Elizabeth Mallon
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Elizabeth Mallon has a total of 52 co-authors. Here are the top co-authors by number of shared publications:
- John M. S. Bartlett
Cancer Research UK Scotland Centre, University of Edinburgh, Edinburgh, United Kingdom
14 shared publications - Karen J. Taylor
Cancer Research UK Scotland Centre, Institute of Genetics and Cancer, University of Edinburgh , Edinburgh
14 shared publications - Annette Hasenburg
Department of Obstetrics and Gynecology, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstraße, 1, 55131 Mainz, Germany
13 shared publications - Christos Markopoulos
Medical School, National and Kapodistrian University of Athens, Athens, Greece
13 shared publications - Cornelis J. H. van de Velde
Department of Surgery, Leiden University Medical Center, 2333 ZA Leiden, The Netherlands
13 shared publications - Gregory Russell Pond
Department of Oncology, McMaster University, Hamilton, ON L8V 5C2, Canada
13 shared publications
Latest Publications
- Open Access
- 14 April 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e88665
Transfer Learning and Machine Learning for Training Five-Year Survival Prognostic Models in Early Breast Cancer: Development and Validation Study
Background: Prognostic information is essential for decision-making in breast cancer management. In recent years, trials and clinical practice have emphasized genomic prognostication tools, despite clinicopathological methods being more affordable and accessible. PREDICT v3 is one such tool with promising results across cohorts. Advances in machine learning (ML), transfer learning, and ensemble methods provide opportunities to enhance these approaches, especially where missing data and model assumptions differ across diverse populations. Objective: This study evaluates the potential to improve survival prognostication in breast cancer. More precisely, we compare de novo ML, transfer learning from the pretrained prognostication model PREDICT v3, and a stacked ensemble approach. Methods: Data from the MA.27 trial (NCT00066573) were used for model training, with external validation on data from the Tamoxifen Exemestane Adjuvant Multinational trial (NCT00279448 and NCT00032136) and a US Surveillance, Epidemiology, and End Results cohort. Transfer learning was applied by re-estimating the parameters of the pretrained prognostic tool PREDICT v3. De novo ML included random survival forests and extreme gradient boosting, and the ensemble was implemented using weighted linear stacking of model predictions. Internal and external validation was assessed in terms of the integrated calibration index and discrimination. Shapley Additive Explanations values were used to explain model predictions and decision-curve analysis to facilitate the interpretation of performance differences. Results: Transfer learning, de novo random survival forest, and the stacked ensemble improved calibration in MA.27 over the pretrained model (integrated calibration index reduced from 0.042 in PREDICT v3 to ≤0.007) while discrimination remained comparable (AUROC increased from 0.738 in PREDICT v3 to 0.744-0.799). In decision-curve analysis, these approaches demonstrated consistently positive net benefit across clinically relevant thresholds, while PREDICT v3 lost net benefit beyond 7.5% predicted risk. Invalid PREDICT v3 predictions were observed in 23.8% to 25.8% of MA.27 individuals due to missing information. In contrast, ML models and the stacked ensemble predicted survival despite missing data. Across all models, patient age, nodal status, pathological grading, and tumor size had the highest Shapley Additive Explanations values, indicating their importance for survival prognostication. External validation in the US Surveillance, Epidemiology, and End Results cohort confirmed the benefits of transfer learning, RSF, and ensemble in terms of calibration while maintaining discrimination at comparable levels. In contrast, generalizability was limited in the Tamoxifen Exemestane Adjuvant Multinational trial, a cohort with a substantially different distribution of clinicopathological characteristics. Conclusions: This study demonstrates that transfer learning, de novo RSF, and a stacked ensemble can improve prognostication compared with the pretrained PREDICT v3, particularly in the presence of missing or uncertain inputs. Transportability may be limited in cohorts with different clinicopathological profiles, requiring local validation before clinical deployment. Ultimately, better survival estimation can provide more meaningful guidance in breast cancer care. Trial Registration: ClinicalTrials.gov NCT00066573; https://clinicaltrials.gov/study/NCT00066573, NCT00279448; https://clinicaltrials.gov/study/NCT00279448, NCT00032136; https://clinicaltrials.gov/study/NCT00032136
