Accurate preoperative assessment of tumor size in breast cancer is a critical step for optimal surgical planning and oncologic management. Indeed, it directly influences the choice between breast-conserving surgery (BCS) and mastectomy, with the objective of achieving negative margins and minimizing the need for re-excision [1]. Positive margins are the strongest predictor of local recurrences [[2], [3], [4]] and are linked to decreased overall survival [5]. For invasive breast carcinoma, the standard definition of a clear margin in Europe and North America is “no tumor on ink,” whereas for ductal carcinoma in situ (DCIS), a wider margin—typically 2 mm—is recommended [[6], [7], [8]]. This stricter margin criterion is driven by the propensity for residual intraductal disease due to the discontinuous, often extensive growth pattern of DCIS and by differences in adjuvant management between pure DCIS and invasive breast cancer with associated DCIS. Moreover, studies report higher local recurrence rates for invasive breast cancer when an associated DCIS component—particularly an extensive intraductal component—is present compared with invasive disease without DCIS [[9], [10], [11], [12]]. DCIS component is often underestimated on mammography [13,14]. That's why lesion size estimates on imaging should reflect total size, including surrounding in-situ disease.

Various imaging modalities play a fundamental role in tumor size assessment beyond initial diagnosis. Conventional mammography (MG) remains the primary imaging tool in breast cancer screening; however, its accuracy in tumor sizing is limited, particularly in dense breast tissue [[15], [16], [17]]. Breast ultrasound (BUS) and breast magnetic resonance imaging (MRI) offer improved measurement precision, yet each presents specific drawbacks. MRI, considered as the most accurate technique for preoperative evaluation, fluctuates according to studies between overestimation [[18], [19], [20]] and underestimation [21,22], while BUS frequently underestimates it [20,23]. The need for a reliable imaging modality that balances sensitivity, specificity, and practical applicability remains crucial.

Contrast-enhanced spectral mammography (CESM) has emerged as a promising alternative, combining the accessibility and cost-effectiveness of mammography with the enhanced sensitivity of MRI [[24], [25], [26]]. CESM uses dual-energy and iodine injection to generate both low-energy (LE) images, providing morphologic characteristics similar to digital MG [[27], [28], [29]] including microcalcifications, and high-energy (HE) images that highlight contrast uptake, revealing neovascularity such as tumor vascularization [30]. Recombined images (RI) are then generated. This technique has already demonstrated superior diagnostic accuracy compared to MG [28,31,32] and showing high sensitivity (>95%) and specificity (>80%) [33], in series pooling various histological types of breast cancer.

This study aims to evaluate the performance of CESM in accurately assessing the size of unifocal invasive ductal carcinoma (IDC) with a DCIS component in the preoperative setting, compared to the histopathological size.