Introduction

External auditory canal carcinoma (EACC) is a rare malignancy arising from the epithelial lining of the external auditory canal (EAC), accounting for fewer than 0.2% of all head and neck cancers [1]. Despite its low incidence, EACC presents significant management challenges due to its aggressive nature and close proximity to critical structures, including the facial nerve, temporal bone, dura mater, and major blood vessels [2]. Clinical outcomes are closely associated with tumor stage at diagnosis, with early-stage disease often being curable with surgical resection, whereas advanced-stage disease is associated with poor prognosis and high morbidity [3]. Surgical resection remains the cornerstone of treatment, often combined with adjuvant radiotherapy (RT) or chemotherapy in advanced cases. The Modified Pittsburgh Staging System is widely used to guide treatment decisions [4].

Achieving complete tumor removal can be technically challenging, particularly with extensive bony invasion [3]. Positive surgical margins and perineural invasion have been linked to recurrence and poor survival [1,5–7], while facial nerve involvement, lymph node metastasis, and distant spread also worsen prognosis [2,4,8]. Adjuvant RT is commonly used in cases with high-risk features such as positive margins, perineural invasion, or regional metastasis [2,9,10]. However, its survival benefit remains uncertain [1]. Chemotherapy is generally reserved for unresectable or metastatic cases, and its optimal role in neoadjuvant or definitive settings remains unclear [1,11]. Despite these multimodal approaches, outcomes for advanced EACC remain poor, with 5-year survival rates ranging from 20% to 40% [3]. These findings underscore the need for reliable prognostic indicators and optimized treatment strategies to improve survival.

Squamous cell carcinoma (SCC) of the external auditory canal (EAC), the most common histologic subtype of EACC, is characterized by local invasiveness and a high propensity for nodal metastasis driven by epithelial–mesenchymal transition and overexpression of matrix metalloproteinases [12–14]. Consequently, SCC of the EAC is often diagnosed at an advanced stage (cT3–T4 in 57.1% of cases), and is associated with poor overall survival (OS) and disease-specific survival (DSS) [15], whereas basal cell carcinoma (BCC), demonstrates more indolent behavior [16]. Given the aggressive clinical course and anatomical complexity of EAC SCC, this study aimed to evaluate treatment outcomes and identify key prognostic factors postoperative survival in a cohort treated over two decades at a single tertiary referral center. By analyzing detailed anatomical invasion patterns and surgical strategies, we sought to refine prognostic assessment and inform future management of this rare but aggressive malignancy.

Materials and methods

This retrospective study included patients diagnosed with SCC of the EAC who underwent surgical treatment between 2000 and 2022 at a single tertiary referral center. A total of 56 patients met the inclusion criteria. Data were obtained from electronic medical records. The data were accessed for research purposes between 07/05/2024 and 01/07/2025. Institutional Review Board approval was obtained before study initiation (2024−0596), and the requirement for informed consent was waived due to the retrospective design.

Results

A total of 56 patients diagnosed with EACC were retrospectively analyzed. The cohort comprised 36 male and 20 female patients, with a mean age of 61.6 years. Table 1 summarizes the distribution of patients by T stage, sex, and mean age.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Distribution of patients by age, sex, and T stage.

