Incorporating progesterone receptor expression into the PREDICT breast prognostic model

ABCTB Investigators

doi:10.1016/j.ejca.2022.06.011

Incorporating progesterone receptor expression into the PREDICT breast prognostic model

ABCTB Investigators

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

Abstract

BACKGROUND: Predict Breast (www.predict.nhs.uk) is an online prognostication and treatment benefit tool for early invasive breast cancer. The aim of this study was to incorporate the prognostic effect of progesterone receptor (PR) status into a new version of PREDICT and to compare its performance to the current version (2.2).

METHOD: The prognostic effect of PR status was based on the analysis of data from 45,088 European patients with breast cancer from 49 studies in the Breast Cancer Association Consortium. Cox proportional hazard models were used to estimate the hazard ratio for PR status. Data from a New Zealand study of 11,365 patients with early invasive breast cancer were used for external validation. Model calibration and discrimination were used to test the model performance.

RESULTS: Having a PR-positive tumour was associated with a 23% and 28% lower risk of dying from breast cancer for women with oestrogen receptor (ER)-negative and ER-positive breast cancer, respectively. The area under the ROC curve increased with the addition of PR status from 0.807 to 0.809 for patients with ER-negative tumours (p = 0.023) and from 0.898 to 0.902 for patients with ER-positive tumours (p = 2.3 × 10-6) in the New Zealand cohort. Model calibration was modest with 940 observed deaths compared to 1151 predicted.

CONCLUSION: The inclusion of the prognostic effect of PR status to PREDICT Breast has led to an improvement of model performance and more accurate absolute treatment benefit predictions for individual patients. Further studies should determine whether the baseline hazard function requires recalibration.

Original language	English
Pages (from-to)	178-193
Number of pages	16
Journal	European journal of cancer (Oxford, England : 1990)
Volume	173
Early online date	4 Aug 2022
DOIs	https://doi.org/10.1016/j.ejca.2022.06.011
Publication status	Published - 1 Sept 2022

Keywords / Materials (for Non-textual outputs)

Breast Neoplasms/pathology
Female
Humans
Progesterone
Prognosis
Receptor, ErbB-2/metabolism
Receptors, Progesterone/metabolism

Access to Document

10.1016/j.ejca.2022.06.011Licence: Creative Commons: Attribution (CC-BY)

Incorporating progesterone receptor expression into the PREDICT breast prognostic modelFinal published version, 4.98 MBLicence: Creative Commons: Attribution (CC-BY)

Cite this

@article{a7b709852ea04c9ba077242db8ae8492,

title = "Incorporating progesterone receptor expression into the PREDICT breast prognostic model",

abstract = "BACKGROUND: Predict Breast (www.predict.nhs.uk) is an online prognostication and treatment benefit tool for early invasive breast cancer. The aim of this study was to incorporate the prognostic effect of progesterone receptor (PR) status into a new version of PREDICT and to compare its performance to the current version (2.2).METHOD: The prognostic effect of PR status was based on the analysis of data from 45,088 European patients with breast cancer from 49 studies in the Breast Cancer Association Consortium. Cox proportional hazard models were used to estimate the hazard ratio for PR status. Data from a New Zealand study of 11,365 patients with early invasive breast cancer were used for external validation. Model calibration and discrimination were used to test the model performance.RESULTS: Having a PR-positive tumour was associated with a 23\% and 28\% lower risk of dying from breast cancer for women with oestrogen receptor (ER)-negative and ER-positive breast cancer, respectively. The area under the ROC curve increased with the addition of PR status from 0.807 to 0.809 for patients with ER-negative tumours (p = 0.023) and from 0.898 to 0.902 for patients with ER-positive tumours (p = 2.3 × 10-6) in the New Zealand cohort. Model calibration was modest with 940 observed deaths compared to 1151 predicted.CONCLUSION: The inclusion of the prognostic effect of PR status to PREDICT Breast has led to an improvement of model performance and more accurate absolute treatment benefit predictions for individual patients. Further studies should determine whether the baseline hazard function requires recalibration.",

keywords = "Breast Neoplasms/pathology, Female, Humans, Progesterone, Prognosis, Receptor, ErbB-2/metabolism, Receptors, Progesterone/metabolism",

