SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The objective of this nationwide case-control study was to evaluate the risk of specific malignancy in diabetic patients who received thiazolidinediones (TZDs). A total of 606,583 type 2 diabetic patients, age 30 years and above, without a history of cancer were identified from the Taiwan National Health Insurance claims database during the period between January 1 2000 and December 31 2000. As of December 31 2007, patients with incident cancer of liver, colorectal, lung, and urinary bladder were included as cases and up to four age- and sex-matched controls were selected by risk-set sampling. Logistic regression models were applied to estimate the odds ratio (OR) and 95% confidence interval (CI) between TZDs and cancer incidence. A total of 10,741 liver cancer cases, 7,200 colorectal cancer cases, and 70,559 diabetic controls were included. A significantly lower risk of liver cancer incidence was found for any use of rosiglitazone (OR: 0.73, 95% CI: 0.65-0.81) or pioglitazone (OR: 0.83, 95% CI: 0.72-0.95), respectively. The protective effects were stronger for higher cumulative dosage and longer duration. For colorectal cancer, rosiglitazone, but not pioglitazone, was associated with a significantly reduced risk (OR: 0.86; 95% CI: 0.76-0.96). TZDs were not associated with lung and bladder cancer incidence, although a potential increased risk for bladder cancer with pioglitazone use ≥3 years could not be excluded (OR: 1.56; 95% CI: 0.51-4.74). Conclusion: The use of pioglitazone and rosiglitazone is associated with a decreased liver cancer incidence in diabetic patients. The effects on occurrence of specific cancer types may be different for pioglitazone and rosiglitazone. (HEPATOLOGY 2012;)

Both diabetes and cancer are common diseases that have tremendous impacts on health worldwide and the prevalence of both diseases is increasing globally. The diagnosis of cancer and diabetes in the same individual occurs more frequently than would be expected by chance. 1, 2 Diabetes has been consistently associated with an increased risk of cancers of the liver, pancreas, and endometrium, despite an association with the occurrence of other cancers being inconclusive. 3 A recent meta-analysis reported that the hazard ratio among persons with diabetes compared with those without diabetes was 1.25 (95% confidence interval [CI]: 1.19 to 1.31) for death from cancer, moderately associated with death from cancers of the liver, pancreas, ovary, colorectal, lung, bladder, and breast. 4

Many factors may affect the positive association between diabetes and cancers. Potential risk factors common to both diseases include age, sex, obesity, physical activity, diet, alcohol, and smoking. 5-10 Furthermore, diabetes treatment might influence cancer risk and cancer prognosis. Evidence from observational studies indicates that oral hypoglycemic agents and insulin are associated with either an increased or reduced risk of cancer. 2

Thiazolidinediones (TZDs) are insulin-sensitizing peroxisome proliferator-activated receptor gamma (PPAR-γ) agonists, available drugs including pioglitazone and rosiglitazone in this class. Laboratory studies showed that PPAR-γ agonists might have anti-cancer activities, such as growth inhibition, induction of apoptosis, and cell differentiation. 11-13 In contrast, preclinical studies showed that bladder tumors were observed in male rats receiving doses of pioglitazone that produced blood drug levels equivalent to those resulting from a clinical dose. 14 Additionally, results from a 3-year placebo-controlled trial of pioglitazone demonstrated a higher incidence of bladder neoplasm in patients receiving pioglitazone. 15 Although a meta-analysis of randomized clinical trials found that rosiglitazone was not associated with a significant modification of cancer risk, epidemiologic data regarding individual sites of cancer risk associated with different TZDs were inconsistent. 16-20 Therefore, the objective of this study was to conduct a nested case-control study based on a nation-wide health insurance claims database in Taiwan to assess the association between TZDs (both pioglitazone and rosiglitazone) and the occurrences of liver, colorectal, lung, and urinary bladder cancers.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Data Source.