- 14 April 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e88665
Open AccessTransfer Learning and Machine Learning for Training Five-Year Survival Prognostic Models in Early Breast Cancer: Development and Validation Study
Background: Prognostic information is essential for decision-making in breast cancer management. In recent years, trials and clinical practice have emphasized genomic prognostication tools, despite clinicopathological methods being more affordable and accessible. PREDICT v3 is one such tool with promising results across cohorts. Advances in machine learning (ML), transfer learning, and ensemble methods provide opportunities to enhance these approaches, especially where missing data and model assumptions differ across diverse populations. Objective: This study evaluates the potential to improve survival prognostication in breast cancer. More precisely, we compare de novo ML, transfer learning from the pretrained prognostication model PREDICT v3, and a stacked ensemble approach. Methods: Data from the MA.27 trial (NCT00066573) were used for model training, with external validation on data from the Tamoxifen Exemestane Adjuvant Multinational trial (NCT00279448 and NCT00032136) and a US Surveillance, Epidemiology, and End Results cohort. Transfer learning was applied by re-estimating the parameters of the pretrained prognostic tool PREDICT v3. De novo ML included random survival forests and extreme gradient boosting, and the ensemble was implemented using weighted linear stacking of model predictions. Internal and external validation was assessed in terms of the integrated calibration index and discrimination. Shapley Additive Explanations values were used to explain model predictions and decision-curve analysis to facilitate the interpretation of performance differences. Results: Transfer learning, de novo random survival forest, and the stacked ensemble improved calibration in MA.27 over the pretrained model (integrated calibration index reduced from 0.042 in PREDICT v3 to ≤0.007) while discrimination remained comparable (AUROC increased from 0.738 in PREDICT v3 to 0.744-0.799). In decision-curve analysis, these approaches demonstrated consistently positive net benefit across clinically relevant thresholds, while PREDICT v3 lost net benefit beyond 7.5% predicted risk. Invalid PREDICT v3 predictions were observed in 23.8% to 25.8% of MA.27 individuals due to missing information. In contrast, ML models and the stacked ensemble predicted survival despite missing data. Across all models, patient age, nodal status, pathological grading, and tumor size had the highest Shapley Additive Explanations values, indicating their importance for survival prognostication. External validation in the US Surveillance, Epidemiology, and End Results cohort confirmed the benefits of transfer learning, RSF, and ensemble in terms of calibration while maintaining discrimination at comparable levels. In contrast, generalizability was limited in the Tamoxifen Exemestane Adjuvant Multinational trial, a cohort with a substantially different distribution of clinicopathological characteristics. Conclusions: This study demonstrates that transfer learning, de novo RSF, and a stacked ensemble can improve prognostication compared with the pretrained PREDICT v3, particularly in the presence of missing or uncertain inputs. Transportability may be limited in cohorts with different clinicopathological profiles, requiring local validation before clinical deployment. Ultimately, better survival estimation can provide more meaningful guidance in breast cancer care. Trial Registration: ClinicalTrials.gov NCT00066573; https://clinicaltrials.gov/study/NCT00066573, NCT00279448; https://clinicaltrials.gov/study/NCT00279448, NCT00032136; https://clinicaltrials.gov/study/NCT00032136
- Open Access
- 29 November 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Transfer Learning and Machine Learning for Training Five-Year Survival Prognostic Models in Early Breast Cancer: Development and Validation Study (Preprint)