https://doi.org/10.1371/journal.pone.0339860.t001

Surgical methods and survival outcomes

Surgical interventions included sleeve resection, LTBR, STBR, and extended LTBR. Sleeve resection was performed exclusively in four patients with T1 lesions. No T2 or more advanced lesions were treated with sleeve resection; all T2 cases underwent LTBR. LTBR was the most common procedure for early-stage disease, used in 14 T1 and 11 T2 lesions. For T1 and T2 lesions, no patients received adjuvant RT. All T1 and T2 tumors achieved negative surgical margins on final pathological assessment. During the entire follow-up period, none of the patients with T1 or T2 disease developed local or regional recurrence or distant metastasis. STBR was the predominant surgical approach for advanced-stage disease, and extended LTBR was performed in two T3 and three T4 cases. All surgical procedures were planned with the intent of en bloc tumor excision. Complete en bloc resection was achieved in all T1–T3 cases. In T4 tumors, initial en bloc removal was attempted; however, when the lesion involved critical adjacent structures including the dura mater, carotid artery abutment, jugular bulb abutment, and Eustachian tube, additional stepwise resection was performed following en bloc excision to remove residual involved tissues and maximize oncologic clearance. Among the 19 patients with T4 disease, negative margins on the initial en bloc resection specimen were achieved in 5 cases, whereas positive margins were reported in 14 cases. Positive margins were mainly observed when tumors abutted critical adjacent structures, particularly the carotid artery, jugular bulb, or tympanic cavity. In these situations, following en bloc resection, additional targeted removal of residual involved tissues was undertaken. Intraoperative frozen-section analysis was used to confirm negative margins, and resection was continued until tumor-free margins were achieved at all accessible sites. For areas adjacent to the carotid artery or jugular bulb, where histologic sampling was not feasible, macroscopic inspection was performed. No cases demonstrated true intraluminal invasion of the carotid artery or jugular bulb. No parotid gland involvement or cervical lymph node metastasis was identified in T1–T2 cases, whereas nodal metastasis in T3–T4 cases was limited to clinically or radiologically suspicious nodes addressed during neck dissection. Most patients with T3 and T4 disease received postoperative adjuvant RT or CRT according to institutional treatment principles. Because the extent and pattern of adjacent structure invasion (such as the TMJ, mastoid, dura, and carotid artery) differ substantially by T stage, detailed comparison of anatomical invasion sites was limited to T4 lesions. Table 2 summarizes the surgical methods and survival outcomes of this advanced-stage cohort according to anatomical involvement. Of the 21 patients initially classified as T4, two were excluded from detailed anatomical-risk analysis because they died from causes unrelated to EAC SCC (acute pneumonia at 29 months and an unidentified medical event at 41 months). Therefore, 19 patients were included in the T4-specific analysis presented in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Comparison of survival outcomes based on involvement areas and the type of surgical intervention.

https://doi.org/10.1371/journal.pone.0339860.t002

Survival analysis

Patients with T1 and T2 tumors demonstrated excellent outcomes, with an OS rate of 100.0%; Kaplan–Meier survival analyses were therefore not performed for these groups. The mean follow-up duration was 86.2 months (range: 24–193 months) for T1 and 85.9 months (range: 32–151 months) for T2 patients. In contrast, patients with T3 and T4 disease showed more variable outcomes. The 5-year OS rate was 100% for early-stage (T1–T2), decreasing to 60.0% for T3, and 42.0% for T4 disease. Advanced T-stage was significantly associated with poorer survival (p < 0.01). Kaplan–Meier survival curves are presented in Fig 1, illustrating lower survival probabilities in T4 compared with T3 patients. Mean follow-up durations for T3 and T4 groups were 21.0 months (range: 9–31 months) and 46.0 months (range 3–218 months), respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Overall survival of patients with external auditory canal SCC (T1-T4).

https://doi.org/10.1371/journal.pone.0339860.g001

Kaplan–Meier overall survival (OS) for the entire cohort, including T1–T4 disease. Number-at-risk values are shown below the plot, and censoring points are indicated on the curves.

In a univariate Cox regression analysis of patients with advanced T4 disease (n = 19), increasing age showed a non-significant trend toward improved survival (hazard ratio [HR] = 0.96, 95% CI: 0.92–1.00, p = 0.079). Variables associated with significantly poorer prognosis included male sex (HR = 4.54, p = 0.023), positive N stage (HR = 3.57, p = 0.028), and carotid artery abutment (HR = 3.67, p = 0.033).

