Mastoid effusion on temporal bone MRI in patients with Bell’s palsy and Ramsay Hunt syndrome

This study aimed to investigate the incidence of mastoid effusion on temporal bone magnetic resonance imaging (MRI) in patients with Bell’s palsy (BP) and Ramsay Hunt syndrome (RHS), and evaluate the usefulness of mastoid effusion in early differential diagnosis between BP and RHS. The incidence of mastoid effusion on 3.0 T—temporal bone MRI, which was conducted within 10 days after the onset of acute facial nerve palsy, was compared between 131 patients with BP and 33 patients with RHS. Findings of mastoid cavity on temporal bone MRI were classified into three groups as normal mastoid, mastoid effusion, and sclerotic change, and the incidence of ipsilesional mastoid effusion was significantly higher in RHS than BP (P < 0.001). Tympanic membrane was normal in 7 of 14 RHS patients with mastoid effusion, and injected without middle ear effusion in 7 patients. This study highlights significantly higher incidence of ipsilesional mastoid effusion in RHS than BP, and suggests that the presence of mastoid effusion may provide additional information for differential diagnosis between RHS and BP.

Bell's palsy (BP) is the most common type of acute peripheral facial nerve palsy (FNP) with an annual incidence of 20-30 per 100,000 persons 1 , and Ramsay Hunt syndrome (RHS) accounts for approximately 10% of acute FNP with an annual incidence of 5 per 100,000 population 2 . RHS is characterized by otalgia and erythematous vesicular rash in the external auditory meatus or auricle in addition to acute FNP. While vertigo or hearing loss is commonly accompanied with FNP in RHS, symptoms of vestibulocochlear deficit is rare in BP 3 . In comparison to BP, the severity of FNP is greater and recovery rate is worse in RHS 4,5 . On the other hand, the gustatory perception and morphology may be bilaterally affected in both BP and RHS at the acute phase 6 . It has been established that combination therapy with systemic steroids and antiviral agents shows benefit over systemic steroids alone in RHS 7 , whereas the additional benefit from combination therapy over steroids alone has not been confirmed in BP 8 . Several prognostic factors for the recovery of FNP have been reported, and early treatment was suggested as one of the most important prognostic factors in RHS 9,10 . Thus, initiating early treatment after proper diagnosis is crucial to improve the recovery rate in RHL. However, differential diagnosis between RHS and BP at an early stage is sometimes difficult because the sequence, in which symptoms appear, may vary among individual patients with RHS. For example, acute FNP may precede erythematous skin rash in some patients with RHS 11,12 . Moreover, in cases such as zoster sine herpete, in which acute FNP develops without erythematous skin rash, differentiation of RHS from BP based on clinical presentation is difficult.
Temporal bone magnetic resonance imaging (MRI) has been utilized to evaluate disease extent of RHS [13][14][15][16][17] and compare imaging characteristics between RHS and BP [18][19][20] . Most of all, temporal bone MRI is helpful for the differential diagnosis between BP and RHS, facilitating early diagnosis of RHS and supporting rapid initiation of combination therapy with systemic steroids and antiviral agents 20 . We accidentally discovered mastoid effusion on the ipsilesional side of FNP in one patient with RHS, and attempted to retrospectively investigate if presence of mastoid effusion in temporal bone MRI may provide additional information for differential diagnosis between RHS and BP. The purpose of the present study was to investigate the incidence of mastoid effusion on temporal bone MRI in patients with BP and RHS, and evaluate the usefulness of mastoid effusion in differential diagnosis between BP and RHS.