The Taiwan National Health Insurance (NHI) claims database includes complete outpatient visits, hospital admissions, prescriptions, disease, and vital status for 99% of the population of 23 million in Taiwan. We established the longitudinal medical history of each beneficiary by linking several computerized administrative and claims datasets to National Cancer and Death Registry through the date of birth and the civil identification number unique to each beneficiary. The protocol of this study was approved by the National Taiwan University Hospital Research Ethics Committee.

Source Population.

Data for all patients with any diabetes diagnostic codes (International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM), ICD-9-CM code 250 and A code 181) in the claims database between January 1 2000 and December 31 2000 were retrieved. An algorithm including age, number of outpatient visits, number of hospitalizations, and the hospital level was used to identify potential diabetic patients with improved accuracy. This definition of diabetes was evaluated by a study sampling 9,000 patients with a diagnosis of diabetes in the NHI claims data in 2000. The diagnostic accuracy of diabetes was assessed based on patient response to a questionnaire concerning (1) being told by doctors they have diabetes or (2) ever use of oral hypoglycemic agents or insulin injections. Subjects who gave negative or uncertain answers but were using hypoglycemic agents in the pharmacy claims database were also classified as diabetic. Validation of this algorithm by which 640,173 patients were identified demonstrated 93.2% sensitivity and 92.3% positive predictive value.

Study Design.

Because diabetic patients may receive highly variable antidiabetic therapies in terms of drug regimens, dosage, duration, and other concomitant drugs, and confounding factors are constantly changing over time in a long-term observational follow-up study, there are complex analytical difficulties for a cohort analysis to be attempted. Instead, a nested case-control approach is a useful alternative of cohort analysis to study time-dependent exposures. 21 The risk estimates from cohort and nested case-control analyses should be identical if confounding is fully controlled in both analyses. The superior computational efficiency by nested case-control study design may be particularly useful in case rare outcomes are encountered. 22

Study Groups.

First, we excluded patients age <30 and >100 years old. To enroll patients with type 2 diabetes, we further excluded those who (1) had a hospital admission with a discharge diagnosis of insulin dependent diabetes mellitus (ICD-9-CM code 250.x1, 250.x3), or (2) received a catastrophic illness certificate issued by the Department of Health for type 1 diabetes (Fig. 1). Patients were classified as having prevalent or newly diagnosed type 2 diabetes according to the criteria in 1999. Those who had a history of cancer recorded in the National Cancer Registry any time before the cohort entry date, that is, date of diabetes diagnosis for newly diagnosed patients and January 1 2000 for prevalent patients, were also excluded. Patients were followed from January 1 2000 (for prevalent type 2 diabetes patients) or the date of diabetes diagnosis in 2000 (for newly diagnosed type 2 diabetes patients) to the earliest of cancer diagnosis, death, disenrollment from the national health insurance, or December 31 2007.

thumbnail image

Figure 1. Study flow.

Download figure to PowerPoint

All individuals in the study cohort with the first occurrence of liver, colorectal, lung, and urinary bladder cancer were included as cases. All potential cases were validated by a linkage through National Cancer Registry.

A risk-set sampling (that is, controls sampled from those in the original study cohort who remained free of outcome at the time point when a case occurred) matched by age (within 5 years), sex, and the number of days of follow-up was used to find controls for the cohort. For newly diagnosed type 2 diabetes patients, cases and controls were also matched on antidiabetic treatment duration (within 30 days) at cancer diagnosis. For newly diagnosed diabetic patients, this scheme that matched follow-up duration would have, by design, also taken diabetes duration into consideration. For prevalent patients with unknown duration, we selected controls with the same follow-up duration to reduce the confounding effect by diabetes duration. Up to four controls were selected for each case.

Exposure Ascertainment and Covariates Adjustment.

