Direct Visualization of Tracer Permeation into the Endolymph in Human Patients Using MR Imaging

Abstract

Purpose

The endolymph of the inner ear, vital for balance and hearing, has long been considered impermeable to intravenously administered gadolinium-based contrast agents (GBCAs) due to the tight blood-endolymph barrier. However, anecdotal observations suggested potential GBCA entry in delayed heavily T2-weighted 3D-real inversion recovery (IR) MRI scans. This study systematically investigated GBCA distribution in the endolymph using this 3D-real IR sequence.

Methods

Forty-one patients suspected of endolymphatic hydrops (EHs) underwent pre-contrast, 4-h, and 24-h post-contrast 3D-real IR imaging. Signal intensity in cerebrospinal fluid (CSF), perilymph, and endolymph was measured and analyzed for temporal dynamics of GBCA uptake, correlations between compartments, and the influence of age and presence of EH.

Results

Endolymph showed a delayed peak GBCA uptake at 24h, contrasting with peaks in perilymph and CSF at 4h. Weak to moderate positive correlations between endolymph and CSF contrast effect were observed at both 4 (r = 0.483) and 24h (r = 0.585), suggesting possible inter-compartmental interactions. Neither the presence of EH nor age significantly influenced endolymph enhancement. However, both perilymph and CSF contrast effects significantly correlated with age at both time points.

Conclusion

This study provides the first in vivo systematic confirmation of GBCA entering the endolymph following intravenous administration. Notably, endolymph uptake peaked at 24h, significantly later than perilymph and CSF. The lack of a link between endolymph contrast and both perilymph and age suggests distinct uptake mechanisms. These findings shed light on inner ear fluid dynamics and their potential implications in Ménière’s disease and other inner ear disorders.

Introduction

The inner ear’s endolymphatic space is tightly shielded from the bloodstream, leading to the long-held belief that intravenously administered gadolinium-based contrast agents (GBCAs) cannot enter this compartment.^1,2 In contrast, the perilymph surrounding the endolymph possesses a more permeable barrier. This allows for visualization of the distinct endolymphatic and perilymphatic spaces 4h after intravenous GBCA administration, due to the resulting concentration difference between the two compartments.^3,4 Notably, GBCAs instilled directly into the tympanic cavity also distribute solely within the perilymph, even after 24 hours, further highlighting the endolymph’s restricted accessibility.⁵

Delayed contrast MRI remains the standard clinical practice for diagnosing endolymphatic hydrops (EHs), a condition characterized by excess endolymph fluid.^2,6–8 In Japan, confirmation of EHs via this method has even become a diagnostic criterion for “the certain Ménière’s disease.”⁹ However, the endolymph’s restricted volume and robust barrier pose significant challenges for studying its fluid dynamics and its role in inner ear diseases. This, in turn, hinders our ability to predict the ototoxicity of systemic drugs reaching the endolymph or monitor the delivery of therapeutic agents for inner ear disorders. Consequently, researchers often rely on perilymph GBCA distribution as a surrogate for understanding endolymph drug exposure.^10–12

Developing a method to directly assess the blood-endolymph barrier and endolymph fluid dynamics would be a transformative tool for unraveling the pathophysiology of inner ear diseases and paving the way for novel treatment strategies. Our extensive experience with over 4000 MRI cases using highly sensitive, heavily T2-weighted 3D-FLAIR (fluid attenuated inversion recovery) and 3D-real inversion recovery (IR) sequences for EHs assessment has led us to observe suggestive evidence of the small amount of GBCA reaching the endolymph in a significant number of cases.^13–16 The heavily T2-weighted 3D-real IR with long TR and high sensitivity for detecting low-concentration GBCA in fluids is used not only for assessing EHs but also for detecting slight migration of GBCA to the vitreous humor of the eyes and cerebrospinal fluid (CSF).^15,17–20

This retrospective study aims to systematically investigate the distribution of GBCA in the endolymph using the heavily T2-weighted 3D-real IR sequence. We will specifically examine:

• Temporal dynamics: GBCA uptake in the endolymph before, 4h after, and 24h after intravenous administration.

• Relationship to age and EHs: Potential influence of these factors on GBCA distribution.

• Comparison with other compartments: GBCA distribution in the CSF and perilymphatic space relative to the endolymph.

