Multistain deep learning for prediction of prognosis and therapy response in colorectal cancer

被引：123

作者：

Foersch, Sebastian ^{[1
]}

Glasner, Christina ^{[1
]}

Woerl, Ann-Christin ^{[1
,2
]}

Eckstein, Markus ^{[3
,4
]}

Wagner, Daniel-Christoph ^{[1
]}

Schulz, Stefan ^{[1
]}

Kellers, Franziska ^{[1
,5
]}

Fernandez, Aurelie ^{[1
]}

Tserea, Konstantina ^{[1
]}

Kloth, Michael ^{[1
]}

Hartmann, Arndt ^{[3
,4
]}

Heintz, Achim ^{[6
]}

Weichert, Wilko ^{[7
]}

Roth, Wilfried ^{[1
]}

Geppert, Carol ^{[3
,4
]}

Kather, Jakob Nikolas ^{[8
,9
,10
]}

Jesinghaus, Moritz ^{[7
]}

机构：

[1] Univ Med Ctr Mainz, Inst Pathol, Mainz, Germany

[2] Johannes Gutenberg Univ Mainz, Inst Comp Sci, Mainz, Germany

[3] Friedrich Alexander Univ Erlangen Nurnberg, Univ Hosp Erlangen, Inst Pathol, Erlangen, Germany

[4] Friedrich Alexander Univ Erlangen Nurnberg, Univ Hosp Erlangen, Comprehens Canc Ctr EMN, Erlangen, Germany

[5] Univ Hosp Schleswig Holstein, Dept Pathol, Kiel, Germany

[6] Marien Hosp Mainz, Dept Gen Visceral & Vasc Surg, Mainz, Germany

[7] Tech Univ Munich, Inst Pathol, Munich, Germany

[8] Univ Hosp RWTH Aachen, Dept Med 3, Aachen, Germany

[9] Univ Leeds, Leeds Inst Med Res St Jamess, Pathol & Data Analyt, Leeds, England

[10] Univ Hosp Marburg, Inst Pathol, Else Kroener Fresenius Ctr Digital Hlth, Marburg, Germany

来源：

NATURE MEDICINE | 2023年 / 29卷 / 02期

关键词：

TUMOR-INFILTRATING LYMPHOCYTES; SPATIAL-ORGANIZATION; IMMUNE CELLS; CLASSIFICATION; DIAGNOSIS;

D O I：

10.1038/s41591-022-02134-1

中图分类号：

Q5 [生物化学]; Q7 [分子生物学];

学科分类号：

071010 ; 081704 ;

摘要：

Although it has long been known that the immune cell composition has a strong prognostic and predictive value in colorectal cancer (CRC), scoring systems such as the immunoscore (IS) or quantification of intraepithelial lymphocytes are only slowly being adopted into clinical routine use and have their limitations. To address this we established and evaluated a multistain deep learning model (MSDLM) utilizing artificial intelligence (AI) to determine the AImmunoscore (AIS) in more than 1,000 patients with CRC. Our model had high prognostic capabilities and outperformed other clinical, molecular and immune cell-based parameters. It could also be used to predict the response to neoadjuvant therapy in patients with rectal cancer. Using an explainable AI approach, we confirmed that the MSDLM's decisions were based on established cellular patterns of anti-tumor immunity. Hence, the AIS could provide clinicians with a valuable decision-making tool based on the tumor immune microenvironment.

引用

页码：430 / 439

页数：22