Patients with spinal cord injury (SCI) may experience several types of chronic pain, including peripheral and central neuropathic pain, pain secondary to overuse, painful muscle spasms, and visceral pain. An accurate classification of the patient's pain is important for choosing the optimal treatment strategy. In particular, neuropathic pain appears to be persistent despite various treatment attempts. In recent years, we have gained increasing knowledge of SCI pain mechanisms from experimental models and clinical studies. Nevertheless, treatment remains difficult and inadequate. In line with the recommendations for peripheral neuropathic pain, evidence from randomized controlled treatment trials suggests that tricyclic antidepressants and pregabalin are first-line treatments. This review highlights the diagnosis and classification of SCI pain and recent improvements in the understanding of underlying mechanisms, and provides an update on treatment of SCI pain.

Pain is a frequent consequence of spinal cord injury (SCI) which may profoundly impair the patients' quality of life. Valid experimental models and methods are therefore desirable in the search for better treatments. Usually, experimental pain assays depend on stimulus-evoked withdrawal responses; however, this spinal-mediated reflex response may be particularly problematic when evaluating below-level SCI pain due to the development of hyperactive reflex circuitries. In this study, we applied and compared assays measuring cold (acetone), static (von Frey filaments), and dynamic mechanical (soft brush) hypersensitivity at different levels of the neuroaxis at and below the level of injury in a rat model of SCI. We induced an experimental SCI (MASCIS 25 mm weight-drop) and evaluated the development of spinal reflexes (withdrawal), spinal-brainstem-spinal reflexes (licking, guarding, struggling, vocalizing, jumping, and biting) and cerebral-dependent behavior (place escape/avoidance paradigm (PEAP)). We demonstrated increased brainstem reflexes and cerebrally mediated aversive reactions to stimuli applied at the level of SCI, suggesting development of at-level evoked pain behavior. Furthermore, stimulation below-level increased innate reflex responses without increasing brainstem reflexes or aversive behavior in the PEAP, suggesting development of the spasticity syndrome rather than pain-like behavior. While spinal reflex measures are acceptable for studying changes in the spinal reflex pathways and spinal cord, they are not suited as nociceptive behavioral measures. Measuring brainstem organized responses eliminates the bias associated with the spastic syndrome, but pain requires cortical involvement. Methods depending on cortical structures, as the PEAP, are therefore optimal endpoints in animal models of central pain.