Full automation of total metabolic tumor volume from FDG-PET/CT in DLBCL for baseline risk assessments.
Artificial Intelligence
Automation
Clinical Trials, Phase III as Topic
Fluorodeoxyglucose F18
Humans
Lymphoma, Large B-Cell, Diffuse
/ diagnostic imaging
Neoplasm Recurrence, Local
Positron Emission Tomography Computed Tomography
/ methods
Positron-Emission Tomography
/ methods
Prognosis
Risk Assessment
Tumor Burden
AI
DLBCL
FDG-PET
Imaging
Journal
Cancer imaging : the official publication of the International Cancer Imaging Society
ISSN: 1470-7330
Titre abrégé: Cancer Imaging
Pays: England
ID NLM: 101172931
Informations de publication
Date de publication:
12 Aug 2022
12 Aug 2022
Historique:
received:
09
03
2022
accepted:
17
07
2022
entrez:
12
8
2022
pubmed:
13
8
2022
medline:
17
8
2022
Statut:
epublish
Résumé
Current radiological assessments of Our aim is to enable real-time assessment of imaging-based risk factors at a large scale and we propose a fully automatic artificial intelligence (AI)-based tool to rapidly extract FDG-PET imaging metrics in DLBCL. On availability of a scan, in combination with clinical data, our approach generates clinically informative risk scores with minimal resource requirements. Overall, 1268 patients with previously untreated DLBCL from the phase III GOYA trial (NCT01287741) were included in the analysis (training: n = 846; hold-out: n = 422). Our AI-based model comprising imaging and clinical variables yielded a tangible prognostic improvement compared to clinical models without imaging metrics. We observed a risk increase for progression-free survival (PFS) with hazard ratios [HR] of 1.87 (95% CI: 1.31-2.67) vs 1.38 (95% CI: 0.98-1.96) (C-index: 0.59 vs 0.55), and a risk increase for overall survival (OS) (HR: 2.16 (95% CI: 1.37-3.40) vs 1.40 (95% CI: 0.90-2.17); C-index: 0.59 vs 0.55). The combined model defined a high-risk population with 35% and 42% increased odds of a 4-year PFS and OS event, respectively, versus the International Prognostic Index components alone. The method also identified a subpopulation with a 2-year Central Nervous System (CNS)-relapse probability of 17.1%. Our tool enables an enhanced risk stratification compared with IPI, and the results indicate that imaging can be used to improve the prediction of central nervous system relapse in DLBCL. These findings support integration of clinically informative AI-generated imaging metrics into clinical workflows to improve identification of high-risk DLBCL patients. Registered clinicaltrials.gov number: NCT01287741.
Sections du résumé
BACKGROUND
BACKGROUND
Current radiological assessments of
METHODS
METHODS
Our aim is to enable real-time assessment of imaging-based risk factors at a large scale and we propose a fully automatic artificial intelligence (AI)-based tool to rapidly extract FDG-PET imaging metrics in DLBCL. On availability of a scan, in combination with clinical data, our approach generates clinically informative risk scores with minimal resource requirements. Overall, 1268 patients with previously untreated DLBCL from the phase III GOYA trial (NCT01287741) were included in the analysis (training: n = 846; hold-out: n = 422).
RESULTS
RESULTS
Our AI-based model comprising imaging and clinical variables yielded a tangible prognostic improvement compared to clinical models without imaging metrics. We observed a risk increase for progression-free survival (PFS) with hazard ratios [HR] of 1.87 (95% CI: 1.31-2.67) vs 1.38 (95% CI: 0.98-1.96) (C-index: 0.59 vs 0.55), and a risk increase for overall survival (OS) (HR: 2.16 (95% CI: 1.37-3.40) vs 1.40 (95% CI: 0.90-2.17); C-index: 0.59 vs 0.55). The combined model defined a high-risk population with 35% and 42% increased odds of a 4-year PFS and OS event, respectively, versus the International Prognostic Index components alone. The method also identified a subpopulation with a 2-year Central Nervous System (CNS)-relapse probability of 17.1%.
