Cancer development in patients hospitalized with systemic lupus erythematosus: A population‐level data linkage study

Abstract

Aim To explore the association between systemic lupus erythematosus (SLE) with the risk of cancer development and subsequent 5‐year mortality in Western Australia (WA). Methods Population‐level, data linkage study of SLE patients (n = 2111) and general population comparators (n = 21 110) hospitalized between 1980 and 2014. SLE patients (identified by ICD‐9‐CM: 695.4, 710.0, and ICD‐10‐AM: L93.0, M32.0) were nearest matched (10:1) for age, sex, Aboriginality, and temporality. Follow up was from time zero (index SLE hospitalization) to cancer development, death or 31 December 2014. We assessed the risk of cancer development and subsequent 5‐year mortality between SLE patients and comparators with univariate and multivariate‐adjusted Cox proportional hazards regression models. Results SLE patients had similar multivariate‐adjusted risk (adjusted hazard ratio [aHR] 1.03, 95% confidence interval [CI] 0.93–1.15; p = .583) of cancer development. Cancer development risk was higher in SLE patients <40 years old (aHR 1.58, 95% CI 1.29–1.94; p < .001), and from 1980 to 1999 (aHR 1.16, 95% CI 1.02–1.31; p < .001). SLE patients had higher risk of developing cancer of the oropharynx (aHR 2.13, 95% CI 1.30–3.50), vulvo‐vagina (aHR 3.22, 95% CI 1.34–7.75), skin (aHR 1.20, 95% CI 1.01–1.43), musculoskeletal tissues (aHR 2.26, 95% CI 1.16–4.40), and hematological tissues (aHR 1.78 95% CI 1.25–2.53), all p < .05. After cancer development, SLE patients had increased risk of 5‐year mortality (aHR 1.31, 95% CI 1.06–1.61); highest in patients <50 years old (aHR 2.03, 95% CI 1.03–4.00), and in those with reproductive system and skin cancers. Conclusions Hospitalized SLE patients had increased risk of multiple cancer sub‐types. Following cancer development, SLE patients had increased risk of 5‐year mortality. There is scope to improve cancer prevention and surveillance in SLE patients. Trial registration Not applicable. This low‐risk risk study used de‐identified administrative linked health data.