Unique Anatomical Imaging Findings in Patients With Patulous Eustachian Tube

ABSTRACT

Objectives

The pathophysiology of patulous Eustachian tube (PET) remains poorly understood. This study characterizes nasopharyngeal anatomy to identify features that distinguish PET patients from controls and determine whether anatomical measurements predict symptom laterality.

Methods

Head CT scans from patients with PET diagnosed by Japanese Otologic Society criteria were compared to scans from demographically matched controls. Anatomical features of the Eustachian tube (ET), turbinates, and nasopharynx were measured. PET patients were stratified into bilateral PET, unilateral affected sides, and unilateral unaffected sides.

Results

Thirty PET patients (23 unilateral, 7 bilateral) were included. Compared with controls, PET patients demonstrated significantly larger ET orifice diameters (5.61 vs. 4.54 mm, mean difference 1.07 mm, p < 0.001), greater ET patent depth (6.51 vs. 4.76 mm, mean difference 1.74 mm, p < 0.001), larger lateral recess length (6.40 vs. 3.86 mm, mean difference 2.54 mm, p = 0.002), larger lateral recess width (4.37 vs. 2.51 mm, mean difference 1.86 mm, p < 0.001), and greater inferior turbinate‐to‐ET asymmetry (2.67 vs. 1.24 mm, mean difference 1.43 mm, p < 0.001). Nasopharyngeal dimensions did not differ between groups. No significant differences were identified across groups in the primary analysis. Sensitivity analyses accounting for within‐patient correlation demonstrated greater ET patent depth and lateral recess width in symptomatic compared with asymptomatic unilateral PET ears.

Conclusion

PET patients demonstrate distinct nasopharyngeal anatomy beyond simple ET patency, including enlarged lateral recesses and greater asymmetry in inferior turbinate positioning. However, anatomical measurements poorly predict symptom laterality, indicating that additional physiologic mechanisms not captured by static CT imaging likely contribute to symptom expression.

Level of Evidence

4.

This retrospective CT‐based study evaluated nasopharyngeal anatomy in patients with patulous Eustachian tube (PET) compared with matched controls. PET patients demonstrated increased ET patency and enlarged lateral recess dimensions; however, no significant anatomical differences were identified between bilateral PET ears, affected unilateral ears, and unaffected unilateral ears. These findings suggest that while anatomy may establish susceptibility to PET, dynamic airflow mechanisms likely contribute to symptom laterality and clinical manifestation.

1. Introduction

Patulous Eustachian tube (PET) is commonly described as abnormal patency of the Eustachian tube (ET) that creates an increased passage of airflow into the middle ear. This air movement can result in symptoms of aural fullness, autophony, and amplified breathing that can be modulated with positional changes [1]. PET affects approximately 0.3%–6.6% of the population, with a higher prevalence in females [2]. Commonly reported comorbidities include environmental allergy, weight loss, laryngopharyngeal reflux, and anxiety, though their mechanistic relationship to PET remains unclear [3].

Despite increasing clinical recognition, the pathophysiology of PET remains poorly defined. Proposed mechanisms include reduced soft tissue volume within the cartilaginous ET and atrophy or sclerosis of the peritubal nasopharyngeal structures, though many cases occur without an identifiable structural cause [4]. Emerging evidence has shifted attention toward airflow dynamics, as maneuvers that temporarily alter airflow or reduce nasopharyngeal turbulence (e.g., Cottle maneuver, nasal tip elevation, and ear canal occlusion) often result in transient symptom relief [5].

Diagnostic criteria for PET, as described by the Japanese Otologic Society (JOS), incorporate subjective symptoms (autophony, aural fullness), improvement with positional changes or pharyngeal orifice obstruction, and objective findings such as tympanic membrane hypermobility with respiration or sonotubometric changes [6]. Computed tomography (CT) has been utilized primarily to exclude other conditions that present similarly, such as superior semicircular canal dehiscence, but its potential role in characterizing nasopharyngeal anatomical features associated with PET has not been systematically evaluated.

