This is Part 3 of a four-part report on selected talks from the 10th IASP Research Symposium, The Genetics of Pain: Science, Medicine and Drug Development, held 7-9 February 2012 in Miami, Florida, US. See also Part 1, Part 2, and Part 4, or download a PDF of the entire report.
In a session on pain assessment, several presenters took up a critical issue in studies looking to relate genes to pain, and that is defining and validating phenotypes. “Poor phenotypes are one of the main reasons genetic studies fail,” said Christopher Sivert Nielsen, Norwegian Institute of Public Health, Oslo, Norway.
What makes a good phenotype? Nielsen said the measure must be reliable, valid, stable over time, heritable, genetically homogeneous, and consistently used across studies. The last two are particular challenges for the pain field, where it is not clear if different types of chronic pain (idiopathic, neuropathic, or nociceptive) have common genetic bases, although recent data from a twin study indicate that pain reporting at different body sites can be explained by a single genetic factor (Williams et al., 2010). In addition, the definition of phenotypes across studies is very inconsistent, making it hard to compare results.
Can experimental pain measures stand in as endophenotypes of chronic pain? While all experimental pain measures are heritable, they are not all equally so. Twin studies show that most clinical pain conditions have a heritability of 40-50 percent, while experimental measures of pain had varying heritability from 22-55 percent in one study (Norbury et al., 2007). In addition, not all experimental pain measures are created equal. Several studies have shown that sensitivity to different painful stimuli in an individual is not consistent across modalities (heat, pressure, cold), suggesting different underlying genetic determinants for each response (Nielsen et al., 2008; Neziri et al., 2011; Lariviere et al., 2002).
That leaves the question of which measures correlate best with clinical pain. Nielsen argued for the cold pressor test, in which subjects immerse their hands in ice water to determine cold pain threshold and tolerance. He presented preliminary data from a study of 10,566 subjects in a population-based sample drawn from citizens of Tromsø, Norway. In the study, subjects with chronic pain (defined as pain lasting three months or longer) were more sensitive in the cold pressor test, and those with longer duration of pain (more than five years) were more sensitive than those with shorter duration. The frequency of pain made little difference, but intensity did—people who experienced moderate or strong pain were more sensitive in the cold pressor test. People who had pain at two or more body sites showed more sensitivity. The results suggest that increased sensitivity to cold pain does mark a population of people with strong, long-lasting pain and widespread pain. In contrast, heat pain threshold (tested in 4,094 people) and pressure pain (n = 4,807) did not show significant correlations with clinical pain. Repeat measures of heat pain and the cold pressor test in a small number of subjects suggested that responses to heat were less stable over time. Nielsen concluded that the cold pressor test may be a better marker for a general pain condition than other measures. Overall, he said, phenotype selection should be based on empirical evidence of measurement characteristics, relationship with clinical pain, and genetic factor structure from twin or family studies.
According to Roger Fillingim, University of Florida, Gainesville, US, the genome is dwarfed by the phenome. “We can train a chimpanzee to do genotyping, but you have to be really smart to do phenotyping,” he said. Like Nielsen, he sees quantitative sensory testing (QST) measures as a source of endophenotypes that may provide increased power to identify genetic interactions. But, in the biopsychosocial model of pain, biological factors, genetic and otherwise, interact with psychological and social factors to produce the ultimate outcome of pain and disability. Other factors also influence genetic effects, including sex, age, and ethnicity. That means that researchers need to model these interactions as well, and remember that psychological, as well as biological, factors fall under genetic control.
How to deal with the complexity? Ignore it, omit it by design, or anticipate and model it, Fillingham said.
Defining phenotypes depends on standardized descriptions of pain. Researchers in the field of spinal cord injury (SCI) have made progress on that front with a recently published pain taxonomy (Bryce et al., 2011), said Eva Widerstrom-Noga, University of Miami, US. Pain after spinal cord injury, consisting of neuropathic and other types of pain, is very common, and is one of the most significant factors that lead to reduced quality of life. The consensus taxonomy covers all the pain that patients experience, whether resulting directly from their injury or not, to allow clinicians and clinical researchers to classify uniformly and to study patient-reported pain after SCI. Widerstrom-Noga showed how both pain reports and QST can be used to define patient subtypes in SCI. She also presented data using magnetic resonance spectrometry to look for changes in brain biochemistry associated with pain. She reported preliminary results in a small group of patients suggesting that SCI patients with severe pain that highly impacted their lives in a negative way showed changes in brain metabolites in the anterior cingulate cortex, consistent with a role for that region in the experience of pain and distress.
One model of chronic pain that is considered the most promising for genetic studies is persistent post-surgical pain, because studies can be done prospectively, before subjects have chronic pain, and subjects are plentiful. In his talk, surgeon Henrik Kehlet, Copenhagen University, Denmark, reviewed work on the factors that contribute to persistent pain after hernia operation. In a prospective study, he identified both patient-related and surgery-related factors. Both are important: Surgical techniques that cause more or less nerve damage can result in higher or lower rates of post-operative pain. But an identical technique can present a different risk to individual patients, depending on patient factors including pre-operative responsiveness to pain (Aasvang et al., 2010). The take-home message, Kehlet said, is that all the details—pre- and post-surgical, patient-specific and procedure-specific—matter to pain outcome, and thus to genetic studies using outcome as a phenotype. “This setup can be used [for genetic studies],” he said, “but you have to take into account many factors.”