CONCLUSION
CONCLUSIONS
Our tool enables an enhanced risk stratification compared with IPI, and the results indicate that imaging can be used to improve the prediction of central nervous system relapse in DLBCL. These findings support integration of clinically informative AI-generated imaging metrics into clinical workflows to improve identification of high-risk DLBCL patients.
TRIAL REGISTRATION
BACKGROUND
Registered clinicaltrials.gov number: NCT01287741.
Identifiants
pubmed: 35962459
doi: 10.1186/s40644-022-00476-0
pii: 10.1186/s40644-022-00476-0
pmc: PMC9373298
doi:
Substances chimiques
Fluorodeoxyglucose F18
0Z5B2CJX4D
Banques de données
ClinicalTrials.gov
['NCT01287741']
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
39Informations de copyright
© 2022. The Author(s).
Références
Blood. 2019 Feb 28;133(9):919-926
pubmed: 30617197
J Hematol Oncol. 2020 Jun 6;13(1):71
pubmed: 32505213
Haematologica. 2022 Jul 01;107(7):1633-1642
pubmed: 34407602
Radiology. 2012 Aug;264(2):559-66
pubmed: 22692034
Blood. 2007 Mar 1;109(5):1857-61
pubmed: 17105812
Bone Marrow Transplant. 2016 Jan;51(1):51-7
pubmed: 26367239
Clin Cancer Res. 2005 Apr 15;11(8):2785-808
pubmed: 15837727
J Nucl Med. 2014 Oct;55(10):1591-7
pubmed: 25214642
Proc Natl Acad Sci U S A. 2013 Jan 22;110(4):1398-403
pubmed: 23292937
Cancer Imaging. 2016 Oct 18;16(1):35
pubmed: 27756360
J Clin Oncol. 2017 Nov 1;35(31):3529-3537
pubmed: 28796588
J Nucl Med. 2021 Jan;62(1):30-36
pubmed: 32532925
Blood. 2017 Oct 19;130(16):1800-1808
pubmed: 28774879
Am J Hematol. 2015 Nov;90(11):1041-6
pubmed: 26260224
EJNMMI Res. 2020 Sep 23;10(1):110
pubmed: 32965554
Q J Nucl Med Mol Imaging. 2011 Aug;55(4):469-75
pubmed: 21150862
Hematology Am Soc Hematol Educ Program. 2016 Dec 2;2016(1):366-378
pubmed: 27913503
Blood. 2014 Feb 6;123(6):837-42
pubmed: 24264230
J Digit Imaging. 2020 Aug;33(4):888-894
pubmed: 32378059
J Clin Oncol. 2022 Jul 20;40(21):2352-2360
pubmed: 35357901
N Engl J Med. 2022 Feb 17;386(7):640-654
pubmed: 34891224
N Engl J Med. 2019 Jan 3;380(1):45-56
pubmed: 30501490
J Nucl Med. 2013 Aug;54(8):1244-50
pubmed: 23674577
Ann Oncol. 2016 Jun;27(6):1095-1099
pubmed: 27002106
Cancer Imaging. 2014 Nov 29;14:34
pubmed: 25608713
Eur J Nucl Med Mol Imaging. 2016 Jul;43(7):1209-19
pubmed: 26902371
Eur J Nucl Med Mol Imaging. 2021 May;48(5):1362-1370
pubmed: 33097974
Am J Hematol. 2019 Jul;94(7):786-793
pubmed: 31006875
Lancet Oncol. 2019 May;20(5):605-606
pubmed: 30948275
Blood. 2020 Apr 16;135(16):1396-1405
pubmed: 31978225
J Stat Softw. 2011 Mar;39(5):1-13
pubmed: 27065756
Radiol Artif Intell. 2020 Sep 02;2(5):e200016
pubmed: 33937842
Pathology. 2018 Jan;50(1):74-87
pubmed: 29167021
Blood. 2013 Jul 4;122(1):61-7
pubmed: 23660958