Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Feb 1.
Published in final edited form as: Cancer Epidemiol. 2017 Jan 2;46:80–84. doi: 10.1016/j.canep.2016.12.006

Proton pump inhibitors on pancreatic cancer risk and survival

Malcolm D Kearns 1, Ben Boursi 3,4,5,6, Yu-Xiao Yang 2,3,4
PMCID: PMC5303431  NIHMSID: NIHMS840336  PMID: 28056391

Abstract

Background

Hypergastrinemia may promote the development and progression of pancreatic cancer. Proton pump inhibitor (PPI) therapy is known to cause hypergastrinemia. We sought to determine the association between PPI therapy and the risk of developing pancreatic cancer as well as survival following pancreatic cancer diagnosis.

Methods

We conducted a nested case-control study and a retrospective cohort study in The Health Improvement Network (THIN), a medical records database representative of the UK population. In the case-control study, each patient with incident pancreatic cancer was matched with up to four controls based on age, sex, practice site and both duration and calendar time of follow-up using incidence density sampling. The odds ratios (ORs) and 95% confidence intervals (CIs) for pancreatic cancer risk associated with PPI use were estimated using multivariable conditional logistic regression. The retrospective cohort study compared the survival of pancreatic cancer patients according to their PPI exposure at the time of diagnosis. The effect of PPI use on pancreatic cancer survival was assessed using a multivariable Cox regression analysis.

Results

The case-control study included 4,113 cases and 16,072 matched controls. PPI use was more prevalent in cases than controls (53% vs. 26% active users). Adjusting for diabetes, smoking, alcohol use and BMI, PPI users including both former users and active users with longer cumulative PPI use had a higher risk of pancreatic cancer compared to non-users. When assessing survival following pancreatic cancer diagnosis, only short-term, active users had a modest decrease in survival.

Conclusions

Long-term PPI therapy may be associated with pancreatic cancer risk. While PPI users recently started on treatment had a slightly worse survival, this result likely is from reverse causation.

Keywords: pancreatic cancer, proton pump inhibitor, risk, survival, epidemiology

1. Introduction

Proton pump inhibitors (PPIs), which are among the most prescribed medications worldwide (1), may influence the risk of gastrointestinal (GI) malignancies, including pancreatic cancer. The mechanism through which PPIs may increase cancer risk is related to the pathway by which they provide therapeutic benefit; PPIs inactivate the H+/K+ ATPase (or proton pump) on parietal cells in the stomach, thus reducing gastric acid secretion. Acid suppression creates a strong stimulus for gastrin (a trophic factor) production by G cells (2) in nearly all patients on long-term PPI therapy (3, 4). Hypergastrinemia may be associated with enterochromaffin-like (ECL) cell hyperplasia (5) and tumorogenesis (6–9), gastric tumors (10–12) and Barrett’s epithelium (13) in in vitro and animal models. Gastrin has been shown to stimulate the growth of human pancreatic cancer cells in cultures (3, 14–17) and pancreatic tumors transplanted into nude mice (18). These effects are likely mediated through the gastrin receptor, which has been found on human pancreatic cancer cells (19). Gastrin-receptor antagonists prevent growth of pancreatic cancer cells (18) and gastrazole, a gastrin inhibitor, increased survival time as a cancer treatment in a small number of patients (20) (though this was refuted in another study (21)). Furthermore, successful antibody production to gastrin (following exposure to a diphtheria toxoid-coupled vaccine) was associated with survival benefits in patients with in pancreatic cancer (22) and colorectal cancer (23).

In spite of a highly suggestive mechanism demonstrated in experimental models linking PPI use, and increased gastrin levels to GI cancers, the results of epidemiologic studies have been mixed; several studies showed increased rates of gastric cancer among PPI users (24–26) while other studies found no link between acid suppression and gastric cancers (27, 28), colorectal cancer (29–32) or pancreatic cancer (33). In this study, we evaluated the impact of PPI use on both the risk of pancreatic cancer and survival after diagnosis in a large, population-based cohort. Elucidating the association between pancreatic cancer and PPI use could help advance our understanding of the pathogenesis of pancreatic cancer, specifically regarding the role of gastrin. It would also provide important data to help patients and prescribers weigh the risk and benefit of long-term PPI therapy.

