基于MFPC-Cox的结直肠癌患者预后动态预测模型

doi:10.6040/j.issn.1671-7554.0.2024.0531

摘要/Abstract

摘要： 目的评估重复测量癌胚抗原(carcinoembryonic antigen, CEA)和糖类抗原19-9(carbohydrate antigen 19-9, CA19-9)对改善结直肠癌(colorectal cancer, CRC)患者预后的应用价值,动态预测患者未来CEA和CA19-9变化趋势及生存概率。方法选取2011年1月至2018年12月在云南省肿瘤医院接受根治性切除术治疗的CRC患者为研究对象,基于患者的临床资料及围手术期CEA和CA19-9纵向测量信息,使用多元函数型主成分分析(multivariate functional principal components analysis, MFPCA)提取患者术后12个月内纵向CEA和CA19-9的轨迹特征,将相应的多元函数型主成分得分作为协变量,纳入Cox比例风险模型,构建MFPC-Cox预后动态预测模型。通过随时间变化的曲线下面积(area under curve, AUC)和Brier评分(Brier score, BS)定量评价模型的预测性能,并与仅考虑基线信息的静态预测模型进行比较。结果对于CEA和CA19-9的MFPCA,选择前7个主成分描述其纵向特征。与静态模型相比,动态预测模型术后60个月生存率的AUC由0.727增加到0.787,BS由0.077下降至0.072。考虑上述标志物的纵向测量信息后,动态模型预测的准确性明显上升。结论考虑CEA和CA19-9围手术期纵向测量信息后,基于MFPC-Cox的CRC预后模型具有较高的准确性,并能够在每一次随访时更新风险,实现动态预测。推荐在CRC患者术后随访过程中重复测量CEA和CA19-9。

关键词: 结直肠癌, 血清肿瘤标志物, MFPC-Cox模型, 动态预测, 预后

Abstract: Objective To evaluated the utility of repeated measurements of carcinoembryonic antigen(CEA)and carbohydrate antigen 19-9(CA19-9)in improving the prognosis of colorectal cancer(CRC)patients, and to predict dynamically the future longitudinal trajectories of CEA and CA19-9, as well as the survival probability. Methods CRC patients who underwent radical resection at Yunnan Cancer Hospital between January 2011 and December 2018 were selected as the study subjects. Based on the clinical data and perioperative longitudinal measurements of CEA and CA19-9 in patients, multivariate functional principal component analysis(MFPCA)was used to extract the trajectory features of longitudinal CEA and CA19-9 measurements within 12 months postoperatively. The corresponding multivariate functional principal component scores were incorporated as covariates into the Cox proportional hazards model to construct a MFPC-Cox dynamic prediction model for the colorectal cancer patients prognosis. The predictive performance of models was quantitatively assessed by the time-varying area under the curve(AUC)and Brier score(BS), and compared with a static prediction model that only considered baseline information. Results The first seven principal components were selected to describe their longitudinal characteristics in the MFPCA of CEA and CA19-9. Compared to the static model, the dynamic prediction model increased the AUC of the 60-month postoperative survival rate from 0.727 to 0.787 and reduced the BS from 0.077 to 0.072. The accuracy of the model prediction improved significantly with the inclusion of aforementioned longitudinal biomarker measurements. Conclusion After considering the perioperative longitudinal measurements of CEA and CA19-9, the prognostic model for CRC based on the MFPC-Cox has high accuracy and can update the risk at each follow-up visit, enabling dynamic prediction. It is recommended to repeatedly measure CEA and CA19-9 during the postoperative follow-up of CRC patients.

Key words: Colorectal cancer, Serum tumor markers, MFPC-Cox model, Dynamic prediction, Prognosis

中图分类号:

R735.3

杜雪,李春霞,刘云霞,张涛. 基于MFPC-Cox的结直肠癌患者预后动态预测模型[J]. 山东大学学报 (医学版), 2025, 63(5): 101-110.

DU Xue, LI Chunxia, LIU Yunxia, ZHANG Tao. Dynamic prediction model for the colorectal cancer patients prognosis based on MFPC-Cox[J]. Journal of Shandong University (Health Sciences), 2025, 63(5): 101-110.

