The present study demonstrates that CE-MRI can accurately depict not only the extent of PD spread on the skin but also its ductal spread. Hence, information from CE-MRI is useful for determining the appropriate surgical procedure for PD.

Historically, mastectomy has been the recommended surgical procedure for treating PD [5]. However, BCS has recently been accepted as a proper surgical procedure in patients with PD to a limited extent [8]. Bijker et al. demonstrated a relatively low 5-year local recurrence rate of 5.2% after central lumpectomy, including NAC and irradiation, suggesting its feasibility for limited-extent PD [9]. Consistent with this report, Marshall et al. recommended local excision and subsequent breast irradiation as alternatives to mastectomy, particularly for patients with no palpable mass or abnormal mammographic density, with a 5-year local control rate of 91% [10]. Subsequent studies have highlighted the role of CE-MRI in the surgical planning of PD. Siponen et al. reported that BCS was performed in 18 (31.0%) of 58 patients with PD and that only 1 patient (5.6%) developed local recurrence. They suggested that CE-MRI might be beneficial when considering BCS. However, in their study, only 14 patients underwent CE-MRI, while 18 underwent BCS [5]. Similarly, Dominici et al. demonstrated the usefulness of CE-MRI in surgical planning of PD. However, in their study of 51 patients with PD, CE-MRI was performed in 27 (52.9%) patients, while 37 (72.5%) patients underwent BCS. Thus, it is possible that CE-MRI findings were not sufficiently used when BCS was chosen, and it was unclear whether CE-MRI was truly useful in determining the surgical procedure for PD based on the results of their study. To date, no reports have described the association between CE-MRI and histopathological findings in patients with PD. Our study fills this gap by providing detailed information on the association between CE-MRI findings and histopathological findings of PD. Although the clinical practice guidelines from the National Comprehensive Cancer Network for breast cancer recommend a preoperative imaging work-up using CE-MRI, the primary goal of this recommendation is to identify otherwise clinically occult breast cancer rather than to assess the extent of ductal spread or skin involvement in PD [11]. In this regard, the results of this study showed that CE-MRI accurately predicts the extent of spread on the skin and ductal spread of PD, providing novel insights into treatment strategies for PD. When CE-MRI does not suggest widespread ductal spread of PD, BCS or central lumpectomy, including NAC, may be sufficient. Conversely, mastectomy was preferred when CE-MRI indicated extensive ductal spread or mammographically occult independent breast cancer.

Central lumpectomy, including NAC, is less invasive than BCS and can be safely performed under local anesthesia with minimal invasion. In an aging society in Japan, there is an increasing demand for minimally invasive surgeries in elderly patients. In this regard, central lumpectomy, including NAC, as a less volumetric procedure may be preferable if the tumor can be completely resected. In line with this context, in PD of the breast, the possibility of axillary lymph node metastasis is extremely low because PD is a malignancy in situ, and omission of axillary surgery is acceptable. Indeed, in our patient cohort, none of the patients had metastases to sentinel lymph nodes. Thus, a complete cure can be achieved if complete tumor resection is achieved using central lumpectomy including NAC alone. Furthermore, patients with PD experience symptoms such as eczema during NAC, necessitating surgical intervention [12]. Hence, central lumpectomy, including NAC, under local anesthesia may be a viable and minimally invasive option, particularly in elderly patients with limited ductal spread of PD. This was evidenced in our cohort, in which patients of > 85 years of age underwent curative operations under local anesthesia guided by preoperative CE-MRI.

In the present study, CE-MRI accurately evaluated not only the ductal spread of PD but also skin spread, with a mean discrepancy of 0.6 mm between CE-MRI and histopathological findings. The mechanisms underlying skin lesion enhancement in PD on CE-MRI are not fully understood. Dissecting these mechanisms may offer new insights into PD pathology. Generally, the enhanced area of breast cancer on CE-MRI is attributed to angiogenesis and inflammatory cell infiltration into the stroma surrounding the ducts affected by cancer [13,14,15]. As shown in Fig. 3b and Fig. 4c, d, such histopathological changes were observed in both the stroma around the ductal spread of PD and the dermis beneath the PD lesions, potentially leading to skin enhancement on CE-MRI in patients with PD.

The present study was associated with several limitations. First, the small sample size (12 patients) limits the generalizability of our findings. More extensive studies are needed to validate the utility of CE-MRI in the surgical planning of PD. However, one strength of this study is that we reported a detailed association between CE-MRI findings and histopathological findings in PD in each case, in contrast to previous studies [5, 8,9,10]. This information enabled us to evaluate the role of CE-MRI in planning surgical procedures. Second, the retrospective nature of the study may have introduced selection bias, as only patients who underwent surgery and preoperative CE-MRI were included. These limitations highlight the need for more comprehensive studies to fully establish the role of CE-MRI in the management of PD. Furthermore, although 83.3% of the patients (10 of 12) included in this study were examined using 1.5 T MRI, 3 T MRI is becoming increasingly common. 3 T MRI provides images with higher resolution and enhanced tissue contrast, and hence may have the potential to improve diagnostic accuracy in PD by better detecting subtle ductal spread or skin involvement. With 3 T MRI, central lumpectomy, including NAC or BCS, with even smaller margins (< 20 mm) may be achievable. Future studies using 3 T MRI are warranted.

In conclusion, CE-MRI is useful for evaluating the extent of PD, including ductal and skin spread, around the NAC. Mastectomy should be considered when CE-MRI findings suggest extensive ductal spread of the PD. However, in patients with limited ductal spread, as predicted by CE-MRI, BCS, or central lumpectomy including NAC, it may be the preferred surgical option.