In the multivariate Cox model, age remained a statistically significant independent protective factor (HR = 0.90, 95% CI: 0.83–0.98, p = 0.019). Carotid artery abutment (HR = 16.41, p = 0.050), jugular bulb abutment (HR = 9.01, p = 0.067), and N stage (HR = 4.40, p = 0.096) showed borderline significance, indicating a trend toward increased risk of mortality. Other variables, including sex, Eustachian tube invasion, facial palsy, and tegmen tympani invasion, were not significantly associated with OS.

The multivariate model demonstrated high predictive accuracy, with a concordance index of 0.873. It was statistically significant based on the likelihood ratio test (p = 0.009) and score (log-rank) test (p = 0.020).

Fig 2 presents a forest plot of multivariate Cox proportional hazard regression analysis, illustrating HRs and 95% CIs for all variables. Results of univariate analysis for all tested variables, including both significant and non-significant factors, are presented in Fig 2. Significant predictors of decreased survival included younger age and carotid artery abutment, while jugular bulb abutment and N stage approached significance.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Forest plot of multivariate cox proportional hazard regression analysis.

https://doi.org/10.1371/journal.pone.0339860.g002

This plot displays hazard ratios (HRs) with 95% confidence intervals for variables included in the multivariate Cox regression model assessing prognostic factors for survival in patients with EACC. A hazard ratio greater than 1 indicates an increased risk, while less than 1 indicates a protective effect. Statistically significant variables (p < 0.050) included age (protective) and carotid artery abutment. Error bars represent 95% confidence intervals. All variables examined in the univariate analysis are displayed irrespective of statistical significance.

Discussion

The prognosis of EAC SCC varies considerably based on several factors, including tumor stage, surgical margins, lymph node involvement, perineural invasion, and adjuvant therapy [17]. Despite advancements in surgical and non-surgical treatments, long-term survival remains poor, particularly in advanced-stage disease [3]. In this study, advanced T-stage, nodal metastasis, and vascular abutment—particularly involving the carotid artery—were associated with worse OS, while older age showed a protective effect. These findings highlight the prognostic value of tumor stage and anatomical involvement.

Tumor staging remains the most significant prognostic factor in EAC SCC. The Modified Pittsburgh classification, which integrates radiologic and pathologic findings, remains the most widely used system and has been validated in several landmark studies [18,19]. In these reports, survival outcomes were primarily stratified by overall T stage, with T1–T2 tumors demonstrating significantly better survival than T3–T4 tumors [17]. Consistent with these findings, our results showed excellent outcomes in early-stage disease (100.0% 5-year OS for T1–T2) and markedly poorer survival in advanced stages (60.0% for T3 and 42.0% for T4). These findings further support the oncologic adequacy of surgery alone for early-stage EAC SCC when negative margins are achieved.

Our study also provides one of the largest single-institution datasets of external auditory canal squamous cell carcinoma, including a substantial number of advanced-stage (T3–T4) cases collected over more than two decades. This large, homogeneous cohort enabled a detailed survival analysis within advanced disease, allowing refined evaluation of specific anatomical invasion sites—such as the TMJ, mastoid, and dura—and their prognostic implications. These findings extend the existing staging framework and offer new insight into the heterogeneity of advanced EAC SCC.

Achieving negative surgical margins is essential for curative resection, but this becomes challenging in advanced tumors with invasion of critical structures. As in previous studies, locally advanced disease was the strongest predictor of poor prognosis. Vascular abutment often precludes en bloc resection and was strongly associated with poorer outcomes in our cohort. Positive surgical margins were associated with a significantly higher recurrence rate and worse survival outcomes [3,20], emphasizing the need for complete tumor resection with adequate margins. In our cohort, initial surgical margin positivity closely reflected site-specific invasion of critical structures such as the carotid artery, jugular bulb, and Eustachian tube. Therefore, these anatomical invasion variables were used as surrogate markers for extent of disease rather than including margin status as an independent covariate to avoid multicollinearity in multivariate modeling. Importantly, no patient was left with gross residual tumor. Any macroscopic remaining tumor was removed by additional stepwise excision guided by intraoperative frozen-section confirmation until all accessible margins were cleared. Thus, initial margin positivity did not indicate inadequate oncologic resection. Perineural invasion in this cohort was clinically manifested almost exclusively as facial nerve involvement. Because isolated histopathologic perineural invasion was extremely rare and not statistically analyzable as a separate variable, facial nerve palsy was used as a clinical surrogate marker of perineural invasion in the prognostic analyses. Carotid artery abutment was an independent negative prognostic factor in our analysis, while jugular bulb abutment showed borderline significance. These results highlight the critical importance of early diagnosis before tumors invade essential anatomical structures and curative resection becomes unfeasible.

