Prediction of recurrence risk in endometrial cancer with multimodal deep learning

Sarah Volinsky-Fremond; Nanda Horeweg; Sonali Andani; Jurriaan Barkey Wolf; Maxime W Lafarge; Cor D de Kroon; Gitte Ørtoft; Estrid Høgdall; Jouke Dijkstra; Jan J Jobsen; Ludy C H W Lutgens; Melanie E Powell; Linda R Mileshkin; Helen Mackay; Alexandra Leary; Dionyssios Katsaros; Hans W Nijman; Stephanie M de Boer; Remi A Nout; Marco de Bruyn; David Church; Vincent T H B M Smit; Carien L Creutzberg; Viktor H Koelzer; Tjalling Bosse

doi:10.1038/s41591-024-02993-w

Prediction of recurrence risk in endometrial cancer with multimodal deep learning

Sarah Volinsky-Fremond, Nanda Horeweg, Sonali Andani, Jurriaan Barkey Wolf, Maxime W Lafarge, Cor D de Kroon, Gitte Ørtoft, Estrid Høgdall, Jouke Dijkstra, Jan J Jobsen, Ludy C H W Lutgens, Melanie E Powell, Linda R Mileshkin, Helen Mackay, Alexandra Leary, Dionyssios Katsaros, Hans W Nijman, Stephanie M de Boer, Remi A Nout, Marco de BruynDavid Church, Vincent T H B M Smit, Carien L Creutzberg, Viktor H Koelzer, Tjalling Bosse^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › Academic › peer-review

19 Citations (Scopus)

597 Downloads (Pure)

Abstract

Predicting distant recurrence of endometrial cancer (EC) is crucial for personalized adjuvant treatment. The current gold standard of combined pathological and molecular profiling is costly, hampering implementation. Here we developed HECTOR (histopathology-based endometrial cancer tailored outcome risk), a multimodal deep learning prognostic model using hematoxylin and eosin-stained, whole-slide images and tumor stage as input, on 2,072 patients from eight EC cohorts including the PORTEC-1/-2/-3 randomized trials. HECTOR demonstrated C-indices in internal (n = 353) and two external (n = 160 and n = 151) test sets of 0.789, 0.828 and 0.815, respectively, outperforming the current gold standard, and identified patients with markedly different outcomes (10-year distant recurrence-free probabilities of 97.0%, 77.7% and 58.1% for HECTOR low-, intermediate- and high-risk groups, respectively, by Kaplan-Meier analysis). HECTOR also predicted adjuvant chemotherapy benefit better than current methods. Morphological and genomic feature extraction identified correlates of HECTOR risk groups, some with therapeutic potential. HECTOR improves on the current gold standard and may help delivery of personalized treatment in EC.

Original language	English
Pages (from-to)	1962–1973
Number of pages	25
Journal	Nature Medicine
Volume	30
Early online date	24-May-2024
DOIs	https://doi.org/10.1038/s41591-024-02993-w
Publication status	Published - 2024

Access to Document

10.1038/s41591-024-02993-wLicence: CC BY

Prediction of recurrence risk in endometrial cancer with multimodal deep learningFinal publisher's version, 24.8 MBLicence: CC BY

Handle.net

Persistent link

19 Citations
1 Erratum

Correction to: Prediction of recurrence risk in endometrial cancer with multimodal deep learning (Nature Medicine, (2024), 10.1038/s41591-024-02993-w)
Volinsky-Fremond, S., Horeweg, N., Andani, S., Barkey Wolf, J., Lafarge, M. W., de Kroon, C. D., Ørtoft, G., Høgdall, E., Dijkstra, J., Jobsen, J. J., Lutgens, L. C. H. W., Powell, M. E., Mileshkin, L. R., Mackay, H., Leary, A., Katsaros, D., Nijman, H. W., de Boer, S. M., Nout, R. A. & de Bruyn, M. & 5 others, Church, D., Smit, V. T. H. B. M., Creutzberg, C. L., Koelzer, V. H. & Bosse, T., Jul-2024, In: Nature Medicine. 30, 7, p. 2092 1 p.
Research output: Contribution to journal › Erratum