参考文献

[1] Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2024, 74(3): 229-263.
[2] Benson AB, Venook AP, Adam M, et al. Colon cancer, Version 3.2024, NCCN clinical practice guidelines in oncology[J]. J Natl Compr Canc Netw, 2024, 22(2D): e240029. doi:10.6004/jnccn.2024.0029
[3] Benson AB, Venook AP, Al-Hawary MM, et al. Rectal cancer, version 2.2022, NCCN clinical practice guidelines in oncology[J]. J Natl Compr Canc Netw, 2022, 20(10): 1139-1167.
[4] 谢雅, 闫文锋, 夏晓博. 基于列线图模型预测结直肠癌淋巴结转移率与临床特征及预后的关系[J]. 癌症进展, 2023, 21(17): 1877-1880. XIE Ya, YAN Wenfeng, XIA Xiaobo. Correlation of metastatic lymph node ratio with clinical characteristics and prognosis in colorectal cancer predicted by nomogram model[J]. Oncology Progress, 2023, 21(17): 1877-1880.
[5] 王廉源, 杨毅, 丛慧文, 等. 基于竞争风险模型的早发性结直肠癌患者预后影响因素分析[J]. 中国医科大学学报, 2023, 52(3): 199-205. WANG Lianyuan, YANG Yi, CONG Huiwen, et al. Ana-lysis of prognostic factors in patients with early-onset colo-rectal cancer based on a competitive risk model[J]. Journal of China Medical University, 2023, 52(3): 199-205.
[6] 翟雅娜, 张敬东. 结直肠癌伴同时肝转移患者同期切除术后预后风险模型的创建及应用[J]. 现代肿瘤医学, 2020, 28(16): 2826-2832. ZHAI Yana, ZHANG Jingdong. Creation and application of prognostic risk model after simultaneous liver and colorectal resection for patients with synchronous colorectal liver metastasis[J]. Journal of Modern Oncology, 2020, 28(16): 2826-2832.
[7] Zhang Q, Wang L, Sun RR, et al. Implications of pretreatment serum carcinoembryonic antigen levels and perineural invasion with staging, prognosis, and management in stage I-III colon cancer after surgery: a retrospective cohort study in the SEER database[J]. Ann Ital Chir, 2024, 95(2): 144-154.
[8] Davis SE, Lasko TA, Chen GH, et al. Calibration drift in regression and machine learning models for acute kidney injury[J]. J Am Med Inform Assoc, 2017, 24(6): 1052-1061.
[9] Kamada T, Ohdaira H, Takahashi J, et al. Novel tumor marker index using carcinoembryonic antigen and carbohydrate antigen 19-9 is a significant prognostic factor for resectable colorectal cancer[J]. Sci Rep, 2024, 14(1): 4192. doi:10.1038/s41598-024-54917-w
[10] Ren GM, Li RK, Zheng GZ, et al. Prognostic value of normal levels of preoperative tumor markers in colorectal cancer[J]. Sci Rep, 2023, 13(1): 22830. doi:10.1038/s41598-023-49832-5
[11] Li ZH, Li CX, Pu HJ, et al. Trajectories of perioperative serum carcinoembryonic antigen and colorectal cancer outcome: a retrospective, multicenter longitudinal cohort study[J]. Clin Transl Med, 2021, 11(2): e293. doi:10.1002/ctm2.293
[12] You WQ, Yan L, Cai ZR, et al. Clinical significances of positive postoperative serum CEA and post-preoperative CEA increment in stage II and III colorectal cancer: a multicenter retrospective study[J]. Front Oncol, 2020, 10: 671. doi:10.3389/fonc.2020.00671
[13] Li ZH, Zhu HB, Pang XL, et al. Preoperative serum CA19-9 should be routinely measured in the colorectal patients with preoperative normal serum CEA: a multicenter retrospective cohort study[J].BMC Cancer, 2022, 22(1): 962. doi:10.1186/s12885-022-10051-2
[14] Zhu J, Hao J, Ma Q, et al.A novel prognostic model and practical nomogram for predicting the outcomes of colorectal cancer: based on tumor biomarkers and log odds of positive lymph node scheme[J]. Front Oncol, 2021, 11: 661040. doi:10.3389/fonc.2021.661040
[15] Kinnier CV, Asare EA, Mohanty S, et al. Risk prediction tools in surgical oncology[J]. J Surg Oncol, 2014, 110(5): 500-508.
[16] Chang GJ, Hu CY, Eng C, et al. Practical application of a calculator for conditional survival in colon cancer[J]. J Clin Oncol, 2009, 27(35): 5938-5943.
