A convolutional attention model for predicting response to chemo-immunotherapy from ultrasound elastography in mouse tumor models.
Journal
Communications medicine
ISSN: 2730-664X
Titre abrégé: Commun Med (Lond)
Pays: England
ID NLM: 9918250414506676
Informations de publication
Date de publication:
17 Oct 2024
17 Oct 2024
Historique:
received:
05
10
2023
accepted:
09
10
2024
medline:
18
10
2024
pubmed:
18
10
2024
entrez:
17
10
2024
Statut:
epublish
Résumé
In the era of personalized cancer treatment, understanding the intrinsic heterogeneity of tumors is crucial. Despite some patients responding favorably to a particular treatment, others may not benefit, leading to the varied efficacy observed in standard therapies. This study focuses on the prediction of tumor response to chemo-immunotherapy, exploring the potential of tumor mechanics and medical imaging as predictive biomarkers. We have extensively studied "desmoplastic" tumors, characterized by a dense and very stiff stroma, which presents a substantial challenge for treatment. The increased stiffness of such tumors can be restored through pharmacological intervention with mechanotherapeutics. We developed a deep learning methodology based on shear wave elastography (SWE) images, which involved a convolutional neural network (CNN) model enhanced with attention modules. The model was developed and evaluated as a predictive biomarker in the setting of detecting responsive, stable, and non-responsive tumors to chemotherapy, immunotherapy, or the combination, following mechanotherapeutics administration. A dataset of 1365 SWE images was obtained from 630 tumors from our previous experiments and used to train and successfully evaluate our methodology. SWE in combination with deep learning models, has demonstrated promising results in disease diagnosis and tumor classification but their potential for predicting tumor response prior to therapy is not yet fully realized. We present strong evidence that integrating SWE-derived biomarkers with automatic tumor segmentation algorithms enables accurate tumor detection and prediction of therapeutic outcomes. This approach can enhance personalized cancer treatment by providing non-invasive, reliable predictions of therapeutic outcomes. In personalized cancer treatment, it is important to understand that not all tumors respond the same way to therapy. While some patients may benefit from a particular treatment, others may not, leading to different outcomes. This study focuses on predicting how tumors will respond to a combination of chemotherapy and immunotherapy. Specifically, we looked at difficult-to-treat tumors with very stiff structures. These tumors can be softened with certain drugs making them more responsive to treatment. We developed a computer method to analyze medical images that measure the stiffness of tumors. Our method was trained on a large set of tumor images and was able to predict how well a tumor would respond to treatment. Overall, this approach could improve personalized cancer treatment using non-invasive medical imaging to predict which therapies will be most effective for each patient.
Sections du résumé
BACKGROUND
BACKGROUND
In the era of personalized cancer treatment, understanding the intrinsic heterogeneity of tumors is crucial. Despite some patients responding favorably to a particular treatment, others may not benefit, leading to the varied efficacy observed in standard therapies. This study focuses on the prediction of tumor response to chemo-immunotherapy, exploring the potential of tumor mechanics and medical imaging as predictive biomarkers. We have extensively studied "desmoplastic" tumors, characterized by a dense and very stiff stroma, which presents a substantial challenge for treatment. The increased stiffness of such tumors can be restored through pharmacological intervention with mechanotherapeutics.
METHODS
METHODS
We developed a deep learning methodology based on shear wave elastography (SWE) images, which involved a convolutional neural network (CNN) model enhanced with attention modules. The model was developed and evaluated as a predictive biomarker in the setting of detecting responsive, stable, and non-responsive tumors to chemotherapy, immunotherapy, or the combination, following mechanotherapeutics administration. A dataset of 1365 SWE images was obtained from 630 tumors from our previous experiments and used to train and successfully evaluate our methodology. SWE in combination with deep learning models, has demonstrated promising results in disease diagnosis and tumor classification but their potential for predicting tumor response prior to therapy is not yet fully realized.
RESULTS
RESULTS
We present strong evidence that integrating SWE-derived biomarkers with automatic tumor segmentation algorithms enables accurate tumor detection and prediction of therapeutic outcomes.