BACKGROUND Prognostic information is essential for decision-making in breast cancer management. In recent years, trials and clinical practice have emphasized genomic prognostication tools, despite clinicopathological methods being more affordable and accessible. PREDICT v3 is one such tool with promising results across cohorts. Advances in machine learning (ML), transfer learning, and ensemble methods provide opportunities to enhance these approaches, especially where missing data and model assumptions differ across diverse populations. OBJECTIVE This study evaluates the potential to improve survival prognostication in breast cancer. More precisely, we compare de novo ML, transfer learning from the pretrained prognostication model PREDICT v3, and a stacked ensemble approach. METHODS Data from the MA.27 trial (NCT00066573) were used for model training, with external validation on data from the Tamoxifen Exemestane Adjuvant Multinational trial (NCT00279448 and NCT00032136) and a US Surveillance, Epidemiology, and End Results cohort. Transfer learning was applied by re-estimating the parameters of the pretrained prognostic tool PREDICT v3. De novo ML included random survival forests and extreme gradient boosting, and the ensemble was implemented using weighted linear stacking of model predictions. Internal and external validation was assessed in terms of the integrated calibration index and discrimination. Shapley Additive Explanations values were used to explain model predictions and decision-curve analysis to facilitate the interpretation of performance differences. RESULTS Transfer learning, de novo random survival forest, and the stacked ensemble improved calibration in MA.27 over the pretrained model (integrated calibration index reduced from 0.042 in PREDICT v3 to ≤0.007) while discrimination remained comparable (AUROC increased from 0.738 in PREDICT v3 to 0.744-0.799). In decision-curve analysis, these approaches demonstrated consistently positive net benefit across clinically relevant thresholds, while PREDICT v3 lost net benefit beyond 7.5% predicted risk. Invalid PREDICT v3 predictions were observed in 23.8% to 25.8% of MA.27 individuals due to missing information. In contrast, ML models and the stacked ensemble predicted survival despite missing data. Across all models, patient age, nodal status, pathological grading, and tumor size had the highest Shapley Additive Explanations values, indicating their importance for survival prognostication. External validation in the US Surveillance, Epidemiology, and End Results cohort confirmed the benefits of transfer learning, RSF, and ensemble in terms of calibration while maintaining discrimination at comparable levels. In contrast, generalizability was limited in the Tamoxifen Exemestane Adjuvant Multinational trial, a cohort with a substantially different distribution of clinicopathological characteristics. CONCLUSIONS This study demonstrates that transfer learning, de novo RSF, and a stacked ensemble can improve prognostication compared with the pretrained PREDICT v3, particularly in the presence of missing or uncertain inputs. Transportability may be limited in cohorts with different clinicopathological profiles, requiring local validation before clinical deployment. Ultimately, better survival estimation can provide more meaningful guidance in breast cancer care. CLINICALTRIAL ClinicalTrials.gov NCT00066573; https://clinicaltrials.gov/study/NCT00066573, NCT00279448; https://clinicaltrials.gov/study/NCT00279448, NCT00032136; https://clinicaltrials.gov/study/NCT00032136
- 29 November 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessTransfer Learning and Machine Learning for Training Five-Year Survival Prognostic Models in Early Breast Cancer: Development and Validation Study (Preprint)
BACKGROUND Prognostic information is essential for decision-making in breast cancer management. In recent years, trials and clinical practice have emphasized genomic prognostication tools, despite clinicopathological methods being more affordable and accessible. PREDICT v3 is one such tool with promising results across cohorts. Advances in machine learning (ML), transfer learning, and ensemble methods provide opportunities to enhance these approaches, especially where missing data and model assumptions differ across diverse populations. OBJECTIVE This study evaluates the potential to improve survival prognostication in breast cancer. More precisely, we compare de novo ML, transfer learning from the pretrained prognostication model PREDICT v3, and a stacked ensemble approach. METHODS Data from the MA.27 trial (NCT00066573) were used for model