This study showed a significant correlation between Node-RADS score and lymph node invasion. Node-RADS score > 2 was identified as the most suitable cut-off point for predicting lymph node invasion. A Node-RADS score of 3, 4, or 5 should be taken into consideration as strongly indicative of metastatic disease and this information should be considered as part of patient management.

The involvement and the number of axillary lymph nodes affected are widely recognized as crucial prognostic factors for BC. A recent meta-analysis demonstrated that the presence of occult metastases, as opposed to their absence, was significantly associated with a poorer 5-year disease-free survival (RR 1.55, 95% CI 1.32–1.82) and overall survival (RR 1.45, 95% CI 1.11–1.88) [22].

Currently, the technique used to evaluate axillary lymph-node status in BC is ultrasound. US is highly specific for the assessment of lymph nodes by using morphological criteria (round shape, absence of fatty hilum, thickening of the cortex > 3 mm) [10], raising a specificity of 98.3% in a large meta-analysis by Houssami et al [4], however, a standard score for their evaluation has not yet been created.

Cross-sectional imaging, such as CT or CE-MRI, is also important for disease evaluation and lymph node staging, but it has some limitations due to the lack of standardized criteria for defining the exact involvement of the lymph nodes [3].

Traditionally, staging has relied on the TNM staging system, which determines the “baseline risk” of BC at the time of diagnosis and after surgery [23]. Clinical nodal involvement (N stage), despite any lack of standardization, remains of paramount importance in guiding management decisions for BC patients.

Node staging plays a crucial role in predicting the likelihood of recurrence in patients who are not receiving systemic therapy. It aids in determining whether a patient should receive adjuvant chemotherapy, endocrine therapy, or anti-HER2 therapy, based on the predicted risk of recurrence [23]. This information is invaluable in tailoring personalized treatment plans for individual patients and ensuring that they receive the most appropriate and effective therapies.

The eighth edition of the TNM staging system aimed to standardize the evaluation of lymph nodes by providing more precise guidelines for the pathological measurement of lymph node metastases [8]. This included specifying different methods of approximation for the size of metastases, such as considering diameters greater than 1 mm and less than 2 mm, or equal to or greater than 2 mm. Additionally, the evaluation of lymph node clusters was addressed, emphasizing that the largest aggregate of contiguous tumor cells should be measured without including separate tumor clusters. However, the TNM staging system did not introduce a definitive diagnostic scoring system for the evaluation of lymphadenopathy based on imaging [8].

In parallel, the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BIRADS) also proposed a qualitative and subjective assessment of lymph nodes. According to the ACR BIRADS guidelines, lymph nodes are described based on their size, without establishing a specific size cut-off and considering any size increase from a previous examination. Furthermore, the loss of the adipose (fatty) hilum appearance of lymph nodes and the evaluation of their margins are taken into account during the assessment [23].

These guidelines, although providing more clarity on the measurement and evaluation of lymph nodes, still rely on subjective and qualitative criteria, lacking a definitive quantitative scoring system for diagnosing lymphadenopathy through imaging.

As a result, clinical judgment and expertise continue to play a significant role in the accurate evaluation and management of lymph node involvement in BC patients, especially if ultrasound is considered the method of choice.

Recently, the Node-RADS scoring system was introduced to address this gap, providing a comprehensive evaluation of the lymph node [11].

Like other RADS, this system aims to improve the distinction between benign and malignant diseases, remove ambiguity from radiology reports, allow for automated auditing of data, and enhance clinical communication with referrers.

In previous studies focused on other parts of the body, such as the prostate, the bladder, the lung, the colon, and the stomach, the Node-RADS score has already been validated and demonstrated significant utility, with favorable outcomes [12,13,14,15,16] comparable to the results obtained in our current research. Moreover, the Node-RADS score emerged as an independent predictor of lymph node involvement with a moderate-to-high overall accuracy in identifying LNI. Additionally, its flexibility in allowing the establishment of different cut-off values based on specific clinical scenarios enhances its clinical applicability [12,13,14,15,16].

In the current study, we assessed the overall diagnostic performance of the Node-RADS scores and hypothesized that the Node-RADS score independently correlates with lymph node involvement.

The Node-RADS scores revealed a positive trend in the rates of LNI. Specifically, Node-RADS scores increased the LNI risk, establishing their status as an independent predictor even after adjusting for multiple variables (p < 0.001). The linear-by-linear association depicted this relationship, showing a progressive rise in LNI risk with increasing Node-RADS scores.

Based on the balanced sensitivity and specificity values for all Reader we concluded that a Node-RADS > 2 could be considered the best cut-off since, from the results obtained, it is possible to determine the presence or absence of suspicious lymph nodes.

Drawing from our collective experience, even when considering various guidelines, national consensus, and corporate recommendations, lymph nodes assigned a Node-RADS score of 1 or 2 are generally benign and do not necessitate further specific assessment.

Therefore, we recommend focusing on malignant Node-RADS scores (3, 4, and 5). For cases falling into this category, we propose a US-guided biopsy prior to surgery. This approach ensures accurate management strategies and facilitates the placement of a clip to assess the lymph nodes’ response to subsequent neoadjuvant chemotherapy.

In situations where lymph nodes receive a Node-RADS > 2 in the absence of a diagnosed breast tumor, we advise conducting a mammographic and US examination, followed by a biopsy.

The agreement among the three readers varied from good to excellent, even for the less experienced breast radiologist. The agreement values evaluated for subclasses of Node-RADS cut-offs, in most cases show even better agreement. This high level of concordance indicates that the Node-RADS scoring system can be reliably applied by different readers, enhancing its practical utility in clinical settings, including for novice radiologists.

In our study, we did not evaluate the timing for scoring, however, subjectively, in the opinion of the three readers, the average reading time per examination was the same for the Node-RADS scores and the non-standardized method of scoring nodes.

The readers found the scoring system useful in doubtful cases where the lack of standardization could lead to confusion, significantly increasing the reading times of the exam and the discrimination of suspicious lymph nodes from non-suspicious ones.

However, it is essential to acknowledge the limitations of our study, which was based on a relatively limited cohort. As such, these results warrant further validation in a larger, more diverse cohort to confirm the robustness and generalizability of the Node-RADS scoring system. Additionally, this study adopted a retrospective design, which may introduce inherent biases and limit the establishment of causal relationships between variables and outcomes.

Furthermore, the Node-RADS scoring system applicability might be specific to patients who undergo surgical treatment for BC, and caution should be exercised when extrapolating the findings to other treatment modalities or patient populations.

To the best of our knowledge, no other scoring system for lymph nodes exists and this study represents the first attempt to test the diagnostic performance of Node-RADS in surgically treated BC patients. Consequently, direct comparisons of our results with those of other studies may not be possible. Nonetheless, our findings contribute to the emerging body of knowledge regarding the role of Node-RADS in the clinical management of BC patients and underscore the need for further research to fully comprehend its potential impact on patient care.