CONCLUSIONS
CONCLUSIONS
This approach can enhance personalized cancer treatment by providing non-invasive, reliable predictions of therapeutic outcomes.
In personalized cancer treatment, it is important to understand that not all tumors respond the same way to therapy. While some patients may benefit from a particular treatment, others may not, leading to different outcomes. This study focuses on predicting how tumors will respond to a combination of chemotherapy and immunotherapy. Specifically, we looked at difficult-to-treat tumors with very stiff structures. These tumors can be softened with certain drugs making them more responsive to treatment. We developed a computer method to analyze medical images that measure the stiffness of tumors. Our method was trained on a large set of tumor images and was able to predict how well a tumor would respond to treatment. Overall, this approach could improve personalized cancer treatment using non-invasive medical imaging to predict which therapies will be most effective for each patient.
Autres résumés
Type: plain-language-summary
(eng)
In personalized cancer treatment, it is important to understand that not all tumors respond the same way to therapy. While some patients may benefit from a particular treatment, others may not, leading to different outcomes. This study focuses on predicting how tumors will respond to a combination of chemotherapy and immunotherapy. Specifically, we looked at difficult-to-treat tumors with very stiff structures. These tumors can be softened with certain drugs making them more responsive to treatment. We developed a computer method to analyze medical images that measure the stiffness of tumors. Our method was trained on a large set of tumor images and was able to predict how well a tumor would respond to treatment. Overall, this approach could improve personalized cancer treatment using non-invasive medical imaging to predict which therapies will be most effective for each patient.
Identifiants
pubmed: 39420199
doi: 10.1038/s43856-024-00634-4
pii: 10.1038/s43856-024-00634-4
doi:
Types de publication
Journal Article
Langues
eng
Pagination
203Subventions
Organisme : EC | EU Framework Programme for Research and Innovation H2020 | H2020 Priority Excellent Science | H2020 European Research Council (H2020 Excellent Science - European Research Council)
ID : 101069207
Organisme : EC | Horizon 2020 Framework Programme (EU Framework Programme for Research and Innovation H2020)
ID : 956201
Informations de copyright
© 2024. The Author(s).
Références
Borrebaeck, C. A. Precision diagnostics: moving towards protein biomarker signatures of clinical utility in cancer. Nat. Rev. Cancer 17, 199–204 (2017).
pubmed: 28154374
doi: 10.1038/nrc.2016.153
Jain, R. K., Martin, J. D. & Stylianopoulos, T. The role of mechanical forces in tumor growth and therapy. Annu Rev. Biomed. Eng. 16, 321–346 (2014).
pubmed: 25014786
pmcid: 4109025
doi: 10.1146/annurev-bioeng-071813-105259
Stylianopoulos, T., Munn, L. L. & Jain, R. K. Reengineering the physical microenvironment of tumors to improve drug delivery and efficacy: from mathematical modeling to bench to bedside. Trends cancer 4, 292–319 (2018).
pubmed: 29606314
pmcid: 5930008
doi: 10.1016/j.trecan.2018.02.005
Martin, J. D., Cabral, H., Stylianopoulos, T. & Jain, R. K. Improving cancer immunotherapy using nanomedicines: progress, opportunities and challenges. Nat. Rev. Clin. Oncol. 17, 251–266 (2020).
pubmed: 32034288
pmcid: 8272676
doi: 10.1038/s41571-019-0308-z
Voutouri, C. et al. Ultrasound stiffness and perfusion markers correlate with tumor volume responses to immunotherapy. Acta Biomater. 167, 121–134 (2023).
