After surgical interventions, of skin, bone, or soft tissues, patients require analgesic therapy because they experience ongoing pain or are sensitive to incident, normally nonpainful stimulation. The intensity and duration of postsurgical pain vary, but with uncomplicated wound healing, the pain typically shows progressive attenuation. However, some patients experience deep pain or pain referred to the dermatomes that correspond to the operated organ, which persists for months or even years.1The nature and properties of persistent postsurgical pain are poorly characterized. Descriptions of postsurgical pain often do not differentiate between spontaneous pain and pain that is dependent on external stimulation, for example, clothing that touches the skin near the scar. We do not know whether there is a distinct transition period between acute and chronic pain or whether the chronic condition constitutes merely an extension of perioperative pain. Are the mechanisms responsible for sustained pain the same as those underlying acute postsurgical pain, or does, in this subgroup of patients, the trauma associated with the surgical intervention provoke different changes in sensory processing?

Surgery causes the release of inflammatory mediators, such as prostaglandins and cytokines,2that activate and sensitize primary sensory afferents. Peripheral sensitization and facilitation of synaptic transmission in the central nervous system normally resolve over time and afferent input is likely to decrease during wound healing. However, in some cases, inflammation may persist. For example, tension-reduced hernia repair with a prosthetic mesh generates a marked inflammatory response around the synthetic material that outlasts the healing of the superficial wound.3Surgical interventions such as thoracotomy or bunionectomy are also often accompanied by varying degrees of stretch or severing of peripheral nerves, risking the development of neuropathic pain. Sustained ectopic activity of primary sensory afferents and changes in spinal inhibition and facilitation may result. Persistent inflammation and nerve injury may lead to changes in gene expression, a phenotypical shift of primary sensory neurons, and activation of peripheral immune and glial cells, all contributing to lasting pain. Combinations of inflammatory and neuropathic mechanisms are likely to occur, for example, after removal of a herniated intervertebral disk or visceral surgery.

To develop strategies for the prevention and treatment of persistent pain after surgery, we need preclinical models that replicate the complexity of the human condition. Only valid surrogate models of pain will provide the mechanistic insight necessary to identify risk factors for chronic pain, improve their management, and compare the efficacy of symptomatic versus  disease-modifying therapies. Most preclinical studies rely on rodent models. They have been created using interventions similar to surgery in humans, for example, incisions of skin and muscles.4In general, these models are used to examine the consequences of wounding on primary afferent activity and characterize rapid changes in signal transduction, including peripheral sensitization of nociceptors. Incision models are useful for determining the efficacy of pharmacologic treatment during the early postsurgical phase. Although not typically associated with long-term changes, incisional injury indeed predisposes both neonate5and adult animals6to enhanced pain sensitivity when a second injury is applied several weeks later.

Subcutaneous, intramuscular, or intraarticular injections of irritants such as formalin or complete Freund's adjuvant induce inflammation associated with transient pain-related behavior (present for minutes to days).7Models of inflammatory pain are therefore used to study the short-term consequences of surgery on sensory function and pain. Longer-lasting behavioral changes (persisting for weeks to months) are observed in models of nerve injury, including partial or complete ligation, transection, or neurectomy of cranial (trigeminal), peripheral or spinal nerves, or nerve roots. Nerve injury provokes neural and immune changes that differ from those observed in models of acute postsurgical pain or inflammation; in fact, exposure of a nerve by skin and muscle incision is a standard control procedure for experimental nerve lesions. Redistribution of the voltage-gated sodium channel subunit Nav1.8, which is predominantly expressed in nociceptors, and microglial activation in the ipsilateral dorsal horn of the spinal cord are examples of nerve injury-induced changes that contribute to the development of neuropathic pain but seem not to be involved in the induction of pain-like behavior after skin and muscle incision.8,9Nerve injury models are valuable tools for studying the consequences of surgical interventions that carry a risk for intraoperative nerve lesion. Such nerve lesions are likely to promote the development of persistent pain because they provoke long-term changes, for example, altered gene expression in the dorsal root ganglia and the spinal cord, central sensitization, and transsynaptic neurodegeneration.

Preclinical models of postsurgical pain, whether acute or persisting, have face validity because they produce increased withdrawal responses to calibrated mechanical or thermal stimulation that are reminiscent of pain elicited by normally nonpainful stimuli (allodynia) and increased sensitivity to painful stimuli (hyperalgesia), both common features of postoperative pain in humans. Assessing spontaneous pain in rats or mice, the two species most frequently used, is difficult, but presumptive parameters of spontaneous pain such as changes in exploratory behavior, vocalization, and conditioned place preference should be considered in the evaluation of animal models.10,11Veterinarians are routinely required to monitor pain in their patients after surgical interventions that are no less complex than those performed on humans. Reports of postsurgical pain in, for example, dogs or horses and guidelines developed for pain assessment in these species12are valuable yet underutilized resources for improving the design of preclinical research on pain after surgery.