2. Materials and Methods

2.1 Study design

We conducted a nested case-control study to determine the effect of PPI exposure on pancreatic cancer risk and a retrospective cohort study to evaluate the impact of PPI use on survival in subjects with pancreatic cancer.

2.2 Data source

The Health Improvement Network (THIN) is a medical records database that contains records from approximately 10 million patients treated in >570 general practices in the UK. Its population has been shown to be representative of the general population of the UK (34). General practitioners have been trained to record their medical diagnoses as READ codes (35) using the Vision general practice computer system (In Practice Systems, London, UK) for the collection of THIN data. The data is entered using a standardized protocol and is routinely analyzed for quality control (34, 36). In our study, we searched for medical diagnoses (e.g. pancreatic cancer, diabetes, alcohol use) using specific READ diagnostic codes (37), and PPI prescriptions were identified using multiplex codes. A recent study in THIN showed that 97% of the incident pancreatic cancer cases identified using READ codes was confirmed based on manual chart review (38).

2.3 The effect of PPI on pancreatic cancer risk, a case-control study

2.3.1 Study population

All patients receiving care from a practitioner using THIN between 1995 and 2013 were potentially eligible for inclusion. Subjects with a diagnosis of inflammatory bowel disease, familial pancreatic cancer syndromes or age below 40 years old at the time of diagnosis were excluded in order to focus on an average risk population. Patients without acceptable medical records (i.e., patients with incomplete documentation or out of sequence date of birth, registration date, date of death, or date of exit from the database) were also excluded.

2.3.2 Cases

Cases were individuals with at least one READ code for pancreatic cancer recorded >183 days after they were either enrolled in a THIN practice (34, 39) or that the practice started using Vision software. The 183-day lag was implemented in order to ensure that only incident pancreatic cancer cases were included (40).

Controls

Up to 4 controls were matched with each case using incidence density sampling (41) based on: age, sex, practice site and both duration and calendar time of follow-up. The controls were assigned the same index date as their matched cases.

2.3.4 Exposure

The exposure of interest was PPI use prior to index date. Individuals without a multiplex code for a PPI were considered unexposed. Reverse causation can occur in case-control studies when a treatment administered for the first symptoms of a disease can appear to cause that disease. We attempted to capture the effect of this bias by stratifying groups based on the timing of their PPI prescriptions prior to pancreatic cancer diagnosis: former users (most recent PPI prescription >6 months prior to index date) and active users (most recent PPI prescription < 6 months prior to index date). Active users were further separated into: 1) short-term, active users (first prescription <12 months before the index date), 2) intermediate-term, active users (first prescription between 12 and 24 months before the index date) and 3) long-term, active users (first prescription >24 months before index date).

2.3.5 Covariates and confounders

We examined a list of variables known or suspected to affect pancreatic cancer risk (e.g., type 2 diabetes (42), cigarette smoking (43), alcohol use (44, 45)) and potential confounders associated with both pancreatic cancer and PPI use (i.e., obesity (46–48)). All variables were measured prior to the index date and defined as follows: obesity (BMI >30 mg/kg2), smoking and alcohol use (as identified by the presence of diagnosis codes entered into THIN by providers). Additional data regarding amount of use (for example number of cigarettes or alcoholic drinks per day) were not extracted given concerns over completeness of this data set and small numbers of individuals in each category. We adjusted our analyses for these variables.

2.3.6 Statistical analysis

The baseline characteristics of cases and controls were compared using Pearson’s chi-squared test for categorical variables and Student’s t test for continuous variables. The association between PPI use and the risk of pancreatic cancer was assessed using univariate and multivariable conditional logistic regressions to estimate odds ratios (ORs) and 95% confidence intervals (CI). All p-values were two-sided and values <0.05 were considered significant.

2.4 The effect of PPI use on pancreatic cancer risk, a retrospective cohort study

2.4.1 Study population

All individuals from the above case-control study with at least one READ code for pancreatic cancer 183 days after they were either enrolled in the clinic or that the practice started using Vision computer system/software were included.

