World Journal of Cancer Research
Vol. 10  No. 01 ( 2020 ), Article ID: 33910 , 8 pages
10.12677/WJCR.2020.101003

Risk Factors for Secondary Brain Metastasis in Lung Adenocarcinoma Based on Competitive Risk Model

Xin Song, Ruiqi Xue, Hongwei Li*, Xiaqin Zhang

Department of Radiation Oncology, Shanxi Cancer Hospital, Taiyuan Shanxi

Received: Dec. 21st, 2019; accepted: Jan. 3rd, 2020; published: Jan. 10th, 2020

ABSTRACT

Objective: To explore the influencing factors of brain metastasis in lung adenocarcinoma using competitive risk model and make predictive analysis. Materials and methods: A total of 938 patients with pathologically confirmed lung adenocarcinoma admitted to Shanxi Cancer Hospital from August 2010 to May 2018 were collected, including 222 patients with brain metastasis and 718 patients without brain metastasis. Brain metastases from lung adenocarcinoma as transient, and brain metastases before death as two absorbing states respectively, brain metastases die before competition risk events for brain metastases, build competitive risk model, get the model parameters, analyze the influence factors of brain metastases from lung adenocarcinoma, and conduct the risk assessment of brain metastasis in lung adenocarcinoma. Results: Univariate analysis showed that gender, smoking status, EGFR mutation status, T stage, N stage, presence or absence of extracranial metastasis and different treatment methods were possible risk factors for brain metastasis in lung adenocarcinoma. Multivariate competitive risk model screening showed that N3 stage patients with extracranial metastasis were significantly correlated with increased risk of brain metastasis in lung adenocarcinoma, and the 2-year brain metastasis rate in 194 patients with N3 stage patients with extracranial metastasis was 80.932% (95%ci = 77.325% - 84.539%). Conclusion: According to the competitive risk model, stage N3 and extracranial metastasis are the high-risk subgroups of brain metastasis in patients with lung adenocarcinoma. Prophylactic brain irradiation is the most likely benefit. Future clinical studies on prophylactic brain irradiation should focus on this high-risk subgroup.

Keywords:Lung Adenocarcinoma, Brain Metastasis, Competitive Risk Model

基于竞争风险模型的肺腺癌继发脑转移危险因素分析

宋欣,薛瑞琪,李红卫*,张霞琴

山西省肿瘤医院放疗科,山西 太原

收稿日期:2019年12月21日;录用日期:2020年1月3日;发布日期:2020年1月10日

摘 要

目的:应用竞争风险模型探讨肺腺癌发生脑转移的影响因素并进行预测分析。材料和方法:收集2010年8月至2018年5月山西省肿瘤医院收治的938例经病理确诊的肺腺癌的患者,其中发生脑转移者222例,未发生脑转移者716例。以肺腺癌作为暂态,脑转移与发生脑转移前死亡分别作为两个吸收态,发生脑转移前死亡为脑转移的竞争风险事件,构建竞争风险模型,获得模型参数,分析肺腺癌发生脑转移的影响因素,从而对肺腺癌发生脑转移进行风险评估。结果:单因素分析结果显示,性别、吸烟状态、EGFR突变状态、T分期、N分期、有无颅外转移以及不同的治疗方法是促进肺腺癌发生脑转移的可能危险因素。多因素竞争风险模型筛选,N3分期且有颅外转移与肺腺癌发生脑转移风险升高显著相关,N3分期且有颅外转移的患者(194人) 2年发生脑转移率为80.932% (95%CI = 77.325%~84.539%)。结论:根据竞争风险模型,肺腺癌患者中,N3期和颅外转移是脑转移的高危亚组。最可能获益于预防性脑照射。将来预防性脑照射的临床研究应该聚焦于该高危亚组。

关键词 :肺腺癌,脑转移,竞争风险模型

Copyright © 2020 by author(s) and Hans Publishers Inc.

This work is licensed under the Creative Commons Attribution International License (CC BY).

