Gastroenterology
Volume 137, Issue 2 , Pages 412-415, August 2009

Do Diabetes Drugs Modify the Risk of Pancreatic Cancer?

  • Yu–Xiao Yang

      Affiliations

    • Corresponding Author InformationReprint requests Address requests for reprints to: Yu-Xiao Yang, MD, MSCE, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, 722 Blockley Hall, 423 Guardian Drive, Philadelphia, Pennsylvania 19104-6021

published online 29 June 2009.

Article Outline

 

See “Antidiabetic therapies affect risk of pancreatic cancer” by Li D, Yeung S–C J, Hassan MM et al, on page 482.

There is a growing body of literature suggesting that type 2 diabetes mellitus (DM) may be associated with the development of several cancers. Among these is pancreatic ductal adenocarcinoma, the 4th leading cause of cancer mortality in the United States.1 Pancreatic ductal adenocarcinoma is a highly lethal cancer, with a median survival of <6 months and a 5-year survival rate of <5%. Currently, there is no effective means to screen or prevent this cancer.

Many epidemiologic studies have investigated the association between DM and pancreatic cancer. However, this association is complex. A number of studies reported that the risk of pancreatic cancer is the highest among those with recent-onset diabetes2 and/or those DM patients with recent initiation of insulin therapy,3, 4 suggesting that pancreatic cancer may induce the onset of DM or worsen existing DM. Frequent resolution of DM after resection of the tumor provides further evidence for such reverse causality.2 By contrast, most studies that examined the risk of pancreatic cancer among patients with long-standing DM reported a moderately increased risk of pancreatic cancer risk in these patients.5 Such observations suggest a causal role of DM in pancreatic carcinogenesis. Possible mechanisms underlying the increased pancreatic cancer risk among patients with DM generally involve insulin resistance and hyperinsulinemia (Figure 1A).

  • View full-size image.
  • Figure 1. 

    (A) Diabetes mellitus, metformin, exogenous insulin and pancreatic cancer. IGF-I, insulin-like growth factor-I; AMPK, AMP-activated protein kinase. (B) Proposed mechanism of actions of metformin on AMPK and pancreatic cancer. AMPK, AMP-activated protein kinase; mTOR, mammalian target of rapamycin.

Beside reverse causality, efforts to elucidate the association between DM and pancreatic cancer face another complicating factor; both the progression of type 2 DM and use of diabetes medications can influence the extent of insulin resistance and hyperinsulinemia, sometimes in opposing directions. Type 2 DM is characterized by 2 fundamental defects: insulin resistance and a progressive impairment in insulin production. Before the clinical diagnosis of type 2 DM, endogenous hyperinsulinemia compensates for insulin resistance. As pancreatic β-cell function deteriorates, increased insulin secretion can no longer be maintained at a level sufficient to overcome insulin resistance, and clinically apparent hyperglycemia manifests. After the clinical onset of type 2 DM, dietary changes and weight reduction, in combination with oral agents, may be effective initially in maintaining glucose control. The mechanisms of action of these oral agents are not fully characterized, but they are thought to enhance insulin sensitivity (thereby reducing hyperinsulinemia; eg, metformin and the thiazolidinediones [TZDs]), decrease hepatic glucose production (eg, metformin), or stimulate pancreatic insulin production (eg, the sulfonylureas). With longer duration of DM, pancreatic β-cell function deteriorates further,6 and insulin resistance persists. A requirement for exogenous insulin then emerges.

