RESUMO
Background: The number of urgent referrals from primary care to specialist one stop breast clinics continues to rise beyond the capacity of the 2-week wait service. This study aims to use artificial intelligence (AI) to identify patients with new breast symptoms requiring a biopsy to identify those who should be prioritised for urgent breast clinic assessment. Methods: Data were collected retrospectively for patients attending one stop triple assessment breast clinic at Broomfield hospital between 1 June and 1 October 2021. PHP machine learning software was used to run AI on the data to identify patients who had a core biopsy in clinic. Results: A total of 794 cases were referred to one stop breast clinic for new breast symptoms-37 male (4.6%) and 757 female (95.3%). The average age of the patients included was 43.2 years. Five hundred thirty-six patients (67.5%) presented with a breast lump, 180 (22.7%) with breast pain, 61 (7.7%) with changes to shape or skin and 13 (1.6%) with a lump identified by their general practitioner. The patients who had a biopsy were of increased age [52.8 (SD 17.9) vs. 44.1 (SD 16.8), P<0.001], and had previous mammogram [n=21, (31.8%) vs. n=148 (20.3%), P 0.03], previous benign breast disease [n=9 (13.6%) vs. n=23 (3.1%), P<0.001], and increased use of HRT [n=13 (19.7%) vs. n=53 (6.4%), P<0.001]. The sensitivity and specificity of AI with neural network algorithms were 84% and 90%, respectively. Conclusion: AI was very effective at predicting the presenting symptoms that are likely to result in biopsy and can therefore be used to identify patients who need to be seen urgently in breast clinic.
RESUMO
Introduction: Increasing evidence suggests that de-escalation of axillary surgery is safe, without significantly impacting patient outcome. Obtaining positive lymph nodes at a sentinel lymph node biopsy (SNB) can guide decisions toward the requirement of axillary nodal clearance (ANC). However, methods to predict how many further nodes will be positive are not available. This study investigates the feasibility of predicting the likelihood of a negative ANC based on the ratio between positive nodes and the total number of lymph nodes excised at SNB. Methods: Retrospective data from January 2017 to March 2022 was collected from electronic medical records. Patients with oestrogen receptor (ER) positive and HER2 negative receptor disease were included in the study. ER-negative and HER2-positive disease was excluded, alongside patients who had chemotherapy before ANC. Results: Of 102 patients, 58.8% (n=60) had no macrometastasis at ANC. On average, 2.76 lymph nodes were removed at SNB. A higher SNB ratio of positive to total nodes [OR 11.09 (CI 95% 2.33-52.72), P=0.002] had a significant association with positive nodes during ANC. SNB ratio less than or equal to 0.33 (1/3) had a specificity of 79.2% in identifying cases that later had a negative completion ANC, with a 95.8% specificity of no further upgrade of nodal staging. Conclusion: A low SNB ratio of less than 0.33 (1/3) has a high specificity in excluding the upgradation of nodal staging on completion of ANC, with a false-negative rate of less than 5%. This may be used to identify patients with a low risk of axillary metastasis, who can avoid ANC.