2.4.2 Exposure

PPI exposure status at the time of the pancreatic cancer diagnosis was categorized using the same approach as the nested case-control study as 1) former users, 2) short-term, active users, 3) intermediate-term, active users, 4) long-term, active users and 5) non-users.

2.4.3 Covariates and confounders

History of smoking, alcohol use, diabetes and obesity were examined in this population as defined above.

2.4.4 Outcomes

This study evaluated survival following pancreatic cancer diagnosis in groups that were exposed and unexposed to PPIs. The follow-up period started on the pancreatic cancer diagnosis date and ended on the date of death recorded in the THIN database, or the earliest of the following: transferring out of practice or end of THIN follow-up.

2.4.5 Statistical analysis

The survival analysis was performed using a multivariate Cox regression analysis, which estimated hazard ratios (HRs) and 95% CI. All p-values were two-sided and values <0.05 were considered significant. Analyses were adjusted as described above.

3. Results

The nested-case-control study included 4,113 patients with pancreatic cancer and 16,072 matched controls. The prevalence of PPI use (active or former) was greater in individuals with a diagnosis of pancreatic cancer, with 52.9% (2175 of 4113 patients) of cases being active users compared to 26.2% (4217 of 16072 patients) of controls and 43.8% (1801 of 4113 patients) of cases being never users compared to 72.0% (11576 of 16072 patients) of controls. As expected, cases were more likely to be obese, smokers, use alcohol, or have a diagnosis of diabetes (Table 1).

Table 1.

Characteristics of pancreatic cancer cases and controls.

Cases Controls ORs (95% CI)
n (%) 4113 (20.4) 16072 (79.6)
age (mean+/− SD) yrs 70.9 +/− 11.5 71.1 +/− 11.4
Gender (n, % female) 1,999 (48.6) 7,794 (48.5)
obesity(n, %)1 861 (20.9)* 3013 (18.7) 1.15 (1.06–1.26)
smoking(n, %)2 2053 (49.9)** 6625 (41.22) 1.52 (1.41–1.64)
alcohol use (n, %)3 2135 (51.9)** 7706 (48.0) 1.23 (1.13–1.34)
diabetes (n,%)4 931 (22.6)** 1598 (9.9) 2.74 (2.50–3.01)
most recent PPI prescription (mean+/− SD)5 1.51+/− 0.85 1.54+/− 0.79 1 (0.94–1.06)
Duration of follow-up (year+/− SD) 6.33+/− 4.09 6.36 +/− 4.10 0.99 (0.68–1.45)
Duration of PPI use (year+/− SD) 2.47+/− 3.07 4.14 +/− 3.19 0.85 (0.83–0.87)
Open in a new tab

n, number of patients; yrs, years; SD, standard deviation; ORs, Odds Ratios; 95% CI, 95% Confidence Intervals;

1

BMI >30 mg/kg2;

2

lifetime smoker;

3

lifetime alcohol use;

4

READ code for diabetes;

5

years prior to index date;

*

p= 0.001;

**

p=<0.0001

All PPI exposure categories were associated with the risk of pancreatic cancer in our unadjusted conditional logistic regression analysis (Table 2). Short-term, active users had the largest OR (10.60, 95% CI 9.36–11.79) while former users (OR 3.54, 95% CI 2.83, 4.43), intermediate-term (OR 2.54, 95% CI 2.16, 3.00), and long-term active users (OR 1.94, 95% CI 1.74, 2.14) all had lower but significantly elevated risks when compared to non-users. When adjusted for diabetes, obesity, alcohol use and smoking, the results were largely unchanged (Table 2). When duration of PPI use was analyzed as a continuous variable, individuals with pancreatic cancer had a significantly shorter duration of use compared to controls (2.47 vs. 4.14 years, OR 0.85, 95% CI 0.83–0.87).

Table 2.