Results
The clinical characteristics in patients with BP and RHS are summarized in Table 1. Male to Female ratio was 65 : 66 in BP and 14 : 19 in RHS, which was not significantly different (P = 0.460, Pearson's chi-squared test). Mean age was 45 years in BP and 53 years in RHS, which was not significantly different (P = 0.369, Student's t-test). Affected side was right in 63 patients with BP (left side in 68 patients) and 15 patients with RHS (left side in 18 patients), which was not significantly different (P = 0.786, Pearson's chi-squared test). The time period from symptom onset to temporal bone MRI was 2.9 ± 2.7 days in BP and 2.7 ± 2.6 days in RHS, which was not significantly different (P = 0.821, Student's t-test). The worst House-Brackmann (HB) grade during an acute stage was significantly higher in RHS (3.6 ± 0.9) than BP (3.2 ± 0.8) (P = 0.005, Student's t-test). Proportion of patients with hearing loss was significantly higher in RHS (88%) than BP (0%) (P < 0.001, Pearson's chi-squared test). Proportion of patients with vertigo was significantly higher in RHS (73%) than BP (0%) (P < 0.001, Pearson's chi-squared test).
While previous studies investigated the imaging findings of facial nerve, inner ear end organs, dura, and vestibulocochlear nerve in patients with acute peripheral FNP [13][14][15][16][17][18][19][20] , the present study focused on the mastoid cavity in the affected side on temporal bone MRI. Findings of mastoid cavity were classified into three groups as normal mastoid (Fig. 1), mastoid effusion (Fig. 2), and sclerotic change (Fig. 3). Figure 1 demonstrates a representative case with normal mastoid in BP patient on the right side. T2-weighted axial image shows normal appearance  1A). Post-contrast axial and coronal T1-weighted images show enhancement of the right facial nerve at the distal canalicular, labyrinthine, geniculate ganglion, and proximal tympanic segments ( Fig. 1B, C).  Figure 3 demonstrates a representative case with sclerotic mastoid in a patient with BP on the right side. Axial T2-weighted image depicts heterogeneous high signal intensity in the mastoids which have similar signal intensity from the adjacent clivus and petrous apex, suggesting bony sclerotic change (Fig. 3A). Temporal bone computed tomography (TBCT) confirms the bony sclerotic change with decreased number of mastoid air cells (Fig. 3B). Post-contrast axial T1-weighted image at the Bill's bar shows enhancement of the distal canalicular, labyrinthine, geniculate ganglion and proximal tympanic segment of the right facial nerve (Fig. 3C). Normal mastoid cavity in temporal bone MRI was observed in 98% (128 of 131) of BP and 48% (16 of 33) of RHS, and ipsilesional mastoid effusion was observed in 0% of BP and 42% (14 of 33) of RHS (Table 2). Bony sclerotic change was observed in 2% (3 of 131) of BP and 9% (3 of 33) of RHS, and the sclerotic change was observed in both mastoid cavities in all patients. It has been suggested that the extent of mastoid pneumatization is determined genetically, rendering its variation an anatomical and etiological factor for chronic otitis media 22 , or determined by environmental factors such as pathologic influences on the middle ear and mastoid air cell system during childhood 23 . Because no evidence for chronic otitis media was suspected in any of six patients with the sclerotic mastoid cavity in temporal bone MRI and the sclerotic change was observed not only on the same side with FNP but also on the contralateral side (Fig. 3A, B) in all six patients, those patients were excluded from the statistical analysis, given that sclerotic bony change in the mastoid is not associated with the acute onset of FNP. Proportion of mastoid effusion was significantly higher in RHS (47%, 14 of 30) than BP (0%, 0 of 128)   Table 3). Air-bone gap was not revealed on a pure tone audiometry in any of 14 RHS patients with mastoid effusion. Otoendoscopic examination revealed normal tympanic membrane in 7 patients (of 14 RHS patients with mastoid effusion on temporal bone MRI), and injected tympanic membrane without recognizable middle ear effusion in 7 patients (of 14 RHS patients with mastoid effusion on temporal bone MRI). TBCT or brain CT, which was conducted in 6 patients of 14 RHS patients with mastoid effusion on temporal bone MRI, demonstrated too small amount of mastoid effusion to be easily detected (Fig. 4).
In the present study, only RHS patients with typical symptoms of acute FNP with vesicular skin eruption were included (see Subjects and Methods). We then evaluated temporal bone MRI findings of another three atypical RHS patients who did not show vesicular eruption or acute FNP (Fig. 5). Temporal bone MRI of a 64-year old female patient, who was diagnosed as zoster sine herpete in the right side, is shown in Fig. 5A, B. Postcontrast 3D T1-weighted MRI at the level of the IAC demonstrated the enhancement of the labyrinthine segment of the right facial nerve, right vestibular nerve and IAC dura (Fig. 5A). Non-contrast axial T2-weighted image at the level of mastoid air cells showed high signal intensity suggesting mastoid effusion on the affected side (Fig. 5B). Temporal bone MRI of a 40-year-old male patient, who developed vesicular skin eruption in the left ear and acute    Fig. 5C, D. Postcontrast 3D T1-weighted MRI showed the enhancement of the labyrinthine segment of the left facial nerve, left vestibular nerve and IAC dura (Fig. 5C). Non-contrast axial T2-weighted image showed normal mastoid cavity in the affected side (Fig. 5D). Temporal bone MRI of a 28-year-old male patient, who developed vesicular skin eruption in the left ear and acute vertigo and hearing loss without acute FNP, is shown in Fig. 5E, F. Postcontrast 3D T1-weighted MRI showed the enhancement of the labyrinthine segment of the left facial nerve, left vestibular nerve, and IAC dura (Fig. 5E). Non-contrast axial T2-weighted image at the level of mastoid air cells showed normal mastoid cavity in the affected side (Fig. 5F).