Materials and Methods

Patients

Out of 42 patients suspected of EHs who underwent 3D-real IR imaging before, 4h after, and 24h after contrast administration at our hospital between 2018 and 2022, one case with an incidental vestibular schwannoma in the internal auditory canal was excluded. This left 41 patients (22 males, 19 females; age range: 17–80 years; median age: 47). Data from 16 of these patients had previously been used in a report on the temporal progression of perilymphatic enhancement, but the enhancement of the endolymph had not been examined at all.²¹ No subjects in this study had an estimated glomerular filtration rate below 50 mL/min/1.73 m². The ethical committee of our institution approved this study (approval number: 2018-0218). Written informed consent was obtained from all patients.

MRI

Imaging was performed using a 32-channel coil on a 3T scanner (Skyra; Siemens Healthineers, Erlangen, Germany). Heavily T2-weighted 3D-real IR was acquired in an axial slab covering the entire brain. The slab position was consistently aligned using the autoalign function for each scan.^22,23 A brief overview of the imaging method is as follows: 3D-real IR images were obtained before, 4h, and 24h after an intravenous injection of a single dose (0.1 mmol/kg) of macrocyclic GBCA (gadobutrol; Bayer, Osaka, Japan). The detailed parameters for the 3D-real IR imaging were the same as in the previous study.^13,21,24

Briefly, a TR of 15130 ms, TE of 549 ms, TI (inversion time) of 2700 ms, and pixel size of 0.5 × 0.5 mm², with 1 mm thickness, phase-sensitive reconstruction (real reconstruction), and scan time of 10 min, were applied. Imaging was performed under the same conditions 4s and 24h after the intravenous injection of GBCA. GBCA administration was carried out between 10 and 11 A.M. in all cases. There were no dietary restrictions and no specific activity restrictions after GBCA administration.

Image evaluation

The presence of EHs was assessed using Nakashima’s Grading scale²⁵ by a radiologist with 16 years of experience in imaging diagnosis of EHs. The assessment for each ear, both vestibule and cochlea, was done as none, mild, or significant on the day of the examination or the following day. In this study, only significant EHs were considered positive, while all others were considered negative.

Signal intensity measurements were performed by the same radiologist, at least 1 yr after the EHs evaluation. Although it was unavoidable to see the degree of EHs when measuring the signal intensity of the lymph fluid in the images. The signal intensity of the CSF was measured by placing a 3-mm circular ROI in the interpeduncular cistern, avoiding vessels and artifacts, as in the previous reports.^13,21,24 The same ROI was copied and measured in other phases.²⁶ The signal intensity of the perilymph was measured as in the previous reports by manually encircling the scala tympani of the cochlear basal turn on the images taken 4h later and copying this ROI to other phases. If there was a positional shift between phases, the ROI was manually moved slightly as appropriate.²¹ For the endolymph, a 1-mm circular ROI was drawn in the posterior part of the utricle in the images taken 4h later and copied to other phases. The posterior part of the utricle was chosen as the location least affected by partial volume effects.

All measurement values were standardized by drawing a 15-mm diameter circular ROI in the cerebellar hemisphere in the slice at the level of the internal auditory canal in each phase, and the signal intensity ratio to that signal value was calculated. This signal intensity ratio was used for subsequent analysis. Measurements were made for each ear, resulting in data for 82 ears in 41 patients. The contrast effect in endolymph, perilymph, and CSF after 24 or 4h was determined by subtracting the pre-contrast signal intensity ratio from the respective post-contrast signal intensity ratio.

Statistical analysis

1. Comparing signal intensity ratios across time points: We used the Friedman test (a nonparametric test for repeated measures) to compare the average signal intensity ratios in the perilymph, endolymph, and CSF before, 4h after, and 24h after GBCA administration in all 82 ears. To account for multiple comparisons, we adjusted the P values using the Bonferroni method.

2. Comparing contrast effects at each time point among three fluids: We focused on the changes (contrast effects) in signal intensity ratios after GBCA administration, comparing 4 and 24h to the baseline. We again used the Friedman test with Bonferroni correction for multiple comparisons. Additionally, we performed pairwise comparisons between the contrast effect for each pair of the fluids at specific time points using the Wilcoxon signed-rank test.

3. Examining correlations between contrast effects: We used Pearson’s correlation coefficient (r) to investigate the relationship between the contrast effect in the endolymph at 4h and the contrast effects in the CSF and perilymph at the same time point. We repeated this analysis for the 24-h time point and also examined the correlation between the endolymph contrast effect at 24h and the CSF and perilymph contrast effects at 4h.

