Advances in Clinical Medicine
Vol. 13  No. 04 ( 2023 ), Article ID: 64155 , 9 pages
10.12677/ACM.2023.134848

应用倾向评分匹配法探讨癌结节对胃癌预后的影响

杨迎新1,2,毛伟征1

1青岛大学附属青岛市立医院普通外科,山东 青岛

2济宁医学院附属金乡医院胃肠腹壁疝外科,山东 济宁

收稿日期:2023年3月17日;录用日期:2023年4月13日;发布日期:2023年4月20日

摘要

目的:探讨癌结节(tumor deposits, TDs)对胃癌预后的影响。方法:研究采取回顾性病例对照研究方法,收集2015年1月1日至2019年10月30日在青岛市立医院行根治性手术治疗的胃癌患者445例的临床病理资料,根据TDs是否为阴性,分为TDs阳性组(61例)和TDs阴性组(384例)。应用倾向评分匹配法(PSM)对患者的年龄、性别、术前CEA、术前Ca199、肿瘤部位、T分期、神经侵犯、脉管癌栓、腹水状态、肿瘤大小和转移淋巴结数量11个因素按照最小半径法1:2进行匹配。采用Kaplan-Meier法进行生存分析,两组间生存率的比较,采用log⁃rank检验;采用Cox回归分析进行多因素分析。分析TDs对胃癌患者3年生存的影响。结果:匹配后,TDs阳性组44例,TDs阴性组73例,两组患者基线资料均衡(均P > 0.05)。全组中位随访时间为43.0 (10.0~86.0)个月,TDs阳性组中有28例(63.6%)死亡,其中28例死于胃癌;TDs阴性组中,30例(41.1%)死亡,其中29例死于胃癌,1例死于其他疾病。TDs阳性组患者3年生存率低于阴性组(54.5%比82.2%),差异有统计学意义(χ2 = 10.404, P = 0.001)。不同TDs数量亚组分析提示:在3年总体生存率方面,多发TDs组低于单发TDs组(37.5%比75.0%,χ2 = 6.148, P = 0.013)。单因素分析结果提示:T分期(P = 0.004)、术前Ca199升高(P = 0.017)、肿瘤大小(P = 0.012)、N分期(P < 0.001)、TDs (P = 0.014)与本组胃癌患者预后有关。多因素分析提示:TDs阳性(HR = 2.258, 95%CI: 1.288~3.960, P = 0.004)、肿瘤直径 > 6 cm (HR = 2.463, 95%CI: 1.351~4.491, P = 0.003)、T分期高(HR = 1.972, 95%CI: 1.160~3.351, P = 0.012)以及N分期高(HR = 6.017, 95%CI: 3.862~9.373, P < 0.001)是影响胃癌预后的独立危险因素。结论:TDs阳性是影响胃癌患者预后的独立危险因素,临床中可采取针对性治疗措施。

关键词

胃肿瘤,癌结节,倾向评分匹配法,预后

Application of Propensity Score Matching Method to Explore the Influence of Tumor Deposits on the Prognosis of Gastric Cancer

Yingxin Yang1,2, Weizheng Mao1

1General Surgery, Qingdao Municipal Hospital Affiliated to Qingdao University, Qingdao Shandong

2Surgery of Gastrointestinal Abdominal Hernia, Jinxiang Hospital Affiliated to Jining Medical College, Jining Shandong