PPI exposure and pancreatic cancer risk

Cases/Controls
N= (4113/16072)		 Crude OR (95%CI) adjusted1 OR (95% CI)
Never users 1801/11,576
Former users 137/279 3.54 (2.83–4.43) 3.36 (2.67–4.22)
Short-term, active users 1109/705 10.50 (9.36–11.79) 10.42 (9.26–11.73)
Intermediate-term, active users 243/629 2.54 (2.16–3.00) 2.47 (2.09–2.92)
Long-term, active users 823/2883 1.94 (1.75–2.14) 1.85 (1.67–2.06)
Open in a new tab

N=, number of individuals; OR, odds ratio; 95% CI, 95% confidence interval;

1

Adjusted for: diabetes, smoking, alcohol use and obesity

Stratification:
  1. former users: most recent PPI prescription >6 months prior to index date
  2. active users: most recent PPI prescription < 6 months prior to index date
    1. short-term, active users: first prescription <12 months before the index date
    2. intermediate-term, active users: first prescription 12 –24 months before the index date
    3. long-term, active users: first prescription >24 months before index date

With regard to the association between PPI use and survival after pancreatic cancer, the unadjusted values showed that only short-term active users experienced a significant increase in death rate (HR 1.11, 95% CI 1.02–1.21). Results were nearly identical when adjusted for diabetes, smoking, alcohol use and obesity (Table 3).

Table 3.

The effect of PPI use on survival in patients with pancreatic cancer

Crude HR (95%CI) adjusted1 HR (95% CI)
Former users 1.15 (0.94–1.40) 1.12 (0.92–1.37)
Short-term, active users 1.11 (1.02–1.21) 1.11 (1.02–1.21)
Intermediate-term, active users 0.96 (0.82–1.11) 0.95 (0.82–1.11)
Long-term, active users 1.03 (0.94–1.14) 1.02 (0.93–1.13)
Open in a new tab

N=, number of individuals; HR, hazard ratio; 95% CI, 95% confidence interval;

1

Adjusted for: diabetes, smoking, alcohol use and obesity

Stratification:
  1. former users: most recent PPI prescription >6 months prior to index date
  2. active users: most recent PPI prescription < 6 months prior to index date
    1. short-term, active users: first prescription <12 months before the index date
    2. intermediate-term, active users: first prescription 12 – 24 months before the index date
    3. long-term, active users: first prescription >24 months before index date

4. Discussion

We evaluated the effect of PPI exposure on both pancreatic cancer risk and survival in a large, population-based study of more than 4,000 incident pancreatic cancer cases. We found PPI use to be associated with pancreatic cancer risk in our population regardless of the duration of PPI therapy. In our retrospective cohort study on PPI and pancreatic cancer survival, only short-term, active PPI users had a modestly increased mortality risk (1.11, 95% CI 1.02–1.21). Our findings are the first population-level study with results in accordance with the compelling link between hypergastrinemia and pancreatic cancer seen in vitro and in animal models. Although the only other large epidemiologic study to evaluate the link between PPI therapy and pancreatic cancer found no association (33), the time period during which this study collected data (between 1995 and 2006) may not include the dramatic increase in PPI use seen in recent years. This pervious study also included far fewer cases compared to our study (1,141 compared to 4, 113) (33).

Of note, the duration of PPI use was significantly shorter when comparing cases to controls (2.47 vs. 4.14 years, OR 0.85, 95% CI 0.83–0.87), likely representing a component of reverse causality. Some cases may have been started on PPI therapy for non-specific symptoms related to pancreatic cancer prior to cancer diagnosis. We accounted for this possible bias in our study design looking at different timing of PPI initiation. The elevated OR in all patients taking PPI, even starting >24 months prior to pancreatic cancer diagnosis argues against reverse causality as the main explanation for our results and suggests that PPIs themselves might increase the risk of pancreatic cancer to some degree. However, even in case of reverse causality, a new PPI prescription may serve as an indicator for an individual at higher risk for pancreatic cancer, even among individuals without known pre-disposing factors. The decreased survival among short term PPI users may be related to the fact that GI symptoms often represent a later stage of pancreatic cancer (49). However in the current study we lacked staging information, thus we were unable to test this hypothesis.

Our study had several strengths. It included a large number of incident pancreatic cases from a population-representative cohort. The exposure and outcome data from the THIN database were complete and accurate. Being a general practice electronic medical records system, THIN captures complete information on chronic medication use. Furthermore, the diagnosis of pancreatic cancer in THIN has been validated previously. Our study also has a duration of follow-up sufficient to capture incident cases of pancreatic cancer in spite of its long latency period (50). Potential limitations of our study include: generalizability (the THIN database is only generalizable to the UK population) and a lack of dosing information. However, inconsistent adherence to prescribed regimens or over-the-counter PPI use that fell outside our analysis would likely bias results towards the null.