4. Comparing contrast effects in hydrops and non-hydrops groups: We divided the ears into two groups: those with significant EHs and those without. We used the Mann-Whitney U test (a nonparametric test for comparing two independent groups) to compare the contrast effects in the CSF, perilymph, and endolymph between these groups at each time point.

5. Investigating correlations with age: We used Spearman’s rank correlation coefficient (ρ) to assess the relationship between each contrast effect (in all compartments and at all time points) and the age of the participants.

All P values were two-sided, and P < 0.05 was considered to represent a significant difference. All confidence intervals were set to 95%. All analyses were performed using R software (version 3.2.2; R Foundation for Statistical Computing).

Results

1. Comparing signal intensity ratios across time points: Our study measured the time course of signal intensity ratios in the perilymph, endolymph, and CSF following GBCA administration. We found that signal intensity ratios significantly differed over time in the perilymph, endolymph, and CSF (P < 0.001). Peak intensity ratios occurred at 4h in the perilymph and CSF, as shown in Fig. 1. A delayed peak was observed at 24h in the endolymph (Fig. 1). Representative images are presented (Fig. 2).

2. Comparing contrast effects at each time point among three fluids: 4h: Significant differences in contrast effect were detected among three fluids (P < 0.001). Pairwise comparisons revealed significant differences between perilymph and both endolymph (P < 0.001) and CSF (P < 0.001). Endolymph and CSF did not show a significant difference (P = 0.34). 24h: Similar to 4h, three fluids exhibited significant differences in contrast effect (P < 0.001). Pairwise comparisons confirmed significant differences between perilymph and both endolymph (P < 0.001) and CSF (P < 0.001). Endolymph and CSF remained statistically indistinct (P = 0.24).

3. Examining correlations between contrast effects: No significant correlation was found between the contrast effect in the endolymph and perilymph at 4h (r = 0.208, P = 0.061, Fig. 3a). However, a significant positive correlation was observed between the endolymph and CSF at 4h (r = 0.483, P < 0.001, Fig. 3b). The endolymph contrast effect at 24h correlated positively with both CSF at 4h (r = 0.508, P < 0.001, Fig. 3c) and 24h (r = 0.585, P < 0.001, Fig. 3d). The endolymph contrast effect at 24h showed a weak positive correlation with both perilymph at 4h (r = 0.2813, P < 0.05, Fig. 3e) and 24h (r = 0.333, P < 0.001, Fig. 3f).

4. Comparing contrast effects in hydrops and non-hydrops groups: No significant differences in contrast effects were observed between the ears with positive or negative cochlear or vestibular EHs at either 4 or 24h (Fig. 4a-l).

5. Investigating correlations with age: Age exhibited no statistically significant association with the endolymph contrast effect (Spearman’s ρ = 0.0122, P = 0.914 at 4h; ρ = 0.108, P = 0.336 at 24h). In contrast, both perilymph and CSF contrast effects displayed significant age-related changes at both 4 and 24h. Perilymph ρ values were 0.357 (P = 0.001) at 4h and 0.359 (P = 0.001) at 24h. CSF ρ values were 0.47 (P = 0.00001) at 4h and 0.39 (P = 0.0004) at 24h.

Fig. 1.

Time course of signal intensity ratio for CSF (a), perilymph (b), and endolymph (c). Mean signal intensity ratio among pre-contrast, 4h after contrast administration, and 24h after contrast administration significantly differed for CSF, perilymph, and endolymph (P < 0.001). The peak of the signal intensity ratio for endolymph was observed at 24h, while that for CSF and perilymph was observed at 4h after contrast administration. CSF, cerebrospinal fluid.

Fig. 2.

Representative images of the right inner ear from a female patient in her 50s. 3D-real IR image before (a), at 4h (b), and at 24h (c) after intravenous administration of single-dose gadolinium-based contrast agent. Window width and level were set uniformly among images. The signal of endolymph (arrows) gradually increases from (a) to (c). In this patient, no significant endolymphatic hydrops were observed in cochlea and vestibule. Note that the signal of the perilymph decreases from 4 to 24h, while that of endolymph increases from 4 to 24h. Although the contrast effect of endolymph is far weaker than that of perilymph, the temporal progression of contrast agent permeation is clearly visible on 3D-real IR. IR, inversion recovery.

Fig. 3.

