Role of axillary ultrasound in the preoperative diagnosis of lymph node metastases in patients affected by breast carcinoma

Machine Learning Model for Predicting Axillary Lymph Node Metastasis in Clinically Node Positive Breast Cancer Based on Peritumoral Ultrasound Radiomics and SHAP Feature Analysis

OBJECTIVE

This study seeks to construct a machine learning model that merges clinical characteristics with ultrasound radiomic analysis-encompassing both the intratumoral and peritumoral-to predict the status of axillary lymph nodes in patients with early-stage breast cancer.

METHODS

The study employed retrospective methods, collecting clinical information, ultrasound data, and postoperative pathological results from 321 breast cancer patients (including 224 in the training group and 97 in the validation group). Through correlation analysis, univariate analysis, and Lasso regression analysis, independent risk factors related to axillary lymph node metastasis in breast cancer were identified from conventional ultrasound and immunohistochemical indicators, and a clinical feature model was constructed. Additionally, features were extracted from ultrasound images of the intratumoral and its 1-5 mm peritumoral to establish a radiomics feature formula. Furthermore, by combining clinical features and ultrasound radiomics features, six machine learning models (Logistic Regression, Decision Tree, Support Vector Machine, Extreme Gradient Boosting, Random Forest, and K-Nearest Neighbors) were compared for diagnostic efficacy, and constructing a joint prediction model based on the optimal ML algorithm. The use of Shapley Additive Explanations (SHAP) enhanced the visualization and interpretability of the model during the diagnostic process.

RESULTS

Among the 321 breast cancer patients, 121 had axillary lymph node metastasis, and 200 did not. The clinical feature model had an AUC of 0.779 and 0.777 in the training and validation groups, respectively. Radiomics model analysis showed that the model including the Intratumor +3 mm peritumor area had the best diagnostic performance, with AUCs of 0.847 and 0.844 in the training and validation groups, respectively. The joint prediction model based on the XGBoost algorithm reached AUCs of 0.917 and 0.905 in the training and validation groups, respectively. SHAP analysis indicated that the Rad Score had the highest weight in the prediction model, playing a significant role in predicting axillary lymph node metastasis in breast cancer.

CONCLUSION

The predictive model, which integrates clinical features and radiomic characteristics using the XGBoost algorithm, demonstrates significant diagnostic value for axillary lymph node metastasis in breast cancer. This model can provide significant references for preoperative surgical strategy selection and prognosis evaluation for breast cancer patients, helping to reduce postoperative complications and improve long-term survival rates. Additionally, the utilization of SHAP enhancing the global and local interpretability of the model.

MRI Findings Suggestive of Metastatic Axillary Lymph Nodes in Patients with Invasive Breast Cancer

Purpose

This study aimed to investigate the diagnostic performance of features suggestive of nodal metastasis on preoperative MRI in patients with invasive breast cancer.

Materials and Methods

We retrospectively reviewed the preoperative breast MRI of 192 consecutive patients with surgically proven invasive breast cancer. We analyzed MRI findings of axillary lymph nodes with regard to the size, long/short ratio, cortical thickness, shape and margin of the cortex, loss of hilum, asymmetry, signal intensity (SI) on T2-weighted images (T2WI), degree of enhancement in the early phase, and enhancement kinetics. Receiver operating characteristic (ROC) analysis, chi-square test, t test, and McNemar’s test were used for statistical analysis.

Results

Increased shorter diameter, uneven cortical shape, increased cortical thickness, loss of hilum, asymmetry, irregular cortical margin, and low SI on T2WI were significantly suggestive of metastasis. ROC analysis revealed the cutoff value for the shorter diameter and cortical thickness as 8.05 mm and 2.75 mm, respectively. Increased cortical thickness (> 2.75 mm) and uneven cortical shape showed significantly higher sensitivity than other findings in McNemar’s test. Irregular cortical margins showed the highest specificity (100%).

Conclusion

Cortical thickness > 2.75 mm and uneven cortical shape are more sensitive parameters than other findings, and an irregular cortical margin is the most specific parameter for predicting axillary metastasis in patients with invasive breast cancer.

