RPS 1008 - Temporal bone and auditory pathway

Purpose:

To investigate the prevalence of cerebral small vessel disease (SVD) in a consecutive cohort of patients presenting with acute symptomatic lacunar ischaemic strokes (ASLIS) and to ascertain how frequently an ASLIS represents the initial radiological manifestation of SVD.

Methods and materials:

In a hospital trust, a database of 831 patients undergoing MRI for acute stroke from May 2018-February 2019 was retrospectively reviewed to identify patients with DWI+ASLIS. Patients with a history of trauma or cancer, haemorrhagic strokes, and large vessel disease were excluded. Patients were split into those with and without SVD by using established scoring systems. Subjects with evidence of SVD were assigned to three categories: arteriolosclerosis, probable/possible cerebral amyloid angiopathy (CAA), and mixed disease. Potential risk factors were evaluated by Fisher’s and Student’s tests, and calculating the odds ratio, between the groups of patients with and without evidence of SVD and among the subgroups of patients with SVD.

Results:

Of 104 subjects included, 29 (28%) showed no evidence of SVD, while of the remaining 75 (72%), 59 (57%) had evidence of arteriolosclerosis, 4 (4%) of CAA, and 12 (11%) of mixed disease. Among the markers of SVD, white matter hyperintensity was the most common. Subjects with evidence of SVD were significantly more prone to be older, smokers, and affected by diabetes, hypercholesterolemia, and hypertension (OR:2.1).

Conclusion:

In 2/3 of patients with ASLIS and a co-existing burden of SVD, arteriolosclerosis was predominant and subjects were likely to present cardiovascular risk factors. DWI+ASLIS was the first imaging manifestation of SVD in about 1/3 of subjects. These patients may have different pathophysiology, different risk of stroke recurrence, and need tailored therapy.

Limitations:

The relative small cohort analysed.

Ethics committee approval

The study was approved by the institutional ethics committee.

Funding:

No funding was received for this work.

Purpose:

High definition magnetic resonance (HDMR) is the gold standard method in the study of the pathologies of the inner ear, however, there are no precise findings to determine a differential diagnosis between them. We investigated the usefulness of dedicated MR sequences in the characterisation of the intralabyrinthine pathology.

Methods and materials:

We included patients evaluated between January 2010 and September 2019. Demographic and imaging data was collected. MRI images included high-resolution T2-weighted (HRT2), pre-enhanced T1-weighted (T1PE) and contrast-enhanced T1-weighted (T1CE). We placed ROIs on both labyrinths and in the fourth ventricle.

Results:

MRI scans of 50 patients were analysed. The median age was 63.5 years (IQR 43-73). We found 18 (36%) patients with inflammatory labyrinthitis, 16 (32%) with intra-labyrinthine neurinoma, 9 (18%) with intra-labyrinthine haemorrhage, and 7 (14%) with ossifying labyrinthitis. All patients with intra-labyrinthine neurinoma had a low signal in HRT2, a mild-high signal in T1PE, and contrast enhancement. 14 patients with inflammatory labyrinthitis (77.8%) had low signal intensity in HRT2 and high signal in T1PE, and 13 patients (72.2%) had contrast enhancement. All patients with ossifying labyrinthitis had low signal intensity in HRT2 and 6 patients (87.5%) had high signal intensity in T1PE. All patients with haemorrhage had a high signal in pre-enhanced T1W. When we compared intra-labyrinthine neurinoma with inflammatory labyrinthitis, we found a significant difference in low signal intensity in HRT2 (p 0.001), T1PE (p 0.014), and in T1CE (p 0.003).

Conclusion:

There are multiple entities that may alter the labyrinth signal in MR, some with similar radiologic patterns. We found differences in all the sequences studied between inflammatory labyrinthitis and intracochlear neurinoma.

Limitations:

A diagnosis based on imaging findings without histopathology.

Ethics committee approval

Approved by an ethics committee.

Funding:

No funding was received for this work.

Purpose:

To evaluate the association of audio-vestibular symptoms with the presence of vascular loops and vascular contact in the cerebellopontine angle and the internal auditory canal using 3-Tesla magnetic resonance imaging.