Because the various DM medications can affect directly the key factors mediating the association between DM and pancreatic cancer, understanding the effect of anti-diabetic therapies on pancreatic cancer is a critical step in fully characterizing the role of type 2 DM in the development of pancreatic cancer. In addition, some of these medications, metformin in particular, may have anti-proliferative effects via mechanisms independent of it effect on insulin resistance and hyperinsulinemia. In this issue of the Gastroenterology, Li et al7 report the results of a hospital-based, case-control study that extended our knowledge on this issue. These authors identified 973 pathologically confirmed cases of pancreatic ductal adenocarcinoma at the University of Texas M.D. Anderson Cancer Center between 2004 and 2008. These cases were frequency matched with 863 controls recruited from healthy individuals accompanying patients being treated at other centers in the same institution. Self-reported information regarding the status and duration of DM diagnosis and DM medication use was collected from all participants. They found that ever use of metformin was associated with a significantly decreased risk of pancreatic cancer compared with metformin non-use (adjusted odds ratio [OR], 0.38; 95% confidence interval [CI], 0.22–0.69). By contrast, long-term insulin use among patients with long-standing DM was associated with a moderately higher risk of pancreatic cancer compared to insulin non-use (OR, 2.78; 95% CI, 1.00–7.73). The results on insulin secretagogues and TZDs were less conclusive.

A novel and riveting finding of this study is the apparent protective effect of metformin against pancreatic cancer. Several previous studies have investigated the association between oral DM medications as a single combined group and the risk of pancreatic cancer, but none of them specifically looked at metformin.3, 4, 8 Given the remarkably different physiologic effects of the various classes of oral DM agents, it is critical that they be examined separately in this context.

Metformin is a biguanide derivative approved for the treatment of type 2 DM. Its primary action is to inhibit hepatic glucose production, but it also increases the sensitivity of peripheral tissue to insulin.9 The observed pancreatic cancer risk reduction associated with metformin use is consistent with prior epidemiologic and preclinical data. In a Syrian hamster model of chemically induced pancreatic cancer, islet cell proliferation, in response to insulin resistance, seems to be a critical premalignant process.10 Metformin treatment, which rapidly improved insulin resistance and reduced hyperinsulinemia in these hamsters, significantly decreased islet cell hyperplasia and pancreatic ductal proliferation, and completely prevented the development of pancreatic adenocarcinoma.11

In addition to its indirect antiproliferative property through reducing insulin resistance and hyperinsulinemia, metformin may suppress protein synthesis and cell proliferation directly via activation of the AMP-activated protein kinase (AMPK) pathway (Figure 1B).12 The AMPK pathway serves fundamental functions in cellular metabolism, cellular energy sensing, and signaling, which may explain the effect of metformin on hepatic gluconeogenesis and obesity. AMPK also plays a critical role in normal cell growth regulation and tumorigenesis. AMPK regulates cell proliferation and apoptosis by phosphorylating key substrates, such as p53 and p27Kip1.13, 14 AMPK also inhibits cell growth and protein synthesis by inactivating the mammalian target of rapamycin.15 However, direct evidence that metformin inhibits pancreatic cancer cell growth via AMPK activation has not yet been reported.

Epidemiologic studies generally lend further support to the hypothesis that metformin may reduce cancer risk and/or improve cancer prognosis. Evans et al16 observed that diabetic patients treated with metformin had a lower incidence of cancer of any type when compared with patients on other treatments. Another observational study involving >10,000 diabetic patients treated with metformin or other antidiabetic agents showed a lower cancer-related mortality in the metformin group when compared with other DM medications, such as the sulfonylureas or insulin.17 The study by Li et al7 provides initial epidemiologic evidence that metformin may be protective against pancreatic cancer.

It is interesting to note that, although both metformin and the TZDs reduce insulin resistance, the TZDs did not seem to be associated with a decreased pancreatic cancer risk in the current study. This argues against the possibility that the protective effect of metformin against pancreatic cancer is mediated solely through improving insulin resistance. Better understanding of the mechanisms of action of these medications is necessary to explain the discrepant results regarding the TZDs and metformin. Clinically, given its superior safety profile and lower cost compared with the TZDs, metformin would be a much better choice as a potential chemopreventive agent, regardless.

Another noteworthy finding is that the magnitude of pancreatic cancer risk reduction associated with short-term (<2 years) metformin use is comparable to that associated with long-term metformin use (>5 years), even among those with long durations of DM. The absence of a duration-response effect raises the suspicion for residual confounding. Alternatively, it may suggest the presence of a premalignant state of pancreatic cancer that is readily and rapidly susceptible to the effect of metformin.

