By Steven Karceski, MD

Assistant Professor of Clinical Neurology, Department of Neurology, Weill Cornell Medical College

Dr. Karceski reports he provides patient education for NeuroPace and Cyberonics.

SYNOPSIS: Olfaction and taste appear to be localized in the insular cortex with bilateral representation.

SOURCE: Mazzola L, Royet JP, Catenoix H, et al. Gustatory and olfactory responses to stimulation of the human insula. Ann Neurol 2017; Aug 10. doi: 10.1002/ana.25010. [Epub ahead of print].

Very specific regions of our brains process the sensations of taste and smell. Animal studies, neuroimaging, metabolic studies (such as PET), and an analysis of cerebral lesions in humans have shown that these sensations are processed bilaterally in the parietofrontal operculum and insula. However, there has been some debate as to which part of the insula is responsible for this important brain function. There are many reasons for this, including the fact that it is very difficult to test one of these sensations without activating another. For instance, noxious stimuli also will activate the trigeminal pathways. A small amount of liquid that is placed in the mouth will go into the nasopharynx, stimulating smell. Direct cortical stimulation eliminates these confounding factors and is likely to be a much more accurate way of identifying the regions of the brain responsible for processing the sensations of taste and smell. Mazzola et al, working at the University Hospital of Saint-Etienne (France), reported their observations of taste and smell sensations using direct electrical stimulation of the insular cortex.

The investigators stimulated the human brain, using stereotactically placed electrodes, which had been placed as part of an evaluation for epilepsy surgery and had two main functions. First, they were used to identify the region of seizure onset in these patients in preparation for surgical removal of the seizure-causing zone. Second, the electrodes were used to stimulate the cortex, thereby identifying regions of eloquent cortex that would need to be spared during the epilepsy surgery.

Mazzola et al identified 221 patients who had insular cortex stimulation at the University Hospital between March 1997 and January 2015. All were undergoing an evaluation for epilepsy surgery. Confirmation of electrode placement occurred by either MRI (after 2009) or by superimposing a radiograph over the patient’s MRI images. The study included 107 women and 114 men, with an average age of 35 years. The investigators used a standard method of cortical stimulation (50 Hz square-wave impulses, pulse duration 0.5 milliseconds, train duration < 5 seconds, intensity 0.2 mA to 3.5 mA). There were 651 electrical stimulations of the insular cortex in this group of patients, and 550 produced a clinical response. The most frequent responses, which occurred in 61% of the stimuli, were somatosensory (pain, thermal sensations, non-painful somatic sensations). Auditory responses occurred in 8%, followed by vestibular in 7%. The taste and smell responses were infrequent, accounting for 2.7% and 1.1%, respectively.

Both the taste and smell responses occurred when stimulating areas in the middorsal region of the insula. The taste (gustatory) responses occurred in both hemispheres equally (3.0% on the right; 2.6% left). Most of the subjective taste sensations were unpleasant. Most people could not identify the taste, just that is was “nasty,” “acidic,” “metallic,” or “salty.” The smells also occurred when stimulating either hemisphere, and were most often described as “metallic” or “chlorine.” Many patients had trouble identifying whether the sensation was a taste or a smell. Interestingly, the location of cortical stimulation for taste and smell overlapped.


It has long been observed that smell, taste, and oropharyngeal sensation are closely associated brain functions. By stimulating various regions of the human insula, Mazzola et al showed that these sensory modalities were closely related to each other anatomically. In fact, they further demonstrated that taste and smell are brain functions that overlap in the insula. Although we think of taste and smell as separate senses, this observation may explain why it is often difficult to distinguish one from the other. Further, oropharyngeal sensations and unusual or unpleasant tastes and smells often are part of a person’s seizure semiology. When patients report these sensations during their seizures, Mazzola’s study results indicated that the insula should be considered as a place where the seizure has started or to which it has propagated.