Methods and materials:

The study included 98 patients (196 ears); 51 females and 47 males, age range 11-73 years, mean 47.6±15 years. The healthy control group with no symptoms in either ear, n=60 (120 ears); 32 females and 28 males, age range 12-69 years, mean 45.3±15.8 years. The non-symptomatic ear in the patients were added to the healthy control group. Patients underwent a neuro-otologic evaluation to exclude an underlying pathologic process. The patients with neuritis or tumours at the CPA were not included in the study. No patients had previous CPA surgery or temporal bone trauma. All MR imaging examinations were performed by using a 3T (Magnetom Verio 3T; Siemens Medical Solutions, Erlangen, Germany). The imaging protocol consisted of axial T2-weighted images of the whole brain and coronal and axial T1-weighted images of the CPA before and after administration of intravenous contrast. 3D constructive interference steady state (CISS) imaging of the CPA was performed. The healthy control group was examined with the 3D CISS sequence in addition to the routine cranial MR imaging protocol.

Results:

No statistically significant association was detected between the studied audiovestibular symptoms; tinnitus, deafness or vertigo and vascular loop (grades I-III) or vascular contact (type I-III) (p<0.05).

Conclusion:

It was concluded that no possible role of the presence of vascular loop or vascular contact with the 8th cranial nerve in causing tinnitus, deafness or vertigo using 3D-CISS sequence assessment. Therefore, these findings are not certainly considered pathological but possibly to be normal anatomical variants.

Limitations:

The sample size.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

Necrotising external otitis (NEO) is a serious complication of the external otitis. NEO can be classified according to anterior, medial, intracranial, and crossed spreading patterns. Currently, there is no consensus on the optimal imaging modality for diagnosis. This study compares NEO spreading patterns and relevant subsites as assessed on MR and HRCT of the temporal bone in order to evaluate diagnostic comparability.

Methods and materials:

We retrospectively examined 21 NEO patients who received an HRCT temporal bone and MR within a 3-month interval. Involvement of subsites and subsequent spreading patterns were assessed by a neuroradiologist blinded for the other imaging modality. The prevalence of spreading patterns by CT and MR were calculated and compared.

Results:

In all patients, an anterior spreading pattern was noted on both MR and CT. Retrocondylar fat infiltration was the most consistent subsite of the anterior spreading pattern. The medial spreading pattern was seen in 7 patients at MR and in 5 of these 7 patients at CT, with parapharyngeal fat infiltration as the most consistent subsite. The intracranial spreading pattern matched in 1 out of 6 patients. The majority of patients showed a single (62%) and matching anterior spreading pattern at CT and MR.

Conclusion:

The most common anterior spreading pattern of NEO can be assessed with CT as well as MR. However, the more complicated medial and intracranial spreading patterns as seen on MR could only be identified on CT in a small number of patients. Diagnosis of NEO at CT should therefore be complemented with additional imaging.

Limitations:

The cohort consisted of a small number of included patients. A wide time interval between CT and MR was present, therefore a disease progression/regression between CT and MR is not excluded.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

To characterise the imaging features of small (<1 cm) endolymphatic sac tumours (ELSTs) in von Hippel-Lindau disease (VHL).

Methods and materials:

19 VHL patients with 22 small ELST (3 with bilateral small ELSTs), 10 males and 9 females with ages between 11 and 74, were enrolled in a longitudinal natural history study of VHL.

High-resolution temporal-bone CT (<1.5 mm) and contrast enhanced MR (<3.0 mm, typically 1.5 thick/0.75 mm spacing T1-weighted) were available in all patients. Endolymphatic hydrops was evaluated in 9 cases using delayed FLAIR obtained 6-12 hours after contrast injection.

Results:

Indirect MR findings: endolymphatic hydrops (4/9), intralabyrinthine haemorrhage (7/22), and enhancement within the ipsilateral vestibular aqueduct (13/22).

Direct MRI findings: intratemporal mass around the aqueduct (11/22) and distension/mass involving the extraosseous portion of the endolymphatic sac (10/22).

CT findings: erosion adjacent to the aqueduct (13/22) and asymmetric enlargement of the ipsilateral aqueduct (6/22).

Four cases were CT negative, no case was MRI negative, but 5 cases exhibited only “indirect” signs on MRI. One of these “indirect MRI only” cases were confirmed by erosion on CT.

