Does an aspirin a day keep pancreas cancer away?

August 27, 2017 | Autor: Luis Lara | Categoria: Gastroenterology, Clinical Sciences, Neurosciences
Share Embed


Descrição do Produto

1002

SELECTED SUMMARIES

Reports of a protective effect of sulfasalazine or the mesalamine conjugates are numerous and seem to outnumber, and outweigh, the negative study reviewed here. Pinczowski et al. (Gastroenterology 1994;107:117–120) reported a protective effect with sulfasalazine therapy for at least 3 months in a case-control review of 102 Swedish UC patients with colon cancer (relative risk [RR], 0.38; 95% confidence interval [CI], 0.20 – 0.69) independent of disease activity. Eaden and colleagues (Aliment Pharmacol Ther 2000;14:145–153) calculated an adjusted 81% decrease in risk for colon cancer in a case-control study of 102 UC CRC patients (odds ratio [OR], 0.19; 95% CI, 0.06 – 0.61; P ⫽ 0.006). Crude colon cancer rates were also higher in a retrospective review of British UC patients (31%) who had either stopped or were noncompliant with aminosalicylate therapy than in those compliant with therapy (3%, P ⬍ 0.001) (Eur J Gastroenterol Hepatol 1996;8:1179 –1183). In contrast, a previous retrospective analysis of surveillance colonoscopies failed to show a protective effect of sulfasalazine therapy on the development of colonic dysplasia or cancer (Dis Colon Rectum 2001;44:77– 85), although insufficient sample size may have resulted in a ␤-error (Am J Gastroenterol 2003;98:1682–1687). Although antineoplastic effects have been shown in animal models of colon cancer for mesalamine (Dig Dis Sci 2000;45:1578 –1584), olsalazine (Dig Dis Sci 2000;45:1578 –1584), balsalazide (Int J Oncol 2000;17:173–179), and sulfasalazine (Br J Cancer 1992;66:777– 780), not all aminosalicylates may be equal in their protective effects on the development of human CRC, and the effects may be dose related. Eaden et al. (Aliment Pharmacol Ther 2000;14:145–153) reported that the protective effect was more apparent for mesalamine than sulfasalazine, and at a lower daily dose (1.2 g vs. 2.0 g, respectively). Preliminary data in UC suggests a dose-related protective effect of mesalamine (ⱖ1.2 g daily) on the development of colonic dysplasia or cancer (Gastroenterology 2003;124:A36) and in the progression from indefinite or low-grade dysplasia to high-grade dysplasia or cancer (ⱖ2.4 g daily) (Gastroenterology 2003;124:A242). What about other anti-inflammatory agents? Corticosteroid use (both systemic and local) was inversely related to CRC development in the Eaden study (Aliment Pharmacol Ther 2000;14:145–153). Other agents shown potentially to decrease colorectal cancer in chronic ulcerative colitis include urseodeoxycholic acid (Ann Intern Med 2001;134:89 –95, Gastroenterology 2003;124:889 – 893) and nonsteroidal anti-inflammatory agents (Am J Gastroenterol 1996;91: 44 – 48). The role of folic acid is somewhat controversial; human studies suggest a protective effect (Gastroenterology 1989;97:255– 259, Gastroenterology 1997;112:29 –32), whereas murine models of intestinal carcinogenesis imply a possible causative effect for supplementation of folate and methionine-deficient tumors (Cancer Res 2000;60:3191–3199, Cancer Res 2000;60:5434 –5440). In summary, most of the recent studies suggest a benefit of the aminosalicylates in decreasing colorectal cancer rates in ulcerative colitis; data on Crohn’s colitis is lacking. However, these retrospective analyses are plagued by referral and selection biases, as well as relatively small numbers of patients with cancer on which to base the statistical comparisons. Confounders such as patient adherence to prescribed mesalamine-based regimens (found to be only 40% in a recent University of Chicago Study [Am J Gastroenterol 2001;96: 2929 –2933]) and variations in timing and frequency of screening and surveillance colonoscopies for this population further cloud the picture. Further investigation into the role of the immunomodulators, 6-mercaptopurine and azathioprine, as potential chemopreventative agents in IBD are also warranted because many physicians discontinue

GASTROENTEROLOGY Vol. 127, No. 3

the first-line mesalamine preparations when patients achieve remission on these more potent agents. RUSSELL D. COHEN, M.D.

