Objectives/hypothesis: Primary carcinoma of the middle ear is a rare clinical entity, best suited for evaluation using a population-based database. The objective of this study was to utilize the Surveillance, Epidemiology, and End Results (SEER) database to determine the incidence, treatment, and survival of middle ear carcinoma.
Study design: Analysis of national cancer database.
Method: Using SEER*Stat software, records for patients diagnosed with middle ear carcinoma between 1973 and 2004 were extracted from the SEER database. Five-year, observed survival was analyzed, with significant differences determined by the Wilcoxon statistic.
Results: The 5-year observed survival rate for the 215 patients in this study was 36.4%. Histologic subtypes included squamous cell carcinoma (62.8%), adenocarcinoma (18.2%), other carcinomas (13.0%), and noncarcinomas (6.0%), with 5-year survival rates of 23.9%, 65.0%, 60.0%, and 38.6%, respectively (P = .003). Of the 123 patients with known stage, 23.6% had local, 69.1% had regional, and 7.3% had distant disease, with their 5-year survival rates being 64.9%, 34.2%, and 0%, respectively (P < .001). Treatment included surgery (31.2%), radiation (16.3%), surgery and radiation (38.6%), or no treatment (8.4%) with 5-year survival of 69.2%, 14.6%, 26.4%, and 0%, respectively (P < .001).
Conclusions: Patients with primary middle ear carcinoma have a relatively poor prognosis. However, subsets of patients, such as those with adenocarcinomas and with localized tumors, demonstrated significantly better survival. Surgery alone had significantly better survival than the other treatment groups, presumably due to less advanced disease in this treatment group. These data are useful in counseling patients and understanding the natural history of middle ear carcinoma.