Immune responses to CCAR1 and other dermatomyositis autoantigens are associated with attenuated cancer emergence
David F. Fiorentino, Christopher A. Mecoli, Matthew C. Rosen, Lorinda S. Chung, Lisa Christopher-Stine, Antony Rosen, Livia Casciola-Rosen
David F. Fiorentino, Christopher A. Mecoli, Matthew C. Rosen, Lorinda S. Chung, Lisa Christopher-Stine, Antony Rosen, Livia Casciola-Rosen
View: Text | PDF
Clinical Research and Public Health Autoimmunity Immunology

Immune responses to CCAR1 and other dermatomyositis autoantigens are associated with attenuated cancer emergence

Abstract

BACKGROUND The temporal clustering of a cancer diagnosis with dermatomyositis (DM) onset is strikingly associated with autoantibodies against transcriptional intermediary factor 1-γ (TIF1-γ). Nevertheless, many patients with anti–TIF1-γ antibodies never develop cancer. We investigated whether additional autoantibodies are found in anti–TIF1-γ–positive patients without cancer.METHODS Using a proteomic approach, we defined 10 previously undescribed autoantibody specificities in 5 index anti–TIF1-γ–positive DM patients without cancer. These were subsequently examined in discovery (n = 110) and validation (n = 142) cohorts of DM patients with anti–TIF1-γ autoantibodies.RESULTS We identified 10 potentially novel autoantibodies in anti–TIF1-γ–positive DM patients, 6 with frequencies ranging from 3% to 32% in 2 independent DM cohorts. Autoantibodies recognizing cell division cycle and apoptosis regulator protein 1 (CCAR1) were the most frequent, and were significantly negatively associated with contemporaneous cancer (discovery cohort OR 0.27 [95% CI 0.7–1.00], P = 0.050; validation cohort OR 0.13 [95% CI 0.03–0.59], P = 0.008). When cancer did emerge, it occurred significantly later in anti-CCAR1–positive compared with anti-CCAR1–negative patients (median time from DM onset 4.3 vs. 0.85 years, respectively; P = 0.006). Cancers that emerged were more likely to be localized (89% of anti-CCAR1–positive cancers presenting at stage 0 or 1 compared with 42% of patients without anti-CCAR1 antibodies, P = 0.02). As the number of additional autoantibody specificities increased in anti–TIF1-γ–positive DM patients, the frequency of cancer decreased (P < 0.001).CONCLUSION As the diversity of immune responses in anti–TIF1-γ DM patients increases, the likelihood of cancer emerging decreases. Our findings have important relevance for cancer risk stratification in DM patients and for understanding natural immune regulation of cancer in humans.TRIAL REGISTRATION Not applicable.FUNDING SOURCES The NIH, the Donald B. and Dorothy L. Stabler Foundation, and the Huayi and Siuling Zhang Discovery Fund.

Authors

David F. Fiorentino, Christopher A. Mecoli, Matthew C. Rosen, Lorinda S. Chung, Lisa Christopher-Stine, Antony Rosen, Livia Casciola-Rosen

×

Figure 5

Model of relationship between cancer fitness and immune response.

Options: View larger image (or click on image) Download as PowerPoint
Model of relationship between cancer fitness and immune response. Model ...
Model is depicted as a spectrum of decreasing cancer fitness (left to right) and its inverse relationship with the antitumor immune response. All scenarios represent DM in association with anti–TIF1-γ antibodies and incipient cancer. Scenario A is a state of high cancer fitness with a paucity of additional immune responses beyond anti–TIF1-γ. This part of the spectrum is associated with rapid (around time of DM onset) and aggressive (e.g., advanced stage) cancer emergence (“immune escape”). Scenario B represents a balance between cancer and immune response (equilibrium), and is characterized by a broader immune response (e.g., anti-CCAR1). In this scenario, cancer eventually manifests (a transition from equilibrium to immune escape), but is less aggressive (e.g., earlier stage) and emerges after a time delay following DM onset. Scenario C is also characterized by a broad (e.g., anti-CCAR1) and effective immune response, but is one in which the antitumor response ultimately deletes (elimination) or maintains the cancer in a subclinical state (equilibrium). Note that other mechanisms exist that might explain the relationship of additional autoantibodies with attenuated cancer emergence (e.g., additional autoantibodies attenuate a procancer property of anti–TIF1-γ).