http://creativecommons.org/licenses/by/4.0/

1. 引言

世界范围内,肺癌是癌症死亡的首要因素。肺癌的发病率及死亡率居恶性肿瘤的首位,其中80%为非小细胞肺癌(non-small cell lung cancer,NSCLC),腺癌是非小细胞肺癌中最常见的病理类型,约占非小细胞肺癌的40%。大约25%~40%的非小细胞肺癌患者会发生脑转移 [1] [2]。脑转移的发生不仅是非小细胞肺癌晚期重要临床表现之一,也是主要的致死因素 [3]。局部晚期非小细胞肺癌脑转移危险因素的探讨是近年NSCLC研究领域的热点课题 [4]。以往对于非小细胞肺癌的研究中,往往把脑转移作为终点事件,而忽略发生脑转移前死亡的患者数据。竞争风险模型(Competing Risk Model)是一种用于处理多种潜在结局生存数据(包含竞争风险事件)的分析方法,这些数据包括失效的时间跨度和导致失效的终点事件,终点事件可能有多个,如果一个事件的发生导致另一个事件不会发生,则前者称为后者的“竞争风险事件” [5]。本研究采用竞争风险模型探讨真实世界肺腺癌发生脑转移的危险因素,对判断预后、指导个体化治疗有重要意义。

2. 资料与方法

2.1. 临床资料

收集2010年8月至2018年5月在我院就诊的1063例肺腺癌患者,采用外科切除、气管镜、CT或彩超引导下穿刺获取肿瘤标本。所有患者均有完整的影像学资料,脑转移的患者均经CT或MRI确诊。剔除125例确诊时即已发生脑转移患者,得到938例符合要求病例。患者临床特征包括年龄、性别、吸烟史、PS评分、原发肿瘤TNM分期(2018版)、EGFR突变状态、治疗方法以及有无颅外转移等。所有患者均进行EGFR组织学检测,2010年8月至2013年3月间采用直接测序法检测EGFR突变,2013年3月至2018年5月采用突变特异性扩增系统法检测EGFR突变。组织标本采用原发病灶或转移灶组织。基因检测位点包括18、19、20、21外显子。

2.2. 治疗方法

手术方法 患者接受了肺叶切除或一侧全肺切除,术中行纵隔淋巴结完全解剖或系统取样。放疗采用直线加速器实施三维适形放疗和调强适形放疗,应用6MV X射线,每次2 Gy,每周5次,总剂量为60 Gy。全身化疗主要采用含铂双药联合方案,包括NP方案(长春瑞滨25 mg/m2 + 顺铂80 mg/Ill2)、TP方案(紫杉醇135~175 mg/1112 + 顺铂75 mg/ITl2)、GP方案(吉西他滨1250 m//m2 + 顺铂75 mg/m2)、PC方案(培美曲塞500 mg/m2 + 顺铂75 mg/Ill2),化疗周期为3~4周期。靶向治疗采用每日口服吉非替尼250 mg (1 7X/d)、厄洛替尼150 mg (1次/d)或埃克替尼125 mg (3次/d),直到病变进展。患者第一年每3个月随访一次,第二、三年每半年一次,之后每年一次。随访截止到2018.8,所有患者行体格检查、全血细胞计数、血生化、胸部CT、腹部B超和其它基于患者症状的必要检查。随访时脑的磁共振影像学检查基于可疑症状的出现。影像学或组织学检查决定疾病进展或转移部位。事件时间为诊断肺腺癌到出现脑转移的时间,如无脑转移则至末次随访。将发生脑转移前死亡作为脑转移的竞争风险事件,即死亡若发生在脑转移之前,则这类人群不会再发生脑转移。

2.3. 统计方法

本研究采用R软件“Cmprsk”软件包实现竞争风险模型分析,单因素采用Gray检验,多因素采用Fine-Gray法。以肺腺癌为唯一暂态。将出现脑转移和死亡定为2个竞争吸收态,纳入可能影响因素,拟合竞争风险的比例风险模型。可能影响因素包括年龄、性别、吸烟状态、EGFR突变状态、原发肿瘤TNM分期、有无颅外转移及不同治疗模式,所有可能影响因素以所收集病例及随访结果为准。

3. 结果

3.1. 临床基线特征

本研究的938例肺腺癌患者男520例。女418例;中位年龄59岁:无吸烟史501例(53.4%),有吸烟史437例(46.6%),其中男性吸烟患者占92.3%,女性占7.7%;EGFR检测敏感突变360例,(19外显子突变190,21外显子突变170)其中123例出现脑转移(34.2%);562人(59.9%)出现远处转移,其中340人(60.5%)仅有颅外转移,累计发生继发脑转移222例,发生脑转移前死亡650例,发生脑转移后死亡193例,222名脑转移患者中T1、T2期109人(49.1%),T3、T4期113人(50.9%),N3期94人(42.3%),所有患者手术+化疗250人,单化疗133人,其中75人(19.6%)出现脑转移,单靶向治疗者134人,放疗+化疗404人,发生脑转移141例(32.2%)。

