Prognostic modeling and validation of middle ear malignancy outcomes on the basis of SEER population data
Yi Ren1, Li-Ping Peng1, Tao Hou1
1Otorhinolaryngology Head and Neck Surgery Department, The Second Affiliated Hospital of Guangxi Medical University, Nanning, China.
Background:
Malignant neoplasms of the middle ear represent a rare yet aggressive entity, with limited evidence regarding prognostic determinants and therapeutic outcomes. This study aimed to evaluate the association between radiotherapy (RT) and survival outcomes, and to develop a prognostic nomogram for middle ear malignancies using the Surveillance, Epidemiology, and End Results (SEER) database.
Methods:
A retrospective cohort analysis was performed on 206 patients diagnosed with middle ear malignancies from the SEER database [1975-2021]. Demographic and clinical variables, including age, sex, race, tumor stage, surgical intervention, RT, and chemotherapy, were analyzed. Survival outcomes were assessed using the Kaplan-Meier (K-M) method and multivariate Cox proportional hazards (PH) regression models. A nomogram predicting 1-, 3-, and 5-year overall survival (OS) was constructed and internally validated.
Results:
RT was administered to 56% of patients (n=112) and was associated with a 72% increased risk of mortality compared to patients not receiving RT [hazard ratio (HR) =1.72, 95% confidence interval (CI): 1.16-2.55; P=0.007]. Patients aged ≥65 years exhibited a 2.6-fold higher mortality risk (HR =2.61; 95% CI: 1.75-3.88; P<0.001). Compared with localized disease, unknown tumor stage was linked to a 75% lower mortality risk (HR =0.25; 95% CI: 0.09-0.71; P=0.009), likely reflecting data limitations rather than a true protective effect. Advanced tumor stage significantly impacted OS (P=0.007). The nomogram incorporating age, race, laterality, tumor stage, surgical intervention, chemotherapy, and RT demonstrated good discriminative ability, with area under the curve (AUC) values of 0.707, 0.740, and 0.752 for 1-, 3-, and 5-year OS prediction, respectively.
Conclusions:
An association between RT and increased mortality risk was observed in patients with middle ear malignancies; however, this finding should be interpreted cautiously due to potential confounding and treatment selection bias inherent in retrospective database analyses. Age and tumor stage were identified as important prognostic factors. The developed nomogram offers a practical tool for individualized survival prediction to aid clinical decision-making. Prospective studies are needed to validate these findings and clarify the role of RT in this patient population.