Discussion
The present study demonstrates that the significantly higher incidence of mastoid effusion was observed in RHS than BP on temporal bone MRI. Previous studies have emphasized that early differential diagnosis between RHS and BP is important due to their different prognoses and treatment modalities 5,7-10,24,25 . However, differentiating RHS from BP based on clinical presentation at an early stage is sometimes difficult in RHS cases with atypical clinical manifestations or zoster sine herpete 11,12,15,26 . Temporal bone MRI has been utilized for the early differential diagnosis between BP and RHS [13][14][15][16][17][18][19][20] . While the enhancement is generally localized within the facial nerve in BP, the enhancement, as shown in the present study (Fig. 2), has been observed in not only the facial nerve but also the vestibulocochlear nerve on temporal bone MRI in RHS [12][13][14][15]20,[27][28][29][30][31][32] . Enhancement of the inner ear organs has been frequently observed in RHS 12,13,15,19,20,27,29,30,33 , and dural enhancement along the IAC, as shown in the present study (Fig. 2), has also been reported 15,20,27,28,30,33 . In RHS with multiple cranial neuropathy, enhancement of involved cranial nerves in the brainstem has been observed in temporal bone MRI 15,34 . Soft tissue swelling and enhancement of the involved auricle has been reported in temporal bone MRI of RHS patients 15,30 . Kuya et al. reported that 3D MRI sequences are useful for the differential diagnosis between RHS and BP, and RHS, compared to BP, shows more thickening of the facial nerve in the fundus of IAC on 3D-contructive interference on steady state sequence (3D-CISS) image 20 . Sugiura et al. reported that high signal intensity of the inner ear organs is observed on non-contrast 3D-fluid-attenuated inversion recovery (3D-FLAIR) images in RHS 16 .
In addition to the above-mentioned characteristic findings of temporal bone MRI in RHS, the presence of mastoid effusion in temporal bone MRI may provide an additional information in differentiating RHS from BP, because, though not all of RHS patients showed mastoid effusion (14 of 33 RHS patients), mastoid effusion was not observed in any of the patients with BP. In addition, among three RHS patients with atypical clinical manifestations, one patient with zoster sine herpete showed mastoid effusion on the affected side. It has been reported that incidental mastoid effusion on MRI, in which isolated fluid in the mastoid bone appears with absent corroborating clinical features, may have little clinical significance [35][36][37][38] . However, mastoid effusion, which was observed in 14 patients with RHS in the present study, may be associated with RHS, because the mastoid effusion was observed on the same side with RHS in all of 14 patients. The mechanism underlying mastoid effusion in RHS is not clear. We speculate that impairment of autonomic nervous function, which is thought to play a role in the regulation of local blood flow and glandular secretion in middle ear and mastoid mucosa, is responsible for the production of mastoid effusion in patients with RHS 39 .  www.nature.com/scientificreports/ While we performed this study with 3.0 T MRI only, it has been reported that facial nerve enhancement can also be detected on 1.5 T MRI with high spatial resolution in pathologic conditions such as BP, RHS and tumorous lesions including lymphoma and leukemia 40 . However, because higher spatial resolution and more sensitivity www.nature.com/scientificreports/ to the contrast materials can be achieved by using the 3.0 T MRI, we can detect facial nerve enhancement in pathologic conditions more rapidly and sensitively. Moreover, it is more convenient to differentiate enhancement between facial, vestibulocochlear nerve and dura by using the 3.0 T MRI. However, considering that mild to moderate enhancement of the intrameatal (15%) and labyrinthine (5%) segments of the normal facial nerve may be observed on 3.0 T MRI 41 , attention should be paid in the interpretation of the presence of facial nerve enhancement by comparing the pattern and degree of the bilateral facial nerve enhancement on MRI.

Conclusions
In conclusion, this study highlighted significantly higher incidence of ipsilesional mastoid effusion in RHS than BP, and suggested that the presence of mastoid effusion may provide additional information for differential diagnosis between RHS and BP. In addition to enhancement of vestibulocochlear nerve and/or dura along the IAC and/or inner ear organs in temporal bone MRI, identification of mastoid effusion may be a valuable MRI finding for early differential diagnosis.

Data availability
The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.