Recent evidence suggests that habitual ear picking and related chronic irritation may contribute to the development of EAC SCC, particularly in East Asian populations [21,22]. However, the causal direction remains unclear—repeated mechanical irritation could promote carcinogenesis, but such behavior may also arise secondary to early neoplastic changes or preexisting local discomfort. In our cohort, tumor laterality did not show a notable predominance (right 25 vs. left 31), and handedness data were unavailable. Therefore, although this behavioral hypothesis is intriguing, additional prospective studies are required to determine whether ear manipulation represents a true etiologic factor or merely an epiphenomenon of early disease.

A previous retrospective cohort study found that perineural invasion, and venous and lymphatic involvement, significantly worsened OS and DSS, supporting the role of aggressive multimodal therapy [23]. Although these histological features did not reach statistical significance in our cohort, nodal metastasis demonstrated a borderline association with poorer outcomes, suggesting a potential role for the selective use of adjuvant therapy in high-risk patients.

Facial nerve palsy is a known adverse prognostic factor. A prior study found that preoperative facial nerve involvement increased the risk of mortality substantially (HR for OS and DSS was 3.80 and 7.63, respectively) [24]. In our study, facial palsy was not an independent predictor of survival. However, five of the six surviving patients with T4 tumors had normal facial function, further supporting the notion that facial nerve involvement reflects aggressive disease and warrants multidisciplinary management.

Surgical resection remains the cornerstone of treatment for EAC SCC, particularly for advanced-stage tumors. Achieving clear surgical margins is often challenging due to the proximity of vital structures. STBR offers the highest chance of achieving negative margins but is associated with substantial morbidity. Preoperative imaging and meticulous surgical planning are essential. In our cohort, no cases involved direct invasion of the carotid artery or jugular bulb. However, when vascular abutment was present, tumors were excised with minimal margins, followed by cauterization of the surrounding tissues. Postoperative adjuvant RT or CRT was administered to ensure complete eradication of any residual disease. It should be emphasized that positive surgical margins in our analysis referred to the initial en bloc resection specimen rather than the final surgical field status. In advanced T4 cases, residual tumor was commonly removed through additional stepwise excision guided by intraoperative frozen-section assessment to achieve clearance of all accessible margins. Therefore, initial margin positivity does not equate to incomplete oncologic treatment but reflects the technical complexity of en bloc removal in tumors abutting critical structures.

Regarding Case 13, which demonstrated synchronous lung metastasis at diagnosis; however, distant metastasis was not an absolute exclusion criterion for surgical intervention in our treatment strategy. This patient was considered to have oligometastatic disease and underwent surgical resection for loco-regional disease control in conjunction with systemic therapy for distant metastasis. Therefore, surgery was performed with therapeutic intent rather than palliative debulking. This approach reflects real-world multidisciplinary management of select advanced EAC SCC cases.

