Trac Quang Thinh, Pawitan Yudi, Mou Tian, Erkers Tom, Östling Päivi, Bohlin Anna, Österroos Albin, Vesterlund Mattias, Jafari Rozbeh, Siavelis Ioannis, Bäckvall Helena, Kiviluoto Santeri, Orre Lukas M, Rantalainen Mattias, Lehtiö Janne, Lehmann Sören, Kallioniemi Olli, Vu Trung Nghia
Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden.
School of Biomedical Engineering, Shenzhen University, Shenzhen, China.
NPJ Precis Oncol. 2023 Mar 24;7(1):32. doi: 10.1038/s41698-023-00374-z.
Despite some encouraging successes, predicting the therapy response of acute myeloid leukemia (AML) patients remains highly challenging due to tumor heterogeneity. Here we aim to develop and validate MDREAM, a robust ensemble-based prediction model for drug response in AML based on an integration of omics data, including mutations and gene expression, and large-scale drug testing. Briefly, MDREAM is first trained in the BeatAML cohort (n = 278), and then validated in the BeatAML (n = 183) and two external cohorts, including a Swedish AML cohort (n = 45) and a relapsed/refractory acute leukemia cohort (n = 12). The final prediction is based on 122 ensemble models, each corresponding to a drug. A confidence score metric is used to convey the uncertainty of predictions; among predictions with a confidence score >0.75, the validated proportion of good responders is 77%. The Spearman correlations between the predicted and the observed drug response are 0.68 (95% CI: [0.64, 0.68]) in the BeatAML validation set, -0.49 (95% CI: [-0.53, -0.44]) in the Swedish cohort and 0.59 (95% CI: [0.51, 0.67]) in the relapsed/refractory cohort. A web-based implementation of MDREAM is publicly available at https://www.meb.ki.se/shiny/truvu/MDREAM/ .
尽管取得了一些令人鼓舞的成功，但由于肿瘤异质性，预测急性髓系白血病（AML）患者的治疗反应仍然极具挑战性。在此，我们旨在开发并验证MDREAM，这是一种基于组学数据（包括突变和基因表达）整合以及大规模药物测试的、用于AML药物反应的强大的基于集成的预测模型。简而言之，MDREAM首先在BeatAML队列（n = 278）中进行训练，然后在BeatAML队列（n = 183）以及两个外部队列中进行验证，这两个外部队列包括一个瑞典AML队列（n = 45）和一个复发/难治性急性白血病队列（n = 12）。最终的预测基于122个集成模型，每个模型对应一种药物。使用置信度评分指标来传达预测的不确定性；在置信度评分>0.75的预测中，良好反应者的验证比例为77%。在BeatAML验证集中，预测的和观察到的药物反应之间的斯皮尔曼相关性为0.68（95%CI：[0.64, 0.68]），在瑞典队列中为-0.49（95%CI：[-0.53, -0.44]），在复发/难治性队列中为0.59（95%CI：[0.51, 0.67]）。MDREAM的基于网络的实现可在https://www.meb.ki.se/shiny/truvu/MDREAM/ 上公开获取。