Of the 4 cases with hydrops, 2 had neither intratemporal mass nor CT findings, and 1 had neither erosion nor intratemporal mass.

Conclusion:

As VHL associated ELSTs cause irreversible audiovestibular morbidity independent of tumour size, early diagnosis and surgery have been advocated. ELSTs are locally invasive so that erosion on CT is diagnostic. Direct MRI tumour findings may be present when CT is negative. In some cases, radiologists must rely on indirect findings, e.g. intralabyrinthine haemorrhage or hydrops (requiring delayed FLAIR imaging) to establish the diagnosis of a small ELST.

Limitations:

n/a

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

The review question was: what is the length of the Corti organ (OC) in humankind?

Current studies offered that cochlear size might affect implant success in patients with hearing loss.

The hypothesis was: are the covariates (such as gender) affected the OC length?

Methods and materials:

This meta-analysis was organised using the PRISMA statement.

The gender, age, country, continent, methods, and materials were determined as covariates, and this data was extracted from the studies.

OC was determined as an outcome. Data was pooled according to a random-effects model.

Meta-regression analysis was performed with Stata software (ver. 14.0) Metareg command. The overall effect was estimated with Metaan command.

Results:

The review included 47 studies (n=3,976 material) and found that covariates weren't predictors of outcome.

Estimation of the between‐study variance for the OC length (τ²=0.7692) and the ratio of residual variance resulting from heterogeneity (I²=0%) was statistically insignificant. Also, the coefficient determination of the model was R²=19.18% (p=0.356). There wasn't a statistically significant relationship between the OC value and covariates.

The overall effect value for the OC length was 32.742 (CI %95:32.270-33.214) mm.

Conclusion:

Evaluating each patient's OC size by imaging prior to cochlear implant surgery is important because of the wide variation.

The fact that there is a difference between spiral coefficient methods with the histological method (the gold standard for measuring the OC) necessitates the development of these methods.

Measurements that are made on images obtained by 3D reconstruction seem to be the future topics of radiology.

Limitations:

The review includes only English language papers (except cites). There was no study from South America and Africa.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

To analyse the degree of interobserver agreement in the detection and grading of endolymphatic hydrops (EH) with different scales and to correlate MRI findings with clinical information.

Methods and materials:

75 patients, diagnosed with unilateral definite Meniere's disease (MD), undergoing 3T-MRI of the inner ear were recruited. Cisternography, T2-FLAIR, and REAL IR imaging were performed 4 hours after intravenous contrast administration. Two independent radiologists blinded to clinical data analysed the images. Vestibular EH was evaluated using 4-stage (none/mild/moderate/severe), 3-stage (none/moderate/severe), and 2-stage (none/present) visual scales. Three-point (none/present/severe) and two-point (none/present) scales were employed for cochlear EH assessment. Discrepancies were solved by consensus. Interobserver agreement was evaluated with weighted-kappa (w-Κ) statistics. The association of pure tone audiometry (PTA) and caloric test with MR findings of EH was analysed with one-way ANOVA (p<0.05).

Results:

Out of 75 patients, we identified EH in 90.6% of the clinically affected ear and in 17.3% on the silent side. Substantial or excellent interobserver agreement was found in all cases (mean w-Κ=0.83; range 0.7-0.91). The agreement was almost perfect with the 4-point vestibular EH scale (0.91±0.1). For cochlear EH, the highest consistency was obtained with the 2-point-scale (0.76±0.1). A statistically significant association was found between PTA and vestibular and cochlear EH, with all the scales. The strongest association was observed with the 2-stage cochlear (49.5±22 vs 25.4±21.5,F=45.1, p<0.001) and vestibular EH (54.8±20.1 vs 27.1±21.2,F=64.9, p<0.001) grading systems. The caloric test was also associated with 2-stage and 4-stage vestibular EH scales (F=4.5, p=0.034 and F=2.9, p=0.035, respectively).

Conclusion:

A 4-stage vestibular EH grading system gave the best interobserver consistency. Vestibular and cochlear EH detected by MRI were associated with PTA and caloric tests, especially with the 2-scale grading system.

Limitations:

A retrospective, single-institution study.

Ethics committee approval

Ethics committee approval obtained and informed consent waived.

Funding:

No funding was received for this work.