Regarding cases with preoperative imaging suggesting carotid artery (CA) involvement, it should be clarified that these represented tumor abutment to the vascular surface rather than true intramural or intraluminal invasion. None of the patients in our cohort demonstrated destructive infiltration of the CA wall requiring vascular sacrifice. In these cases, en bloc tumor resection was initially performed, followed by additional targeted excision and careful scraping of the adherent tumor tissue from the CA surface until macroscopic clearance was achieved. Thus, these procedures were undertaken with curative intent rather than palliative debulking. Although all patients with CA abutment ultimately succumbed to disease, this reflects the aggressive nature and extensive skull-base or intracranial involvement of advanced T4 tumors rather than technical futility of surgical intervention itself.

With respect to jugular fossa and jugular bulb involvement, we differentiated bony invasion of the jugular fossa from venous surface abutment of the jugular bulb based on both preoperative imaging and intraoperative findings. Jugular fossa involvement, characterized by skull-base bone destruction, can be reliably identified on preoperative CT and MRI. In contrast, jugular bulb “involvement” in our study referred to tumor abutment or adherence to the venous surface without evidence of intraluminal invasion or venous wall penetration. None of our cases demonstrated true jugular bulb invasion requiring venous sacrifice or resection. When jugular bulb abutment was encountered intraoperatively, additional careful surface excision and scraping were performed to achieve macroscopic clearance; in one case, partial surface reconstruction using absorbable hemostatic materials (e.g., Gelfoam) was required. Therefore, the term “no direct invasion of the jugular bulb” used in this manuscript indicates the absence of intraluminal venous invasion, which is consistent with our surgical management and pathological findings.

The role of adjuvant RT and CRT in improving survival outcomes remains controversial. However, in patients with locally advanced disease (T3–T4), postoperative RT is widely practiced even after negative-margin resection, with the primary goal of improving local disease control rather than overall survival [18,19]. Given the aggressive infiltrative behavior of EAC SCC and the difficulty in ensuring true histological clearance from surrounding critical structures, this multimodal approach is generally considered standard care rather than overtreatment. Adjuvant RT may improve local control without significantly affecting OS [25]. A 5-year LRC of 73.9% with adjuvant RT compared with 45.8% for definitive RT highlights its value in postoperative settings. CRT showed modest benefit in inoperable T2–3 tumors, although prior radiation exposure negatively impacted outcomes, highlighting the need for judicious patient selection [26]. Within the scope of EAC SCC, RT and CRT are generally reserved for adjuvant therapy following surgical resection. Chemotherapy regimens varied substantially across the 20-year study period, as many patients received treatment at local affiliated hospitals. Therefore, protocol heterogeneity limited the ability to compare systemic therapies directly or assess their independent prognostic impact. This highlights the challenges in establishing standardized multimodal treatment strategies for such a rare malignancy.

Balancing oncological control with functional preservation is central in managing advanced EAC SCC. En bloc resections such as LTBR or STBR maximize margin clearance but are associated with significant morbidity, including facial nerve paralysis, hearing loss, and cerebrospinal fluid leaks. Treatment decisions must weigh tumor stage, patient comorbidities, and anticipated quality of life. All surviving T4 patients in our study underwent STBR. In cases of vascular abutment, tumors were resected with minimal margins, followed by cauterization and adjuvant therapy. Given the aggressive nature of advanced EACC, a multidisciplinary approach is essential. Future advancements in RT and molecular-targeted therapies may improve outcomes, but current strategies continue to rely heavily on early detection and complete surgical resection. Prospective studies and molecular profiling are warranted to establish standardized treatment protocols and guide personalized therapy.

Conclusion

EACC, particularly in advanced stages, poses a formidable clinical challenge due to its complex anatomical location and aggressive behavior. Optimal management requires a multidisciplinary approach, with complete surgical resection and negative margins being critical for improving outcomes. Tumor stage, nodal status, and vascular abutment, particularly involving the carotid artery or jugular bulb, are key prognostic indicators. While early-stage tumors are associated with favorable outcomes, advanced-stage disease is marked by high recurrence and poor long-term survival. Given the challenges in achieving clear margins in anatomically constrained regions, early diagnosis remains essential to enable curative en bloc resection and improve prognosis.

Supporting information