RT Journal Article
SR Electronic
T1 Collaborative Federated Learning behind Hospitals’ Firewalls for Predicting Histological Response to Neoadjuvant Chemotherapy in Triple-Negative Breast Cancer
JF medRxiv
FD Cold Spring Harbor Laboratory Press
SP 2021.10.27.21264834
DO 10.1101/2021.10.27.21264834
A1 du Terrail, Jean Ogier
A1 Léopold, Armand
A1 Joly, Clément
A1 Beguier, Constance
A1 Andreux, Mathieu
A1 Maussion, Charles
A1 Schmauch, Benoît
A1 Tramel, Eric W.
A1 Bendjebbar, Etienne
A1 Zaslavskiy, Mikhail
A1 Wainrib, Gilles
A1 Milder, Maud
A1 Gervasoni, Julie
A1 Guérin, Julien
A1 Durand, Thierry
A1 Livartowski, Alain
A1 Moutet, Kelvin
A1 Gautier, Clément
A1 Djafar, Inal
A1 Moisson, Anne-Laure
A1 Marini, Camille
A1 Galtier, Mathieu
A1 Bataillon, Guillaume
A1 Heudel, Pierre-Etienne
YR 2021
UL http://medrxiv.org/content/early/2021/10/28/2021.10.27.21264834.abstract
AB Triple-Negative Breast Cancer (TNBC) is a rare cancer, characterized by high metastatic potential and poor prognosis, and has limited treatment options compared to other breast cancers. The current standard of care in non-metastatic settings is neoadjuvant chemotherapy (NACT), with the goal of breast-conserving surgery and for an in vivo assessment of chemosensitivity. However, the efficacy of this treatment varies significantly across patients, and this histological response heterogeneity is still poorly understood partly due to the paucity of available curated TNBC data. Motivated by this problem, we investigate the use of machine learning (ML) to predict at diagnosis the histological response to NACT for early TNBC patients. To overcome the known biases of related small scale studies while respecting data privacy, we conduct, for the first time, a TNBC study in a multi-centric fashion behind hospitals’ firewalls using collaborative Federated Learning (FL). Thereby allowing access to enough TNBC data to sustain a complete response heterogeneity investigation. We show evidence that local ML models relying on Whole-Slide Images (WSIs) at diagnosis are able to predict the histological response to NACT as accurately as current clinical approaches, which rely on time-consuming expert annotations. We demonstrate that collaborative training further improves performance over single-center training outperforming clinical methods. Our ML model is interpretable by design, and we show that it is sensitive to specific histological patterns. While we identify known predictive biomarkers among them, this proof of concept for real-world collaborative FL paves the way for future biomarker discovery using unprecedently large datasets.Competing Interest StatementFundings: As described in the acknowledgments this work received the financial support of the French BPI through the HealthChain project (https://www.substra.ai/en/healthchain-project). Some authors are employed by Owkin, Inc.. (J.O.d.T., C.B., M.A., Ch.M., B.S., E.B., M.Z., G.W., K.M., C.G., I.D., A.-L.M., Ca.M., M.G.) Personal fundings: G.W. is Co-founder of Owkin. P.-E.H. reports grants, personal fees and non-financial support from PFIZER, grants and non-financial support from NOVARTIS, grants and non-financial support from ROCHE, grants, personal fees and non-financial support from Astrazeneca, personal fees and non-financial support from Mylan, grants, personal fees and non-financial support from Pierre Fabre, personal fees and non-financial support from AMGEN and personal fees and non-financial support from SEAGEN outside the submitted work. E.W.T. did this work while employed at Owkin, Inc. and is now employed by Amazon.Funding StatementFundings: As described in the acknowledgments this work received the financial support of the French BPI through the HealthChain project (https://www.substra.ai/en/healthchain-project). Some authors are employed by Owkin, Inc.. (J.O.d.T., C.B., M.A., Ch.M., B.S., E.B., M.Z., G.W., K.M., C.G., I.D., A.-L.M., Ca.M., M.G.) Personal fundings: G.W. is Co-founder of Owkin. P.-E.H. reports grants, personal fees and non-financial support from PFIZER, grants and non-financial support from NOVARTIS, grants and non-financial support from ROCHE, grants, personal fees and non-financial support from Astrazeneca, personal fees and non-financial support from Mylan, grants, personal fees and non-financial support from Pierre Fabre, personal fees and non-financial support from AMGEN and personal fees and non-financial support from SEAGEN outside the submitted work. E.W.T. did this work while employed at Owkin, Inc. and is now employed by Amazon.Author DeclarationsI confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.YesThe details of the IRB/oversight body that provided approval or exemption for the research described are given below:The IRB of Institut Curie (Comité de Revue Institutionnelle - CRI Data) and the IRB of Centre Léon Bérard (Artificial Intelligence Project Evaluation Board) both validated the ethics of the current study inside the HealthChain project. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.YesData is available upon reasonable and GDPR compliant request from Institut Curie and Centre Léon Bérard