Diagnostic value of sinus CT score combined with serum allergen sIgE for postoperative recurrence in patients with eosinophilic CRSwNP
-
摘要: 目的 探讨鼻窦CT评分联合血清变应原sIgE构建嗜酸粒细胞(EOS)型慢性鼻窦炎伴鼻息肉(CRSwNP)患者术后复发的风险模型。方法 收集2016年1月—2019年1月在漯河市中心医院接受治疗的183例EOS型CRSwNP患者的临床资料，术后1年评估疗效，根据术后复发情况分为复发组和未复发组。单因素分析临床和病理因素对患者术后的疗效，XGboost模型及多因素Cox分析术后复发的影响因素。绘制两种模型的受试者工作特征(ROC)曲线比较两种模型的预测效果。Kaplan-Meier法绘制生存曲线，比较不同危险级别患者的存活率。结果 Cox多因素分析结果显示术后未坚持综合治疗、组织EOS比例、组织中性粒细胞(NEU)比例、组织淋巴细胞比例、组织浆细胞比例、外周血NEU比例、变应原sIgE、鼻窦CT总评分是影响患者复发的独立危险因素。XGboost模型排名前六的术后复发的影响因素为变应原sIgE、鼻窦CT总评分、组织EOS比例、术后未坚持综合治疗、组织淋巴细胞比例、组织浆细胞比例。ROC曲线显示，XGboost模型的ROC曲线下面积为0.818，较多因素Cox分析(0.789)增加了3.68%，且模型灵敏度、特异度和约登指数明显高于多因素Cox分析模型。将纳入XGboost模型的因素构建术后复发风险模型，高危组存活率显著低于低危组和中危组(log-rank检验值：21.946，P < 0.001)。结论 鼻窦CT评分联合血清变应原sIgE建立的术后复发风险模型可有效预测患者的术后复发率，XGboost模型对EOS型CRSwNP患者术后复发的预测效果优于多因素Cox分析模型，可应用于术后复发的预测。Abstract: Objective To explore the combination of sinus CT score and serum allergen sIgE to construct a postoperative recurrence risk model for patients with eosinophilic CRSwNP.Methods The clinical data of 183 patients with eosinophilic CRSwNP who were treated in Luohe Central Hospital from January 2016 to January 2019 were collected. The curative effect was evaluated one year after the operation. According to the postoperative recurrence, they were divided into recurrence group and non-recurrence group. Single factor analysis of clinical and pathological factors on the postoperative curative effect of patients, XGboost model and multivariate Cox analysis of factors affecting postoperative recurrence. Draw the receiver operating characteristic(ROC) curves of the two models to compare the prediction effects of the XGboost model. The Kaplan-Meier method draws survival curve and compares the recurrence-free survival rate of patients with different risk levels.Results The results of Cox multivariate analysis showed postoperative adherence to comprehensive treatment, tissue EOS ratio, tissue NEU ratio, tissue lymphocyte ratio, tissue plasma cell ratio, peripheral blood NEU ratio, Allergen sIgE and total sinus CT score were independent risk factors for recurrence. The top six factors influencing postoperative recurrence in the XGboost model were allergen sIgE, total sinus CT score, tissue EOS ratio, postoperative adherence to comprehensive treatment, tissue lymphocyte ratio, and tissue plasma cell ratio. The ROC curve showed that the area under the ROC curve of the XGboost model was 0.818. Cox analysis (0.789) with more factors increased by 3.68%, and the sensitivity, specificity and Youden index of the model were significantly higher than the multivariate Cox analysis model. The factors included in the XGboost model were used to construct a postoperative recurrence risk model. The recurrence-free survival rate of high-risk group was significantly lower than that of low-risk group and intermediate-risk group (log-rank test value: 21.946, P < 0.001).Conclusion The postoperative recurrence risk model established by the sinus CT score combined with serum allergen sIgE can effectively predict the incidence of postoperative recurrence in patients. The XGboost model is better than the multivariate Cox analysis model in predicting postoperative recurrence in patients with eosinophilic CRSwNP. It can be used to predict postoperative recurrence.