The main exposure of interest was the use of rosiglitazone and pioglitazone, which entered Taiwan's market in March 2000 and June 2001, respectively. We collected information of prescribed drug types (according to the anatomic therapeutic chemical [ATC] classification system, A10BG02 for rosiglitazone and A10BG03 for pioglitazone), dosage, date of prescription, supply days, and total number of pills dispensed from the outpatient pharmacy prescription database. The mean daily dose for each individual was calculated as dividing the cumulative number of pills by the follow-up duration. Subsequently, the defined daily dose (DDD) was then established by an expert panel according to the relative amount compared to the typical maintenance dose for an adult. Other medications, including biguanides (ATC code A10BA), sulfonylurea (A10BB), alpha glucosidase inhibitors (A10BF), glinides (A10BX02, A10BX03), fast-acting insulins (A10AB01, A10AB02, A10AB03, A10AB30), insulin glargine (A10AE04), statins (C10AA), and low-dose aspirin (B01AC06), angiotensin converting enzyme inhibitors (C09AA), angiotensin receptor blockers (C09CA), beta-blockers (C07A), and calcium channel blockers (C08C, C08D, C08E), were also recorded and reported.

We also incorporated inpatient and outpatient diagnosis files to ascertain the history of cardiovascular disease, peripheral vascular disease, cerebrovascular disease, retinopathy, nephropathy, neuropathy, depression, chronic kidney disease, chronic liver disease, and chronic lung disease based on ICD-9-CM codes. Patients were classified as having chronic liver disease if they had at least one hospital admission or outpatient visit with a diagnostic code of hepatitis B virus infection (ICD-9-CM codes 070.2x, 070.3x, V02.61), hepatitis C virus infection (070.41, 070.44, 070.51, 070.54, V02.62), chronic hepatitis (571.4), liver cirrhosis (571.2, 571.5, 571.6), or alcoholic liver disease (571.0x, 571.1x, 571.2, 571.3x). A previous validation study using hospital administrative database reported a positive predictive value of 90% with this definition. 23 Covariate information included age, gender, and socioeconomic status (i.e., using monthly income as a proxy).

Statistical Analysis.

Conditional logistic regression was used to estimate the crude and adjusted odds ratio (OR) and 95% confidence interval (CI) for the association between rosiglitazone/pioglitazone and cancer occurrence with “nonuse” as the reference group. Potential covariates, including socioeconomic status (monthly income level), diabetes complications and comorbidities at cancer diagnosis, other antidiabetic agents, antihypertensive medications, statin, and aspirin were examined. In the multivariate analysis, we adjusted for the use of short-acting human insulin, sulfonylurea, metformin, as these antidiabetic agents were reported to be associated with cancer risks and could potentially confound the association. Other variables were chosen by using stepwise selection with P values < 0.10 for model entry and > 0.05 for removal. The association between rosiglitazone/pioglitazone and individual cancer incidence was separately estimated after adjustment for potential confounders specific to that cancer type. In the dose- and duration-response analyses, we calculated the ORs for higher (≥120 DDD) and lower cumulative dose (<120 DDD) use, and for cumulative treatment duration ≥3, 2-3, 1-2, and ≤1 years. A two-sided P value < 0.05 was considered statistically significant.

Approximately 15% participants claimed at lease one prescription for pioglitazone. Assuming a correlation coefficient for pioglitazone use between case and control was 0.5 and an ORs was 0.8, a study of 2,500 cases and 4 controls for each case would have a power ≥80% at α = 0.05.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

During the period between January 1 2000 and December 31 2000, a total of 640,173 patients were initially identified by the algorithm (Fig. 1). Among them, 606,583 patients, including 497,663 prevalent type 2 diabetes and 108,920 newly diagnosed type 2 diabetes, were included in the analysis after excluding those who were age <30 or >100 years, who were type 1 diabetes, or who already had prevalent cancer. These patients were followed for a median of 7.9 years. Meanwhile, a total of 174,800 (27.3%) patients died, whereas only 1,566 (0.2%) were lost to follow-up due to discontinuation from or drop-out of health insurance.

During the study period the number of oral antidiabetic agents (mean ± standard deviation) was 2.62 ± 1.07 and the mean daily dosage was 1.18 ± 0.92 DDD per day. Metformin and sulfonylurea were the most commonly used oral antidiabetic medications (83.5% and 88.4% of the study population, respectively). In the diabetic cohort, 324,773 (50.7%) had ever used insulin therapy during the study period.

Approximately 26.1% of the patients ever received rosiglitazone and 14.1% pioglitazone. The mean cumulative duration was 522 days and the mean daily dosage was 0.14 DDD/day for rosiglitazone, as compared with 375 days and 0.11 DDD/day for pioglitazone. Because of the concern that physicians might preferentially prescribe TZDs to patients with normal liver function, we compared the proportion of diabetic patients with chronic liver disease (hepatitis B virus infection, hepatitis C virus infection, chronic hepatitis, liver cirrhosis, and alcoholic liver disease) among control subjects (a representative sample of the study population) who received different types of antidiabetic therapies. A significantly higher proportion of patients with chronic liver disease were found to have received insulin, rosiglitazone, and/or pioglitazone than those receiving sulfonylureas, metformin, or diet therapy (Supporting Table A).

A total of 10,741 incident liver cancer, 7,200 colorectal cancer, 5,361 lung cancer, and 1,583 bladder cancer cases were identified. These cases were age- and sex-matched with 99,538 controls (at least one and up to four eligible controls for each case) by the risk-set sampling scheme. In general, cancer cases were more likely to be of lower socioeconomic status and more likely to have diabetes-associated complications (retinopathy, neuropathy, and nephropathy), cardiovascular disease, chronic kidney diseases, liver diseases, and lung diseases. The cases were also more likely to have received fast-acting insulin and insulin glargine and glinides, whereas fewer of them have received statins before cancer diagnosis as compared with controls (Table 1 for liver cancer and Table 2 for colorectal cancer). Despite a similar proportion of overall cancer cases and controls who received metformin and sulfonylurea, the mean daily dosage of these two antidiabetic agents in overall cancer cases were significantly higher than those for matched controls (data not shown). (Associations with specific cancer sites are reported below.) Regarding antihypertensive therapy, a higher proportion of cancer cases had ever used beta-blockers and calcium channel blockers.