Reply. The potential use of 5-ASA as a prophylactic agent in ulcerative colitis (UC) against cancer development has tremendous implications for long-term patient management. 5-ASA drugs are somewhat costly and tedious to take over years because they are generally dosed 3– 4 times per day. However, if patients could be convinced that the use of these agents could prevent cancer development, they would potentially be popular because these drugs are mostly benign and easily tolerated. I remain unconvinced that the current evidence proves that 5-ASA is chemopreventative in UC. The two most recently published studies have failed to show a preventative effect of 5-ASA (Am J Gastroenterol 2003;98:2784 –2788, Gastroenterology 2004;126:451– 459). Although our study could analyze only for short-term use of 5-ASA (⬍4 years) and could access only 4 years of data to assess for colorectal cancer incidence (Am J Gastroenterol 2003;98:2784 – 2788), it is a population-based study and free of any possible selection biases of referral center populations. A case-controlled study from St. Mark’s Hospital’s longstanding UC surveillance program does access patients with 5-ASA for ⬎10 years (Gastroenterology 2004;126:451– 459). This study did not show a statistically significant benefit of mesalamine use in preventing colorectal cancer (up to 10 years of use: OR, 0.79; 95% CI, 0.4 –1.55; and ⬎10 years of use: OR, 0.65; 95% CI, 0.26 –1.62). Two recent studies from tertiary referral centers suggest some possible benefit of 5-ASA in preventing colon cancer (Gastroenterology 2003;124: A242, Gastroenterology 2003;124:A36), and while we await the full publication of these results, they will be biased by the nature of the specialty IBD clinics at each center from where the study subjects are drawn. It is plausible that patients followed at specialty clinics are generally more compliant with other aspects of their health care and use other therapies (as reviewed by Dr. Cohen) that may possibly lower colon cancer risk. Even the study most often cited for evidence in favor of 5-ASA use as chemopreventative in UC (Aliment Pharmacol Ther 2000;14:145–153) also showed a number of other factors (that imply physician contact and other therapy) as chemopreventative. Use of systemic steroids at some time (OR, 0.26; 95% CI, 0.01– 0.70), 1–2 visits with doctor/year (compared to none) (OR, 0.32; 95% CI, 0.14 – 0.76), ⬎2 doctor visits/year (OR, 0.098; 95% CI, 0.03– 0.29), and having 1 or 2 colonoscopies over the course of the disease (versus none) (OR, 0.22; 95% CI, 0.09 – 0.55) also reduced colon cancer risk. Hence, while I share Dr. Cohen’s enthusiasm in the potential of 5-ASA as a colorectal cancer preventative agent in UC, I also share his concern that the current literature on the topic does not yet lend itself to a translation from potential to reality. CHARLES N. BERNSTEIN, M.D.

DOES AN ASPIRIN A DAY KEEP PANCREAS CANCER AWAY? Schernhammer ES, Kang JH, Chan AT, Michaud DS, Skinner HG, Giovannucci E, Colditz, GA (Department of Medicine, Channing Laboratory, Harvard Medical School, Boston, Massachusetts; National Cancer Institute, Rockville, MD). A prospec-

