Zhang Ming-Yi, Tang Lian-Sha, Qin Zhao-Juan, Hao Ya-Ting, Cheng Ke, Zheng Ai
Department of Obstetrics and Gynecology, West China Second University Hospital, Sichuan University, Chengdu, Sichuan, China.
Department of Biotherapy, Cancer Center, West China Hospital, Sichuan University, Chengdu, Sichuan, China.
Front Med (Lausanne). 2022 Oct 18;9:988830. doi: 10.3389/fmed.2022.988830. eCollection 2022.
Pulmonary carcinosarcoma (PCS) is a rare but aggressive malignant disease in the lung. It is characterized by coexisting histologic elements of carcinomatous and sarcomatous components. This study aimed to comprehensively understand the clinical features of PCS and develop a nomogram for prognostic prediction of PCS patients.
Data were collected from the Surveillance Epidemiology and End Results (SEER) database between 1975 and 2018. Propensity-score matching (PSM) was used to match the demographic characteristic of the PCS vs. pulmonary sarcoma (PS). Cancer-specific survival (CSS) and overall survival (OS) were the main endpoints of the survival of patients and were evaluated using the Kaplan Meier curves and Cox proportional hazards regression. We further randomly split enrolled PCS patients from SEER into the training and validation sets. All independent predictors for OS of the training set were integrated to create a predictive nomogram. The performance of the nomogram was determined by discrimination, calibration ability, clinical usefulness, and risk stratification ability both in the training and validation cohorts. In addition, the clinical data of PCS patients from the West China Hospital were also retrospectively analyzed by this model.
A total of 428 PCS patients and 249 PS patients were enrolled from SEER. Compared to pure PS, PCS was associated with significantly better survival in the unmatched cohorts, whereas non-significantly better survival after PSM. In subgroup analysis, PCS showed significantly worse survival than pure PS in subgroups among the race, marital status, and radiation treatment. A nomogram was constructed for PCS patients' survival prediction by combining the independent risk factors, including gender, stage, surgery, radiation, and chemotherapy. The nomogram showed good discrimination, calibration, and predictive power in the training and validation sets. Risk stratification analysis indicated that the nomogram scores efficiently divided PCS patients into low and high-risk groups.
PCS is a rare malignant disease of the lung with distinct clinical features. It had a comparable survival compared with pure PS in the matched cohorts. In addition, a nomogram was developed and validated for predicting the OS in PCS patients.
肺肉瘤样癌（PCS）是一种罕见但侵袭性强的肺部恶性疾病。其特征是同时存在癌性和肉瘤性组织学成分。本研究旨在全面了解PCS的临床特征，并建立一个用于预测PCS患者预后的列线图。
收集1975年至2018年监测、流行病学和最终结果（SEER）数据库中的数据。采用倾向评分匹配（PSM）来匹配PCS与肺肉瘤（PS）的人口统计学特征。癌症特异性生存（CSS）和总生存（OS）是患者生存的主要终点，并使用Kaplan-Meier曲线和Cox比例风险回归进行评估。我们进一步将从SEER纳入的PCS患者随机分为训练集和验证集。整合训练集所有OS的独立预测因素以创建一个预测列线图。通过训练队列和验证队列中的区分度、校准能力、临床实用性和风险分层能力来确定列线图的性能。此外，还使用该模型对来自华西医院的PCS患者的临床数据进行了回顾性分析。
从SEER纳入了428例PCS患者和249例PS患者。与单纯PS相比，在未匹配队列中，PCS的生存情况明显更好，而在PSM后生存情况改善不显著。在亚组分析中，在种族、婚姻状况和放射治疗亚组中，PCS的生存情况比单纯PS显著更差。通过结合包括性别、分期、手术、放疗和化疗在内的独立危险因素，构建了用于PCS患者生存预测的列线图。该列线图在训练集和验证集中显示出良好的区分度、校准度和预测能力。风险分层分析表明，列线图评分有效地将PCS患者分为低风险和高风险组。
PCS是一种具有独特临床特征的罕见肺部恶性疾病。在匹配队列中，其生存情况与单纯PS相当。此外，开发并验证了一个用于预测PCS患者OS的列线图。