In summary, while epidemiological data has been variable, our study shows a significant association between PPI use and pancreatic cancer although part of this association is secondary to reverse causality and non-specific abdominal symptoms prior to pancreatic cancer diagnosis. These data require confirmation in future studies, but they are consistent with the effect of gastrin on the pancreas and pancreatic cancer cells demonstrated in in vitro and animal models. The potential association between long-term PPI therapy and the risk of pancreatic cancer adds another reason for the judicious prescribing of PPI therapy in general. As many as two-thirds of hospitalized patients are prescribed PPIs without an appropriate indication (51) and furthermore are often continued on PPIs after discharge without an indication (52). On the other hand, our retrospective cohort study among pancreatic cancer patients seems to suggest that these patients should not be deprived of the benefit of PPI therapy if they have acid-related symptoms.

Highlights.

  • -Hypergastrinemia is mechanistically linked to pancreatic cancer risk.

  • -Proton pump inhibitors (PPIs) cause hypergastrinemia.

  • PPIs may be associated with an increased risk of pancreatic cancer.

  • -PPI use is not associated with survival in pancreatic cancer patients.

  • -PPIs should be prescribed judiciously but not withheld in patients with pancreatic cancer.

Acknowledgments

Funding: The work was supported by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1TR000003. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflicts of interest: none

Authorship Contribution Statement

Re: Proton pump inhibitor use on pancreatic cancer risk and survival: a case-control and cohort study
  1. Drs. Yang, Boursi and Kearns each made substantial contributions to all aspects of this study, including conception and design, acquisition of data, analysis and interpretation of data
  2. Drs. Yang, Boursi and Kearns jointly drafted the initial version of the manuscript and subsequently critically revised the manuscript for important intellectual content. They also jointly revised the manuscript for the resubmission.
  3. Drs. Yang, Boursi and Kearns give final approval of the version to be published.

