Joint AI-driven event prediction and longitudinal modeling in newly diagnosed and relapsed multiple myeloma.


Journal

NPJ digital medicine
ISSN: 2398-6352
Titre abrégé: NPJ Digit Med
Pays: England
ID NLM: 101731738

Informations de publication

Date de publication:
29 Jul 2024
Historique:
received: 23 08 2023
accepted: 15 07 2024
medline: 30 7 2024
pubmed: 30 7 2024
entrez: 29 7 2024
Statut: epublish

Résumé

Multiple myeloma management requires a balance between maximizing survival, minimizing adverse events to therapy, and monitoring disease progression. While previous work has proposed data-driven models for individual tasks, these approaches fail to provide a holistic view of a patient's disease state, limiting their utility to assist physician decision-making. To address this limitation, we developed a transformer-based machine learning model that jointly (1) predicts progression-free survival (PFS), overall survival (OS), and adverse events (AE), (2) forecasts key disease biomarkers, and (3) assesses the effect of different treatment strategies, e.g., ixazomib, lenalidomide, dexamethasone (IRd) vs lenalidomide, dexamethasone (Rd). Using TOURMALINE trial data, we trained and internally validated our model on newly diagnosed myeloma patients (N = 703) and externally validated it on relapsed and refractory myeloma patients (N = 720). Our model achieved superior performance to a risk model based on the multiple myeloma international staging system (ISS) (p < 0.001, Bonferroni corrected) and comparable performance to survival models trained separately on each task, but unable to forecast biomarkers. Our approach outperformed state-of-the-art deep learning models, tailored towards forecasting, on predicting key disease biomarkers (p < 0.001, Bonferroni corrected). Finally, leveraging our model's capacity to estimate individual-level treatment effects, we found that patients with IgA kappa myeloma appear to benefit the most from IRd. Our study suggests that a holistic assessment of a patient's myeloma course is possible, potentially serving as the foundation for a personalized decision support system.

Identifiants

pubmed: 39075240
doi: 10.1038/s41746-024-01189-3
pii: 10.1038/s41746-024-01189-3
doi:

Types de publication

Journal Article

Langues

eng

Pagination

200

Informations de copyright

© 2024. The Author(s).

