H. Aytac1, M. Nazli2, S. Ozbas3, B. Yigit4, M. Tokocin4, E. Mendelson5, W. Berg6, A. Soran7 1Baskent University, Department Of General Surgery, Adana, Turkey 2Basaksehir Cham and Sakura City Hospital, Department Of Radiology, Istabul, Turkey 3Private Practice, Breast Surgery, Ankara, Turkey 4Bagcilar Training and Research Hospital, Department Of General Surgery, Istanbul, Turkey 5Feinberg School of Medicine, Department Of Radiology, Chicago, IL, USA 6University of Pittsburgh, Department Of Radiology, Pittsburgh, PA, USA 7University Of Pittsburg, Department Of Surgery, Pittsburgh, PA, USA
Introduction: Idiopathic granulomatous mastitis (IGM) is one of the most difficult diseases to treat in surgical clinics, with options ranging from simple observation to frequent abscess aspirations, intralesional steroid injections, systemic treatments, surgical interventions, or a combination of above. The absence of consensus on clinical and radiologic classifications has prevented the standardization of treatment algorithms. In this study, we proposed Pittsburgh classifications of clinical and radiologic findings and associated treatment algorithms for IGM then assessed concordance with current treatment approaches.
Methods: We conducted an ethics-committee approved multicenter retrospective medical record and ultrasound image review of 172 patients treated for IGM. Proposed clinical classification ranges from Type 1 (minimal skin irritation), Type 2 (abscess), Type 3 (palpable mass) to Type 5 (widespread involvement). US categorizations range from Type A (localized mass ≤ 2 cm), Type B (localized mass> 2 cm), Type C (regional + fistula) through Type E (diffuse disease). A treatment algorithm was developed to align with clinical and US findings, and actual patient management was compared to the proposed standard. In 3 patients (1.8%) who had multiple lesions, the treatment chosen for some of the lesions was compatible, but not for others; these were assessed as partial concordance and not included in the analysis
Results: Data from 156 women treated for IGM were evaluable. Mean patient age was 36.9 + 8.9 years, with average follow-up duration of 11.5 + 8.7 months. Bilateral disease was seen in 10 women (6.4%) of the women. Eighty-seven (55.7%) women had a single IGM focus, while 72 (45.3%) had 2-4 foci. Of the 156 women, 120 (76.9%), received treatment concordant with clinical and imaging findings based on the Pittsburgh treatment algorithm. Of 120 women, 20 (16.7%) underwent surgical excision; 89/120 (74.2%) had a complete response, 24 (20%) had a near-complete response, and seven (5.6%) showed no response. Surgical excision was performed on 5 (13.8%) of 36 patients with treatment discordant with the Pittsburgh algorithm. Eight of these 36 (22.2%) showed a complete response to treatment, 21 (58.3%) showed a partial response, and seven (19.4%) showed no response. Complete response was more likely achieved in those with treatment concordant with the Pittsburgh treatment algorithm than discordant group (74.2% vs 22.2%, p<0.001).
Conclusion: We found the treatment algorithm created based on the Pittsburgh clinical and ultrasound classifications to be validated by current treatment approaches. Importantly, our results suggest that concordance with the Pittsburgh treatment algorithm improves complete response rates in IGM; further study is warranted and is planned.