Open Access
File
1 Citation (Scopus)

40 Downloads (Pure)

Cite this

Volinsky-Fremond, S., Horeweg, N., Andani, S., Barkey Wolf, J., Lafarge, M. W., de Kroon, C. D., Ørtoft, G., Høgdall, E., Dijkstra, J., Jobsen, J. J., Lutgens, L. C. H. W., Powell, M. E., Mileshkin, L. R., Mackay, H., Leary, A., Katsaros, D., Nijman, H. W., de Boer, S. M., Nout, R. A., ... Bosse, T. (2024). Prediction of recurrence risk in endometrial cancer with multimodal deep learning. Nature Medicine, 30, 1962–1973. https://doi.org/10.1038/s41591-024-02993-w

@article{fd409c3910b048379b0d69c3a6e3b10d,

title = "Prediction of recurrence risk in endometrial cancer with multimodal deep learning",

abstract = "Predicting distant recurrence of endometrial cancer (EC) is crucial for personalized adjuvant treatment. The current gold standard of combined pathological and molecular profiling is costly, hampering implementation. Here we developed HECTOR (histopathology-based endometrial cancer tailored outcome risk), a multimodal deep learning prognostic model using hematoxylin and eosin-stained, whole-slide images and tumor stage as input, on 2,072 patients from eight EC cohorts including the PORTEC-1/-2/-3 randomized trials. HECTOR demonstrated C-indices in internal (n = 353) and two external (n = 160 and n = 151) test sets of 0.789, 0.828 and 0.815, respectively, outperforming the current gold standard, and identified patients with markedly different outcomes (10-year distant recurrence-free probabilities of 97.0%, 77.7% and 58.1% for HECTOR low-, intermediate- and high-risk groups, respectively, by Kaplan-Meier analysis). HECTOR also predicted adjuvant chemotherapy benefit better than current methods. Morphological and genomic feature extraction identified correlates of HECTOR risk groups, some with therapeutic potential. HECTOR improves on the current gold standard and may help delivery of personalized treatment in EC.",

author = "Sarah Volinsky-Fremond and Nanda Horeweg and Sonali Andani and {Barkey Wolf}, Jurriaan and Lafarge, {Maxime W} and {de Kroon}, {Cor D} and Gitte {\O}rtoft and Estrid H{\o}gdall and Jouke Dijkstra and Jobsen, {Jan J} and Lutgens, {Ludy C H W} and Powell, {Melanie E} and Mileshkin, {Linda R} and Helen Mackay and Alexandra Leary and Dionyssios Katsaros and Nijman, {Hans W} and {de Boer}, {Stephanie M} and Nout, {Remi A} and {de Bruyn}, Marco and David Church and Smit, {Vincent T H B M} and Creutzberg, {Carien L} and Koelzer, {Viktor H} and Tjalling Bosse",

note = "{\textcopyright} 2024. The Author(s).",

year = "2024",

doi = "10.1038/s41591-024-02993-w",

language = "English",

volume = "30",

pages = "1962–1973",

journal = "Nature Medicine",

issn = "1078-8956",

publisher = "Nature Publishing Group",

}

Volinsky-Fremond, S, Horeweg, N, Andani, S, Barkey Wolf, J, Lafarge, MW, de Kroon, CD, Ørtoft, G, Høgdall, E, Dijkstra, J, Jobsen, JJ, Lutgens, LCHW, Powell, ME, Mileshkin, LR, Mackay, H, Leary, A, Katsaros, D, Nijman, HW, de Boer, SM, Nout, RA, de Bruyn, M, Church, D, Smit, VTHBM, Creutzberg, CL, Koelzer, VH & Bosse, T 2024, 'Prediction of recurrence risk in endometrial cancer with multimodal deep learning', Nature Medicine, vol. 30, pp. 1962–1973. https://doi.org/10.1038/s41591-024-02993-w