Références

Ludwig, H. et al. International myeloma working group recommendations for global myeloma care. Leukemia 28, 981–992 (2014).
doi: 10.1038/leu.2013.293 pubmed: 24177258
Sonneveld, P. et al. Treatment of multiple myeloma with high-risk cytogenetics: a consensus of the international myeloma working group. Blood J. Am. Soc. Hematol. 127, 2955–2962 (2016).
Dimopoulos, M. A. et al. International myeloma working group recommendations for the diagnosis and management of myeloma-related renal impairment. J. Clin. Oncol. 34, 1544–1557 (2016).
doi: 10.1200/JCO.2015.65.0044 pubmed: 26976420
Guerrero, C. et al. A machine learning model based on tumor and immune biomarkers to predict undetectable MRD and survival outcomes in multiple myeloma. Clin. Cancer Res. 28, 2598–2609 (2022).
doi: 10.1158/1078-0432.CCR-21-3430 pubmed: 35063966
Kuiper, R. et al. Prediction of high-and low-risk multiple myeloma based on gene expression and the international staging system. Blood J. Am. Soc. Hematol. 126, 1996–2004 (2015).
Park, S.-S. et al. Ml-based sequential analysis to assist selection between vmp and rd for newly diagnosed multiple myeloma. NPJ Precis. Oncol. 7, 46 (2023).
doi: 10.1038/s41698-023-00385-w pubmed: 37210456 pmcid: 10199943
Mosquera Orgueira, A. et al. Unsupervised machine learning improves risk stratification in newly diagnosed multiple myeloma: an analysis of the spanish myeloma group. Blood Cancer J. 12, 76 (2022).
doi: 10.1038/s41408-022-00647-z pubmed: 35468898 pmcid: 9038663
Maura, F. et al. Genomic classification and individualized prognosis in multiple myeloma. J. Clin. Oncol. 42, 1229–1240 (2024).
doi: 10.1200/JCO.23.01277 pubmed: 38194610
Avet-Loiseau, H. et al. Carfilzomib significantly improves the progression-free survival of high-risk patients in multiple myeloma. Blood J. Am. Soc. Hematol. 128, 1174–1180 (2016).
Fotiou, D., Gavriatopoulou, M. & Terpos, E. Multiple myeloma and thrombosis: prophylaxis and risk prediction tools. Cancers 12, 191 (2020).
doi: 10.3390/cancers12010191 pubmed: 31940972 pmcid: 7016665
Alaa, A. M. & van der Schaar, M. Attentive state-space modeling of disease progression. Advances in Neural Information Processing Systems 32 (2019).
Vaswani, A. et al. Attention is all you need. Advances in Neural Information Processing Systems 30 (2017).
Li, Y. & Cai, Z. Clinical characteristics and survival of patients with igd multiple myeloma. Blood Sci. 3, 57–58 (2021).
doi: 10.1097/BS9.0000000000000066 pubmed: 35402831 pmcid: 8975086
Habermehl, G. K., Nakashima, M. O. & Cotta, C. V. Iga plasma cell neoplasms are characterized by poorer long-term survival and increased genomic complexity compared to igg neoplasms. Ann. Diagn. Pathol. 44, 151449 (2020).
doi: 10.1016/j.anndiagpath.2019.151449 pubmed: 31865247
Greipp, P. R. et al. International staging system for multiple myeloma. J. Clin. Oncol. 23, 3412–3420 (2005).
doi: 10.1200/JCO.2005.04.242 pubmed: 15809451
Lee, C., Yoon, J. & Van Der Schaar, M. Dynamic-deephit: a deep learning approach for dynamic survival analysis with competing risks based on longitudinal data. IEEE Trans. Biomed. Eng. 67, 122–133 (2019).
doi: 10.1109/TBME.2019.2909027 pubmed: 30951460
Jamet, B. et al. Random survival forest to predict transplant-eligible newly diagnosed multiple myeloma outcome including fdg-pet radiomics: a combined analysis of two independent prospective European trials. Eur. J. Nucl. Med. Mol. Imaging 48, 1005–1015 (2021).
doi: 10.1007/s00259-020-05049-6 pubmed: 33006656
Mosquera Orgueira, A. et al. Survival prediction and treatment optimization of multiple myeloma patients using machine-learning models based on clinical and gene expression data. Leukemia 35, 2924–2935 (2021).
doi: 10.1038/s41375-021-01286-2 pubmed: 34007046
Barlow, H., Mao, S. & Khushi, M. Predicting high-risk prostate cancer using machine learning methods. Data 4, 129 (2019).
doi: 10.3390/data4030129