Table 1. Characteristics, Comorbidities, and Medication Use Among Incident Liver Cancer Cases and Matched Controls
 Cases (N = 10,741)Controls (N = 41,847)Crude Odds Ratio (95% CI)
Age at cancer diagnosis (mean ± SD)66.66 ± 9.8866.71 ± 9.98
Male (%)64.0363.86
Diagnosis year   
 20001,3145,144
 20011,4585,613
 20021,3865,397
 20031,3535,277
 20041,3645,379
 20051,3415,225
 20061,3495,162
 20071,1764,650
Socioeconomic status (monthly income in NTD, %)   
 ≤17,28036.9137.15Reference
 17,281∼22,80045.9541.991.10 (1.05-1.15)
 22,801∼28,8005.055.490.92 (0.83-1.02)
 28,801∼36,3005.015.610.89 (0.81-0.99)
 36,301∼45,8004.115.130.80 (0.72-0.89)
 >45,8002.974.630.64 (0.56-0.72)
Comorbidities (%)   
Cardiovascular disease79.6780.900.91 (0.86-0.97)
Peripheral vascular disease6.285.461.17 (1.07-1.28)
Cerebrovascular disease25.9527.250.93 (0.89-0.98)
Retinopathy27.9926.181.10 (1.05-1.16)
Neuropathy43.6340.341.16 (1.11-1.21)
Nephropathy61.1351.051.60 (1.53-1.68)
Depression8.006.311.30 (1.20-1.41)
Chronic kidney disease11.678.251.48 (1.38-1.59)
Chronic liver disease82.5930.0412.80 (12.04-13.61)
Chronic lung disease15.8414.631.10 (1.04-1.17)
Charlson's index (mean ± SD)5.25 ± 2.893.43 ± 2.261.39 (1.38-1.40)
Medication use before cancer diagnosis (%)   
Metformin79.0079.980.93 (0.88-0.99)
 Mean daily dosage among users0.26 ± 0.290.18 ± 0.215.59 (5.09-6.14)
Sulfonylurea91.6190.581.16 (1.07-1.25)
 Mean daily dosage among users0.68 ± 0.580.41 ± 0.433.37 (3.21-3.53)
Alpha-glucosidase inhibitors17.6415.641.18 (1.11-1.25)
Thiazolidinediones17.0417.790.94 (0.88-1.00)
 Pioglitazone4.415.030.87 (0.78-0.96)
 Rosiglitazone14.7815.560.93 (0.87-0.99)
Glinides15.3611.891.38 (1.29-1.47)
Mean number of oral anti-diabetic agents (mean ± SD)2.22 ± 1.102.17 ± 1.081.05 (1.03-1.08)
Short-acting human insulin47.3126.572.64 (2.53-2.77)
Insulin glargine0.800.591.38 (1.07-1.77)
Statins16.5329.020.45 (0.43-0.48)
Aspirin42.0746.740.81 (0.77-0.85)
ACE inhibitors52.1452.260.99 (0.95-1.04)
Angiotensin receptor blockers28.4128.251.01 (0.96-1.06)
Beta-blockers56.2148.901.38 (1.32-1.44)
Calcium channel blockers64.1060.681.18 (1.13-1.24)
Table 2. Characteristics, Comorbidities, and Medication Use Among Incident Colorectal Cancer Cases and Matched Controls
 Cases (N = 7,200)Controls (N = 28,712)Crude Odds Ratio (95% CI)
Age at cancer diagnosis (mean ± SD)70.17 ± 9.4170.24 ± 9.41
Male (%)54.2154.16
Diagnosis year   
 20008643,424
 20018183,283
 20028983,574
 20039223,657
 20049713,888
 20058903,542
 20069003,626
 20079373,718
Socioeconomic status (monthly income in NTD, %)   
 ≤17,28042.4740.85Reference
 17,281∼22,80039.8641.760.91 (0.86-0.97)
 22,801∼28,8004.464.620.92 (0.81-1.05)
 28,801∼36,3004.945.160.92 (0.81-1.04)
 36,301∼45,8004.394.091.03 (0.90-1.18)
 >45,8003.883.521.05 (0.92-1.21)
Comorbidities (%)   
Cardiovascular disease86.6984.591.21 (1.12-1.31)
Peripheral vascular disease5.586.030.92 (0.82-1.03)
Cerebrovascular disease31.7130.371.07 (1.01-1.13)
Retinopathy27.8127.251.03 (0.97-1.10)
Neuropathy42.4042.281.01 (0.95-1.06)
Nephropathy61.8554.291.43 (1.35-1.51)
Depression6.217.150.86 (0.77-0.96)
Chronic kidney disease11.408.651.37 (1.26-1.49)
Chronic liver disease32.3828.931.19 (1.12-1.26)
Chronic lung disease16.3116.790.96 (0.90-1.03)
Charlson's index (mean ± SD)4.86 ± 3.193.60 ± 2.311.24 (1.22-1.25)
Medication use before cancer diagnosis (%)   
Metformin80.0779.601.03 (0.97-1.11)
 Mean daily dosage among users0.29 ± 0.300.18 ± 0.217.97 (7.12-8.93)
Sulfonylurea91.5790.261.20 (1.09-1.33)
Mean daily dosage among users0.69 ± 0.590.42 ± 0.443.42 (3.23-3.62)
Alpha-glucosidase inhibitors16.1715.861.03 (0.95-1.11)
Thiazolidinediones17.3517.780.97 (0.90-1.04)
 Pioglitazone5.255.091.03 (0.92-1.17)
 Rosiglitazone14.9015.480.95 (0.88-1.03)
Glinides15.3112.721.26 (1.17-1.36)
Mean number of oral anti-diabetic agents (mean ± SD)2.22 ± 1.092.18 ± 1.091.05 (1.02-1.08)
Short-acting human insulin48.3127.702.62 (2.48-2.77)
Insulin glargine0.880.621.42 (1.06-1.90)
Statins27.6329.520.90 (0.85-0.96)
Aspirin51.0650.181.04 (0.99-1.10)
ACE inhibitors56.5855.401.05 (1.00-1.11)
Angiotensin receptor blockers31.9430.571.07 (1.01-1.14)
Beta-blockers57.0051.941.25 (1.18-1.32)
Calcium channel blockers71.5165.731.35 (1.27-1.43)

