Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Feb 5:216:115199.
doi: 10.1016/j.ejca.2024.115199. Epub 2024 Dec 26.

Predicting benefit from PARP inhibitors using deep learning on H&E-stained ovarian cancer slides

Frederik Marmé  1 , Eva I Krieghoff-Henning  2 , Lennard Kiehl  2 , Christoph Wies  3 , Jan Hauke  4 , Eric Hahnen  4 , Philipp Harter  5 , Philip C Schouten  6 , Tobias Brodkorb  1 , Mohamad Kayali  4 , Florian Heitz  5 , Claudio Zamagni  7 , Antonio González-Martin  8 , Isabelle Treilleux  9 , Stefan Kommoss  10 , Katharina Prieske  11 , Timo Gaiser  12 , Stefan Fröhling  13 , Isabelle Ray-Coquard  9 , Eric Pujade-Lauraine  14 , Titus J Brinker  15
Affiliations
Free article

Predicting benefit from PARP inhibitors using deep learning on H&E-stained ovarian cancer slides

Frederik Marmé et al. Eur J Cancer. .
Free article

Abstract

Purpose: Ovarian cancer patients with a Homologous Recombination Deficiency (HRD) often benefit from polyadenosine diphosphate-ribose polymerase (PARP) inhibitor maintenance therapy after response to platinum-based chemotherapy. HR status is currently analyzed via complex molecular tests. Predicting benefit from PARP inhibitors directly on histological whole slide images (WSIs) could be a fast and cheap alternative.

Patients and methods: We trained a Deep Learning (DL) model on H&E stained WSIs with "shrunken centroid" (SC) based HRD ground truth using the AGO-TR1 cohort (n = 208: 108 training, 100 test) and tested its ability to predict HRD as evaluated by the Myriad classifier and the benefit from olaparib in the PAOLA-1 cohort (n = 447) in a blinded manner.

Results: In contrast to the HRD prediction AUROC of 72 % on hold-out, our model only yielded an AUROC of 57 % external. Kaplan-Meier analysis showed that progression free survival (PFS) in the PARP inhibitor treated PAOLA-1 patients was significantly improved in the HRD positive group as defined by our model, but not in the HRD negative group. PFS improvement in PARP inhibitor-treated patients was substantially longer in our HRD positive group, hinting at a biologically meaningful prediction of benefit from PARP inhibitors.

Conclusion: Together, our results indicate that it might be possible to generate a predictor of benefit from PARP inhibitors based on the DL-mediated analysis of WSIs. However, further studies with larger cohorts and further methodological improvements will be necessary to generate a predictor with clinically useful accuracy across independent patient cohorts.

Keywords: Digital pathology; Homologous recombination deficiency (HRD); Ovarian cancer; PARP inhibitor.

PubMed Disclaimer

Conflict of interest statement

Declaration of Competing Interest TJB would like to disclose that he is the owner of Smart Health Heidelberg GmbH (Handschuhsheimer Landstr. 9/1, 69120 Heidelberg, Germany, https://smarthealth.de) which develops teledermatology mobile apps, outside of the submitted work. EPL is an employee of ARCAGY GINECO, has received honoraria from AstraZeneca, GSK and Agenus and has participated on Data Safety Monitoring Boards for Incyte, Roche and Pfizer. PH has received honoraria from Amgen, AstraZeneca, GSK, Roche, Sotio, Stryker, Zai Lab, MSD, Clovis, Eisai, Mersana, Exscientia, is a member of the Advisory Board for AstraZeneca, Roche, GSK, Clovis, Immunogen, MSD, Miltenyi, Novartis and Eisai and has received institutional research funding from AstraZeneca, Roche, GSK, Genmab, DFG, European Union, DKH, Immunogen, Seagen, Clovis, and Novartis. PCS is (one of the) named inventors on the patent 'Methods for assessing homologous recombination deficiency in ovarian cancer cells', EP4141127). IRC reports honoraria (self) from AbbVie, Agenus, Advaxis, BMS, PharmaMar, Genmab, Pfizer, AstraZeneca, Roche, GSK, MSD, Deciphera, Mersena, Merck Sereno, Novartis, Amgen, Tesaro, and Clovis; honoraria (institution) from GSK, MSD, Roche, and BMS; advisory/consulting fees from AbbVie, Agenus, Advaxis, BMS, PharmaMar, Genmab, Pfizer, AstraZeneca, Roche/Genentech, GSK, MSD, Deciphera, Mersena, Merck Serono, Novartis, Amgen, Tesaro, and Clovis; research grant/funding (self) from MSD, Roche, and BMS; research grant/funding (institution) from MSD, Roche, BMS, Novartis, AstraZeneca, and Merck Serono; travel support from Roche, AstraZeneca, and GSK. All other authors report no competing interests.

LinkOut - more resources