Given the potential role of airflow dynamics in PET pathophysiology, anatomical differences in the nasopharynx may contribute to symptom development. This study utilizes CT imaging to identify nasopharyngeal anatomical differences between PET patients and matched controls, and to determine whether these anatomical features predict symptom laterality.

2. Methods

This retrospective cohort study was conducted within the clinical practice of a single board‐certified, fellowship‐trained otologist at a tertiary‐care institution and approved by the hospital Institutional Review Board. Patients were included if they met JOS diagnostic criteria for PET and had a head or sinus CT scan obtained at the time of diagnosis [5]. All PET patients reported one or more of the following subjective symptoms: aural fullness, voice autophony, and/or breathing autophony (Criterion 1) and demonstrated improvement of symptoms with posture change to a supine/lordotic position or with pharyngeal orifice obstruction procedure (Criterion 2). Some patients also had objective findings of PET such as tympanic membrane movement with respiration, but sonotubometry was not performed (Criterion 3). Patients with other causes of autophony, including superior semicircular canal dehiscence or external auditory canal pathology were excluded based on clinical evaluation and CT imaging. Demographic and clinical variables, including otolaryngologic comorbidities, were obtained by electronic medical record review.

A control cohort was assembled from patients undergoing CT imaging for non‐ET‐related indications who had no documented history of autophony, aural fullness attributed to ET dysfunction, or PET diagnosis. Controls were matched to PET patients on age, sex, race, and body mass index (BMI). CT examinations obtained at the time of diagnosis or within 6 months of diagnosis were performed using Xoran Technologies MiniCAT and GE Medical Systems Revolution HD scanners with institutional head and sinus CT protocols. Acquisition parameters varied between patients, with slice thicknesses of 0.4, 1.25, or 2.5 mm. All measurements were performed on multiplanar reconstructions using a standardized measurement protocol. Most scans were obtained in the supine position; however, 6 PET patients and 4 control patients underwent CT imaging in the upright position.

2.1. Outcome Measurements

The primary outcomes were linear measurements (in millimeters) of structures in the nasopharynx that influence airflow near the ET orifice. To confirm ET patency, the width (a) and patent depth (b) of the ET were recorded for both PET and control cohorts. Additional structures measured included the length (c) and width (d) of the lateral recess (fossa of Rosenmüller), distance from the inferior turbinate to the ET orifice (e), length of the nasopharynx (f) measured from the posterior nasal septum to the posterior wall of the nasopharynx at the level of the ET orifice, and width of the nasopharynx at the level of the ET orifice (g). Asymmetry in the inferior turbinate‐ET distance was calculated as |e_right − e_left|, and cross‐sectional area of the nasopharynx was estimated as the product of measurements (f) and (g). Figure 1 indicates the location of each measurement. All measurements were performed manually by a single trained reviewer following this standardized protocol under supervision of the senior author to minimize inter‐rater variability.

FIGURE 1.

Axial CT image depicting measurement locations of nasopharyngeal structures that influence airflow across the ET orifice. (a) ET orifice diameter, (b) ET patent depth, (c) lateral recess length, (d) lateral recess width, (e) inferior turbinate‐to‐ET distance, (f) nasopharyngeal length measured from posterior nasal septum to posterior nasopharyngeal wall at the level of the ET orifice, (g) nasopharyngeal width at the level of the ET orifice.

For comparative analyses between PET and control cohorts, right and left measurements for variables (a) through (d) were grouped together, yielding two measurements per patient. For PET patients, measurements were obtained bilaterally, including both symptomatic and asymptomatic sides in unilateral cases. Variables representing overall nasopharyngeal dimensions, including inferior turbinate‐ET asymmetry, nasopharyngeal length, nasopharyngeal width, and cross‐sectional area, were recorded once per patient. Within the PET cohort, anatomic measurements were compared among bilateral PET patients, affected sides of unilateral PET patients, and unaffected sides of unilateral PET patients.

