Prognostic Model and Nomogram for Early-Onset Adenoid Cystic Carcinoma of Head and Neck: A Retrospective SEER-Based Analysis

BACKGROUND: Adenoid cystic carcinoma of head and neck (ACCHN) is an uncommon head and neck cancers, whose predilection age is 40 to 60. Some studies have revealed that early-onset cancers, such as colorectal cancers and esophageal adenocarcinoma, might present some unique clinicopathological features and have different prognosis with late-onset ones. However, little is known about the early-onset ACCHN. This study aimed to develop a prognostic nomogram for overall survival (OS) of patients younger than 40 with ACCHN. METHODS: Cases with ACCHN from 1975 to 2016 were retrieved from SEER-18 program. Demographic, clinical, and survival outcomes data of patients were identified for further analysis. The caret package was used to randomly divide early-onset patients into a training cohort and a validation cohort. A prognostic nomogram was constructed based on the univariate and multivariate Cox analysis. The discriminative ability and calibration power of the nomogram was evaluated by the concordance index (C-index), calibration curve, and receiver operating characteristic (ROC) curve. RESULTS: A total of 5858 cases with ACCHN were selectively retrieved from SEER program in this study. The number of patients younger than 40, which was defined as early-onset ACCHN in this study, was 825. Based on the outcomes of multivariate analysis, tumor size, chemotherapy, surgery, and stage were selected for the construction of nomogram to predict 10-year OS. The C-index was 0.792 (95%CI 0.760-0.823) and 0.776 (95%CI 0.720-0.832) in the training and validation set, respectively. The area under the ROC curve values were 0.875 (95%CI 0.810-0.940) and 0.833(95%CI 0.754-0.912). The calibration plot indicated that this nomogram had proper calibration in both the training and validation cohorts. CONCLUSION: A novel prognostic nomogram for early-onset ACCHN was constructed and validated in this study. This nomogram could be applied for assisting clinicians to more accurately assess the prognosis of young patients, which might facilitate clinical decision-making and subsequent follow-up.