[17] Parr H, Porta N, Tree AC, et al. A personalized clinical dynamic prediction model to characterize prognosis for patients with localized prostate cancer: analysis of the CHHiP phase 3 trial[J]. Int J Radiat Oncol Biol Phys, 2023, 116(5): 1055-1068.
[18] Wijnands AM, Penning de Vries BBL, Lutgens MWMD, et al. Dynamic prediction of advanced colorectal neoplasia in inflammatory bowel disease[J]. Clin Gastroenterol Hepatol, 2024, 22(8): 1697-1708.
[19] Li K, Luo S. Dynamic prediction of Alzheimers disease progression using features of multiple longitudinal outcomes and time-to-event data[J]. Stat Med, 2019, 38(24): 4804-4818.
[20] Ren XH, Lin J, Stebbins GT, et al. Prognostic modeling of Parkinsons disease progression using early longitudinal patterns of change[J]. Mov Disord, 2021, 36(12): 2853-2861.
[21] Happ C, Greven S. Multivariate functional principal component analysis for data observed on different(dimensional)domains[J]. J Am Stat Assoc, 2018, 113(522): 649-659.
[22] Wichitaksorn N, Boris Choy ST, Gerlach R. A genera-lized class of skew distributions and associated robust quantile regression models[J]. Can J Statistics, 2014, 42(4): 579-596.
[23] Sène M, Taylor JM, Dignam JJ, et al. Individualized dynamic prediction of prostate cancer recurrence with and without the initiation of a second treatment: development and validation[J]. Stat Methods Med Res, 2016, 25(6): 2972-2991.
[24] 李春霞. 结直肠癌围手术期血清肿瘤标志物轨迹分析及动态预测研究[D]. 济南: 山东大学, 2022.
[25] Yao F, Müller HG, Wang JL. Functional data analysis for sparse longitudinal data[J]. J Am Stat Assoc, 2005, 100(470): 577-590.
[26] Lyu Y, Liu Y, Xiao X, et al. High level of intraoperative lactate might predict acute kidney injury in aortic arch surgery via minimally invasive approach in patients with type A dissection[J]. Front Cardiovasc Med, 2023, 10: 1188393. doi:10.3389/fcvm.2023.1188393
[27] Alba AC, Agoritsas T, Walsh M, et al. Discrimination and calibration of clinical prediction models: users guides to the medical literature[J]. JAMA, 2017, 318(14): 1377-1384.
[28] Happ C, Greven S. Multivariate functional principal component analysis for data observed on different(dimensional)domains[J]. J Am Stat Assoc, 2018, 113(522): 649-659.
[29] Therneau TM. A package for survival analysis in R[EB/OL].(2023-03-12)[2024-05-01]. R package version 3.5-5, 2023. https://cran.r-project.org/web/packages/survival/vignettes/survival.pdf
[30] Ushigome M, Shimada H, Miura Y, et al. Changing pattern of tumor markers in recurrent colorectal cancer patients before surgery to recurrence: serum p53 antibodies, CA19-9 and CEA[J]. Int J Clin Oncol, 2020, 25(4): 622-632.
[31] Sonoda H, Yamada T, Matsuda A, et al. Elevated serum carcinoembryonic antigen level after curative surgery is a prognostic biomarker of stage II-III colorectal cancer[J]. Eur J Surg Oncol, 2021, 47(11): 2880-2887.
[32] Zhou SY, Sheng NQ, Ren JZ, et al. Clinical significance of and predictive risk factors for the postoperative elevation of carcinoembryonic antigen in patients with non-metastatic colorectal cancer[J]. Front Oncol, 2021, 11: 741309. doi:10.3389/fonc.2021.741309
[33] Lee JO, Kim M, Lee JH, et al. Carbohydrate antigen 19-9 plus carcinoembryonic antigen for prognosis in colorectal cancer: an observational study[J]. Colorectal Dis, 2023, 25(2): 272-281.
[34] Proust-Lima C, Dartigues JF, Jacqmin-Gadda H. Joint modeling of repeated multivariate cognitive measures and competing risks of dementia and death: a latent process and latent class approach[J]. Stat Med, 2016, 35(3): 382-398.
[35] Elashoff RM, Li G, Li N. A joint model for longitudinal measurements and survival data in the presence of multiple failure types[J]. Biometrics, 2008, 64(3): 762-771.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed