430 Page 1

Pain: Mechanism and Activity at the Spinal Cord

James C. Eisenach, M.D. Winston-Salem, North Carolina

Normal Pain Transduction Pain is a sensori-emotional experience which results from stimulation of nerve fibers which report from structures interacting with our external environment or which report from internal, visceral structures. In the former case, our mind localizes the source of the pain very precisely and conscious and subconscious systems cause us to examine the painful area of the body, protect it, and stop the painful stimulus. As wonderfully reviewed by Basbaum (1), scientists have identified the proteins embedded in nerve endings which monitor the local environment and cause nerves to fire in a manner which is perceived as painful. Thus, there are a series of related proteins which respond to temperature, some in the noxious cold range, some at body temperature, and some in the noxious heat range, by opening a channel, allowing cations, especially Calcium, to enter, and depolarizing the nerve terminal, leading to generation of an action potential. Some of these same proteins also respond to chemicals which are perceived by the mind as hot (capsaicin, wasabe) or cold (menthol), and it was actually from this property of binding specifically to these chemicals that originally allowed these proteins to be recognized and identified by scientists. In addition to these temperature- and chemical-sensitive proteins, other proteins open their cation channel when exposed to hydrogen ions, are termed Acid Sensing Ion Channels (ASICs), and transduce the pain related to local falls in pH, such as would occur during ischemia of skeletal or myocardial muscle, during inflammation of tissue, or during exposure to environmental acids. Finally, there are a series of proteins, including some ASICs, which open their cation channel when the membrane of the nerve terminal is deformed by pressure. Some are exquisitely sensitive to small deformations of the membrane, and transduce the sensation of light touch, whereas others only respond to more marked deformations of the membrane, and transduce the sensation of noxious pressure or pinch. In addition to these specialized proteins, pain is also selectively elicited in the normal condition by activation of certain types of nerve fibers the unmyelinated, small diameter C fibers and the sparsely myelinated, small diameter A fibers. Stimulation of large myelinated fibers, the A fibers, normally is not perceived as painful. Postoperative Pain Pain related to tissue inflammation has been extensively studied in animals and humans for several decades, and, until recently, it was tacitly assumed that postoperative pain was essentially the pain of inflammation plus, perhaps, some direct pain from cutting nerves. Over a decade ago, Tim Brennan, an anesthesiologist at the University of Iowa, described a simple model of surgery, incision of the rat paw, in order to test whether this was true (2). He subsequently has demonstrated important differences between pain after surgery and that from inflammation. For example, there is sensitization of nerve endings, leading to spontaneous firing of nerve fibers, which constantly drives a pain system in the spinal cord after surgery, whereas this is not the case with simple inflammation. More important from a clinical perspective, some drugs are effective to treat either surgery or inflammation, such as the opioids, whereas others are unique to the setting. At the spinal cord, for example, glutamate receptors of the n-methyl-d-aspartate (NMDA) subtype are essential to driving pain after inflammation, but are not involved in pain after surgery. Other drugs may be active in both settings, but for different reasons. For example, cyclo-oxygenase (COX) inhibitors are effective analgesics in both inflammatory and surgical conditions. Whereas some of this analgesia occurs by actions in the periphery in areas of inflammation and immune response, there is also a central site of action for COX inhibitors (3). Recent work by our laboratory suggests that, at the spinal cord level, the isoenzyme of COX activated by inflammation (COX-2) is different from that activated by surgery (COX-1). Whether inhibition of COX at the spinal level plays an important part in acute and chronic pain after surgery is under study. Opioids are commonly administered to treat surgical pain, and some have suggested that pre-emptive administration of large doses of opioids during surgery may prevent some of the amplification or sensitization of pain processing which contributes to postoperative pain. Certainly, opioids should be administered for the treatment of moderate to severe postoperative pain. However, whether large doses of opioids help or hurt the situation is less

Copyright 2011 American Society of Anesthesiologists. All rights reserved.

430 Page 2

[Type text]

[Type text]

clear, since opioids themselves stimulate a sensitizing system and can result in prolonged increases rather than decreases in pain (4). Transition of acute to chronic pain: Can chronic pain be prevented? Chronic pain often begins as acute pain, and as much as 1/3 of the population seen in chronic pain clinics date the onset of their pain to the trauma and acute pain of surgery (5). Risk factors for chronic pain following surgery include the type of surgery itself (e.g., major limb amputation resulting in a greater risk than abdominal hysterectomy), anxiety state, degree of somatization, and response to experimental pain (6). Additionally, severe postoperative pain carries an increased risk of chronic pain after surgery, and its is tempting to speculate that better treatment of pain in these individuals would not only benefit them in the peri-operative period, but also for many months or years to come. There is some experimental evidence to suggest that the spinal cord plays a key role in the transition of acute to chronic pain. For example, surgery in rats results in activation of spinal microglia and stimulation of cyclooxygenase (COX) enzymes and release of prostaglandins (7). Inhibition of this process by spinally administered COX inhibitor, especially COX-1 inhibitors, at the time of surgery in animals reduces acute hypersensitivity and, more importantly, prevents permanently the development of chronic hypersensitivity (8). We recently introduce intrathecal ketorolac, a COX-1 preferring inhibitor, into clinical trials. Other studies suggest a relationship between hypersensitivity to mechanical stimuli surrounding the wound after surgery and the risk of chronic pain after surgery. De Kock and colleagues have led this field of investigation, asking whether systemic administration of the NMDA receptor antagonist ketamine (9), or epidural or spinal administration of opioids, local anesthetics, or clonidine (10,11) alters acute perioperative hypersensitivity and chronic pain in patients undergoing colectomy. These studies have resulted in two major conclusions. First, there is a highly significant correlation between the ability of these therapies to reduce hypersensitivity surrounding the wound 48 hr after surgery and their ability to reduce pain 1 year after surgery. Some of these therapies, like low dose ketamine infusion, have minimal impact on acute perioperative pain and morphine use, yet have a large effect on acute hypersensitivity surrounding the wound and on the incidence of chronic pain. This suggests that we should pay attention to hypersensitivity phenomena in addition to pain in the acute perioperative period in studies to prevent chronic pain after surgery. Second, several manipulations to reduce spinal cord activation, including intraoperative neuraxial opioids, clonidine, and local anesthetics dramatically reduce the incidence of chronic pain after surgery. Should these results be further replicated, they would provide a strong rationale for more widespread use of regional anesthesia, at least epidural or spinal anesthesia and analgesia, for surgery. Changes in the nervous system associated with chronic pain Alterations in sensory neurotransmission are observed in 4 general sites after peripheral nerve injury, and are these alterations are argued to underlie chronic neuropathic pain. Some investigators believe that nerve injury, whether from physical trauma, invasion of local or metastatic cancer, viral infection, diabetes, or chemotherapy results in fundamental changes that drive changes elsewhere in the central nervous system in states of chronic pain (12). According to this argument, blocking abnormal activity of injured peripheral nerves makes pain and hypersensitivity disappear, and the key to novel and effective analgesia for chronic pain should target these peripheral sites. Two fundamental changes occur in the periphery after injury. First, there is an increase in excitation state of peripheral nerves. This reflects increased expression and activity of excitatory ion channels and receptors along the nerve and at nerve endings, such as the transient receptor potential ion channel, TRPV-1, which responds to capsaicin, heat, and low pH. The result is depolarization of the resting membrane potential, bringing it closer to the threshold for action potential firing. This in turn could lead to spontaneous firing of single pulses or short bursts, and increased sensitivity to normal stimuli, resulting in allodynia and hyperalgesia. Its important to recognize, however, that these changes will never result in abnormal, repetitive firing, such as occurs in humans at the site of neuroma formation and is thought to underlie spontaneous, ongoing pain. A second process, increased excitability, is thought to underlie this ability to spontaneously fire. This change reflects altered or increased expression of voltage gated ion channels, leading to abnormal pacemaker activity, or oscillations of resting membrane potential Copyright 2011 American Society of Anesthesiologists. All rights reserved.

430 Page 3

[Type text]

[Type text]

causing spontaneous and repetitive firing. Both of these phenomena may be driven, not just from the injury itself, but by the immune response at the site of injury and Wallerian degeneration. As such, recruited immune cells release growth factors, cytokines, and prostaglandins, some of which are taken up by nerves and result in change in their expression of ion channels and receptors and the sensitization processes described above (13). A second site of neuroplasticity thought to result in chronic pain is the spinal cord. Following chronic inflammation or nerve injury in the periphery, several aspects of the spinal cord view of afferent input change (14). Changes occur both in neurons as well as in glia, in general increasing excitatory mechanisms and decreasing inhibitory mechanisms. A fibers may change anatomic location in the cord, and upregulate synthesis of excitatory neurotransmitters such as CGRP, which normally is only present in nociceptors. Abnormal small and perhaps large diameter fiber input not only leads to spontaneous pain by the normal transmission system, but stimulates a series of changes in the spinal cord which are associated with central sensitization. These processes classically involve NK1 and NMDA receptor activation by ongoing release of substance P and glutamate, respectively. These result in abnormal expression of activated transcription factors in spinal neurons, including CREB and cFos, altering ion channel and receptor expression. There may additionally be a loss of inhibitory tone in the cord, either through a permanent loss of GABA containing neurons and opioid receptors or a reduction in their activation. Neurotrophins, especially BDNF, are transported from the periphery and released by afferents into the spinal cord, further enhancing excitation. Interestingly, recent data from our laboratory suggests that BDNF also stimulates sprouting of inhibitory noradrenergic fibers in the spinal cord. Finally, resident immune / support cells, first microglia and later astrocytes, become activated, resulting in further release of BDNF, cytokines, and neurotransmitters which maintain the sensitized state. The spinal cord receives descending inhibitory and excitatory influences from higher centers, and these modulate the pain gate in the cord, according to the gate control theory of pain (15). Peripheral nerve injury also alters descending influences to the spinal cord which can lead to central sensitization. Medullary centers project to the cord, releasing both inhibitory (norepinephrine) and excitatory (serotonin) catecholamines and other neurotransmitters. The relative role of descending inhibition and facilitation after inflammation or nerve injury is complex, depending on time after initiation of the injury and phenomenon of pain studied. In general with long standing nerve injury there is little change in descending inhibition with a large increase in descending facilitation (16). As such, hypersensitivity behaviors in animals after peripheral nerve injury are abolished by destruction of facilitatory centers in the brainstem. Thus, the essentials of chronic pain transmission include abnormal small and large diameter afferent input, altered gene transcription in spinal cord neurons, loss of inhibitory tone, activation of microglia and astrocytes, and increased descending facilitation. Finally, cortical neuroplasticity occurs in chronic pain states. Although much less is known regarding changes in the brain then in the spinal cord or periphery, two general phenomena have been described. The first relates to cortical plasticity in the representation of body parts associated with pain. In an elegant series of studies in non human primates (17), repetitive motion injury of the hand results in hypersensitivity to light touch and inability to use the fingers for fine motions. Receptive fields of cortical neurons which respond to stimuli in the hand become enlarged, disorganized, and overlapping. Physical therapy by training the monkey to discern small differences in strength or location of tactile stimuli to the hand result not only in return to the ability to use the hand for fine motions and alleviation of hypersensitivity, but also in a remarkable return of receptive fields of cortical neurons to the normal state. These findings underscore the ability of the adult brain to undergo remarkable plasticity, and may represent the basis for improvement in function and pain by physical therapy. Targeting peripheral nerves for the treatment of pain A central problem from peripheral nerve block injection of local anesthetics is the toxicity associated with their absorption, or more likely, accidental injection into the circulation. This risk is inherent in the local anesthetics themselves. Recent research on the transient receptor potential vanilloid-1 (TRPV-1) channel, the so called capsaicin receptor, suggests that another approach might be taken to reduce the risk inherent in local anesthetic molecules themselves. TRPV-1 channels are expressed on nerve axons, and when stimulated by agonists such as capsaicin, they open with a pore size large enough to allow entry of local anesthetics (18). Since local anesthetics block axonal conduction by entering the nerve, where they bind to the channel on the cytoplasmic side, addition of capsaicin Copyright 2011 American Society of Anesthesiologists. All rights reserved.

