Publications
APS Bulletin • Volume 14, Number 1, 2004
Innovations in Practice
Norman Harden, M.D., Department Editor
Molecular Mechanisms of Pain in Pancreatitis
John H. Winston, PhD, and Pankaj J. Pasricha, MD
Acute or chronic pancreatitis (CP) is a significant contributor to the “burden of gastrointestinal disease” in the United States, according to The Burden of Gastrointestinal Diseases, a survey conducted by the American Gastroenterological Association (American Gastroenterological Association, 2001). In 1998, there were about 1.2 million cases, with 327,000 inpatient admissions and 530,000 physician office visits. The estimated total direct cost for this group of diseases was $2.1 billion in 1998.
Unfortunately, progress in our understanding of the biology of these diseases has been slow, particularly with respect to the pathogenesis of the cardinal symptom of pancreatitis—pain. Any physician who has worked with these patients knows pain is not only the most important symptom of CP, but also the most difficult to treat. Painful CP is poorly understood and its management is controversial (DiMagno, 1999). A serious obstacle to improved care for these patients is our lack of knowledge of the causes of pain in pancreatitis. Consequently, the approach to treatment is usually empirical (American Gastroenterological Association, 1998). Despite a wide variety of approaches that are innocuous (enzyme therapy), minimally invasive (endoscopic decompression and nerve blocks), and highly aggressive (surgical decompression, pancreatectomy), no consensus has emerged and no single form of treatment can be considered satisfactory at this time.
Clinical Patterns of Pain in CP
CP pain varies widely from patient to patient and over time in the same patient; this makes it difficult to reliably interpret the published literature, much of which is in the form of uncontrolled clinical trials. Several factors may contribute to this variability, including etiology, natural history of pancreatitis, and the presence or absence of associated complications (e.g., pseudocysts, biliary obstruction, etc.). In addition, psychosocial factors such as secondary gain and chemical dependency and abuse issues may significantly add to the complexity of management for these patients. Pain is present for approximately 75% of patients with alcoholic CP, 50% of patients with late-onset idiopathic CP, and 100% of patients with early-onset idiopathic CP (DiMagno, 1999).
Most of the published literature deals with pain in patients with alcoholic pancreatitis, the most common form of the disease. Even in these patients, pain is not a uniform symptom, and its pattern may have implications for therapy and prognosis. In a prospective longitudinal study of 207 patients, Ammann et al. (Ammann, 2001; Ammann & Muellhaupt, 1999) identified two categories of patients in approximately equal numbers. The first group reported an intermittent, but recurrent, pattern of short episodes of pain; none of these patients required surgery. The second group complained of constant pain occurring 2 or more days per week for at least 2 months at a time, and everyone in this group underwent surgery. The majority of these patients had a pseudocyst, and some others had biliary obstruction (this was particularly common later in the course of the disease). Most patients (80%) in either group had pain relief after 10 years.
Although it appears pain will subside spontaneously over time in at least 50% of patients with alcoholic CP, in practice such an outcome remains difficult to predict in individual patients. Several studies suggest most patients will continue to experience pain despite glandular failure, withdrawal of alcohol, or even after pancreatic surgery (Di Sebastiano, Friess, Di Mola, Innocenti, & Buchler, 2000; Malfertheiner, Buchler, Stanescu, & Ditschuneit, 1987; Malfertheiner, Pieramico, Buchler, & Ditschuneit, 1990). Consequently, a decision to withhold surgical or other invasive therapies in the hope of eventual “burnout” is probably not warranted. This is particularly true of patients with idiopathic forms of CP, with one series reporting spontaneous cessation of pain in only about 25% of patients (Lankisch, Seidensticker, Lohr-Happe, Otto, & Creutzfeldt, 1995).
Pathogenesis of Pain in CP
A minority of CP patients with pain have readily identifiable lesions, such as pseudocysts, which are relatively easy to treat surgically or endoscopically. In other patients, pain has traditionally been thought to result from a variety of causes, including elevated intrapancreatic pressures, ischemia, and fibrosis. It is likely that these phenomena, while clearly associated with the disease, are not the root cause of the pain. Instead, they probably are inciting factors against a background of neuronal sensitization induced by damage to the perineurium and subsequent exposure of the nerves to mediators and products of inflammation. This is in keeping with evidence in the literature that shows persistent pain associated with peripheral tissue injury or inflammation results from long-term changes in nociceptive processing that can involve both primary sensory neurons (peripheral sensitization) and neurons in the spinal cord and higher structures (central sensitization). Many of the elements of the “inflammatory soup” described in somatic pain models, including ions (K+ and H+), amines (5-HT and histamine), kinins (bradykinin), prostanoids (PGE2), purines (ATP), cytokines (TNF, IL-1, and IL-6), nitric oxide, and caloric activity (heat) are likely to result in early sensitization of pancreatic nociceptors in patients with pancreatitis, as well. The evidence for neuroimmune interactions in the pathogenesis of pain in humans with CP has recently been reviewed (Di Sebastiano, et al., 2003), and is summarized in Table 1. Given the difficulty in medically treating the pain, the neurochemical changes induced in the afferent nervous system may be different than those of somatic or visceral inflammation in hollow organs.