References

  • 1.Fitzgerald RC, Omary MB, Triadafilopoulos G. Dynamic effects of acid on Barrett's esophagus. An ex vivo proliferation and differentiation model. J Clin Invest. 1996;98(9):2120–2128. doi: 10.1172/JCI119018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Klinkenberg-Knol EC, Festen HP, Jansen JB, et al. Long-term treatment with omeprazole for refractory reflux esophagitis: efficacy and safety. Ann Intern Med. 1994;121(3):161–167. doi: 10.7326/0003-4819-121-3-199408010-00001. [DOI] [PubMed] [Google Scholar]
  • 3.Lamberts R, Creutzfeldt W, Stockmann F, et al. Long-term omeprazole treatment in man: effects on gastric endocrine cell populations. Digestion. 1988;39(2):126–135. doi: 10.1159/000199615. [DOI] [PubMed] [Google Scholar]
  • 4.Jansen JB, Klinkenberg-Knol EC, Meuwissen SG, et al. Effect of long-term treatment with omeprazole on serum gastrin and serum group A and C pepsinogens in patients with reflux esophagitis. Gastroenterology. 1990;99(3):621–628. doi: 10.1016/0016-5085(90)90946-x. [DOI] [PubMed] [Google Scholar]
  • 5.Hakanson R, Bottcher G, Sundler F, et al. Activation and hyperplasia of gastrin and enterochromaffin-like cells in the stomach. Digestion. 1986;35(Suppl 1):23–41. doi: 10.1159/000199380. [DOI] [PubMed] [Google Scholar]
  • 6.Kidd M, Tang LH, Modlin IM, et al. Gastrin-mediated alterations in gastric epithelial apoptosis and proliferation in a mastomys rodent model of gastric neoplasia. Digestion. 2000;62(2–3):143–151. doi: 10.1159/000007806. [DOI] [PubMed] [Google Scholar]
  • 7.Qvigstad G, Falkmer S, Westre B, et al. Clinical and histopathological tumour progression in ECL cell carcinoids ("ECLomas") APMIS. 1999;107(12):1085–1092. doi: 10.1111/j.1699-0463.1999.tb01513.x. [DOI] [PubMed] [Google Scholar]
  • 8.Waldum HL, Sandvik AK, Idle JR. Gastrin is the most important factor in ECL tumorigenesis. Gastroenterology. 1998;114(5):1113–1115. doi: 10.1016/s0016-5085(98)70346-4. [DOI] [PubMed] [Google Scholar]
  • 9.Mattsson H, Havu N, Brautigam J, et al. Partial gastric corpectomy results in hypergastrinemia and development of gastric enterochromaffin like-cell carcinoids in the rat. Gastroenterology. 1991;100(2):311–319. doi: 10.1016/0016-5085(91)90197-s. [DOI] [PubMed] [Google Scholar]
  • 10.Henwood M, Clarke PA, Smith AM, et al. Expression of gastrin in developing gastric adenocarcinoma. Br J Surg. 2001;88(4):564–568. doi: 10.1046/j.1365-2168.2001.01716.x. [DOI] [PubMed] [Google Scholar]
  • 11.Ho AC, Horton KM, Fishman EK. Gastric carcinoid tumors as a consequence of chronic hypergastrinemia: spiral CT findings. Clin Imaging. 2000;24(4):200–203. doi: 10.1016/s0899-7071(00)00199-6. [DOI] [PubMed] [Google Scholar]
  • 12.Poynter D, Pick CR, Harcourt RA, et al. Association of long last ing unsurmountable histamine H2 blockade and gastric carcinoid tumours in the rat. Gut. 1985;26(12):1284–1295. doi: 10.1136/gut.26.12.1284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Haigh CR, Attwood SE, Thompson DG, et al. Gastrin induces proliferation in Barrett's metaplasia through activation of the CCK2 receptor. Gastroenterology. 2003;124(3):615–625. doi: 10.1053/gast.2003.50091. [DOI] [PubMed] [Google Scholar]
  • 14.Bordi C, D'Adda T, Azzoni C, et al. Hypergastrinemia and gastric enterochromaffin-like cells. Am J Surg Pathol. 1995;19(Suppl 1):S8–S19. [PubMed] [Google Scholar]
  • 15.Cadiot G, Lehy T, Mignon M. Gastric endocrine cell proliferation and fundic argyrophil carcinoid tumors in patients with the Zollinger-Ellison syndrome. Acta Oncol. 1993;32(2):135–140. doi: 10.3109/02841869309083902. [DOI] [PubMed] [Google Scholar]
  • 16.Delle Fave G, Marignani M, Moretti A, et al. Hypergastrinemia and enterochromaffin-like cell hyperplasia. Yale J Biol Med. 1998;71(3–4):291–301. [PMC free article] [PubMed] [Google Scholar]
  • 17.Lehy T, Cadiot G, Mignon M, et al. Influence of multiple endocrine neoplasia type 1 on gastric endocrine cells in patients with the Zollinger-Ellison syndrome. Gut. 1992;33(9):1275–1279. doi: 10.1136/gut.33.9.1275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Smith JP, Fantaskey AP, Liu G, et al. Identification of gastrin as a growth peptide in human pancreatic cancer. Am J Physiol. 1995;268(1 Pt 2):R135–R141. doi: 10.1152/ajpregu.1995.268.1.R135. [DOI] [PubMed] [Google Scholar]
