Alcohol and cigarettes: Partners in crime in chronic pancreatitis

Abstract

Maisonneuve P, Lowenfels AB, Müllhaupt B, Cavallini G, Lankisch PG, Andersen JR, DiMagno EP, Andrén-Sandberg Å, Domellöf L, Frulloni L, Ammann RW (Epidemiology Unit, European Institute of Oncology, Milan, Italy). Cigarette smoking accelerates progression of alcoholic chronic pancreatitis. Gut 2005;54:510–514. There exists a strong correlation between alcohol and tobacco use, with the prevalence of cigarette smoking in the alcohol-dependent population being as high as 80%–90% (Addict Behav 1988;13:185–190, Addict Behav 1989;14:273–278, Alcohol Clin Exp Res 1994;18:867–872, J Stud Alcohol 1990;51:130–135, J Addict Dis 1998;17:55–66). Cardiologists, gastroenterologists, otolaryngologists, and primary care physicians are faced daily with the problem of dual dependence on alcohol and tobacco. The reality that cigarette smoking and alcohol consumption occur so frequently together emphasizes the need to further investigate their interaction. The effect of the combined exposure to alcohol and tobacco smoke on normal body function and in causation of disease is complex and still poorly understood. Esophageal, hepatocellular, and head and neck cancers have been associated with tobacco and alcohol exposure. A multiplicative risk has been identified with oral/throat cancers, while in the other cancers such an association has not been as clearly elicited (Semin Oncol 1999;14:131–138, Int J Gastrointest Cancer 2003;33:3–14, Cancer Causes Control 2003;14:245–250). This study by Maisonneuve et al (Gut 2005;54:510–514) shows that smoking accelerates the progression of the natural history of alcohol-induced chronic pancreatitis. From their original historical cohort of 1552 chronic pancreatitis subjects created to determine the risk of pancreatic cancer in chronic pancreatitis (N Engl J Med 1993;328:1433-–437), the authors retrospectively analyzed the data collected on the 934 patients whose chronic pancreatitis was due to alcohol use and for whom smoking status was known. Using this data set, they investigated the effect of tobacco smoking on the onset and progression of alcoholic chronic pancreatitis. Smoking and alcohol consumption data were recorded at the onset of clinical disease. Patients were classified as “ever” and “never” smokers and “ever” smokers were further subclassified into < and > 1 pack of cigarettes per day. The duration of smoking (pack-year data) was not available and in 20% of smokers, no quantitation of smoking was possible due to lack of information. Alcohol consumption was quantified as low-moderate <5 drinks/day, heavy 5–9 drinks/day, very heavy >10 drinks/day, and the quantity was unknown in 27% of subjects. Information on the presence or absence of pancreatic calcification and diabetes was noted because these were considered the primary end-points for progression of disease. Calcification was identified on abdominal radiographs and diabetes was diagnosed by elevated blood sugars or abnormal glucose tolerance tests. Student t test was used to compare age at diagnosis of pancreatitis in smokers versus nonsmokers. Logistic regression analysis was used to evaluate the association between smoking and the presence of calcification or diabetes either at the time of diagnosis of chronic alcoholic pancreatitis or at the end of the follow-up period. Cox proportional hazards regression analysis was used to evaluate the role of smoking on the time to development of calcification or diabetes on follow-up. Of the 934 subjects with chronic alcoholic pancreatitis, 870 (93%) were smokers (93% of males and 91% of females). The majority of patients were heavy or very heavy alcohol drinkers (65% of nonsmokers vs 77% of smokers; P = .06). Smokers were diagnosed with chronic pancreatitis at a younger age than nonsmokers (mean age, 44.6 ± 10.9 vs 49.3 ± 14.3 years; P = .001). Smokers were more likely to have calcification at the time of diagnosis compared to nonsmokers (odds ratio, 2.0; 95% CI, 1.1–3.8). The prevalence of diabetes at the time of diagnosis in smokers and nonsmokers was similar. The development of calcification and diabetes on follow-up after diagnosis was significantly associated with tobacco smoking (hazards ratio: 4.9, 95% CI, 2.3–10.5; and 2.3, 95% CI 1.2–4.2), respectively. When quantifying the tobacco used as less