Department of General Surgery, Wayne State University School of Medicine, Detroit, MI, USA.
Department of Otolaryngology-Head and Neck Surgery, Wayne State University School of Medicine, Detroit, MI, USA.
Am J Otolaryngol. 2020 May-Jun;41(3):102419. doi: 10.1016/j.amjoto.2020.102419. Epub 2020 Feb 7.
The incidence of lymph node metastases (LNM) in squamous cell carcinoma of the lip is modest (8%), making it challenging to identify patients that may benefit from elective pathologic staging evaluation of the neck. We evaluated predictors of LNM in patients with lip cancer in order to potentially refine selection of patients for pathologic staging evaluation of the neck.
Retrospective cohort study.
Clinically N0 patients with squamous cell carcinoma of the lip that underwent definitive surgical resection and pathologic evaluation of lymph node metastases in the National Cancer Data Base from 2006 to 2013.
Multivariable binomial logistic-regression was used to assess the relationship between occult pathologic lymph node metastasis and potential preoperative predictors including; patient demographics, T-stage, location, and pathologic details.
Among 786 patients the overall rate of LNM was 12.1%. Patients were more likely to have LNM with T2 (odds ratio (OR) 2.05; (95% confidence interval (CI) 1.19-3.54) or T3-4 (OR 2.36; CI 1.32-4.22) moderately differentiated (OR 2.65; CI 1.30-5.38) or poorly differentiated (OR 4.37; CI 1.97-9.71), or involvement of the mucosal surface (OR 1.82; CI 1.09-3.03). We created a prediction model based on proportional odd ratios from multivariant binomial logistic-regression analysis from statistically significant factors; incorporating T2-4, moderate/poorly differentiated, or mucosal site.
Our prediction model found that patients with two or more risk factors were the best candidates for elective pathologic nodal evaluation.
唇鳞状细胞癌的淋巴结转移(LNM)发生率适中(8%),因此难以确定可能受益于颈部选择性病理分期评估的患者。我们评估了唇癌患者中 LNM 的预测因素,以便有可能完善颈部病理分期评估患者的选择。
回顾性队列研究。
2006 年至 2013 年期间在国家癌症数据库中接受确定性手术切除和淋巴结转移病理评估的临床 N0 唇鳞状细胞癌患者。
使用多变量二项逻辑回归评估隐匿性病理淋巴结转移与潜在术前预测因子之间的关系,包括患者人口统计学、T 期、位置和病理细节。
在 786 例患者中,LNM 的总体发生率为 12.1%。T2(优势比(OR)2.05;95%置信区间(CI)1.19-3.54)或 T3-4(OR 2.36;CI 1.32-4.22)、中度分化(OR 2.65;CI 1.30-5.38)或低分化(OR 4.37;CI 1.97-9.71)、或黏膜表面受累(OR 1.82;CI 1.09-3.03)的患者更有可能发生 LNM。我们根据多变量二项逻辑回归分析中统计学显著因素的比例优势比创建了一个预测模型;纳入 T2-4、中/低分化或黏膜部位。
我们的预测模型发现,有两个或更多危险因素的患者是选择性病理淋巴结评估的最佳候选者。