Introduction
External auditory canal (EAC) cancer is a rare head and neck malignancy with an estimated annual incidence of approximately 1-6 cases per million people [1,2]. The most common pathology of EAC cancer is squamous cell carcinoma (SCCa), followed by adenoid cystic carcinoma (ACCa) [3-5]. Although no single study has directly compared the survival rates of SCCa and ACCa, the prognosis has been reported to differ substantially between these two pathologic types, with SCCa demonstrating poorer survival outcomes than ACCa. The 5-year survival of patients with ACCa ranges from 59% to 92% [6-9]. On the other hand, SCCa showed 61.8% of 5-year survival in a meta-analysis study [10].
The largest cohort study reported by Ogawa, et al. [11] in 2007 demonstrated a 5-year overall survival (OS) rate of 55% among 87 patients with SCCa. Given the rarity of EAC cancer, data accumulation is essential for future systematic reviews and development of clinical guidelines. Our cohort comprised 137 Korean patients with EAC cancer who visited a single institution, representing a relatively large cohort of patients with this rare disease. Therefore, we were able to directly compare the clinical characteristics of SCCa and ACCa, including OS.
This study aimed to share our institutional experience with the clinical outcomes of EAC cancer. In particular, reports of EAC cancer in the Korean population are limited [12-15]. In addition, we aimed to clarify the differences between SCCa and ACCa.
Subjects and Methods
Patient enrollment and data curation
The medical records of patients diagnosed with EAC cancer between January 1, 2015, and January 1, 2026, at a single tertiary referral hospital were retrospectively reviewed. In total, 203 patients were initially screened, of whom 151 met the inclusion criteria for pathologically confirmed malignancies. Fourteen patients were excluded based on the following criteria: 1) tumors originating from nearby structures (auricle, n=5; middle ear, n=1); 2) metastasis from other organs (nasopharyngeal cancer, n=1; lung cancer, n=1; breast cancer, n=1; brain cancer, n=1); and 3) direct invasion from adjacent organs (parotid gland cancer, n=4). Finally, 137 patients with primary EAC cancer patients were enrolled in this study.
Demographic data were collected from medical records. The date of initial diagnosis was defined as the date of the first biopsy-confirmed diagnosis before treatment at any institution, regardless of the referral status or recurrence. Tumor staging was performed according to the Pittsburgh TNM staging system for EAC cancer [16]; however, ACCa was staged according to the updated staging system specific to ACCa [17]. Clinical staging was determined based on pretreatment imaging studies performed immediately after initial diagnosis, including temporal bone computed tomography, magnetic resonance imaging, and positron emission tomography. OS was determined by reviewing medical records and supplemented by National Health Insurance database queries for patients lost to follow-up.
This study was approved by the Institutional Review Board of the Yonsei University College of Medicine (Project No. 2025-1280-001). The board waived the requirement for informed consent. This study was conducted in accordance with the principles of the Declaration of Helsinki.
Statistical analysis
Statistical analyses were performed using SPSS version 28.0 (IBM Corp.). Descriptive statistics were used to summarize the patient demographics and clinical characteristics. Continuous variables, such as age, were expressed as mean± standard deviation and compared using an independent t-test. Categorical variables including sex, tumor side, and clinical stage were compared between the SCCa and ACCa groups using Fisher’s exact test. Survival curves were constructed using the Kaplan-Meier method and compared between groups using the log-rank (Mantel-Cox) test. Survival rates were calculated at 1, 3, and 5 years of age. Statistical significance was set at p<0.05. All graphical representations and survival plots were generated using GraphPad Prism 8.0 (GraphPad Software).
Results
The mean age of the EAC cancer cohort was 65.6±14.1 years, and 50.4% of patients were male (Table 1). The right side was more frequently affected (62.8%) than the left side (37.2%). The most common histologic subtype was SCCa (73.0%), followed by ACCa (20.4%), and basal cell carcinoma (2.9%). At the initial visit, Stage I disease was the most frequent (41.2%), followed by Stages IV (30.9%), II (16.2%), and III (11.8%).
Next, we compared the clinical characteristics of the patients with SCCa and ACCa (Table 2). The mean age was significantly younger in ACCa (55.5±13.4 years) than in SCCa (67.7±13.1 years, p<0.001). Other clinicopathological factors did not differ significantly between the two groups, although ACCa showed a female predominance (60.7%). Interestingly, both SCCa (60.0%) and ACCa (78.6%) showed a right-sided predominance. Clinical N1 disease was observed in 13.3% of patients with SCCa, whereas none of the patients with ACCa presented with nodal involvement. Distant metastasis was identified in one patient with ACCa (lung metastasis), but was not observed in patients with SCCa. Most patients in both the groups were treated surgically.