3.2. 竞争风险模型分析

拟合竞争风险的比例风险模型,按a = 0.05检验水准,单因素和多因素分析发现了与脑转移相关的临床病理因素,见表1。单因素分析显示,性别、吸烟状态、EGFR突变状态、T分期、N分期、有无颅外转移以及不同的治疗方法在竞争风险模型中有统计学意义(P < 0.05)。在多因素分析中,N分期(SHR = 0.629, 95%CI: 0.502~0.787) (图1)、颅外转移(SHR = 0.123, 95%CI: 0.039~0.393) (图2)与脑转移风险升高显著相关。

Table 1. Regression results of competitive risk model for brain metastasis of lung adenocarcinoma

表1. 肺腺癌脑转移的竞争风险模型回归结果

Figure 1. Risk comparison of brain metastasis in different N stages

图1. 不同N分期脑转移的风险比较

Figure 2. Risk comparison of brain metastasis with and without extracranial metastasis

图2. 有颅外转移和没有颅外转移的脑转移的风险比较

3.3. 转归预测

基于竞争风险模型预测2年(24个月)脑转移概率:938例患者出现脑转移32.753% (95%CI = 32.649%~32.856%) (图3)。

Figure 3. Risk of brain metastasis in the whole group

图3. 全组患者的脑转移风险

无颅外转移且淋巴结分期小于3的患者(人数 = 380)脑转移率是20.821%。无颅外转移且淋巴结分期等于3的患者(人数 = 94)脑转移率是40.135%。有颅外转移且淋巴结分期小于3的患者(人数 = 270)脑转移率是40.345%。有颅外转移且淋巴结分期等于3的患者(人数=194)脑转移率是80.932%。两个危险因素不同组合的患者脑转移风险的差别有统计学意义(P < 0.001) (图4)。

Figure 4. Comparison of brain metastasis risk in patients with different combinations of two risk factors

图4. 两个危险因素不同组合的患者脑转移风险的比较

4. 讨论

脑转移是晚期肺癌病人最容易出现的主要症状之一,脑转移的发生,直接影响肺癌患者的生存时间和预后 [6] [7] [8]。因此,降低脑转移的发生对肺癌患者生存有直接影响。研究显示,非小细胞肺癌的脑转移率明显低于小细胞肺癌,如果不经选择行全脑预防照射,可能引起患者认知功能的下降,很难达到临床获益的结果;选择高危患者给予脑预防照射,可能会降低脑转移的发生率 [9] [10],从而延长肺癌患者的生存时间。因此,分析引起脑转移的高危因素,对预防非小细胞肺癌脑转移具有重要意义。

传统研究报道的非小细胞肺癌脑转移的危险因素包括病理组织学类型 [11] 、EGFR突变类型和状态 [12] 、TNM分期 [13] [14] 不同治疗方法 [15] [16] 等。Hubbs15等发现在早期非小细胞肺癌患者中,原发肿瘤大小与脑转移风险正相关。Mujoomdar [17] 等分析了264名I-IV期非小细胞肺癌患者,提示原发肿瘤大小与肺癌脑转移预测概率成正相关(p < 0.001),原发肿瘤大小不仅是非小细胞肺癌的主要预后因素,也是脑转移的危险因素。Mujoomdar等的研究同时表明:患者的淋巴结分期和非小细胞肺癌脑转移成正相关,认为转移淋巴结的分期、个数、大小、区域等淋巴结转移的程度越严重,脑转移的风险越高。随着对肺癌分子机制研究的深入,EGFR基因突变给非小细胞肺癌尤其是腺癌患者带来了新的治疗希望,但有研究显示EGFR突变的患者有着更高的脑转移率 [18],同时EGFR-TKI治疗可能增加非小细胞肺癌尤其是肺腺癌患者的脑转移率 [19] [20]。

尽管肺腺癌脑转移发生的几率比较高,但在随访过程中发现会有一定数量的腺癌患者在出现脑转移前死亡,死亡竞争风险比较高,因此在肺腺癌脑转移研究中竞争风险模型的应用对于准确评估脑转移出现风险和制定临床决策至关重要。对于肺腺癌脑转移的研究,通常采用常规生存分析方法(单因素分析采用χ2检验,多因素分析用Cox回归模型,生存率计算用Kaplan-Meier法),这种统计分析法对脑转移风险评估不客观,不真实 [21] [22],因为这些传统的统计学方法只考虑单个终点事件(脑转移)。竞争风险模型是多状态模型中的一种标准状态结构 [23]。它可以有效处理竞争风险资料,对所有的状态、结局等可能影响因素的随机过程进行连续性动态研究,以进一步认识影响疾病进展的因素。本研究将发生脑转移前死亡作为脑转移的竞争风险事件,采用竞争风险模型进行统计学分析,探讨肺腺癌脑转移的影响因素并进行转归预测。研究结果表明,938例患者出现脑转移率为32.753%;而N3伴有颅外转移的患者出现脑转移率为80.932%,说明N3伴有颅外转移的患者发生脑转移的风险更大,是肺腺癌发生脑转移的高危因素。最可能获益于预防性脑照射。将来预防性脑照射的临床研究应该聚焦于该高危亚组。

