Search

Purpose We analysed incidence, treatment, survival, occurrence of ductal carcinoma in situ (DCIS) and invasive breast cancer (IBC) after lobular carcinoma in situ (LCIS) in the Netherlands. Methods All women diagnosed with classic LCIS between 1989 and 2017 were identified from the Netherlands Cancer Registry. We calculated overall (OS), relative survival (RS) and cumulative incidence functions (CIF, accounting for competing risks) of mortality, DCIS and IBC. For IBC, standardised incidence ratios (SIR) of IBC were calculated. Analyses were stratified for surgical treatment. Results We included 1890 patients. Median age was 51 years. Median follow-up was 8.5 years. In 1989–2017, LCIS incidence increased from 41 to 124, surgical treatment decreased from 100% to 41.1 % – mostly BCS. 10-year OS and 20-year RS exceeded 90 % in all subgroups. Overall, 48 (2.5 %) and 270 (14.3 %) patients were diagnosed with DCIS and IBC. IBCs were mostly early-stage. After mastectomy, 13 of 14 IBCs presented contralaterally. In the other groups, 64.8–70.9 % of IBCs presented ipsilaterally, 34.5–53.9 % of these were lobular. The SIR of ipsilateral IBC was highest after no surgery (6.9, 95%CI:4.9–9.4), lowest after mastectomy (0.2, 95%CI:0.4–0.8). Conclusion LCIS incidence increased, surgical treatment decreased. The low mortality risks support consideration of active surveillance. However, the increased IBC incidence suggests careful monitoring., Highlights • The incidence of LCIS increased, while surgical treatment decreased over time. • 10- and 20-year relative survival rates exceeded 90 %, irrespective of surgery. • The risk of breast cancer was highest after no surgery, lowest after mastectomy. • Subsequent breast cancers were generally of low stage. • Active surveillance for LCIS is advised, with the current follow-up five years.

Background: Regional recurrence (RR), also known as lymph node recurrence, is an endpoint in several trials concerning reducing axillary treatment in cT1-2N0 breast cancer patients. The risk of RR may decrease with each subsequent event-free year, affecting the yield and consequently usefulness of long (er) follow-up. The aim of this study is to determine the risk of RR as a first event within five years after diagnosis in subtypes of breast cancer, conditional to being event-free for one, two, three and four years.Methods: From the Netherlands Cancer Registry, cT1-2N0 breast cancer patients diagnosed from 2005 to 2008 were analyzed. Subgroup analysis was performed for pT1-2N+(sn) patients. RR risk was calculated with Kaplan-Meier analysis. Conditional RR (assuming x event-free years) was determined by selecting patients without an event at x years, and calculating the remaining risk for RR within five years after diagnosis.Results: A total of 18,009 cT1-2N0 (all pN stages) breast cancer patients were included. RR occurred in 1.3% of cT1-2N0 and 1.5% of pT1-2N+(sn) patients. The risk of RR varied between subtypes; it was highest for triple negative tumors and lowest for ER + PR + Her2-and ER + Her2+ tumors. After event-free years, the risk of RR decreased subsequently in both groups and in all subtypes. After two event-free years, the risk of RR was 0.8%.Conclusion: The absolute yield of follow-up to detect RR beyond two years is low; for every 125 event-free patients, one RR can be expected until five years. This suggests that follow-up longer than two years is of limited value for detecting RR in both clinical and research setting.(c) 2020 Published by Elsevier Ltd.

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 $$c$$ c -index. We demonstrated that in our dataset, ML-based models can perform at least as good as the classical CPH regression ($$c$$ c -index $$\sim \,0.63$$ ∼ 0.63 ), and in the case of XGB even better ($$c$$ c -index $$\sim 0.73$$ ∼ 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.

Background Breast cancer follow-up aims at early detection of locoregional recurrences (LRR) and second primary breast cancer (SP). Predicting a patient’s time-dependent risk for such an event may improve follow-up strategies. Furthermore, many clinicians and patients are interested in the risk of distant metastasis (DM) as well, as it provides additional information on prognosis. The existing INFLUENCE nomogram predicts the yearly risk of LRR up to five years from diagnosis using logistic regression analyses. However, this model was based on patients diagnosed between 2003 and 2006 and lacked information on HER2 status, which is nowadays an important prognostic variable. In this population-based study, we aimed to improve the INFLUENCE model with more flexible predictions over time, more recent information and additional outcomes - risk of LRR, distant metastasis (DM) and SP - using different underlying models. Methods Patient-, tumor- and treatment-related characteristics of all female patients diagnosed with invasive adenocarcinoma of the breast in 2007, 2008 and 2012, of who active 5-year follow-up was performed, were selected from the Netherlands Cancer Registry. Models were developed for LRR, DM and SP as a first event separately, and for each three-monthly time interval to predict time-dependent risks as accurately as possible. Follow-up was calculated from date of definite surgery to date of event or last observation. We compared three different models to estimate the three outcomes: a Cox regression model, a flexible parametric spline model and a random survival forest (RF) model. To assess calibration of each outcome, three-monthly risk predictions during the 5-year follow-up period were compared to corresponding observed event rates. To assess discrimination, the areas under the curves (AUC) were calculated for each timeframe in the 5-year period and averaged to one combined measure (average 5-year AUC). To correct for optimism we performed bootstrapping on the entire cohort with 200 replicates and updated the calibration measures and AUCs accordingly. Results In total, 13,494 patients were included. LRR, DM and SP within five years were experienced by 2.8%, 6.3% and 3.1% of the patients, respectively. The following variables were included in the final models: age, tumor grade, tumor stage, nodal stage, multifocality, HER2 status, hormonal receptor status, type of surgery, chemotherapy, radiotherapy, hormonal therapy, anti-HER2 therapy. The optimism-corrected mean time-dependent prediction errors for individual risk predictions ranged between 0.12% and 0.46% for all three models and all three month timeframes. The average 5-year AUCs for LRR, SP and DM using the Cox model were 0.71, 0.63 and 0.77, respectively. The average 5-year AUCs for LRR, SP and DM using the parametric spline model were 0.71, 0.63 and 0.78, respectively. When using the RF model, the average 5-year AUCs were 0.74, 0.68 and 0.77 for LRR, SP and DM, respectively (Table). Conclusion Although differences were very small, the RF model had the best performance overall. This INFLUENCE 2.0 model provides us with updated information on the individual time-dependent risks of LRR, DM and SP within five years following surgery. This model is freely accessible at https://tinyurl.com/influence-2 and can be used as an instrument to improve follow-up strategies and to give patients better insights in their prognosis. Table. Optimism-corrected average AUCs over the 5-year follow-up periodModel typeLRDMSPCox regression0.710.770.63Flexible parametric spline0.710.780.63Random survival forest0.740.770.68 Citation Format: Sabine Siesling, Vinzenz Völkel, Tom Hueting, Teresa Draeger, Marissa C van Maaren, Luc JA Strobbe, Marjanka K Schmidt, Gabe S Sonke, Marjan van Hezewijk, Catharina GM Groothuis-Oudshoorn. Risk estimation of locoregional recurrence, distant metastasis and second primary breast cancer in early stage breast cancer patients: The INFLUENCE 2.0 nomogram [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-10.