430 Page 4

[Type text]

[Type text]

allows rapid and selective entry of local anesthetics into these pain fibers. Clinical trials will ultimately determine whether perineural injection of capsaicin or other activators or TRPV-1 channels are safe and effective to increase permeability of drugs to enter the peripheral nerve. Should they prove to be so, their combination with local anesthetics could enhance the safety of peripheral nerve block. There is also an emerging literature to suggest that reduction in inflammation in nerves as a cause of pain could be treated by peripheral nerve injections. Mechanical nerve injury, either from trauma or by tumor enlargement, results in cell death, Wallerian degeneration, often followed by regeneration. These processes result in activation and release of pro-inflammatory compounds from cells which reside in the peripheral nerve, ultimately leading to sensitization by upregulating ion channels and by inducing further spinal cord release of cytokines (19). Animal studies and some anecdotal clinical experience suggests that neuropathic pain associated with peripheral nerve injury and presumed inflammation can be effectively treated by perineural administration of drugs which modulate the immune response. One example is clonidine, which acts on 2-adrenoceptors localized to macrophages and T cells in states of neuritis (20,21). Stimulation of these receptors by clonidine downregulates their pro-inflammatory products while upregulating their anti-inflammatory products, and these effects are accompanied by a reduction in hypersensitivity (22). Whether prolonged peripheral nerve blockade alters the risk of chronic pain after surgery is unknown. Most studies have focused on blockade for a few hr to a few days, with contradictory results. Perhaps this is not surprising, since we dont know the time at which a transition from acute to chronic pain occurs. One study with very long (weeks) treatment with peripheral nerve blockade in patients with lower limb amputation showed a remarkable reduction in phantom limb pain (23), but these results have not to date been reproduced. Targeting the spinal cord for treatment of chronic pain There are 5 essential targets for intrathecal and epidurally administered drugs. I will discuss in this lecture these 5 targets in turn, focusing on clinically used drugs and using examples from both the acute and chronic pain literature. 1. Primary afferent fibers Both nerve conduction and synaptic release mechanisms of primary afferent fibers can be affected by spinal drugs. Local anesthetics block nerve conduction after epidural injection by action on nerve roots near their exit from dural sleeves and, to a lesser extent by diffusion into the intrathecal space and direct access to nerve roots. Being non-selective NaV channel blockers, local anesthetics are capable of blocking nerve conduction of all fiber types, including all sensory modalities, motor fibers, and sympathetic fibers. As such, the most commonly observed complications of intrathecal / epidural administration of local anesthetics are motor block and cardiovascular depression. Ideally, blockade of only certain subtypes of NaV channels which are expressed primarily or exclusively on sensory C fibers would produce analgesia without other proprioceptive, motor, or sympathetic blockade. For this reason the pharmaceutical industry has spent decades attempting to create small molecules to selectively block certain channels or to decrease their level of expression. Alternatively, there is considerable interest in the recent observation that TRP channels, when activated, increase their pore size large enough to allow local anesthetics access to the cytoplasm, where they reach their site of binding to ion channels. This line of investigation suggests that combination of TRP channel activators and very low concentrations of local anesthetics might produce selective blockade which is not possible from local anesthetics alone. As regards synaptic release from primary afferents, three classes of agents acting at this site are approved for epidural and/or intrathecal administration opioids (primarily morphine and fentanyl), 2-adrenoceptor agonists (clonidine and, to a much lesser extent, epinephrine), and Ca2+ channel modulators (ziconitide). These drugs also have actions at other sites within the spinal cord. Numerous G-protein coupled receptors inhibit neurotransmitter release from primary afferents in animals, although their relevance to analgesia in humans has been assessed in only two cases. Intrathecal injection of adenosine reduces hypersensitivity phenomena in humans with experimentally induced hypersensitivity and also reduces areas of allodynia and hyperalgesia in patients with chronic neuropathic pain. It does not, however, reduce spontaneous pain in patients with chronic pain. Intrathecal injection of neostigmine reduces pain in volunteers and in settings of acute pain, and animal studies suggest that this Copyright 2011 American Society of Anesthesiologists. All rights reserved.