First, studies on the prevention of persistent pain need to be distinguished from those focusing on preemptive or short-term perioperative analgesia. Preclinical models involving long-term neurobiologic and behavioral changes must reflect a time course that exceeds the duration of wound healing. Although we hesitate to specify an interval, 14–21 days after surgery would seem to be minimal. Second, analgesic treatment applied long enough to prevent the transition from acute to chronic pain—if the two processes can be separated and are not independent of each other—needs to be distinguished from disease-modifying interventions aimed to block, for example, central sensitization or the immune response to nerve injury. Third, timing of intervention is crucial. Drugs delivered before and early after surgery are preemptive if they reduce perioperative pain, and preventive if they block the development of sustained pain. Treatment initiated later after the surgical intervention may reverse persistent pain and prevent a relapse if the reversal is sustained beyond the pharmacokinetic activity of the drug at its molecular target.

Chronic pain after surgery is likely to result from a complex combination of mechanisms, and the relative contribution of each of these mechanisms to the persistence of pain will differ between surgical procedures. Most preclinical studies on postsurgical pain have focused on the immediate perioperative period. To make progress, we must characterize these models with respect to the sequence of associated inflammatory or neuropathic changes and their potential to induce sustained pain. Testing pain-related behavior should include withdrawal responses after stimulation and parameters of spontaneous pain to validate animal models against the phenotypes of postsurgical pain in humans. Ultimately, validation will require similar treatment success or failure in the surrogate model and the human condition, and the ability to predict treatment efficacy in humans. Genetic factors, gender, and age have an effect on the risk for developing chronic pain and should therefore be evaluated.13,14The epidemiologic significance of chronic pain after surgery is enormous. Successful prevention and effective management will depend on the identification of targets for disease-modifying intervention rather than analgesia alone.

*Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Charlestown, Massachusetts. †Department of Anesthesiology and Pharmacology, University of California, San Diego, LaJolla, California.

Kehlet H, Jensen TS, Woolf CJ: Persistent postsurgical pain: Risk factors and prevention. Lancet 2006; 367:1618–25
Clark JD, Qiao Y, Li X, Shi X, Angst MS, Yeomans DC: Blockade of the complement C5a receptor reduces incisional allodynia, edema, and cytokine expression. Anesthesiology 2006; 104:1274–82
Nathan JD, Pappas TN: Inguinal hernia: An old condition with new solutions. Ann Surg 2003; 238:S148–57
Xu J, Brennan TJ: Comparison of skin incision vs.  skin plus deep tissue incision on ongoing pain and spontaneous activity in dorsal horn neurons. Pain 2009; 144:329–39
Walker SM, Franck LS, Fitzgerald M, Myles J, Stocks J, Marlow N: Long-term impact of neonatal intensive care and surgery on somatosensory perception in children born extremely preterm. Pain 2009; 141:79–87
Laboureyras E, Chateauraynaud J, Richebe P, Simonnet G: Long-term pain vulnerability after surgery in rats: Prevention by nefopam, an analgesic with antihyperalgesic properties. Anesth Analg 2009; 109:623–31
Mogil JS: Animal models of pain: Progress and challenges. Nat Rev Neurosci 2009; 10:283–94
Ito N, Obata H, Saito S: Spinal microglial expression and mechanical hypersensitivity in a postoperative pain model: Comparison with a neuropathic pain model. Anesthesiology 2009; 111:640–8
Joshi SK, Mikusa JP, Hernandez G, Baker S, Shieh CC, Neelands T, Zhang XF, Niforatos W, Kage K, Han P, Krafte D, Faltynek C, Sullivan JP, Jarvis MF, Honore P: Involvement of the TTX-resistant sodium channel Nav 1.8 in inflammatory and neuropathic, but not post-operative, pain states. Pain 2006; 123:75–82
King T, Vera-Portocarrero L, Gutierrez T, Vanderah TW, Dussor G, Lai J, Fields HL, Porreca F: Unmasking the tonic-aversive state in neuropathic pain. Nat Neurosci 2009; 12:1364–6
Mogil JS, Crager SE: What should we be measuring in behavioral studies of chronic pain in animals? Pain 2004; 112:12–5
Auer JA, Goodship A, Arnoczky S, Pearce S, Price J, Claes L, von Rechenberg B, Hofmann-Amtenbrinck M, Schneider E, Muller-Terpitz R, Thiele F, Rippe KP, Grainger DW: Refining animal models in fracture research: Seeking consensus in optimising both animal welfare and scientific validity for appropriate biomedical use. BMC Musculoskelet Disord 2007; 8:72
Greenspan JD, Craft RM, LeResche L, Arendt-Nielsen L, Berkley KJ, Fillingim RB, Gold MS, Holdcroft A, Lautenbacher S, Mayer EA, Mogil JS, Murphy AZ, Traub RJ: Studying sex and gender differences in pain and analgesia: A consensus report. Pain 2007; 132(suppl 1):S26–45
Ririe DG, Vernon TL, Tobin JR, Eisenach JC: Age-dependent responses to thermal hyperalgesia and mechanical allodynia in a rat model of acute postoperative pain. Anesthesiology 2003; 99:443–8.