Feasibility of Using the Video-Head Impulse Test to Detect the Involved Canal in Benign Paroxysmal Positional Vertigo Presenting With Positional Downbeat Nystagmus

Positional downbeat nystagmus (pDBN) represents a relatively frequent finding. Its possible peripheral origin has been widely scertained. Nevertheless, distinguishing features of peripheral positional nystagmus, including latency, paroxysm and torsional components, may be missing, resulting in challenging differential diagnosis with central pDBN. Moreover, in case of benign paroxysmal positional vertigo (BPPV), detection of the affected canal may be challenging as involvement of the non-ampullary arm of posterior semicircular canal (PSC) results in the same oculomotor responses generated by contralateral anterior canal (ASC)-canalolithiasis. Recent acquisitions suggest that patients with persistent pDBN due to vertical canal-BPPV may exhibit impaired vestibulo-ocular reflex (VOR) for the involved canal on video-head impulse test (vHIT). Since canal hypofunction normalizes following proper canalith repositioning procedures (CRP), an incomplete canalith jam acting as a “low-pass filter” for the affected ampullary receptor has been hypothesized. This study aims to determine the sensitivity of vHIT in detecting canal involvement in patients presenting with pDBN due to vertical canal-BPPV. We retrospectively reviewed the clinical records of 59 consecutive subjects presenting with peripheral pDBN. All patients were tested with video-Frenzel examination and vHIT at presentation and after resolution of symptoms or transformation in typical BPPV-variant. BPPV involving non-ampullary tract of PSC was diagnosed in 78%, ASC-BPPV in 11.9% whereas in 6 cases the involved canal remained unidentified. Presenting VOR-gain values for the affected canal were greatly impaired in cases with persistent pDBN compared to subjects with paroxysmal/transitory nystagmus (p < 0.001). Each patient received CRP for BPPV involving the hypoactive canal or, in case of normal VOR-gain, the assumed affected canal. Each subject exhibiting VOR-gain reduction for the involved canal developed normalization of vHIT data after proper repositioning (p<0.001), proving a close relationship with otoliths altering high-frequency cupular responses. According to our results, overall vHIT sensitivity in detecting the affected SC was 72.9%, increasing up to 88.6% when considering only cases with persistent pDBN where an incomplete canal plug is more likely to occur. vHIT should be routinely used in patients with pDBN as it may enable to localize otoconia within the labyrinth, providing further insights to the pathophysiology of peripheral pDBN.