training, with external validation on data from the Tamoxifen Exemestane Adjuvant Multinational trial (NCT00279448 and NCT00032136) and a US Surveillance, Epidemiology, and End Results cohort. Transfer learning was applied by re-estimating the parameters of the pretrained prognostic tool PREDICT v3. De novo ML included random survival forests and extreme gradient boosting, and the ensemble was implemented using weighted linear stacking of model predictions. Internal and external validation was assessed in terms of the integrated calibration index and discrimination. Shapley Additive Explanations values were used to explain model predictions and decision-curve analysis to facilitate the interpretation of performance differences. RESULTS Transfer learning, de novo random survival forest, and the stacked ensemble improved calibration in MA.27 over the pretrained model (integrated calibration index reduced from 0.042 in PREDICT v3 to ≤0.007) while discrimination remained comparable (AUROC increased from 0.738 in PREDICT v3 to 0.744-0.799). In decision-curve analysis, these approaches demonstrated consistently positive net benefit across clinically relevant thresholds, while PREDICT v3 lost net benefit beyond 7.5% predicted risk. Invalid PREDICT v3 predictions were observed in 23.8% to 25.8% of MA.27 individuals due to missing information. In contrast, ML models and the stacked ensemble predicted survival despite missing data. Across all models, patient age, nodal status, pathological grading, and tumor size had the highest Shapley Additive Explanations values, indicating their importance for survival prognostication. External validation in the US Surveillance, Epidemiology, and End Results cohort confirmed the benefits of transfer learning, RSF, and ensemble in terms of calibration while maintaining discrimination at comparable levels. In contrast, generalizability was limited in the Tamoxifen Exemestane Adjuvant Multinational trial, a cohort with a substantially different distribution of clinicopathological characteristics. CONCLUSIONS This study demonstrates that transfer learning, de novo RSF, and a stacked ensemble can improve prognostication compared with the pretrained PREDICT v3, particularly in the presence of missing or uncertain inputs. Transportability may be limited in cohorts with different clinicopathological profiles, requiring local validation before clinical deployment. Ultimately, better survival estimation can provide more meaningful guidance in breast cancer care. CLINICALTRIAL ClinicalTrials.gov NCT00066573; https://clinicaltrials.gov/study/NCT00066573, NCT00279448; https://clinicaltrials.gov/study/NCT00279448, NCT00032136; https://clinicaltrials.gov/study/NCT00032136
- Open Access
- 18 October 2024
- Published by Oxford University Press (OUP) in JNCI: Journal of the National Cancer Institute
- Vol. 117 (3) , 486-495
Validation of a breast cancer assay for radiotherapy omission: an individual participant data meta-analysis
There are currently no molecular tests to identify individual breast cancers where radiotherapy (RT) offers no benefit. Profile for the Omission of Local Adjuvant Radiotherapy (POLAR) is a 16-gene molecular signature developed to identify low risk cancers where RT will not further reduce recurrence rates. An individual participant data meta-analysis was performed in 623 cases of node-negative ER+/HER2-negative early breast cancer enrolled in three RT randomized trials for whom primary tumor material was available for analysis. A Cox proportional hazards model on time to locoregional recurrence (LRR) was used to test the interaction between POLAR score and RT. 429 (69%) patients’ tumors had a high POLAR score and 194 (31%) had a low score. Patients with high POLAR score had, in the absence of RT, a 10-year cumulative incidence of LRR: 20% (15%-26%) vs 5% (2%-11%) for those with a low score. Patients with a high POLAR score had a large benefit from RT (hazard ratio [HR] for RT vs no RT: 0.37 [0.23-0.60], p < .001). In contrast, there was no evidence of benefit from RT for patients with a low POLAR score (HR: 0.92 [0.42-2.02], p = .832). The test for interaction between RT and POLAR was statistically significant (p = .022). POLAR is not only prognostic for locoregional recurrence but also predictive of benefit from radiotherapy in selected patients. Patients ≥ 50 years with ER+/HER2-negative disease and a low POLAR score could consider omitting adjuvant RT. Further validation in contemporary clinical cohorts is required.
