Kaiser Foundation Hospital, Hayward, California 94545
Probably the greatest fear of individuals suspecting they may have cancer or being advised that they do is that of intractable pain during their illness, a fear even greater than that of a shortened life expectancy. Although there have been many advances in pain management, pain accompanying malignancy has been most difficult to treat.
Sometimes, it is true, the basis of cancer pain is obvious and quite amenable to therapy, such as the pain of a pathological fracture controllable by orthopedic treatment and the pain of neural compression relieved by surgical, chemotherapeutic, or radiation therapy. Unfortunately, however, cancer pain is often due to an ill-defined local process at the tumor site, not well understood and poorly responsive to specific malignancy-directed therapies. In this regard, the pain of malignancy resembles pain experienced in other forms of chronic disease, such as chronic rheumatologic disorders, chronic pelvic disorders, and chronic inflammation of other organ systems, such as the gastrointestinal tract, skin, or cranial vasculature. In these syndromes, morbidity is not severe from pain spontaneously felt, but rather from the increased pain after a mild stimulus such as a small joint movement, gentle touch, simple walking, or normal bowel activity. Often the pain of malignancy also has this characteristic, i.e., the pain is not so severe as long as there is no stimulation of the area involved. This distinction is an important one since the underlying neural phenomena are different for spontaneous pain (in which there is enhanced spontaneous neural activity) and mild-stimulus-induced (or -exacerbated) pain, or tenderness, also termed hyperalgesia, in which there is pain after a stimulus that is normally nonpainful, a frequent accompaniment of inflammation. Whereas both peripheral factors causing spontaneous pain and central factors affecting the sensation of pain likely play a role in the pain of malignancy, we are going to focus, in this discussion, on mechanisms underlying inflammatory hyperalgesia that may be relevant to cancer pain.
The inflammatory process is activated in response to tissue injury or to the presence of foreign substances; both of these are likely to be present at sites of malignancy. In fact, inflammation facilitates the destruction and removal of necrotic tissue at tumor sites and contributes to the ongoing local immunological response that often accompanies malignancy. The process of inflammation has now been at least partly understood in terms of the action of different inflammatory mediators present at sites of tissue injury. In this chapter, we will focus on mechanisms by which several of these inflammatory mediators might act to contribute to the pain of malignancy.
In the presence of local inflammation, which is common at sites of malignancy, individual nociceptors may become sensitized so that activation is produced by stimuli usually not intense enough to do so (i.e., threshold is lowered).
Various techniques have been useful in studying mechanisms by which inflammatory mediators lower nociceptive threshold, including the study of isolated C fibers (1,2,4,17,52,61,92) and dissociated dorsal root ganglion neurons in culture (1,84,85). Behavioral studies employing nociceptive reflex tests in intact animals have also been employed (35,36,67,104). Their advantage over single-fiber studies is that they allow direct evaluation of nociception. In addition, they can be performed on awake, normally functioning animals, without the confounding effects of anesthesia. All these techniques have been used extensively in the study of hyperalgesia.
The first inflammatory mediator noted to have potent hyperalgesic properties and suggested to contribute to inflammatory pain was bradykinin (BK), a nonapeptide cleaved from a globulins by kallikreins circulating in the plasma and activated at sites of tissue injury (43). BK has been found to be present in large amounts in inflammatory exudates (26,75,89) and would be expected to be present at tumor sites characterized by surrounding inflammation. BK can activate primary afferent neurons, producing overt pain (28,49,62,114), but BK also sensitizes nociceptors producing hyperalgesia, in the absence of activation (2,51,60,68,77,82).
Interestingly, the lowering of mechanical nociceptive thresholds elicited by intradermal injection of BK in rats was observed to be markedly attenuated in rats that had been chemically sympathectomized with 6-hydroxydopamine (68,107), suggesting that BK hyperalgesia is dependent on postganglionic sympathetic neurons. Since postganglionic sympathetic terminals are present in the fine peripheral vasculature, they would be expected to be present in any vascularized tumor. The synthesized prostaglandins (PGs) appear to be the direct agents mediating BK hyperalgesia (48). BK can induce PG synthesis (5,11,16,19) and BK hyperalgesia is very significantly attenuated by the PG synthesis inhibitor indomethacin (62,64,68).
