Breast tumour-secreted ADAM10 mediates atrial fibrogenesis and fibrillation.

BACKGROUND AND AIMS: Breast cancer (BC)-related atrial fibrillation (AF) contributes to an overall poor prognosis and raises great concerns. However, beyond cancer treatment-related AF, it is unknown whether BC itself is sufficient to directly induce AF. This study aimed to explore the role of BC in the development of AF, focusing on the tumour itself rather than therapy-related cardiotoxicity. METHODS: A retrospective analysis was conducted on 1224 female BC patients before therapy and 18 159 healthy female participants. The prevalence of AF was assessed, and markers of atrial remodelling, including P-wave indices on electrocardiogram (ECG), were measured. An orthotopic BC model was established by implanting E0771 cells to mammary fat pads of female C57BL/6 mice. An electrophysiological study was employed to assess atrial arrhythmia vulnerability. Masson's trichrome staining, coculture assays and molecular biological techniques were used to determine the mechanisms of BC-induced AF. RESULTS: BC patients (n = 1217) were individually matched 1:1 to non-cancer female controls (n = 1217) by propensity score matching. Even before treatment, BC patients had a significantly higher prevalence of AF (6.41%) than controls (0.90%). BC patients exhibited more marked atrial remodelling evaluated by abnormal P-wave terminal force in ECG lead V1 (7.37% vs 3.90%, P < .001), compared to non-cancer female controls. Similarly, BC mice, as well as mice administered with tumour-conditioned media or plasma from BC mice, developed atrial fibrosis and had increased AF susceptibility. After screening tumour-secreted factors, both BC patients with AF and BC mice displayed elevated levels of soluble ADAM10 (sADAM10) in tumour tissue and plasma. Moreover, sADAM10 promoted the cleavage of Ephrin B2 into soluble Ephrin B2 (sEphrin B2) in atrial cardiofibroblasts (ACFs), which then bound with EphB3 and EphB4 receptors and activated the atrial fibrosis pathway. Cardiac fibroblast-specific knockdown of Ephrin B2 suppressed AF development in BC mice. The plasma levels of sEphrin B2 were especially higher in pre-treatment breast cancer patients with AF compared to controls, in parallel with increasing sADAM10 levels (R2 = 0.67, P < .0001). The inhibition of tumour-originated sADAM10 with its specific inhibitor (GI254023X) or its shRNA attenuated atrial fibrosis and AF susceptibility in BC mice. CONCLUSIONS: BC patients before treatment present a significantly higher prevalence of AF and atrial remodelling, which correlate with elevated levels of tumour-originated sADAM10 and ACF-secreted sEphrin B2. Tumour-originated sADAM10 directly promotes AF pathogenesis by inducing atrial fibrosis, independent of cancer treatment. These findings imply that the inhibition of sADAM10 represents a potentially new therapeutic option for BC-associated AF.