- 18 October 2024
- Published by Oxford University Press (OUP) in JNCI: Journal of the National Cancer Institute
- Vol. 117 (3) , 486-495
Open AccessValidation of a breast cancer assay for radiotherapy omission: an individual participant data meta-analysis
There are currently no molecular tests to identify individual breast cancers where radiotherapy (RT) offers no benefit. Profile for the Omission of Local Adjuvant Radiotherapy (POLAR) is a 16-gene molecular signature developed to identify low risk cancers where RT will not further reduce recurrence rates. An individual participant data meta-analysis was performed in 623 cases of node-negative ER+/HER2-negative early breast cancer enrolled in three RT randomized trials for whom primary tumor material was available for analysis. A Cox proportional hazards model on time to locoregional recurrence (LRR) was used to test the interaction between POLAR score and RT. 429 (69%) patients’ tumors had a high POLAR score and 194 (31%) had a low score. Patients with high POLAR score had, in the absence of RT, a 10-year cumulative incidence of LRR: 20% (15%-26%) vs 5% (2%-11%) for those with a low score. Patients with a high POLAR score had a large benefit from RT (hazard ratio [HR] for RT vs no RT: 0.37 [0.23-0.60], p < .001). In contrast, there was no evidence of benefit from RT for patients with a low POLAR score (HR: 0.92 [0.42-2.02], p = .832). The test for interaction between RT and POLAR was statistically significant (p = .022). POLAR is not only prognostic for locoregional recurrence but also predictive of benefit from radiotherapy in selected patients. Patients ≥ 50 years with ER+/HER2-negative disease and a low POLAR score could consider omitting adjuvant RT. Further validation in contemporary clinical cohorts is required.
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Table S2 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Table 2. Distribution of patients as defined by BCI prognostic (BCI/BCIN+) and BCI predictive (BCI (H/I)).
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S2 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Table 2. Distribution of patients as defined by BCI prognostic (BCI/BCIN+) and BCI predictive (BCI (H/I)).
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Data from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Purpose:Patients with early-stage hormone receptor–positive (HR+) breast cancer face a prolonged risk of recurrence even after adjuvant endocrine therapy. The Breast Cancer Index (BCI) is significantly prognostic for overall (0–10 years) and late (5–10 years) distant recurrence (DR) risk in N0 and N1 patients. Here, BCI prognostic performance was evaluated in HR+ postmenopausal women from the Tamoxifen and Exemestane Adjuvant Multinational (TEAM) trial.Experimental Design:3,544 patients were included in the analysis (N = 1,519 N0, N = 2,025 N+). BCI risk groups were calculated using pre-specified cutoff points. Kaplan–Meier analyses and log-rank tests were used to assess the prognostic significance of BCI risk groups based on DR. Hazard ratios (HR) and confidence intervals (CI) were calculated using Cox models with and without clinical covariates.Results:For overall 10-year DR, BCI was significantly prognostic in Ni0 (N = 1,196) and N1 (N = 1,234) patients who did not receive prior chemotherapy (P < 0.001). In patients who were DR-free for 5 years, 10-year late DR rates for low- and high-risk groups were 5.4% and 9.3% (N0 cohort, N = 1,285) and 4.8% and 12.2% (N1 cohort, N = 1,625) with multivariate HRs of 2.25 (95% CI, 1.30–3.88; P = 0.004) and 2.67 (95% CI, 1.53–4.63; P < 0.001), respectively. Late DR performance was substantially improved using previously optimized cutoff points, identifying BCI low-risk groups with even lower 10-year late DR rates of 3.8% and 2.7% in N0 and N1 patients, respectively.Conclusions:The TEAM trial represents the largest prognostic validation study for BCI to date and provides a more representative assessment of late DR risk to guide individualized treatment decision-making for HR+ patients with early-stage breast cancer.
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessData from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Purpose:Patients with early-stage hormone receptor–positive (HR+) breast cancer face a prolonged risk of recurrence even after adjuvant endocrine therapy. The Breast Cancer Index (BCI) is significantly prognostic for overall (0–10 years) and late (5–10 years) distant recurrence (DR) risk in N0 and N1 patients. Here, BCI prognostic performance was evaluated in HR+ postmenopausal women from the Tamoxifen and Exemestane Adjuvant Multinational (TEAM) trial.Experimental Design:3,544 patients were included in the analysis (N = 1,519 N0, N = 2,025 N+). BCI risk groups were calculated using pre-specified cutoff points. Kaplan–Meier analyses and log-rank tests were used to assess the prognostic significance of BCI risk groups based on DR. Hazard ratios (HR) and confidence intervals (CI) were calculated using Cox models with and without clinical covariates.Results:For overall 10-year DR, BCI was significantly prognostic in Ni0 (N = 1,196) and N1 (N = 1,234) patients who did not receive prior chemotherapy (P < 0.001). In patients who were DR-free for 5 years, 10-year late DR rates for low- and high-risk groups were 5.4% and 9.3% (N0 cohort, N = 1,285) and 4.8% and 12.2% (N1 cohort, N = 1,625) with multivariate HRs of 2.25 (95% CI, 1.30–3.88; P = 0.004) and 2.67 (95% CI, 1.53–4.63; P < 0.001), respectively. Late DR performance was substantially improved using previously optimized cutoff points, identifying BCI low-risk groups with even lower 10-year late DR rates of 3.8% and 2.7% in N0 and N1 patients, respectively.Conclusions:The TEAM trial represents the largest prognostic validation study for BCI to date and provides a more representative assessment of late DR risk to guide individualized treatment decision-making for HR+ patients with early-stage breast cancer.