Background: In breast cancer treatment decision-making, PREDICT is one of the most frequently used tools in clinical practice. The tool estimates the expected benefit of adjuvant systemic therapy on 5- and 10-year overall survival (OS). The model is based on an UK population and was previously validated on several other populations. OS is accurately predicted in these populations. However, PREDICT under- or overestimated OS in several predefined subgroups, which may lead to inadequate treatment decisions. A possible solution for this may be to use CancerMath. This model is based on a UK population and estimates not only OS, but also long-term breast cancer specific survival (BCSS). In this study, we aimed to validate the CancerMath model to determine its accuracy in several predefined subgroups based on patient-, tumor- and treatment-related characteristics. Methods: All women diagnosed with stage I-III primary invasive breast cancer in 2005 who received surgery were identified from the Netherlands Cancer Registry. The accuracy of CancerMath was evaluated by its calibration and discrimination overall, and in several predefined subgroups based on patient-, tumor- and treatment-related characteristics. For BCSS, we used distant metastasis followed by death as a proxy for breast cancer mortality. Calibration was assessed by comparing 5- and 10-year predicted and observed OS and BCSS using chi-squared tests. A difference >3% was considered to be clinically relevant, as the Dutch guidelines advise adjuvant systemic therapy in case of an at least 3-5% expected benefit on survival. Discrimination was assessed by area under the receiver operating characteristic (AUC) curves. Results: In total, 8,032 women were included. CancerMath underestimated 5- and 10-year OS by 2.2% and 1.9%, respectively. AUCs of 5- and 10-year OS were both 0.77. Divergence between predicted and observed OS was most pronounced in grade II tumors, patients without positive nodes, tumors 1.01-2.00 cm, hormonal receptor positive disease, and patients 60-69 years, in which OS was significantly underestimated. CancerMath underestimated 5- and 10-year BCSS by 0.5% and 0.6%, respectively. AUCs were 0.78 and 0.73, respectively. No significant difference was found in any subgroup. Conclusion: CancerMath predicts OS accurately for most patients with early breast cancer although outcomes should be interpreted with care in some subgroups. BCSS is predicted accurately in all subgroups. Importantly, a 3% difference may not be clinically relevant for someone with a predicted benefit of 12% and an observed benefit of 15%, as adjuvant systemic therapy will be indicated in both cases. However, in case these percentages are 1% and 4%, the indication for adjuvant systemic therapy will change. It is therefore crucial to interpret the over- or underestimations for individual patients with consideration of the initial predicted survival without adjuvant systemic therapy. In conclusion, CancerMath can reliably be used in (Dutch) clinical practice to estimate 5- and 10-year OS and BCSS, with a careful interpretation of OS in some subgroups. We encourage health care professionals and researchers to validate clinical prediction models in the specific target population before using them in that population. Citation Format: Marissa Corine van Maaren, Liza Hoveling, Tom Hueting, Luc JA Strobbe, Mathijs P Hendriks, Gabe S Sonke, Sabine Siesling. Validation of the online prediction model CancerMath in the Dutch breast cancer population [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-08-09.

Background: The optimal time between surgery and initiation of adjuvant chemotherapy is unknown, though delayed time to chemotherapy (TTC) is associated with decreased outcomes of breast cancer patients. Recently, studies have suggested that the association might be subtype-dependent and that TTC within 30 days should be warranted particularly in high-risk triple-negative breast cancer (TNBC) patients. Objective: To determine the extent to which TTC beyond 30 days is associated with reduced overall survival (OS) in TNBC patients. Methods: Using the population-based nationwide Netherlands Cancer Registry we identified TNBC patients diagnosed between 2006 and 2014 who received adjuvant chemotherapy. We distinguished between patients who underwent breast-conserving surgery (BCS) versus mastectomy given the difference in preoperative characteristics and outcomes. Median (95% confidence interval) follow-up was 82.9 (80.5-86.5) and 81.4 (79.5-83.9) months, respectively. Main outcomes and measures: The association between TTC beyond 30 days and OS was estimated with hazard ratios (HR) using propensity-score matched Cox proportional hazard analyses separately for patients who underwent BCS and mastectomy. Results: In total, 3016 patients were included, of whom 1079 (35.8%) underwent BCS and 1937 (64.2%) underwent a mastectomy. In matched patients who underwent BCS, 10-year OS was significantly better for patients with TTC within 30 days compared to patients with TTC beyond 30 days (84.4% vs. 76.9%, P=0.001). Patients with TTC beyond 30 days were more likely than those with TTC within 30 days to die within 10 years after surgery (HR 1.69 (95% CI 1.22-2.34), P=0.002). In matched patients who underwent mastectomy, there was no difference in 10-year OS between those with TTC within or beyond 30 days (74.5% vs. 74.7%, P=0.716), nor an increased risk of death for those with TTC beyond 30 days (HR 1.04 (95%-CI 0.84-1.28), P=0.716). In both populations, the associations were independent of adjuvant radiotherapy. Conclusions: The current results suggest that initiation of chemotherapy beyond 30 days is associated with decreased OS in TNBC patients who underwent BCS; no association was observed for patients who underwent a mastectomy. Therefore, timelier initiation of chemotherapy in TNBC patients undergoing BCS seems warranted. Citation Format: Erik Heeg, Perla J. Marang-van de Mheen, Marissa C. Van Maaren, Kay Schreuder, Rob A.E.M. Tollenaar, Sabine Siesling, Monique E.M.M. Bos, Marie-Jeanne T.F.D. Vrancken Peeters. Association between initiation of adjuvant chemotherapy beyond 30 days following surgery and overall survival among patients with triple-negative breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-14-07.

