Comparative validation of PREDICT versions 3.1 and 2.2 for overall survival in the Dutch breast cancer population

Lara W.A. Vreven; Elfi M. Verheul; Marissa C. van Maaren; Frank Doornkamp; Robert Jan Schipper; Sabine Siesling; Paul D.P. Pharoah; Vivianne C.G. Tjan-Heijnen; Adri C. Voogd

doi:10.1016/j.breast.2025.104681

Comparative validation of PREDICT versions 3.1 and 2.2 for overall survival in the Dutch breast cancer population

Lara W.A. Vreven^*
, Elfi M. Verheul
, Marissa C. van Maaren
, Frank Doornkamp
, Robert Jan Schipper
, Sabine Siesling
, Paul D.P. Pharoah
, Vivianne C.G. Tjan-Heijnen
, Adri C. Voogd

^*Corresponding author for this work

Public Health

Research output: Contribution to journal › Article › Academic › peer-review

Abstract

Background:

PREDICT Breast is a clinical decision-support tool estimating prognosis and the absolute benefit of adjuvant systemic therapies in early breast cancer. PREDICT v2.2 is recommended in Dutch guidelines. Both v2.2 and the recently updated v3.1 have not been validated in the Dutch population. This study compares the predictive performance of PREDICT v3.1 and v2.2 for 10-year OS in Dutch breast cancer patients.

Methods:

Women diagnosed between 2005 and 2013 with primary invasive breast cancer were selected from the Netherlands Cancer Registry. Ten-year OS predictions from v2.2 and v3.1 were compared with observed OS for the overall cohort and 36 subgroups defined by oestrogen receptor (ER) status, HER2-status, age, and tumour stage. Discrimination (ability to distinguish patients with different outcomes) and calibration (agreement between predicted and observed outcomes) of both models were assessed.

Results:

Among 101,282 patients, both versions showed moderate discrimination (AUC v2.2 = 0.768; v3.1 = 0.775) and calibration (v2.2 intercept: 0.07; slope: 1.09; v3.1 intercept: 0.12, slope: 1.00). V3.1 slightly overestimated (1.9%), whereas v2.2 slightly underestimated (1.6%) 10-year OS. Across subgroups, v3.1 generally outperformed v2.2 except in patients aged >75 years, where v2.2 provided more accurate estimates. In ER-/HER- patients aged 50–75 years, v3.1 overestimated (1.5–2.8%) and v2.2 underestimated (2.8–5.3%) 10-year OS.

Conclusion:

Both PREDICT v2.2 and v3.1 accurately predict 10-year OS in Dutch breast cancer patients, with small differences between versions that vary by subgroup. No single model is optimal for all patients highlighting the need for subgroup-specific recalibration and careful interpretation when applying PREDICT.

Original language	English
Article number	104681
Journal	Breast
Volume	85
DOIs	https://doi.org/10.1016/j.breast.2025.104681
Publication status	Published - Feb 2026

Bibliographical note

Publisher Copyright:
© 2025 The Authors

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1016/j.breast.2025.104681Licence: CC BY

Comparative validation of PREDICT versions 3.1 and 2.2 for overall survival in the Dutch breast cancer populationFinal published version, 1.78 MBLicence: CC BY

Cite this

@article{ac42d6c4c3c24d9599b4b2399db94c31,

title = "Comparative validation of PREDICT versions 3.1 and 2.2 for overall survival in the Dutch breast cancer population",

abstract = "Background: PREDICT Breast is a clinical decision-support tool estimating prognosis and the absolute benefit of adjuvant systemic therapies in early breast cancer. PREDICT v2.2 is recommended in Dutch guidelines. Both v2.2 and the recently updated v3.1 have not been validated in the Dutch population. This study compares the predictive performance of PREDICT v3.1 and v2.2 for 10-year OS in Dutch breast cancer patients. Methods: Women diagnosed between 2005 and 2013 with primary invasive breast cancer were selected from the Netherlands Cancer Registry. Ten-year OS predictions from v2.2 and v3.1 were compared with observed OS for the overall cohort and 36 subgroups defined by oestrogen receptor (ER) status, HER2-status, age, and tumour stage. Discrimination (ability to distinguish patients with different outcomes) and calibration (agreement between predicted and observed outcomes) of both models were assessed. Results: Among 101,282 patients, both versions showed moderate discrimination (AUC v2.2 = 0.768; v3.1 = 0.775) and calibration (v2.2 intercept: 0.07; slope: 1.09; v3.1 intercept: 0.12, slope: 1.00). V3.1 slightly overestimated (1.9\%), whereas v2.2 slightly underestimated (1.6\%) 10-year OS. Across subgroups, v3.1 generally outperformed v2.2 except in patients aged >75 years, where v2.2 provided more accurate estimates. In ER-/HER- patients aged 50–75 years, v3.1 overestimated (1.5–2.8\%) and v2.2 underestimated (2.8–5.3\%) 10-year OS. Conclusion: Both PREDICT v2.2 and v3.1 accurately predict 10-year OS in Dutch breast cancer patients, with small differences between versions that vary by subgroup. No single model is optimal for all patients highlighting the need for subgroup-specific recalibration and careful interpretation when applying PREDICT.",

author = "Vreven, \{Lara W.A.\} and Verheul, \{Elfi M.\} and \{van Maaren\}, \{Marissa C.\} and Frank Doornkamp and Schipper, \{Robert Jan\} and Sabine Siesling and Pharoah, \{Paul D.P.\} and Tjan-Heijnen, \{Vivianne C.G.\} and Voogd, \{Adri C.\}",

note = "Publisher Copyright: {\textcopyright} 2025 The Authors",

year = "2026",

month = feb,

doi = "10.1016/j.breast.2025.104681",

language = "English",

volume = "85",

journal = "Breast",

issn = "0960-9776",

publisher = "Churchill Livingstone",

}

TY - JOUR

T1 - Comparative validation of PREDICT versions 3.1 and 2.2 for overall survival in the Dutch breast cancer population