In the univariate analyses, we found a negative association between pioglitazone/rosiglitazone and liver cancer incidence, and a positive one between rosiglitazone and lung cancer. After controlling for potential confounding variables including short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea (mean daily dosage in quartiles), number of oral antidiabetic agents, chronic liver disease, statins, aspirin, beta-blockers, chronic kidney disease, glinides, nephropathy, cerebrovascular disease, calcium channel blockers, cardiovascular disease, and chronic lung disease, a significantly decreased risk of liver cancer incidence was found for any use of rosiglitazone (OR: 0.73, 95% CI: 0.65-0.81) and pioglitazone (OR: 0.83, 95% CI: 0.72-0.95), respectively (Table 3), in contrast to the adjusted ORs of 2.35 (95% CI: 2.21-2.49) for short-acting insulin, 1.05 (95% CI: 0.93-1.18) for sulfonylurea, and 0.77 (95% CI: 0.69-0.85) for metformin. The protective effects were even stronger for higher cumulative dosage ≥120 DDD (OR 0.64; 95% CI: 0.56-0.72 for rosiglitazone and OR 0.80; 95% CI: 0.67-0.95 for pioglitazone) and for cumulative treatment duration ≥3 years (OR 0.64; 95% CI: 0.49-0.85 for rosiglitazone and OR 0.44; 95% CI: 0.23-0.86 for pioglitazone). Risk estimates were similar between prevalent and newly diagnosed type 2 diabetes patients. Due to the high prevalence of hepatitis B and C infection, the analysis was further stratified to those with and without chronic liver disease. The risk reduction (especially in high and prolonged dosage of rosiglitazone and pioglitazone) in liver cancer was mostly seen in the patients with prevalent chronic liver disease (Table 4).

Table 3. Risk of Liver Cancer Associated With Pioglitazone or Rosiglitazone Use in Type 2 Diabetic Patients
 PioglitazoneRosiglitazone
 Number of CasesNumber of ControlsCrude Odds RatioAdjusted* Odds RatioNumber of CasesNumber of ControlsCrude Odds RatioAdjusted* Odds Ratio
  • *

    Multivariate model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea (mean daily dosage in quartiles), number of oral antidiabetic agents, chronic liver disease, statins, aspirin, beta-blockers, chronic kidney disease, glinides, nephropathy, cerebrovascular disease, calcium channel blockers, cardiovascular disease, chronic lung disease.

Nonuse10,26739,741ReferenceReference9,15435,336ReferenceReference
Any use4742,1060.87 (0.78-0.96)0.83 (0.72-0.95)1,5876,5110.93 (0.87-0.99)0.73 (0.65-0.81)
Cumulative dosage        
 < 120 DDD2259150.95 (0.81-1.10)0.87 (0.72-1.05)7252,4831.11 (1.02-1.22)0.86 (0.75-0.98)
 ≥ 120 DDD2491,1910.81 (0.70-0.93)0.80 (0.67-0.95)8624,0280.81 (0.75-0.88)0.64 (0.56-0.72)
Cumulative duration        
 ≤1 year3521,4630.93 (0.82-1.04)0.87 (0.74-1.02)1,0343,8561.02 (0.95-1.10)0.78 (0.69-0.88)
 1-2 years793870.78 (0.61-1.00)0.80 (0.59-1.07)3301,4570.86 (0.76-0.98)0.66 (0.56-0.79)
 2-3 years301590.73 (0.49-1.08)0.71 (0.45-1.14)1356970.73 (0.60-0.88)0.59 (0.46-0.74)
 ≥3 years13970.52 (0.29-0.93)0.44 (0.23-0.86)885010.65 (0.52-0.82)0.64 (0.49-0.85)
Table 4. Risk of Liver Cancer Associated With Pioglitazone or Rosiglitazone Use in Type 2 Diabetic Patients With and Without Chronic Liver Disease
 Diabetic Patients Without Chronic Liver DiseaseDiabetic Patients With Chronic Liver Disease
 PioglitazoneRosiglitazonePioglitazoneRosiglitazone
 Crude Odds RatioAdjusted* Odds RatioCrude Odds RatioAdjusted* Odds RatioCrude Odds RatioAdjusted* Odds RatioCrude Odds RatioAdjusted* Odds Ratio
  • *