than or greater than 1 pack per day, there was not a significant difference in the development of either pancreatic calcifications or diabetes. To avoid confounding variables, a stratified analysis of alcohol consumption was performed that showed similar hazard ratios in all 3 groups of alcohol consumer (moderate, heavy, and very heavy) for both calcification and diabetes. The authors conclude that tobacco smoking was associated with earlier age at diagnosis of chronic alcoholic pancreatitis and greater risk of development of pancreatic calcifications and diabetes during the course of chronic pancreatitis, independent of alcohol consumption. Because the majority of patients with alcoholic chronic pancreatitis are moderate to heavy drinkers, it is not surprising that smoking is highly prevalent (>90%) among subjects with chronic alcoholic pancreatitis. Alcoholic chronic pancreatitis patients also smoke much more than either controls or patients with pancreatic adenocarcinoma (Dig Dis Sci 1999;44:1303–1311). The issues to consider are: (1) Does smoking alone alter pancreatic function and cause pancreatic disease? (2) What are the effects of concurrent smoking and alcohol consumption on the pancreas in general and the course of chronic pancreatitis in particular? (3) What are the clinical implications of the high prevalence of smoking in subjects with alcohol-induced chronic pancreatitis? Compared with the large number of studies on the effect of alcohol on the pancreas, the effects of smoking on pancreatic structure and function have received relatively little attention. Studies done in 1970s showed that smoking decreased pancreatic function (Gut 1972;13:361–365, Gastroenterology 1971;60:1098–1105, Gastroenterology 1977;73:758–761). A few studies have investigated the adverse effects of nicotine on pancreatic acini (Am J Physiol 1985;248:G158–163, Pancreas 1990;5:222–229). The effect of smoking on pancreatic function, both acutely and long term, clearly needs further investigation. Smoking has consistently been shown to be a risk factor for pancreatic cancer. Similarly, smoking has also been suggested to be an independent risk factor for chronic pancreatitis (Am J Epidemiol 1982;116:1407–1414, Gut 1991;32:1392–1395, Dig Dis Sci 1999;44:1303–1311, Pancreas 1996;12:131–137) although the strength of the evidence is much weaker. These studies have mostly been conducted in patients with alcoholic pancreatitis, and no study has directly answered the question of whether cigarette smoking is an independent risk factor for chronic pancreatitis in the absence of chronic alcoholism. Only 5%–10% of alcoholics ever develop evidence of pancreatitis. In an accompanying editorial to the study by Maisonneuve et al (Gut 2005;54:510–514), Apte et al (Gut 2005;54:446–447) summarize the (negative) results so far in the search for the co-factor(s) to explain why only a small minority of alcoholics develop pancreatitis. These include genetic factors, beverage use, drinking pattern, dietary nutrients, lipid profile, and adiposity. The question of whether smoking is necessary for the development of alcoholic chronic pancreatitis has been raised before (Pancreas 1993;8:568–572), but remains unanswered. Because of the strong association between heavy alcohol consumption and smoking, it is hard to separate the role of smoking from the effect of alcohol in causation of chronic pancreatitis. One way to overcome this problem is to study a large cohort with sufficient numbers of nonsmokers, nonalcoholics, and patients with pancreatitis. A recently published study did exactly this. Morton et al (Am J Gastroenterol 2004;99:731–738) studied 129,000 prepaid health plan workers who had satisfactorily completed a research questionnaire, including a special alcohol questionnaire. Among this group they identified 439 persons with pancreatitis. The pancreatitis was deemed to be due to gallstones in 168, due to alcohol in 125, and was considered to be idiopathic in 110. In this study (Am J Gastroenterol 2004;99:731–738), smoking was strongly associated with alcoholic pancreatitis even after for adjustment for confounding factors (age, sex, race, and alcohol amount). The relative risk (RR) was 4.9 (95% CI, 2.2–11.2) for subjects smoking ≥1 pack/day and the RR was lower in those with lesser tobacco use (RR in ex-smokers, 1.7 [CI, 0.9–3.1], and