Received: Mar. 17th, 2023; accepted: Apr. 13th, 2023; published: Apr. 20th, 2023

ABSTRACT

Objective: To investigate the influence of tumor deposits (TDs) on the prognosis of gastric cancer. Methods: A retrospective case-control study was conducted to collect the clinicopathological data of 445 patients with gastric cancer who underwent radical surgery in Qingdao Municipal Hospital from January 1, 2015 to October 30, 2019. According to the TDs status, they were divided into TDs positive group (61 cases) and TDs negative group (384 cases). Eleven factors including age, sex, preoperative CEA, preoperative CA19-9, tumor site, T stage, nerve invasion, vascular tumor thrombus, ascites status, tumor size, and the number of metastatic lymph nodes were matched according to the minimum radius method (PSM) of 1:2. Kaplan Meier method was used to analyze the survival rate, and log-rank test was used to compare the survival rate between the two groups. Cox regression analysis was used for multivariate analysis, to analyze the effects of TDs on gastric cancer patients for 3 years survival impact. Results: After matching, there were 44 cases in TDs positive group and 73 cases in TDs negative group. The baseline data of patients in the two groups were balanced (P > 0.05). The median follow-up time was 43.0 (10.0~86.0) months in the whole group. Twenty-eight cases (63.6%) in the TDs positive group died, of which 28 cases died of gastric cancer. In the TDs-negative group, 30 cases (41.1%) died, including 29 cases of gastric cancer and 1 case of other diseases. The 3-year survival rate of TDs positive group was lower than that of TDs negative group (54.5% vs 82.2%) (χ2 = 10.404, P = 0.001). Subgroup analyses with different numbers of TDs suggested that in terms of 3-year overall survival rate, the multiple-TDs group was lower than the single-TDs group (37.5% vs 75.0%, χ2 = 6.148, P = 0.013). The results of univariate analysis suggested that T stage (P = 0.004), preoperative increased CA19-9 (P = 0.017), tumor size (P = 0.012), N stage (P < 0.001), and TDs (P = 0.014) were related to the prognosis of gastric cancer in this group. Multivariate analysis revealed that positive TDs (HR = 2.258, 95%CI: 1.288~3.960, P = 0.004), tumor diameter > 6 cm (HR = 2.463, 95%CI: 1.351~4.491, P = 0.003), high T stage (HR = 1.972, 95%CI: 1.160~3.351, P = 0.012) and high N stage (HR = 6.017, 95%CI: 3.862~9.373, P < 0.001) were independent prognostic factors for gastric cancer. Conclusion: According to propensity score matching, perigastric TDs positive is an independent risk factor affecting the prognosis of gastric cancer patients. Targeted treatment measures can be taken in clinic.

Keywords:Gastric Neoplasms, Tumor Deposits, Propensity Score Matching, Prognosis

Copyright © 2023 by author(s) and Hans Publishers Inc.

This work is licensed under the Creative Commons Attribution International License (CC BY 4.0).

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

1. 引言

胃癌是严重威胁人类健康的恶性肿瘤之一,在世界范围内,发病率在全球排名第五,死亡率在全球排名第四 [1] 。随着手术技术规范、生物基因技术、化疗,靶向治疗及免疫治疗的发展,胃癌的治疗水平已有明显提高。了解与胃癌预后相关的因素,对患者预后的正确判断、选择最合适的诊疗方案是巩固和提高治疗效果的重要环节。诸多研究表明,肿瘤TNM分期、神经侵犯、血管侵犯和肿瘤部位等与胃癌预后有关 [2] [3] [4] 。癌结节(tumor deposits, TDs)也是近些年来学者们研究较多的胃癌患者的显著病理特征。2017年AJCC发布的第8版胃癌TNM分期系统中,首次推荐将TDs进行记录,并作为区域淋巴结计算,但这只是仅作为经验性推荐,并没有高等级循证医学证据支持,其对分期和预后的影响仍存在诸多争议。因此,本研究通过应用倾向性评分匹配法(propensity score matching, PSM)克服选择性偏倚,减少这些偏差和混杂变量的影响,探讨TDs与胃癌预后的关系,为临床分期和治疗提供参考。

2. 资料和方法

2.1. 一般资料

病例纳入标准:1) 病理学确诊为胃腺癌;2) 行胃癌根治手术治疗;3) 无远处转移;4) 临床病理资料完整;5) 按照第8版TNM分期指南推荐方案行标准化疗。排除标准:1) 其他病理类型的胃癌,如鳞癌、腺鳞癌和神经内分泌癌;2) 合并其他恶性肿瘤者;3) 术前接受新辅助化疗者;4) 残胃癌。根据上述标准收集2015年1月至2019年10月期间在青岛市立医院东院区普外科收治的胃癌患者445例。其中男117例(78.5%),女32例(21.5%);年龄28~89岁,平均年龄64.0 ± 10.8岁。所有患者术前对治疗内容知情并签署手术同意书,本研究经我院伦理委员会讨论通过。

