ARTICLE The influence of glucose-lowering therapies on cancer risk in type 2 diabetes C. J. Currie & C. D. Poole & E. A. M. Gale Received: 19 May 2009 /Accepted: 18 June 2009 /Published online: 2 July 2009 # Springer-Verlag 2009 Abstract Aims/hypothesis The risk of developing a range of solid tumours is increased in type 2 diabetes, and may be influenced by glucose-lowering therapies. We examined the risk of development of solid tumours in relation to treatment with oral agents, human insulin and insulin analogues. Methods This was a retrospective cohort study of people treated in UK general practices. Those included in the analysis developed diabetes 40 years of age, and started treatment with oral agents or insulin after 2000. A total of 62,809 patients were divided into four groups according to whether they received monotherapy with metformin or sulfonylurea, combined therapy (metformin plus sulfonyl- urea), or insulin. Insulin users were grouped according to treatment with insulin glargine, long-acting human insulin, biphasic analogue and human biphasic insulin. The out- come measures were progression to any solid tumour, or cancer of the breast, colon, pancreas or prostate. Confound- ing factors were accounted for using Cox proportional hazards models. Results Metformin monotherapy carried the lowest risk of cancer. In comparison, the adjusted HR was 1.08 (95% CI 0.96���1.21) for metformin plus sulfonylurea, 1.36 (95% CI 1.19���1.54) for sulfonylurea monotherapy, and 1.42 (95% CI 1.27���1.60) for insulin-based regimens. Adding metfor- min to insulin reduced progression to cancer (HR 0.54, 95% CI 0.43���0.66). The risk for those on basal human insulin alone vs insulin glargine alone was 1.24 (95% CI 0.90���1.70). Compared with metformin, insulin therapy increased the risk of colorectal (HR 1.69, 95% CI 1.23���2.33) or pancreatic cancer (HR 4.63, 95% CI 2.64���8.10), but did not influence the risk of breast or prostate cancer. Sulfonylureas were associated with a similar pattern of risk as insulin. Conclusions/interpretation Those on insulin or insulin secretagogues were more likely to develop solid cancers than those on metformin, and combination with metformin abolished most of this excess risk. Metformin use was associated with lower risk of cancer of the colon or pancreas, but did not affect the risk of breast or prostate cancer. Use of insulin analogues was not associated with increased cancer risk as compared with human insulin. Keywords Cancer . Insulin . Insulin analogues . Metformin . Sulfonylureas . Survival . Type 2 diabetes Abbreviations LVD Large vessel disease OHA Oral hypoglycaemic agent THIN The Health Information Network Introduction Type 2 diabetes is associated with an increased risk of mortality from a range of solid tumours, including cancers of the colon, breast and pancreas [1]. Similar associations C. J. Currie (*) School of Medicine, Cardiff University, The Pharma Research Centre, Cardiff MediCentre, Cardiff CF14 4UJ, UK e-mail: currie@cardiff.ac.uk C. D. Poole Department of Epidemiology, Pharmatelligence, Cardiff, UK E. A. M. Gale Diabetes and Metabolism, School of Medicine, Bristol University, Bristol, UK Diabetologia (2009) 52:1766���1777 DOI 10.1007/s00125-009-1440-6

have been noted with central obesity and other conditions associated with increased levels of circulating insulin. These observations have given rise to the hypothesis that growth of these tumours, which are characterised by abnormal expression and function of the insulin���IGF-1 series of receptors [2, 3], is promoted by the trophic action of insulin interacting with these receptors. The cancer risk associated with diabetes may also be influenced by therapy: for example, the risk of colon cancer is higher in individuals on insulin [4], patients on metformin are less likely to be diagnosed with cancer [5], and the risk of mortality from solid tumours is lower for metformin than for exogenous insulin or sulfonylureas [6]. As recognition dawns that cancer should be numbered among the compli- cations of diabetes, the possibility that therapies for diabetes may influence tumour progression is likely to attract increasing interest and concern. Furthermore, the observation that both endogenous insulin and exogenous insulin therapy are associated with tumour progression raises questions as to the safety of the insulin analogues, which have subtly modified receptor binding properties and accelerate the growth and proliferation of both healthy and tumour cell lines in culture [7, 8]. The present study examined the relative frequency with which cancer was diagnosed in patients receiving a range of therapies for type 2 diabetes, including human and analogue insulins. To establish the overall pattern of cancer risk using alternative glucose-lowering therapies, we compared four common treatment regimens for type 2 diabetes: metformin monotherapy, sulfonylurea monother- apy, combination therapy with metformin plus sulfonylur- eas, and all insulin-based therapies combined. We tested the hypotheses that therapies other than metformin would be associated with increased development of solid tumours in general, and of diabetes-associated cancers of the breast, colon and pancreas in particular. We also consid- ered carcinoma of the prostate, which has a weak negative association with diabetes [9], and looked to see if combined treatment with metformin influenced these risks. Finally, we subdivided the insulin-based therapies to test the hypothesis that insulin glargine (A21Gly, B31Arg,B32Arg human insulin) or analogue mixes were associated with a greater risk of cancer than insulin of human origin. Methods Data source This was a retrospective cohort study of people treated in UK general practices participating in The Health Information Network (THIN). Created in 2002, THIN includes data from approximately 300 UK practices, and is similar in structure and scope to the General Practice Research Database [10���14]. Patients included in THIN are similar in age, sex and geographic characteristics to the general UK population [10���12]. THIN includes records on 4.78 million patients, of which 2.26 million are currently active. Approximately 3% of patients are lost annually because of leaving a practice or death. The data are collected in a non-interventional way from the daily record keeping of physicians. The records are anonymised at the collection stage so that researchers have access to only encrypted identifiers for the physician���s office and the patient. The database contains information on all past and current medical diagnoses, both acute and chronic (coded using Read codes), and prescribed medications (coded using British National Formulary codes). Unlike many other databases, THIN also includes laboratory values, which are electronically captured, and some aspects of physical examinations. Validation studies have been pub- lished and data collected in this way have previously been used to study diabetes [10, 11, 15]. Patients were selected if they achieved cohort member- ship after 2000, shortly before the introduction of insulin glargine in the UK in mid-2000. Study population Patients selected for analysis had a diagnosis of diabetes presenting later than 40 years of age, had received six or more sequential prescriptions for oral hypoglycaemic agents (OHAs), and other potential causes of secondary diabetes had not been recorded. Four primary cohorts were defined. Cohort 1 and Cohort 2 were defined by newly initiated OHA monotherapy (preceded by a wash-in period of 6 months, during which no OHAs were prescribed) with either metformin or a sulfonylurea, respectively. Cohort 3 was defined by a newly identified switch from OHA monotherapy with either metformin or sulfonylurea to an oral regimen involving both drugs in combination but no other concomitant class of OHA. Cohort 4 included those previously treated with OHAs who had been newly initiated on any insulin-based regimen. Cohort 4 was then subdivided when appropriate to assess the association between treatment subgroups and risk of cancer. The insulin subclasses were as follows: 4a, insulin glargine with no other concomitant insulin 4b, long-acting human insulin with no other concomitant insulin 4c, biphasic insulin of human origin and 4d, analogue biphasic insulin. In addition we wanted to provide some indication of the cancer risk of the various alternative diabetes therapies with regard to untreated diabetes. To this end, we included a preliminary analysis of the cancer risk in the aforementioned cohorts vs those with a diagnosis of type 2 diabetes who had no record of a diabetes-related medication, or those in other cohorts before prescription of their first diabetes-related medication. This comparison group thus represented a group of people with diet-treated Diabetologia (2009) 52:1766���1777 1767