  • 19.Smith JP, Liu G, Soundararajan V, et al. Identification and characterization of CCK-B/gastrin receptors in human pancreatic cancer cell lines. Am J Physiol. 1994;266(1 Pt 2):R277–R283. doi: 10.1152/ajpregu.1994.266.1.R277. [DOI] [PubMed] [Google Scholar]
  • 20.Chau I, Cunningham D, Russell C, et al. Gastrazole (JB95008), a novel CCK2/gastrin receptor antagonist, in the treatment of advanced pancreatic cancer: results from two randomised controlled trials. Br J Cancer. 2006;94(8):1107–1115. doi: 10.1038/sj.bjc.6603058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Chu M, Kullman E, Rehfeld JF, et al. Effect of chronic endogenous hypergastrinaemia on pancreatic growth and carcinogenesis in the hamster. Gut. 1997;40(4):536–540. doi: 10.1136/gut.40.4.536. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Brett BT, Smith SC, Bouvier CV, et al. Phase II study of anti-gastrin-17 antibodies, raised to G17DT, in advanced pancreatic cancer. J Clin Oncol. 2002;20(20):4225–4231. doi: 10.1200/JCO.2002.11.151. [DOI] [PubMed] [Google Scholar]
  • 23.Smith AM, Justin T, Michaeli D, et al. Phase I/II study of G17-DT, an anti-gastrin immunogen, in advanced colorectal cancer. Clin Cancer Res. 2000;6(12):4719–4724. [PubMed] [Google Scholar]
  • 24.Bateman DN, Colin-Jones D, Hartz S, et al. Mortality study of 18 000 patients treated with omeprazole. Gut. 2003;52(7):942–946. doi: 10.1136/gut.52.7.942. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Garcia Rodriguez LA, Lagergren J, Lindblad M. Gastric acid suppression and risk of oesophageal and gastric adenocarcinoma: a nested case control study in the UK. Gut. 2006;55(11):1538–1544. doi: 10.1136/gut.2005.086579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Poulsen AH, Christensen S, McLaughlin JK, et al. Proton pump inhibitors and risk of gastric cancer: a population-based cohort study. Br J Cancer. 2009;100(9):1503–1507. doi: 10.1038/sj.bjc.6605024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Ahn JS, Eom CS, Jeon CY, et al. Acid suppressive drugs and gastric cancer: a meta-analysis of observational studies. World J Gastroenterol. 2013;19(16):2560–2568. doi: 10.3748/wjg.v19.i16.2560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Eslami L, Nasseri-Moghaddam S. Meta-analyses: does long-term PPI use increase the risk of gastric premalignant lesions? Arch Iran Med. 2013;16(8):449–458. [PubMed] [Google Scholar]
  • 29.van Soest EM, van Rossum LG, Dieleman JP, et al. Proton pump inhibitors and the risk of colorectal cancer. Am J Gastroenterol. 2008;103(4):966–973. doi: 10.1111/j.1572-0241.2007.01665.x. [DOI] [PubMed] [Google Scholar]
  • 30.Chubak J, Boudreau DM, Rulyak SJ, et al. Colorectal cancer risk in relation to use of acid suppressive medications. Pharmacoepidemiol Drug Saf. 2009;18(7):540–544. doi: 10.1002/pds.1749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Robertson DJ, Larsson H, Friis S, et al. Proton pump inhibitor use and risk of colorectal cancer: a population-based, case-control study. Gastroenterology. 2007;133(3):755–760. doi: 10.1053/j.gastro.2007.06.014. [DOI] [PubMed] [Google Scholar]
  • 32.Yang YX, Hennessy S, Propert K, et al. Chronic proton pump inhibitor therapy and the risk of colorectal cancer. Gastroenterology. 2007;133(3):748–754. doi: 10.1053/j.gastro.2007.06.022. [DOI] [PubMed] [Google Scholar]
  • 33.Bradley MC, Murray LJ, Cantwell MM, et al. Proton pump inhibitors and histamine-2-receptor antagonists and pancreatic cancer risk: a nested case-control study. Br J Cancer. 2012;106(1):233–239. doi: 10.1038/bjc.2011.511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Lewis JD, Schinnar R, Bilker WB, et al. Validation studies of the health improvement network (THIN) database for pharmacoepidemiology research. Pharmacoepidemiol Drug Saf. 2007;16(4):393–401. doi: 10.1002/pds.1335. [DOI] [PubMed] [Google Scholar]