Pretherapeutic Imaging for Axillary Staging in Breast Cancer: A Systematic Review and Meta-Analysis of Ultrasound, MRI and FDG PET

BACKGROUND

This systematic review aimed at comparing performances of ultrasonography (US), magnetic resonance imaging (MRI), and fluorodeoxyglucose positron emission tomography (PET) for axillary staging, with a focus on micro- or micrometastases.

METHODS

A search for relevant studies published between January 2002 and March 2018 was conducted in MEDLINE database. Study quality was assessed using the QUality Assessment of Diagnostic Accuracy Studies checklist. Sensitivity and specificity were meta-analyzed using a bivariate random effects approach; Results: Across 62 studies (n = 10,374 patients), sensitivity and specificity to detect metastatic ALN were, respectively, 51% (95% CI: 43-59%) and 100% (95% CI: 99-100%) for US, 83% (95% CI: 72-91%) and 85% (95% CI: 72-92%) for MRI, and 49% (95% CI: 39-59%) and 94% (95% CI: 91-96%) for PET. Interestingly, US detects a significant proportion of macrometastases (false negative rate was 0.28 (0.22, 0.34) for more than 2 metastatic ALN and 0.96 (0.86, 0.99) for micrometastases). In contrast, PET tends to detect a significant proportion of micrometastases (true positive rate = 0.41 (0.29, 0.54)). Data are not available for MRI.

CONCLUSIONS

In comparison with MRI and PET Fluorodeoxyglucose (FDG), US is an effective technique for axillary triage, especially to detect high metastatic burden without upstaging majority of micrometastases.

Does axillary lymph node size predict better metastatic involvement than apparent diffusion coefficient (ADC) value in women with newly diagnosed breast cancer?

BACKGROUND

It has been demonstrated that the number of metastatic axillary lymph nodes (mALNs) influence disease-free and overall survival in patients with breast cancer.

PURPOSE

To determine if the ALN size is more accurate than the ALN apparent diffusion coefficient (ADC) value to predict metastatic involvement in newly diagnosed breast cancer.

MATERIAL AND METHODS

A total of 44 patients with breast cancer were included. Magnetic resonance imaging (MRI) examinations were performed on a 1.5-T system with sagittal T1-weighted fast spin-echo non-fat saturated, sagittal T2-weighted fast spin-echo non-fat saturated, axial diffusion-weighted imaging echo-planar (b values of 0 and 700 s/mm²), and non-contrast axial VIBRANT sequences. The size and the ADC value were obtained for ALN ipsilateral and contralateral to breast cancer. The reference standard was the histopathologic lymph node status.

RESULTS

mALN had a greater cortical thickness compared to contralateral non-mALN (10.3 ± 5.32 mm vs. 4 ± 1.17 mm, P ≤ 0.001). The threshold of ≥6.7 mm for predicting axillary metastatic involvement had a sensitivity and a specificity of 80.0% and 97.7%, respectively. The ADC value of mALN was significantly higher than the contralateral non-mALN (0.90 ± 0.12 × 10^-3mm²/s vs. 0.78 ± 0.12 × 10^-3mm²/s; P = 0.001). The threshold of ≥0.86 × 10^-3mm²/s had a sensitivity and a specificity of 66.7% and 76.7%, respectively.

CONCLUSION

Our results indicate that the cortical thickness has a better diagnostic performance in the differentiation of metastatic and non-metastatic lymph nodes than the lymph node ADC.

Diagnostic Accuracy of Fine-Needle Aspiration Cytology and Core-Needle Biopsy in the Assessment of the Axillary Lymph Nodes in Breast Cancer-A Meta-Analysis

BACKGROUND

The present study aims to evaluate the diagnostic accuracy between ultrasonography-guided fine-needle aspiration cytology (US-FNAC) and core needle biopsy (CNB) of axillary lymph nodes (ALNs) in patients with breast cancer through a meta-analysis and a diagnostic test accuracy (DTA) review.

