The myocardial effects of proton vs. photon radiotherapy in left-sided breast cancer patients: a cardiac MRI study

Background Conventional photon therapy for breast cancer (BC) is associated with cardiac injury, cardiac fibrosis, and an increased risk for heart failure. Proton RT is capable of more precise dose deposition than photon RT, resulting in lower mean heart (MHD) and cardiac substructure doses, including the left ventricle. In theory, this physical dose advantage of proton RT may limit the extent of myocardial injury and, consequently, the development of fibrosis. Cardiac MRI can quantify myocardial fibrosis by measuring T1 times and, more sensitively, extracellular volume (ECV). A direct comparison of the myocardial effects of conventional photon and proton RT using cardiac MRI has not yet been performed. Purpose To compare the myocardial effects of photon versus proton RT using cardiac MRI-based functional and tissue analysis over a 12-month period in patients with stage II-III, left-sided BC receiving comprehensive radiotherapy. Methods Gadolinium contrast-enhanced cardiac MRI was performed at baseline, before RT, and 12 months after enrollment among persons treated with photon and proton RT. Standard cines were acquired to evaluate left (LV) and right ventricular (RV) volumetry and function. Global LV strain values were calculated using a feature tracking method. T1 maps were acquired in a mid-ventricular slice of the LV to calculate native T1 times and ECV. T2 maps were captured in the same mid-LV slice. Results The cohort included 13 participants, with a median age of 54 years [IQR 45-62] in the proton group (n=7) and 47 years [44-48] in the photon group (n=6). An anthracycline was administered to 3/7 in the proton group and to 5/6 in the photon group, and trastuzumab to 2/7 in the proton group and to 2/6 in the photon group, preceding RT. The median MHD was 0.9 [0.6-1.2] Gy among proton RT recipients, and 3.5 [2.8-4.2] Gy in the photon group (p=0.001). LV and RV volumetric, strain, and T2 values were similar between the two groups at baseline and 12-month follow-up. Native T1 times at baseline (1,125 [1,095-1,154] msec for protons vs. 1,120 [1,109-1,125] msec for photons, p=0.53) and at 12 months (1,084 [1,063-1,149] msec vs. 1,129 [1,101-1,169] msec, p=0.53) were similar between the groups. Baseline ECV was similar between the groups (median 24% [23-26] for protons vs. 26% [25-27] for photons, p=0.15). At 12 months, the ECV was significantly higher in the photon group (median 25% [22-27] vs. 30% [29-31]; p=0.005). The interval change in ECV in the proton group was -0.9% [-1.2, 0.7] vs. 4.4% [3.7-5.8] for photons (p=0.005). Conclusions Myocardial fibrosis, measured as MRI-derived ECV, was preserved in a small cohort receiving proton RT for left-sided BC, while ECV significantly increased in photon RT recipients over 12 months. These findings warrant further investigation with long-term follow-up to elucidate the potential cardiac benefits of proton RT.