Multimodal survival prediction in advanced pancreatic cancer using machine learning
Standard
Multimodal survival prediction in advanced pancreatic cancer using machine learning. / Keyl, J; Kasper, S; Wiesweg, M; Götze, J; Schönrock, M; Sinn, M; Berger, A; Nasca, E; Kostbade, K; Schumacher, B; Markus, P; Albers, D; Treckmann, J; Schmid, K W; Schildhaus, H-U; Siveke, J T; Schuler, M; Kleesiek, J.
In: ESMO OPEN, Vol. 7, No. 5, 100555, 10.2022.Research output: SCORING: Contribution to journal › SCORING: Journal article › Research › peer-review
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - JOUR
T1 - Multimodal survival prediction in advanced pancreatic cancer using machine learning
AU - Keyl, J
AU - Kasper, S
AU - Wiesweg, M
AU - Götze, J
AU - Schönrock, M
AU - Sinn, M
AU - Berger, A
AU - Nasca, E
AU - Kostbade, K
AU - Schumacher, B
AU - Markus, P
AU - Albers, D
AU - Treckmann, J
AU - Schmid, K W
AU - Schildhaus, H-U
AU - Siveke, J T
AU - Schuler, M
AU - Kleesiek, J
N1 - Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.
PY - 2022/10
Y1 - 2022/10
N2 - BACKGROUND: Existing risk scores appear insufficient to assess the individual survival risk of patients with advanced pancreatic ductal adenocarcinoma (PDAC) and do not take advantage of the variety of parameters that are collected during clinical care.METHODS: In this retrospective study, we built a random survival forest model from clinical data of 203 patients with advanced PDAC. The parameters were assessed before initiation of systemic treatment and included age, CA19-9, C-reactive protein, metastatic status, neutrophil-to-lymphocyte ratio and total serum protein level. Separate models including imaging and molecular parameters were built for subgroups.RESULTS: Over the entire cohort, a model based on clinical parameters achieved a c-index of 0.71. Our approach outperformed the American Joint Committee on Cancer (AJCC) staging system and the modified Glasgow Prognostic Score (mGPS) in the identification of high- and low-risk subgroups. Inclusion of the KRAS p.G12D mutational status could further improve the prediction, whereas radiomics data of the primary tumor only showed little benefit. In an external validation cohort of PDAC patients with liver metastases, our model achieved a c-index of 0.67 (mGPS: 0.59).CONCLUSIONS: The combination of multimodal data and machine-learning algorithms holds potential for personalized prognostication in advanced PDAC already at diagnosis.
AB - BACKGROUND: Existing risk scores appear insufficient to assess the individual survival risk of patients with advanced pancreatic ductal adenocarcinoma (PDAC) and do not take advantage of the variety of parameters that are collected during clinical care.METHODS: In this retrospective study, we built a random survival forest model from clinical data of 203 patients with advanced PDAC. The parameters were assessed before initiation of systemic treatment and included age, CA19-9, C-reactive protein, metastatic status, neutrophil-to-lymphocyte ratio and total serum protein level. Separate models including imaging and molecular parameters were built for subgroups.RESULTS: Over the entire cohort, a model based on clinical parameters achieved a c-index of 0.71. Our approach outperformed the American Joint Committee on Cancer (AJCC) staging system and the modified Glasgow Prognostic Score (mGPS) in the identification of high- and low-risk subgroups. Inclusion of the KRAS p.G12D mutational status could further improve the prediction, whereas radiomics data of the primary tumor only showed little benefit. In an external validation cohort of PDAC patients with liver metastases, our model achieved a c-index of 0.67 (mGPS: 0.59).CONCLUSIONS: The combination of multimodal data and machine-learning algorithms holds potential for personalized prognostication in advanced PDAC already at diagnosis.
KW - Humans
KW - C-Reactive Protein
KW - Retrospective Studies
KW - CA-19-9 Antigen
KW - Proto-Oncogene Proteins p21(ras)
KW - Neoplasm Staging
KW - Prognosis
KW - Pancreatic Neoplasms/diagnosis
KW - Adenocarcinoma/pathology
KW - Machine Learning
U2 - 10.1016/j.esmoop.2022.100555
DO - 10.1016/j.esmoop.2022.100555
M3 - SCORING: Journal article
C2 - 35988455
VL - 7
JO - ESMO OPEN
JF - ESMO OPEN
SN - 2059-7029
IS - 5
M1 - 100555
ER -