METHODS

The present meta-analysis and DTA review included 67 eligible studies. The diagnostic accuracy of various preoperative assessments, including US-FNAC and CNB, was evaluated for ALNs assessments in patients with breast cancer. In addition, a subgroup analysis based on methods of cytologic preparation was performed. In the DTA review, the sensitivity, specificity, diagnostic odds ratio (OR) and area under the curve (AUC) on the summary receiver operating characteristic (SROC) curve were calculated.

RESULTS

The diagnostic accuracy of the preoperative assessments of ALNs was 0.850 (95% confidence interval (CI) 0.833-0.866) for patients with breast cancer. The diagnostic accuracy of CNB was significantly higher than that of US-FNAC (0.896, 95% CI 0.844-0.932 vs. 0.844, 95% CI 0.825-0.862; p = 0.044 in a meta-regression test). In the subgroup analysis based on cytologic preparation, the diagnosis accuracies were 0.860, 0.861 and 0.859 for the methods of conventional smear, liquid-based preparation and cell block, respectively. In the DTA review, CNB showed higher sensitivity than US-FNAC (0.849 vs. 0.760). However, there was no difference in specificity between US-FNAC and CNB (0.997 vs. 1.000). US-FNAC with liquid-based preparation and CNB showed the highest diagnostic OR and AUC on the SROC, respectively.

CONCLUSION

Both US-FNAC and CNB are useful in preoperative assessments of ALNs in patients with breast cancer. Although the most sensitive test was found to be CNB in this study, there was no difference in specificity between various preoperative evaluations and the application of US-FNAC or CNB may be impacted by various factors.

Improved false negative rate of axillary status using sentinel lymph node biopsy and ultrasound-suspicious lymph node sampling in patients with early breast cancer

Background

The false negative rate of sentinel lymph node biopsy (SLNB) is 5-10%, and results in improper patient management. The study was to assess the value of ultrasound-suspicious axillary lymph node biopsy (USALNB) in patients with early breast cancer, and to compare SLNB combined with USALNB (SLNB + USALNB) with SLNB alone.

Methods

From January 2010 to July 2013, 216 patients with early breast cancer were enrolled consecutively at the Department of Breast and Thyroid Surgery, Qianfoshan Hospital, Shandong University. All patients underwent wire localization of the suspicious node by color Doppler ultrasonography, followed by SLNB 2–3 hours later, suspicious node lymphadenectomy, and level ≥ II axillary dissection (as the gold standard). The predictive values of node status between SLNB + USALNB and SLNB alone were compared.

Results

The success rate of SLNB was 99.1% (214/216). After axillary dissection, 71 patients were confirmed with axillary lymph node metastases by pathological examinations. Eight false negatives were observed using SLNB alone, resulting in sensitivity of 88.7%, specificity of 100%, false negative rate of 11.3%, and false positive rate of 0% in predicting the axillary node status. SLNB + USALNB resulted in sensitivity of 97.2%, specificity of 100%, false negative rate of 2.8%, and false positive rate of 0%. The false negative rate of SLNB + USALNB was significantly different from that of SLNB alone (P = 0.031).

Conclusions

SLNB + USALNB seems to be a low-risk procedure that might be useful in reducing the false negative rate of SLNB, improving the accuracy of axillary nodes evaluation in early breast cancer.

Axillary lymph nodes in breast cancer patients: sonographic evaluation

Axillary staging of patients with early-stage breast cancer is essential in the treatment planning. Currently such staging is intraoperatively performed, but there is a tendency to seek a preoperative and less invasive technique to detect lymph node metastasis. Ultrasonography is widely utilized for this purpose, many times in association with fine-needle aspiration biopsy or core needle biopsy. However, the sonographic criteria for determining malignancy in axillary lymph nodes do not present significant predictive values, producing discrepant results in studies evaluating the sensitivity and specificity of this method. The present study was aimed at reviewing the literature approaching the utilization of ultrasonography in the axillary staging as well as the main morphological features of metastatic lymph nodes.