September 2004

tive study of aspirin use and the risk of pancreatic cancer in women. J Natl Cancer Inst 2004;96:22–28. In epidemiologic studies, use of aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) have been shown to reduce the risk of gastrointestinal cancers. This association is robust for colon cancer (N Engl J Med 1995;333:609 – 614). Fewer cases of gastric and esophageal cancer have been reported in regular aspirin users compared with the expected incidence (Cancer Res 1993;53:1322–1327). The effect of aspirin use on pancreatic cancer is not clear. Schernhammer et al. (J Natl Cancer Inst 2004;96:22–28) report a prospective study of female registered nurses in the United States in which periodic assessment of aspirin use was recorded. These data were obtained from information compiled as part of the Nurses’ Health Study that was initiated in 1976 (J Womens Health 1997;6:49 – 62). Aspirin use was recorded at baseline in 1980 and updated every 2 years, except in 1986. Aspirin use was recorded differently at different times, but an estimation of usage frequency and duration could be obtained. Current aspirin use was defined as 2 or more standard tablets per week and subanalysis stratified frequency and duration of aspirin use. Nonaspirin NSAID use was not routinely assessed. Current or past smoking history was well documented. The body mass index recorded at baseline in 1976 was used for analysis. Incident pancreatic cancers were ascertained by answers on the questionnaires, medical records, or through the National Death Index from 1980 to June 1, 1998. Pancreatic cancers that occurred within 2 years of the first reported use of aspirin were excluded from the analyses. Relative risk (incidence of pancreatic cancer among participants who reported use of aspirin divided by the incidence of pancreatic cancer among participants without such a report) was used to measure association. Other potential risk factors for pancreatic cancer (age, cigarette smoking, diabetes mellitus, body mass index [BMI], and low physical activity) were adjusted for by Cox proportional hazards models. Participants who did not respond to baseline questionnaire, did not provide information on aspirin use on the baseline questionnaire, or reported a history of cancer at baseline were excluded from analyses. Of 121,700 enrolled female registered nurses, 88,378 (72.6%) were found eligible for follow-up. Over an 18-year period, there was a total of 1,475,262 person-years in the cohort. At baseline, 43% of all participants reported any current, regular aspirin use (ⱖ2 tablets/week). This group had a higher prevalence of obesity and diabetes mellitus compared with those who did not report aspirin use. There were a total of 161 incident pancreatic cancers among eligible participants during the 18-year follow-up. Overall, there was no significant increase in risk of pancreatic cancer in all those who reported regular (ⱖ2 tablets/week) aspirin use compared with non-users. Significant increase in risk for pancreatic cancer was noted with aspirin use of ⱖ14 tablets/week (RR, 1.86; 95% CI, 1.03–3.35), regular aspirin use of ⱖ2 tablets/week for ⬎20 years (RR, 1.58; 95% CI, 1.03–2.43), or

SELECTED SUMMARIES

1003

ⱖ5 aspirin tablets/week for ⬎10 years (RR, 1.75; 95% CI, 1.18 –2.60). There was no association between pancreatic cancer and aspirin use in patients with a normal BMI (⬍25 kg/m2). Patients with BMI ⬎25 kg/m2 who used aspirin had a higher risk for pancreatic cancer compared with nonusers with a BMI ⬎25 kg/m2 (RR, 1.78; 95% CI, 1.08 –2.94). The authors conclude that regular use of aspirin for extended periods of time seems to be associated with a statistically significant increase in risk of pancreatic cancer among women. Comment. Aspirin and other NSAIDs inhibit the enzymatic activity of cyclooxygenase (COX), which is the rate-limiting enzyme for the synthesis of eicosanoids, such as prostaglandins, from arachidonic acid. Two isoforms of COX have been identified: constitutively expressed COX-1 and inducible COX-2. COX-1 is a housekeeping gene and has an important role in protecting the gastroduodenal mucosa. The COX-2 gene, an immediate early-response gene, is rapidly induced in response to tumor promoters, cytokines, and growth factors. Aspirin nonselectively inhibits both COX-1 and COX-2. COX-2 contributes to carcinogenesis through several mechanisms: increase production of prostaglandins, convert procarcinogens to carcinogens, inhibit apoptosis, promote angiogenesis, increase invasiveness of cancer cells, and modulate inflammation and immunoresponse (J Cancer Res Clin Oncol 2001;127:411– 417). The anticarcinogenic effect of NSAIDs involves their capacity to inhibit COX activity, leading to reduced immunosuppressive prostaglandin synthesis, reduced production of potential mutagens, and cell signaling substances from arachidonic acid metabolism. Other proposed antiproliferative pathways include cell cycle arrest, induction of apoptosis, inhibition of angiogenesis, and regulation of the mismatch repair protein (Clin Cancer Res 2003;9:383–390). NSAIDs are an accepted chemotherapeutic option for the prevention of colon cancer, but evidence for protection of other forms of cancer is debated (BMJ 2000;320:1642– 1646). In the Nurses’ Study reported by Schernhammer et al. (J Natl Cancer Inst 2004;96:22–28), after correcting for age, smoking, body mass index, and level of activity, consistent and long-term aspirin use was associated with a modest increase in the risk of pancreatic cancer in those reporting use of 7–13 tablets of aspirin/week, consistent users of aspirin reporting ⬎14 tablets a day, and those reporting ⬎10 years of aspirin use. The risk of pancreatic cancer was not increased in those reporting current, regular use of aspirin, in those reporting current use of up to 6 tablets/week or ⱖ14 tablets/week, and those reporting use of aspirin for 10 years or less. The results of this study are in direct contrast to the results of the 4 case control studies and 2 prospective cohort studies that have previously evaluated the risk of pancreatic cancer in aspirin and NSAID users. A case control study of 504 patients with pancreatic cancer and 5952 controls reported a nonsignificant lower odds ratio for NSAID or aspirin users of 5 years or more (Cancer Epidemiol Biomarkers Prev 2000;9:119 –123). However, the small number of aspirin users in the study limited assessment of the effect of duration of use. An analysis of the cancer risk in the Cancer Prevention Study II showed no association between pancreas cancer risk and aspirin use (Cancer Res 1993;53:1322–1327). A case-control study using the UK general research database showed a nonsignificant dose-related increase in the risk of pancreatic cancer (BMJ 2000;320:1642–1646). A hospital-based case control study from Roswell Park Cancer Institute showed no change in pancreatic cancer risk in aspirin users, but data

