The first-line approach for incidentally detected MRI lesions is a targeted second-look examination using MX and US. In most cases, abnormal findings on breast MRI, particularly those with a high risk of malignancy and mass-like lesions, can be identified through experienced physicians conducting second-look examinations using US and MX [4]. However, in cases where suspicious lesions lack corresponding findings on US/MX, there is still a low but significant probability of malignancy. Therefore, MRI-guided biopsy is necessary to investigate such lesions [11, 12].

In our study, our primary objective was to calculate the diagnostic accuracy of 9G percutaneous MRI-VABB specifically for detectable breast lesions. Additionally, we aimed to determine the rates of upgrading to malignancy for MRI-VABB-detected breast lesions and assess the corresponding underestimation rates of high-risk lesions. To establish a reliable gold standard, we used the results of the final histological examination. Obtaining data on the risk of malignancy for upgraded B3 lesions is essential for making appropriate management recommendations when high-risk lesions are diagnosed through MRI-VABB.

To the best of our knowledge, a freehand VABB technique has been described in only a few other studies in literature [13, 14]. Our results, especially the rates of benign (48.4%), high-risk (26.2%), and malignant lesions (25.2%), are in line with recent literature [8, 9]. In our report, MRI-guided VABB was not performed in 3.3% of cases, which aligns with the range reported in the literature (3–13%) [15]. This was primarily due to the non-visualization of the target at the grid position, which could be attributed to hormonal changes or indicating a benign lesion. For these patients, management was carried out through MRI follow-up.

A VABB diagnosis of a benign lesion is considered reliable when there is agreement between radiological suspicion and histological results [20, 21]. In the literature, it is recommended to conduct an MRI examination at 6 months after a benign concordant VABB [22, 23]. In contrast, based on our experience, we consider the biopsy result to be adequate without the need for a subsequent MRI 6-month follow-up, depending on factors such as the initial radiological suspicion, imaging findings before and after the biopsy, and concordant histological results. Due to the low rate of false negative results, we recommend a standard annual MX/US follow-up after a benign-concordant MRI VABB. However, it is important to note that, unlike stereotactic and US-guided VABB, real-time confirmation of the removed target lesion is unavailable in MRI-biopsy. Additionally, the presence of a residual lesion cannot be completely ruled out in the biopsy cavity after VABB, as gadolinium is not present and post-procedural bleeding can obscure visibility. A close MRI follow-up is advised in the case of a discordant result from the MRI VABB, as it enables the detection of enhanced residual lesions that may require a second biopsy or further follow-up.

Therefore, based on our current experience in 2023, a 6-month MRI follow-up after a negative VABB is not recommended. We acknowledge that this approach might not be applicable on a broader scale due to the extremely low number of false negative cases. This data could be extended to other hospital settings characterized by a significant number and extensive expertise of MRI-guided VABB biopsies.

The management of B3 lesions remains controversial.

According to the UK guidelines [24], vacuum-assisted excision (VAE) is recommended as the gold standard for managing all B3 lesions without atypia (B3a). Surgical management, on the other hand, is recommended for lesions with atypia (B3b) due to a significantly higher risk of underestimation. In the Third International Consensus Conference on lesions of uncertain malignant potential in the breast (B3 lesions) [25], the majority of panelists suggest open excision for ADH and PT, while for other B3 lesions (RS, FEA, PL, LIN) VAE is considered an alternative to open excision. High-risk lesions that were diagnosed by MRI VABB, and in which a subsequent diagnosis of invasive cancer or DCIS lesion was made at surgical excision were considered underestimates [23]. In our study, no upgrades to carcinoma were observed in the MRI-guided biopsies for B3 lesions without atypia (B3a). Based on these findings, in our institution, we recommend surgical excision for B3a lesions only if it is recommended by the surgeon, based on patient preferences, or if these lesions have a size larger than 15 mm. No cases of FEA and PA have been upgraded to malignant lesions. However, according to Lourenco et al. [26] there is a significant risk of underestimation for RS (23.1%) and PA (5.9%) identified during MRI-guided breast biopsy. Based on our analysis, we recommend surgical intervention for PA lesions as well. In fact, it’s worth noting that we only analyzed one PA lesion in our study. On the other hand, we examined 10 cases of FEA, and all FEA lesions had dimensions less than 10 mm. Based on this observation, it can be concluded that surgery could be avoided for FEA lesions with dimensions below this threshold especially if the MRI VABB procedure was considered adequate.

Our study found an ADH upgrade rate of 23.1%, which falls within the lower range of upgrades reported in the literature (Table 3—16.7–100%) [30,31,32]. Consistent with our findings, other studies have also demonstrated that ADH is the most commonly identified B3 lesion on MRI and has the highest upgrade rate [33, 34]. Based on these results, we recommend excision of ADH lesions detected on MRI-guided core biopsy. This recommendation is supported by the findings of Michaels et al. [29], who confirmed that ADH lesions are more likely to upgrade compared to other B3 lesions (Table 4).

In our experience, we found that LIN1-2 lesions have an upgrade risk of 9.1%, which falls within the lower range reported in other studies in the literature (16%) [29].

Although the upgrade rate of LIN lesions to malignancy is lower compared to ADH, it is still considered high. These results, in line with the existing literature, indicate that B3b lesions detected on MRI-VABB should be considered for surgical excision, particularly in cases of ADH or LIN diagnosis.

In conclusion, our study reveals a relatively high rate of upgrade to malignancy for high-risk lesions, specifically ADH or LIN, detected through MRI-VABB. This finding suggests that surgical excision is warranted for these lesions. It is worth noting that there were no specific MRI imaging characteristics that could reliably predict the upgrade to malignancy, emphasizing the importance of pathology in guiding the decision for surgical excision. Furthermore, our malignancy rate aligns with those reported in the literature, indicating consistency with previous studies (25.2% vs. 21% in other studies) [33].

MRI-guided VABB procedures have proven to be effective in characterizing suspicious non-palpable breast lesions detected exclusively through MRI. Our study demonstrates that these procedures are well-tolerated, simple to perform, and reliable in their results. This consistency in performance allows for the identification of new disease, enabling healthcare professionals to make crucial decisions for accurate surgical planning.

This study has main limitations. The first limitation is its retrospective design.

Additionally, the study was conducted at a single center, which may limit the generalizability of the findings to other settings. Furthermore, we were unable to evaluate the negative predictive value of MRI-guided biopsy due to the loss of some patients during the follow-up period. Another limitation is the small sample size, particularly for each subtype of B3b lesions, which may affect the statistical power and precision of the results. However, by comparing our findings to those of similar studies, we have contributed additional data and information that can help guide the management of B3 lesions detected through MRI.