Diagnostic Value of Axillary Ultrasound, MRI, and 18F-FDG-PET/ CT in Determining Axillary Lymph Node Status in Breast Cancer Patients

Deep learning-based breast MRI for predicting axillary lymph node metastasis: a systematic review and meta-analysis

BACKGROUND

To perform a systematic review and meta-analysis that assesses the diagnostic performance of deep learning algorithms applied to breast MRI for predicting axillary lymph nodes metastases in patients of breast cancer.

METHODS

A systematic literature search in PubMed, MEDLINE, and Embase databases for articles published from January 2004 to February 2025. Inclusion criteria were: patients with breast cancer; deep learning using MRI images was applied to predict axillary lymph nodes metastases; sufficient data were present; original research articles. Quality Assessment of Diagnostic Accuracy Studies-AI and Checklist for Artificial Intelligence in Medical Imaging was used to assess the quality. Statistical analysis included pooling of diagnostic accuracy and investigating between-study heterogeneity. A summary receiver operating characteristic curve (SROC) was performed. R statistical software (version 4.4.0) was used for statistical analyses.

RESULTS

A total of 10 studies were included. The pooled sensitivity and specificity were 0.76 (95% CI, 0.67-0.83) and 0.81 (95% CI, 0.74-0.87), respectively, with both measures having moderate between-study heterogeneity (I2 = 61% and 60%, respectively; p < 0.01). The SROC analysis yielded a weighted AUC of 0.788.

CONCLUSION

This meta-analysis demonstrates that deep learning algorithms applied to breast MRI offer promising diagnostic performance for predicting axillary lymph node metastases in breast cancer patients. Incorporating deep learning into clinical practice may enhance decision-making by providing a non-invasive method to more accurately predict lymph node involvement, potentially reducing unnecessary surgeries.

Development and validation of a machine learning model to predict the risk of lymph node metastasis in early-stage supraglottic laryngeal cancer

Background

Cervical lymph node metastasis (LNM) is a significant factor that leads to a poor prognosis in laryngeal cancer. Early-stage supraglottic laryngeal cancer (SGLC) is prone to LNM. However, research on risk factors for predicting cervical LNM in early-stage SGLC is limited. This study seeks to create and validate a predictive model through the application of machine learning (ML) algorithms.

Methods

The training set and internal validation set data were extracted from the Surveillance, Epidemiology, and End Results (SEER) database. Data from 78 early-stage SGLC patients were collected from Fujian Provincial Hospital for independent external validation. We identified four variables associated with cervical LNM and developed six ML models based on these variables to predict LNM in early-stage SGLC patients.

Results

In the two cohorts, 167 (47.44%) and 26 (33.33%) patients experienced LNM, respectively. Age, T stage, grade, and tumor size were identified as independent predictors of LNM. All six ML models performed well, and in both internal and independent external validations, the eXtreme Gradient Boosting (XGB) model outperformed the other models, with AUC values of 0.87 and 0.80, respectively. The decision curve analysis demonstrated that the ML models have excellent clinical applicability.

Conclusions

Our study indicates that combining ML algorithms with clinical data can effectively predict LNM in patients diagnosed with early-stage SGLC. This is the first study to apply ML models in predicting LNM in early-stage SGLC patients.

Can we skip invasive biopsy of sentinel lymph nodes? A preliminary investigation to predict sentinel lymph node status using PET/CT-based radiomics

BACKGROUND

Sentinel lymph node (SLN) biopsy (SLNB) is considered the gold standard for detecting SLN metastases in patients with invasive ductal breast cancer (IDC). However, SLNB is invasive and associated with several complications. Thus, this study aimed to evaluate the diagnostic performance of a non-invasive radiomics analysis utilizing 2-deoxy-2-[18F]fluoro-d-glucose positron emission tomography/computed tomography (18F-FDG-PET/CT) for assessing SLN metastasis in IDC patients.

METHODS

This retrospective study included 132 patients with biopsy-confirmed IDC, who underwent 18F-FDG PET/CT scans prior to mastectomy or breast-conserving surgery with SLNB. Tumor resection or SLNB was conducted within one-week post-scan. Clinical data and metabolic parameters were analyzed to identify independent SLN metastasis predictors. Radiomic features were extracted from each PET volume of interest (VOI) and CT-VOI. Feature selection involved univariate and multivariate logistic regression analysis, and the least absolute shrinkage and selection operator (LASSO) method. Three models were developed to predict SLN status using the random forest (RF), decision tree (DT), and k-Nearest Neighbors (KNN) classifiers. Model performance was assessed using the area under the receiver operating characteristic curve (AUC).

