Explore the vast range of titles available.

Cox Proportional Hazards (CPH) analysis is the standard for survival analysis in oncology. Recently, several machine learning (ML) techniques have been adapted for this task. Although they have shown to yield results at least as good as classical methods, they are often disregarded because of their lack of transparency and little to no explainability, which are key for their adoption in clinical settings. In this paper, we used data from the Netherlands Cancer Registry of 36,658 non-metastatic breast cancer patients to compare the performance of CPH with ML techniques (Random Survival Forests, Survival Support Vector Machines, and Extreme Gradient Boosting [XGB]) in predicting survival using the c -index. We demonstrated that in our dataset, ML-based models can perform at least as good as the classical CPH regression ( c -index  ∼ 0.63 ), and in the case of XGB even better ( c -index  ∼ 0.73 ). Furthermore, we used Shapley Additive Explanation (SHAP) values to explain the models’ predictions. We concluded that the difference in performance can be attributed to XGB’s ability to model nonlinearities and complex interactions. We also investigated the impact of specific features on the models’ predictions as well as their corresponding insights. Lastly, we showed that explainable ML can generate explicit knowledge of how models make their predictions, which is crucial in increasing the trust and adoption of innovative ML techniques in oncology and healthcare overall.

[...]hospitals might have postponed diagnostic evaluation or have longer turnaround times for diagnostic evaluation because many hospital-based resources are being allocated to tackle COVID-19. [...]national screening programmes for breast, colorectal, and cervical cancer are temporarily halted as of March 16, 2020, to alleviate the demand on the health-care system due to COVID-19. [...]misconceptions were eliminated about a heightened risk of contracting COVID-19 in a health-care setting because of inadequate policies for infection control at the institutional level and resource constraints in the delivery of essential oncological care.

Investigators of registry-based studies report improved survival for breast-conserving surgery plus radiotherapy compared with mastectomy in early breast cancer. As these studies did not present long-term overall and breast cancer-specific survival, the effect of breast-conserving surgery plus radiotherapy might be overestimated. In this study, we aimed to evaluate 10 year overall and breast cancer-specific survival after breast-conserving surgery plus radiotherapy compared with mastectomy in Dutch women with early breast cancer. In this population-based study, we selected all women from the Netherlands Cancer Registry diagnosed with primary, invasive, stage T1–2, N0–1, M0 breast cancer between Jan 1, 2000, and Dec 31, 2004, given either breast-conserving surgery plus radiotherapy or mastectomy, irrespective of axillary staging or dissection or use of adjuvant systemic therapy. Primary outcomes were 10 year overall survival in the entire cohort and breast cancer-specific survival in a representative subcohort of patients diagnosed in 2003 with characteristics similar to the entire cohort. We estimated breast cancer-specific survival by calculating distant metastasis-free and relative survival for every tumour and nodal category. We did multivariable Cox proportional hazard analysis to estimate hazard ratios (HRs) for overall and distant metastasis-free survival. We estimated relative survival by calculating excess mortality ratios using life tables of the general population. We did multiple imputation to account for missing data. Of the 37 207 patients included in this study, 21 734 (58%) received breast-conserving surgery plus radiotherapy and 15 473 (42%) received mastectomy. The 2003 representative subcohort consisted of 7552 (20%) patients, of whom 4647 (62%) received breast-conserving surgery plus radiotherapy and 2905 (38%) received mastectomy. For both unadjusted and adjusted analysis accounting for various confounding factors, breast-conserving surgery plus radiotherapy was significantly associated with improved 10 year overall survival in the whole cohort overall compared with mastectomy (HR 0·51 [95% CI 0·49–0·53]; p<0·0001; adjusted HR 0·81 [0·78–0·85]; p<0·0001), and this improvement remained significant for all subgroups of different T and N stages of breast cancer. After adjustment for confounding variables, breast-conserving surgery plus radiotherapy did not significantly improve 10 year distant metastasis-free survival in the 2003 cohort overall compared with mastectomy (adjusted HR 0·88 [0·77–1·01]; p=0·07), but did in the T1N0 subgroup (adjusted 0·74 [0·58–0·94]; p=0·014). Breast-conserving surgery plus radiotherapy did significantly improve 10 year relative survival in the 2003 cohort overall (adjusted 0·76 [0·64–0·91]; p=0·003) and in the T1N0 subgroup (adjusted 0·60 [0·42–0·85]; p=0·004) compared with mastectomy. Adjusting for confounding variables, breast-conserving surgery plus radiotherapy showed improved 10 year overall and relative survival compared with mastectomy in early breast cancer, but 10 year distant metastasis-free survival was improved with breast-conserving surgery plus radiotherapy compared with mastectomy in the T1N0 subgroup only, indicating a possible role of confounding by severity. These results suggest that breast-conserving surgery plus radiotherapy is at least equivalent to mastectomy with respect to overall survival and may influence treatment decision making for patients with early breast cancer. None.