2.2. Statistical Analysis

Descriptive statistics were conducted to report demographic information and confirm appropriate matching between PET and control cohorts. Independent‐samples t‐tests were performed when comparing PET and control patients. One‐way analysis of variance (ANOVA) was performed when comparing bilateral PET patients, affected unilateral sides, and unaffected unilateral sides. Because bilateral ear measurements were included for individual patients, sensitivity analyses were performed using linear mixed‐effects models with a random intercept for patient to account for non‐independence of ear‐level measurements. Effect sizes are reported as mean differences with 95% confidence intervals (CI). Statistical analyses were performed using Python (Python Software Foundation, Wilmington, DE, USA). All tests were two‐tailed, and a p‐value < 0.05 was considered statistically significant.

3. Results

3.1. Patient and Disease Characteristics

Thirty PET patients and 30 matched control patients were included. There were no significant differences between PET and control cohorts in age, sex, race, or BMI. The PET cohort was 63% female, had a mean age of 59.8 ± 17.2 years, and race identified as 70% White, 17% as Black, 7% as Asian, and 7% as Hispanic. The mean BMI was 28.5 kg/m². Twenty‐three patients (77%) had unilateral PET and 7 patients (23%) had bilateral PET.

3.2. Radiologic Characteristics

CT measurement analyses demonstrated that patients with PET, including both symptomatic and asymptomatic ears in unilateral cases, had larger ET orifice diameters (5.61 vs. 4.54 mm; mean difference 1.07 mm, 95% CI: 0.56–1.57; p < 0.001) and patent depth (6.51 vs. 4.76 mm; mean difference 1.74 mm, 95% CI: 0.98–2.51; p < 0.001) than controls. Similarly, PET patients demonstrated larger lateral recess length (6.40 vs. 3.86 mm; mean difference 2.54 mm, 95% CI: 0.94–4.13; p = 0.002) and width (4.37 vs. 2.51 mm; mean difference 1.86 mm, 95% CI: 1.27–2.45; p < 0.001) than controls. Asymmetry in inferior turbinate‐to‐ET distance was greater in PET patients (2.67 vs. 1.24 mm; mean difference 1.43 mm, 95% CI: 0.69–2.17; p < 0.001). Nasopharyngeal length, width, and cross‐sectional area did not differ significantly between the PET and control cohorts (Table 1).

TABLE 1.

Nasopharyngeal measurements (in mm) obtained from CT images for patients with patulous eustachian tube (PET) and matched control patients with corresponding Wilcoxon comparisons. Thirty patients in total with 2 measurements (each side) for a–d (n = 60) and 1 measurement for e–g.

Measurement location	Anatomical region	PET group (in mm) (n = 60)	Control (in mm) (n = 60)	p
a	ET diameter	5.61	4.54	< 0.001*
b	ET patent depth	6.51	4.76	< 0.001*
c	Lateral recess length	6.40	3.86	0.002*
d	Lateral recess width	4.37	2.51	< 0.001*

Measurement Location	Anatomical Region	PET (n = 30)	Control (n = 30)	p
|e_right − e_left|	Absolute value of difference between right and left inferior turbinate to ET	2.67	1.24	< 0.001*
f	NP length	21.0	19.5	0.148
g	NP width	20.6	20.3	0.719
f × g	NP cross‐sectional area	433.9	393.4	0.152

Note: Bold and asterisk indicate statistical significance values.

When comparing bilateral PET patients, affected unilateral sides, and unaffected unilateral sides using ANOVA, no significant anatomical differences were detected across all five measurements (ET diameter, ET patent depth, lateral recess length, lateral recess width, and inferior turbinate‐to‐ET distance; Table 2).

TABLE 2.

CT measurement comparisons between bilateral PET, affected unilateral PET, and unaffected unilateral PET with corresponding ANOVA comparisons.

Anatomic region measurements	Bilateral PET (in mm) (n = 14)	Unilateral PET—affected side (in mm) (n = 23)	Unilateral PET—unaffected side (in mm) (n = 23)	p
ET diameter (a)	6.19	5.33	5.53	0.164
ET patent depth (b)	6.16	7.07	6.15	0.389
Lateral recess length (c)	5.39	6.60	6.82	0.663
Lateral recess width (d)	4.22	4.77	4.06	0.344
Inferior turbinate to ET (e)	11.58	11.63	10.71	0.633

Sensitivity analyses using linear mixed‐effects models accounting for clustering by patient yielded findings similar to the primary analyses. PET patients continued to demonstrate significantly larger ET diameter (p = 0.001), ET patent depth (p = 0.001), lateral recess length (p = 0.020), and lateral recess width (p < 0.001) compared with controls. Within unilateral PET patients, affected ears demonstrated greater ET patent depth (p = 0.006) and lateral recess width (p = 0.008) than unaffected ears, whereas no other measurements differed significantly.