Interleukin-1 is a cytokine, produced by leukocytes in response to exposure to bacterial toxins or to inflammatory mediators (25). Both interleukin-1b and interleu kin-1a polypeptides were found to produce hyperalgesia (33). Like BK, interleukin-1 induces E-type PG (PGE) production in nonneuronal cells (23) and interleukin-1 hyperalgesia is attenuated by indomethacin. Interleukin-1 can also enhance nerve growth factor (NGF) production (53,71), which is also markedly increased following nerve injury (53). The amino terminal octapeptide fragment of NGF (13), which is structurally similar to BK (9), can also induce a hyperalgesia attenuated by sympathectomy and by indomethacin (112). All these inflammatory mediators are capable of contributing to inflammatory pain in the presence of malignancy, especially in the presence of bacterial infection at the tumor site or when there is local neural injury.
The polymorphonuclear leukocyte (neutrophil) is a principal effector cell in the acute inflammatory reaction. It is attracted to sites of inflammation in high numbers to destroy and evacuate antigenic material and facilitate removal of cellular debris; it is frequently seen in significant numbers at sites of malignancy. Since an accumulation of neutrophils is commonly associated with a marked hyperalgesia, this process probably contributes to cancer pain. Leukotriene B (LTB), a specific inflammatory mediator, is a potent neutrophil attractant (46,47) and has been shown to produce hyperalgesia both in animals (66,67,86) and humans (6,70). It was found that lowering of nociceptive threshold following intradermal injection of LTB in rats is dependent on the presence of circulating leukocytes (65).
Since the latency to onset of hyperalgesia induced by LTB is long, approximately 10 min, it was thought that a secondary mediator directly sensitizes the primary afferent nociceptors and not LTB itself. Study of supernatants of neutrophils yielded such a hyperalgesic ligand, 8(R),15(S)-diHETE (65). Two additional substances, Ca, the anaphylactoid fragment of the fifth component of the complement cascade, and formylmethionyleucylphenylalanine (fMLP), the tripeptide bacterial cell wall fragment, have been shown, like LTB, to produce neutrophil-dependent hyperalgesia mediated via 8(R),15(S)-diHETE. These hyperalgesic substances would be expected to be present at malignant sites characterized by infection or immunological response.
PGs, which, as mentioned above, mediate BK, interleukin, and NGF-octapeptide hyperalgesia, are considered prototypic sensitizing agents. Their administration does not elicit overt pain (20) yet they decrease nociceptive threshold in behavioral tests in animals (37,105,108) and produce tenderness in humans (32). In electrophysiological studies, PGE sensitizes high-threshold somatic and visceral afferents when administered systemically (41,51,58) or when injected directly into the receptive field of a nociceptive afferent (74,83). The clinical importance of PG hyperalgesia is evidenced by the analgesic properties of aspirin, indomethacin, and other nonsteroidal anti-inflammatory analgesics (NSAIAs), all of which inhibit the cyclooxygenase pathway of arachidonic acid metabolism, which mediates the endogenous synthesis of hyperalgesic prostaglandins (34,113).
The most direct evidence that prostanoids act on the primary afferent neurons and do not require intermediary cells has come from studies done on cultured neurons (1). Whole-cell patch clamp electrophysiology is performed on cells identified as nociceptors. Both PGE (and PGI) significantly enhance the magnitude of the inward current produced by a constant dose of capsaicin (the pungent portion of the chili pepper that selectively activates nociceptors) (1,12,38,57,79,102).
Adenosine, which is generated in large amounts in hypoxic and ischemic tissue, such as occurs in inflammatory lesions (29,42), and would be expected to be present at sites of tumor necrosis and inflammation, has been shown to activate unmy[chelin[chated afferents (14,59,78,90), to produce pain in humans (7,99), and to elicit nociceptive behaviors in animals (18,109). Like PGs and 8(R),15(S)-diHETE, adenosine appears to produce hyperalgesia by a direct action on the primary afferent neuron (109). The hyperalgesia appears to be mediated by action on an adenosine A type receptor, based on studies using selective antagonists and agonists for A and A action.
