作者：陈树清，屈开新，张宗仁，王守玉，常喜豹，刘长志，张宗峰

单位：

236300 安徽省阜阳市，阜南县人民医院医学影像科

Units：

Medical Imaging Department, Funan County People's Hospital, Funan 236300, China

关键词：

孤立性肺结节；CT 引导；肺穿刺活检术；咯血；影响因素分析；列线图

Keywords：

Solitary pulmonary nodule; CT-guided; Lung puncture biopsy; Hemoptysis; Root cause analysis; Nomogram

CLC：

R 521.6

DOI：

10.12114/j.issn.1008-5971.2021.00.138

Funds：

摘要：

背景　孤立性肺结节（SPN）是肺癌的早期表现，目前肺穿刺活检术对肺癌的诊断准确率高达95%以上， 但其对早期SPN 良、恶性鉴别诊断的相关报道较少。另外，CT 引导下肺穿刺活检术（以下简称CT 肺穿刺）是一种有 创的检查手段，易引发咯血、气胸、血胸等并发症，其中关于咯血的相关报道较少。目的　探讨CT肺穿刺对早期SPN良、 恶性患者的诊断价值，分析术后咯血的危险因素，并构建风险预测列线图模型，以期为临床诊疗提供参考。方法　选 取2017—2019 年在阜南县人民医院行CT 肺穿刺的328 例早期SPN 患者为研究对象，根据病灶直径分为A 组（10~ 20 mm）182 例和B 组（20~30 mm）146 例。比较A、B 组患者1 次穿刺成功率、病理诊断、病理类型及并发症发生 情况。根据患者咯血发生情况将其分为咯血组和未咯血组，比较两组患者的临床资料。采用多因素Logistic 回归分析 探讨早期SPN 患者CT 肺穿刺后并发咯血的影响因素；采用R 3.6.1 软件包结合rms 程序包建立早期SPN 患者CT 肺穿 刺后并发咯血的风险预测列线图模型，计算一致性指数（CI）；绘制受试者工作特征曲线（ROC 曲线）以评估列线图 模型对早期SPN 患者CT 肺穿刺后并发咯血的预测价值。结果　328 例患者均穿刺成功。A、B 组患者1 次穿刺成功率 比较，差异无统计学意义（P ＞ 0.05）；B 组患者病理诊断恶性率高于A 组（P ＜ 0.05）。246 例恶性早期SPN 患者 中，以腺癌检出率最高，为33.7%，其次为鳞癌（24.8%）、小细胞癌（15.4%）、转移瘤（6.5%）；82 例良性早期 SPN 中，以肺炎检出率最高，为52.4%，其次为肺结核（40.2%）。B 组患者咯血、气胸、肺内出血发生率低于A 组 （P ＜ 0.05）。多因素Logistic 回归分析结果显示，年龄≥ 60 岁〔OR=3.346，95%CI（1.469，7.623）〕、病灶直径 ≥ 20 mm〔OR=4.764，95%CI（1.750，12.973）〕、穿刺深度≥ 20 mm〔OR=5.554，95%CI（2.403，12.833）〕、穿 刺次数＞ 1 次〔OR=3.443，95%CI（1.489，7.963）〕、病灶- 胸膜距离≤ 1 cm〔OR=5.798，95%CI（1.978，8.026）〕 是早期SPN 患者CT 肺穿刺后并发咯血的独立危险因素（P ＜ 0.05）。将多因素Logistic 回归分析中有统计学差异的 指标纳入列线图模型，Calibration 验证及Bootstrap 内部验证结果显示，CI 为0.841〔95%CI（0.780，0.902）〕，其中 校正曲线与理想曲线贴合良好。ROC 曲线分析结果显示，列线图模型预测早期SPN 患者CT 肺穿刺后并发咯血的ROC 曲线下面积为0.841〔95%CI（0.780，0.902）〕，灵敏度为95%，特异度为90%。结论　CT 肺穿刺对早期SPN 具有 较高的穿刺成功率，但直径为＞ 20~30 mm 的早期SPN 患者肺结节恶性率较高。年龄≥ 60 岁、病灶直径≥ 20 mm、 穿刺深度≥ 20 mm、穿刺次数＞ 1 次、病灶- 胸膜距离≤ 1 cm 是早期SPN 患者CT 肺穿刺后并发咯血的独立危险因素， 且根据上述危险因素建立的风险预测列线图模型具有良好的精准度、区分度及预测能力。

