Ketamine and the brain’s anti-reward system

Ketamine turns down the brain’s response to things we expect to be unpleasant.

Most antidepressants take weeks before depression starts to improve, which delays effective treatment and complicates finding the right treatment for each individual. However, the drug ketamine has been found to work more quickly in depression, with effects emerging over a few hours, when administered at sub-anaesthetic dosage. This effect was discovered incidentally, and we don’t know why this drug is able to reduce depression so quickly. This would be useful to know in order to pinpoint other potential treatments which also have this fast acting profile in depression.

In our study, we showed that ketamine turns down the brain’s response to things we expect to be unpleasant. We have found evidence of this effect in a small but important brain region called the habenula, which is normally involved in processing disappointment, punishment and other negative experiences not only in humans but also other mammals. More interestingly, our findings suggest that changing activity in this region may also influence how negative experiences are remembered and subsequently guide our choices.

The habenula has attracted considerable attention in depression research. It is sometimes described as part of the brain’s “anti-reward” system. It becomes active when things go wrong, when we anticipate something unpleasant, or when we fail to receive something rewarding and need to learn from these experiences. Previous studies in rodents had shown that excessive activity in the lateral habenula is associated with depression-like behaviour, and that ketamine can suppress this activity rapidly. An outstanding question was whether something similar happens in the human brain.

We gave healthy volunteers either ketamine or a placebo and, 24 hours later, asked them to learn associations between abstract images and different outcomes. Some images predicted winning money, others losing money, receiving an electric shock, or experiencing nothing. While participants learned these associations, we measured their brain activity using an ultra-high-resolution 7-tesla MRI scanner.

People who received ketamine showed reduced activity in the habenula when anticipating negative outcomes. We also found reduced responses in this region when negative outcomes were delivered, relative to positive ones.

There was another intriguing finding. Participants with lower habenula activity during learning were more likely to subsequently prefer images associated with electric pain. This relationship suggested a possible connection between brain activity during learning and how people remember negative memories later on.

Our study suggests that ketamine works similarly in the human brain as earlier animal work has shown. Our study was randomised and placebo-controlled, and we used 7-tesla imaging to examine a particularly small brain structure with greater precision than conventional fMRI allows. However, our sample was modest, and all participants were healthy volunteers.

The important next step is to establish whether the same mechanism operates in people with depression, and whether changes in habenula activity relate to actual improvements in symptoms. Although we cannot claim that we have solved ketamine’s antidepressant mechanism, we took an important step in finding a key piece of the puzzle and provided human evidence for a mechanism previously suggested only by animal research.

Ongoing thread. More from Erdem Pulcu to follow.