author = "\{ABCTB Investigators\} and Isabelle Grootes and Renske Keeman and Blows, \{Fiona M\} and Milne, \{Roger L\} and Giles, \{Graham G\} and Swerdlow, \{Anthony J\} and Fasching, \{Peter A\} and Mustapha Abubakar and Andrulis, \{Irene L\} and Hoda Anton-Culver and Beckmann, \{Matthias W\} and Carl Blomqvist and Bojesen, \{Stig E\} and Bolla, \{Manjeet K\} and Bernardo Bonanni and Ignacio Briceno and Barbara Burwinkel and Camp, \{Nicola J\} and Castelao, \{Jose E\} and Ji-Yeob Choi and Clarke, \{Christine L\} and Couch, \{Fergus J\} and Angela Cox and Cross, \{Simon S\} and Kamila Czene and Peter Devilee and Thilo D{\"o}rk and Dunning, \{Alison M\} and Miriam Dwek and Easton, \{Douglas F\} and Eccles, \{Diana M\} and Mikael Eriksson and Kristina Ernst and Evans, \{D Gareth\} and Figueroa, \{Jonine D\} and Visnja Fink and Giuseppe Floris and Stephen Fox and Marike Gabrielson and Manuela Gago-Dominguez and Garc{\'i}a-S{\'a}enz, \{Jos{\'e} A\} and Anna Gonz{\'a}lez-Neira and Lothar Haeberle and Haiman, \{Christopher A\} and Per Hall and Ute Hamann and Harkness, \{Elaine F\} and Mikael Hartman and Alexander Hein and Hooning, \{Maartje J\} and Ming-Feng Hou and Howell, \{Sacha J\} and Hidemi Ito and Anna Jakubowska and Wolfgang Janni and John, \{Esther M\} and Audrey Jung and Daehee Kang and Kristensen, \{Vessela N\} and Ava Kwong and Diether Lambrechts and Jingmei Li and Jan Lubi{\'n}ski and Mehdi Manoochehri and Sara Margolin and Keitaro Matsuo and Taib, \{Nur Aishah Mohd\} and Mulligan, \{Anna Marie\} and Heli Nevanlinna and Newman, \{William G\} and Kenneth Offit and Ana Osorio and Park, \{Sue K\} and Tjoung-Won Park-Simon and Patel, \{Alpa V\} and Nadege Presneau and Katri Pylk{\"a}s and Brigitte Rack and Paolo Radice and Gad Rennert and Atocha Romero and Emmanouil Saloustros and Sawyer, \{Elinor J\} and Andreas Schneeweiss and Fabienne Schochter and Schoemaker, \{Minouk J\} and Chen-Yang Shen and Rana Shibli and Peter Sinn and Tapper, \{William J\} and Essa Tawfiq and Teo, \{Soo Hwang\} and Teras, \{Lauren R\} and Diana Torres and Vachon, \{Celine M\} and \{van Deurzen\}, \{Carolien H M\} and Camilla Wendt and Williams, \{Justin A\} and Robert Winqvist and Mark Elwood and Schmidt, \{Marjanka K\} and Montserrat Garc{\'i}a-Closas and Pharoah, \{Paul D P\}",

year = "2022",

month = sep,

day = "1",

doi = "10.1016/j.ejca.2022.06.011",

language = "English",

volume = "173",

pages = "178--193",

journal = "European journal of cancer (Oxford, England : 1990)",

issn = "0959-8049",

publisher = "Elsevier",

}

TY - JOUR

T1 - Incorporating progesterone receptor expression into the PREDICT breast prognostic model