Staging the axilla in women with breast cancer: the utility of preoperative ultrasound-guided needle biopsy

Preoperative staging of the axilla in women with invasive breast cancer using ultrasound-guided needle biopsy (UNB) identifies approximately 50% of patients with axillary nodal metastases prior to surgical intervention. Although moderately sensitive, it is a highly specific staging strategy that is rarely falsely-positive, hence a positive UNB allows patients to be triaged to axillary lymph-node dissection (ALND) avoiding potentially unnecessary sentinel node biopsy (SNB). In this review, we extend our previous work through an updated literature search, focusing on studies that report data on UNB utility. Based on data for 10,934 breast cancer patients, sourced from 35 studies, a positive UNB allowed triage of 1,745 cases (simple proportion 16%) to axillary surgical treatment: the utility of UNB was a median 19.8% [interquartile range (IQR) 11.6%-26.7%] across these studies. We also modelled data from a subgroup of studies, and estimated that amongst patients with metastases to axillary nodes, the odds ratio (OR) for high nodal disease burden for a positive UNB versus a negative UNB was 4.38 [95% confidence interval (95% CI): 3.13, 6.13], P<0.001. From this model, the estimated proportion with high nodal disease burden was 58.9% (95% CI: 50.2%, 67.0%) for a positive UNB, whereas the estimated proportion with high nodal disease burden was 24.6% (95% CI: 17.7%, 33.2%) if UNB was negative. Overall, axillary UNB has good clinical utility and a positive UNB can effectively triage to ALND. However, the evolving landscape of axillary surgical treatment means that UNB will have relatively less utility where surgeons have modified their practice to omission of ALND for minimal nodal metastatic disease.

Performance characteristics of ultrasound-guided fine-needle aspiration of axillary lymph nodes for metastatic breast cancer employing rapid on-site evaluation of adequacy: analysis of 136 cases and review of the literature

BACKGROUND

It has been demonstrated that axillary ultrasound-guided fine-needle aspiration (US-FNA) has excellent positive predictive value for the axillary lymph node status of patients with breast cancer before surgery or neoadjuvant therapy and, thus, can obviate the need for sentinel lymph node biopsy in FNA-positive patients. However, US-FNA has only moderate sensitivity, in part because of the collection of nondiagnostic or equivocal specimens. Rapid on-site evaluation for adequacy (ROSE) can improve definitive diagnosis rates but has not been well characterized in this setting.

METHODS

One hundred thirty-three patients with breast carcinoma were identified who underwent 136 US-FNAs of axillary lymph nodes, all with ROSE, and the results were correlated with the diagnosis on a subsequent surgical procedure.

RESULTS

The adequacy rate was 95.6% (130 of 136 FNAs), and a definitive diagnosis was made in 91.2% (124 of 136 FNAs). Among definite diagnoses, sensitivity was 75%, specificity was 100%, the positive predictive value was 100%, and the negative predictive value was 79%. Sources of false-negative and potential false-positive diagnoses were evaluated among these cases and in the literature.

CONCLUSIONS

Small metastasis size is the most common cause of false-negative results, whereas interpretation errors by pathologists are quite rare. ROSE appears to improve adequacy and definitive diagnosis rates and, thus, can more accurately triage patients to appropriate care.

Applicability of two different validated models to predict axillary non-sentinel lymph node status by sentinel node biopsy in a single Italian center

BACKGROUND

The necessity of complete axillary lymph node dissection (CALND) after sentinel lymph node biopsy (SLNB) for women with sentinel lymph node metastases is a matter of debate because non-sentinel lymph nodes after CALND contain no further metastases in about 50 % of cases. Our study aims to determine the applicability in our setting of two different validated nomograms to predict axillary lymph node status after SLNB.

METHODS

We collected data about all women who underwent SLNB in our Department of Surgery from 2007 to 2010, focusing on tumor, patient, and breast characteristics. Data was analyzed by R (version 2.15.2); p < 0.05 was considered significant.

RESULTS

Among 511 women who underwent SLNB, 126 received CALND due to sentinel lymph node metastasis, and 73.0 % of these had no further metastatic non-sentinel lymph node. The area under the receiver operating characteristic (ROC) curves for the Memorial Sloan-Kettering Cancer Center (MSKCC) nomogram and the Tenon score were 78.5 % (95 % CI 70.1-86.8 %) and 77.0 % (95 % CI 67.9-86.0 %) (p = 0.678), respectively.