The correlations between contrast effects. (a) There was no significant correlation of contrast effect between endolymph and perilymph at 4h (r = 0.208, P = 0.061). (b) A significant positive correlation was observed between the endolymph and CSF at 4h (r = 0.483, P < 0.001). The contrast effect of endolymph at 24h correlated positively with both that of CSF at 4h (r = 0.508, P < 0.001, c) and 24h (r = 0.585, P < 0.001, d). (e) A weak positive correlation was observed between the endolymph at 24h and the perilymph at 4h (r = 0.2813, P < 0.05). (f) A weak positive correlation was observed between the endolymph and perilymph at 24h (r = 0.333, P < 0.001). CSF, cerebrospinal fluid.

Fig. 4.

Box-and-whisker plots comparing contrast effects in the group of the ears with significant cochlear hydrops and that without cochlear hydrops (a–f): the contrast effect of CSF (a), perilymph (b), and endolymph (c) at 4h; that of CSF (d), perilymph (e), and endolymph (f) at 24h. Box-and-whisker plots comparing contrast effects in the group of the ears with significant vestibular hydrops and that without vestibular hydrops (g–l): the contrast effect of CSF (g), perilymph (h), and endolymph (i) at 4h; that of CSF (j), perilymph (k), and endolymph (l) at 24h. No significant difference was observed in all comparisons. CSF, cerebrospinal fluid.

Discussion

This study demonstrated for the first time in humans that a single dose of intravenously administered GBCAs gradually permeated into the endolymphatic space over time, as detected by 3D-real IR. Unlike the perilymph and CSF, the contrast effect in the endolymph was more pronounced at 24h than at 4h postinjection. Unlike the CSF and perilymph, no correlation was observed between the contrast effect in the endolymph and age. Also, no relation was found between the contrast effect in the endolymph and the presence of EHs.

Anatomically, the endolymph is in extensive contact with the perilymph. The stronger contrast effect in the perilymph could potentially be inferred as causing an apparent signal increase in the endolymph due to partial volume effects. However, the lack of correlation between the contrast effects in the perilymph and endolymph at 4h makes this inference less likely. Additionally, while the perilymph’s contrast effect is stronger at 4h than at 24h, the opposite is true for the endolymph. Therefore, the signal increase in the endolymph cannot be solely explained by partial volume effects from the enhanced perilymph.

There are reports suggesting communication between the perilymph and endolymph,²⁷ and that the endolymph is produced from the perilymph.²⁸ However, the dynamics of water and drugs, having different molecular weights, suggest that the distribution of GBCAs in the endolymph and perilymph may not be entirely parallel. The present results show that while there is a weak positive correlation between the contrast effects in the endolymph and perilymph at 24h, there is no correlation at 4h. It is more natural to consider that the migration of intravenously administered GBCAs into the endolymphatic space primarily occurs through leakage from the surrounding capillaries rather than from the perilymphatic space.

Supporting this notion, it has been reported that systemically administered aminoglycosides are trafficked via the endolymph into cochlear hair cells. This trafficking route is predominant compared to uptake via the basolateral membranes of hair cells during perilymph infusion.²⁹ This indicates that systemically administered drugs reach the hair cells predominantly through the endolymph from blood vessels rather than via the perilymph. Aminoglycosides are ototoxic, so understanding their dynamics is useful. The molecular weights of GBCAs, steroids used for the treatment of sudden hearing loss, and ototoxic drugs like gentamicin and platinum compounds are as follows:

• – GBCA: Gadobutrol, 604.71; Gadoteridol, 558.6848; and Dotarem: 753.8553

• – Steroid: Prednisolone, 360.444 and Dexamethasone, 392.464

• – Ototoxic agents: Gentamicin (an aminoglycoside), 477.5954; Carboplatin, 371.25; and Cisplatin, 300.05.

As mentioned earlier, the molecular weights of these drugs are relatively close to those of gadolinium compounds. While the permeability of barriers is not determined solely by molecular weight, having similar molecular weights is a necessary condition for using one drug as a surrogate marker for the distribution of another.

The blood-perilymph barrier is reported to increase in permeability with age and in various diseases such as sudden deafness, vestibular neuritis, vestibular schwannoma, otosclerosis, and Ménière’s disease.^11,30,31 The contrast effect in the perilymphatic space has been suggested as a biomarker in EHs.³²

There are few reports on the increased permeability of the blood-endolymph barrier in diseases. There have been case reports of contrast enhancement in the endolymphatic space in patients with severe sudden hearing loss.³³ In animal models of autoimmune diseases, changes in the tight junctions of capillaries in the stria vascularis and increased permeability of the blood-endolymph barrier have been reported.³⁴

Clinically, the ability to assess increased permeability of the blood-endolymph barrier could be important for predicting the distribution of drugs such as gentamicin and steroids after systemic administration.²⁹ Additionally, gathering data on healthy individuals and various diseases may uncover the clinical utility of the contrast effect in the endolymph as a biomarker.