TY - JOUR

T1 - Prediction of recurrence risk in endometrial cancer with multimodal deep learning

AU - Volinsky-Fremond, Sarah

AU - Horeweg, Nanda

AU - Andani, Sonali

AU - Barkey Wolf, Jurriaan

AU - Lafarge, Maxime W

AU - de Kroon, Cor D

AU - Ørtoft, Gitte

AU - Høgdall, Estrid

AU - Dijkstra, Jouke

AU - Jobsen, Jan J

AU - Lutgens, Ludy C H W

AU - Powell, Melanie E

AU - Mileshkin, Linda R

AU - Mackay, Helen

AU - Leary, Alexandra

AU - Katsaros, Dionyssios

AU - Nijman, Hans W

AU - de Boer, Stephanie M

AU - Nout, Remi A

AU - de Bruyn, Marco

AU - Church, David

AU - Smit, Vincent T H B M

AU - Creutzberg, Carien L

AU - Koelzer, Viktor H

AU - Bosse, Tjalling

PY - 2024

Y1 - 2024

N2 - Predicting distant recurrence of endometrial cancer (EC) is crucial for personalized adjuvant treatment. The current gold standard of combined pathological and molecular profiling is costly, hampering implementation. Here we developed HECTOR (histopathology-based endometrial cancer tailored outcome risk), a multimodal deep learning prognostic model using hematoxylin and eosin-stained, whole-slide images and tumor stage as input, on 2,072 patients from eight EC cohorts including the PORTEC-1/-2/-3 randomized trials. HECTOR demonstrated C-indices in internal (n = 353) and two external (n = 160 and n = 151) test sets of 0.789, 0.828 and 0.815, respectively, outperforming the current gold standard, and identified patients with markedly different outcomes (10-year distant recurrence-free probabilities of 97.0%, 77.7% and 58.1% for HECTOR low-, intermediate- and high-risk groups, respectively, by Kaplan-Meier analysis). HECTOR also predicted adjuvant chemotherapy benefit better than current methods. Morphological and genomic feature extraction identified correlates of HECTOR risk groups, some with therapeutic potential. HECTOR improves on the current gold standard and may help delivery of personalized treatment in EC.

AB - Predicting distant recurrence of endometrial cancer (EC) is crucial for personalized adjuvant treatment. The current gold standard of combined pathological and molecular profiling is costly, hampering implementation. Here we developed HECTOR (histopathology-based endometrial cancer tailored outcome risk), a multimodal deep learning prognostic model using hematoxylin and eosin-stained, whole-slide images and tumor stage as input, on 2,072 patients from eight EC cohorts including the PORTEC-1/-2/-3 randomized trials. HECTOR demonstrated C-indices in internal (n = 353) and two external (n = 160 and n = 151) test sets of 0.789, 0.828 and 0.815, respectively, outperforming the current gold standard, and identified patients with markedly different outcomes (10-year distant recurrence-free probabilities of 97.0%, 77.7% and 58.1% for HECTOR low-, intermediate- and high-risk groups, respectively, by Kaplan-Meier analysis). HECTOR also predicted adjuvant chemotherapy benefit better than current methods. Morphological and genomic feature extraction identified correlates of HECTOR risk groups, some with therapeutic potential. HECTOR improves on the current gold standard and may help delivery of personalized treatment in EC.

U2 - 10.1038/s41591-024-02993-w

DO - 10.1038/s41591-024-02993-w

M3 - Article

C2 - 38789645

SN - 1078-8956

VL - 30

SP - 1962

EP - 1973

JO - Nature Medicine

JF - Nature Medicine

ER -

Prediction of recurrence risk in endometrial cancer with multimodal deep learning

Abstract

Access to Document

Handle.net

Fingerprint

Research output

Correction to: Prediction of recurrence risk in endometrial cancer with multimodal deep learning (Nature Medicine, (2024), 10.1038/s41591-024-02993-w)

Cite this