AU - ABCTB Investigators

AU - Grootes, Isabelle

AU - Keeman, Renske

AU - Blows, Fiona M

AU - Milne, Roger L

AU - Giles, Graham G

AU - Swerdlow, Anthony J

AU - Fasching, Peter A

AU - Abubakar, Mustapha

AU - Andrulis, Irene L

AU - Anton-Culver, Hoda

AU - Beckmann, Matthias W

AU - Blomqvist, Carl

AU - Bojesen, Stig E

AU - Bolla, Manjeet K

AU - Bonanni, Bernardo

AU - Briceno, Ignacio

AU - Burwinkel, Barbara

AU - Camp, Nicola J

AU - Castelao, Jose E

AU - Choi, Ji-Yeob

AU - Clarke, Christine L

AU - Couch, Fergus J

AU - Cox, Angela

AU - Cross, Simon S

AU - Czene, Kamila

AU - Devilee, Peter

AU - Dörk, Thilo

AU - Dunning, Alison M

AU - Dwek, Miriam

AU - Easton, Douglas F

AU - Eccles, Diana M

AU - Eriksson, Mikael

AU - Ernst, Kristina

AU - Evans, D Gareth

AU - Figueroa, Jonine D

AU - Fink, Visnja

AU - Floris, Giuseppe

AU - Fox, Stephen

AU - Gabrielson, Marike

AU - Gago-Dominguez, Manuela

AU - García-Sáenz, José A

AU - González-Neira, Anna

AU - Haeberle, Lothar

AU - Haiman, Christopher A

AU - Hall, Per

AU - Hamann, Ute

AU - Harkness, Elaine F

AU - Hartman, Mikael

AU - Hein, Alexander

AU - Hooning, Maartje J

AU - Hou, Ming-Feng

AU - Howell, Sacha J

AU - Ito, Hidemi

AU - Jakubowska, Anna

AU - Janni, Wolfgang

AU - John, Esther M

AU - Jung, Audrey

AU - Kang, Daehee

AU - Kristensen, Vessela N

AU - Kwong, Ava

AU - Lambrechts, Diether

AU - Li, Jingmei

AU - Lubiński, Jan

AU - Manoochehri, Mehdi

AU - Margolin, Sara

AU - Matsuo, Keitaro

AU - Taib, Nur Aishah Mohd

AU - Mulligan, Anna Marie

AU - Nevanlinna, Heli

AU - Newman, William G

AU - Offit, Kenneth

AU - Osorio, Ana

AU - Park, Sue K

AU - Park-Simon, Tjoung-Won

AU - Patel, Alpa V

AU - Presneau, Nadege

AU - Pylkäs, Katri

AU - Rack, Brigitte

AU - Radice, Paolo

AU - Rennert, Gad

AU - Romero, Atocha

AU - Saloustros, Emmanouil

AU - Sawyer, Elinor J

AU - Schneeweiss, Andreas

AU - Schochter, Fabienne

AU - Schoemaker, Minouk J

AU - Shen, Chen-Yang

AU - Shibli, Rana

AU - Sinn, Peter

AU - Tapper, William J

AU - Tawfiq, Essa

AU - Teo, Soo Hwang

AU - Teras, Lauren R

AU - Torres, Diana

AU - Vachon, Celine M

AU - van Deurzen, Carolien H M

AU - Wendt, Camilla

AU - Williams, Justin A

AU - Winqvist, Robert

AU - Elwood, Mark

AU - Schmidt, Marjanka K

AU - García-Closas, Montserrat

AU - Pharoah, Paul D P

PY - 2022/9/1

Y1 - 2022/9/1

N2 - BACKGROUND: Predict Breast (www.predict.nhs.uk) is an online prognostication and treatment benefit tool for early invasive breast cancer. The aim of this study was to incorporate the prognostic effect of progesterone receptor (PR) status into a new version of PREDICT and to compare its performance to the current version (2.2).METHOD: The prognostic effect of PR status was based on the analysis of data from 45,088 European patients with breast cancer from 49 studies in the Breast Cancer Association Consortium. Cox proportional hazard models were used to estimate the hazard ratio for PR status. Data from a New Zealand study of 11,365 patients with early invasive breast cancer were used for external validation. Model calibration and discrimination were used to test the model performance.RESULTS: Having a PR-positive tumour was associated with a 23% and 28% lower risk of dying from breast cancer for women with oestrogen receptor (ER)-negative and ER-positive breast cancer, respectively. The area under the ROC curve increased with the addition of PR status from 0.807 to 0.809 for patients with ER-negative tumours (p = 0.023) and from 0.898 to 0.902 for patients with ER-positive tumours (p = 2.3 × 10-6) in the New Zealand cohort. Model calibration was modest with 940 observed deaths compared to 1151 predicted.CONCLUSION: The inclusion of the prognostic effect of PR status to PREDICT Breast has led to an improvement of model performance and more accurate absolute treatment benefit predictions for individual patients. Further studies should determine whether the baseline hazard function requires recalibration.

AB - BACKGROUND: Predict Breast (www.predict.nhs.uk) is an online prognostication and treatment benefit tool for early invasive breast cancer. The aim of this study was to incorporate the prognostic effect of progesterone receptor (PR) status into a new version of PREDICT and to compare its performance to the current version (2.2).METHOD: The prognostic effect of PR status was based on the analysis of data from 45,088 European patients with breast cancer from 49 studies in the Breast Cancer Association Consortium. Cox proportional hazard models were used to estimate the hazard ratio for PR status. Data from a New Zealand study of 11,365 patients with early invasive breast cancer were used for external validation. Model calibration and discrimination were used to test the model performance.RESULTS: Having a PR-positive tumour was associated with a 23% and 28% lower risk of dying from breast cancer for women with oestrogen receptor (ER)-negative and ER-positive breast cancer, respectively. The area under the ROC curve increased with the addition of PR status from 0.807 to 0.809 for patients with ER-negative tumours (p = 0.023) and from 0.898 to 0.902 for patients with ER-positive tumours (p = 2.3 × 10-6) in the New Zealand cohort. Model calibration was modest with 940 observed deaths compared to 1151 predicted.CONCLUSION: The inclusion of the prognostic effect of PR status to PREDICT Breast has led to an improvement of model performance and more accurate absolute treatment benefit predictions for individual patients. Further studies should determine whether the baseline hazard function requires recalibration.

KW - Breast Neoplasms/pathology

KW - Female

KW - Humans

KW - Progesterone

KW - Prognosis

KW - Receptor, ErbB-2/metabolism

KW - Receptors, Progesterone/metabolism

U2 - 10.1016/j.ejca.2022.06.011

DO - 10.1016/j.ejca.2022.06.011

M3 - Article

C2 - 35933885

SN - 0959-8049

VL - 173

SP - 178

EP - 193

JO - European journal of cancer (Oxford, England : 1990)

JF - European journal of cancer (Oxford, England : 1990)

ER -

Incorporating progesterone receptor expression into the PREDICT breast prognostic model

Abstract

Keywords / Materials (for Non-textual outputs)

Access to Document

Fingerprint

Cite this