2.2. TDs的判定

根据第8版UICC/AJCC胃癌TNM分期系统,将TDs定义为:在胃周淋巴结引流区域内,与胃周脂肪组织相邻,独立存在的肿瘤结节,并且无可识别的淋巴结、淋巴管、神经或血管的结构 [5] 。由2名病理科医师对每例患者的术后组织切片阅片,进行TDs病理诊断。并根据TDs阳性和阴性进行分组。

2.3. 临床病理资料收集

查阅患者病历,收集患者的性别、年龄、术前CEA、术前Ca199、肿瘤大小、肿瘤位置、浸润深度、淋巴结转移、脉管浸润、神经浸润、转移淋巴结数量(N分期)。

2.4. 随访方法

通过电话、社交软件、门诊复查、住院检查等方式对患者进行随访,术后第1个月随访;术后2年内每3个月随访1次;术后3年每6个月随访1次;总生存时间定义为病理确诊到患者死亡或随访终点的时间。截止随访时间为2022年10月。

2.5. 统计学方法

采用SPSS 26.0软件进行数据分析。除观察因素外,对预后有影响、且基本资料不平衡的可观察变量作为协变量,纳入二分类的logistic回归模型中,产生0~1间的匹配评分,匹配精度设置为0.02,以最小半径法1:2进行匹配后,对患者进行亚组分析。可能与预后相关的基本资料变量是年龄、性别、术前CEA、术前Ca199、肿瘤部位、T分期、神经侵犯、脉管癌栓、腹水状态、肿瘤大小、和转移淋巴结数量。计数资料用[例(%)]表示,独立两组间比较采用χ2检验;独立的两组有序分类资料则采用Mann Whitney U检验。生存分析采用Kaplan-Meier法进行描述,两组间生存率的比较采用log-rank检验。采用Cox回归模型进行单因素及多因素分析,将单因素分析P < 0.1的变量纳入多因素分析。P < 0.05表示差异有统计学意义。

3. 结果

本研究445例患者中,TDs阳性者61例(13.7%)。两组患者的年龄、性别、肿瘤部位、腹水状态的比较,差异均无统计学意义(均P > 0.05);但两组患者的术前CEA,术前CA19-9、神经浸润、脉管浸润、肿瘤大小、浸润深度(T分期)及转移淋巴结(N分期)比较差异均有统计学意义(均P < 0.05),见表1。根据以上11个因素匹配后,TDs阳性组44例、TDs阴性组73例患者;两组间临床基本资料比较,差异无统计学意义(均P > 0.05),说明两组资料均衡,具有可比性,见表1

Table 1. Comparison of clinical basic data of patients with gastric cancer between TDs negative group and positive group before and after propensity score matching [case (%)]

表1. 倾向评分匹配前后TDs阴性组与阳性组胃癌患者临床基本资料的比较[例(%)]

(一) 匹配后TDs阴性组与阳性组患者预后比较

全组中位随访时间为43.0 (10.0~86.0)个月,TDs阳性组中有28例(63.6%)死亡,其中28例死于胃癌;TDs阴性组中,30例(41.1%)死亡,其中29例死于胃癌,1例死于其他疾病。TDs阳性组患者3年生存率低于阴性组(54.5%比82.2%),差异有统计学意义(χ2 = 10.404, P = 0.001),见图1。不同TDs数量亚组分析提示:在3年总体生存率方面,多发癌结节组低于单发癌结节组(37.5%比75.0%,χ2 = 6.148,P = 0.013),见图2

Figure 1. Comparison of overall survival curves between positive and negative groups of TDs after matching

图1. 匹配后癌结节阳性组和阴性组总体生存曲线比较

Figure 2. Comparison of survival curves between multiple TDs group and single TDs group