1004

SELECTED SUMMARIES

on doses and other NSAIDs was lacking (BMC Public Health 2002; 2:18). A nonsignificant reduction in standardized incidence ratio for pancreatic cancer in women was reported in a cohort of 11,683 patients with rheumatoid arthritis followed for up to 20 years. Only 32 incident pancreatic cancers were observed in this study and results were not adjusted for smoking and other variables (J Natl Cancer Inst 1993;85:307–311). In the Iowa Women’s Health Study, an evaluation of a cohort 28,283 postmenopausal women showed a significant decrease in relative risk of pancreatic cancer in subjects who took aspirin, and the decrease was more significant with increased frequency of aspirin use (J Natl Cancer Inst 2002;94:1168 –1171). In this study, nonaspirin NSAIDs had no effect on the risk of pancreatic cancer. Thus, the 2 largest prospective cohort studies on aspirin use and risk of pancreatic cancer, the Iowa Women’s Health Study (J Natl Cancer Inst 2002;94:1168 –1171) and the Nurses’ Health Study (J Natl Cancer Inst 2004;96:22–28), have reached diametrically opposite conclusions. Both were carried out in women. Approximately 40% of both populations reported use of 2 or more tablets of aspirin/week. However, there are some striking differences between the 2 populations. While in the Iowa Women’s Health Study all participants were between 55 and 69 years old at entry, the women in the Nurses’ Health Study were between 30 and 55 years at baseline, with the average age at entry into the study being ⬃45 years. Not surprisingly, the incidence of pancreatic cancer in the Iowa Women’s Health Study is nearly 4 times that in the Nurses’ Health Study (40.4 vs. 10.9 per 100, 000 person-years of follow-up calculated from the number of incident cancers and number of person-years of follow-up reported in the studies). Although the mean age at diagnosis of pancreatic cancer is not mentioned in either study, one suspects that the cancers in the Nurses’ Health Study were diagnosed at a significantly younger age compared with those in the Iowa Women’s Health Study. The percentage of participants who were current smokers in the Nurses’ Health Study is approximately twice that in the Iowa Women’s Health Study (12%–15% vs. 28.5%–30.8% in the various subgroups of aspirin users). Additionally, the groups in which the Nurses’ Health Study showed an increased risk of pancreatic cancer, i.e., those who consistently used excessive amounts of aspirin (ⱖ14 tablets/week) and those who used it for prolonged periods of time (⬎20 years) were not examined in the Iowa Women’s Health Study. It is unclear how the differences in the study populations affected the results reported in the 2 studies. Could it be, as the authors contend, that the Nurses’ study reflects a procarcinogenic effect of persistent use of high doses of aspirin in middle-aged women? Unfortunately, the authors’ biologic explanations for why aspirin might predispose to pancreatic cancer are weak at best. The authors quote 2 studies on the role of aspirin and NSAIDs in causing acute pancreatitis and extrapolate this to state that aspirin might be causing “subclinical” chronic pancreatitis and thereby predispose to pancreatic cancer. This line of argument is purely speculative and not based on any scientific evidence. Aspirin is not a drug commonly associated with acute pancreatitis, and aspirin (or any other drug) has never been shown to be the cause of chronic pancreatitis. The role of aspirin in disturbing the balance between cyclooxygenase and lipoxygenase and enhanced expression of procarcinogenic lipoxygenases, another possible explanation for the carcinogenic effect of aspirin suggested by the authors, is also purely hypothetical. In summary, a review of the results of epidemiologic studies done so far, both case control and cohort studies, does not provide a