The contrast effect in the endolymph, as mentioned earlier, is expected to have applications in elucidating the pathophysiology of inner ear diseases, monitoring disease progression, and assessing treatment efficacy. To achieve this, prospective studies in various diseases are necessary. Furthermore, technical improvements to enhance the detection sensitivity of low-concentration contrast agents will also be required.

Stable isotope ¹⁷O-water, which has a T2 shortening effect on proton MRI, can be indirectly detected with clinical MRI devices. Administered ¹⁷O-water, having a molecular weight of only one unit greater than regular water, essentially exhibits similar dynamics to normal water.³⁵ Preliminary studies in healthy individuals and patients with EHs have shown that ¹⁷O-water administered into the tympanic cavity rapidly distributes not only to the perilymphatic space but also to the endolymphatic space.³⁶ Reports of tracers reaching the endolymphatic space in healthy individuals are rare. The evaluation of water dynamics in the endolymphatic space using ¹⁷O-water is a promising method. However, ¹⁷O-water is currently not clinically available, and intratympanic administration is invasive. Moreover, vertigo and nystagmus presumably due to the buoyancy effect of ¹⁷O-water have been reported following intratympanic administration, and widespread use of intratympanic ¹⁷O-water has not been achieved.³⁶ Furthermore, the dynamics of water may not coincide with those of drugs. Analysis of the dynamics of endolymphatic fluid would likely require evaluation of both water and drug dynamics.

The limitations of this study include that all cases were suspected of EHs, with no healthy individuals included. In EHs, the permeability of the blood-perilymph barrier is increased.^21,30,32,37 It is possible that the results are due to increased permeability of the blood-endolymph barrier in EHs. However, since there was no significant difference in the contrast effect in the endolymph between those with and without EHs, this result does not strongly support this idea. Additionally, like the blood-brain barrier, the blood-perilymph barrier showed age-related increased permeability, consistent with reported findings. However, no correlation was found with age in the endolymph. Further investigation in various age groups, including healthy individuals, is necessary.

Future studies should aim to quantify GBCA concentration, as the current evaluation was based on signal values in 3D-real IR measured by a single observer. For example, the contrast effect of endolymph at 4h showed negative values of −2.8, −3.1, and −4.3 in three ears (Fig. 3a), suggesting that some signal fluctuations are included in the measurements of subtle contrast enhancement. Further technical improvements are needed for a more reliable assessment of the contrast enhancement in endolymph.

Considering patient burden, evaluation was only possible for up to 24h. Further assessment at 36 or 48h might evaluate the drainage function of GBCAs from the endolymphatic space. We may need to enroll healthy volunteers for the longer-term study.

Impairments of the drainage functions in the brain, such as the glymphatic system and the intramural periarterial drainage system, are being actively researched as potential causes of neurodegenerative diseases like Alzheimer’s.^38,39 Similarly, researches on the ocular glymphatic system and its relation to glaucoma are underway.^17,40,41 Regarding the inner ear, although the dynamics of drainage of the perilymph have been investigated by Yoshida et al., further research is needed on the drainage of the endolymph, which contains the sensory organs of the inner ear.²¹ This study has set the stage for such research.

In summary, this study revealed distinct temporal patterns of contrast uptake in the perilymph, endolymph, and CSF. Notably, the endolymph exhibited a delayed peak compared to the other fluids.

Positive correlations were observed between the endolymph and CSF at both 4 and 24h, suggesting potential interactions between these compartments. Neither hydrops type nor age had a significant influence on the contrast effect. These findings offer valuable insights into the dynamics of inner ear fluid exchange and their potential role in Ménière’s disease and other inner ear disorders.

Conclusion

The distribution of GBCAs into the endolymphatic space after intravenous administration of conventional doses of GBCA was first confirmed over time. GBCAs were found to distribute more into the endolymphatic space at 24h postinjection than at 4h. The contrast effect in the endolymph showed little correlation with the effect in the perilymph, as well as with age and the presence of EHs. This study opens the door to evaluating the pharmacodynamics in the endolymphatic space. Further research on the contrast effect in the endolymphatic space is anticipated.