图2. 多发癌结节组和单发癌结节组生存曲线比较

(二) TDs与胃癌预后的关系

Cox单因素分析提示:T分期、术前Ca199升高、肿瘤大小、腹水、N分期、TDs与本组胃癌患者预后有关(均P < 0.05)。多因素分析提示:TDs阳性、肿瘤直径大、T分期高以及N分期高是影响胃癌预后的独立危险因素(P < 0.05)。见表2

Table 2. Univariate and multivariate analysis of gastric cancer patients before and after propensity score matching

表2. 倾向评分匹配前后胃癌患者单因素和多因素分析

注:“-”表示无数据;将单因素分析P < 0.1的因素纳入多因素分析中。

4. 讨论

TDs首次由Gabriel等于1935年在结直肠癌标本中发现并报道 [6] ,随着研究的逐渐深入,诸多研究表明TDs是影响结直肠癌患者预后的重要因素之一 [7] [8] 。1997年第5版AJCC结直肠癌分期指南首次纳入TDs的概念。然后随着研究的深入,TDs的判定标准及其组织特征被多次修订,并在AJCC第7版结直肠癌TNM分期指南将肠周发现TDs但无局部淋巴结转移的结直肠癌纳入到病理Nc分期。而AJCC第8版胃癌TNM分期系统推荐将TDs按转移淋巴结计算 [5] ,通过合并TDs增加了TNM分期的准确性,但目前这些研究存在样本量小、单中心研究等局限性,而且这些研究中患者的基线可比性很差。倾向性评分匹配法(propensity score matching, PSM)是由Rosenbaum等提出 [9] ,PSM的工作原理是通过从原始对照组中选择一部分受试者来创建一个新的对照组——新组通过匹配消除队列中的协变量不平衡,从而将观察性研究近似于随机对照研究。本研究通过PSM评分,两组临床基本资料对比,差异无统计学意义(均P > 0.05),说明两组资料均衡,具有可比性,减少这些偏差和混杂变量的影响。

TDs的成因和转移机制的机制尚不明确。Goldstein等研究认为TDs是一种比较特殊类型的脉管侵袭,并逐渐增殖形成的肿瘤性结节 [10] 。而Wünsch等通过连续病理切片,根据TDs的镜下形态和组织来源,将TDs分为血管侵袭型、淋巴管侵袭型、神经浸润型、孤立结节型和脉管内TDs等 [11] 。在不同的研究中证实,TDs的形成与肿瘤血管浸润、淋巴结转移、神经浸润、肿瘤浸润深度等因素相关 [12] [13] [14] 。目前对TDs形成的观点主要有以下几种:1) 淋巴结转移,癌细胞侵蚀了肿瘤周围的淋巴结,最终使淋巴结原有的特征组织丧失。2) 血管侵袭,癌细胞在血行转移过程中侵出静脉管壁外增殖形成。3) 淋巴侵袭,癌细胞在淋巴管转移途中,侵及管壁并于管壁外增殖形成。4) 神经周围浸润。5) 肿瘤的直接浸润或散播 [15] 。

