Background and purposeTo assess implant stability and identify causes of implant variations during high-dose-rate multi-catheter breast brachytherapy. Materials and methodsPlanning-CTs were compared to control-CTs acquired halfway through the treatment for 100 patients. For assessing geometric stability, Fréchet-distance and button-to-button distance changes of all catheters as well as variations of Euclidean distances and convex hulls of all dwell positions were determined. The CTs were inspected to identify the causes of geometric changes. Dosimetric effects were evaluated by target volume transfers and re-contouring of organs at risk. The dose non-uniformity ratio (DNR), 100% and 150% isodose volumes (V100 and V150), coverage index (CI), and organ doses were calculated. Correlations between the examined geometric and dosimetric parameters were assessed. ResultsFréchet-distance and dwell position deviations >2.5 mm as well as button-to-button distance changes >5 mm were detected for 5%, 2%, and 6.3% of catheters, but for 32, 17, and 37 patients, respectively. Variations occurred enhanced in the lateral breast and close to the ribs, e.g. due to different arm positions. Only small dosimetric effects with median DNR, V100, and CI variations of −0.01 ± 0.02, (−0.5 ± 1.3)ccm, and (–1.4 ± 1.8)% were observed in general. Skin dose exceeded recommended levels for 12 of 100 patients. Various correlations between geometric and dosimetric implant stability were found, based on which decision-tree regarding treatment re-planning was established. ConclusionMulti-catheter breast brachytherapy shows a high implant stability in general, but considering skin dose changes is important. To increase implant stability for individual patients, we plan to investigate patient immobilization aids during treatments.