Why is this topic important?
Formal recognition of the human ability to modify our own pain experience arose in the early 19th century, anteceding most pain relief medications and surgical treatments that remain widely accepted today. Despite its long recognition throughout clinical medicine and a variety of experimental domains, we still have yet to fully untangle the methods by which the human brain translates expectations and beliefs into pain relief.
Indeed, we have come a long way from Perkin’s tractors and animal magnetism, and through ever advancing brain imaging technologies we have a tight grasp of the important nodes in the cortical and subcortical pain systems which are involved in placebo analgesia. The prefrontal and cingulate cortices, the midbrain periaqueductal gray matter (PAG), and the rostral ventromedial medulla (RVM) all emerge consistently throughout the pain modulation field regardless of the elected placebo or method of induction, and provide a strong schematic for which research can be based on. However, what was lacking at the time of conducting our study was methodologies which provided direct support of top-down modulation – that is, instead of a map of harbours around the world, we set out to identify the exact routes that ships take between them, and whether they go only from point A to B, or freely conduct trips between ports.
What did we do?
We recruited 47 healthy participants to engage in a well-established response conditioning method for inducing placebo analgesia involving deceptively tweaking thermal pain applied to two adjacent creams – a control vaseline cream and placebo cream participants believed to contain the topical analgesic lidocaine. All participants then partook in a high field (7-Tesla) MRI scanning session where both creams received identical intensity stimuli to trigger the phenomenon. In 23 participants (49%) we observed significant placebo analgesia, and we then investigated these participants scans relative to those that did not express the phenomena to determine connections behaving differently between our two groups. Using the PAG as a seed region, Pain dependent (psycho-physiological interaction) and -independent (functional connectivity) analyses were conducted, as well as Dynamic Causal Modelling and mediation to pinpoint the most integral connections between brain regions, as well as which direction these connectivity signals were being conveyed in, responsible for the manifestation of significant placebo analgesia.
What did we find out?
By considering the placebo analgesia phenomena as one which does not only cause changes in brain systems strictly during the perception of pain, we were able to classify two distinct brain networks – one stimulus-dependent and the other -independent which we believe work in duality to drive the phenomenon. Both of these brain networks appear to centre on top-down contact of the lateral PAG column, a subregion we had previously identified as critical in a brainstem-focussed investigation of brainstem activity patterns during placebo analgesia and nocebo hyperalgesia.
In this investigation however, the first network, which contacted the PAG strictly during periods where painful stimuli were applied comprised of the cingulate and insular cortices, as well as the nucleus accumbens (NAc) – a region long described as a dopaminergic hub involved in driving error-predictions. Indeed, dynamic causal modelling revealed the most important connections in this network were from the NAc to the cingulate, and from the cingulate down to the PAG – suggesting an active process of correcting incoming noxious information to match expected and believed intensities.
It was in our second network identified however that our results become particularly interesting. Regions of the stimulus-independent network, those which altered their contact with the PAG outside periods of pain stood in striking contrast to the first, and instead comprised of subcortical and limbic sites such as the hypothalamus and central amygdaloid nucleus, which all decreased in their PAG coupling in those showing significant placebo analgesia. The location of individual nodes in this network overlapped entirely with known autonomic and pain-regulatory circuitry identified in preclinical studies over 30 years ago, with an overarching function in animals of maintaining PAG sensitivity to enable appropriate behavioural responses to external stimuli. To the best of our knowledge, this investigation represents the first time this same circuit in its entirety has been revealed in humans, with our results directly suggesting a role for this network in facilitating dis-inhibition of the PAG to allow the active network to drive changes in top-down pain modulatory systems. Specifically, it was coupling from the left and right posterior hypothalamic nuclei to the PAG that displayed significant differences between responders and non-responders, suggesting these specific connections as most important in establishing an optimal excitatory tone of the PAG for receiving active pain modulatory signals.
What makes our findings important?
It is rare for a manuscript to make inferences on an encompassing neural network responsible for any given pathology or phenomenon. Indeed, we make no claim that our work here does that, but rather we reinforce known signalling pathways – ie. From the rACC to the PAG – and further shed light on some of the less recognised pain modulatory circuits in the human brain – ie. From the hypothalamus to the PAG - evolutionarily conserved from animals which may potentiate or assist in driving top down pain modulatory effects. Acknowledgement of their presence and role in a phenomena like placebo analgesia paves the way for the investigation and scrutiny of these connections in alternative phenomena or chronic pain conditions, which may further explain the well-recognised disruptions in pain pathway signalling known to underlie difference pathologies / manifest pain modulatory effects.