Development and Validation of a Predictive Model for Metastatic Melanoma Patients Treated with Pembrolizumab Based on Automated Analysis of Whole-Body [
PET/CT
[18F]FDG
melanoma
prognosis
survival
whole-body
Journal
Cancers
ISSN: 2072-6694
Titre abrégé: Cancers (Basel)
Pays: Switzerland
ID NLM: 101526829
Informations de publication
Date de publication:
13 Aug 2023
13 Aug 2023
Historique:
received:
04
07
2023
revised:
04
08
2023
accepted:
11
08
2023
medline:
26
8
2023
pubmed:
26
8
2023
entrez:
26
8
2023
Statut:
epublish
Résumé
Antibodies that inhibit the programmed cell death protein 1 (PD-1) receptor offer a significant survival benefit, potentially cure (i.e., durable disease-free survival following treatment discontinuation), a substantial proportion of patients with advanced melanoma. Most patients however fail to respond to such treatment or acquire resistance. Previously, we reported that baseline total metabolic tumour volume (TMTV) determined by whole-body [18F]FDG PET/CT was independently correlated with survival and able to predict the futility of treatment. Manual delineation of [18F]FDG-avid lesions is however labour intensive and not suitable for routine use. A predictive survival model is proposed based on automated analysis of baseline, whole-body [18F]FDG images. Lesions were segmented on [18F]FDG PET/CT using a deep-learning approach and derived features were investigated through Kaplan-Meier survival estimates with univariate logrank test and Cox regression analyses. Selected parameters were evaluated in multivariate Cox survival regressors. In the development set of 69 patients, overall survival prediction based on TMTV, lactate dehydrogenase levels and presence of brain metastases achieved an area under the curve of 0.78 at one year, 0.70 at two years. No statistically significant difference was observed with respect to using manually segmented lesions. Internal validation on 31 patients yielded scores of 0.76 for one year and 0.74 for two years. Automatically extracted TMTV based on whole-body [18F]FDG PET/CT can aid in building predictive models that can support therapeutic decisions in patients treated with immune-checkpoint blockade.
Sections du résumé
BACKGROUND
BACKGROUND
Antibodies that inhibit the programmed cell death protein 1 (PD-1) receptor offer a significant survival benefit, potentially cure (i.e., durable disease-free survival following treatment discontinuation), a substantial proportion of patients with advanced melanoma. Most patients however fail to respond to such treatment or acquire resistance. Previously, we reported that baseline total metabolic tumour volume (TMTV) determined by whole-body [18F]FDG PET/CT was independently correlated with survival and able to predict the futility of treatment. Manual delineation of [18F]FDG-avid lesions is however labour intensive and not suitable for routine use. A predictive survival model is proposed based on automated analysis of baseline, whole-body [18F]FDG images.
METHODS
METHODS
Lesions were segmented on [18F]FDG PET/CT using a deep-learning approach and derived features were investigated through Kaplan-Meier survival estimates with univariate logrank test and Cox regression analyses. Selected parameters were evaluated in multivariate Cox survival regressors.
RESULTS
RESULTS
In the development set of 69 patients, overall survival prediction based on TMTV, lactate dehydrogenase levels and presence of brain metastases achieved an area under the curve of 0.78 at one year, 0.70 at two years. No statistically significant difference was observed with respect to using manually segmented lesions. Internal validation on 31 patients yielded scores of 0.76 for one year and 0.74 for two years.
CONCLUSIONS
CONCLUSIONS
Automatically extracted TMTV based on whole-body [18F]FDG PET/CT can aid in building predictive models that can support therapeutic decisions in patients treated with immune-checkpoint blockade.
Identifiants
pubmed: 37627111
pii: cancers15164083
doi: 10.3390/cancers15164083
pmc: PMC10452475
pii:
doi:
Types de publication
Journal Article
Langues
eng
Subventions
Organisme : Innoviris
ID : BHG/2017-PFS-15
Références
Comput Methods Programs Biomed. 2022 Jun;221:106902
pubmed: 35636357
J Clin Oncol. 2019 Feb 1;37(4):318-327
pubmed: 30557521
J Clin Oncol. 2022 May 1;40(13):1428-1438
pubmed: 35030011
J Clin Oncol. 2013 Jun 20;31(18):2337-46
pubmed: 23690423
J Clin Oncol. 2016 Dec;34(34):4102-4109
pubmed: 27863197
Phys Med Biol. 2022 Sep 29;67(19):
pubmed: 36096113
Med Phys. 2023 May 3;:
pubmed: 37134002
Lancet Oncol. 2023 Jan;24(1):33-44
pubmed: 36460017
J Nucl Med. 2015 Jan;56(1):38-44
pubmed: 25500829
Nat Methods. 2021 Feb;18(2):203-211
pubmed: 33288961
IEEE J Biomed Health Inform. 2022 Dec 19;PP:
pubmed: 37015600
Curr Oncol Rep. 2022 Jul;24(7):905-915
pubmed: 35347590
Sci Transl Med. 2018 Jul 18;10(450):
pubmed: 30021886
Diagnostics (Basel). 2022 Aug 30;12(9):
pubmed: 36140504
Ann Oncol. 2019 Apr 1;30(4):582-588
pubmed: 30715153
Sci Data. 2022 Oct 4;9(1):601
pubmed: 36195599
Expert Opin Pharmacother. 2019 Jun;20(9):1135-1152
pubmed: 31025594
J Digit Imaging. 2020 Aug;33(4):888-894
pubmed: 32378059
Cancers (Basel). 2021 Jan 06;13(2):
pubmed: 33418936
J Clin Oncol. 2022 Jan 10;40(2):127-137
pubmed: 34818112
Ann Intern Med. 2015 Jan 6;162(1):55-63
pubmed: 25560714
Clin Cancer Res. 2009 Dec 1;15(23):7412-20
pubmed: 19934295
Am J Public Health. 2020 May;110(5):731-733
pubmed: 32191523
Diagnostics (Basel). 2022 Feb 02;12(2):
pubmed: 35204479
Biometrics. 1988 Sep;44(3):837-45
pubmed: 3203132
IEEE Trans Pattern Anal Mach Intell. 2018 Apr;40(4):834-848
pubmed: 28463186
Ann Oncol. 2019 Jul 1;30(7):1154-1161
pubmed: 30923820
Eur J Cancer. 2021 Feb;144:182-191
pubmed: 33360855
J Clin Oncol. 2020 Nov 20;38(33):3937-3946
pubmed: 32997575
J Clin Oncol. 2018 Jun 10;36(17):1668-1674
pubmed: 29283791