Table 1.
Cluster composition compared with therapeutic interventions by the time of biopsy
| Therapy | Patients, % |
P | ||
|---|---|---|---|---|
| Cluster 1 | Cluster 2 | Cluster 3 | ||
| No therapy | 33 | 47 | 23 | .400 |
| Topical corticosteroids | 62 | 53 | 73 | .719 |
| Topical tacrolimus | 14 | 16 | 0 | .550 |
| Topical nitrogen mustard | 10 | 11 | 0 | .672 |
| Topical bexarotene | 14 | 5 | 9 | .653 |
| PUVA | 29 | 26 | 23 | .930 |
| UVB (narrow or broadband) | 10 | 16 | 18 | .746 |
| Electron beam radiation (total skin or local) | 14 | 5 | 9 | .653 |
| Oral bexarotene | 29 | 11 | 23 | .449 |
| ONTAK (denileukin diftitox) | 24 | 5 | 14 | .304 |
| Extracorporeal photopheresis | 14 | 11 | 18 | .814 |
| Interferon | 19 | 5 | 23 | .351 |
| Oral corticosteroids | 29 | 5 | 23 | .223 |
| Chemotherapy (single or multiple agent) | 19 | 11 | 23 | .642 |
| Suberoylanilide hydroxamic acid (SAHA) | 0 | 0 | 5 | .556 |
Table shows percentage of patients in each cluster who had been treated with the given therapy by the time of biopsy. Clustering did not correlate with particular therapies at the time of biopsy. The chi-squared distribution was used to calculate the P value.