AU - Vreven, Lara W.A.

AU - Verheul, Elfi M.

AU - van Maaren, Marissa C.

AU - Doornkamp, Frank

AU - Schipper, Robert Jan

AU - Siesling, Sabine

AU - Pharoah, Paul D.P.

AU - Tjan-Heijnen, Vivianne C.G.

AU - Voogd, Adri C.

PY - 2026/2

Y1 - 2026/2

N2 - Background: PREDICT Breast is a clinical decision-support tool estimating prognosis and the absolute benefit of adjuvant systemic therapies in early breast cancer. PREDICT v2.2 is recommended in Dutch guidelines. Both v2.2 and the recently updated v3.1 have not been validated in the Dutch population. This study compares the predictive performance of PREDICT v3.1 and v2.2 for 10-year OS in Dutch breast cancer patients. Methods: Women diagnosed between 2005 and 2013 with primary invasive breast cancer were selected from the Netherlands Cancer Registry. Ten-year OS predictions from v2.2 and v3.1 were compared with observed OS for the overall cohort and 36 subgroups defined by oestrogen receptor (ER) status, HER2-status, age, and tumour stage. Discrimination (ability to distinguish patients with different outcomes) and calibration (agreement between predicted and observed outcomes) of both models were assessed. Results: Among 101,282 patients, both versions showed moderate discrimination (AUC v2.2 = 0.768; v3.1 = 0.775) and calibration (v2.2 intercept: 0.07; slope: 1.09; v3.1 intercept: 0.12, slope: 1.00). V3.1 slightly overestimated (1.9%), whereas v2.2 slightly underestimated (1.6%) 10-year OS. Across subgroups, v3.1 generally outperformed v2.2 except in patients aged >75 years, where v2.2 provided more accurate estimates. In ER-/HER- patients aged 50–75 years, v3.1 overestimated (1.5–2.8%) and v2.2 underestimated (2.8–5.3%) 10-year OS. Conclusion: Both PREDICT v2.2 and v3.1 accurately predict 10-year OS in Dutch breast cancer patients, with small differences between versions that vary by subgroup. No single model is optimal for all patients highlighting the need for subgroup-specific recalibration and careful interpretation when applying PREDICT.

AB - Background: PREDICT Breast is a clinical decision-support tool estimating prognosis and the absolute benefit of adjuvant systemic therapies in early breast cancer. PREDICT v2.2 is recommended in Dutch guidelines. Both v2.2 and the recently updated v3.1 have not been validated in the Dutch population. This study compares the predictive performance of PREDICT v3.1 and v2.2 for 10-year OS in Dutch breast cancer patients. Methods: Women diagnosed between 2005 and 2013 with primary invasive breast cancer were selected from the Netherlands Cancer Registry. Ten-year OS predictions from v2.2 and v3.1 were compared with observed OS for the overall cohort and 36 subgroups defined by oestrogen receptor (ER) status, HER2-status, age, and tumour stage. Discrimination (ability to distinguish patients with different outcomes) and calibration (agreement between predicted and observed outcomes) of both models were assessed. Results: Among 101,282 patients, both versions showed moderate discrimination (AUC v2.2 = 0.768; v3.1 = 0.775) and calibration (v2.2 intercept: 0.07; slope: 1.09; v3.1 intercept: 0.12, slope: 1.00). V3.1 slightly overestimated (1.9%), whereas v2.2 slightly underestimated (1.6%) 10-year OS. Across subgroups, v3.1 generally outperformed v2.2 except in patients aged >75 years, where v2.2 provided more accurate estimates. In ER-/HER- patients aged 50–75 years, v3.1 overestimated (1.5–2.8%) and v2.2 underestimated (2.8–5.3%) 10-year OS. Conclusion: Both PREDICT v2.2 and v3.1 accurately predict 10-year OS in Dutch breast cancer patients, with small differences between versions that vary by subgroup. No single model is optimal for all patients highlighting the need for subgroup-specific recalibration and careful interpretation when applying PREDICT.

UR - https://www.scopus.com/pages/publications/105024878542

U2 - 10.1016/j.breast.2025.104681

DO - 10.1016/j.breast.2025.104681

M3 - Article

C2 - 41418492

AN - SCOPUS:105024878542

SN - 0960-9776

VL - 85

JO - Breast

JF - Breast

M1 - 104681

ER -

Comparative validation of PREDICT versions 3.1 and 2.2 for overall survival in the Dutch breast cancer population

Abstract

Bibliographical note

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this