The significantly increased risk of pancreatic cancer observed among long-term insulin users by Li et al7 may represent an important finding. Increased risk of pancreatic cancer associated with long-term insulin use was observed also in 2 of the 3 previous studies that reported on the association between long-term (>5 years) insulin use and the risk of pancreatic cancer.3, 4, 8 Insulin has been shown to induce cell proliferation and reduces apoptosis by increasing the bioavailability of insulin-like growth factor-I.18 Data regarding the effect of hyperinsulinemia on pancreatic cancer are inconsistent. Although insulin promotes growth of human pancreatic cell lines in vitro,19 administration of exogenous insulin in the hamster pancreatic cancer model inhibits tumor induction.20 Nevertheless, the Alpha-Tocopherol Beta-Carotene (ATBC) Cancer Prevention Study reported a 2-fold increase in pancreatic cancer risk among those with insulin concentration in the highest quintile versus those in the lowest quintile.21

Because of its inefficiency in maintaining glucose homeostasis, exogenous insulin therapy leads to a substantially higher level of systemic insulin hyperinsulinemia than does endogenous hyperinsulinemia induced by insulin resistance.6 Insulin therapy has been shown in patients with DM to be associated with an increased risk of colorectal cancer.22 Thus, the role of long-term insulin use in pancreatic carcinogenesis deserves further investigation in a larger set of patients with more reliable insulin use data.

For any observational study reporting on the effect of a medication on disease outcomes other than its primary indication, confounding by indication is a concern. In the current context, the relevant question is whether the altered risk of pancreatic cancer among metformin or long-term insulin users was confounded by severity of the DM. Indeed, the prevalence of insulin use, a potential marker of DM severity, was significantly lower among metformin users than among non-users in this study, suggesting milder severity or shorter duration of DM in metformin users compared with metformin non-users. It is somewhat reassuring that the authors adjusted for insulin use and DM duration and that the risk reduction associated with metformin use among insulin non-users was similar to that in the primary analysis. Nevertheless, until there is a prospective study incorporating longitudinal data on degree of glycemic control and pancreatic β-cell function, the possibility of confounding by severity of DM cannot be excluded completely.

In conclusion, the report by Li et al7 describes an intriguing negative association between metformin use and the risk of pancreatic cancer. The study also raises the possibility that long-term exogenous insulin therapy may increase the risk of pancreatic cancer. This single-center, case-control study included a relatively small number of patients with DM, with the conclusions generated from a few users of these mediations, but the associations appeared consistent across various subgroups of the study cohort. At the very least, the observations reported raise testable hypotheses about the effects of metformin and insulin on the risk of pancreatic cancer. The observations also underscore the need for more information about how these widely used medications influence carcinogenesis in general and pancreatic carcinogenesis in particular. The current consensus treatment algorithm in type 2 DM proposed by the American Diabetes Association and the European Association for the Study of Diabetes advocates the inclusion of metformin in all treatment regimens in patient without contraindications.23 A possible chemopreventive effect of metformin on pancreatic cancer may provide an additional incentive for patients and physicians to follow this recommendation.