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Data from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Purpose:Patients with early-stage hormone receptor–positive (HR+) breast cancer face a prolonged risk of recurrence even after adjuvant endocrine therapy. The Breast Cancer Index (BCI) is significantly prognostic for overall (0–10 years) and late (5–10 years) distant recurrence (DR) risk in N0 and N1 patients. Here, BCI prognostic performance was evaluated in HR+ postmenopausal women from the Tamoxifen and Exemestane Adjuvant Multinational (TEAM) trial.Experimental Design:3,544 patients were included in the analysis (N = 1,519 N0, N = 2,025 N+). BCI risk groups were calculated using pre-specified cutoff points. Kaplan–Meier analyses and log-rank tests were used to assess the prognostic significance of BCI risk groups based on DR. Hazard ratios (HR) and confidence intervals (CI) were calculated using Cox models with and without clinical covariates.Results:For overall 10-year DR, BCI was significantly prognostic in Ni0 (N = 1,196) and N1 (N = 1,234) patients who did not receive prior chemotherapy (P < 0.001). In patients who were DR-free for 5 years, 10-year late DR rates for low- and high-risk groups were 5.4% and 9.3% (N0 cohort, N = 1,285) and 4.8% and 12.2% (N1 cohort, N = 1,625) with multivariate HRs of 2.25 (95% CI, 1.30–3.88; P = 0.004) and 2.67 (95% CI, 1.53–4.63; P < 0.001), respectively. Late DR performance was substantially improved using previously optimized cutoff points, identifying BCI low-risk groups with even lower 10-year late DR rates of 3.8% and 2.7% in N0 and N1 patients, respectively.Conclusions:The TEAM trial represents the largest prognostic validation study for BCI to date and provides a more representative assessment of late DR risk to guide individualized treatment decision-making for HR+ patients with early-stage breast cancer.
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessData from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Purpose:Patients with early-stage hormone receptor–positive (HR+) breast cancer face a prolonged risk of recurrence even after adjuvant endocrine therapy. The Breast Cancer Index (BCI) is significantly prognostic for overall (0–10 years) and late (5–10 years) distant recurrence (DR) risk in N0 and N1 patients. Here, BCI prognostic performance was evaluated in HR+ postmenopausal women from the Tamoxifen and Exemestane Adjuvant Multinational (TEAM) trial.Experimental Design:3,544 patients were included in the analysis (N = 1,519 N0, N = 2,025 N+). BCI risk groups were calculated using pre-specified cutoff points. Kaplan–Meier analyses and log-rank tests were used to assess the prognostic significance of BCI risk groups based on DR. Hazard ratios (HR) and confidence intervals (CI) were calculated using Cox models with and without clinical covariates.Results:For overall 10-year DR, BCI was significantly prognostic in Ni0 (N = 1,196) and N1 (N = 1,234) patients who did not receive prior chemotherapy (P < 0.001). In patients who were DR-free for 5 years, 10-year late DR rates for low- and high-risk groups were 5.4% and 9.3% (N0 cohort, N = 1,285) and 4.8% and 12.2% (N1 cohort, N = 1,625) with multivariate HRs of 2.25 (95% CI, 1.30–3.88; P = 0.004) and 2.67 (95% CI, 1.53–4.63; P < 0.001), respectively. Late DR performance was substantially improved using previously optimized cutoff points, identifying BCI low-risk groups with even lower 10-year late DR rates of 3.8% and 2.7% in N0 and N1 patients, respectively.Conclusions:The TEAM trial represents the largest prognostic validation study for BCI to date and provides a more representative assessment of late DR risk to guide individualized treatment decision-making for HR+ patients with early-stage breast cancer.