4. Discussion

PET remains a challenging and often debilitating disorder whose underlying mechanisms are not fully understood [3, 7]. Traditional approaches have focused on reducing tubal patency through surgical interventions targeting the ET orifice, yet these procedures demonstrate inconsistent efficacy and high recurrence rates [8, 9]. Our findings demonstrate that PET patients exhibit distinct nasopharyngeal anatomy compared with controls; however, anatomy alone does not fully explain symptom laterality or distinguish unilateral from bilateral disease. Notably, PET patients demonstrated anatomic differences from controls on both symptomatic and asymptomatic sides, suggesting that anatomy may establish a predisposed state while symptom expression may be influenced by additional dynamic physiologic factors not captured by static CT imaging.

4.1. Anatomic Differences Between PET and Controls

Our findings confirm prior literature demonstrating anatomical differences between PET patients and controls. PET patients had significantly larger ET orifice diameters, greater patent depths, and larger lateral recess dimensions compared to controls. Yoshida et al. demonstrated that visualized ET lumen accounted for 83% of total tube length in PET patients versus 59% in controls [10], while Kikuchi et al. found most PET patients showed fully patent ETs not observed in controls [11]. Together, these findings support the concept that PET patients possess a baseline anatomic predisposition characterized by increased tubal patency and altered peritubal geometry.

We also noted that PET patients exhibited significantly greater asymmetry in inferior turbinate‐to‐ET distances (2.67 vs. 1.24 mm), suggesting that air currents entering the nasopharynx from each side of the nose differ in PET patients, while are likely equivalent in controls. Importantly, nasopharyngeal dimensions (length, width, and cross‐sectional area) did not differ between groups, indicating that overall nasopharyngeal volume alone does not determine PET susceptibility. Instead, these findings suggest that localized structural features around the ET orifice and lateral recess may exert greater influence on airflow behavior relevant to symptom development in PET.

4.2. Anatomy Does Not Predict Symptom Laterality

A critical observation is that anatomical measurements poorly predict symptom laterality within PET patients. Despite clearly different symptom patterns, most anatomical measurements did not differ between bilateral PET ears, affected unilateral ears, and unaffected unilateral ears. Sensitivity analyses accounting for within‐patient correlation demonstrated greater ET patent depth and lateral recess width in symptomatic compared with asymptomatic unilateral PET ears, although the overall pattern of findings remained remarkably similar across groups. The magnitude of these differences was modest relative to the larger anatomical differences observed between PET patients and controls, suggesting that they are unlikely to fully account for symptom laterality. This finding has important implications. If anatomy alone determined symptom expression, affected sides, unaffected sides, and bilateral cases would be expected to demonstrate distinct anatomic profiles. Instead, our results indicate that PET patients exhibit anatomical symmetry even when symptoms are unilateral.

These findings clarify a long‐standing clinical paradox in PET. Prior morphometric studies have noted anatomic symmetry despite unilateral symptoms [12, 13], and our results extend this observation by demonstrating that even when PET patients are stratified by symptom laterality and bilaterality, anatomical parameters alone do not differentiate clinically distinct states. Instead, the data indicate that anatomy may create a predisposed environment, but additional dynamic factors likely determine which side becomes clinically active and whether disease manifests unilaterally or bilaterally. Grimmer et al. emphasize that while anatomical patency is necessary for PET, dynamic factors such as airflow, mucosal changes, and functional tube behavior are likely required for symptom manifestation and laterality [14]. Thus, our findings indicate that anatomic patency establishes susceptibility to PET but does not determine which side becomes symptomatic.