Purpose To 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. Methods Data 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. Results Age, 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 Conclusions INFLUENCE 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.

Delayed time to chemotherapy (TTC) is associated with decreased outcomes of breast cancer patients. Recently, studies suggested that the association might be subtype‐dependent and that TTC within 30 days should be warranted in patients with triple‐negative breast cancer (TNBC). The aim of the current study is to determine if TTC beyond 30 days is associated with reduced 10‐year overall survival in TNBC patients. We identified all TNBC patients diagnosed between 2006 and 2014 who received adjuvant chemotherapy in the Netherlands. We distinguished between breast‐conserving surgery (BCS) vs. mastectomy given the difference in preoperative characteristics and outcomes. The association was estimated with hazard ratios (HRs) using propensity‐score matched Cox proportional hazard analyses. In total, 3,016 patients were included. In matched patients who underwent BCS (n = 904), 10‐year overall survival was favorable for patients with TTC within 30 days (84.4% vs. 76.9%, p = 0.001). Patients with TTC beyond 30 days were more likely than those with TTC within 30 days to die within 10 years after surgery (HR 1.69 (95% CI 1.22–2.34), p = 0.002). In matched patients who underwent mastectomy (n = 1,568), there was no difference in 10 years overall survival between those with TTC within or beyond 30 days (74.5% vs. 74.7%, p = 0.716), nor an increased risk of death for those with TTC beyond 30 days (HR 1.04 (95% CI 0.84–1.28), p = 0.716). Initiation of adjuvant chemotherapy beyond 30 days is associated with decreased 10 years overall survival in TNBC patients who underwent BCS. Therefore, timelier initiation of chemotherapy in TNBC patients undergoing BCS seems warranted., What's new? Current breast cancer treatment guidelines recommend that chemotherapy is initiated between 6‐12 weeks after surgery. Delayed treatment can lead to poorer outcome, but there's no precise definition of the optimal window. Recent work suggests that for patients with triple negative breast cancer (TNBC), it's best to initiate chemotherapy within 30 days. Here, the authors evaluated outcomes for TNBC patients correlated with time to chemotherapy. They found that in patients undergoing breast‐conserving surgery, an interval of more than 30 days before chemotherapy was associated with decreased survival. If the patient had undergone a mastectomy, a longer delay before chemotherapy did not impact survival.

Background Distant metastatic disease is frequently observed in inflammatory breast cancer (IBC), with a poor prognosis as a consequence. The aim of this study was to analyze the association of hormone receptor (HR) and human epidermal growth factor receptor-2 (HER2) based breast cancer subtypes in stage IV inflammatory breast cancer (IBC) with preferential site of distant metastases and overall survival (OS). Methods For patients with stage IV IBC, diagnosed in the Netherlands between 2005 and 2016, tumors were classified into four breast cancer subtypes: HR+/HER2−, HR+/HER2+, HR−/HER2+, and HR−/HER2−. Patient, tumor, and treatment characteristics and sites of metastases were compared. OS of the subtypes was compared using Kaplan-Meier curves and the log-rank test. Association between subtype and OS was assessed in multivariable models using logistic regression. Results In total, 744 eligible patients were included: 340 (45.7%) tumors were HR+/HER2−, 148 (19.9%) HR−/HER2+, 131 (17.6%) HR+/HER2+, and 125 (16.8%) HR−/HER2−. Bone was the most common metastatic site in all subtypes. A significant predominance of bone metastases was found in HR+/HER2− IBC (71.5%), and liver and lung metastases in the HR−/HER2+ (41.2%) and HR−/HER2− (40.8%) subtypes, respectively. In multivariable analysis, the HR−/HER2− subtype was associated with significantly worse OS as compared to the other subtypes. Conclusion Breast cancer subtypes in stage IV IBC are associated with distinct patterns of metastatic spread and display notable differences in OS. The use of breast cancer subtypes can guide a more patient-tailored staging directed to metastatic site and extend of disease.

Background Survival estimates valid at the time of diagnosis are of limited value for (ex-)breast cancer patients who survived several years, as it includes information on already deceased patients. This study analyzed the 10-year conditional risk of recurrent breast cancer in specific prognostic subgroups according to T and N stage and breast cancer subtypes. Secondly, we investigated 10-year conditional overall (OS) and relative survival (RS), adjusted for confounding. Patients and methods We selected all women diagnosed in 2005 with operated T1-2N0-1 breast cancer from the Netherlands Cancer Registry. Patients were classified into T1N0, T1N1, T2N0 and T2N1 stage. Ten-year conditional recurrence rates were calculated for every year from diagnosis for patients without an event (local (LR), regional recurrence (RR), distant metastasis (DM) or death). Ten-year conditional OS was calculated using multivariable Cox regression. RS was estimated by dividing patient survival rates by those of the general Dutch population. Results We included 7,969 patients: 52.3% had T1N0, 15.3% T1N1, 19.9% T2N0 and 12.5% T2N1 stage. For T1N0, 10-year LR rates changed from 4.6% at diagnosis to 0.5% in year 10. RR rates decreased from 2.3% to 0.2% and DM rates decreased from 7.8% to 0.6%. For T2N1 stage, the LR, RR and DM rates decreased from 6.2% to 0.8%, 5.2% to 0.4% and 19.6% to 1.5%, respectively. Of all patients, 1,702 patients (21.4%) had an unknown breast cancer subtype and were consequently excluded from the analyses according to subtype. Of the remaining 6,267 patients, 3,774 (60.2%) had luminal A, 1,465 (23.4%) had luminal B, 314 (5.0%) had HER2 positive and 714 (11.4%) had triple negative disease For the luminal A subtype, LR, RR and DM rates ranged from 3.9% to 0.4%, 1.7% to 0.5% and 7.3% to 1.1%, while for triple negative these rates ranged between 5.6% to 0.7%, 4.9% to 0.2% and 16.7% to 0%, respectively. Differences between subgroups attenuated over time and all recurrence rates became ≤1.5% in year 10. Ten-year OS and RS, adjusted for confounding, showed diminishing risk differences between subgroups over time. Conclusion Differences in recurrence rates, OS and RS between prognostic subgroups decreased as years passed by. These results highlight the importance of taking into account disease-free years to more accurately predict (ex-)breast cancer patients' prognosis over time. Citation Format: van Maaren MC, Strobbe LJ, Smidt ML, Moossdorff M, Poortmans PM, Siesling S. 10-year conditional recurrence risks, overall and relative survival for breast cancer patients in the Netherlands: Taking account of event-free years [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-08-19.