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Table S2 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Table 2. Distribution of patients as defined by BCI prognostic (BCI/BCIN+) and BCI predictive (BCI (H/I)).
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S2 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Table 2. Distribution of patients as defined by BCI prognostic (BCI/BCIN+) and BCI predictive (BCI (H/I)).
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S1 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Figure 1. Continuous risk curves of BCI and BCIN+ for overall 10-year DR in patients who did not receive adjuvant chemotherapy. A: risk of overall 10-year DR as a function of continuous BCI for N0 patients (N=1196); B: risk of overall 10-year DR as a function of continuous BCIN+ for N1 patients (N=1234); C: risk of overall 10-year DR as a function of continuous BCI for N0 HER2- patients (N=978); D: risk of overall 10-year DR as a function of continuous BCIN+ for N1 HER2- patients (N=1067).
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S1 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Figure 1. Continuous risk curves of BCI and BCIN+ for overall 10-year DR in patients who did not receive adjuvant chemotherapy. A: risk of overall 10-year DR as a function of continuous BCI for N0 patients (N=1196); B: risk of overall 10-year DR as a function of continuous BCIN+ for N1 patients (N=1234); C: risk of overall 10-year DR as a function of continuous BCI for N0 HER2- patients (N=978); D: risk of overall 10-year DR as a function of continuous BCIN+ for N1 HER2- patients (N=1067).
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S1 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Figure 1. Continuous risk curves of BCI and BCIN+ for overall 10-year DR in patients who did not receive adjuvant chemotherapy. A: risk of overall 10-year DR as a function of continuous BCI for N0 patients (N=1196); B: risk of overall 10-year DR as a function of continuous BCIN+ for N1 patients (N=1234); C: risk of overall 10-year DR as a function of continuous BCI for N0 HER2- patients (N=978); D: risk of overall 10-year DR as a function of continuous BCIN+ for N1 HER2- patients (N=1067).
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S1 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Figure 1. Continuous risk curves of BCI and BCIN+ for overall 10-year DR in patients who did not receive adjuvant chemotherapy. A: risk of overall 10-year DR as a function of continuous BCI for N0 patients (N=1196); B: risk of overall 10-year DR as a function of continuous BCIN+ for N1 patients (N=1234); C: risk of overall 10-year DR as a function of continuous BCI for N0 HER2- patients (N=978); D: risk of overall 10-year DR as a function of continuous BCIN+ for N1 HER2- patients (N=1067).
- Open Access
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S2 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Figure 2. Continuous risk curves of BCI and BCIN+ for late 10-year DR in patients DR-free at 5 years. A: risk of 10-year late DR as a function of continuous BCI for N0 patients (N=1285); B: risk of 10-year late DR as a function of continuous BCIN+ for N1 patients (N=1625); C: risk of 10-year late DR as a function of continuous BCI for N0 HER2- patients (N=1063); D: risk of 10-year late DR as a function of continuous BCIN+ for N1 HER2- patients (N=1394).
- 15 April 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S2 from Validation of the Prognostic Performance of Breast Cancer Index in Hormone Receptor–Positive Postmenopausal Breast Cancer Patients in the TEAM Trial
Supplemental Figure 2. Continuous risk curves of BCI and BCIN+ for late 10-year DR in patients DR-free at 5 years. A: risk of 10-year late DR as a function of continuous BCI for N0 patients (N=1285); B: risk of 10-year late DR as a function of continuous BCIN+ for N1 patients (N=1625); C: risk of 10-year late DR as a function of continuous BCI for N0 HER2- patients (N=1063); D: risk of 10-year late DR as a function of continuous BCIN+ for N1 HER2- patients (N=1394).