4.3. Integrating Anatomy With Airflow Dynamics

Taken together, these findings support a model in which anatomic predisposition alone may be insufficient to explain PET symptom expression. One potential explanation is that dynamic factors, including nasopharyngeal airflow behavior, contribute to the development and laterality of symptoms. The nasopharynx can be conceptualized as a resonant chamber where variance in airflow velocity, pressure, and directionality patterns may create turbulence and influence sound transmission to the ET lumen [15]. As supported by Ikeda et al. increased ET patent depth and enlarged lateral recesses may reduce resistance to airflow and allow fluctuations to propagate more effectively into the ET [16]. Greater asymmetry in inferior turbinate positioning implies airflow trajectories entering the nasopharynx differ on each side of the nose, potentially creating turbulence as these air currents meet. Similarly, the depth and shape of the lateral recess, a cul‐de‐sac known to promote flow separation, may amplify pressure fluctuations that reach the ET orifice.

Computational fluid dynamics (CFD) studies of nasal airflow have demonstrated that anatomical irregularities, such as turbinate hypertrophy or abrupt geometric expansions, create turbulence in otherwise laminar flow patterns [17, 18]. Regions with sudden recesses, such as the lateral nasopharyngeal wall, are particularly prone to flow separation and eddy formation. Clinically, PET symptoms worsen during exercise and high respiratory rates, conditions that increase airflow velocity and turbulence potential [3]. Conversely, symptoms improve in the recumbent position, where decreased airflow velocity and vascular engorgement of the torus tubarius may dampen turbulence [10]. These clinical and aerodynamic observations are consistent with a mechanism in which anatomy defines the boundary conditions for airflow, but dynamic airflow behavior ultimately governs symptom generation. Ultimately, future studies utilizing computational fluid dynamics are essential to test this airflow hypothesis.

4.4. Implications for Treatment Failures

This airflow hypothesis may explain the disappointing results of procedures aimed solely at narrowing the ET orifice. Ward et al. reported recurrence rates of 39%–68% within 1 year following various ET reduction procedures, with only complete tube obliteration providing reliable symptom resolution, at the cost of persistent middle ear effusion requiring tympanostomy tubes in 82% of cases [9]. If turbulence rather than absolute lumen size drives symptoms, interventions that fail to address the upstream airflow conditions in the nasopharynx would predictably have limited efficacy. This framework may also explain anecdotal reports of PET symptom onset or worsening following nasal surgeries that alter turbinate anatomy or nasal airflow patterns, as such changes could modify aerodynamic forces acting at the ET orifice.

4.5. Limitations

This study has limitations. Measurements were obtained manually and are subject to inter‐rater variability, though standardized protocols were employed to minimize measurement error. While most CT scans were obtained in supine position with only a subset obtained upright, matching PET and control cohorts for patient positioning mitigates this as a confounding factor. Additionally, CT acquisition parameters were not standardized because imaging was obtained retrospectively as part of routine clinical care. Scans were acquired on multiple scanner platforms using slice thicknesses of 0.4, 1.25, or 2.5 mm. However, all measurements were performed using a standardized protocol, and the observed differences between PET and control cohorts remained significant despite this variability. The relatively small sample size, particularly for bilateral PET patients (n = 7), limits statistical power for detecting smaller effect sizes in subgroup analyses. Despite these limitations, the clear anatomical differences between PET patients and controls, combined with the poor correlation between anatomy and symptom laterality, provide valuable insights into PET pathophysiology.

5. Conclusion

This study demonstrates that patients with PET exhibit distinct nasopharyngeal anatomical features compared with matched controls, including larger ET orifices, greater patent tubal depth, and enlarged lateral recess dimensions. However, these structural differences do not reliably predict symptom laterality or distinguish unilateral from bilateral disease. These findings suggest that anatomic variations may create a predisposed state for PET, while additional physiologic factors not captured by static CT imaging likely contribute to symptom expression.

Funding

The authors have nothing to report.

Disclosure

The article was presented as a poster presentation at the Combined Otolaryngology Spring Meetings: Triological Society, New Orleans, Louisiana, USA, May 15–16, 2025.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.