Learned placebo responses in neuroendocrine and immune functions - Wendt et al. - 2014 - Article

How do the central nervous system and peripheral immune system communicate?

Stress induces several humoral and cellular immune responses and influences disease outcome. Opposite to this, immune responses influence mood and behaviour. The brain and peripheral immune system are mainly connected by the sympathetic nervous system as an efferent neural pathway. The immune system is also affected by the central nervous system (CNS) through the humoral pathway of the hypothalamus-pituitary-adrenal (HPA) axis. Lastly, an efferent cholinergic pathway mediated by the vagus nerve influences the immune system and proinflammatory responses.

The immune system, in turn, informs the CNS about the peripheral immune status. This happens via the vagus nerve, but also through an efferent humoral pathway, which uses neurotransmitters to transfer messages to the CNS by crossing the blood-brain barrier or via circumventricular organs. These messages have to be translated into neuronal signals in the afferent neural pathways.

What happens in behavioural conditioning?

Classical conditioning means that a property of an unconditioned stimulus (US) leads to a certain response. This response is later also elicited by a conditioned stimulus (CS), which does not possess this property. The CS predicts the occurrence of the US, which later results in a different immune response. So, in classical conditioning we learn about the temporal relationship between stimuli in order to prepare for important biological events. This is done in three steps:

1. In the acquisition or learning phase the CNS notices the US directly or indirectly because of immune response changes.
2. Signals caused by the US and sensory information provided by the CS are integrated and associated by the CNS.
3. In the evocation or memory phase the immune response is changed via efferent pathways. This happens when the brain regions, that integrated the CS/US association, are activated by re-exposure to the CS.

Which brain areas are involved in behavioural conditioning?

Regarding the afferent pathways, the insular cortex is important in the acquisition and evocation phase of behavioural conditioning. The amygdala is important in acquisition, because it mediates the input of visceral information. The ventro-medial hypothalamic nucleus is important in the output pathway to the immune system, leading to a conditioned immune response. The sympathetic nervous system is an important efferent neural pathway between the CNS and peripheral immune system. The conditioned immune response is also mediated via the splenic nerve, but there appear to be many more efferent neural routes.

Can conditioning be therapeutically used?

Studies find that the outcome of a disease is influenced by conditioned immune responses. These conditioned responses might decrease the magnitude of the disease and its mortality rate. Research in mice found that conditioned immune responses influence mast cell functions. Conditioning can be used as supportive therapy, for instance, to reduce tumour growth and prolong survival time in mice.

In humans, it also seems possible to create learning-induced placebo responses. For instance, an olfactory cue (CS) which was formerly paired with an allergen challenge also led to an increased histamine release. In chemotherapy, research found that formerly neutral stimuli (CS) are later able to induce certain side effects, as they were present during the infusion of medication (US). Studies state that behavioural conditioning can thus be used in treatment supportive to pharmacological therapies. Besides effects on the immune system, learned placebo responses also affect neuroendocrine functions.

What are the clinical implications of the learned placebo response?

Learned placebo responses could be used to reduce the amount of medication, leading to less side effects and maximised therapeutic outcome. However, we first need to learn more about reproducibility, predictability and extinction of learned placebo responses. For instance, we know that re-exposing someone once to the CS is not enough to cause an immunosuppressive response. The conditioned response will be stronger, when someone is exposed to the CS multiple times.

Furthermore, every individual is different in their ability to develop conditioned immune responses. We should identify which factors cause a conditioned immune response to occur. In addition, for a learned placebo response to be clinically implementable, the response should not be evoked by only a single event, but should be retrievable and able to be evoked in several ways. However, if this is not the case, we could still use it to examine the bidirectional communication between the CNS and peripheral immune system.

Where should future studies focus on?

A conditioned response will extinct when multiple re-exposures to the CS are not reinforced. However, we don’t know yet which processes are involved in extinction learning from conditioned immune responses. Knowing how to modify or control extinction of these responses could be extremely helpful in treatments. Future research should investigate how the extinction process can be overcome.