Summary Background Genetic testing in breast cancer patients is important for the patient’s local and systemic treatment choices and follow-up, as well as for their family members. Not all eligible patients currently undergo genetic testing and disparities persist in genetic testing uptake. It is unknown on the large scale whether pre-test counselling by non-genetic healthcare professionals (HCPs)–mainstream genetic testing (MGT) – improves overall genetic testing uptake and reduces disparities. We examined the impact of MGT on germline genetic testing uptake in general and in subgroups of socioeconomic status (SES) in particular. Methods In this retrospective nationwide cohort study, we selected all breast cancer patients from the Netherlands Cancer Registry who were eligible for genetic testing according to patient and tumour characteristics under the Dutch guidelines and who were diagnosed between 1-Jan-2017 and 31-Dec-2022. The primary outcome was genetic testing uptake. The influence of MGT and SES on overall uptake and uptake across different SES levels was evaluated using chi-squared tests and multivariable logistic regression analyses. Results A total of 12,071 breast cancer patients were included. Overall genetic testing uptake was 67%: 78% for MGT versus 63% in referral to a genetics department (RGD) ( p  < 0.001) with significantly higher odds of receiving genetic testing for MGT versus RGD (OR 2.48, 95% CI 2.14–2.87). Patients with low SES showed significantly lower odds of receiving genetic testing compared to those with a high SES (OR 0.71, 95% CI 0.61–0.83). In MGT, no significant difference was found between low and high SES in the likelihood of receiving genetic testing (OR 0.75, 95% CI 0.50–1.13). Conclusions MGT significantly increases genetic testing uptake among all eligible patients and across all SES subgroups, strongly encouraging further implementation of MGT. Educating HCPs about current disparities in genetic testing is essential to improve health equity in breast cancer care.

Oncological care was largely derailed due to the reprioritisation of health care services to handle the initial surge of COVID-19 patients adequately. Cancer screening programmes were no exception in this reprioritisation. They were temporarily halted in the Netherlands (1) to alleviate the pressure on health care services overwhelmed by the upsurge of COVID-19 patients, (2) to reallocate staff and personal protective equipment to support critical COVID-19 care, and (3) to mitigate the spread of COVID-19. Utilising data from the Netherlands Cancer Registry on provisional cancer diagnoses between 6 January 2020 and 4 October 2020, we assessed the impact of the temporary halt of national population screening programmes on the diagnosis of breast and colorectal cancer in the Netherlands. A dynamic harmonic regression model with ARIMA error components was applied to assess the observed versus expected number of cancer diagnoses per calendar week. Fewer diagnoses of breast and colorectal cancer were objectified amid the early stages of the initial COVID-19 outbreak in the Netherlands. This effect was most pronounced among the age groups eligible for cancer screening programmes, especially in breast cancer (age group 50–74 years). Encouragingly enough, the observed number of diagnoses ultimately reached and virtually remained at the level of the expected values. This finding, which emerged earlier in age groups not invited for cancer screening programmes, comes on account of the decreased demand for critical COVID-19 care since early April 2020, which, in turn, paved the way forward to resume screening programmes and a broad range of non-critical health care services, albeit with limited operating and workforce capacity. Collectively, transient changes in health-seeking behaviour, referral practices, and cancer screening programmes amid the early stages of the initial COVID-19 epidemic in the Netherlands conjointly acted as an accelerant for fewer breast and colorectal cancer diagnoses in age groups eligible for cancer screening programmes. Forthcoming research is warranted to assess whether the decreased diagnostic scrutiny of cancer during the COVID-19 pandemic resulted in stage migration and altered clinical management, as well as poorer outcomes.