Background The incidence of ductal carcinoma in situ (DCIS) has drastically increased over the past decades. Since DCIS is resected after diagnosis similarly to invasive breast cancer, the natural cause and behaviour of DCIS is not well known. We aimed to determine breast cancer-specific (BCSS) and overall survival (OS) according to grade in DCIS patients after surgical treatment in the Netherlands. Patients and methods All DCIS patients diagnosed between 1999-2012 were selected from the Netherlands Cancer Registry. Cause of death was obtained from 'Statistics Netherlands'. BCSS and OS were estimated using multivariable Cox regression in the entire cohort and stratified for grade. Results In total, 12,256 patients were included, of whom 1,509 (12.3%) presented with grade I, 3,675 (30.0%) with grade II, 6,064 (49.5%) with grade III and 1,008 (8.2%) with an unknown grade. During a median follow-up of 7.8 years, 1,138 (9.3%) deaths were observed, and 179 (1.5%) were breast cancer-related. Of these, 10 patients had grade I, 46 grade II, 95 grade III and 28 an unknown grade. After adjustment for confounding, grade II and III were related to worse BCSS compared to grade I with HRs of 1.92 (95% CI:0.97-3.81) and 2.14 (95% CI:1.11-4.12), respectively. No association between grade and OS was observed. Conclusion BCSS and OS rates in DCIS patients are excellent. Since superior rates were observed for low-grade DCIS, and earlier studies have shown that low-grade DCIS have a very low chance on recurrence or upstage to invasive cancer, it seems justified to investigate whether active surveillance may be a balanced alternative for conventional surgical treatment. Citation Format: van Maaren MC, Lagendijk M, Tilanus-Linthorst MM, de Munck L, Pijnappel RM, Schmidt MK, Wesseling J, Koppert LB, Siesling S. Breast cancer-related deaths according to grade in ductal carcinoma in situ: A Dutch population-based study on patients diagnosed between 1999 and 2012 [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-08-09.

Our study aimed to provide a comprehensive overview of trends in incidence, survival, mortality and treatment of first primary invasive breast cancer (BC), according to age, stage and receptor subtype in the Netherlands between 1989 and 2017. Data from all women diagnosed with first primary stage I to IV BC (N = 320 249) were obtained from the Netherlands Cancer Registry. BC mortality and general population data were retrieved from Statistics Netherlands. Age‐standardised incidence and mortality rates were calculated with annual percentage change (APC) and average annual percentage change (AAPC) statistics. The relative survival (RS) was used as estimator for disease‐specific survival. The BC incidence for all BC patients combined significantly increased until 2013 from 126 to 158 per 100 000 person‐years, after which a declining trend was observed. Surgery became less extensive, but (neo‐)adjuvant systemic treatments and their combinations were given more frequently. The RS improved for all age groups and for most stages and receptor subtypes, but remained stable for all subtypes since 2012 to 2013 and since 2000 to 2009 for Stage IV BC at 15 years of follow‐up. Overall, the 5‐ and 10‐year RS increased from 76.8% (95% confidence interval [CI]: 76.1, 77.4) and 55.9% (95% CI: 54.7, 57.1) in 1989 to 1999 to 91.0% (95% CI: 90.5, 91.5) and 82.9% (95% CI: 82.2, 83.5), respectively, in 2010 to 2016. BC mortality improved regardless of age and overall decreased from 57 to 35 per 100 000 person‐years between 1989 and 2017. In conclusion, the BC incidence in the Netherlands has steadily increased since 1989, but the latest trends show promising declines. Survival improved markedly for most patients and the mortality decreased regardless of age., What's new? Studies that simultaneously capture incidence, survival, and mortality trends in breast cancer are scarce, and receptor subtype‐specific trends have remained largely unexplored. This study provides an up‐to‐date and comprehensive overview of first primary invasive breast cancer trends in the Netherlands in 1989 to 2017. Breast cancer incidence increased for all breast cancer patients combined until 2013, but the latest trends show a promising decline. Treatment strategies became more complex. Relative survival improved for all age groups and for most stages and receptor subtypes, and mortality decreased overall. The results may be useful in supporting healthcare management and informing current clinical practice.

Background Despite the potential for residual lymph node metastases after a negative or positive sentinel lymph node biopsy (SLNB), breast cancer patients rarely experience regional recurrences (RRs). This study aimed to quantify the effects of nonsurgical treatments on RR incidence among SLNB-negative (SLNB N0) breast cancer patients. Methods All primary SLNB N0-staged breast cancer patients with a diagnosis between 2005 and 2008 and 5-year follow-up data on recurrences were selected from the Netherlands Cancer Registry. The cumulative incidence function (CIF) for RR was calculated as the first event at 5 years, taking into account any other first-event (local or distant recurrence, contralateral breast cancer, or death) as competing risk. Cox regression analysis was used to model the cause-specific hazard of RR developing as the first event to quantify the effect of adjuvant systemic therapy and whole-breast radiotherapy (RT) on RR incidence at 5 years. Results The study included 13,512 patients. Of these patients, 162 experienced an RR. The CIF of RR at 5 years was 1.3% (95% confidence interval [CI], 1.1–1.5%), whereas the CIFs for death and other events were 4.4% and 9.5%, respectively. Cox regression analysis showed hazard ratios (HRs) of 0.46 (95% CI 0.33–0.64), 0.31 (95% CI 0.18–0.55), and 0.40 (95% CI 0.24–0.67) respectively for patients treated by RT as a routine part of breast-conserving therapy (BCT), chemotherapy, and hormonal therapy. Conclusion RT as routine part of BCT, chemotherapy, and hormonal therapy independently exerted a mitigating effect on the risk for the development of RR. The three methods at least halved the risk.

