The components of progression as explanatory variables for overall survival in the Response Evaluation Criteria in Solid Tumours 1.1 database

Saskia Litière, Elisabeth G E De Vries, Lesley Seymour, Dan Sargent, Lalitha Shankar, Jan Bogaerts

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Purpose Progressive disease (PD) per Response Evaluation Criteria in Solid Tumours (RECIST) 1.1 is defined as growth of measurable target lesions, presence of new lesions or unequivocal progression of non-target disease. In this manuscript we explored whether a more refined categorisation of tumour response and/or these components of progression, varying over time, can improve prediction of overall survival (OS) in the RECIST database. Methods Data were randomly selected from 13 randomised clinical trials (3758 patients with breast, lung or colorectal cancer). A maximum of five target lesions contributed to the sum of longest diameters. At each measurement time we determined: best target response as best % improvement from baseline; tumour growth of target lesions as worst % change and worst rate of increase (mm/week) from nadir; presence of new lesions and occurrence of non-target PD. OS was analysed by tumour type using Cox regression, adjusting for baseline sum and including these parameters as time-dependent covariates. Results 36% of patients had new lesions, 28% non-target PD and 49% experienced target lesion growth (median strongest growth 1.5 mm/week). Regardless of tumour type, presence of new lesions (hazard ratio (HR) ranging 1.5-2.3) and non-target PD (HR 1.5-2.0) were strongly associated with worse OS. The explanatory value of tumour growth for OS was low compared to the other components. Conclusion Modelling target lesion tumour growth did not show a marked improvement in OS prediction over and above the other components. These analyses enable a better understanding of the role of each component in PD evaluation. Work is ongoing to incorporate this information into an updated version of RECIST with enhanced prediction of subsequent survival.

Original languageEnglish (US)
Pages (from-to)1847-1853
Number of pages7
JournalEuropean Journal of Cancer
Volume50
Issue number10
DOIs
StatePublished - 2014

Fingerprint

Databases
Survival
Growth
Neoplasms
Response Evaluation Criteria in Solid Tumors
Colorectal Neoplasms
Lung Neoplasms
Randomized Controlled Trials
Breast Neoplasms

Keywords

  • Breast cancer
  • Colorectal cancer
  • Goodness-of-fit
  • Lung cancer
  • Time-dependent model
  • Tumour growth

ASJC Scopus subject areas

  • Cancer Research
  • Oncology

Cite this

The components of progression as explanatory variables for overall survival in the Response Evaluation Criteria in Solid Tumours 1.1 database. / Litière, Saskia; De Vries, Elisabeth G E; Seymour, Lesley; Sargent, Dan; Shankar, Lalitha; Bogaerts, Jan.

In: European Journal of Cancer, Vol. 50, No. 10, 2014, p. 1847-1853.

Research output: Contribution to journalArticle

Litière, Saskia ; De Vries, Elisabeth G E ; Seymour, Lesley ; Sargent, Dan ; Shankar, Lalitha ; Bogaerts, Jan. / The components of progression as explanatory variables for overall survival in the Response Evaluation Criteria in Solid Tumours 1.1 database. In: European Journal of Cancer. 2014 ; Vol. 50, No. 10. pp. 1847-1853.
@article{d40efc7731c14191ae8282f4913abca2,
title = "The components of progression as explanatory variables for overall survival in the Response Evaluation Criteria in Solid Tumours 1.1 database",
abstract = "Purpose Progressive disease (PD) per Response Evaluation Criteria in Solid Tumours (RECIST) 1.1 is defined as growth of measurable target lesions, presence of new lesions or unequivocal progression of non-target disease. In this manuscript we explored whether a more refined categorisation of tumour response and/or these components of progression, varying over time, can improve prediction of overall survival (OS) in the RECIST database. Methods Data were randomly selected from 13 randomised clinical trials (3758 patients with breast, lung or colorectal cancer). A maximum of five target lesions contributed to the sum of longest diameters. At each measurement time we determined: best target response as best {\%} improvement from baseline; tumour growth of target lesions as worst {\%} change and worst rate of increase (mm/week) from nadir; presence of new lesions and occurrence of non-target PD. OS was analysed by tumour type using Cox regression, adjusting for baseline sum and including these parameters as time-dependent covariates. Results 36{\%} of patients had new lesions, 28{\%} non-target PD and 49{\%} experienced target lesion growth (median strongest growth 1.5 mm/week). Regardless of tumour type, presence of new lesions (hazard ratio (HR) ranging 1.5-2.3) and non-target PD (HR 1.5-2.0) were strongly associated with worse OS. The explanatory value of tumour growth for OS was low compared to the other components. Conclusion Modelling target lesion tumour growth did not show a marked improvement in OS prediction over and above the other components. These analyses enable a better understanding of the role of each component in PD evaluation. Work is ongoing to incorporate this information into an updated version of RECIST with enhanced prediction of subsequent survival.",
keywords = "Breast cancer, Colorectal cancer, Goodness-of-fit, Lung cancer, Time-dependent model, Tumour growth",
author = "Saskia Liti{\`e}re and {De Vries}, {Elisabeth G E} and Lesley Seymour and Dan Sargent and Lalitha Shankar and Jan Bogaerts",
year = "2014",
doi = "10.1016/j.ejca.2014.03.014",
language = "English (US)",
volume = "50",
pages = "1847--1853",
journal = "European Journal of Cancer",
issn = "0959-8049",
publisher = "Elsevier Limited",
number = "10",

