by looking at some of these genes," Dr. Thomas S. Kupper said at the annual meeting of the American Academy of Dermatology.
The great majority of patients with stage IA and many with stage IB cutaneous T-cell lymphoma (CTCL) have indolent, skin-limited disease, but others have rapidly progressive and often fatal disease. Lesional skin gene expression profiling may eventually make it possible to employ, in proactive fashion, the more aggressive and toxic systemic therapies in selected high-risk patients with early-stage disease, according to Dr. Kupper, Thomas B. Fitzpatrick Professor of Dermatology at Harvard Medical School and chair of the department of dermatology at Brigham and Women's Hospital, Boston.
He and a multidisciplinary team profiled gene expression on 63 lesional skin biopsies from 62 patients with all stages of CTCL. Thirteen patients had stage IA diseasethat is, limited patches and plaques over less than 10% of their body surface areawhile 29 had stage IB CTCL, 8 had stage IIB, and 12 had stage III.
The investigators identified 593 markedly upregulated genes grouped within three distinct patterns or clusters of gene expression, with roughly one-third of patients falling within each cluster.
The dominant genes upregulated in cluster 1 included many genes involved in T-cell activation and the immune response, including the T-cell alpha and beta chains, interleukin-2 receptor, lymphocyte-specific protein kinase, CD8+ T-cell markers, tumor necrosis factor pathway genes, downstream members of the WNT signaling pathway, and several B-cell-related genes.
Cluster 2 was rich in upregulated genes involved in keratinocyte and epidermal differentiation and proliferation. All but two patients in cluster 2 had stage IA or IB disease. Cluster 2 patients were generally extremely treatment-responsive; indeed, all but three cluster 2 patients were maintained in remission or under good control with only topical therapies. The sole stage IIB patient in cluster 2 went into remission on PUVA alone.
Cluster 3 featured upregulation of several T-cell-specific genes as well as genes related to keratinocyte function. Like cluster 1, cluster 3 was distinguished by more aggressive, less treatment-responsive disease. Seven of 12 stage III patients fell into cluster 3, as did only 1 patient with IA disease.
During roughly 3 years of follow-up, there were seven deaths, three cases of disease progression despite systemic therapies, and one large-cell transformation.
None occurred in cluster 2 patients. In contrast, the event-free survival rate was 80% in cluster 3 and less than 60% in cluster 1. One cluster 1 patient with stage IA disease at the time of gene profiling progressed to large-cell transformation unresponsive to total skin electron beam therapy, denileukin diftitox (Ontak), and suberoylanilide hydroxamic acid.
Nine of 11 stage IB patients in cluster 2 were maintained on intermittent topical therapies. In contrast, the more difficult to treat stage IB patients tended to fall into clusters 1 and 3; indeed, 7 of 18 patients with IB CTCL in clusters 1 and 3 required systemic therapies after a variety of topical therapies failed. One stage IB patient in cluster 1 developed metastatic disease and died despite oral bexarotene, denileukin diftitox, total skin electron beam irradiation, interferon, photopheresis, and multiagent chemotherapy.
Prospective studies with larger numbers of CTCL patients will be required to characterize the gene profile clusters more fully and zero in on individual genes having particularly potent predictive power, Dr. Kupper said.
The gene profiling study was sponsored by the National Institutes of Health.