RESULTS

The study included 91 cases (32 SLN-positive and 59 SLN-negative patients) in the training cohort and 41 cases (29 SLN-positive and 12 SLN-negative patients) in the validation cohort. Multivariate logistic regression analysis identified Ki 67 and TLG as independent predictors of SLN status. Five PET-derived features, three CT-derived features, and two clinical variables were selected for model development. The AUC values of the RF, KNN, and DT models for the training cohort were 0.887, 0.849, and 0.824, respectively, and for the validation cohort were 0.856, 0.830, and 0.819, respectively. The RF model demonstrated the highest accuracy for the preoperative prediction of SLN metastasis in IDC patients.

CONCLUSION

The PET-CT radiomics approach may offer robust and non-invasive predictors for SLN status in IDC patients, potentially aiding in the planning of personalized treatment strategies for IDC patients.

Disease extent according to baseline [18F]fluorodeoxyglucose PET/CT and molecular subtype: prediction of axillary treatment response after neoadjuvant systemic therapy for breast cancer

BACKGROUND

Axillary disease extent according to baseline [18F]fluorodeoxyglucose PET/CT combined with pathological axillary treatment response has been proposed to guide de-escalation of axillary treatment for clinically node-positive breast cancer patients treated with neoadjuvant systemic therapy. The aim of this study was to assess whether axillary disease extent according to baseline [18F]fluorodeoxyglucose PET/CT and breast cancer molecular subtype are predictors of axillary pCR.

METHODS

This study included clinically node-positive patients treated with neoadjuvant systemic therapy in the prospective Radioactive Iodine Seed placement in the Axilla with Sentinel lymph node biopsy ('RISAS') trial (NCT02800317) with baseline [18F]fluorodeoxyglucose PET/CT imaging available. The predictive value of axillary disease extent according to baseline [18F]fluorodeoxyglucose PET/CT and breast cancer molecular subtype to estimate axillary pCR was evaluated using logistic regression analysis. Discriminative ability is expressed using ORs with 95% confidence intervals.

RESULTS

Overall, 185 patients were included, with an axillary pCR rate of 29.7%. The axillary pCR rate for patients with limited versus advanced baseline axillary disease according to [18F]fluorodeoxyglucose PET/CT was 31.9% versus 26.1% respectively. Axillary disease extent was not a significant predictor of axillary pCR (OR 0.75 (95% c.i. 0.38 to 1.46) (P = 0.404)). There were significant differences in axillary pCR rates between breast cancer molecular subtypes. The lowest probability (7%) was found for hormone receptor+/human epidermal growth factor receptor 2- tumours. Using this category as a reference group, significantly increased ORs of 14.82 for hormone receptor+/human epidermal growth factor receptor 2+ tumours, 40 for hormone receptor-/human epidermal growth factor receptor 2+ tumours, and 6.91 for triple-negative tumours were found (P < 0.001).

CONCLUSION

Molecular subtype is a significant predictor of axillary pCR after neoadjuvant systemic therapy, whereas axillary disease extent according to baseline [18F]fluorodeoxyglucose PET/CT is not.

Predicting axillary lymph node metastasis in breast cancer patients: A radiomics-based multicenter approach with interpretability analysis

PURPOSE

To develop a MRI-based radiomics model, integrating the intratumoral and peritumoral imaging information to predict axillary lymph node metastasis (ALNM) in patients with breast cancer and to elucidate the model's decision-making process via interpretable algorithms.

METHODS

This study included 376 patients from three institutions who underwent contrast-enhanced breast MRI between 2021 and 2023. We used multiple machine learning algorithms to combine peritumoral, intratumoral, and radiological characteristics with the building of radiological, radiomics, and combined models. The model's performance was compared based on the area under the curve (AUC) obtained from the receiver operating characteristic analysis and interpretable machine learning techniques to analyze the operating mechanism of the model.

RESULTS

The radiomics model, incorporating features from both intratumoral tissue and the 3 mm peritumoral region and utilizing the backpropagation neural network (BPNN) algorithm, demonstrated superior diagnostic efficacy, achieving an AUC of 0.820. The AUC of the combination of the RAD score, clinical T stage, and spiculated margin was as high as 0.855. Furthermore, we conducted SHapley Additive exPlanations (SHAP) analysis to evaluate the contributions of RAD score, clinical T stage, and spiculated margin in ALNM status prediction.