Etoh等研究发现TDs阳性胃癌患者较阴性患者预后更差,更多伴随肝转移和腹膜转移 [14] 。Tan等研究发现TDs是影响胃癌患者预后的独立危险因素,TDs阳性患者5年生存率仅19.7%,并认为TDs可能是晚期胃癌的表现特点 [16] 。Jiang等研究认为TDs阳性肿瘤更具有侵袭性,是预后独立危险因素,并且是远处复发、腹膜复发的独立危险因素 [17] 。我们通过PSM后研究发现:TDs阳性、肿瘤直径大、T分期晚以及N分期晚是影响胃癌预后的独立危险因素(P < 0.05)。TDs阳性组患者3年生存率低于阴性组(54.5%比82.2%),并且在3年总体生存率方面,多发组低于单发TDs组(37.5%比75.0%,χ2 =6.148,P = 0.013)。TDs的数量同样影响胃癌患者的预后,TDs以转移淋巴结计入TNM分期增加了对胃癌患者预后预测的准确性,并且在一些研究中得到了证实 [17] [18] 。但这同样不能非常精确的预测预后,与相同数量的转移淋巴结相比,TDs对胃癌患者的预后的影响明显是不同的,Etoh等研究发现随着TDs数目的增加,胃癌患者的5年生存率迅速下降.而不是表现为线性关系 [14] 。许多学者将TDs以不同的形式纳入TNM分期中,以期提高TDs对胃癌患者预后预测的准确性:Anup等研究发现合并胃周TDs的每个T分期和T4期预后相似,类似于浆膜侵犯,建议将胃周TDs患者以T4a纳入TNM分期 [19] 。Gu等则建议将TDs阳性患者除ⅢC期外,TNM分期依次提升一级 [20] ;Chen等同样建议将TDs纳入N分期,将TDs阳性患者的N分期提升一级 [12] 。而Wang等则根据TDs的数量纳入TNM,建议将TDs数量为1~2枚,定为N3;若TDs数量为3枚或以上则定为远处转移(M1) [21] 。但目前这些研究存在样本量小、回顾性研究、单中心研究等局限性,仍不能准确预测TDs对胃癌预后影响,有待于更多的大样本量、多中心的研究证实。

5. 结论

综上所述,本研究经过PSM研究发现,TDs是影响胃癌预后的独立危险因素。但本研究为回顾性分析,虽然采用PSM克服选择偏移,但TDs阳性病例数较少,随访时间较短,如何准确的评价TDs的预后价值,合理的纳入临床分期中,仍有需要大样本量的临床研究分析,同时也需要在TDs的形成及转移机制上进行更深入的基础研究和探索。

文章引用

杨迎新,毛伟征. 应用倾向评分匹配法探讨癌结节对胃癌预后的影响
Application of Propensity Score Matching Method to Explore the Influence of Tumor Deposits on the Prognosis of Gastric Cancer[J]. 临床医学进展, 2023, 13(04): 6012-6020. https://doi.org/10.12677/ACM.2023.134848

参考文献

  1. 1. Sung, H., Ferlay, J., Siegel, R.L., et al. (2021) Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: A Cancer Journal for Clinicians, 71, 209-249. https://doi.org/10.3322/caac.21660

  2. 2. 刘传绪, 王熙勋, 胡金晨, 等. 神经浸润与胃癌其他临床病理因素的关系及预后分析[J]. 中华内分泌外科杂志, 2022, 16(2): 216-220.

  3. 3. 庞芳宁, 白槟, 杨鹏, 等. 不同部位原发性胃癌临床病理特征对比研究[J]. 陕西医学杂志, 2020, 49(5): 626-629.

  4. 4. 陈昊, 唐兆庆, 刘凤林. 癌结节的定义及其在胃癌诊疗中的临床意义[J]. 中华胃肠外科杂志, 2019(1): 94-97.

  5. 5. Edge, S.B. and Edge, S.B. (2017) AJCC Cancer Staging Manual. 8th Edition, Springer, New York.

  6. 6. Gabriel, W.B., Dukes, C. and Bussey, H.J.R. (1935) Lymphatic Spread in Cancer of the Rectum. Journal de Chirurgie (Paris), 23, 395-413. https://doi.org/10.1002/bjs.1800239017

  7. 7. Jin, M., Roth, R., Rock, J.B., et al. (2015) The Impact of Tumor De-posits on Colonic Adenocarcinoma AJCC TNM Staging and Outcome. The American Journal of Surgical Pathology, 39, 109-115. https://doi.org/10.1097/PAS.0000000000000320

  8. 8. Zheng, K., Zheng, N., Xin, C., et al. (2020) The Prognostic Significance of Tumor Deposit Count for Colorectal Cancer Patients after Radical Surgery. Gastroenterology Research and Practice, 2020, Article ID: 2052561. https://doi.org/10.1155/2020/2052561

  9. 9. Rosenbaum, P.R. and Rubin, D.B. (1983) The Central Role of the Pro-pensity Score in Observational Studies for Causal Effects. Biometrika, 70, 41-55. https://doi.org/10.1093/biomet/70.1.41