Background: The incidence of ductal carcinoma in situ (DCIS) has drastically increased over the past decades. Because DCIS is resected after diagnosis similar to invasive breast cancer, the natural cause and behaviour of DCIS is not well known. We aimed to determine breast cancer-specific survival (BCSS) and overall survival (OS) according to grade in DCIS patients after surgical treatment in the Netherlands.Patients and methods: All DCIS patients diagnosed between 1999 and 2012 were selected from the Netherlands Cancer Registry. The cause of death was obtained from 'Statistics Netherlands'. BCSS and OS were estimated using multivariable Cox regression in the entire cohort and stratified for grades.Results: In total, 12,256 patients were included, of whom 1509 (12.3%) presented with grade I, 3675 (30.0%) with grade II, 6064 (49.5%) with grade III and 1008 (8.2%) with an unknown grade. During a median follow-up of 7.8 years, 1138 (9.3%) deaths were observed, and 179 (1.5%) were breast cancer-related. Of these, 10 patients had grade I; 46 grade II; 95 grade III and 28 an unknown grade. After adjustment for confounding, grade II and III were related to worse BCSS than grade I with hazard ratios of 1.92 (95% confidence interval [CI]: 0.97-3.81) and 2.14 (95% CI: 1.11-4.12), respectively. No association between grades and OS was observed.Conclusion: BCSS and OS in DCIS patients were excellent. Because superior rates were observed for low-grade DCIS, it seems justified to investigate whether active surveillance may be a balanced alternative for conventional surgical treatment. (C) 2018 Elsevier Ltd. All rights reserved.

Background Landmark trials have shown breast-conserving surgery (BCS) combined with radiotherapy to be as safe as mastectomy in breast cancer treatment. This population-based study aimed to evaluate trends in BCS from 1989 to 2015 in nine geographical regions in the Netherlands. Methods All women diagnosed between 1989 and 2015 with primary T1–2 N0–1 breast cancer, treated with BCS or mastectomy, were identified from the Netherlands Cancer Registry. Crude and case mix-adjusted rates of BCS were evaluated and compared between nine Dutch regions for two time intervals: 1989–2002 and 2003–2015. The annual percentage change in BCS per region over time was assessed by means of Joinpoint regression analyses. Explanatory variables associated with the choice of initial surgery were evaluated using multivariable logistic regression. Results A total of 202 934 patients were included, 82 200 treated in 1989–2002 and 120 734 in 2003–2015. During 1989–2002, the mean rate of BCS was 50·6 per cent, varying significantly from 39·0 to 71·7 per cent between the nine regions. For most regions, a marked rise in BCS was observed between 2002 and 2003. During 2003–2015, the mean rate of BCS increased to 67·4 per cent, but still varied significantly between regions from 58·5 to 75·5 per cent. A significant variation remained after case-mix correction. Conclusion This large nationwide study showed that the use of BCS increased from 1989 to 2015 in the Netherlands. After adjustment for explanatory variables, a large variation still existed between the nine regions. This regional variation underlines the need for implementation of a uniform treatment and decision-making strategy.

Purpose: This study aimed to validate and update a model for predicting the risk of axillary lymph node (ALN) metastasis for assisting clinical decision-making.Methods: We included breast cancer patients diagnosed at six Dutch hospitals between 2011 and 2015 to validate the original model which includes six variables: clinical tumor size, tumor grade, estrogen receptor status, lymph node longest axis, cortical thickness and hilum status as detected by ultrasonography. Subsequently, we updated the original model using generalized linear model (GLM) tree analysis and by adjusting its intercept and slope. The area under the receiver operator characteristic curve (AUC) and calibration curve were used to assess the original and updated models. Clinical usefulness of the model was evaluated by false-negative rates (FNRs) at different cut-off points for the predictive probability.Results: Data from 1416 patients were analyzed. The AUC for the original model was 0.774. Patients were classified into four risk groups by GLM analysis, for which four updated models were created. The AUC for the updated models was 0.812. The calibration curves showed that the updated model predictions were better in agreement with actual observations than the original model predictions. FNRs of the updated models were lower than the preset 10% at all cut-off points when the predictive probability was less than 12.0%. ER Conclusions: The original model showed good performance in the Dutch validation population. The updated models resulted in more accurate ALN metastasis prediction and could be useful preoperative tools in selecting low-risk patients for omission of axillary surgery. (C) 2018 Elsevier Ltd, BASO - The Association for Cancer Surgery, and the European Society of Surgical Oncology. All rights reserved.

This large population-based study compared breast-conserving surgery with radiation therapy (BCT) with mastectomy on (long-term) breast cancer-specific (BCSS) and overall survival (OS), and investigated the influence of several prognostic factors. Patients with primary T1-2N0-2M0 breast cancer, diagnosed between 1999 and 2012, were selected from the Netherlands Cancer Registry. We investigated the 1999-2005 (long-term outcome) and the 2006-2012 cohort (contemporary adjuvant systemic therapy). Cause of death was derived from the Statistics Netherlands (CBS). Multivariable analyses, per time cohort, were performed in T1-2N0-2, and separately in T1-2N0-1 and T1-2N2 stages. The T1-2N0-1 stages were further stratified for age, hormonal receptor and HER2 status, adjuvant systemic therapy and comorbidity. In total, 129,692 patients were included. In the 1999-2005 cohort, better BCSS and OS for BCT than mastectomy was seen in all subgroups, except in patients 50 years, not treated with chemotherapy and with comorbidity. Both treatments led to similar BCSS in patients

This abstract was withdrawn by the authors.