Stylianopoulos, T. et al. Causes, consequences, and remedies for growth-induced solid stress in murine and human tumors. Proc. Natl Acad. Sci. USA 109, 15101–15108 (2012).
pubmed: 22932871
pmcid: 3458380
doi: 10.1073/pnas.1213353109
Voutouri, C. & Stylianopoulos, T. Accumulation of mechanical forces in tumors is related to hyaluronan content and tissue stiffness. PloS One 13, e0193801 (2018).
pubmed: 29561855
pmcid: 5862434
doi: 10.1371/journal.pone.0193801
Angeli, S. & Stylianopoulos, T. Biphasic modeling of brain tumor biomechanics and response to radiation treatment. J. Biomech. 49, 1524–1531 (2016).
pubmed: 27086116
pmcid: 4921059
doi: 10.1016/j.jbiomech.2016.03.029
Vavourakis, V. et al. A validated multiscale in-silico model for mechano-sensitive tumour angiogenesis and growth. PLoS Comput. Biol. 13, e1005259 (2017).
pubmed: 28125582
pmcid: 5268362
doi: 10.1371/journal.pcbi.1005259
Jain, R. K. Antiangiogenesis strategies revisited: from starving tumors to alleviating hypoxia. Cancer Cell 26, 605–622 (2014).
pubmed: 25517747
pmcid: 4269830
doi: 10.1016/j.ccell.2014.10.006
Mpekris, F. et al. Combining microenvironment normalization strategies to improve cancer immunotherapy. Proc. Natl Acad. Sci. USA 117, 3728–3737 (2020).
pubmed: 32015113
pmcid: 7035612
doi: 10.1073/pnas.1919764117
Chauhan, V. P. et al. Angiotensin inhibition enhances drug delivery and potentiates chemotherapy by decompressing tumor blood vessels. Nat. Commun. 4, https://doi.org/10.1038/ncomms.3516 (2013).
Papageorgis, P. et al. Tranilast-induced stress alleviation in solid tumors improves the efficacy of chemo- and nanotherapeutics in a size-independent manner. Sci. Rep. 7, 46140 (2017).
pubmed: 28393881
pmcid: 5385877
doi: 10.1038/srep46140
Polydorou, C., Mpekris, F., Papageorgis, P., Voutouri, C. & Stylianopoulos, T. Pirfenidone normalizes the tumor microenvironment to improve chemotherapy. Oncotarget 8, 24506–24517 (2017).
pubmed: 28445938
pmcid: 5421866
doi: 10.18632/oncotarget.15534
Panagi, M. et al. TGF-β inhibition combined with cytotoxic nanomedicine normalizes triple negative breast cancer microenvironment towards anti-tumor immunity. Theranostics 10, 1910–1922 (2020).
pubmed: 32042344
pmcid: 6993226
doi: 10.7150/thno.36936
Mpekris, F. et al. Normalizing the microenvironment overcomes vessel compression and resistance to nano-immunotherapy in breast cancer lung metastasis. Adv. Sci. 8, 2001917 (2021).
doi: 10.1002/advs.202001917
Voutouri, C. et al. Endothelin inhibition potentiates cancer immunotherapy revealing mechanical biomarkers predictive of response. Adv. Ther. 4, 2000289 (2021).
Panagi, M. et al. Polymeric micelles effectively reprogram the tumor microenvironment to potentiate nano-immunotherapy in mouse breast cancer models. Nat. Commun. 13, 7165 (2022).
pubmed: 36418896
pmcid: 9684407
doi: 10.1038/s41467-022-34744-1
Murphy, J. E. et al. Total neoadjuvant therapy with FOLFIRINOX in combination with Losartan followed by chemoradiotherapy for locally advanced pancreatic cancer: A Phase 2 clinical trial. JAMA Oncol. 5, 1020–1027 (2019).
pubmed: 31145418
pmcid: 6547247
doi: 10.1001/jamaoncol.2019.0892
Sheridan, C. Pancreatic cancer provides testbed for first mechanotherapeutics. Nat. Biotechnol. 37, 829–831 (2019).
pubmed: 31375797
doi: 10.1038/d41587-019-00019-2
Cui, X. W. et al. Ultrasound elastography. Endosc. Ultrasound 11, 252–274 (2022).
pubmed: 35532576
pmcid: 9526103
doi: 10.4103/EUS-D-21-00151
Mislati, R. et al. Shear wave elastography can stratify rectal cancer response to short-course radiation therapy. Sci. Rep. 13, 16149 (2023).
pubmed: 37752156
pmcid: 10522682
doi: 10.1038/s41598-023-43383-5
Wang, H. et al. Shear wave elastography can differentiate between radiation-responsive and non-responsive pancreatic tumors: an ex vivo study with murine models. Ultrasound Med. Biol. 46, 393–404 (2020).