-
Key words:
- nasal polyps /
- sinusitis /
- recurrence /
- XGboost model
-
-
表 1 两组EOS型CRSwNP患者临床资料比较
项目 复发组(n=52) 未复发组(n=131) P值 年龄/岁 46.62±8.02 49.45±9.68 0.817 性别(男/女) 36/16 91/40 0.975 平均病程/年 3.12±1.74 2.38±1.99 0.064 吸烟史/例 38 79 0.105 饮酒史/例 27 52 0.132 合并哮喘/例 12 20 0.053 合并变应性鼻炎/例 27 48 0.058 前期鼻窦炎手术史/例 39 75 0.025 术后坚持综合治疗/例 13 98 ＜0.001 鼻塞评分 7.58±0.60 6.72±0.84 0.041 头/面痛评分 2.20±0.54 1.63±0.23 0.069 嗅觉损失评分 5.36±1.71 1.86±0.74 0.034 组织EOS比例/% 14.23±5.86 11.33±6.05 0.011 组织NEU比例/% 0.92±1.10 6.05±8.86 0.012 组织淋巴细胞比例/% 14.60±9.85 47.23±21.46 0.004 组织浆细胞比例/% 6.87±4.54 31.39±16.21 0.006 外周血EOS比例/% 8.62±5.23 3.59±2.78 0.029 外周血NEU比例/% 45.17±10.99 56.01±19.42 0.016 sIgE/(U·mL-1) 3.02±1.87 0.71±0.60 0.003 鼻窦CT总评分 16.87±6.24 10.85±2.26 0.007 
下载: 导出CSV
表 2 EOS型CRSwNP患者复发的单因素及多因素Cox分析
变量 单因素分析 多因素分析 95%CI P值 95%CI P值 年龄 0.793~1.290 0.817 - - 性别 0.654~1.324 0.975 - - 平均病程 0.586~1.719 0.064 - - 吸烟史 0.813~1.532 0.105 - - 饮酒史 0.762~1.634 0.132 - - 合并哮喘 0.845~1.770 0.053 - - 合并变应性鼻炎 0.763~1.692 0.058 - - 前期鼻窦炎手术史 1.034~4.301 0.025 0.637~1.895 0.524 术后未坚持综合治疗 1.589~4.922 0.000 1.963~2.401 0.002 鼻塞评分 1.125~2.854 0.041 0.564~1.573 0.862 头/面痛评分 0.934~1.679 0.069 - - 嗅觉损失评分 1.073~2.581 0.034 0.685~1.741 0.065 组织EOS比例 1.698~3.884 0.000 1.872~4.653 0.001 组织NEU比例 1.536~3.007 0.012 1.255~2.036 0.023 组织淋巴细胞比例 1.602-4.158 0.004 1.659~3.757 0.009 组织浆细胞比例 1.351~3.216 0.006 1.682~3.092 0.011 外周血EOS比例 1.284~2.495 0.029 0.763~1.707 0.098 外周血NEU比例 1.063~1.982 0.016 1.052~1.941 0.042 sIgE 1.912~4.846 0.003 2.326~5.137 < 0.001 鼻窦CT总评分 1.832~4.610 0.007 1.835~4.354 < 0.001
下载: 导出CSV
表 3 XGboost模型与多因素Cox分析模型预测复发的效能比较
模型 AUC 95%CI 约登指数 灵敏度/% 特异度/% XGboost模型 0.818 0.756~0.881 0.432 82.93 71.62 多因素Cox分析模型 0.789 0.723~0.856 0.401 77.12 62.49
下载: 导出CSV
-
[1] Matsuno O, Minamoto S. Rapid effect of benralizumab for severe asthma with chronic rhinosinusitis with nasal polyps[J]. Pulm Pharmacol Ther, 2020, 64: 101965. doi: 10.1016/j.pupt.2020.101965
[2] Schatz M, Sicherer SH, Khan D, et al. The Journal of Allergy and Clinical Immunology: In Practice 2018 Highlights[J]. J Allergy Clin Immunol Pract, 2019, 7(2): 393-411. doi: 10.1016/j.jaip.2018.12.007
