Can machine learning be used to reduce overtreatment of the axilla in breast cancer? Results from a retrospective cohort study (Preprint)

BACKGROUND Patients with early breast cancer undergoing primary surgery who have low axillary nodal burden can safely forego axillary node clearance (ANC). However, routine use of axillary ultrasound (AUS) leads to 43% of patients in this group having ANC unnecessarily following a positive AUS. The intersection of machine learning with medicine can provide innovative ways to understand specific risk within large patient data sets, but this has not yet been trialled in the arena of axillary node management in breast cancer. OBJECTIVE To assess if machine learning techniques could be used to improve pre-operative identification of patients with low and high axillary metastatic burden. METHODS A single-centre retrospective analysis was performed on patients with breast cancer who had a preoperative axillary ultrasound, and the specificity and sensitivity of AUS were calculated. Machine learning and standard statistical methods were applied to the data to see if, when used preoperatively, they could have improved the accuracy of AUS to better discern between high and low axillary burden. RESULTS The study included 459 patients; 31% (n=142) had a positive AUS, and, among this group, 62% (n=88) had two or fewer macrometastatic nodes at ANC. When applied to the dataset, logistic regression outperformed AUS and machine learning methods with a specificity of 0.950, correctly identifying 66 patients in this group who had been incorrectly classed as having high axillary burden by AUS alone. Of all the methods, the artificial neural network had the highest accuracy (0.919). Interestingly, AUS had the highest sensitivity of all methods (0.777), underlining its utility in this setting. CONCLUSIONS Machine learning greatly improves identification of the important subgroup of patients with no palpable axillary disease, positive ultrasound, and more than two metastatically involved nodes. A negative ultrasound in patients with no palpable lymphadenopathy is highly indicative of low burden and it is unclear if sentinel node biopsy adds value in this situation. CLINICALTRIAL n/a

Predictors of sentinel lymph node metastasis in Chinese women with clinical T1-T2 N0 breast cancer and a normal axillary ultrasound

Background The clinicopathological predictors of sentinel lymph node (SLN) metastasis in clinical T1-T2 N0 (cT1-T2 N0) patients with a normal axillary ultrasound (AUS) are unclear. Purpose To assess the association between clinicopathological characteristics of a primary tumor and SLN metastasis in cT1-T2 N0 patients with a normal AUS. Material and Methods Patients who were diagnosed with cT1-T2 N0 invasive breast cancer and who obtained normal AUS results between October 2016 and September 2018 in a single hospital were included. Clinicopathological data were collected to explore the predictors of SLN metastasis using a multivariate logistic regression model. Results SLN metastasis occurred in 26 patients (18.4%) among 141 AUS-normal patients, of which 24 cases (17.0%) had one or two nodal involvements. In the univariate analysis, tumor location, estrogen receptor (ER) status, progesterone receptor (PR) status, and lymphovascular invasion (LVI) were significantly associated with SLN metastasis ( P < 0.05). The multivariate analysis showed that tumor location in the upper outer quadrant (odds ratio [OR] = 4.49, 95% confidence interval [CI] = 1.63–12.37; P = 0.004), positive PR status (OR = 13.35, 95% CI = 1.60–111.39; P = 0.017), and positive LVI (OR = 8.66, 95% CI = 2.20–34.18; P = 0.002) were independent high-risk factors for SLN metastasis. The area under the receiver operating characteristic curve of the regression model was 0.787 (95% CI = 0.694–0.881; P < 0.001). Conclusion Tumor location in the upper outer quadrant, positive PR, and LVI status were found to be significantly high-risk factors for SLN metastasis among cT1-T2 N0 breast cancer patients with a normal AUS result.

