Abstract
Background
Despite rapid advances in treatment, breast cancer remains the leading cause of cancer mortality in women, with triple negative breast cancers having a particularly poor prognosis. Some tumors have (epi)genetic alterations causing homologous recombination deficiency, providing opportunities for targeted therapeutics including poly (ADP-ribose) polymerase inhibitors. However, the effects of targeted treatments are variable; therefore, functional assays are needed to predict the best personalized treatment options.
Methods
We developed a high-throughput spheroid-based assay using patient-derived breast cancer xenograft models sensitive and resistant to cisplatin. Methods were developed for automatic spheroid segmentation using deep learning to measure response of spheroids to treatment with cisplatin, olaparib and radiotherapy. We developed a method to distinguish between sensitive and resistant tumors based on predicting the percentage of responding and non-responding spheroids.
Results
Here we show that differences in treatment response between cisplatin-sensitive and resistant tumors faithfully correspond with the expected in vivo responses. The assay is able to discriminate between olaparib-sensitive and resistant tumors based on predicting the percentage of responding and non-responding spheroids.
Conclusions
We demonstrate that this assay, guided by automatic spheroid segmentation using deep learning, may report on the tumor’s sensitivity to therapies with the potential to be applied to functional precision oncology for breast cancer.
Plain language summary
Many women die of breast cancer, especially if tumors are more difficult to treat or spread more easily to other organs. Some of these tumors can be successfully treated with drugs because of faults in their DNA repair. To choose the best treatment for the patient, it would be helpful to know which tumors are vulnerable to treatment. We therefore developed a test. We grew small spheres from tumor material (spheroids) and treated them with different drugs used in cancer therapies. To measure the response, we took pictures of the spheroids and we taught a computer program to recognize the spheroids and measure their size. We tested our assay with patient tumors grown on mice (sensitive and resistant to a drug). We demonstrate that this assay may report on the tumor’s sensitivity to therapies and may be used in the future to help choose the best treatment for breast cancer.
Similar content being viewed by others
Data availability
Training data and fine-tuned models are accessible through Zenodo (https://doi.org/10.5281/zenodo.14832406)44. The numerical data plotted (source data) in Figs. 1–4 are provided in Supplementary Data 1 and 2.
Code availability
All code, including the training notebook and analysis scripts, is available on GitHub (https://github.com/maartenpaul/spheroidAnalysis; https://doi.org/10.5281/zenodo.15239560)45. Code development was assisted by the Anthropic Claude 3.5 Sonnet large language model. All code was thoroughly validated and verified by the authors. We take full responsibility for its functionality and results.
References
Arnold, M. et al. Current and future burden of breast cancer: global statistics for 2020 and 2040. Breast 66, 15–23 (2022).
Newman, L. Oncologic anthropology: global variations in breast cancer risk, biology, and outcome. J. Surg. Oncol. 128, 959–966 (2023).
Lund, M. J. et al. Race and triple negative threats to breast cancer survival: a population-based study in Atlanta, GA. Breast Cancer Res. Treat. 113, 357–370 (2009).
Hartman, A.-R. et al. Prevalence of BRCA mutations in an unselected population of triple-negative breast cancer. Cancer 118, 2787–2795 (2012).
Heeke, A. L. et al. Prevalence of homologous recombination–related gene mutations across multiple cancer types. JCO Precis. Oncol. 1–13, https://doi.org/10.1200/po.17.00286 (2018).
van Wijk, L. M., Nilas, A. B., Vrieling, H. & Vreeswijk, M. P. G. RAD51 as a functional biomarker for homologous recombination deficiency in cancer: a promising addition to the HRD toolbox?. Expert Rev. Mol. Diagn. 22, 185–199 (2022).
Melinda, L. T. et al. Homologous recombination deficiency (hrd) score predicts response to platinum-containing neoadjuvant chemotherapy in patients with triple-negative breast cancer. Clin. Cancer Res. 22, 3764–3773 (2016).
Mangogna, A. et al. Homologous recombination deficiency in ovarian cancer: from the biological rationale to current diagnostic approaches. J. Pers. Med. 13, 284 (2023).
Joosse, S. A. et al. Prediction of BRCA2-association in hereditary breast carcinomas using array-CGH. Breast Cancer Res. Treat. 132, 379–389 (2012).