Uno, H., Cai, T., Pencina, M. J., D’Agostino, R. B. & Wei, L.-J. On the c-statistics for evaluating overall adequacy of risk prediction procedures with censored survival data. Stat. Med. 30, 1105–1117 (2011).
doi: 10.1002/sim.4154 pubmed: 21484848 pmcid: 3079915
Facon, T. et al. Oral ixazomib, lenalidomide, and dexamethasone for transplant-ineligible patients with newly diagnosed multiple myeloma. Blood 137, 3616–3628 (2021).
doi: 10.1182/blood.2020008787 pubmed: 33763699 pmcid: 8462404
Jones, G. S. & Baldwin, D. R. Recent advances in the management of lung cancer. Clin. Med. 18, s41 (2018).
doi: 10.7861/clinmedicine.18-2-s41
Scherr, D., Swindle, P. W. & Scardino, P. T. National comprehensive cancer network guidelines for the management of prostate cancer. Urology 61, 14–24 (2003).
doi: 10.1016/S0090-4295(02)02395-6 pubmed: 12667883
McDonald, E. S., Clark, A. S., Tchou, J., Zhang, P. & Freedman, G. M. Clinical diagnosis and management of breast cancer. J. Nucl. Med. 57, 9S–16S (2016).
doi: 10.2967/jnumed.115.157834 pubmed: 26834110
Ibrahim, J. G., Chu, H. & Chen, L. M. Basic concepts and methods for joint models of longitudinal and survival data. J. Clin. Oncol. 28, 2796 (2010).
doi: 10.1200/JCO.2009.25.0654 pubmed: 20439643 pmcid: 4503792
Rizopoulos, D. Joint Models for Longitudinal and Time-to-Event Data: With Applications in R (CRC press, 2012).
Tsiatis, A. A. & Davidian, M. Joint modeling of longitudinal and time-to-event data: an overview. Stat. Sin. 14, 809–834 (2004).
Proust-Lima, C. & Taylor, J. M. Development and validation of a dynamic prognostic tool for prostate cancer recurrence using repeated measures of posttreatment psa: a joint modeling approach. Biostatistics 10, 535–549 (2009).
doi: 10.1093/biostatistics/kxp009 pubmed: 19369642 pmcid: 2697347
Liu, S. et al. Pyraformer: low-complexity pyramidal attention for long-range time series modeling and forecasting. In Proc. International Conference on Learning Representations (2021).
Zhou, T. et al. Fedformer: frequency-enhanced decomposed transformer for long-term series forecasting. In Proc. International Conference on Machine Learning, 27268–27286 (PMLR, 2022).
Shalit, U., Johansson, F. D. & Sontag, D. Estimating individual treatment effect: generalization bounds and algorithms. In Proc. International Conference on Machine Learning, 3076–3085 (PMLR, 2017).
Künzel, S. R., Sekhon, J. S., Bickel, P. J. & Yu, B. Metalearners for estimating heterogeneous treatment effects using machine learning. Proc. Natl Acad. Sci. 116, 4156–4165 (2019).
doi: 10.1073/pnas.1804597116 pubmed: 30770453 pmcid: 6410831
Hahn, P. R., Murray, J. S. & Carvalho, C. M. Bayesian regression tree models for causal inference: regularization, confounding, and heterogeneous effects (with discussion). Bayesian Anal. 15, 965–1056 (2020).
doi: 10.1214/19-BA1195
Kingma, D. P. & Ba, J. Adam: a method for stochastic optimization. In Proc. International Conference on Learning Representations (2015).
Breiman, L. Bagging predictors. Mach. Learn. 24, 123–140 (1996).
doi: 10.1007/BF00058655
Palumbo, A. et al. Revised international staging system for multiple myeloma: a report from international myeloma working group. J. Clin. Oncol. 33, 2863 (2015).
doi: 10.1200/JCO.2015.61.2267 pubmed: 26240224 pmcid: 4846284

Auteurs

Zeshan Hussain (Z)

CSAIL, MIT, Cambridge, MA, USA.
Harvard Medical School, Boston, MA, USA.

Edward De Brouwer (E)

CSAIL, MIT, Cambridge, MA, USA.

Rebecca Boiarsky (R)

CSAIL, MIT, Cambridge, MA, USA.

Sama Setty (S)

CSAIL, MIT, Cambridge, MA, USA.

Neeraj Gupta (N)

Takeda LLC, Cambridge, MA, USA.

Guohui Liu (G)

Takeda LLC, Cambridge, MA, USA.

Cong Li (C)

Takeda LLC, Cambridge, MA, USA.

Jaydeep Srimani (J)

Takeda LLC, Cambridge, MA, USA.

Jacob Zhang (J)

Takeda LLC, Cambridge, MA, USA.

Rich Labotka (R)

Takeda LLC, Cambridge, MA, USA.

David Sontag (D)

CSAIL, MIT, Cambridge, MA, USA. dsontag@csail.mit.edu.

Classifications MeSH