CONCLUSIONS

Both the MSKCC nomogram and the Tenon score were predictive for the axillary non-sentinel lymph node status by SLNB. The MSKCC nomogram was the more accurate of the two and the Tenon score was the easier one to apply.

The role of preoperative axillary lymph node fine needle aspiration in locoregional staging of breast cancer

Preoperative ultrasound-guided fine needle aspiration cytology (UG-FNAC) of axillary lymph nodes in breast cancer emerged after the onset of the surgical sentinel node (SN) procedure. Today it is established as one of the preoperative routine procedures in patients with a cytological or histological confirmation or strong suspicion of breast carcinoma, the interest being that a positive UG-FNAC allows to avoid SLN biopsy or two-stage surgical procedure. Our article reviews the recent data in the literature regarding the diagnostic accuracy of lymph node FNAC in breast cancer staging, and presents the experience of the Breast Diagnostic Centre of Oslo University Hospital Ullevaal, Norway, in this context. Nowadays, UG-FNAC is indicated whenever the breast radiologist finds a suspicious or otherwise abnormal axillary lymph node, regardless of the size of the primary tumour. UG-FNAC is a cost effective and safe method. A diagnosis of metastatic malignancy has a very high accuracy and false-positives are virtually non-existent. False-negatives do occur, especially in lymph nodes with partial involvement as micrometastases and isolated tumor cells (ITC), and recent recommendations advocate that in these particular situations the axillary dissection is not necessary.

Ultrasound and fine needle aspiration cytology of axillary lymph nodes in breast cancer. To do or not to do?

AIM

The purpose of our study was to evaluate the accuracy of axillary ultrasound and ultrasound-guided fine needle aspiration cytology (FNAC) in the preoperative diagnosis of axillary metastases.

METHODS

Between 2004 and 2009, 1132 female patients were evaluated and treated in our clinic for histologically proven breast carcinoma. Preoperative axillary ultrasound with subsequent FNAC in case of suspicious lymph nodes was performed in 1150 axillae (18 bilateral breast carcinomas). We analyzed the results of axillary ultrasound and FNAC retrospectively. Pathological node status was used as the reference standard (based on axillary dissection or sentinel node biopsy).

RESULTS

Axillary ultrasound showed suspicious lymph nodes in 327 axillae (28.4%). FNAC showed axillary metastases in 107 of these 327 axillae. Final histological analysis confirmed 106 metastases (one false positive). Histological analysis showed metastatic disease in 429 of 1150 axillae (37.3%). Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of axillary ultrasound alone were 43.8% (188/429), 80.7% (582/721), 57.5% (188/327) and 70.7% (582/823), respectively. When combining axillary ultrasound with FNAC of suspicious lymph nodes, sensitivity was 24.7% (106/429), specificity was 99.9% (720/721), PPV was 99.1% (106/107) and NPV was 69.0% (720/1043).

CONCLUSIONS

106/429 (24.7%) Node-positive axillae were identified by ultrasound-guided FNAC and spared unnecessary sentinel node biopsy. Unfortunately, the percentage of false negative results of ultrasound-guided FNAC (28.1%, 323/1150) was very high.

Preoperative staging of the axilla in women with invasive breast cancer

SUMMARY Axillary node status is an important prognostic factor in invasive breast cancer. Axillary lymph node dissection (ALND) was previously the primary management approach; however, sentinel node biopsy (SNB) has largely replaced ALND, which is currently predominantly used to manage SNB-positive patients. Preoperative imaging-based staging of the axilla (with ultrasound, PET or MRI) for the detection of suspicious nodes has been applied to inform and potentially streamline axillary surgical management. Ultrasound with ultrasound-guided needle biopsy is the most accurate imaging-based strategy and has been shown to have clinical utility for preoperative axillary staging. A meta-analysis has reported a median sensitivity of 79.4% and a specificity of 100% for ultrasound-based staging, and estimated that ultrasound-guided needle biopsy triages 55.2% of women with metastatic axillary nodes (or a median of 17.7% of patients) directly to ALND, thereby avoiding unnecessary SNB. Although ultrasound-based staging has had a role in preoperative axillary assessment, it appears likely that this role will be shaped in the future by the ongoing evolution in surgical management of the axilla, particularly in light of recent evidence that challenges the need for ALND in some SNB-positive patients.

