Enhancing clinical decision support with genomic tools in breast cancer: A Scottish perspective.

INTRODUCTION: The Oncotype DX Breast RS test has been adopted in Scotland and has been the subject of a large population-based study by a Scottish Consensus Group to assess the uptake of the recurrence score (RS), evaluate co-variates associated with the RS and to analyse the effect it may have had on clinical practice. MATERIALS & METHODS: Pan-Scotland study between August 2018-August 2021 evaluating 833 patients who had a RS test performed as part of their diagnostic pathway. Data was extracted retrospectively from electronic records and analysis conducted to describe change in chemotherapy administration (by direct comparison with conventional risk assessment tools), and univariate/multivariate analysis to assess relationship between covariates and the RS. RESULTS: Chemotherapy treatment was strongly influenced by the RS (p < 0.001). Only 30 % of patients received chemotherapy treatment in the intermediate and high risk PREDICT groups, where chemotherapy is considered. Additionally, 55.5 % of patients with a high risk PREDICT had a low RS and did not receive chemotherapy. There were 17 % of patients with a low risk PREDICT but high RS who received chemotherapy. Multivariate regression analysis showed the progesterone receptor Allred score (PR score) to be a strong independent predictor of the RS, with a negative PR score being associated with high RS (OR 4.49, p < 0.001). Increasing grade was also associated with high RS (OR 3.81, p < 0.001). Classic lobular pathology was associated with a low RS in comparison to other tumour pathology (p < 0.01). Nodal disease was associated with a lower RS (p = 0.012) on univariate analysis, with menopausal status (p = 0.43) not influencing the RS on univariate or multivariate analysis. CONCLUSIONS: Genomic assays offer the potential for risk-stratified decision making regarding the use of chemotherapy. They can help reduce unnecessary chemotherapy treatment and identify a subgroup of patients with more adverse genomic tumour biology. A recent publication by Health Improvement Scotland (HIS) has updated guidance on use of the RS test for NHS Scotland. It suggests to limit its use to the intermediate risk PREDICT group. Our study shows the impact of the RS test in the low and high risk PREDICT groups. The implementation across Scotland has resulted in a notable shift in practice, leading to a significant reduction in chemotherapy administration in the setting of high risk PREDICT scores returning low risk RS. There has also been utility for the test in the low risk PREDICT group to detect a small subgroup with a high RS. We have found the PR score to have a strong independent association with high risk RS. This finding was not evaluated by the key RS test papers, and the potential prognostic information provided by the PR score as a surrogate biomarker is an outstanding question that requires more research to validate.

Clinical Experience of Axillary Radiotherapy for Node-positive Breast Cancer.

AIMS: Patients with breast cancer who have positive lymph nodes are currently recommended axillary node clearance (ANC) or regional nodal irradiation (RNI). ANC is associated with complications such as lymphoedema, brachial plexopathy and shoulder stiffness. The AMAROS Group showed RNI to be non-inferior to ANC with regards to survival and recurrence, and with a better quality of life. We conducted a large real-world population study to show our centre's experience with the use of RNI and to contribute to the current discussion around the management of node-positive breast cancer. MATERIALS AND METHODS: We evaluated patients who received RNI as opposed to ANC between 2006 and 2009 (n = 190). Patients had a range of cancer subtypes/grades. All had positive axillary disease, identified by axillary node sampling or sentinel lymph node biopsy. Systemic therapy was given as per standard protocol. Our data were compared with those of patients who had RNI (n = 681) in AMAROS. Patients were followed up retrospectively and overall survival, breast cancer-specific survival, distant metastasis-free survival, locoregional recurrence and toxicity were recorded, including lymphoedema, brachial plexopathy and shoulder stiffness. Survival analysis was performed on R via the Kaplan-Meier method. Univariate and multivariate analyses were also performed. Toxicity data were reported as percentages. Patients meeting POSNOC trial criteria (one to two positive sentinel lymph nodes, macrometastasis, receiving adjuvant chemotherapy) including if oestrogen receptor-positive (stratified POSNOC) were identified for subgroup analysis in the regression model. RESULTS: Locoregional recurrence was 3.16% versus AMAROS RNI of 1.82%. Overall survival was slightly lower in our population, but cancer-specific survival was higher than AMAROS. Lymphoedema rates were 5.8% versus AMAROS 11% in RNI and 23% in ANC arms, respectively. Brachial plexopathy was 1.6% and arm/shoulder stiffness 7.4%. AMAROS conducted a quality of life survey pertaining to arm/shoulder stiffness, mobility and function, which seemed to affect about 18% in the RNI arm. Univariate analysis revealed POSNOC status, especially if also oestrogen receptor-positive, to be a low risk group with hazard ratio 0.42 (0.20-0.83, P = 0.015). Extracapsular extension of lymph node metastasis was a poor prognostic factor; hazard ratio 4.39 (1.45-14.0, P = 0.009). CONCLUSION: We support the conclusion of the AMAROS trial with survival and recurrence following RNI being non-inferior to ANC, and with similarly favourable toxicity data. We support the continuing use of RNI as a treatment option for patients with node-positive breast cancer. Further research is required to answer the key questions regarding personalised management for node-positive breast cancer, with regards to de-escalation and also intensification for the patients exhibiting adverse tumour biology.

Association of need for cognition with judgments of height, weight, and body fat covariation.

559 college students, assessed for Need for Cognition, judged whether height, weight, and body fat were correlated using judgment probes that controlled for framing and conditional format. A principal components analysis of Need for Cognition scores identified two factors underlying the scale, which may be important in judgment outcome. In addition, judgments about correlations among height, weight, and body fat were similar to those of previous studies. Furthermore, the hypothesis that Need for Cognition would be related to the tendency to judge a negative correlation between height and body fat, i.e., a possible illusory correlation, was confirmed. Results are discussed relative to the Elaboration Likelihood Model of Persuasion.