![]()
免疫细胞是人类体内天然的癌细胞杀手，但是癌细胞通过免疫细胞表面的程序性死亡受体PD-1及其配体PD-L1等免疫检查点，可以防止自己被人体免疫细胞杀死。免疫检查点抑制剂对部分晚期癌症患者可诱发持久的抗癌免疫反应，不过这些药物非常昂贵，而且大多数患者的临床获益并不持久。预测免疫检查点抑制剂疗效的方法可以帮助医生确定哪些患者较有可能或者较不可能对这些药物获益，将对精准医疗产生重大影响，可是这些预测方法（例如肿瘤突变负荷或者PD-L1免疫染色）通常需要借助高级、复杂而且同样昂贵的基因组学或免疫学检测。
2025年1月6日，英国《自然》旗下《自然医学》在线发表美国纽约西奈山医院伊坎医学院、纽约纪念医院斯隆凯特林癌症中心、罗斯威尔帕克综合癌症中心的研究报告，通过人工智能机器学习，根据常规血液检测数据以及临床数据，即可预测免疫检查点抑制剂治疗癌症的疗效。
该研究首先利用免疫检查点抑制剂治疗21种癌症（膀胱癌、乳腺癌、子宫颈癌、中枢神经系统肿瘤、结直肠癌、子宫内膜癌、食管癌、胃癌、头颈癌、肝胆癌、黑色素瘤、间皮瘤、非小细胞肺癌、卵巢癌、胰腺癌、前列腺癌、肾癌、骨肉瘤、小细胞肺癌、非黑色素瘤皮肤癌、原发灶不明癌）9745例患者的临床特征以及常规血液检测（全血细胞计数、各种代谢指标）开发出人工智能机器学习系统天蝎（SCORPIO）并进行验证。
![]()
SCORPIO: Standard Clinical and labOratory featuRes for Prognostication of Immunotherapy Outcomes
天蝎系统建模数据来自纽约纪念医院斯隆凯特林癌症中心17种癌症1628例患者。两个内部验证数据集包括19种癌症2511例患者，天蝎系统预测6、12、18、24和30个月总生存时，真假阳性率曲线下面积分别达0.763和0.759，显著优于肿瘤突变负荷的0.503和0.543；天蝎系统预测临床获益（肿瘤缩小或者长期稳定）时，真假阳性率曲线下面积分别达0.714和0.641，也显著优于肿瘤突变负荷的0.546和0.573。 利用全球10项3期随机对照试验（6种癌症4447例患者）以及西奈山医院真实世界队列（18种癌症1159例患者）进行外部验证，天蝎系统预测免疫检查点抑制剂治疗后患者结局仍然表现出色，显著优于PD-L1免疫染色。 因此，该研究结果表明了天蝎系统的可靠性和适应性，凸显其用于各种癌症类型以及医疗卫生环境预测免疫检查点抑制剂治疗后患者结局的潜力，故有必要进一步开展前瞻临床研究进行验证。运行天蝎系统所需程序代码下载网址：www.zenodo.org/records/13646737Nat Med. 2025 Jan 6. IF: 58.7Prediction of checkpoint inhibitor immunotherapy efficacy for cancer using routine blood tests and clinical data.Seong-Keun Yoo, Conall W. Fitzgerald, Byuri Angela Cho, Bailey G. Fitzgerald, Catherine Han, Elizabeth S. Koh, Abhinav Pandey, Hannah Sfreddo, Fionnuala Crowley, Michelle Rudshteyn Korostin, Neha Debnath, Yan Leyfman, Cristina Valero, Mark Lee, Joris L. Vos, Andrew Sangho Lee, Karena Zhao, Stanley Lam, Ezekiel Olumuyide, Fengshen Kuo, Eric A. Wilson, Pauline Hamon, Clotilde Hennequin, Miriam Saffern, Lynda Vuong, A. Ari Hakimi, Brian Brown, Miriam Merad, Sacha Gnjatic, Nina Bhardwaj, Matthew D. Galsky, Eric E. Schadt, Robert M. Samstein, Thomas U. Marron, Mithat Gonen, Luc G. T. Morris, Diego Chowell.Icahn School of Medicine at Mount Sinai, New York, NY, USA; Memorial Sloan Kettering Cancer Center, New York, NY, USA; Roswell Park Comprehensive Cancer Center, Buffalo, NY, USA; Pathos, New York, NY, USA.Predicting whether a patient with cancer will benefit from immune checkpoint inhibitors (ICIs) without resorting to advanced genomic or immunologic assays is an important clinical need. To address this, we developed and evaluated SCORPIO, a machine learning system that utilizes routine blood tests (complete blood count and comprehensive metabolic profile) alongside clinical characteristics from 9,745 ICI-treated patients across 21 cancer types. SCORPIO was trained on data from 1,628 patients across 17 cancer types from Memorial Sloan Kettering Cancer Center. In two internal test sets comprising 2,511 patients across 19 cancer types, SCORPIO achieved median time-dependent area under the receiver operating characteristic curve (AUC(t)) values of 0.763 and 0.759 for predicting overall survival at 6, 12, 18, 24 and 30 months, outperforming tumor mutational burden (TMB), which showed median AUC(t) values of 0.503 and 0.543. Additionally, SCORPIO demonstrated superior predictive performance for predicting clinical benefit (tumor response or prolonged stability), with AUC values of 0.714 and 0.641, compared to TMB (AUC = 0.546 and 0.573). External validation was performed using 10 global phase 3 trials (4,447 patients across 6 cancer types) and a real-world cohort from the Mount Sinai Health System (1,159 patients across 18 cancer types). In these external cohorts, SCORPIO maintained robust performance in predicting ICI outcomes, surpassing programmed death-ligand 1 immunostaining. These findings underscore SCORPIO's reliability and adaptability, highlighting its potential to predict patient outcomes with ICI therapy across diverse cancer types and healthcare settings.DOI: 10.1038/s41591-024-03398-5
