Abstract Background: Adjuvant locoregional radiotherapy (LRRT) reduces risk of locoregional failures (LRF) and improves survival for node positive breast cancer (BC) patients. However, LRRT increases the risk for toxicity as edema in the arm, lung cancer and cardiac mortality. Modern radiotherapy allows a more conform therapy which makes knowledge of LRF-patterns very important, in order not to underdose volumes with high risk for microscopic disease and simultaneously restrict dose to risk organs. In addition, BC-subtype may affect radiosensitivity and could possibly be used to individualize LRRT in the future. Methods: We investigated outcome for BC-patients receiving LRRT in the Southwest Sweden (2004-2008) in order to identify LRFs. During this period patients with >3 positive lymph nodes were given LRRT (50Gy in 2Gy fractions) to the breast/thoracic wall, axilla level II+III, supra- and infraclavicular lymph nodes according to a target definition atlas introduced in 2002. Patients with LRF as first event were identified, with distant failures and death considered as competing risks. The anatomical distribution of LRF was compared with the contouring atlas, radiotherapy given, and biological subtype based on immunohistochemistry. Results: 904 patients received LRRT. Median follow-up time was 9.8 years (0.2-14.6) for patients without an event. 59 patients (6,5%) developed a LRF, 30 of which were local failures (LF) and 31 regional failures (RF) (2 simultaneous LF/RF). Median time to LRF was 2.8 years. 37 of the 59 (63%) LRF-patients developed generalized disease within 3 months from the LRF. Of the 845 patients without LRF 316 developed distant metastases as first recurrence, 1 an isolated RF in the contralateral axilla, 64 died from other causes, and for 5 patients recurrence-status was unclear. 459 were alive at end of follow-up. Of the LF 19 developed after mastectomy (MRM) and 11 after breast conserving surgery (BCS). LF-location after MRM: 9 in-field, 6 at field margin, 2 both in/out of field, 1 out of field, and for 1 not yet determined. After BCS: 8 in-field, 1 at field margin, and for 2 not yet determined. Of the RF 28 developed after MRM and 3 after BCS. Location of RF after MRM: 11 in-field, 6 at field margin, 1 both in/out of field, 2 out of field, and for 8 not yet determined. After BCS: 1 in-field, 1 at field margin, and 1 not yet determined. The most common location for RF close to /out of field was superior to the treated area in fossa supraclavicularis. Biological approximate subtype was available for 885 of the primary tumours; luminal (ER+ and/or PR+ HER2-/?) 67% (589/885), HER2+ 19% (169/885), triple negative (ER- and PR- HER2-/?) (TN) 14% (127/885). Subtype distribution of BC later causing LRF despite LRRT was: luminal 44% (26/59), HER2+ 27% (16/59), TN 29% (17/59). Among primary tumours causing a LRF within irradiated volume 77% (24/31) were HER2+ or TN. Conclusion: In this high risk cohort of BC-patients, we found a low incidence of LRFs. The majority of LRFs developed within irradiated volume. BCs of the HER2+ and TN-subtype may be more radioresistant and have a higher risk of LRF. Updated information as well as figures mapping all recurrences in relation to previous LRRT will be presented at the symposium. Citation Format: Adra J, Karlsson P, Killander F, Lundstedt D, Alkner S. Distribution of locoregional breast cancer recurrences in relation to previous radiotherapy and biological subtype [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr PD8-11.