430 Page 5

[Type text]

[Type text]

cholinesterase inhibitor acts in large part by increasing availability of acetylcholine at inhibitory muscarinic receptors on presynaptic terminals for this effect. Neither of these drugs are in widespread clinical use, due to relatively poor efficacy or intolerable side effects. 2. Second order neurons in the spinal cord which project to supraspinal structures Of clinically used drugs, opioids, 2-adrenoceptor agonists, and Ca2+ channel modulators, in addition to actions on primary afferents, also act on second order neurons to reduce excitability or hyperexcitability. A host of targets have been identified at this site in animal studies, including other inhibitory G-protein coupled receptors, chloride transporters, growth factor stimulated kinases, cytokine receptors, and excitatory G-protein coupled and ionotropic receptors, including prostaglandin receptors and the AMPA receptor. None of these have been systematically tested for relevancy after epidural or intrathecal drug delivery in the human, with the exception of ketorolac, which inhibits synthesis of prostaglandins (and potentially increases concentrations of endogenous cannabinoids). Acute intrathecal bolus of ketorolac in humans fails to reduce hypersensitivity after acute or chronic experimental inflammation (as opposed to rodents, where it is effective by this route), and has only a minor effect on pain and hypersensitivity after surgery or in patients with chronic pain. These observations, coupled with data indicating that most of these pain states are not associated with increased concentrations of prostaglandins in CSF, question the relevance of prostaglandin receptors in the spinal cord of the human in the treatment of pain. 3. Modulatory cells and terminals in the spinal cord (whether from intrinsic interneurons or descending inhibitory or facilitatory pathways) Action of acetylcholine on excitatory muscarinic receptors located on GABA containing inhibitor interneurons may partially underlie analgesia from intrathecal neostigmine in humans, although other sites of action are possible, as noted above. Descending noradrenergic pathways represent a major inhibitory influence on pain transmission at the spinal level, and analgesia from spinally administered 2-adrenoceptor agonists is thought to occur through mimicking the action of spinal norepinephrine. It is possible that analgesia from inhibitors of norepinephrine reuptake, commonly used in the treatment of chronic pain, reflects amplification of spinal norepinephrine release, none of the clinically used oral agents are approved for spinal administration. Indeed, it may be dangerous to administer these drugs spinally. For example, amitriptyline, and other tricyclic antidepressants, has multiple effects in addition to inhibition of norepinephrine reuptake, including blockade of NMDA receptors. We have shown that amitriptyline, like other NMDA receptor blockers, produces overt spinal neurotoxicity when administered intrathecally. 4. Spinal cord glia Glia in the spinal cord, especially astrocytes and microglia, respond to peripheral injury or high intensity afferent traffic, by releasing a host of compounds, including prostaglandins, cytokines, and growth factors, which contribute to central sensitization and, presumably increased pain (24). Microglia are activated quickly following a peripheral injury, including in the postoperative period, whereas astroglia can remain activated for prolonged periods following nerve injury, at least in rodent models of postoperative and neuropathic pain. Spinal injection of a variety of glial inhibitors reduces hypersensitivity in rodents following these injury, but the relevance of these observations to clinical pain is highly uncertain. We recently introduced the COX inhibitor, ketorolac into clinical trials for spinal analgesia to probe the role of spinally produced prostaglandins, presumably from activated glia, in clinical pain states. As noted earlier, spinal concentrations of prostaglandins were rarely elevated following experimental pain or inflammation or in patients with neuropathic pain, and ketorolac had minimal to no effect on pain. The one exception was patients who were receiving high doses of spinal opioids, who had markedly elevated CSF concentrations of prostaglandins and who did respond to intrathecal ketorolac with analgesia. Thus, there is only modest evidence that prostaglandin products of activated glia play an important or even modulatory role in clinical pain. 5. Spinal meninges, primarily the metabolically active arachnoid membrane The arachnoid membrane is a metabolically active group of cells important to transport of substances as well as their metabolism. This may explain a paradox observed with bolus and infusion therapy using the cholinesterase inhibitor, neostigmine, for acute pain. Intrathecal administration of this compound results in clinically unacceptable nausea and vomiting, yet analgesia is obtained after epidural administration without these side effects. It is conceivable that epidural administration of neostigmine acts, not so much by diffusion into the

Copyright 2011 American Society of Anesthesiologists. All rights reserved.

