As neurosurgeons, many of the patients we encounter suffer from pain. As such it can play a central role in our decision making process. For example, the judgment to perform spine surgery is frequently dependent on the patient’s narrative of their pain. Furthermore, neurosurgeons commonly encounter clinical scenarios in which, patients are unable to report pain due to head trauma or sedation. In these situations we cannot determine with any certainty whether we have treated their pain either inadequately or excessively. In the past, pain could only be assessed as a subjective report given by patients. However, self-reporting has obvious limitations and may be affected by secondary gain. The unreliability of such an important variable in our decision making process may have far reaching effects on the outcome of our interventions.
In an article published in the April 2013 edition of the New England Journal of Medicine, Wager and colleagues attempted to determine brain signatures to quantify pain objectively using modern functional magnetic resonance imaging (fMRI) technology.1 The authors conducted several different studies. The first identified the brain regions (neurological signature) activated by thermal pain applied to the forearm, as the dorsal posterior insula, the secondary somatosensory cortex the anterior insula, the ventrolateral and medial thalamus, the hypothalamus and the dorsal anterior cingulate cortex (Figure). In the second study, activation in these brain regions by non-painful warmth was compared to painful heat. The third study investigated the brains’ activity during somatic stimulus as compared with social/emotional stimulations. The final branch of the study investigated the effect of remifentanil on brain activity during both social and heat-induced pain.
In contrast to the possibility that different individuals manifest pain in different brain regions, the authors noticed a universal pattern of pain activation (neurological signature) across individuals that could be used to detect pain objectively across all study participants. fMRI scans were able to discriminate between painful heat and non-painful warmth with a sensitivity and specificity of 94%. More importantly, fMRI was able to discriminate between relative differences in pain with a sensitivity and specificity of 93%, with pain ratings differing by 2 or more points on the 10-point Visual Analogue Scale. As expected from wide-dynamic-range neurons and nociceptive-specific neurons, signal intensity increased on a continuum from warmth to painful heat stimulation. Interestingly, the signal strength correlation with the reported level of pain was r = 0.73 compared to r = 0.65 with temperature change, indicating the strength of neural signature holds a stronger correlation to subjective pain perception than to the intensity of the thermal stimulation. Remifentanil infusion was associated with a 53% reduction in the signature response (P = .94). This shows that fMRI signals may be clinically useful for monitoring the adequacy of pain control.
This benchmark paper expanded our understanding of acute pain in healthy individuals and provided an important method to visualize and quantify pain objectively using brain imaging. However, we need to remember that the pain signature described in this study were the result of moderate thermal pain applied to the forearm of young healthy volunteers. Severe pain, pain in other locations, pathological pain, or chronic pain may all produce variations in pain signature patterns. Additionally, modulating factors such as anxiety, fatigue, and depression were not studied. Specifically, central pain processing in chronic pain states from certain diseases may differ significantly from the acute pain in healthy individuals. The authors rightly pointed out that further investigation of different anatomical locations and pain modalities is needed to assess the generalizability of this method before it can be applied to clinical practice.
The publication of this paper will undoubtedly trigger additional fMRI studies for various pain conditions in the near future. For spine surgeons, it would be very interesting to see whether fMRI can reliably detect certain pain signatures in chronic lower back pain caused by common pathologies such as spinal stenosis, nerve roots compression and/or pain caused by mechanical spine instability (the most common area of severe chronic pain in the US population). Time will tell whether fMRI can be used by neurosurgeons as a reliable and clinically applicable tool for objective pain measurement. If this is possible, its application will have a momentous impact on our clinical decision-making process and ability to optimize treatments to assure the best therapeutic outcome.
1. Wager TD, Atlas LY, Lindquist MA, Roy M, Woo CW, Kross E. An fMRI-based neurologic signature of physical pain. N Engl J Med. 2013;368(15):1388–1397.