The survival status was successfully identified in 124 of 137 (90.5%) patients in the entire EAC cancer cohort, including 91 of 100 (91%) patients with SCCa and 25 of 28 (89.3%) patients with ACCa. In the entire EAC cancer cohort, the 1-year survival rate was 84.4%, and the 3- and 5-year survival rates were 78.9% (Fig. 1). Survival outcomes differed significantly between the two groups (p=0.011, Mantel-Cox test). None of the patients with ACCa died during the follow-up. In contrast, patients with SCCa showed a 1-year survival rate of 80.1% and 3- and 5-year survival rates of 72.4%. No morality event occured between 3-5 years. In patients with SCCa, the clinical stage at initial diagnosis was a significant prognostic factor for OS (p=0.003, Mantel-Cox test). The 5-year survival rates were 97.2%, 75.0 %, 61.7 %, and 44.1% for Stages I, II, III, and IV, respectively (Fig. 2).
Discussion
In the Korean cohort, SCCa was the dominant histologic subtype, accounting for 73.0% of all cases. This is in line with previous studies that included different ethnicities as well as the East Asian population [18-20]. Both SCCa and ACCa showed right-sided predominance, and the mean age was significantly lower in patients with ACCa than in those with SCCa. OS was significantly better in patients with ACCa than in those with SCCa, with 5-year survival rates of 100% and 72.4%, respectively. The clinical staging system accurately reflected the prognosis of SCCa. However, the prognostic value of ACCa staging could not be evaluated because there were no deaths in the ACCa group in our cohort.
The proportion of SCCa and ACCa has been reported in various studies, and Park, et al. [18] reported that 57.6% of 33 Korean patients with EAC cancer had SCCa, followed by 27.3% with ACCa. Moody, et al. [19] reported that 76.1% of 46 patients with EAC cancer in the United States had SCCa. Moore, et al. [20] reported that SCCa accounted for 57.1% and ACCa for 20.0% among 35 patients with EAC cancer. Our results show that the proportions of SCCa (73.0%) and ACCa (20.4%) were within the ranges reported in previous studies.
In this study, patients with ACCa (mean age, 55.5 years) were significantly younger than those with SCCa (mean age, 67.7 years), a finding that has not been conclusively demonstrated in previous studies. The mean age at SCCa diagnosis was 61.2 years in systematic reports including 282 patients with SCCa [10]. In contrast, several studies that included patients with ACCa reported a mean age of less than 60 years. Dong, et al. [9] reported a mean age of 42 years in 22 patients with ACCa; Jiang, et al. [21] reported a mean age of 51 years in 23 patients, and Wang, et al. [22] reported a median age of 50.5 years in a cohort of 68 ACCa patients. Considering these previous reports, the significant age difference between patients with ACCa and SCCa observed in this study appears plausible.
No sex predominance was observed; however, a clear predominance of right-sided involvement was observed. Interestingly, this phenomenon was previously reported by Tsunoda, et al. [23], who found that 76.5% of the 68 patients had right external auditory involvement. In our cohort, 62.8% of EAC cancers occurred on the right side, and ACCa (78.6%) showed a more pronounced right-sided predominance than SCCa (60.0%). Tsunoda, et al. [23] suggested that habitual earpicking in the Japanese population may contribute to the development of EAC cancer, as most individuals are righthanded. Similar situations in the Korean population may have affected our data. However, no definitive conclusion can be drawn regarding this laterality owing to the limited data available in the existing literature.
The OS rate in our cohort was notably higher than that reported in previous studies. All 25 patients with ACCa in our cohort showed a 100% 5-year OS rate. To date, the highest OS rate was reported by Wang, et al. [22], who reported a 5-year OS rate of 96.7% in a cohort of 68 patients with ACCa. A systematic review by Cazzador, et al. [10] reported 3-year and 5-year OS rates of 65.7% and 61.8%, respectively, in SCCa. Our cohort showed 3-year and 5-year OS rates of 72.4% each. A recent study by Kang, et al. [15], which included 56 Korean patients with SCCa, also reported a better prognosis than previous studies. In their report, the 5-year OS rates were 100% for T1 and T2 stages, 60.0% for T3, and 42.0% for T4. This implies that our data did not overestimate OS.
This study has several limitations. Our analysis was restricted to OS and did not include disease-specific survival or recurrence rates. This was because we used National Health Insurance data to compensate for missing medical records, which only indicated survival status (alive or deceased). However, this approach allowed us to include a substantial number of patients who were lost to follow-up at our institution. In addition, applying Ear-picking or mechanical irritation for right-sided predominace may be a possible explanation for SCCa, but it is less directly applicable to ACCa, which originates from ceruminous glands rather than squamous epithelium. So mechanism of right-sided predominance in ACCa remains unclear. Also, survival rate of ACCa has some limitation, too. ACCa usually known for its high rate of late recurrence and distant metastasis [24]. As we use 5-year OS, survival rate may be overestimated due to limited long-term follow-up and lack of locoreginal recurrance data. Survival comparison between SCCa and ACCa (p=0.011) may have some statistical limitation, because zero events occurring in the ACCa group.
In conclusion, SCCa was the dominant histologic subtype of EAC cancer in Korean patients, accounting for 73.0% of cases, followed by ACCa at 20.4%. Patients with ACCa were younger than those with SCCa, and both histologic types showed right-sided predominance. The 5-year OS rate was significantly higher in patients with ACCa than in those with SCCa (100% and 72.4%, respectively).