Department of Radiology, Tianjin Medical Imaging Institute, Tianjin First Central Hospital, School of Medicine, Nankai University, No. 24 Fukang Road, Nankai District, Tianjin, 300192, China.
Department of Otorhinolaryngology Head and Neck Surgery, Tianjin First Central Hospital, School of Medicine, Nankai University, No. 24 Fukang Road, Nankai District, Tianjin, 300192, China.
Eur Arch Otorhinolaryngol. 2023 Jun;280(6):2885-2896. doi: 10.1007/s00405-023-07851-y. Epub 2023 Jan 25.
To developed a model to screen distant metastatic laryngeal carcinoma (DMLC) patients who would benefit from the primary tumor resection.
The propensity score matching (PSM) was utilized to avoid disproportionate distributions of the confounding factors. We hypothesized that patients who underwent surgery would benefit from surgery by having a longer median cancer-specific survival (CSS) than patients without surgery. Multivariable Cox model was used to explore the independent factors of CSS and overall survival (OS) among PSM population. We used these factors to construct a nomogram to identify surgery benefit patients. The predictive performance and clinical practicability of the nomogram were determined by area under the curve (AUC), calibration curve, and decision curve.
The CSS and OS for patients who received primary tumor resection were significantly longer than those without resection (median CSS: 19 months vs. 10 months, P = 0.009; median OS: 21 months vs. 10 months, P = 0.001). The nomogram displayed a good degree of discrimination and calibration. The mean AUC of the nomogram was 0.70 (95% confidence interval [CI] 0.66-0.76) through threefold cross-validation.
A predictive model was established and might be used to screen the optimal candidates for primary tumor surgery in DMLC patients.
开发一种模型，以筛选可能从原发肿瘤切除中获益的远处转移性喉癌（DMLC）患者。
采用倾向评分匹配（PSM）避免混杂因素分布不均。我们假设手术组患者的癌症特异性生存（CSS）中位数比未手术组患者更长，从而从手术中获益。多变量 Cox 模型用于探索 PSM 人群中 CSS 和总生存（OS）的独立因素。我们使用这些因素构建一个列线图，以确定手术获益的患者。通过曲线下面积（AUC）、校准曲线和决策曲线来确定列线图的预测性能和临床实用性。
接受原发肿瘤切除术的患者的 CSS 和 OS 明显长于未接受切除术的患者（中位 CSS：19 个月 vs. 10 个月，P=0.009；中位 OS：21 个月 vs. 10 个月，P=0.001）。列线图显示出良好的区分度和校准度。通过三折交叉验证，列线图的平均 AUC 为 0.70（95%置信区间 [CI] 0.66-0.76）。
建立了一种预测模型，可用于筛选 DMLC 患者中接受原发肿瘤手术的最佳候选者。