Purpose We performed a systematic review to document the spatial location of local recurrences (LR) after mastectomy. Methods A PubMed search was conducted in August 2019 for the following terms: breast [Title/Abstract] AND cancer [Title/Abstract] AND recurrence [Title/Abstract] AND mastectomy [Title/Abstract]. The search was filtered for English language. Exclusion criteria included studies that did not specify the LR location or studies reporting LR associated with inflammatory breast cancer, or other breast cancers such as phyllodes tumours, lymphoma or associated with sarcoma/angiosarcoma. Results A total of 3922 titles were identified, of which 21 publications were eligible for inclusion in the final analysis. A total of 6901 mastectomy patients were included (range 25–1694). The mean LR proportion was 3.5%. Among the total of 351 LR lesions, 81.8% were in the subcutaneous tissue and the skin, while 16% were pectoral muscle recurrences. Conclusion Local recurrences are mostly located within the subcutaneous tissue and the skin, assumed to result from unrecognized/subclinical tumour foci left behind after mastectomy, surgical implantation of tumour cells in the wound/scar and/or tumour emboli within the subcutaneous lymphatics. Pectoral muscle recurrences are less frequent and may be attributed to residual disease along the posterior surgical margin and/or lymphatic involvement.

To systematically review the currently available prediction models that may support treatment decision-making in breast cancer. Literature was systematically searched to identify studies reporting on development of prediction models aiming to support breast cancer treatment decision-making, published between January 2010 and December 2020. Quality and risk of bias were assessed using the Prediction model Risk Of Bias (ROB) Assessment Tool (PROBAST). After screening 20,460 studies, 534 studies were included, reporting on 922 models. The 922 models predicted: mortality (n = 417 45%), recurrence (n = 217, 24%), lymph node involvement (n = 141, 15%), adverse events (n = 58, 6%), treatment response (n = 56, 6%), or other outcomes (n = 33, 4%). In total, 285 models (31%) lacked a complete description of the final model and could not be applied to new patients. Most models (n = 878, 95%) were considered to contain high ROB. A substantial overlap in predictor variables and outcomes between the models was observed. Most models were not reported according to established reporting guidelines or showed methodological flaws during the development and/or validation of the model. Further development of prediction models with thorough quality and validity assessment is an essential first step for future clinical application.

Background To assess proportions of metastatic recurrence in women initially diagnosed with non-metastatic breast cancer by stage at diagnosis, breast cancer subtype, calendar period and age. Methods A systematic search of MEDLINE and Web of Science databases (January 2010–12 May 2022) was conducted. Studies reporting the proportion of distant metastatic recurrence in women with non-metastatic breast cancer were identified and outcomes and characteristics were extracted. Risk of bias was assessed independently by two reviewers. Random-effects meta-analyses of proportions were used to calculate pooled estimates and 95% confidence intervals (CIs). Results 193 studies covering over 280,000 patients were included in the main analysis. Pooled proportions of metastatic recurrence increased with longer median follow-up time from 12.2% (95% CI 10.5–14.0%) at 1–4 years post diagnosis, 14.3% (95% CI 12.9–15.7%) at 5–9 years to 23.3% (95% CI 20.1–26.8) at 10 years or more. Regional variation was observed with pooled estimates ranging from 11.0% (95% CI 8.5–13.7%) in Europe to 26.4% (95% CI 16.7–37.4%) in Africa (1–4 years follow-up). Proportions of recurrence were higher in studies with diagnosis before 2000 (22.2%, 95% CI 15.1–30.3) compared to studies with diagnosis from 2000 onwards (12.8%, 95% CI 11.7–14.0). At 1–4 years median follow-up, pooled proportions of metastatic recurrence were higher in women with hormone receptor negative (15.2%, 95% CI 12.0–18.7%) compared with receptor positive disease (9.6%, 95% CI 6.2–13.6%) and in women with locally advanced (33.2%, 95% CI 24.7–42.3%) relative to early disease at initial diagnosis (4.8%, 95% CI 2.5–7.8%). Proportions were higher in those under 50 years compared with 70+ years, 18.6% (95% CI 15.9–21.4%) versus 13.3% (95% CI 9.2, 18.0%), respectively. Heterogeneity was high in all meta-analyses and results should be interpreted with caution. Conclusions Higher proportions of metastatic recurrence in patients initially diagnosed at an advanced stage and in earlier calendar period emphasises the importance of early detection and treatment advancements. As the global number of breast cancer survivors increases, research and health policy efforts should be directed towards timely diagnosis and access to effective treatments and care. Study registration : PROSPERO CRD42022314500.