  10. 10. Goldstein, N.S. and Turner, J.R. (2000) Pericolonic Tumor Deposits in Patients with T3N+MO Colon Adenocarcinomas: Markers of Reduced Disease Free Survival and Intra-Abdominal Me-tastases and Their Implications for TNM Classification. Cancer, 88, 2228-2238. https://doi.org/10.1002/(SICI)1097-0142(20000515)88:10<2228::AID-CNCR5>3.0.CO;2-1

  11. 11. Wünsch, K., Müller, J., Jähnig, H., et al. (2010) Shape Is Not Associated with the Origin of Pericolonic Tumor Deposits. American Journal of Clinical Pathology, 133, 388-394. https://doi.org/10.1309/AJCPAWOLX7ADZQ2K

  12. 12. Chen, H., Tang, Z., Chen, L., et al. (2018) Evaluation of the Impact of Tumor Deposits on Prognosis in Gastric Cancer and a Pro-posal for Their Incorporation into the AJCC Staging System. European Journal of Surgical Oncology, 44, 1990-1996. https://doi.org/10.1016/j.ejso.2018.10.062

  13. 13. Zhang, H.D., Tang, P., Duan, X.F., et al. (2013) Extranodal Me-tastasis Is a Powerful Prognostic Factor in Patients with Adenocarcinoma of the Esophagogastric Junction. Journal of Surgical Oncology, 108, 542-549. https://doi.org/10.1002/jso.23430

  14. 14. Etoh, T., Sasako, M., Ishikawa, K., Katai, H., Sano, T. and Shimoda, T. (2006) Extranodal Metastasis Is an Indicator of Poor Prognosis in Patients with Gastric Carcinoma. British Journal of Surgery, 93, 369-373. https://doi.org/10.1002/bjs.5240

  15. 15. Ersen, A., Unlu, M.S., Akman, T., et al. (2014) Tumor Deposits in Gastric Carcinomas. Pathology—Research and Practice, 210, 565-570. https://doi.org/10.1016/j.prp.2014.03.006

  16. 16. Tan, J., Yang, B., Xu, Z., et al. (2019) Tumor Deposit Indicates Worse Prognosis than Metastatic Lymph Node in Gastric Cancer: A Propensity Score Matching Study. Annals of Translational Medicine, 7, 671. https://doi.org/10.21037/atm.2019.10.33

  17. 17. Jiang, N., Deng, J.Y., Ding, X.W., Ke, B., Liu, N. and Liang, H. (2014) Node-Extranodal Soft Tissue Stage Based on Extranodal Metastasis Is Associated with Poor Prognosis of Pa-tients with Gastric Cancer. Journal of Surgical Research, 192, 90-97. https://doi.org/10.1016/j.jss.2014.05.053

  18. 18. Lee, H.S., Lee, H.E., Yang, H.K. and Kim, W.H. (2013) Perigastric Tumor Deposits in Primary Gastric Cancer: Implications for Patient Prognosis and Staging. Annals of Surgical Oncology, 20, 1604-1613. https://doi.org/10.1245/s10434-012-2692-9

  19. 19. Anup, S., Lu, J., Zheng, C.H., et al. (2017) Prognostic Signifi-cance of Perigastric Tumor Deposits in Patients with Primary Gastric Cancer. BMC Surgery, 17, 84. https://doi.org/10.1186/s12893-017-0280-4

  20. 20. Gu, L., Chen, P., Su, H., et al. (2020) Clinical Significance of Tumor Deposits in Gastric Cancer: A Retrospective and Propensity Score-Matched Study at Two Institutions. Journal of Gastrointestinal Surgery, 24, 2482-2490. https://doi.org/10.1007/s11605-019-04421-8

  21. 21. Wang, W., Li, Y., Zhang, Y., et al. (2011) Incorporation of Ex-tranodal Metastasis of Gastric Carcinoma into the 7th Edition UICC TNM Staging System. PLOS ONE, 6, e19557. https://doi.org/10.1371/journal.pone.0019557

期刊菜单