pubmed: 31727378
doi: 10.1016/j.ultrasmedbio.2019.10.005
Wang, H. et al. Elastography can map the local inverse relationship between shear modulus and drug delivery within the pancreatic ductal adenocarcinoma microenvironment. Clin. Cancer Res. 25, 2136–2143 (2019).
pubmed: 30352906
doi: 10.1158/1078-0432.CCR-18-2684
Chen, L. D. et al. Assessment of rectal tumors with shear-wave elastography before surgery: comparison with endorectal US. Radiology 285, 279–292 (2017).
pubmed: 28640694
doi: 10.1148/radiol.2017162128
Berg, W. A. et al. Shear-wave elastography improves the specificity of breast US: the BE1 multinational study of 939 masses. Radiology 262, 435–449 (2012).
pubmed: 22282182
doi: 10.1148/radiol.11110640
Liu, B. J. et al. Quantitative shear wave velocity measurement on acoustic radiation force impulse elastography for differential diagnosis between benign and malignant thyroid nodules: a meta-analysis. Ultrasound Med Biol. 41, 3035–3043 (2015).
pubmed: 26371402
doi: 10.1016/j.ultrasmedbio.2015.08.003
Evans, A. et al. Prediction of pathological complete response to neoadjuvant chemotherapy for primary breast cancer comparing interim ultrasound, shear wave Elastography and MRI. Ultraschall Med. 39, 422–431 (2018).
pubmed: 28934812
doi: 10.1055/s-0043-111589
Gu, J. et al. Early assessment of shear wave elastography parameters foresees the response to neoadjuvant chemotherapy in patients with invasive breast cancer. Breast Cancer Res. 23, 52 (2021).
pubmed: 33926522
pmcid: 8082810
doi: 10.1186/s13058-021-01429-4
Hayashi, M., Yamamoto, Y. & Iwase, H. Clinical imaging for the prediction of neoadjuvant chemotherapy response in breast cancer. Chin. Clin. Oncol. 9, 31 (2020).
pubmed: 32594748
doi: 10.21037/cco-20-15
Fujioka, T. et al. Classification of breast masses on ultrasound shear wave elastography using convolutional neural networks. Ultrason Imaging 42, 213–220 (2020).
pubmed: 32501152
doi: 10.1177/0161734620932609
Liao, W.-X. et al. Automatic identification of breast ultrasound image based on supervised block-based region segmentation algorithm and features combination migration deep learning model. IEEE J. Biomed. Health Inform. 24, 984–993 (2019).
pubmed: 31869809
doi: 10.1109/JBHI.2019.2960821
Zhang, X. et al. Deep learning-based radiomics of b-mode ultrasonography and shear-wave elastography: Improved performance in breast mass classification. Front. Oncol. 10, 1621 (2020).
pubmed: 32984032
pmcid: 7485397
doi: 10.3389/fonc.2020.01621
Zhou, Y. et al. A radiomics approach with CNN for shear-wave elastography breast tumor classification. IEEE Trans. Biomed. Eng. 65, 1935–1942 (2018).
pubmed: 29993469
doi: 10.1109/TBME.2018.2844188
Li, H. et al. Deep learning in ultrasound elastography imaging: A review. Med. Phys. 49, 5993–6018 (2022).
pubmed: 35842833
doi: 10.1002/mp.15856
Misra, S. et al. Bi-modal transfer learning for classifying breast cancers via combined B-mode and ultrasound strain imaging. IEEE Trans. Ultrason Ferroelectr. Freq. Control 69, 222–232 (2022).
pubmed: 34633928
doi: 10.1109/TUFFC.2021.3119251
Papanastasiou, G., Dikaios, N., Huang, J., Wang, C. & Yang, G. Is attention all you need in medical image analysis? A review. IEEE J. Biomed. Health Inform 28, 1398–1411 (2023).