PurposeTo extend the functionality of the existing INFLUENCE nomogram for locoregional recurrence (LRR) of breast cancer toward the prediction of secondary primary tumors (SP) and distant metastases (DM) using updated follow-up data and the best suitable statistical approaches.MethodsData on women diagnosed with non-metastatic invasive breast cancer were derived from the Netherlands Cancer Registry (n = 13,494). To provide flexible time-dependent individual risk predictions for LRR, SP, and DM, three statistical approaches were assessed; a Cox proportional hazard approach (COX), a parametric spline approach (PAR), and a random survival forest (RSF). These approaches were evaluated on their discrimination using the Area Under the Curve (AUC) statistic and on calibration using the Integrated Calibration Index (ICI). To correct for optimism, the performance measures were assessed by drawing 200 bootstrap samples.ResultsAge, tumor grade, pT, pN, multifocality, type of surgery, hormonal receptor status, HER2-status, and adjuvant therapy were included as predictors. While all three approaches showed adequate calibration, the RSF approach offers the best optimism-corrected 5-year AUC for LRR (0.75, 95%CI: 0.74–0.76) and SP (0.67, 95%CI: 0.65–0.68). For the prediction of DM, all three approaches showed equivalent discrimination (5-year AUC: 0.77–0.78), while COX seems to have an advantage concerning calibration (ICI < 0.01). Finally, an online calculator of INFLUENCE 2.0 was created.ConclusionsINFLUENCE 2.0 is a flexible model to predict time-dependent individual risks of LRR, SP and DM at a 5-year scale; it can support clinical decision-making regarding personalized follow-up strategies for curatively treated non-metastatic breast cancer patients.

Purpose For the implementation of personalised surveillance, it is important to create more awareness among HCPs with regard to the risk for locoregional recurrences (LRRs). The aim of this study is to evaluate the current awareness and estimations of individual risks for LRRs after completion of primary treatment for breast cancer among health care professionals (HCPs) in the Netherlands, without using any prediction tools. Methods A cross-sectional survey was performed among 60 HCPs working in breast cancer care in seven Dutch hospitals and 25 general practitioners (GPs). The survey consisted of eleven realistic surgically treated breast cancer cases. HCPs were asked to estimate the 5-year risk for LRRs for each case, which was compared to the estimations by the INFLUENCE-nomogram using one-sample Wilcoxon tests. Differences in estimations between HCPs with different specialities were determined using Kruskal–Wallis tests and Dunn tests. Results HCPs tended to structurally overestimate the 5-year risk for LRR on each case. Average overestimations ranged from 4.8 to 26.1%. Groups of HCPs with varying specialities differed significantly in risk estimations. GPs tended to overestimate the risk for LRRs on average the most (15.0%) and medical oncologists had the lowest average overestimation (2.7%). Conclusions It is important to create more awareness of the risk for LRRs, which is a pre-requisite for the implementation of personalised surveillance after breast cancer. Besides education for HCPs, the use of prediction models such as the INFLUENCE-nomogram can support in estimating an objective estimate of each individual patient’s risk.