Risk factors for cervical lymph node metastasis in endoscopically resected superficial hypopharyngeal cancers

2019 
Abstract Objective Hypopharyngeal cancer is a head and neck cancer with a poor prognosis, and most cases show metastases on diagnosis. Cervical lymph node (LN) metastasis is a poor prognostic factor in hypopharyngeal cancer patients. The identification of risk factors for LN metastasis can help guide surgical treatment strategies for these patients. Methods This retrospective study included 93 superficial hypopharyngeal cancer patients with 109 histopathologically examined lesions treated by endoscopic resection between January 2007 and December 2017. Tumor thickness quantification, quantification of budding nests, immunostaining and other histopathological analyses in paraffin-embedded, formalin-fixed tissue sections (3-μm) of surgical specimens were performed by a certified pathologist. Results Cervical LN metastasis was positive in 18 out of 93 cases (19.3%) and 18 out of 109 lesions (16.5%). No differences were detected in patient characteristics between LN-positive and LN-negative cases, except for tumor thickness, which was significantly larger in LN-positive cases (3119.4 ± 602.2 μm vs. 1015.5 ± 129.6 μm, respectively; p   0.0001). Univariate analysis showed that tumor thickness ≥1000 μm (odds ratio: 5.559, p  = 0.003), lesions with high budding grade (odds ratio: 5.188, p  = 0.01) and vascular invasion (odds ratio: 12.710, p  = 0.007) were significantly associated with cervical LN metastasis. Multivariate analysis revealed tumor thickness ≥ 1000 μm as the most significant risk factor for cervical LN metastasis in superficial hypopharyngeal cancer (odds ratio: 3.639, p  = 0.04). Conclusions We demonstrate for the first time that high budding grade may serve as powerful predictors of LN metastasis and tumor thickness ≥1000 μm is a significant risk factor for LN metastasis of superficial hypopharyngeal cancer. These results should be further examined in future larger scale studies.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    30
    References
    3
    Citations
    NaN
    KQI
    []