Abstract：

　Background　Solitary pulmonary nodule (SPN) is early manifestations of lung cancer. At present, the diagnostic accuracy of lung cancer with lung puncture biopsy is more than 95%, but there are few reports on the differentialdiagnosis of benign and malignant for early SPN. CT-guided lung biopsy (hereinafter referred to as CT lung puncture) is an invasive examination method, which is easy to cause complications such as hemoptysis, pneumothorax and hemothorax, among which there are few reports about hemoptysis. Objective　To explore the diagnostic value of CT lung puncture for early SPN patients, and analyse the risk factors of hemoptysis in patients with early SPN after CT lung puncture, and establish the Nomogram model of postoperative hemoptysis risk prediction, in order to provide reference for clinical diagnosis and treatment. Methods　 A total of 328 cases of SPN patients were selected from 2017 to 2019 in Funan County People's Hospital, and they were divided into group A (n=182, 10-20 mm) and group B (n=146, > 20-30 mm) according to lesion diameter. Success rate of one puncture, pathological diagnosis, pathological types and incidence of complications were compared between group A and group B. Clinical data were compared between the hemoptysis group and non-hemoptysis group, which were divided according to incidence of hemoptysis. Multivariate Logistic regression analysis was used to discuss the influencing factors of hemoptysis in patients with early SPN after CT lung puncture; the Nomogram model of hemoptysis risk prediction in patients with early SPN after CT lung puncture was constructed through R 3.6.1 software package combined with rms package, and concordance index (CI) was calculated; receiver operating characteristic curve (ROC curve) was used to evaluate the predictive value of hemoptysis in patients with early SPN after CT lung puncture. Results　All 328 patients were successfully punctured, and there was no significant difference of success rate of one puncture between group A and group B (P > 0.05) ; malignant diagnosis rate in the group B was higher than that in the group A (P < 0.05) . Among 246 cases of patients with malignant SNP, the highest detection rate was adenocarcinoma (33.7%) , followed by squamous cell carcinoma (24.8%) , small cell carcinoma (15.4%) , metastatic tumor (6.5%) ; among 82 cases of patients with benign early SPN, the highest detection rate was pneumonia (52.4%) , followed by tuberculosis (40.2%) . Incidence of hemoptysis, pneumothorax and intrapulmonary hemorrhage in the group B were lower than those in the group A (P < 0.05) . Multivariate Logistic regression analysis showed that, age ≥ 60 years old [OR=3.346, 95%CI (1.469, 7.623) ] , lesion diameter ≥ 20 mm [OR=4.764, 95%CI (1.750, 12.973) ] , puncture puncture ≥ 20 mm [OR=5.554, 95%CI (2.403, 12.833) ] , puncture times > 1 [OR=3.443, 95%CI (1.489, 7.963) ] , lesion-pleural distance ≤ 1 cm [OR=5.798, 95%CI (1.978, 8.026) ] were independent risk factors of hemoptysis in patients with early SPN after CT lung puncture (P < 0.05) . The indicators with statistical differences in multivariate Logistic regression analysis were included in the Nomogram model, and the results of Calibration verification and Bootstrap internal verification showed that, CI was 0.841 [95%CI (0.780, 0.902) ] , the calibration curve fit well with the ideal curve. ROC curve showed that, AUC of Nomogram model in predicting hemoptysis risk in patients with early SPN after CT lung puncture was 0.841 [95%CI (0.780, 0.902) ] , sensitivity was 95%, specificity was 90%. Conclusion　CT lung puncture has a high puncture success rate for early SPN, but the malignant rate of pulmonary nodules is higher in early SPN patients with diameter > 20-30 mm. Age ≥ 60 years old, lesion diameter ≥ 20 mm, puncture puncture ≥ 20 mm, puncture times > 1, lesion-pleural distance ≤ 1 cm are independent risk factors of hemoptysis in patients with early SPN after CT lung puncture, and the risk prediction Nomogram model established based on the above risk factors has good accuracy, differentiation and predictive ability.

ReferenceList：