Background: Lobular carcinoma in situ (LCIS) is a non-invasive proliferation of epithelial cells of the terminal-duct lobular unit. Although LCIS is non-invasive itself, studies in the past showed that patients with LCIS have a 4-12 times higher risk of developing invasive breast cancer than the general population. Therefore, the optimal management of LCIS is still subject of debate. Despite lack of information about LCIS in the Dutch population, the Dutch guidelines recommend active surveillance for classic LCIS. This study aims to describe incidence and management of classic LCIS over time in the Netherlands, and to understand its behavior in terms of survival and the risk of developing subsequent ductal carcinoma in situ (DCIS) or invasive breast cancer. Methods: All female patients diagnosed with classic LCIS between 1989-2017 in the Netherlands were identified from the nationwide population-based Netherlands Cancer Registry. Patients previously diagnosed with DCIS or invasive breast cancer were excluded. Standardized incidence ratios (SIR) were calculated to estimate the risk of subsequent DCIS and invasive breast cancer, using data on population size from Statistics Netherlands (CBS). Multivariable Cox proportional hazards regression was performed to estimate hazard ratios (HR) of overall mortality and risk of subsequent invasive breast cancer for different patient-, tumor- and treatment-related characteristics. Competing risks were taken into account. Relative survival was estimated using the Ederer II method, with expected mortality of the general population (matched by age, gender, calendar year) as a reference. Results: We included 1,890 patients. The median age at diagnosis was 51 years (range 24-91 years). The incidence of classic LCIS increased from 186 patients between 1989-1993 to 497 patients between 2014-2017. Of all patients, 50.5% were treated with breast-conserving surgery, 10.2% were treated with mastectomy, 12.6% were treated with an unknown type of surgery and 26.7% did not receive surgery. While in 1989 all patients were surgically treated, only 41% were surgically treated in 2017. Radiotherapy was administered in 3.1% and endocrine therapy in 0.2% of the patients. Median follow-up to death, and subsequent DCIS or invasive breast cancer was 8.5 years (range 0-29 years) and 6.9 years (range 0-27 years). Unadjusted 10- and 20-year overall survival rates were 91% (95%CI:90-93%) and 76% (95%CI:73-79%). Unadjusted 10- and 20-year relative survival rates were 99% (95%CI:97-100%) and 95% (95%CI:91-99%). In total, 270 patients was diagnosed with subsequent invasive breast cancer and 48 were diagnosed with DCIS. Of these, 60% were ipsilateral and 40% were contralateral. The SIRs for DCIS and invasive breast cancer were 4.4 (95%CI:3.2-5.8) and 24.4 (95%CI:21.6-27.5) as compared to the general population, respectively. The SIR was higher for younger patients than for older patients, and the SIR of ipsilateral invasive breast cancer was higher in case no surgical management was performed. In multivariable analysis, increasing age was the most important predictor for overall mortality, while no surgical management was related to a higher incidence of subsequent invasive breast cancer. Conclusion: LCIS incidence has increased over time, while surgical management has decreased. This study shows that the mortality risk is very low compared to the general population. In contrast, the incidence of invasive breast cancer, over a median follow-up of 6.9 years (range 0-27 years), is over 24 times more common in LCIS patients than in the general population. These results suggest that LCIS patients have a good prognosis, even when a subsequent invasive breast cancer is diagnosed. However, the high incidence of second tumors suggest that we carefully monitor these patients. Citation Format: Marissa Corine van Maaren, Agustin Ortega Avila, Jos Bart, Pieter J Westenend, Sabine Siesling. Classic lobular carcinoma in situ in the Netherlands: A population-based analysis [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr PD3-04.

Contains fulltext : 204285.pdf (Publisher’s version ) (Open Access) PURPOSE: To analyze the influence of hormone receptors (HR) and Human Epidermal growth factor Receptor-2 (HER2)-based molecular subtypes in stage III inflammatory breast cancer (IBC) on tumor characteristics, treatment, pathologic response to neoadjuvant chemotherapy (NACT), and overall survival (OS). METHODS: Patients with stage III IBC, diagnosed in the Netherlands between 2006 and 2015, were classified into four breast cancer subtypes: HR+/HER2- , HR+/HER2+ , HR-/HER2+ , and HR-/HER2- . Patient-, tumor- and treatment-related characteristics were compared. In case of NACT, pathologic complete response (pCR) was compared between subgroups. OS of the subtypes was compared using Kaplan-Meier curves and the log-rank test. RESULTS: 1061 patients with stage III IBC were grouped into subtypes: HR+/HER2- (N = 453, 42.7%), HR-/HER2- (N = 258, 24.3%), HR-/HER2+ (N = 180,17.0%), and HR+/HER2+ (N = 170,16.0%). In total, 679 patients (85.0%) received NACT. In HR-/HER2+ tumors, pCR rate was highest (43%, (p < 0.001). In case of pCR, an improved survival was observed for all subtypes, especially for HR+/HER2+ and HR-/HER2+ tumor subtypes. Trimodality therapy (NACT, surgery, radiotherapy) improved 5-year OS as opposed to patients not receiving this regimen: HR+/HER2- (74.9 vs. 46.1%), HR+/HER2+ (80.4 vs. 52.6%), HR-/HER2+ (76.4 vs. 29.7%), HR-/HER2- (47.6 vs. 27.8%). CONCLUSIONS: In stage III IBC, breast cancer subtypes based on the HR and HER2 receptor are important prognostic factors of response to NACT and OS. Patients with HR-/HER2- IBC were less likely to achieve pCR and had the worst OS, irrespective of receiving most optimal treatment regimen to date (trimodality therapy).

Purpose: Propensity trimming, hierarchical modelling and instrumental variable (IV) analysis are statistical techniques dealing with confounding, cluster-related variation or confounding by severity. This study aimed to explain (dis)advantages of these techniques in estimating the effect of breast-conserving therapy (BCT) and mastectomy on 10-year distant metastasis-free survival (DMFS). Methods: All women diagnosed in 2005 with primary T1-2N0-1 breast cancer treated with BCT or mastectomy were selected from the Netherlands Cancer Registry. We used multivariable Cox regression to correct for confounding. Propensity trimming was used to create a more homogeneous population for which the treatment choice was not self-evident. Hospital of surgery was used as hierarchical level to handle hospital-related variation, and as IV to deal with unmeasured confounding. Results: Multivariable Cox regression showed higher 10-year DMFS for BCT than mastectomy [HR 0.70 (95% CI 0.60–82)]. Propensity trimming on the 10–90th and the 20–80th percentile of the propensity score distribution and hierarchical modelling showed similar HRs. IV analysis showed no significant difference between BCT and mastectomy. Conclusion: Unmeasured confounding is very difficult to eliminate in observational research. We cannot conclude that BCT or mastectomy has a causal relationship with 10-year DMFS. It is crucial to critically evaluate all model’s assumptions, and to be careful in drawing firm conclusions.