  • 35.Chisholm J. The Read clinical classification. BMJ. 1990;300(6732):1092. doi: 10.1136/bmj.300.6732.1092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Bourke A, Dattani H, Robinson M. Feasibility study and methodology to create a quality-evaluated database of primary care data. Inform Prim Care. 2004;12(3):171–177. doi: 10.14236/jhi.v12i3.124. [DOI] [PubMed] [Google Scholar]
  • 37.Read J. Read clinical classification. BMJ. 1990;301(6742):45. doi: 10.1136/bmj.301.6742.45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lu Y, Garcia Rodriguez LA, Malgerud L, et al. New-onset type 2 diabetes, elevated HbA1c, anti-diabetic medications, and risk of pancreatic cancer. Br J Cancer. 2015;113(11):1607–1614. doi: 10.1038/bjc.2015.353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Garcia-Rodriguez LA, Huerta-Alvarez C. Reduced risk of colorectal cancer among long-term users of aspirin and nonaspirin nonsteroidal antiinflammatory drugs. Epidemiology. 2001;12(1):88–93. doi: 10.1097/00001648-200101000-00015. [DOI] [PubMed] [Google Scholar]
  • 40.Lewis JD, Bilker WB, Weinstein RB, et al. The relationship between time since registration and measured incidence rates in the General Practice Research Database. Pharmacoepidemiol Drug Saf. 2005;14(7):443–451. doi: 10.1002/pds.1115. [DOI] [PubMed] [Google Scholar]
  • 41.Lubin JH, Gail MH. Biased selection of controls for case-control analyses of cohort studies. Biometrics. 1984;40(1):63–75. [PubMed] [Google Scholar]
  • 42.Klein AP, Lindstrom S, Mendelsohn JB, et al. An absolute risk model to identify individuals at elevated risk for pancreatic cancer in the general population. PLoS One. 2013;8(9):e72311. doi: 10.1371/journal.pone.0072311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Bosetti C, Lucenteforte E, Silverman DT, et al. Cigarette smoking and pancreatic cancer: an analysis from the International Pancreatic Cancer Case-Control Consortium (Panc4) Ann Oncol. 2012;23(7):1880–1888. doi: 10.1093/annonc/mdr541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Zheng W, McLaughlin JK, Gridley G, et al. A cohort study of smoking, alcohol consumption, and dietary factors for pancreatic cancer (United States) Cancer Causes Control. 1993;4(5):477–482. doi: 10.1007/BF00050867. [DOI] [PubMed] [Google Scholar]
  • 45.Silverman DT, Brown LM, Hoover RN, et al. Alcohol and pancreatic cancer in blacks and whites in the United States. Cancer Res. 1995;55(21):4899–4905. [PubMed] [Google Scholar]
  • 46.Nothlings U, Wilkens LR, Murphy SP, et al. Body mass index and physical activity as risk factors for pancreatic cancer: the Multiethnic Cohort Study. Cancer Causes Control. 2007;18(2):165–175. doi: 10.1007/s10552-006-0100-0. [DOI] [PubMed] [Google Scholar]
  • 47.Michaud DS, Giovannucci E, Willett WC, et al. Physical activity, obesity, height, and the risk of pancreatic cancer. JAMA. 2001;286(8):921–929. doi: 10.1001/jama.286.8.921. [DOI] [PubMed] [Google Scholar]
  • 48.Hvid-Jensen F, Nielsen RB, Pedersen L, et al. Lifestyle factors among proton pump inhibitor users and nonusers: a cross-sectional study in a population-based setting. Clin Epidemiol. 2013;5:493–499. doi: 10.2147/CLEP.S49354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Porta M, Fabregat X, Malats N, et al. Exocrine pancreatic cancer: symptoms at presentation and their relation to tumour site and stage. Clin Transl Oncol. 2005;7(5):189–197. doi: 10.1007/BF02712816. [DOI] [PubMed] [Google Scholar]
  • 50.Brat DJ, Lillemoe KD, Yeo CJ, et al. Progression of pancreatic intraductal neoplasias to infiltrating adenocarcinoma of the pancreas. Am J Surg Pathol. 1998;22(2):163–169. doi: 10.1097/00000478-199802000-00003. [DOI] [PubMed] [Google Scholar]
  • 51.Forgacs I, Loganayagam A. Overprescribing proton pump inhibitors. BMJ. 2008;336(7634):2–3. doi: 10.1136/bmj.39406.449456.BE. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Zink DA, Pohlman M, Barnes M, et al. Long-term use of acid suppression started inappropriately during hospitalization. Aliment Pharmacol Ther. 2005;21(10):1203–1209. doi: 10.1111/j.1365-2036.2005.02454.x. [DOI] [PubMed] [Google Scholar]

RESOURCES