in those with <1 pack/day, 2.7 [CI, 1.5–4.9]). A significant but weaker association was seen in subjects with idiopathic pancreatitis (RR, 3.1 [CI 1.4–7.2] for subjects smoking ≥1 pack/day) but not with gallstone pancreatitis (RR, 1.3 [CI 0.5–3.1] for subjects smoking >1 pack/day) (Am J Gastroenterol 2004;99:731–738). In a study of the Native American population in Phoenix, Arizona, a surprisingly low incidence of chronic pancreatitis was observed despite a high prevalence of chronic alcoholism and many alcohol-related disorders, such as cirrhosis (Pancreas 1987;2:694–697). Interestingly, prevalence of cigarette smoking is low in this population. The strong dose-dependent association between smoking and alcoholic pancreatitis in the Kaiser Permanante study (Am J Gastroenterol 2004;99:731–738) and the observation of low prevalence of alcoholic pancreatitis in a population with a low prevalence of smoking (Pancreas 1987;2:694–697) suggest that smoking is an important risk factor for pancreatitis in alcoholic subjects. What are the ill effects of smoking in alcoholic subjects developing pancreatitis? The current and previous studies (Gut 1991;32:1392–1395, Pancreas 1994;9:42–46) have shown an association between smoking and earlier development of calcification in alcoholic subjects who smoke compared to nonsmokers. A similar association between smoking and calcification has been reported in patients with idiopathic chronic pancreatitis (Pancreas 2000;21:115–119). Cavallini et al (Pancreas 1994;9:42–46) have suggested that this may be related to the reduction of lithostatin-S (LS-S) protein which is responsible for stabilizing pancreatic calcium. Nicotine, acetaldehyde, and cyanide, all metabolites of tobacco smoke, have been proposed as factors responsible for a reduction in the levels of LS-S protein, thereby favoring calcification. Further studies are needed to clarify this interaction. The risk of cancer, both pancreatic and extra-pancreatic, is increased in all forms of chronic pancreatitis (Dig Dis Sci 1999;44:1303–1311). Much of this excess risk of cancer can be attributed to smoking (N Engl J Med 1979;300:213–217, Ann Intern Med 1981;95:139–145, Arch Intern Med 1996;156:22552260, JAMA 2001;286:169–170). Pancreatic cancer is the most dreaded complication of chronic pancreatitis. It carries a poor prognosis and, despite the knowledge that chronic pancreatitis predisposes patients to pancreatic cancer, there is currently no reliable screening test to detect pancreatic cancer early in patients with chronic pancreatitis. However, there is a measure that patients can take to reduce their risk of developing cancer, ie, stop smoking. In summary, there is evidence that smoking adversely affects pancreatic function and structure and it clearly increases the risk of both pancreatic and extra-pancreatic cancers. It is, therefore, important for physicians to educate patients with pancreatitis about abstaining not only from alcohol, but also from smoking. However, multiple studies have reported the difficulty with concurrent treatment of alcohol and tobacco dependency (Drug Alcohol Depend 2003;69:1–7, J Stud Alcohol 2004;65:681–691). These studies report limited success when attempting to make patients refrain from tobacco smoking while undergoing alcohol treatment programs. Although studies have shown the difficulty of simultaneous abstinence, the interest in quitting both alcohol and tobacco has been reported as high as 75% among subjects coming to alcohol rehabilitation (Subst Use Misuse 1998;33:2021–2047, Subst Use Misuse 2002;37:1275–1298). If properly educated on the role that tobacco smoking plays in the development and course of chronic alcoholic pancreatitis, patients may be more inclined to participate in smoking cessation programs. Reply to ChariGastroenterologyVol. 129Issue 5PreviewThere is no doubt that the strong correlation existing between smoking and drinking makes it difficult to evaluate whether tobacco alone is a risk factor or risk modifier for alcoholic chronic pancreatitis. In other conditions such as upper digestive tract cancer, alcohol and tobacco can act synergistically (Cancer Research 1990;50:6502–6507); this synergistic risk effect could be, in part, associated with the markedly increased exposure of upper digestive tract mucosa to carcinogenic salivary acetaldehyde. Full-Text PDF