Here we report for the first time the relation between breast cancer subtypes and 10-year recurrence rates and mortality in the Netherlands. All operated women diagnosed with invasive non-metastatic breast cancer in 2005 in the Netherlands were included. Patients were classified into breast cancer subtypes according to ER, PR, HER2 status and grade: luminal A, luminal B, HER2 positive and triple negative. Percentages and hazards of recurrence were compared among subtypes. Adjusted 10-year overall (OS) and recurrence-free survival (RFS) were calculated using multivariable Cox regression. Of 8,062 patients, 4,482 (56%) were luminal A, 2,090 (26%) luminal B, 504 (6%) HER2 positive and 986 (12%) triple negative. Local recurrences (7.5%) and distant metastases (25.6%) occurred most often in HER2 positive disease and the least often in luminal A (3.7% and 9.5%, respectively). Regional recurrences were most often diagnosed in triple negative disease (5.2%), and the least often in luminal A (1.7%). HER2 positive and triple negative subtypes had the highest recurrence rates in the second year, while luminal A and B showed a more continuous pattern over time, with lobular tumours recurring more often. After adjustment for differences in baseline characteristics, triple negative disease showed worse 10-year OS and triple negative and HER2 positive disease had the lowest 10-year RFS. In the Netherlands, breast cancer subtypes are important predictors for 10-year recurrence rates. Knowledge on recurrence and survival rates according to these different subtypes, in combination with other prognostic factors, can support patient-tailored treatment and individualised follow-up.

Objective Many cancer survivors are facing difficulties in getting a life insurance; raised premiums and declinatures are common. We generated a prediction model estimating the conditional extra mortality risk of breast cancer patients in the Netherlands. This model can be used by life insurers to accurately estimate the additional risk of an individual patient, conditional on the years survived. Methodology All women diagnosed with stage I-III breast cancer in 2005–2006, treated with surgery, were selected from the Netherlands Cancer Registry. For all stages separately, multivariable logistic regression was used to estimate annual mortality risks, conditional on the years survived, until 10 years after diagnosis, resulting in 30 models. The conditional extra mortality risk was calculated by subtracting mortality rates of the general Dutch population from the patient mortality rates, matched by age, gender and year. The final model was internally and externally validated, and tested by life insurers. Results We included 23,234 patients: 10,101 stage I, 9,868 stage II and 3,265 stage III. The final models included age, tumor stage, nodal stage, lateralization, location within the breast, grade, multifocality, hormonal receptor status, HER2 status, type of surgery, axillary lymph node dissection, radiotherapy, (neo)adjuvant systemic therapy and targeted therapy. All models showed good calibration and discrimination. Testing of the model by life insurers showed that insurability using the newly-developed model increased with 13%, ranging from 0%-24% among subgroups. Conclusion The final model provides accurate conditional extra mortality risks of breast cancer patients, which can be used by life insurers to make more reliable calculations. The model is expected to increase breast cancer patients’ insurability and transparency among life insurers.

Background: Survival estimates from diagnosis are of limited importance for (ex-) breast cancer patients who survived several years, as it includes information on already deceased patients. This study analysed the 10-year conditional risk of recurrent breast cancer in specific prognostic subgroups. Second, we investigated 10-year conditional overall survival (OS) and relative survival (RS), adjusted for confounding. Patients and methods: All women diagnosed in 2005 with operated T1-2N0-1 breast cancer were selected from the Netherlands Cancer Registry. Patients were classified into T1N0, T1N1, T2N0 and T2N1 stage. Ten-year conditional recurrence rates were calculated from diagnosis, and for patients without an event (local [LR], regional recurrence [RR], distant metastasis [DM] or death) every year following diagnosis. Ten-year conditional OS was calculated using multivariable Cox regression. RS was estimated by dividing patient survival rates by those of the general Dutch population. Results: We included 7969 patients: 52.3% had T1N0, 15.3% T1N1, 19.9% T2N0 and 12.5% T2N1 stage. For T1N0, 10-year LR rates changed from 4.6% at diagnosis to 0.5% in year 10. RR rates changed from 2.3% to 0.2%, and DM rates changed from 7.8% to 0.6%. For T2N1 stage, the LR, RR and DM rates changed from 6.2% to 0.8%, 5.2%-0.4% and 19.6% e1.5%, respectively. For the luminal A subtype, LR, RR and DM rates changed from 3.9% to 0.4%, 1.7%-0.5% and 7.3%-1.1%, while for triple negative, these rates changed from 5.6% to 0.7%, 4.9%-0.2% and 16.7%-0%, respectively. Differences between subgroups attenuated over time, and all recurrence rates became

Background: Randomised controlled trials have shown that breast conserving therapy (conserving surgery with radiation therapy, BCT) has equal overall survival (OS) rates as mastectomy without radiation therapy (MAST) in early stage breast cancer. However, 10-year disease-free survival (DFS) in a population-based study was not investigated before. The aim of this study was to compare 10-year OS and DFS after BCT with MAST in Dutch women with early stage breast cancer. Methods: Data of all women diagnosed with primary invasive T1-2N0-1M0 stage breast cancer between 1 January 2000 and 31 December 2004, treated with either BCT or MAST, were selected from the Netherlands Cancer Registry. Multivariable Cox proportional hazard analysis was performed to estimate 10-year OS, stratified for T and N stage. Ten-year DFS was determined in a subgroup of patients diagnosed in 2003, of which an active follow-up was conducted registering all recurrent events within 10 years. Multiple imputation was performed to account for missing data. Results: Of in total 37,207 patients, 21,734 patients (58.4%) received BCT and 15,473 patients (41.6%) received MAST. The subcohort of 2003 consisted of 7,552 patients, with similar distributions of treatments and characteristics. In the total cohort, 10-year OS was 76.8% (99% CI: 76.1-77.5%) after BCT and 59.7 (99% CI: 58.7-60.7%) after MAST. After correction for confounding, 10-year OS was better after BCT than after MAST (HRadjusted: 0.79 [99% CI 0.75-0.83]). In the 2003 cohort, 10-year DFS was 83.6% (99% CI: 82.5-84.7%) after BCT and 81.5% (99% CI: 79.6-83.4%) after MAST. After correction for confounding, 10-year DFS was comparable for both treatments (HRadjusted 0.91 [99% CI 0.77-1.07]). All results were similar for all subgroups (Table). In the 2003 cohort, 11.0% of the patients experienced distant metastases (DM) after BCT compared to 14.7% after MAST (p Table. Adjusted hazard ratios of breast-conserving surgery with radiation therapy vs. mastectomy on 10-year overall and disease-free survival in T1-2N0-1 staged breast cancer patients n10-year OS [99% CI]10-year DFS [99% CI]Overall cohort MAST15,47311BCT21,7340.79 [0.75-0.83]0.91 [0.79-1.07]Subgroups T1N0 MAST6,09211BCT18,8600.80 [0.74-0.87]0.81 [0.62-1.08]T1N1 MAST2,18511BCT3,7410.80 [0.69-0.92]1.02 [0.64-1.63]T2N0 MAST4,17411BCT3,1650.79 [0.70-0.88]0.99 [0.72-1.35]T2N1 MAST3,02211BCT2,0600.77 [0.67-0.88]0.82 [0.58-1.15]All hazard ratios are adjusted for all relevant confounders. Abbreviations: MAST = mastectomy; BCT = breast conserving surgery with post-operative radiation therapy; OS = overall survival; DFS = disease-free survival Conclusion: BCT showed substantially improved OS compared to MAST. However, while DFS was similar, patients treated with BCT less often developed RR and DM. Although residual factors might explain part of the difference in recurrences, we hypothesise that radiation therapy might largely be responsible for better OS by eliminating residual tumour cells. Citation Format: van Maaren MC, de Munck L, de Bock GH, Jobsen JJ, van Dalen T, Poortmans P, Linn SC, Strobbe LJA, Siesling S. Higher 10-year overall survival after breast conserving therapy compared to mastectomy in early stage breast cancer: A population-based study with 37,207 patients. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr S3-05.

