RESUMO
BACKGROUND: Due to recent changes in breast cancer treatment strategy, significantly more patients are treated with neoadjuvant systemic therapy (NST). Radiological methods do not precisely determine axillary lymph node status, with up to 30% of patients being misdiagnosed. Hence, supplementary methods for lymph node status assessment are needed. This study aimed to apply and evaluate machine learning models on clinicopathological data, with a focus on patients meeting NST criteria, for lymph node metastasis prediction. METHODS: From the total breast cancer patient data (n = 8381), 719 patients were identified as eligible for NST. Machine learning models were applied for the NST-criteria group and the total study population. Model explainability was obtained by calculating Shapley values. RESULTS: In the NST-criteria group, random forest achieved the highest performance (AUC: 0.793 [0.713, 0.865]), while in the total study population, XGBoost performed the best (AUC: 0.762 [0.726, 0.795]). Shapley values identified tumor size, Ki-67, and patient age as the most important predictors. CONCLUSION: Tree-based models achieve a good performance in assessing lymph node status. Such models can lead to more accurate disease stage prediction and consecutively better treatment selection, especially for NST patients where radiological and clinical findings are often the only way of lymph node assessment.
RESUMO
BACKGROUND: Our objective was to assess the effects of COVID-19 antiepidemic measures and subsequent changes in the function of the health care system on the number of newly diagnosed breast cancers in the Republic of Croatia. SUBJECTS, MATERIALS, AND METHODS: We performed a retrospective, population- and registry-based study during 2020. The comparator was the number of patients newly diagnosed with breast cancer during 2017, 2018, and 2019. The outcome was the change in number of newly diagnosed breast cancer cases. RESULTS: The average monthly percent change after the initial lockdown measures were introduced was -11.0% (95% confidence interval - 22.0% to 1.5%), resulting in a 24% reduction of the newly diagnosed breast cancer cases in Croatia during April, May, and June compared with the same period of 2019. However, during 2020, only 1% fewer new cases were detected than in 2019, or 6% fewer than what would be expected based on the linear trend during 2017-2019. CONCLUSION: It seems that national health care system measures for controlling the spread of COVID-19 had a detrimental effect on the number of newly diagnosed breast cancer cases in Croatia during the first lockdown. As it is not plausible to expect an epidemiological change to occur at the same time, this may result in later diagnosis, later initiation of treatment, and less favorable outcomes in the future. However, the effect weakened after the first lockdown and COVID-19 control measures were relaxed, and it has not reoccurred during the second COVID-19 wave. Although the COVID-19 lockdown affected the number of newly diagnosed breast cancers, the oncology health care system has shown resilience and compensated for these effects by the end of 2020. IMPLICATIONS FOR PRACTICE: It is possible to compensate for the adverse effects of COVID-19 pandemic control measures on breast cancer diagnosis relatively promptly, and it is of crucial importance to do it as soon as possible. Moreover, as shown by this study's results on the number of newly diagnosed breast cancer cases during the second wave of the pandemic, these adverse effects are preventable to a non-negligible extent.