当然,作为回顾性研究,本研究存在不足。1) 研究为回顾性分析,治疗方案及剂量无统一标准,给结果带来一定影响;2) 本研究中远处转移靠影像学来确诊,可能存在一些假阳性或假阴性情况。3) 进行EGFR突变状态检测的患者大多来自优势人群(即亚裔、女性、腺癌、非吸烟患者),不可避免产生选择偏倚,对实验结果可能产生一定影响。

基金项目

山西省科技厅自然科学基金(201701D121169),山西省卫生计生委科研课题(2014053)。

文章引用

宋 欣,薛瑞琪,李红卫,张霞琴. 基于竞争风险模型的肺腺癌继发脑转移危险因素分析
Risk Factors for Secondary Brain Metastasis in Lung Adenocarcinoma Based on Competitive Risk Model[J]. 世界肿瘤研究, 2020, 10(01): 16-23. https://doi.org/10.12677/WJCR.2020.101003

参考文献

  1. 1. Press, R.H., Zhang, C., et al. (2018) Targeted Sequencing and Intracranial Outcomes of Patients with Lung Adenocarcinoma Brain Metastases Treated with Radiotherapy. Cancer, 124, 3586-3595.
    https://doi.org/10.1002/cncr.31589

  2. 2. Tsakonas, G., De Petris, L. and Ekman, S. (2017) Management of Brain Metastasized Non-Small Cell Lung Cancer (NSCLC) from Local Treatment to New Systemic Therapies. Cancer Treatment Reviews, 54, 122-131.
    https://doi.org/10.1016/j.ctrv.2017.02.004

  3. 3. Schuler, M., Wu, Y.L., Hirsh, V., et al. (2016) First-Line Afatinib versus Chemotherapy in Patients with Non-Small Cell Lung Cancer and Common Epidermal Growth Factor Receptor Gene Mutations and Brain Metastases. Journal of Thoracic Oncology, 11, 380-390.
    https://doi.org/10.1016/j.jtho.2015.11.014

  4. 4. Cai, Y., Wang, J.Y. and Liu, H. (2013) Clinical Observation of Whole Brain Radiotherapy Concomitant with Targeted Therapy for Brain Metastasis in Non-Small Cell Lung Cancer Patients with Chemotherapy Failure. Asian Pacific Journal of Cancer Prevention, 14, 5699-5703.
    https://doi.org/10.7314/APJCP.2013.14.10.5699

  5. 5. Kretowska, M. (2018) Tree-Based Models for Survival Data with Competing Risks. Computer Methods & Programs in Biomedicine, 159, 185-198.
    https://doi.org/10.1016/j.cmpb.2018.03.017

  6. 6. Magnuson, W.J., Lester-Coll, N.H., Wu, A.J., et al. (2017) Management of Brain Metastases in Tyrosine Kinase Inhibitor-Naïve Epidermal Growth Factor Receptor-Mutant Non-Small Cell Lung Cancer: A Retrospective Multi-Institutional Analysis. Journal of Clinical Oncology, 35, 1070-1077.
    https://doi.org/10.1200/JCO.2016.69.7144

  7. 7. Shin, D.Y., Na, I.I., Kim, C.H., et al. (2014) EGFR Mutation and Brain Metastasis in Pulmonary Adenocarcinomas. Journal of Thoracic Oncology, 9, 195-199.
    https://doi.org/10.1097/JTO.0000000000000069

  8. 8. Gow, C.H., Chien, C.R., Chang, Y.L., et al. (2008) Radiotherapy in Lung Adenocarcinoma with Brain Metastases: Effects of Activating Epidermal Growth Factor Receptor Mutations on Clinical Response. Clinical Cancer Research, 14, 162-168.
    https://doi.org/10.1158/1078-0432.CCR-07-1468

  9. 9. 张矛, 陈忠民, 崔洪霞, 孙涛. 非小细胞肺癌脑预防照射研究进展[J]. 中华肺部疾病杂志(电子版), 2016, 9(6): 666-670.