}

TY - JOUR

T1 - The components of progression as explanatory variables for overall survival in the Response Evaluation Criteria in Solid Tumours 1.1 database

AU - Litière, Saskia

AU - De Vries, Elisabeth G E

AU - Seymour, Lesley

AU - Sargent, Dan

AU - Shankar, Lalitha

AU - Bogaerts, Jan

PY - 2014

Y1 - 2014

N2 - Purpose Progressive disease (PD) per Response Evaluation Criteria in Solid Tumours (RECIST) 1.1 is defined as growth of measurable target lesions, presence of new lesions or unequivocal progression of non-target disease. In this manuscript we explored whether a more refined categorisation of tumour response and/or these components of progression, varying over time, can improve prediction of overall survival (OS) in the RECIST database. Methods Data were randomly selected from 13 randomised clinical trials (3758 patients with breast, lung or colorectal cancer). A maximum of five target lesions contributed to the sum of longest diameters. At each measurement time we determined: best target response as best % improvement from baseline; tumour growth of target lesions as worst % change and worst rate of increase (mm/week) from nadir; presence of new lesions and occurrence of non-target PD. OS was analysed by tumour type using Cox regression, adjusting for baseline sum and including these parameters as time-dependent covariates. Results 36% of patients had new lesions, 28% non-target PD and 49% experienced target lesion growth (median strongest growth 1.5 mm/week). Regardless of tumour type, presence of new lesions (hazard ratio (HR) ranging 1.5-2.3) and non-target PD (HR 1.5-2.0) were strongly associated with worse OS. The explanatory value of tumour growth for OS was low compared to the other components. Conclusion Modelling target lesion tumour growth did not show a marked improvement in OS prediction over and above the other components. These analyses enable a better understanding of the role of each component in PD evaluation. Work is ongoing to incorporate this information into an updated version of RECIST with enhanced prediction of subsequent survival.

AB - Purpose Progressive disease (PD) per Response Evaluation Criteria in Solid Tumours (RECIST) 1.1 is defined as growth of measurable target lesions, presence of new lesions or unequivocal progression of non-target disease. In this manuscript we explored whether a more refined categorisation of tumour response and/or these components of progression, varying over time, can improve prediction of overall survival (OS) in the RECIST database. Methods Data were randomly selected from 13 randomised clinical trials (3758 patients with breast, lung or colorectal cancer). A maximum of five target lesions contributed to the sum of longest diameters. At each measurement time we determined: best target response as best % improvement from baseline; tumour growth of target lesions as worst % change and worst rate of increase (mm/week) from nadir; presence of new lesions and occurrence of non-target PD. OS was analysed by tumour type using Cox regression, adjusting for baseline sum and including these parameters as time-dependent covariates. Results 36% of patients had new lesions, 28% non-target PD and 49% experienced target lesion growth (median strongest growth 1.5 mm/week). Regardless of tumour type, presence of new lesions (hazard ratio (HR) ranging 1.5-2.3) and non-target PD (HR 1.5-2.0) were strongly associated with worse OS. The explanatory value of tumour growth for OS was low compared to the other components. Conclusion Modelling target lesion tumour growth did not show a marked improvement in OS prediction over and above the other components. These analyses enable a better understanding of the role of each component in PD evaluation. Work is ongoing to incorporate this information into an updated version of RECIST with enhanced prediction of subsequent survival.

KW - Breast cancer

KW - Colorectal cancer

KW - Goodness-of-fit

KW - Lung cancer

KW - Time-dependent model

KW - Tumour growth

UR - http://www.scopus.com/inward/record.url?scp=84901789986&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901789986&partnerID=8YFLogxK

U2 - 10.1016/j.ejca.2014.03.014

DO - 10.1016/j.ejca.2014.03.014

M3 - Article

C2 - 24726734

AN - SCOPUS:84901789986

VL - 50

SP - 1847

EP - 1853

JO - European Journal of Cancer

JF - European Journal of Cancer

SN - 0959-8049

IS - 10

ER -