Precautions for breast ultrasound examination following COVID-19 vaccination

Background: Coronavirus disease 2019 (COVID-19) vaccine-induced lymphadenopathy is a critical side effect that should be a concern to clinicians, patients, radiologists, and oncologists. Vaccine-induced lymphadenopathy causes a diagnostic dilemma, especially for breast radiologists who examine both axillary regions during breast ultrasound examinations. Appropriate imaging guidelines are needed to manage vaccine-induced lymphadenopathy for patients undergoing screening examinations or symptomatic patients, including cancer patients.Current Concepts: For patients with axillary lymphadenopathy in the setting of recent ipsilateral vaccination, clinical follow-up is recommended. In other scenarios, short-term follow-up axillary ultrasound examinations are recommended if the clinical concerns persist for more than 6 weeks after vaccination. To mitigate the diagnostic dilemma of vaccine-induced lymphadenopathy, patients should schedule screening imaging examinations before the first vaccination or at least six weeks following the second vaccination. For clinicians and radiologists, documenting the patients’ vaccination status is critical to decreasing unnecessary follow-up imaging, biopsies, and patient’s anxiety.Discussion and Conclusion: Our proposal can help reduce patient anxiety, provider burden, and costs of unnecessary evaluation of enlarged lymph nodes in the setting of recent COVID-19 vaccination. Further, it can avoid delays in vaccination and breast cancer screening during the COVID-19 pandemic.

TP8.1.1 Does ultrasound quantification of suspicious axillary nodes correlate with pathological results?

Aims Fast-track axillary node clearance (ANC) leads to overtreatment of axilla. Improved quantification by axillary US (AUS) is suggested to avoid unnecessary ANC and proceed with ANC or SLNB based on the number of abnormal axillary nodes. This retrospective study was aimed to evaluate whether ANC can be omitted based on AUS quantification in patients with low axillary burden. Methods Retrospective data of breast cancer patients who underwent ANC following a positive pre-operative axillary nodal biopsy between 1 January 2017 and 31 December 2018 were included in this study. The patients who received neoadjuvant chemotherapy, those having ANC following positive SLNB and those with axillary recurrence were excluded. The histopathology results of ANC were correlated with axillary ultrasound findings. Results 45 patients underwent fast-track ANC following positive axillary core biopsy. On pre-operative AUS, 18 of these patients were reported to have a single abnormal node, while 8 had two abnormal nodes and 19 patients had multiple abnormal nodes. The comparison of the number of metastatic nodes following ANC, and the reported abnormal nodes on pre-operative AUS, showed that 57.3% of patients with 1 – 2 abnormal nodes on AUS had 3 or more metastatic nodes and 26.3% of patients with multiple abnormal nodes on AUS had 1 – 2 metastatic nodes following ANC. Conclusions The quantification of the axillary burden with pre-operative AUS does not correlate with the number of metastatic axillary nodes. The reported relevant axillary burden on AUS is not sufficiently specific to form the basis of omission of ANC.

Axillary ultrasound and fine-needle aspiration cytology to predict clinically relevant nodal burden in breast cancer patients

Background Axillary lymph node involvement is one important prognostic factor in breast cancer, but the way to access this information has been modified over the years. This study evaluated if axillary ultrasound (US) coupled with fine-needle aspiration cytology (FNAC) can accurately predict clinically relevant node metastasis in patients with breast cancer, and thus assist clinical decisions Methods This is a cross-sectional study with retrospective data collection of 241 individuals (239 women and 2 men) with unilateral operable breast cancer who were submitted to preoperative axillary assessment by physical exam, US and FNAC if suspicious nodes by imaging. We calculated sensitivity, specificity, and accuracy of the methods. We compared the patient's characteristics using chi-square test, parametrics and non-parametrics statistics according to the variable. Results The most sensible method was US (0.59; 95% CI, 0.50–0.69), and the most specific was US coupled with FNAC (0.97; 95% CI, 0.92–0.99). Only 2.7% of the patients with normal axillary US had more than 2 metastatic nodes in the axillary lymph node dissection, against 50% of the patients with suspicious lymph nodes in the US and positive FNAC. Conclusions Axillary US coupled with FNAC can sort patients who have a few metastatic nodes at most from those with heavy axillary burden and could be one more tool to initially evaluate patients and define treatment strategies.

