A great article published in diagnostics by Emmanuel Bäckryd highlights the need for a good mechanism based diagnosis for clinical pain medicine.

Everyone understands that pain is unpleasant. Pain is nonetheless necessary information: humans need to know when they are doing things that will injure their body. People who lack the ability to feel pain are at very high risk for injury or death. The sad case I member reading about was a little girl who was born without the ability to feel pain. When her eye itched, she rubbed her eye until she went blind. She lacked the painful feedback to warn her that what she was doing was damaging her eye.

The trouble is that in some cases pain persists even when there is no injurious condition. This kind of pain does not help people avoid injury. In fact, this kind of pain may actually increase injury. Because of chronic, uninformative pain, people avoid healthy activities like physical therapy or exercise. In some cases, pain may interfere with hygiene and self-care. These are expensive problems.

Additionally, many forms of pain are not well treated by opiates. Patients seeking care for chronic, unmanageable pain may end up on high doses of opiates to compensate for their low efficacy. I suspect that this significantly contributes to the opiate epidemic in the United States. Pharmaceutical marketing campaigns have led doctors to feel that they can prescribe opiates for chronic pain. According to Frontline, MDs were reluctant to do so until fairly recently. Prescribing opiates for chronic pain is a fine idea when they work. When opiates are available, but don’t work, it is a gateway to addiction.

If we had a good blood test to determine if a person would respond well to opiates (or any other kind of pain medication, for that matter) we could guide treatment more effectively. A blood test for the existence of pain seems unlikely (pain is a subjective experience). But I think that guiding treatment by biochemistry instead of by subjective assessment is usually a good idea.

The review discusses several interesting candidates that have failed as biomarkers for pain. Cystatin C was a potential biomarker identified in animal studies. A study in humans did not confirm biomarker as correlating with acute labor pain. Bäckryd notes that their group did not detect a significant differences between patients and controls in any of the other potential biomarkers listed (e.g. neurotrophic factors, cytokines, neuropeptides, beta endorphins, and substance P). And of course, none of these biomarkers are listed in the common blood panels used by pain management physicians. Perhaps no single protein is adequate to give useful information to guide pain treatment.

Ultimately, Bäckryd confirms my view that the field of biomarkers for pain conditions is still young. He also shares my view that it would be good to have a blood test to help guide the prescription of pain medicine or disease modifying drugs. He brings up an interesting point of skepticism: if pain is sensitized or mediated by central nervous system conditions, it’s not clear that the blood brain barrier would allow for biomarkers to enter the blood. In such a case, it might be more reasonable to look for degradation products or other downstream effects. Undoubtedly, this would complicate diagnosis.