Davies, H. et al. HRDetect is a predictor of BRCA1 and BRCA2 deficiency based on mutational signatures. Nat. Med. 23, 517–525 (2017).
Nguyen, L. et al. Pan-cancer landscape of homologous recombination deficiency. Nat. Commun. 11, 1–12 (2020).
van Renterghem, A. W. J., van de Haar, J. & Voest, E. E. Functional precision oncology using patient-derived assays: bridging genotype and phenotype. Nat. Rev. Clin. Oncol. 20, 305–317 (2023).
Stover, E. H., Fuh, K., Konstantinopoulos, P. A., Matulonis, U. A. & Liu, J. F. Clinical assays for assessment of homologous recombination DNA repair deficiency. Gynecol. Oncol. 159, 887–898 (2020).
Meijer, T. G. et al. Functional RECAP (REpair CAPacity) assay identifies homologous recombination deficiency undetected by DNA-based BRCAness tests. Oncogene 41, 3498–3506 (2022).
Naipal, K. A. T. et al. Functional Ex vivo assay to select homologous recombination-deficient breast tumors for PARP inhibitor treatment. Clin. Cancer Res. 20, 4816–4826 (2014).
van Wijk, L. M. et al. The recap test rapidly and reliably identifies homologous recombination-deficient ovarian carcinomas. Cancers 12, 1–16 (2020).
Meijer, T. G. et al. Functional ex vivo assay reveals homologous recombination deficiency in breast cancer beyond BRCA gene defects. Clin. Cancer Res. 24, 6277–6287 (2018).
Kapałczyńska, M. et al. 2D and 3D cell cultures – a comparison of different types of cancer cell cultures. Arch. Med. Sci. 14, 910–919 (2018).
Costa, E. C., Gaspar, V. M., Coutinho, P. & Correia, I. J. Optimization of liquid overlay technique to formulate heterogenic 3D co-cultures models. Biotechnol. Bioeng. 111, 1672–1685 (2014).
Haspels, B. & Kuijten, M. M. P. Protocol for formation, staining, and imaging of 3D breast cancer models using MicroTissues mold systems. STAR Protoc 5, 103250 (2024).
Grexa, I. et al. SpheroidPicker for automated 3D cell culture manipulation using deep learning. Sci. Rep. 11, 1–11 (2021).
Trossbach, M., Åkerlund, E., Langer, K., Seashore-Ludlow, B. & Joensson, H. N. High-throughput cell spheroid production and assembly analysis by microfluidics and deep learning. SLAS Technol. 28, 423–432 (2023).
Mukundan, S. et al. Automated assessment of cancer drug efficacy on breast tumor spheroids in aggrewellTM400 plates using image cytometry. J. Fluoresc. 32, 521–531 (2022).
Linxweiler, J. et al. Patient-derived, three-dimensional spheroid cultures provide a versatile translational model for the study of organ-confined prostate cancer. J. Cancer Res. Clin. Oncol. 145, 551–559 (2019).
Della Corte, C. M. et al. Antitumor activity of dual blockade of PD-L1 and MEK in NSCLC patients derived three-dimensional spheroid cultures. J. Exp. Clin. Cancer Res. 38, 1–12 (2019).
Piccinini, F., Tesei, A., Arienti, C. & Bevilacqua, A. Cancer multicellular spheroids: Volume assessment from a single 2D projection. Comput. Methods Programs Biomed. 118, 95–106 (2015).
Celli, J. P. et al. An imaging-based platform for high-content, quantitative evaluation of therapeutic response in 3D tumour models. Sci. Rep. 4, 1–10 (2014).
Kirillov, A. et al. Segment anything. In Proceedings of the IEEE Internation Conference on Computer Vision 3992–4003, https://doi.org/10.1109/ICCV51070.2023.00371 (2023).
Benning, L., Peintner, A., Finkenzeller, G. & Peintner, L. Automated spheroid generation, drug application and efficacy screening using a deep learning classification: a feasibility study. Sci. Rep. 10, 1–11 (2020).
Ter Brugge, P. et al. Mechanisms of therapy resistance in patient-derived xenograft models of brca1-deficient breast cancer. J. Natl. Cancer Inst. 108, 1–12 (2016).