Morris, D. M. et al. A novel deep learning method for large-scale analysis of bone marrow adiposity using UK Biobank Dixon MRI data. Comput. Struct. Biotechnol. J. 24, 89–104 (2024).
pubmed: 38268780
doi: 10.1016/j.csbj.2023.12.029
Mpekris, F. et al. Translational nanomedicine potentiates immunotherapy in sarcoma by normalizing the microenvironment. J. Control Rel. 353, 956–964 (2022).
doi: 10.1016/j.jconrel.2022.12.016
Mpekris, F. et al. Normalizing tumor microenvironment with nanomedicine and metronomic therapy to improve immunotherapy. J. Control Rel. 345, 190–199 (2022).
doi: 10.1016/j.jconrel.2022.03.008
Brigato, L. & Iocchi, L. In 2020 25th International Conference on Pattern Recognition (ICPR). 2490–2497 (IEEE).
Eisenhauer, E. A. et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur. J. Cancer 45, 228–247 (2009).
pubmed: 19097774
doi: 10.1016/j.ejca.2008.10.026
Ronneberger, O., Fischer, P. & Brox, T. in Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18. 234-241 (Springer).
Chollet, F. in Proceedings of the IEEE conference on computer vision and pattern recognition. 1251–1258.
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. & Wojna, Z. in Proceedings of the IEEE conference on computer vision and pattern recognition. 2818–2826.
He, K., Zhang, X., Ren, S. & Sun, J. in Proceedings of the IEEE conference on computer vision and pattern recognition. 770–778.
Selvaraju, R. R. et al. in Proceedings of the IEEE international conference on computer vision. 618–626.
Riegler, J. et al. Tumor elastography and its association with collagen and the tumor microenvironment. Clin. Cancer Res. 24, 4455–4467 (2018).
pubmed: 29798909
doi: 10.1158/1078-0432.CCR-17-3262
Zheng, D. et al. Biomimetic nanoparticles drive the mechanism understanding of shear-wave elasticity stiffness in triple negative breast cancers to predict clinical treatment. Bioact. Mater. 22, 567–587 (2023).
pubmed: 36382024
Chang, J. M. et al. Clinical application of shear wave elastography (SWE) in the diagnosis of benign and malignant breast diseases. Breast Cancer Res. Treat. 129, 89–97 (2011).
pubmed: 21681447
doi: 10.1007/s10549-011-1627-7
Chang, J. M. et al. Comparison of shear-wave and strain ultrasound elastography in the differentiation of benign and malignant breast lesions. Am. J. Roentgenol. 201, W347–W356 (2013).
doi: 10.2214/AJR.12.10416
Olgun, D. Ç. et al. Use of shear wave elastography to differentiate benign and malignant breast lesions. Diagn. Interven. Radiol. 20, 239 (2014).
doi: 10.5152/dir.2014.13306
Brassart-Pasco, S. et al. Tumor microenvironment: extracellular matrix alterations influence tumor progression. Front. Oncol. 10, 397 (2020).
pubmed: 32351878
pmcid: 7174611
doi: 10.3389/fonc.2020.00397
Eble, J. A. & Niland, S. The extracellular matrix in tumor progression and metastasis. Clin. Exp. Metastasis 36, 171–198 (2019).
pubmed: 30972526
doi: 10.1007/s10585-019-09966-1
Bi, W. L. et al. Artificial intelligence in cancer imaging: clinical challenges and applications. CA: Cancer J. Clin. 69, 127–157 (2019).
pubmed: 30720861
Jiang, X., Hu, Z., Wang, S. & Zhang, Y. Deep learning for medical image-based cancer diagnosis. Cancers 15, 3608 (2023).
pubmed: 37509272
pmcid: 10377683
doi: 10.3390/cancers15143608
Kumar, Y., Gupta, S., Singla, R. & Hu, Y.-C. A systematic review of artificial intelligence techniques in cancer prediction and diagnosis. Arch. Comput. Methods Eng. 29, 2043–2070 (2022).
pubmed: 34602811
doi: 10.1007/s11831-021-09648-w
Voutouri, C. et al. A convolutional attention model for predicting response to chemo-immunotherapy from ultrasound elastography in mouse tumor models. Zenodo, https://doi.org/10.5281/zenodo.13771359 (2024).