GASTROENTEROLOGY Vol. 127, No. 3

definitive answer as to the effect of aspirin consumption on the risk of pancreatic cancer. LUIS F. LARA, M.D. SURESH T. CHARI, M.D.

LYSINE AS A SEROTONIN RECEPTOR ANTAGONIST: USING THE DIET TO MODULATE GUT FUNCTION Smriga M, Torii K (Ajinomoto Co., Inc., Institute of Life Sciences, Kawasaki, Japan). L-Lysine acts like a partial serotonin receptor 4 antagonist and inhibits serotonin-mediated intestinal pathologies and anxiety in rats. Proc Natl Acad Sci U S A 2003;100:15370 –15375. The ability of nutrient components of the diet to modulate gut motor function has long been recognized. In most cases, these effects are mediated by physical characteristics or by nonselective action on intestinal chemo- or osmoreceptors. Other nutrients serve as neurotransmitter precursors; however, no research to date has suggested that any dietary component acts to modify gut contractile activity by action on a specific receptor subtype. In the present study, Smriga and Torii investigate the effects of L-lysine in animal models of serotonin-mediated, stress-induced diarrhea. Their findings suggest that this essential amino acid serves as a partial antagonist of gut serotonin 5-HT4 receptors to reduce stress-related diarrhea as well as anxiety. In the first series of experiments, in vitro effects of L-lysine on intestinal serotonin receptor function were tested. In radioligand binding studies, L-lysine (0.8 mmol/dL) inhibited serotonin binding to guinea pig corpus striatum 5-HT4 receptors by 9.7% but had no effects on binding to other serotonin receptors. At 100fold higher concentrations (80 mmol/dL), which are nonphysiologic, lysine blocked ⬎40% of serotonin binding to 5-HT4 receptors but also inhibited binding to 5-HT1A, 5-HT2A, and 5-HT2B receptors. Binding results were compared to effects of L-lysine on serotonin-evoked ileal contractions. By itself, L-lysine had no contractile action on guinea pig ileal segments indicating a lack of serotonin receptor agonism. Conversely, L-lysine at a concentration (0.7 mmol/dL) similar to that which blocked 5-HT4 receptors reduced serotonin-evoked ileal contractions by 25% (P ⬍ 0.01), confirming its action as a serotonin receptor antagonist. A related amino acid, L-leucine, did not modify serotonin-elicited contractions showing the specificity of L-lysine actions. In vitro actions of lysine on serotonin receptors were correlated with its effects on serotonin-mediated, stress-induced diarrhea. L-lysine had no effect on fecal output in rats exposed to no stress. Experimental stress induced by wrapping the extremities and thorax of rats with tape increased fecal weight but did not induce diarrhea. Oral L-lysine (1 g/kg) decreased fecal weight in the second hour of restraint to 0.15 g vs. 0.45 g in rats given water (P ⬍ 0.05). Eighty percent of rats first exposed to restraint stress and then given a subcutaneous

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.