¹Department of Dermatology and Center for Clinical Epidemiology and Biostatistics University of Pennsylvania, Philadelphia, Pennsylvania, USA

Correspondence: Dr Joel M. Gelfand, Department of Dermatology and Center for Clinical Epidemiology and Biostatistics University of Pennsylvania, 3600 Spruce Street, 2 Maloney Building, Philadelphia, Pennsylvania 19104, USA. E-mail: joel.gelfand@uphs.upenn.edu

Received 5 February 2006; Revised 20 April 2006; Accepted 20 April 2006; Published online 1 June 2006.

Abstract

Psoriasis is a common, chronic, inflammatory disease. Psoriasis has been hypothesized to be associated with an increased risk of lymphoma due to its pathophysiology, its treatments, or a combination of these factors. We performed a large population-based cohort study of the risk of lymphoma in psoriasis patients using the General Practice Research Database. We identified 153,197 patients with psoriasis and 765,950 corresponding subjects without psoriasis. Psoriasis patients who received a systemic treatment consistent with extensive disease were classified as severe (N=3,994) and those who did not receive systemic therapies were classified as mild (N=149,203). The analyses were adjusted for age, gender, and person–time using a Cox proportional hazards model. For mild and severe psoriasis patients, the respective adjusted relative risks for lymphoma and its subtypes were as follows: all lymphoma 1.34 (1.16, 1.54) and 1.59 (0.88, 2.89); non-Hodgkin's lymphoma 1.15 (0.97, 1.37) and 0.73 (0.28, 1.96); Hodgkin's lymphoma (HL) 1.42 (1.00, 2.02) and 3.18 (1.01, 9.97); cutaneous T-cell lymphoma (TCL) 4.10 (2.70, 6.23) and 10.75 (3.89, 29.76). Psoriasis is associated with an increased risk of lymphoma. The association is strongest for HL and CTCL. The excess risk of lymphoma attributed to psoriasis was 7.9/100,000 psoriasis patients per year. Although patients with psoriasis have an increased relative risk of lymphoma, the absolute risk attributable to psoriasis is low given that lymphoma is a rare disease and the magnitude of association is modest.