Pain processing and pain perception in the human brain: a neural signals approach

Pain is an almost universal experience. Its importance for health, well-being and disease processes is undeniable. One can experience more or less pain, and measuring pain intensity is important in diagnosis, assessment and therapy. However, this measurement is surprisingly difficult, because pain levels depend on complex factors such as expectation, attention, and even personality. Moreover, robust, valid measures of pain intensity are elusive. Current methods, such as numerical pain rating scales, rely strongly on patients' verbal understanding - e.g., the phrase "worst pain imaginable" is generally used to define the top of the scale. Perhaps for these reasons, ratings show notoriously poor correlation with the energy of a noxious stimulus, and also with activation of putative 'pain centres' or 'pain networks' in the brain.

This project develops a novel approach to understanding pain perception. Using a state-of-the-art feedback-controlled laser stimulator, we expose the skin on the back of the hand of healthy volunteers to brief, quantified pulses of radiant heat, causing a sharp 'pinprick' pain. This reflects signals in a specific neurophysiological pain pathway ("A-delta pathway"). Participants are stimulated randomly with either of two energies, both above the individual's threshold to activate the A-delta pathway. They judge whether they experienced 'higher' or 'lower' intensity. We use standard methods to calculate two distinct aspects of pain perception: the SENSITIVITY with which a person's judgements track the actual stimulation energy, and their BIAS in preferring to respond 'higher', or 'lower', irrespective of actual stimulation energy. For example, someone who simply fears strong pain might judge ALL pain stimuli as 'higher' intensity: they would then show strong bias and low sensitivity.

We also record the brain activity evoked by laser stimulation, using EEG electrodes non-invasively placed on the scalp. By dividing the distribution of EEG amplitudes in two at different cutoff points, we can simulate how effective the EEG signal at that particular moment would be in classifying the higher and lower intensity stimuli. We will thus identify the components of the brain's response to noxious stimulation that best approximate the sensitivity and bias components of pain perception. This provides a novel, well-controlled, physiologically-specific, and fully objective method to search for the neural basis of pain perception. It does not require any subjective definition or instruction regarding what counts as 'pain', and merely requires participants to judge whether a stimulus is more or less intense. By testing a large sample of participants, we address individual differences between people in pain perception. We may identify psychological profiles associated with high and low pain bias, and with high and low pain sensitivity.

Two further large experiments leverage these important methodological developments. Experiment 2 aims to show how our method can be used to quantify and understand the potency of a popular analgesic drug (remifentanil). Double-blind infusions of drug or placebo are compared to address whether the drug influences bias or sensitivity, and which EEG components underlie this change.

Experiment 3 uses these new measurement approaches to investigate the basis of the well-known placebo effect - perhaps the most striking finding in pain research. Participants learn that a specific visual stimulus, presented before laser stimulation, predicts whether the next stimulus will be more or less intense. Again, we assess whether this predictive cue changes sensitivity or bias. We will identify the EEG components that are responsible for how the cue influences these two aspects of pain perception.

Overall, the results provide a novel and objective approach to pain perception, useful for future clinical studies of pain and analgesia.

Evidence-based medicine requires rigorous methods to measure pain experience and its neural bases, yet these methods remain elusive. Noxious laser heat conveyed by the Adelta pathway provides a key experimental model of human pain.

We focus on perceived intensity of 'pinprick' laser pain, identified by reaction times as Adelta-mediated. Noxious stimuli are presented at two preselected levels within the Adelta range. Volunteers report each stimulus as 'higher' or 'lower' intensity. Signal detection measures of d' (sensitivity) and C (bias) provide key indices of pain processing WITHIN the Adelta pathway.

Different features of the EEG responses to nociceptive stimuli (e.g., N1, N2, P2 waves, gamma power) are extracted. Bisecting the amplitude distribution for each component is equivalent to a 'higher/lower' perceptual judgement, defining NEUROMETRIC d' and C. By varying cutoff values, we identify WHICH neural component best approximates perceptual sensitivity and bias. This produces a highly-controlled, pathway-specific model of pain intensity, able to distinguish bias from sensitivity at both perceptual and neural levels.

Experiments 2 and 3 put the method to work. Infusions of remifentanil, or saline are given under clinical supervision in Expt 2. We investigate whether analgesia involves changes in perceptual sensitivity or bias, and associated EEG neurometric changes. Experiment 3 extends the method to placebo and nocebo conditioning. Visual cues signal impending lower or higher intensity, while neutral cues have no predictive value. We assess whether placebo/nocebo cues change perceptual and neurometric sensitivity, relative to neutral cues, or rather change biases. Neural events after the cue but before the laser could also influence pain perception. We thus investigate the interesting hypotheses that reduced pain perception under placebo is based on DIFFERENT neural signals, and not simply reduced levels of 'normal' nociceptive signals.

Impact and Exploitation:
Robust pain measurement protocols, and methods for establishing neurometric-perceptual equivalence may interest pharmaceutical and other life-science companies. Potential commercial exploitation will be discussed with UCL Business PLC (Dr Abigail Watts), and UCL's dedicated Translational Research Office (https://www.ucl.ac.uk/translational-research/). The suggest relevant industrial contacts, and provide appropriate dedicated funding streams for pump-priming of translational work with qualified partners.

Dissemination:
A long-term dissemination target of our work is chronic pain. Experimental pain provides a rigorous scientific model of pain perception and its underlying neural mechanisms, but our results will have increased societal and economic impact if they can be translated to the chronic pain state, which is a complex and costly concern in the NHS. After completion of experiment 1, we will make contact with experts in central chronic pain, with a view to developing an additional, future proposal for further project funding. One particular focus might be experimental pain testing on unaffected body parts of chronic pain patients, versus healthy controls, in order to identify changes in sensitivity and bias caused by chronic pain. A preliminary interest in collaboration has been expressed by Dr S Chong.

Wider audiences:
The dissemination of our research outcomes to general public will include a number of channels, such as giving talks in Café Scientifiques (events organised by The Royal Society), participating in educational activities (e.g. teaching in primary schools) and interviews in major UK and international media (e.g. BBC TV and radio, newspapers, etc). Both PIs have a track record in these avenues of disseminating results to the general public (e.g., TED talks, BBC TV and radio coverage; please see "Pathways to impact" for details). The dissemination of results to the general public is facilitated by Dr Harry Dayantis of the UCL media relations team.