  10. 10. Wang, X., Ye, X., Tang, W., et al. (2018) Efficacy and Safety of Radiotherapy plus EGFR-TKIs in NSCLC Patients with Brain Metastases: A Meta-Analysis of Published Data. Translational Oncology, 11, 1119-1127.
    https://doi.org/10.1016/j.tranon.2018.07.003

  11. 11. Franzoi, M.A. and Hortobagyi, G.N. (2019) Leptomeningeal Carcinomatosis in Patients with Breast Cancer. Critical Reviews in Oncology/Hematology, 135, 85-94.
    https://doi.org/10.1016/j.critrevonc.2019.01.020

  12. 12. Barnholtzsloan, J.S., Sloan, A.E., Davis, F.G., et al. (2004) Incidence Proportions of Brain Metastases in Patients Diagnosed (1973 to 2001) in the Metropolitan Detroit Cancer Surveillance System. Journal of Clinical Oncology, 22, 2865-2872.
    https://doi.org/10.1200/JCO.2004.12.149

  13. 13. Gore, E.M., Bae, K., Wong, S.J., et al. (2011) Phase III Comparison of Prophylactic Cranial Irradiation versus Observation in Patients with Locally Advanced Non-Small-Cell Lung Cancer: Primary Analysis of Radiation Therapy Oncology Group Study RTOG 0214. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology, 29, 272-278.
    https://doi.org/10.1200/JCO.2010.29.1609

  14. 14. Jonnalagadda, S., Arcinega, J., Smith, C., et al. (2011) Validation of the Lymph Node Ratio as a Prognostic Factor in Patients with N1 Non-Small Cell Lung Cancer. Cancer, 117, 4724-4731.
    https://doi.org/10.1002/cncr.26093

  15. 15. Wang, S.Y., Ye, X., Ou, W., et al. (2009) Risk of Cerebral Metastases for Postoperative Locally Advanced Non-Small-Cell Lung Cancer. Lung Cancer, 64, 238-243.
    https://doi.org/10.1016/j.lungcan.2008.08.012

  16. 16. 刘宇, 张幸平, 刘辉, 等. 非小细胞肺癌脑转移放射治疗预后因素分析[J]. 中华肿瘤防治杂志, 2014, 21(21): 1719-1722.

  17. 17. Mujoomdar, A., Austin, J.H.M., Malhotra, R., et al. (2007) Clinical Predictors of Metastatic Disease to the Brain from Non-Small Cell Lung Carcinoma: Primary Tumor Size, Cell Type, and Lymph Node Metastases. Radiology, 243, 882-888.
    https://doi.org/10.1148/radiol.2423051707

  18. 18. Iuchi, T., et al. (2015) Frequency of Brain Metastases in Non-Small-Cell Lung Cancer, and Their Association with Epidermal Growth Factor Receptor Mutations. International Journal of Clinical Oncology, 20, 674-679.
    https://doi.org/10.1007/s10147-014-0760-9

  19. 19. Lee, D.W., Shin, D.Y., Kim, J.W., et al. (2014) Additional Prognostic Role of EGFR Activating Mutations in Lung Adenocarcinoma Patients with Brain Metastasis: Integrating with Lung Specific GPA Score. Lung Cancer, 86, 363-368.
    https://doi.org/10.1016/j.lungcan.2014.10.001

  20. 20. Stani, K., Zwitter, M., Turnsek Hitij, N., et al. (2014) Brain Metastases in Lung Adenocarcinoma: Impact of EGFR Mutation Status on Incidence and Survival. Radiology and Oncology, 48, 173-183.
    https://doi.org/10.2478/raon-2014-0016

  21. 21. Zhang, Z.H., Geskus, R.B., Kattan, M.W., Zhang, H.Y., Liu, T.Y., et al. (2017) Nomogram for Survival Analysis in the Presence of Competing Risks. Annals of Translational Medicine, 5, 403.
    https://doi.org/10.21037/atm.2017.07.27

  22. 22. Veena, M., Siwar, A., Gordon, G., et al. (2017) Criteria for Use of Composite End Points for Competing Risks: A Systematic Survey of the Literature with Recommendations. Journal of Clinical Epidemiology, 82, 4-11.
    https://doi.org/10.1016/j.jclinepi.2016.12.001

  23. 23. Zhou, H.Q. and Zhang, Y.X. (2018) Nomogram to Predict Cause-Specific Mortality in Patients with Surgically Resected Stage I Non-Small-Cell Lung Cancer: A Competing Risk Analysis. Clinical Lung Cancer, 19, e195-e203.
    https://doi.org/10.1016/j.cllc.2017.10.016

  24. NOTES

    *通讯作者。

期刊菜单