Site‑specific timing and mode of detection of locoregional recurrence after breast cancer surgery

Locoregional recurrence after breast cancer surgery remains potentially curable; however, current guidelines recommend annual mammography (MMG) alone, and recommend against routine ultrasonography (US) or computed tomography (CT).Despite arising in different anatomic locations, intramammary and regional nodal recurrences are typically analyzed together.The present study aimed to examine when and how locoregional recurrence after surgical treatment was detected under a defined institutional surveillance protocol.Among 61 locoregional recurrences verified at Fukushima Medical University Hospital (Fukushima, Japan) between February 2016 and August 2025, 13 were excluded, leaving 48 patients, each of whom contributed one recurrence.The disease-free interval (DFI), mode of detection and examination that first identified the recurrence were extracted from medical records, and intramammary and regional nodal recurrences were compared.Among the 48 recurrences, 32 were intramammary, 13 involved regional lymph nodes and 3 involved the skin/soft tissue.The median DFI was 156 months for intramammary recurrences and 42 months for regional nodal recurrences (P=0.0009), with 59.4% compared with 15.4% of recurrences being diagnosed >10 years after primary surgery.Overall, US was the most frequent detection modality (47.9%), followed by MMG (18.8%),CT (12.5%), positron emission tomography (PET)-CT (10.4%) and palpation (10.4%).Among intramammary recurrences, 75.0% were detected by US or MMG whereas 9.4% were detected by CT or PET-CT.By contrast, among regional nodal recurrences, 46.2% were detected by CT or PET-CT (P=0.0106).Tumor size at detection did not differ between the recurrence sites.In conclusion, in the present single-center series, intramammary and regional nodal recurrences demonstrated distinct patterns of timing and detection modality.Because the number of examinations performed was unavailable, the findings characterize detection patterns rather than the diagnostic performance of individual examinations.Further validation in larger, multicenter studies is warranted.