Letter to the Editor: “Visualization of the saccule and utricle with non-contrast-enhanced FLAIR sequences”
by Michael Eliezer (michael.eliezer@aphp.fr)
Visualization of the saccule and utricle with non-contrast-enhanced FLAIR sequencesDear Editor-in-Chief,
We read with great interest the study of Fukutomi et al in European Radiology [1]. The authors performed a step-by-step study on healthy subjects to assess the saccule and utricle with non-contrast-enhanced FLAIR sequences. Despite the well conducted study, we were very confused by some of their interpretation and results. From a methodological point of view, the authors performed their study on 22 healthy subjects by evaluating the effect of T2-prep and time inversion (TI) on the contrast between the perilymphatic and endolymphatic spaces. However, we assume that the authors cannot state that the utricle and saccule might be truly observed with non-contrast 3D-FLAIR sequences since no comparison with delayed inner ear MRI has been performed. Up to date, the gold standard for the MRI evaluation of the endolymphatic space remains by using gadolinium contrast agents, which cross progressively the blood-perilymph barrier, thereby inducing a high contrast between the perilymphatic space (with gadolinium uptake) and the endolymphatic space (without gadolinium uptake) [2]. Even if this method provides an evaluation of the endolymphatic space with a high reliability, we agree that a more physiological approach without contrast administration would be highly required. Nevertheless, a delayed acquisition would still be required since various inner ear disorders could only present with a slight blood-labyrinth barrier impairment that could not be observed with routine inner ear MRI [3-5].
As the authors have stated, the evaluation of the saccule and utricle without gadolinium administration by using heavily-T2-weighted sequences, questions the reliability of this method because of the very low contrast within the vestibule and the thin membrane of the membranous labyrinth [6,7]. To our knowledge, only one study enabled to assess clearly the saccule and utricle without gadolinium administration on T2-weighted sequences in patients with obstructive vestibular schwannomas, based on the high protein content in the perilymphatic space [8]. The latter providing a high contrast between the perilymphatic and the endolymphatic spaces because of the T1-shortening induced by the high protein content in the perilymphatic space.
The authors have stated that a 7.2 times difference in protein concentration exists between the perilymphatic and the endolymphatic spaces in pig tissue. Yet, this difference in protein concentration is smaller in human, approximately 3 times (0.6 g/l for the endolymphatic space, 2 g/l for the perilymphatic space) [9]. Based on our experience, we also agree with the authors about the similar T1 values of the perilymphatic and endolymphatic spaces should be close. Indeed, we have previously performed high-resolution (slice thickness of 0.8 mm) T1-mapping in many patients (unpublished results) without contrast media administration and found close values between these two spaces. Therefore, we also agree with the authors that the T2 values should be different between these two spaces. Unfortunately, because of the very small size of the structures and the high-resolution required, T2-mapping could not be performed, since the slice thickness could not be lower than 2 mm.
By using a T2-preparation module before the inversion pulse enables to reduce the unwanted T1-weighting for a given repetition time. The T2-preparation module contains a 90° excitation pulse, a variable number of refocusing pulses and a -90° flip back pulse. The time between excitation and flip-back pulse is chosen such that the transversal magnetization of tissues with comparatively T2 times significantly decays. This results in more complete recovery of these tissues during the time inversion period and hence less unwanted T1-weighting. Hence, it seems logical that by combing T2-preparation module with the inversion recovery pulse of a 3D-FLAIR sequence could provide sufficient contrast between the perilymphatic and endolymphatic spaces. However, based on our own experience, this approach should be used very carefully. First, the thin band of perilymph that separates the saccule and the utricle is the major issue in most cases. It is true that, by increasing the TI, therefore the perilymphatic signal, this thin band of perilymph could be captured. Unfortunately, the consequence of this increased TI leads to an imperfect suppression of some part of the endolymphatic space, such as the anterior part of the utricle and most of the ampullas. We would kindly ask the authors to provide some images acquired in this specific area to demonstrate whether these parts of the endolymphatic space are well suppressed without gadolinium injection.
As a reminder, 4 hours inner ear MRI enables to assess the endolymphatic space in patients with primary and secondary hydropic ear diseases, but also in patients with unilateral and bilateral vestibular areflexia. In these patients, a vestibular atelectasis of the pars superior is frequently involved, consisting in an absence of visualization of the utricle and the ampullas [10,11]. For this reason, combining high T2-prep values and time inversion with non-contrast 3D-FLAIR sequences remains problematic since the whole anatomy of the endolymphatic space could not be well depicted with this method. This observation is in concordance with the study of Conte et al. that evaluated the membranous labyrinth in 10 infants using a heavily T2-weighted 3D-FLAIR sequence without contrast media administration [12]. However, the T2-prep value was not described in their study. The authors stated that the saccule was well observed in 83.3% the utricle in 100% but the ampullas were clearly visible in half of cases, which questions the reliability of this method to assess adequately the membranous labyrinth.