Tragus or cymba conchae? Investigating the anatomical foundation of transcutaneous auricular vagus nerve stimulation (taVNS)

Dear Editor

We thank Burger and Verkuil [1] for bringing attention to the only published cadaver ear dissection studies in the literature, published in 2002 by Peuker and Filler [2]. This cadaver study is indeed important, and has fueled the increased interest in modulating the auricular branch of the vagus nerve (ABVN). Researchers clearly want to know which ear target is optimal for taVNS e tragus or cymba conchae? Burger and Verkuil have found and investigated an apparent discrepancy in Peuker and Filler’s work, and then contacted Prof. Filler for clarification. Unfortunately, this discrepancy was difficult to resolve. We have several comments.

First, although the Peuker and Filler manuscript is important, there are other studies describing vagus innervation of the ear, best exemplified with the Eastern medicine field of therapeutic auricular acupuncture. Auricular acupuncture has been used for centuries and these studies add to this body of information [3–5]. Thus, it is critical to acknowledge that the field of taVNS does not necessarily rest on one table in one manuscript.

Second, we do not necessarily see the same conflict. See the figure from Peuker and Filler (Fig. 1a) as well as a complementary figure from He et al. (Fig. 1b). Although the arrow of the ABVN is unclear and may point to the tragus as well as the conchae, the table (Fig. 1c) describes the vagus as being 45% innervated by the ABVN (with the conchae having 100% of its fibers from the vagus, but slightly more proximal). Both locations likely engage some vagus fibers. Although the cymba conchae targets may be 100% innervated by the ABVN according to the table, ABVN fibers innervate the anterior wall of the external ear canal, landmarked by the tragus. The tragus also offers some advantages in terms of ease of applying electricity to the anterior wall of the external ear canal by being able to clip onto the tragus, as opposed to having to insert and hold an electrode against the conchae.

Fig. 1.

Figures and table taken and reproduced here from the initial Peuker & Filler cadaver paper (a,c) as well as an anatomical drawing from He et al. (b) demonstrating ABVN innervation of the human ear.

Finally, our imaging study, and others [6,7], combined with heart rate data in adults [8] and now infants (Badran, Jenkins et al., under review, Brain Stimulation) indicate that tragus stimulation engages vagal afferents. Other groups have also demonstrated vagus sensory evoked potentials (VSEP) from tragus stimulation [9] as well as heart rate variability and reduced sympathetic activity [10].

We agree, however, with Burger and Verkuil that we do not know the absolute best ear target for taVNS, as the innervation of the ABVN remains unclear. More anatomical dissection studies would help resolve this inconsistency and we hope that this correspondence spurs further research. These cadaver studies should be conducted by individuals with advanced training in microdissection of the head and neck, as the ABVN diameter is on a micrometer resolution. Teams should also consider age, ethnicity, trauma, and prior health history in their selection of cadavers. A large, diverse, cross-section of individuals would aid in optimizing taVNS.

Advanced In Vivo peripheral nerve imaging using highresolution ultrasonography (US) may also be used to track nerve innervation, although poor resolution may be a limitation. Alternatively, the magnetic resonance imaging (MRI) scanning protocol known as diffusion tensor imaging (DTI) could potentially be used to image the peripheral nerves. In vivo imaging approaches might eventually permit individualized ABVN targeting and placement of taVNS.

In order to optimize noninvasive brain stimulation modalities as therapeutics, it is critical to consider 4 basic fundamentals: 1) anatomy; 2) target engagement; 3) parameter space; and 4) behavioral effects. taVNS is still in its infancy with groups investigating target and parameter engagement via central [6,7,9] and peripheral [8,10] biomarkers. Behavioral effects are being demonstrated in numerous neuropsychiatric disorders. We need more work on the anatomical basis of taVNS in order to better guide our future trials.