CONCLUSIONS

The interpretable radiomics model we propose can better predict the ALNM status of breast cancer and help inform clinical treatment decisions.

Does F-18 FDG-PET/CT Have an Additional Impact on Axillary Approach in Early-Stage Breast Cancer?

Objective

Breast cancer (BC) is a significant health concern and one of the most diagnosed cancers in women, both in Turkey and globally. Despite advances in the management of BC, axillary lymph node involvement remains a significant consideration for treatment planning, local recurrence, and prognosis. We aimed to evaluate the contribution of F-18 fluorodeoxyglucose-positron emission tomography/computed tomography (F-18 FDG-PET/CT) in detecting axillary lymph node metastasis compared to ultrasound (US).

Materials and Methods

Eighty patients who were diagnosed with stage I and II BC and underwent US and F-18 FDG-PET/CT scans before surgery were enrolled in this study. Those who did not undergo F-18 FDG-PET/CT imaging, patients with distant metastases at the time of diagnosis and patients with micrometastases in the axilla were excluded from the analysis. Imaging results of the status of axillary lymph nodes were verified with the final pathology report of axillary lymph nodes.

Results

The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of F-18 FDG-PET/CT for the detection of ipsilateral axillary lymph node metastases were 75%, 77.27%, 72.97%, 79.07%, and 76.25%. The corresponding values for US were 72.22%, 81.82%, 76.47%, 78.26%, and 77.50%, respectively. When US finding is negative or suspicious in axillary lymph node evaluation, the accuracy of F-18 FDG-PET/CT for the detection of ipsilateral axillary lymph node metastases were 65.38%, 83.33%, 70.83%, and 79.55%, respectively.

Conclusion

This study found that F-18 FDG-PET/CT does not provide an additional advantage over US in assessing the axilla in early-stage disease.

Ultrasound Pretreatment Lymph Node Evaluation in Early-Stage Breast Cancer: Should We Biopsy High Suspicion Nodes?

BACKGROUND

With the growing incidence of breast cancer, efficient and correct staging is essential for further treatment decisions. Axillary ultrasound (US) remains the most common method for regional nodal involvement assessment. The aim of this study was to evaluate whether high-risk US features can accurately predict axillary lymph node metastasis.

METHODS

A total of 150 early-stage breast cancer patients (T1 or T2) were prospectively included in the study. Based on axillary US, patients were classified as normal, low-risk, or high-risk, with all patients in the low-risk and high-risk groups undergoing fine-needle aspiration (FNAB) and core-needle biopsies.

RESULTS

For the low-risk US group, a lower prediction rate of axillary nodal metastasis was achieved than for the group with high-risk features, recording a sensitivity of 66.6% vs. 89.2%, a specificity of 57.1% vs. 100%, a positive predictive value (PPV) of 26.6% vs. 100%, a negative predictive value (NPV) of 88% for both groups, and an accuracy of 58.9% vs. 94%, respectively. FNAB resulted in more false-negative results compared to core-needle biopsy in both low-risk and high-risk US groups.

CONCLUSIONS

Our findings suggest that high-risk US features can predict axillary lymph node metastasis with high accuracy.

Tumor Characteristics Associated with Axillary Nodal Positivity in Triple Negative Breast Cancer

Larger-size primary tumors are correlated with axillary metastases and worse outcomes. We evaluated the relationships among tumor size, location, and distance to nipple relative to axillary node metastases in triple-negative breast cancer (TNBC) patients, as well as the predictive capacity of imaging. We conducted a single-institution, retrospective chart review of stage I-III TNBC patients diagnosed from 1998 to 2019 who underwent upfront surgery. Seventy-three patients had a mean tumor size of 20 mm (range 1-53 mm). All patients were clinically node negative. Thirty-two patients were sentinel lymph node positive, of whom 25 underwent axillary lymph node dissection. Larger tumor size was associated with positive nodes (p < 0.001): the mean tumor size was 14.30 mm in node negative patients and 27.31 mm in node positive patients. Tumor to nipple distance was shorter in node positive patients (51.0 mm) vs. node negative patients (73.3 mm) (p = 0.005). The presence of LVI was associated with nodal positivity (p < 0.001). Tumor quadrant was not associated with nodal metastasis. Ultrasound yielded the largest number of suspicious findings (21/49), with sensitivity of 0.25 and specificity of 0.40. On univariate analysis, age younger than 60 at diagnosis was also associated with nodal positivity (p < 0.002). Comparative analyses with other subtypes may identify biologic determinants.

