Smoking as a Cofactor for Causation of Chronic Pancreatitis A Meta-Analysis Angelo Andriulli, MD,* Edoardo Botteri, MSc,Þ Piero L. Almasio, MD,þ Italo Vantini, MD,§ Generoso Uomo, MD,|| and Patrick Maisonneuve, Eng,Þ on behalf of an ad hoc committee of the Italian Association for the Study of the Pancreas Objectives: To assess the evidence for tobacco smoking as a risk factor for the causation of chronic pancreatitis. Methods: We performed a meta-analysis with random-effects models to estimate pooled relative risks (RRs) of chronic pancreatitis for cur- rent, former, and ever smokers, in comparison to never smokers. We also performed dose-response, heterogeneity, publication bias, and sensitivity analyses. Results: Ten case-control studies and 2 cohort studies that evaluated, overall, 1705 patients with chronic pancreatitis satisfied the inclusion criteria. When contrasted to never smokers, the pooled risk estimates for current smokers was 2.8 (95% confidence interval [CI], 1.8Y4.2) overall and 2.5 (95% CI, 1.3Y4.6) when data were adjusted for alcohol consumption. A dose-response effect of tobacco use on the risk was ascertained: the RR for subjects smoking less than 1 pack per day was 2.4 (95% CI, 0.9Y6.6) and increased to 3.3 (95% CI, 1.4Y7.9) in those smoking 1 or more packs per day. The risk diminished significantly after smoking cessation, as the RR estimate for former smokers dropped to a value of 1.4 (95% CI, 1.1Y1.9). Conclusions: Tobacco smoking may enhance the risk of developing chronic pancreatitis. Recommendation for smoking cessation, besides al- cohol abstinence, should be incorporated in the management of patients with chronic pancreatitis. Key Words: chronic pancreatitis, smoking, cigarette, nicotine, alcohol, meta-analysis (Pancreas 2010;39: 1205Y1210) D uring the past 2 decades, several experimental evidences indicated that cigarette/nicotine smoking induces marked pathological and functional changes in the exocrine pancreas. 1 In humans, serum pancreatic enzymes after intravenous secre- tin injection increase substantially in smokers, but not in non- smokers. 2 Mechanisms by which nicotine induces such changes are mediated via signal transduction pathways in the pancreatic acinar cell, leading to enhanced levels of intracellular calcium release 1 and/or impairment of pancreatic blood flow. 3 Yet, only few pancreatic specialists advise alcoholic patients with chronic pancreatitis for smoking cessation treatment, besides alcohol abstinence. Cigarette smoking has been suggested as the single most important factor for the development of pancreatic cancer, caus- ing a 75% increase in the risk compared with nonsmoking. 4 In contrast, epidemiological data linking cigarette smoking to the development of chronic pancreatitis offer conflicting con- clusions. 5Y20 Most studies indicate smoking as one of the can- didate susceptibility factors for pancreatitis, 5Y7,10Y17,20 whereas others have denied such an association. 8,9,18 The close inter- action between tobacco consumption and alcohol intake 21 with heavy drinkers often being heavy smokers 22,23 renders prob- lematic to ascertain the relative pathogenic role of the 2 factors in tissue injury. Moreover, methodological differences among existing epidemiological studies cast doubt on the findings in this topic. Many of the positive reports have been confounded by the retrospective nature of data collection, which impacts significantly on the ability to accurately assess alcohol intake or cigarette-smoke exposure. 24 Furthermore, the lack of appro- priate controls is crucial for any study examining individual susceptibility to alcoholic pancreatitis: it has been argued that the best comparison for such studies must be between alco- holics with the disease and alcoholics without the disease so that the index and the control groups differ only in 1 variable, that is, the presence or absence of pancreatitis. 8,25 Unfortunately, numerous studies in the literature have used only the general population as controls, making the interpretation of results controversial. 7,12,14,19,20 The current review focuses on important epidemiological observation linking tobacco smoking to chronic pancreatitis development and provides insightful interpretation of what is known. The aims of this study were to provide a statistical evaluation using a meta-analytic approach of the strength of the association between tobacco smoking and chronic pancreatitis and to assess whether the association depends on study char- acteristics such as the control population, study designs, or level of exposure. MATERIALS AND METHODS Search Strategy, Inclusion Criteria, and Data Abstraction We searched PubMed from 1966 to July 15, 2009, for the following terms: ([smoke OR cigarette OR tobacco OR smoking] AND [(chronic pancreatitis OR alcoholic chronic pancreatitis OR pancreatitis)]. Our search was restricted to human studies published in English language. We searched the reference lists of articles and reviews on the topic to identify additional relevant studies. Only reports complying with the following inclusion cri- teria were included in the meta-analysis. First, studies should ORIGINAL ARTICLE Pancreas & Volume 39, Number 8, November 2010 www.pancreasjournal.com 1205 From the *Division of Gastroenterology, BCasa Sollievo Sofferenza[ Hospital, IRCCS, San Giovanni Rotondo; †Division of Epidemiology and Biostatistics, European Institute of Oncology, Milan; ‡Gastroenterology Unit, University of Palermo, Palermo; §Gastroenterology Unit, University of Verona, Verona; and ||Department of Internal Medicine, BA. Cardarelli[ Hospital, Naples, Italy. Received for publication November 17, 2009; accepted March 11, 2010. Reprints: Angelo Andriulli, MD, Division of Gastroenterology, BCasa Sollievo Sofferenza[ Hospital, IRCCS, viale Cappuccini 1, 71013 San Giovanni Rotondo, Italy (e-mail: a.andriulli@operapadrepio.it). The authors declare no competing interests. Writing assistance: None. Copyright * 2010 by Lippincott Williams & Wilkins Copyright @ 20 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. 10