Discovery of Pre-Treatment FDG PET/CT-Derived Radiomics-Based Models for Predicting Outcome in Diffuse Large B-Cell Lymphoma.
diffuse large B-cell lymphoma
lymphoma
machine learning
predictive modelling
radiomics
Journal
Cancers
ISSN: 2072-6694
Titre abrégé: Cancers (Basel)
Pays: Switzerland
ID NLM: 101526829
Informations de publication
Date de publication:
28 Mar 2022
28 Mar 2022
Historique:
received:
07
03
2022
revised:
23
03
2022
accepted:
25
03
2022
entrez:
12
4
2022
pubmed:
13
4
2022
medline:
13
4
2022
Statut:
epublish
Résumé
Approximately 30% of patients with diffuse large B-cell lymphoma (DLBCL) will have recurrence. The aim of this study was to develop a radiomic based model derived from baseline PET/CT to predict 2-year event free survival (2-EFS). Patients with DLBCL treated with R-CHOP chemotherapy undergoing pre-treatment PET/CT between January 2008 and January 2018 were included. The dataset was split into training and internal unseen test sets (ratio 80:20). A logistic regression model using metabolic tumour volume (MTV) and six different machine learning classifiers created from clinical and radiomic features derived from the baseline PET/CT were trained and tuned using four-fold cross validation. The model with the highest mean validation receiver operator characteristic (ROC) curve area under the curve (AUC) was tested on the unseen test set. 229 DLBCL patients met the inclusion criteria with 62 (27%) having 2-EFS events. The training cohort had 183 patients with 46 patients in the unseen test cohort. The model with the highest mean validation AUC combined clinical and radiomic features in a ridge regression model with a mean validation AUC of 0.75 ± 0.06 and a test AUC of 0.73. Radiomics based models demonstrate promise in predicting outcomes in DLBCL patients.
Sections du résumé
BACKGROUND
BACKGROUND
Approximately 30% of patients with diffuse large B-cell lymphoma (DLBCL) will have recurrence. The aim of this study was to develop a radiomic based model derived from baseline PET/CT to predict 2-year event free survival (2-EFS).
METHODS
METHODS
Patients with DLBCL treated with R-CHOP chemotherapy undergoing pre-treatment PET/CT between January 2008 and January 2018 were included. The dataset was split into training and internal unseen test sets (ratio 80:20). A logistic regression model using metabolic tumour volume (MTV) and six different machine learning classifiers created from clinical and radiomic features derived from the baseline PET/CT were trained and tuned using four-fold cross validation. The model with the highest mean validation receiver operator characteristic (ROC) curve area under the curve (AUC) was tested on the unseen test set.
RESULTS
RESULTS
229 DLBCL patients met the inclusion criteria with 62 (27%) having 2-EFS events. The training cohort had 183 patients with 46 patients in the unseen test cohort. The model with the highest mean validation AUC combined clinical and radiomic features in a ridge regression model with a mean validation AUC of 0.75 ± 0.06 and a test AUC of 0.73.
CONCLUSIONS
CONCLUSIONS
Radiomics based models demonstrate promise in predicting outcomes in DLBCL patients.
Identifiants
pubmed: 35406482
pii: cancers14071711
doi: 10.3390/cancers14071711
pmc: PMC8997127
pii:
doi:
Types de publication
Journal Article
Langues
eng
Subventions
Organisme : Wellcome Trust
ID : 104688
Pays : United Kingdom
Organisme : the Royal Academy of Engineering
ID : CiET1819\19
Références
J Nucl Med. 2020 Jan;61(1):40-45
pubmed: 31201248
Leuk Res. 2016 Mar;42:1-6
pubmed: 26851438
Eur J Haematol. 2015 Jun;94(6):532-9
pubmed: 25311082
J Clin Oncol. 2014 Sep 20;32(27):3059-68
pubmed: 25113753
Eur J Nucl Med Mol Imaging. 2019 Dec;46(13):2790-2799
pubmed: 31482428
Nat Rev Clin Oncol. 2022 Feb;19(2):132-146
pubmed: 34663898
Radiology. 2021 Mar;298(3):505-516
pubmed: 33399513
Clin Radiol. 2021 Jan;76(1):78.e9-78.e17
pubmed: 33036778
Eur J Nucl Med Mol Imaging. 2016 Jul;43(7):1209-19
pubmed: 26902371
Eur J Nucl Med Mol Imaging. 2018 May;45(5):680-688
pubmed: 29344718
Eur J Nucl Med Mol Imaging. 2021 Sep;48(10):3198-3220
pubmed: 33604689
Ann Hematol. 2012 May;91(5):697-703
pubmed: 22071570
Eur Radiol. 2020 Jan;30(1):523-536
pubmed: 31350588
Neuroimage. 2018 Feb 15;167:104-120
pubmed: 29155184
Clin Cancer Res. 2016 Aug 1;22(15):3801-9
pubmed: 26936916
Blood Adv. 2020 May 26;4(10):2286-2296
pubmed: 32453838
Br J Haematol. 2021 Feb;192(3):504-513
pubmed: 32621535
Blood. 2010 Sep 23;116(12):2040-5
pubmed: 20548096
Curr Treat Options Oncol. 2018 Sep 10;19(11):52
pubmed: 30203318
Eur Radiol. 2020 Aug;30(8):4623-4632
pubmed: 32248365
Mol Imaging Biol. 2020 Aug;22(4):1102-1110
pubmed: 31993925
Lancet. 2017 Jul 15;390(10091):298-310
pubmed: 28153383
PLoS One. 2015 Dec 15;10(12):e0145063
pubmed: 26669541
Oncotarget. 2017 Aug 24;8(59):99587-99600
pubmed: 29245926
Hematol Oncol. 2022 Feb;40(1):11-21
pubmed: 34714558
Ann Hematol. 2011 Jul;90(7):797-802
pubmed: 21181163
Eur J Nucl Med Mol Imaging. 2018 Sep;45(10):1672-1679
pubmed: 29705879
Eur J Nucl Med Mol Imaging. 2019 Feb;46(2):520-521
pubmed: 30386867
Nat Rev Clin Oncol. 2017 Dec;14(12):749-762
pubmed: 28975929
Eur Radiol. 2021 Jan;31(1):1-4
pubmed: 32767103
Eur J Nucl Med Mol Imaging. 2022 Mar;49(4):1298-1310
pubmed: 34651227