|
Table 1. Pathological Correlates of Pain in Patients with Chronic Pancreatitis (CP)
Adapted from Di Sebastiano, et al., 2003 |
Experimental Models of Painful Pancreatitis
Although pathological studies on resected human specimens discussed above offer valuable insight, mechanistic studies require robust experimental models. In this regard, a significant barrier to understanding the biological basis of pain in pancreatitis has been the lack of suitable animal models and assays. This is due in part to the relative inaccessibility of the pancreas, and also to the lack of validated behavioral or other surrogate markers for pancreatic pain in animals. One approach to studying pancreatic pain is to take advantage of the phenomenon of referred hypersensitivity, which is characteristic of visceral pain. Humans with both acute and chronic pancreatitis can develop severe pain that often is associated with areas of referred cutaneous hyperalgesia extending across the upper abdomen and around the back. These findings were exploited to develop two recently described rat models of pancreatitic pain. Lu, Vera-Protocarrero, and Westlund (2003) induced “persistent” pancreatitis by systemic dibutylin dichloride in rats and showed an increase in withdrawal events after von Frey filament stimulation of the abdomen and decreased withdrawal latency after thermal stimulation during a period of 7 days, indicating a “sensitized nociceptive” state accompanied by increased levels of substance P, but not CGRP levels in spinal cords (Lu, et al., 2003).
The efficacy of VFF testing as a means to measure nociceptive behavior also was shown by Winston et al. from our group, who induced acute pancreatitis by systemic L-arginine. In this model, referred tactile sensitization occurred during the period of maximum tissue damage and inflammation in the pancreas, developing on Day 1 after L-arginine administration and persisting at least through Day 7 (Winston, et al., 2003). Lu et al. (2003) also have described direct behavioral assays for pancreatic pain using acute noxious stimulation of the pancreas via an indwelling ductal canula in awake and freely moving rats. These assays included cage crossing, rearing, and hind limb extension in response to intrapancreatic bradykinin infusion. Intrathecal administration of either D-APV (NMDA receptor antagonist) or morphine alone partially reduced visceral pain behaviors in this model. Combinations of both reduced pain behaviors to baseline. These findings demonstrate the feasibility of making quantitative measures of nociceptive behavior in rats with pancreatitis.
A major unanswered question is clinical relevance of rat models. Chemically induced inflammatory models may not completely reproduce all aspects of human pancreatic pathology, including changes in the duct and gross destruction of the architecture of the pancreatic lobules. Nevertheless, these models represent useful systems to investigate molecular changes in the pancreas and the afferent nervous system that contribute to pain in pancreatitis and to test potential novel analgesics.
Mechanisms of Neuronal Sensitization in Experimental Pancreatitis: Potential Role for Neurotrophins and Protein Kinases
We previously described an increased expression of nerve growth factor (NGF) as well as other neurotrophic factors in the pancreas of animals with acute L-arginine- induced inflammation (Toma, Winston, Micci, Shenoy, & Pasricha, 2000; Toma, et al., 2002). Subsequently, we also observed increases in mRNA levels of PPT and CGRP in thoracic DRG (T9-T11). No changes were found in cervical or lumbar DRG, suggesting this effect was secondary to the induction of pancreatitis and not a nonspecific response to systemic or peritoneal L-arginine by itself (Winston, et al., 2003). Initial studies with this model have produced findings similar to those found in some human studies—increased neuropeptide content of nerve fibers and increased NGF in the pancreas (Toma, et al., 2000; Winston, et al., 2003).
Nociceptive sensitization in several somatic pain models has been associated with activation of protein kinases including tyrosine kinase A (trkA, the high affinity receptor for NGF), PKC, and PKA. In the L-arginine model, increased sensitivity to abdominal stimulation with VFF observed in rats with pancreatitis was associated with an eight-fold increase in levels of phosphorylated trkA in the pancreas. We tested the hypothesis that systemic treatment with a kinase inhibitor, k252a, known to inhibit all of these kinases, would alleviate pain in an animal model of pancreatitis. Treatment with the kinase inhibitor k252a suppressed the phosphorylation of trkA in the pancreas and reversed both the abdominal sensitivity and the increase in neuropeptide expression associated with pancreatitis (Winston, et al., 2003).