Back to Article Outline

References 

  1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008. CA: a Cancer Journal for Clinicians. 2008;58:71–96
  2. Pannala R, Leirness JB, Bamlet WR, et al. Prevalence and clinical profile of pancreatic cancer-associated diabetes mellitus. Gastroenterology. 2008;134:981–987
  3. Wang F, Gupta S, Holly EA. Diabetes mellitus and pancreatic cancer in a population-based case-control study in the San Francisco Bay Area, California. Cancer Epidemiology, Biomarkers & Prevention. 2006;15:1458–1463
  4. Bonelli L, Aste H, Bovo P, et al. Exocrine pancreatic cancer, cigarette smoking, and diabetes mellitus: a case-control study in northern Italy. [see comment] Pancreas. 2003;27:143–149
  5. Huxley R, Ansary-Moghaddam A, de Gonzalez ABerrington, et al. Type-II diabetes and pancreatic cancer: a meta-analysis of 36 studies. Br J Cancer. 2005;92:2076–2083
  6. Genuth S. Insulin use in NIDDM. Diabetes Care. 1990;13:1240–1264
  7. Li D, Yeung S–CJ, Hassan MM, et al. Antidiabetic therapies affect risk of pancreatic cancer. Gastroenterology. 2009;137:482–488
  8. Silverman DT, Schiffman M, Everhart J, et al. Diabetes mellitus, other medical conditions and familial history of cancer as risk factors for pancreatic cancer. Br J Cancer. 1999;80:1830–1837
  9. Bailey CJ, Turner RC. Metformin. [see comment] N Engl J Med. 1996;334:574–579
  10. Pour PM, Kazakoff K. Stimulation of islet cell proliferation enhances pancreatic ductal carcinogenesis in the hamster model. Am J Pathol. 1996;149:1017–1025
  11. Schneider MB, Matsuzaki H, Haorah J, et al. Prevention of pancreatic cancer induction in hamsters by metformin. [see comment] Gastroenterology. 2001;120:1263–1270
  12. Zhou G, Myers R, Li Y, et al. Role of AMP-activated protein kinase in mechanism of metformin action. [see comment] J Clin Invest. 2001;108:1167–1174
  13. Jones RG, Plas DR, Kubek S, et al. AMP-activated protein kinase induces a p53-dependent metabolic checkpoint. [see comment] Molecular Cell. 2005;18:283–293
  14. Liang J, Shao SH, Xu ZX, et al. The energy sensing LKB1-AMPK pathway regulates p27(kip1) phosphorylation mediating the decision to enter autophagy or apoptosis. Nat Cell Biol. 2007;9:218–224
  15. Zakikhani M, Dowling R, Fantus IG, et al. Metformin is an AMP kinase-dependent growth inhibitor for breast cancer cells. Cancer Res. 2006;66:10269–10273
  16. Evans JMM, Donnelly LA, Emslie-Smith AM, et al. Metformin and reduced risk of cancer in diabetic patients. BMJ. 2005;330:1304–1305
  17. Bowker SL, Majumdar SR, Veugelers P, et al. Increased cancer-related mortality for patients with type 2 diabetes who use sulfonylureas or insulin. [see comment] Diabetes Care. 2006;29:254–258
  18. Sandhu MS, Dunger DB, Giovannucci EL. Insulin, insulin-like growth factor-I (IGF-I), IGF binding proteins, their biologic interactions, and colorectal cancer. J Natl Cancer Inst. 2002;94:972–980
  19. Fisher WE, Boros LG, Schirmer WJ. Insulin promotes pancreatic cancer: evidence for endocrine influence on exocrine pancreatic tumors. J Surg Res. 1996;63:310–313
  20. Pour PM, Kazakoff K, Carlson K. Inhibition of streptozotocin-induced islet cell tumors and N-nitrosobis(2-oxopropyl)amine-induced pancreatic exocrine tumors in Syrian hamsters by exogenous insulin. Cancer Res. 1990;50:1634–1639
  21. Stolzenberg-Solomon RZ, Graubard BI, Chari S, et al. Insulin, glucose, insulin resistance, and pancreatic cancer in male smokers. JAMA. 2005;294:2872–2878
  22. Yang YX, Hennessy S, Lewis JD. Insulin therapy and colorectal cancer risk among type 2 diabetes mellitus patients. Gastroenterology. 2004;127:1044–1050
  23. Nathan DM, Buse JB, Davidson MB, et al. Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. [see comment] Diabetes Care. 2009;32:193–203

 Conflicts of interest The author discloses no conflicts.

PII: S0016-5085(09)00991-3

doi:10.1053/j.gastro.2009.06.022

Refers to article:

  • Editorial Accompanies this Article Antidiabetic Therapies Affect Risk of Pancreatic Cancer , 17 April 2009

    Donghui Li, Sai–Ching J. Yeung, Manal M. Hassan, Marina Konopleva, James L. Abbruzzese
    Gastroenterology August 2009 (Vol. 137, Issue 2, Pages 482-488)

Gastroenterology
Volume 137, Issue 2 , Pages 412-415, August 2009