[3] 中华耳鼻咽喉头颈外科杂志编辑委员会鼻科组, 中华医学会耳鼻咽喉头颈外科学分会鼻科学组. 中国慢性鼻窦炎诊断和治疗指南(2018)[J]. 中华耳鼻咽喉头颈外科杂志, 2019, 54(2): 81-100. doi: 10.3760/cma.j.issn.1673-0860.2019.02.001
[4] Shah SA, Ishinaga H, Takeuchi K. Pathogenesis of eosinophilic chronic rhinosinusitis[J]. J Inflamm(Lond), 2016, 13: 11. doi: 10.1186/s12950-016-0121-8
[5] Akyigit A, Keles E, Etem EO, et al. Genetic polymorphism of antioxidant enzymes in eosinophilic and non-eosinophilic nasal polyposis[J]. Eur Arch Otorhinolaryngol, 2017, 274(1): 267-273. doi: 10.1007/s00405-016-4259-z
[6] 杨钦泰, 孙悦奇, 吴庆武, 等. 2020年欧洲鼻窦炎和鼻息肉意见书解读[J]. 中华耳鼻咽喉头颈外科杂志, 2020, 55(3): 304-308. doi: 10.3760/cma.j.issn.1673-0860.2020.03.024
[7] 许源, 马健勇, 葛艳秋, 等. XGboost模型对缺血性脑卒中出院后90 d内复发再入院风险的预测效果分析[J]. 中华神经医学杂志, 2018, 17(8): 813-818. doi: 10.3760/cma.j.issn.1671-8925.2018.08.010
[8] 李伟, 彭启旺, 林曜, 等. 结合术前纤维蛋白原及血液炎性指标构建中高危胃肠间质瘤复发预测模型[J]. 中华胃肠外科杂志, 2020, 23(9): 896-903. doi: 10.3760/cma.j.cn.441530-20200613-00355
[9] 张丽川, 胡春华, 韩星雨, 等. 鼻息肉患者组织嗜酸粒细胞计数与改良鼻窦CT嗅区评分及嗅觉功能的关系[J]. 临床耳鼻咽喉头颈外科杂志, 2019, 33(12): 1142-1147. https://www.cnki.com.cn/Article/CJFDTOTAL-LCEH201912008.htm
[10] DeCondeAS, MaceJC, LevyJM, et al. Prevalence of polyp recurrence after endoscopic sinus surgery for chronic rhinosinusitis with nasal polyposis[J]. Laryngoscope, 2017, 127(3): 550-555. doi: 10.1002/lary.26391
[11] 杨焕焕, 方红, 尤权杰, 等. 组织嗜酸粒细胞及鼻窦CT对慢性鼻-鼻窦炎伴鼻息肉术后复发的预测研究[J]. 中华耳鼻咽喉头颈外科杂志, 2018, 53(11): 842-846. doi: 10.3760/cma.j.issn.1673-0860.2018.11.009
[12] 徐征, 邹玉华, 张舒. 嗜酸粒细胞对慢性鼻-鼻窦炎伴鼻息肉术后复发的预测价值研究[J]. 全科医学临床与教育, 2018, 16(5): 524-527. https://www.cnki.com.cn/Article/CJFDTOTAL-QKYL201805013.htm
[13] 张洪侠, 郭贺, 王金霞, 等. 基于XGBoost算法的2型糖尿病精准预测模型研究[J]. 中国实验诊断学, 2018, 22(3): 408-412. doi: 10.3969/j.issn.1007-4287.2018.03.008
[14] Maali Y, Perez-Concha O, Coiera E, et al. Predicting 7-day, 30-day and 60-day all-cause unplanned readmission: a case study of a Sydney hospital[J]. BMC Med Inform Decis Mak, 2018, 18(1): 1. doi: 10.1186/s12911-017-0580-8
-
图(2)
表(3)
计量
- 文章访问数: 812
- PDF下载数: 400
- 施引文献: 0