Clinical and epidemiological project in cancer-associated dermatomyositis : the role of anti-TIF1γ antibodies

Abstract

Idiopathic inflammatory myopathies (IIM) are a heterogeneous group of rare chronic autoimmune inflammatory diseases causing a debilitating muscle weakness but encountering even other symptoms such as arthritis, skin rash, lung involvement or dysphagia. IIM are a diagnostic challenge because of the clinical variability and the need to early identify the subgroups of disease with worst prognosis. A late diagnosis delays treatment start, which is detrimental to disease outcome. One of these subgroups is dermatomyositis (DM) which is characterized by a particular skin rash, specific histopathological findings, and typical autoantibodies. DM can be associated with cancer but clinically it is very difficult to identify patients with DM who are at risk to develop cancer. Anti-TIF1γ autoantibody has been identified as a marker of cancer risk in patients with DM and can be helpful to identify which patients with DM should undergo cancer screening. However, only half of the anti-TIF1γ positive patients develop cancer. Moreover, it is unclear if these autoantibodies are present before the cancer or develop as a consequence of cancer and if they have a prognostic role in cancer treatment.

The aims of my thesis were 1) to update the epidemiology of IIM and cancer in Sweden, 2) to elucidate the temporal relationship between anti-TIF1γ autoantibody, DM diagnosis and cancer; 3) to study the prognostic value of anti-TIF1γ autoantibody levels; 4) to compare the reliability of different laboratory assays to detect anti-TIF1γ autoantibodies.

In paper I we performed a registry based study of the risk of cancer in IIM in Sweden between 2002 and 2016 and observed that especially patients with DM presented a high cancer risk within one year from diagnosis compared to other forms of IIM and that cancer subtypes differed in the period before or after IIM diagnosis where ovarian and lung cancer were more frequent before IIM diagnosis while oral, cervical and skin cancers were more frequent after diagnosis. In paper II we studied a cohort of cancer associated IIM in Sweden and Spain and showed that anti-TIF1γ autoantibodies are specific for DM and associated with cancer within 3 years from DM diagnosis. Anti-TIF1γ autoantibodies may be detected before clinical symptoms of cancer and may disappear after successful treatment of cancer. Higher autoantibody titers were associated with a high mortality rate within one year of DM diagnosis. In paper III we investigated longitudinally collected sera from patients with DM with and without cancer and found that the levels of anti-TIF1γ autoantibodies were slightly higher in cancer associated DM and decreased over time but almost never became negative regardless of cancer status. Furthermore, anti-TIF1γ autoantibodies were in most cases present in the first available serum sample and rarely appeared after DM diagnosis. This means that re-testing for anti-TIF1γ is not useful during follow-up. In paper IV we analyzed the same cohort as in paper 3 for anti-TIF1γ antibodies with four different autoantibody assays (immunoprecipitation, in-house and commercial ELISA and lineblot) and confirmed that ELISA is a sensitive and reliable assay to find anti-TIF1γ antibody positive cancer associated DM.

In conclusion, my thesis has contributed by describing a contemporary picture of cancer associated IIM and DM in Sweden, adding new knowledge about the risk for different cancer subtypes before and after IIM diagnosis. It has helped understanding the temporal behavior of anti-TIF1γ autoantibodies and the most reliable laboratory assays to measure them. Thanks to this thesis we today have more information on how to interpret the cancer risk in patients with DM and anti-TIF1γ antibodies.