Rapid on-site evaluation of axillary fine-needle aspiration cytology in breast cancer

BACKGROUND

Axillary ultrasonography (AUS) and fine-needle aspiration cytology (FNAC) can establish axillary lymph node status before surgery, although this technique is hampered by poor adequacy rates. To achieve consistently high rates of FNAC adequacy, rapid on-site evaluation (ROSE) of FNAC samples was introduced.

METHODS

This single-centre, retrospective observational study of patients with newly diagnosed breast cancer undergoing preoperative AUS and FNAC between February 2008 and November 2010 examined the effect of the introduction of ROSE.

RESULTS

A total of 381 patients were included. AUS revealed 152 axillae with suspicious radiological features. FNAC was positive for malignant cells in 75 (49·3 per cent) of 152 samples. Sentinel lymph node mapping was avoided in 75 patients, representing 19·7 per cent of the entire study population. Adequacy rates increased from 78 per cent to 96 per cent following the introduction of ROSE (P = 0·001). The overall sensitivity and specificity of AUS and FNAC was 80·6 and 100 per cent respectively. A lymph node diameter equal to or larger than 10 mm and extranodal extension were significantly associated with positive FNAC (P < 0·001 and P = 0·012 respectively). Maximum lymph node diameter of at least 10 mm was an independent predictor of positive FNAC (odds ratio 11·2, 95 per cent confidence interval 3·32 to 37·76; P < 0·001).

CONCLUSION

AUS with FNAC provided accurate preoperative staging of the axilla for metastatic breast disease and avoided unnecessary sentinel lymph node mapping. The introduction of ROSE ensured the efficiency of AUS and FNAC.

Preoperative ultrasound-guided needle biopsy of axillary nodes in invasive breast cancer: meta-analysis of its accuracy and utility in staging the axilla

OBJECTIVE

Systematic evidence synthesis of ultrasound-guided needle biopsy (UNB) of axillary nodes in breast cancer.

SUMMARY BACKGROUND DATA

Women affected by invasive breast cancer undergo initial staging with sentinel node biopsy, generally progressing to axillary node dissection (AND) if metastases are found. Preoperative UNB can potentially identify and triage women with node metastases directly to AND.

METHODS

Review and meta-analysis of studies reporting UNB accuracy: we estimated sensitivity, specificity, and PPV, using bivariate random-effects models and examined the effect of covariates; we calculated UNB utility (effect on axillary surgery).

RESULTS

Thirty-one studies provided 2874 UNB data from 6166 subjects (median proportion with metastatic nodes 47.2%; IQR 39.5%, 61.2%). Modeled estimates for UNB were: sensitivity 79.6% (95% confidence intervals [CI] 74.1-84.2), specificity 98.3% (95%CI 97.2-99.0), PPV 97.1% (95%CI 95.2-98.3); median UNB insufficiency was 4.1% (IQR0%-10.9%). UNB sensitivity increased with increasing ultrasound sensitivity, and was higher in studies performing UNB for "suspicious" than for "visible" nodes. Specificity was higher in studies of consecutive (vs. selected) subjects, in studies reporting ultrasound data, and in more recent studies. Median proportion of women triaged directly to AND (attributed to UNB) was 19.8% (IQR11.6%-28.1%) or 17.7% (IQR11.6%-27.1%) if restricted to clinically node-negative series. Median proportion of women with metastatic axillary nodes potentially triaged to AND was 55.2% (IQR41.8%-68.2%) and was higher (65.6%; IQR48.9%-69.7%) in the subgroup of studies with median tumor size ≥21 mm.

CONCLUSIONS

Preoperative UNB of the axilla is accurate for initial staging of women with invasive breast cancer. Meta-analysis indicates that UNB provides better utility in women with average or higher underlying risk of node metastases.