Potential Role for Trypsin
In addition to the somewhat ubiquitous inflammatory elements described above, pancreatitis also is uniquely associated with a significant release and activation of endogenous proteases such as trypsin that may contribute to pain by causing damage to afferent nerves. In addition, these activated proteases exert direct effects on sensory neurons, mediated by specific receptors such as protease-activated receptor-2 (PAR-2), a member of a unique family of G-protein coupled receptors (Schmidlin & Bunnett, 2001). We demonstrated the presence of PAR-2 mRNA and protein expression in adult rat thoracic DRG, as well as an increase in intracellular calcium in response to treatment of cultured DRG neurons with either trypsin or the PAR-2 agonist-activating peptide (AcPeP) (Hoogerwerf, et al., 2001). We then showed that activation of PAR-2 in vitro resulted in sensitization of the TRPV1 receptor in the form of enhanced capsaicin-evoked release of calcitonin gene-related peptide, a marker for nociceptive signaling. Further, when injected directly into the pancreatic duct, AcPeP was able to directly activate pancreas-specific afferent neurons in vivo, as well as sensitize them to subsequent applications of capsaicin, as measured by Fos expression in the dorsal horn of the spinal cord. These observations suggest PAR-2 contributes to nociceptive signaling and may provide a novel link between inflammation and pain in pancreatitis.
Conclusions
Painful CP remains a major clinical challenge. Recent insights from both human and experimental animal studies are beginning to shed light on potential mechanisms of pain. There is growing evidence for the role of neurotrophins, and, possibly, trypsin in neuronal sensitization in this condition. If validated, these findings may have major implications for the pathogenesis of pain in CP and will provide novel targets for analgesic therapy.
References
American Gastroenterological Association. (1998). Medical position statement: Treatment of pain in chronic pancreatitis. Gastroenterology, 115(3), 763–764.
American Gastroenterological Association. (2001). The burden of gastrointestinal diseases, Bethesda, MD, 48–50:Author.
Ammann, R. W. (2001). The natural history of alcoholic chronic pancreatitis. Internal Medicine, 40(5), 368–375.
Ammann, R. W. & Muellhaupt, B. (1999). The natural history of pain in alcoholic chronic pancreatitis. Gastroenterology, 116(5), 1132–1140.
Di Sebastiano, P., Di Mola, F. F., Bockman, D. E., Friess, H., & Buchler, M. W. (2003). Chronic pancreatitis: The perspective of pain generation by neuroimmune interaction. Gut, 52(6), 907–911.
Di Sebastiano, P., Friess, H., Di Mola, F. F., Innocenti, P., & Buchler, M. W. (2000). Mechanisms of pain in chronic pancreatitis. Annali Italiani di Chirurgia, 71(1), 11–16.
DiMagno, E. P. (1999). Toward understanding (and management) of painful chronic pancreatitis. Gastroenterology, 116(5), 1252– 1257.
Hoogerwerf, W. A., Zou, L., Shenoy, M., Sun, D., Micci, M. A., Lee-Hellmich, H., et al. (2001). The proteinase-activated receptor 2 is involved in nociception. The Journal of Neuroscience, 21(22), 9036–9042.
Lankisch, P. G., Seidensticker, F., Lohr-Happe, A., Otto, J., & Creutzfeldt, W. (1995). The course of pain is the same in alcohol- and nonalcohol-induced chronic pancreatitis. Pancreas, 10(4), 338–341.
Lu, Y., Vera-Portocarrero, L. P., & Westlund, K. N. (2003). Intrathecal coadministration of D-APV and morphine is maximally effective in a rat experimental pancreatitis model. Anesthesiology, 98(3), 734–740.
Malfertheiner, P., Buchler, M., Stanescu, A., & Ditschuneit, H. (1987). Pancreatic morphology and function in relationship to pain in chronic pancreatitis. International Journal of Pancreatology, 2(1), 59–66.
Malfertheiner, P., Pieramico, O., Buchler, M., & Ditschuneit, H. (1990). Relationship between pancreatic function and pain in chronic pancreatitis. Acta Chirurgica Scandinavica, 156(4), 267–270; discussion 270–261.
Schmidlin, F., & Bunnett, N. W. (2001). Protease-activated receptors: How proteases signal to cells. Current Opinions in Pharmacology, 1(6), 575–582.
Toma, H., Winston, J., Micci, M. A., Shenoy, M., & Pasricha, P. J. (2000). Nerve growth factor expression is up-regulated in the rat model of L-arginine-induced acute pancreatitis. Gastroenterology, 119(5), 1373–1381.
Toma, H., Winston, J.H, Micci, M.A., Li, H., Hellmich, H.L., & Pasricha, P.J. (2002). Characterization of the neurotrophic response to acute pancreatitis. & Pancreas 25 (1), 31–38.
Winston, J.H., Toma, H., Shenoy, M., He, Z.J., Zou, L., Xiao, S.Y., et al. (2003). Acute pancreatitis results in referred mechanical hypersensitivity and neuropeptide up-regulation that can be suppressed by the protein kinase inhibitor k252a. Journal of Pain 4(6), 329–337.
John H. Winston, PhD, is assistant professor, Division of Gastroenterology and Hepatology, Internal Medicine, at the University of Texas Medical Branch in Galveston, TX.
Pankaj J. Pasricha, MD, is chief, Division of Gastroenterology and Hepatology, Bassel and Frances Blanton Distinguished Professor of Internal Medicine, and professor of anatomy and neurosciences and biomedical engineering at the University of Texas Medical Branch, Galveston TX.
Please direct your comments or suggestions about this article or department to Norman Harden, MD, Department Editor, at nharden@rehabchicago.org.