What question did this study set out to answer?

March 6, 2026Scientific Reports0 citationsOpen Access

Predicting the risk for distant metastasis in hypopharyngeal squamous cell carcinoma and assessing the survival benefit of induction therapy

YZYu ZhangDalian Medical University YWYì WángBGI Group (China)ZWZhao WkDalian Medical University

Key Points

The goal is to predict distant metastasis risk in hypopharyngeal squamous cell carcinoma and assess the survival benefit of induction therapy.
Data were obtained from the SEER database and divided into training and validation sets.
A nomogram was developed to predict the risk of distant metastasis.
Multivariate logistic regression identified independent risk factors for distant metastasis.
Patients were categorized into high- and low-risk groups based on the nomogram's predictions.
The nomogram showed good discriminative ability with AUCs ranging from 0.702 to 0.863 in different validation sets.
In the high-risk group, induction therapy significantly improved overall survival (HR 0.364) and progression-free survival (HR 0.420).
No significant survival benefit was observed from induction therapy in the low-risk group.

Abstract

Hypopharyngeal squamous cell carcinoma (HPSCC) has a poor prognosis due, in large part, to distant metastasis (DM). Although induction therapy (IT) can reduce DM rates, its translation to an overall survival (OS) benefit remains unclear, highlighting the need for tools to identify patients who would benefit most from IT. This study developed a nomogram to predict the risk for DM in patients with HPSCC, and assessed the survival benefit of IT across risk groups. Data from patients obtained from the Surveillance, Epidemiology, and End Results (i.e., “SEER”) database (2004–2015) were randomly assigned to 1 of 2 groups at a ratio of 7:3: training; and internal validation. The external validation set comprised patients from 2 medical centers in China. Risk factors were identified using multivariate logistic regression analysis. Male sex, T classification ≥ 2, N classification ≥ 1, and poorer histological grade were independent risk factors for DM. The nomogram demonstrated good discriminative ability, with areas under the receiver operating characteristic curve of 0.702 (95% confidence interval CI: 0.669–0.735) in the training set, 0.704 (95% CI: 0.648–0.759) in the internal validation set, and 0.863 (95% CI: 0.804–0.923) in the external validation set. Based on the optimal cut-off value, patients were stratified into high- and low-risk groups. In the high-risk group, patients who received IT exhibited significantly improved OS (hazard ratio HR 0.364; 95% CI: 0.165–0.805; P = 0.040) and progression-free survival (PFS; HR 0.420; 95% CI: 0.190–0.928; P = 0.042) compared with those who did not receive IT. There was no significant survival benefit from IT in the low-risk group (OS: HR 0.881; 95% CI: 0.414–1.875, P = 0.095; PFS: HR 0.544 95% CI: 0.182–1.626, P = 0.250). This study constructed and preliminarily validated a nomogram for predicting DM in patients with HPSCC, and may serve as an exploratory tool for screening high-risk patients who are likely to benefit from IT, thereby providing a reference for the design of future prospective intervention trials.

KI fragen

Bookmark

View Full Paper