430 Page 6

[Type text]

[Type text]

intrathecal space and hence into the spinal cord to inhibit cholinesterase, but rather by local inhibition of cholinesterase in the lumbar meninges, allowing for a fairly restricted buildup of acetylcholine and analgesia. References 1. Basbaum AI, Bautista DM, Scherrer G, Julius D. Cellular and molecular mechanisms of pain. Cell 2009;139:267-84. 2. Brennan TJ, Vandermeulen EP, Gebhart GF. Characterization of a rat model of incisional pain. Pain 1996;64:493-501. 3. Malmberg AB, Yaksh TL. Hyperalgesia mediated by spinal glutamate or substance P receptor blocked by spinal cyclooxygenase inhibition. Science 1992;257:1276-9. 4. Clrier E, Laulin JP, Corcuff JB, et al. Progressive enhancement of delayed hyperalgesia induced by repeated heroin administration: A sensitization process. J Neurosci 2001;21:4074-80. 5. Crombie IK, Davies HT, Macrae WA. Cut and thrust: antecedent surgery and trauma among patients attending a chronic pain clinic. Pain 1998;76:167-71. 6. Perkins FM, Kehlet H. Chronic pain as an outcome of surgery - A review of predictive factors. Anesthesiology 2000;93:1123-33. 7. Zhu XY, Conklin D, Eisenach JC. Cyclooxygenase-1 in the spinal cord plays an important role in postoperative pain. Pain 2003;104:15-23. 8. Hefferan MP, O'Rielly DD, Loomis CW. Inhibition of spinal prostaglandin synthesis early after L5/L6 nerve ligation prevents the development of prostaglandin-dependent and prostaglandin-independent allodynia in the rat. Anesthesiology 2003;99:1180-8. 9. De Kock M, Lavand'homme P, Waterloos H. 'Balanced analgesia' in the perioperative period: is there a place for ketamine? Pain 2001;92:373-80. 10. De Kock M, Lavand'homme P, Waterloos H. The short-lasting analgesia and long-term antihyperalgesic effect of intrathecal clonidine in patients undergoing colonic surgery. Anesth Analg 2005;101:566-72. 11. Lavand'homme P, De Kock M, Waterloos H. Intraoperative epidural analgesia combined with ketamine provides effective preventive analgesia in patients undergoing major digestive surgery. Anesthesiology 2005;103:813-20. 12. Devor M. Neuropathic pain: what do we do with all these theories? Acta Anaesth Scand 2001;45:1121-7. 13. DeLeo JA, Yezierski RP. The role of neuroinflammation and neuroimmune activation in persistent pain. Pain 2001;90:1-6. 14. Yaksh TL. Regulation of spinal nociceptive processing: where we went when we wandered onto the path marked by the gate. Pain 1999;Suppl 6:S149-S152. 15. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;150:971-5. Copyright 2011 American Society of Anesthesiologists. All rights reserved.

430 Page 7

[Type text]

[Type text]

16. Vanegas H, Schaible HG. Descending control of persistent pain: inhibitory or facilitatory? Brain Research Reviews 2004;46:295-309. 17. Blake DT, Byl NN, Cheung S, et al. Sensory representation abnormalities that parallel focal hand dystonia in a primate model. Somatosens Mot Res 2002;19:347-57. 18. Bautista D, Julius D. Fire in the hole: pore dilation of the capsaicin receptor TRPV1. Nat Neurosci 2008;11:528-9. 19. Kieseier BC, Hartung HP, Wiendl H. Immune circuitry in the peripheral nervous system. Curr Opin Neurol 2006;19:437-45. 20. Lavand'homme PM, Ma W, De Kock M, Eisenach JC. Peri-neural 2A -adrenoceptor activation inhibits spinal cord neuroplasticity and tactile allodynia after nerve injury. Anesthesiology 2002;97:972-80. 21. Lavand'homme PM, Eisenach JC. Perioperative administration of the 2-adrenoceptor agonist clonidine at the site of nerve injury reduces the development of mechanical hypersensitivity and modulates local cytokine expression. Pain 2003;105:247-54. 22. Romero-Sandoval EA, McCall C, Eisenach JC. 2-Adrenoceptor stimulation transforms immune responses in neuritis and blocks neuritis induced pain. J Neurosci 2005;25:8988-94. 23. Borghi B, D'Addabbo M, White PF, et al. The use of prolonged peripheral neural blockade after lower extremity amputation: the effect on symptoms associated with phantom limb syndrome. Anesth Analg 2010;111:1308-15. 24. Ren K, Dubner R. Interactions between the immune and nervous systems in pain. Nat Med 2010;16:126776.

Copyright 2011 American Society of Anesthesiologists. All rights reserved.

Disclosure Medtronic, consulting fees; Endo, consulting fees; Solvay, consulting fees; King Pharmaceuticals, funded research Pharmaceuticals, funded research