As local surgery is currently the most important component of curative breast cancer treatment, a large part of the study by focused on trends in breast surgery and survival following breast-conserving therapy and mastectomy. I n the Netherlands, the use of breast-conserving surgery had increased over time. Right after publication of landmark trials in 2002, demonstrating the safety of breast-conserving surgery, a steep rise in the use of breast-conserving surgery was seen. However, even after case mix correction, there was large variation between the nine Dutch regions. Part of this variation can be explained by unmeasured factors such as the patient’ and doctors’ preference or the degree of shared decision-making and are subject of further research. In the Dutch early stage breast cancer population, breast-conserving treatment led to at least equal overall, disease-free and breast cancer-specific survival as compared to mastectomy. This was confirmed in all subgroups based on age, tumour and nodal stage, receptor status, administration of adjuvant systemic therapy and presence of comorbidities. Besides, it was shown that starting radiation therapy as soon as possible after breast-conserving surgery did not affect survival rates. The second part of the study by Marissa van Maaren concentrated on clinical risk prediction of survival in breast cancer. First, the online prediction tool PREDICT was validated on the Dutch breast cancer population. The validation study showed that PREDICT accurately predicts overall survival in most Dutch breast cancer patients, but that results for should be interpreted cautiously in some specific patient groups. Furthermore, this thesis emphasised the importance of classifying patients according to breast cancer subtype, as these subtypes are independently related to survival and recurrence risks. In addition, the number of (disease-free) years survived (conditional survival) is a determining factor for the remaining risk of recurrences. CONCLUSIONS Breast-conserving treatment is at least equal to mastectomy in early stage breast cancer in terms of 10-year overall, disease-free and breast cancer-specific survival. Combined with results of randomised controlled trials, this thesis contributes to increased awareness that a mastectomy does not reduce the risk of recurrence and mortality, as compared to breast-conserving therapy. Besides, this thesis revealed the importance of specific risk profiles and use of conditional survival in clinical risk prediction. Overall, communication to patients, based on relevant data, plays a central role and should form the basis of the shared decision-making process. Ultimately, this will lead to better treatment decisions and more realistic risk perceptions.

This large population-based study compared breast-conserving surgery with radiation therapy (BCT) with mastectomy on (long-term) breast cancer-specific (BCSS) and overall survival (OS), and investigated the influence of several prognostic factors. Patients with primary T1-2N0-2M0 breast cancer, diagnosed between 1999 and 2012, were selected from the Netherlands Cancer Registry. We investigated the 1999–2005 (long-term outcome) and the 2006–2012 cohort (contemporary adjuvant systemic therapy). Cause of death was derived from the Statistics Netherlands (CBS). Multivariable analyses, per time cohort, were performed in T1-2N0-2, and separately in T1-2N0-1 and T1-2N2 stages. The T1-2N0-1 stages were further stratified for age, hormonal receptor and HER2 status, adjuvant systemic therapy and comorbidity. In total, 129,692 patients were included. In the 1999–2005 cohort, better BCSS and OS for BCT than mastectomy was seen in all subgroups, except in patients < 40 years with T1-2N0-1 stage. In the 2006–2012 cohort, superior BCSS and OS were found for T1-2N0-1, but not for T1-2N2. Subgroup analyses for T1-2N0-1 showed superior BCSS and OS for BCT in patients >50 years, not treated with chemotherapy and with comorbidity. Both treatments led to similar BCSS in patients

Item does not contain fulltext BACKGROUND: PREDICT version 2.0 is increasingly used to estimate prognosis in breast cancer. This study aimed to validate this tool in specific prognostic subgroups in the Netherlands. METHODS: All operated women with non-metastatic primary invasive breast cancer, diagnosed in 2005, were selected from the nationwide Netherlands Cancer Registry (NCR). Predicted and observed 5- and 10-year overall survival (OS) were compared for the overall cohort, separated by oestrogen receptor (ER) status, and predefined subgroups. A >5% difference was considered as clinically relevant. Discriminatory accuracy and goodness-of-fit were determined using the area under the receiver operating characteristic curve (AUC) and the Chi-squared-test. RESULTS: We included 8834 patients. Discriminatory accuracy for 5-year OS was good (AUC 0.80). For ER-positive and ER-negative patients, AUCs were 0.79 and 0.75, respectively. Predicted 5-year OS differed from observed by -1.4% in the entire cohort, -0.7% in ER-positive and -4.9% in ER-negative patients. Five-year OS was accurately predicted in all subgroups. Discriminatory accuracy for 10-year OS was good (AUC 0.78). For ER-positive and ER-negative patients AUCs were 0.78 and 0.76, respectively. Predicted 10-year OS differed from observed by -1.0% in the entire cohort, -0.1% in ER-positive and -5.3 in ER-negative patients. Ten-year OS was overestimated (6.3%) in patients >/=75 years and underestimated (-13.%) in T3 tumours and patients treated with both endocrine therapy and chemotherapy (-6.6%). CONCLUSIONS: PREDICT predicts OS reliably in most Dutch breast cancer patients, although results for both 5-year and 10-year OS should be interpreted carefully in ER-negative patients. Furthermore, 10-year OS should be interpreted cautiously in patients >/=75 years, T3 tumours and in patients considering endocrine therapy and chemotherapy.