Koob, L. et al. MND1 enables homologous recombination in somatic cells primarily outside the context of replication. Mol. Oncol. 17, 1192–1211 (2023).
Archit, A. et al. Segment Anything for Microscopy. Nature Methods, vol. 22 (Springer US, 2025).
Wang, P. et al. Cisplatin induces hepG2 cell cycle arrest through targeting specific long noncoding RNAs and the p53 signaling pathway. Oncol. Lett. 12, 4605–4612 (2016).
Wagner, J. M. & Karnitz, L. M. Cisplatin-induced DNA damage activates replication checkpoint signaling components that differentially affect tumor cell survival. Mol. Pharmacol. 76, 208–214 (2009).
Mahaney, B. L., Meek, K. & Lees-Miller, S. P. Repair of ionizing radiation-induced DNA double strand breaks by non-homologous end-joining. Biochem. J. 417, 639–650 (2009).
Motulsky, H. J. & Brown, R. E. Detecting outliers when fitting data with nonlinear regression - a new method based on robust nonlinear regression and the false discovery rate. BMC Bioinformatics 7, 1–20 (2006).
Ayuso, J. M., Virumbrales-Muñoz, M., Lang, J. M. & Beebe, D. J. A role for microfluidic systems in precision medicine. Nat. Commun. 13, 1–12 (2022).
Yin, S. et al. Patient-derived tumor-like cell clusters for drug testing in cancer therapy. Sci. Transl. Med. 12, eaaz1723 (2020).
Aref, A. R. et al. 3D microfluidic ex vivo culture of organotypic tumor spheroids to model immune checkpoint blockade. Lab Chip 18, 3129–3143 (2019).
Sachs, N. et al. A living biobank of breast cancer organoids captures disease heterogeneity. Cell 172, 373–386.e10 (2018).
Kretzschmar, K. Cancer research using organoid technology. J. Mol. Med. 99, 501–515 (2021).
Dekkers, J. F. et al. Long-term culture, genetic manipulation and xenotransplantation of human normal and breast cancer organoids. Nat. Protoc. 16, 1936–1965 (2021).
Ireson, C. R., Alavijeh, M. S., Palmer, A. M., Fowler, E. R. & Jones, H. J. The role of mouse tumour models in the discovery and development of anticancer drugs. Br. J. Cancer 121, 101–108 (2019).
Paul, M. Training data and fine-tuned models. https://doi.org/10.5281/zenodo.14832406 (2025).
Paul, M. Code on GitHub. https://github.com/maartenpaul/spheroidAnalysis; https://doi.org/10.5281/zenodo.15239560 (2025).
Acknowledgements
The authors thank the people from the Preclinical Intervention Unit of the Mouse Clinic for Cancer and Ageing (MCCA) at the NKI for performing the PDX tumor outgrowth. This work was supported by the Oncode Institute, which is partly financed by the Dutch Cancer Society, by HollandPTC-Varian consortium-confined call 2019 (project number 2019011), the Dutch Cancer Society (KWF project 13651), and the NWO (Building Blocks of Life grant 737.016.011). Figures 1 and 3 were partially created with Biorender.com.
Author information
Authors and Affiliations
Contributions
Conceptualization: M.M.P.K. Funding acquisition: R.K and D.vG. Investigation: B.H, J.J.T., M.B. and M.M.P.K. Methodology: B.H., M.W.P., M.M.P.K. Visualization: M.W.P., B.H. and M.M.P.K. Supervision: R.K., J.J., D.vG. and M.M.P.K. Data curation: M.B. and Z.M.K. Software: M.W.P. Formal analysis: B.H., M.W.P. and M.M.P.K. Writing: M.W.P. and M.M.P.K. Review and editing: M.W.P., R.K., J.J, D.vG. and M.M.P.K.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Communications Medicine thanks Brian Karlberg, Evan F. Cromwel, Elad Katz, Zixuan Zhao and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Haspels, B., Paul, M.W., Jagessar Tewari, J. et al. High-throughput spheroid-based assay for functional breast cancer precision medicine facilitated by deep learning. Commun Med (2026). https://doi.org/10.1038/s43856-025-01359-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s43856-025-01359-8