Axillary ultrasound-guided approach for the brachial plexus in pig cadavers: A descriptive study

Effective multimodal analgesia techniques are required when pigs are used as models in orthopedic human research. Regional anesthesia is a widely used technique to provide perioperative analgesia in animals undergoing orthopedic surgery. The brachial plexus (BP) block is indicated to desensitize the forelimb in many species but has not been yet described in pigs. The main aims of this study were to develop an ultrasound-guided axillary approach for the BP and to evaluate injectate spread and nerve staining in pig cadavers. Eight fresh F1 cross breed German Large White and German Landrace male pig cadavers were enrolled. Two cadavers were used for anatomical dissection of the axillary space and to determine the disposition of the BP. Six cadavers were used to perform a bilateral axillary ultrasound-guided approach for the BP, and after injecting 0.3 ml/kg of a solution of 2% lidocaine and new methylene blue (L-NMB), these were dissected to determine injectate spread and nerve staining. Upon dissection, the BP was observed in all the cases surrounded by the axillary sheath and in close proximity to the axillary artery. Ultrasonographic scanning and guidance for the approach to the BP was feasible in all cadavers and upon dissection, all the nerves forming the BP were stained in all their quadrants and in all the cases. In conclusion, the injection of 0.3 ml/kg of L-NMB through an ultrasound-guided axillary approach to the BP was feasible and adequate to entirely stain the BP in all pig cadavers. Further clinical studies are required to evaluate the effectiveness of this technique in live animals.

Magnetic-Guided Axillary UltraSound (MagUS) Sentinel Lymph Node Biopsy and Mapping in Patients with Early Breast Cancer. A Phase 2, Single-Arm Prospective Clinical Trial

Lymph Node Dissection (SLND) is standard of care for diagnosing sentinel lymph node (SLN) status in patients with early breast cancer. Study aim was to determine whether the combination of Superparamagnetic iron oxide nanoparticles (SPIO) MRI-lymphography (MRI-LG) and a Magnetic-guided Axillary UltraSound (MagUS) with biopsy can allow for minimally invasive, axillary evaluation to de-escalate surgery. Patients were injected with 2 mL of SPIO and underwent MRI-LG for SN mapping. Thereafter MagUS and core needle biopsy (CNB) were performed. Patients planned for neoadjuvant treatment, the SLN was clipped and SLND was performed after neoadjuvant with the addition of isotope. During surgery, SLNs were controlled for signs of previous biopsy or clip. The primary endpoint was MagUS SLN detection rate, defined as successful SLN detection of at least one SLN of those retrieved in SLND. In 79 patients, 48 underwent upfront surgery, 12 received neoadjuvant and 19 had recurrent cancer. MagUS traced the SLN in all upfront and neoadjuvant cases, detecting all patients with macrometastases (n = 10). MagUS missed only one micrometastasis, outperforming baseline axillary ultrasound AUS (AUC: 0.950 vs. 0.508, p < 0.001) and showing no discordance to SLND (p = 1.000). MagUS provides the niche for minimally invasive axillary mapping that can reduce diagnostic surgery.

Axillary Ultrasound and Fine-Needle Aspiration Cytology to Predict Clinically Relevant Nodal Burden in Breast Cancer Patients .

BackgroundAxillary lymph node involvement is one important prognostic factor in breast cancer, but the way to access this information has modified over the years.This study evaluated if axillary ultrasound (US) coupled with fine needle aspiration cytology (FNAC) can better predict clinically relevant node metastasis than physical exam, in patients with breast cancer.MethodsThis is a cross-sectional study with retrospective data collection of 241 women with operable breast cancer who were submitted to preoperative axillary assessment by physical exam, US and FNAC if suspicious nodes by imaging. We calculated sensitivity, specificity, and accuracy of the methods. We compared the patient's characteristics using chi-square test, parametrics and non parametrics statistics according to the variable.ResultsThe most sensible method was US (0.59; 95% CI, 0.50-0.69) and the most specific was US coupled with FNAC (0.97; 95% CI, 0.92-0.99 ). Only 2,7% of the patients with normal axillary US had more than 2 metastatic nodes in the axillary lymph node dissection, against 50% of the patients with suspicious lymph nodes in the US and positive FNAC. ConclusionsWe found that axillary US and FNAC are more accurate than physical examination in detecting node metastasis in breast cancer patients. Moreover, axillary US coupled with FNAC can sort patients that have a few metastatic nodes at most from those with heavy axillary burden and could be one more tool to initially evaluate patients and define treatment´s strategies.