Factors Influencing Lymph Node Positivity in HER2/neu+ Breast Cancer Patients

Axillary lymph node metastases are a key prognostic factor in breast cancer treatment. Our aim was to evaluate how tumor size, tumor location, and imaging results correlate to axillary lymph node diseases for patients with stage I-III HER2/neu+ breast cancer. This is a single-institution retrospective chart review of female breast cancer patients diagnosed with primary invasive Her2/neu+ breast cancer who were treated with upfront surgical resection from 2000–2021. Of 75 cases, 44/75 (58.7%) had nodal metastasis, and there was a significant association of larger tumor size to nodal metastases (p ≤ 0.001). Patients with negative nodes had a smaller mean tumor size (n = 30; 15.10 mm) than patients with positive nodes (n = 45; 23.9 mm) (p = 0.002). Preoperative imaging detected suspicious nodes in 36 patients, and ultrasound detected the most positive nodes (14/18; p = 0.027). Our data confirms that tumor size at diagnosis is correlated with a higher likelihood of axillary involvement in patients with Her2/neu+ breast cancer; notably, a large proportion of Her2/neu+ breast cancers have metastatic involvement of axillary lymph nodes even with small primary lesions.

Radiomics model of diffusion-weighted whole-body imaging with background signal suppression (DWIBS) for predicting axillary lymph node status in breast cancer

BACKGROUND

In breast cancer diagnosis and treatment, non-invasive prediction of axillary lymph node (ALN) metastasis can help avoid complications related to sentinel lymph node biopsy.

OBJECTIVE

This study aims to develop and evaluate machine learning models using radiomics features extracted from diffusion-weighted whole-body imaging with background signal suppression (DWIBS) examination for predicting the ALN status.

METHODS

A total of 100 patients with histologically proven, invasive, clinically N0 breast cancer who underwent DWIBS examination consisting of short tau inversion recovery (STIR) and DWIBS sequences before surgery were enrolled. Radiomic features were calculated using segmented primary lesions in DWIBS and STIR sequences and were divided into training (n = 75) and test (n = 25) datasets based on the examination date. Using the training dataset, optimal feature selection was performed using the least absolute shrinkage and selection operator algorithm, and the logistic regression model and support vector machine (SVM) classifier model were constructed with DWIBS, STIR, or a combination of DWIBS and STIR sequences to predict ALN status. Receiver operating characteristic curves were used to assess the prediction performance of radiomics models.

RESULTS

For the test dataset, the logistic regression model using DWIBS, STIR, and a combination of both sequences yielded an area under the curve (AUC) of 0.765 (95% confidence interval: 0.548-0.982), 0.801 (0.597-1.000), and 0.779 (0.567-0.992), respectively, whereas the SVM classifier model using DWIBS, STIR, and a combination of both sequences yielded an AUC of 0.765 (0.548-0.982), 0.757 (0.538-0.977), and 0.779 (0.567-0.992), respectively.

CONCLUSIONS

Use of machine learning models incorporating with the quantitative radiomic features derived from the DWIBS and STIR sequences can potentially predict ALN status.

Conservative axillary surgery is emerging in the surgical management of breast cancer

Axillary lymph node dissection (ALND) has been the standard axillary treatment for breast cancer for a long time. However, ALND is associated with postoperative morbidities, including local sensory dysfunction, reduced shoulder mobility and most notably arm lymphedema. Recently, ALND can be avoided not only in clinically node-negative (cN0) patients with negative sentinel lymph nodes (SLNs), but also in patients with less than 3 positive SLNs receiving breast radiation, axillary radiation, or a combination of the two. Moreover, SLN biopsy has been adopted for use in clinically node-positive (cN +) patients presenting as cN0 after neoadjuvant chemotherapy (NAC); ALND may be avoided in cN + patients who convert to SLN-negative following NAC. Patients who undergo SLN biopsy alone have less postsurgical morbidities than those who undergo ALND. Nevertheless, ALND is still required in a select group of patients. A variety of conservative approaches to ALND have been developed to spare arm lymphatics to minimize arm lymphedema. These conservative procedures seem to decrease the incidence of lymphedema without increasing axillary recurrence. In the era of effective multimodality therapy, full conventional ALND removing all microscopic axillary disease may now be unnecessary in both cN0 patients and cN + patients. Regardless, emerging procedures for ALND should still be considered as investigational approaches, as further studies with longer follow-up are necessary to determine the safety of conservative ALND to spare arm lymphatics.