Jin Chen, Jiefang Dai, Ruiyang Wu, Haiyan Zhang
Histologically confirmed microabscess independently lowers recurrence risk in GLM. For patients with microabscess, a combined regimen of quinolones, abscess drainage and anti-mycobacterial therapy yields the best anti-relapse outcomes. Corticosteroid use markedly increases recurrence likelihood. Treatment decisions for GLM should be adjusted based on whether microabscess is present on pathology.
OBJECTIVE: We aimed to clarify whether microabscess observed on pathological biopsy independently predicts recurrence in granulomatous lobular mastitis (GLM), and to compare recurrence risks across different treatment regimens stratified by microabscess status. Our findings may help clinicians design individualized anti-recurrence therapy for GLM patients.
METHODS: This single-center retrospective cohort enrolled 520 women with pathologically confirmed GLM from a non-puerperal mastitis registry maintained at a provincial maternal and child health hospital. Patient records spanned January 2017 to December 2024, with all follow-up completed September 1, 2025. We collected demographic, laboratory, ultrasound, medication and surgical data for all participants. Cox proportional hazards regression models identified independent recurrence predictors. Kaplan-Meier curves with log-rank tests were used to compare 1-year recurrence rates between patients with and without histologic microabscess. Subgroup analyses further tested whether quinolone use or surgical resection modified the prognostic impact of microabscess.
RESULTS: The cohort's average age at GLM onset was 31.5 years; 268 patients (51.5%) had microabscess on histology, while 252 (48.5%) did not. Multivariate Cox regression identified affected breast laterality, histologic microabscess (HR = 0.605, 95% CI 0.413-0.886, P = 0.010), peripheral white blood cell count, anti-mycobacterial therapy, and corticosteroid use as independent recurrence predictors. Microabscess acted as a protective factor against relapse. The 1-year recurrence rate was lower in patients with microabscess (20.1%) than those without (28.6%, P = 0.021). Baseline comparisons revealed higher rates of quinolone administration and complete surgical excision among microabscess-positive individuals (P = 0.007 and P = 0.004, respectively). When restricted to patients receiving quinolones, the recurrence gap between microabscess-positive and negative cases widened significantly (19.7% vs 28.7%, P = 0.030); no meaningful intergroup difference emerged for patients who did not take quinolones (P = 0.432). Anti-mycobacterial medication reduced recurrence risk regardless of microabscess status, while corticosteroid use raised relapse risk. Abscess drainage only lowered 1-year recurrence for patients with microabscess (17.4% vs 27.0%, P = 0.036), with no measurable benefit for those without microabscess. Surgical resection, cephalosporins and bromocriptine showed no independent ability to reduce recurrence in either subgroup.
CONCLUSIONS: Histologically confirmed microabscess independently lowers recurrence risk in GLM. For patients with microabscess, a combined regimen of quinolones, abscess drainage and anti-mycobacterial therapy yields the best anti-relapse outcomes. Corticosteroid use markedly increases recurrence likelihood. Treatment decisions for GLM should be adjusted based on whether microabscess is present on pathology.