Serotonin (5-HT), which is released from activated platelets, is also significantly elevated in inflammatory exudates and has been reported to cause pain in humans, probably by action at the 5-HT receptor (44,45,87,88). 5-HT may potentiate the pain induced by other inflammatory mediators such as BK (2,27,39,40,54,76,80,81,88,98), and recent behavioral studies suggest that 5-HT also produces hyperalgesia, probably by direct action on the primary afferent, at a different receptor, probably the 5-HTa receptor (111). In summary, there are a number of endogenous mediators present in inflammation, likely to be present at sites of malignancy, that can lower nociceptor mechanical threshold and, consequently, substantially increase nociceptor input to the CNS after trivial stimuli, such as a movement or light touch.
Of particular interest to us recently have been the mechanisms by which these direct hyperalgesic mediators affect primary afferents to lower threshold. Interestingly, PGs (18,35,50), 5-HT (30,31,73,97), and adenosine, acting at the A receptor (21,22,91,99) have all been suggested to elevate intracellular cAMP (35). In recent studies, a membrane-permeable analogue of cAMP, 8-bromo cAMP, was found to produce hyperalgesia that appears to be direct (103). Forskolin, which directly activates adenylate cyclase and leads to increased cAMP, also produces hyperalgesia (24,93-95); phosphodiesterase inhibitors, such as IBMX and rolipram, significantly prolong 8-bromo cAMP hyperalgesia and also that produced by other directly acting agents (103,106,109). In using cultured dorsal root ganglion neurons, 8-bromo cAMP was shown to increase the inward current induced by capsaicin. A role of the cAMP second messenger system is further suggested by the attenuation of hyperalgesia by inhibitors of cAMP-dependent protein kinase (8,10).
This coupling of receptors for hyperalgesic agents, located on the primary afferent nociceptor, to adenylate cyclase activity appears to be via a guanine nucleotide regulatory stimulatory G protein (Gs). Agents that activate inhibitory G proteins (Gi), such as opioids and, interestingly, also adenosine acting at the A receptor (69,108), were found to inhibit hyperalgesia. Clearly mechanisms in the primary afferent to lower threshold after exposure to hyperalgesic inflammatory mediators are complex. The elucidation of specific pathways may allow the development of very specific novel therapeutic agents for cancer pain.
The identification of inflammatory mediators that act directly on nociceptors to cause pain or hyperalgesia (PGE, PGI, 8(R),15(S)-diHETE, adenosine (A), and 5-HTa) allows for the rational development of novel therapeutic strategies for treating inflammatory pain. These strategies may be particularly relevant novel approaches for the treatment of the pain of malignancy. Since multiple mediators probably simultaneously contribute to inflammatory pain, a multiple drug approach, not unfamiliar to oncologists, may be needed.
The understanding of the second messenger system mediating primary afferent hyperalgesia suggests an entirely new approach to the treatment of inflammatory pain of malignancy, which could obviate the need for multiple agents. Antagonizing the cAMP second messenger system in the nociceptor at the level of G-protein activation, if possible, would be a promising therapeutic approach. Agents activating inhibitory G proteins would also be of potential use. Mu opioids, which decrease intracellular cAMP via activation of inhibitory G proteins (15,63,71,96), are also good candidates (72); they in fact have already been shown to produce naloxone antagonizable analgesia when injected into the site of inflammation (hyperalgesic tissue) (52,69,100). The use of opioids at local sites of malignancy might have a similar therapeutic effect.
Delta and kappa receptor specific opioid agonists can also produce antinociception in inflamed tissue (100), but by a different mechanism. They do not affect the primary afferent and inhibit direct hyperalgesia, such as PGE hyperalgesia (69), but can inhibit sympathetic postganglionic neuron-dependent hyperalgesia (110), probably by acting on kappa and delta opioid receptors known to be present on SPGN terminals (3,55,56,115). These agents and other specific antagonists=mfor example, those directed at inhibiting receptor action of interleukin and BK=mare all potential therapies for inflammatory pain in patients with cancer. Some cancers can present insidiously with little pain, whereas others cause pain early, often by producing inflammation. In addition, complications at sites of malignancy, by infection, can be a cause of exacerbation of cancer pain. Indeed, it may be that the degree of inflammation on tumor biopsy specimens may help distinguish those patients likely to be responsive to treatments for the control of inflammatory pain.