Systemic lupus erythematosus (SLE) and cancer : epidemiological and immunohistochemical studies on SLE patients with malignant lymphoma or myeloid leukaemia

Abstract

Malignancy as a cause of death was reported in occasional SLE patients, but the question whether patients with SLE have an increased risk of developing cancer compared to the general population has remained unanswered. To address this question, we created a national Swedish SLE cohort from the Hospital Discharge Register where all patients with an SLE diagnosis between 1964 and 1994 were included. The number of observed cancer cases in this cohort was identified by register linkage with the Cancer Register 1964-1995 and was compared with the expected numbers in the general population. We found a 25% overall increased risk of cancer in SLE. Haematological malignancies constituted the major excess risk. A doubled increased risk of respiratory cancer and a tripled of squamous cell skin cancer - most pronounced after 15 years of follow-up - were also observed.

Non-Hodgkin´s lymphoma (NHL) represented the most outstanding (a tripled) cancer risk. To investigate the lymphoma subtype and to identify risk factors we performed a nested case control study comparing SLE patients who developed NHL during the observation period developed NHL with those SLE patients without malignancy. Lymphoma tissues were stained with new classification markers and reclassified. The NHL subtype diffuse large B cell lymphoma (DLBCL) dominated - 10 out of total 16 cases. Two of these were subtyped into germinal centre (GC) (better prognosis) and eight into non-GC. There were no indications of treatment-induced lymphomas, but lymphoma risk was elevated if haematological or sicca symptoms, or pulmonary involvement were present in the SLE disease.

For myeloid leukaemia, another haematological malignancy, the SLE patients had a doubled risk. In a nested case control study eight SLE patients in our cohort developed acute or chronic myeloid leukaemia. Leucopenia was a risk factor for leukaemia development whereas low-dose chemotherapy was not a major cause in our cohort -or in the reported cases we found in a Medline search - but a preceding myelodysplastic syndrome was frequently seen.

Finally, with the hypothesis that some factors related to rheumatic disease may contribute to the risk to develop lymphoma we investigated the presence of a costimulator for B-cell activation, A PRoliferating-Inducing Ligand (APRIL), in lymphoma tissue of patients with SLE, rheumatoid arthritis (RA), and patients without a chronic inflammatory disease and correlated to clinical variables. We found an overexpression of APRIL mainly in lymphomas of the DLBCL type. Moreover, APRIL was higher up regulated in the DLBCLs of the SLE patients, and in the RA subset with high cumulative RA disease activity suggesting a particular importance for the DLBCL development in these patient groups but possibly also reflecting the APRIL dysregulation per se seen in these diseases.

In conclusion; patents with SLE have an increased risk to develop malignancies, particularly haematological types. This could be related to disease specific risk factors such as chronic activation of the immune system.