    Multivariate model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea (mean daily dosage in quartiles), number of oral antidiabetic agents, statins, aspirin, beta-blockers, chronic kidney disease, glinides, nephropathy, cerebrovascular disease, calcium channel blockers, cardiovascular disease, chronic lung disease.

NonuseReferenceReferenceReferenceReferenceReferenceReferenceReferenceReference
Any use1.11 (0.85-1.44)0.99 (0.72-1.36)1.00 (0.85-1.18)0.85 (0.66-1.09)0.81 (0.71-0.93)0.76 (0.64-0.90)0.88 (0.81-0.96)0.73 (0.64-0.83)
Cumulative dosage        
 < 120 DDD0.94 (0.63-1.40)0.85 (0.54-1.34)1.07 (0.84-1.37)0.92 (0.67-1.27)0.89 (0.73-1.08)0.88 (0.70-1.11)1.10 (0.97-1.24)0.93 (0.79-1.10)
 ≥ 120 DDD1.25 (0.89-1.74)1.10 (0.74-1.63)0.96 (0.78-1.17)0.80 (0.61-1.07)0.75 (0.62-0.91)0.68 (0.54-0.84)0.74 (0.67-0.83)0.60 (0.51-0.70)
Cumulative duration        
 ≤1 year1.08 (0.79-1.47)0.98 (0.68-1.42)1.02 (0.83-1.24)0.87 (0.65-1.15)0.87 (0.74-1.02)0.82 (0.68-0.99)0.96 (0.87-1.07)0.80 (0.69-0.93)
 1-2 years1.05 (0.60-1.86)0.85 (0.45-1.61)0.95 (0.68-1.31)0.83 (0.56-1.22)0.82 (0.59-1.14)0.76 (0.53-1.08)0.83 (0.70-0.98)0.64 (0.52-0.79)
 2-3 years1.26 (0.54-2.94)1.16 (0.44-3.06)0.93 (0.60-1.44)0.81 (0.49-1.35)0.59 (0.35-1.00)0.51 (0.28-0.91)0.66 (0.51-0.85)0.51 (0.38-0.69)
 ≥3 years1.93 (0.53-6.99)1.70 (0.40-7.26)1.13 (0.66-1.95)0.83 (0.45-1.53)0.38 (0.17-0.82)0.30 (0.13-0.68)0.64 (0.47-0.87)0.63 (0.45-0.90)

For colorectal cancer, rosiglitazone was associated with a significantly decreased risk (OR: 0.86; 95% CI: 0.76-0.96) with a more prominent effect among those with the highest cumulative dose (OR: 0.83; 95% CI: 0.73-0.95) (Table 5). In contrast, pioglitazone was not associated with a significantly protective effect for colorectal cancer, although a trend of negative association was also found. The characteristics for cases and controls of lung and bladder cancer are summarized in Supporting Tables B and C. No relation between rosiglitazone/pioglitazone and lung or bladder cancer was found, although the ORs were above 1.0 (Supporting Tables D, E). A dosage and duration response relationship was not evident between the use of rosiglitazone and pioglitazone and these two types of cancer. An increase in bladder cancer risk was observed with pioglitazone use ≥3 years (OR: 1.56, 95% CI: 0.51-4.74), which did not reach statistical significance.

Table 5. Risk of Colorectal Cancer Associated With Pioglitazone or Rosiglitazone Use in Type 2 Diabetic Patients
 PioglitazoneRosiglitazone
 Number of CasesNumber of ControlsCrude Odds RatioAdjusted* Odds RatioNumber of CasesNumber of ControlsCrude Odds RatioAdjusted* Odds Ratio
  • *

    Multivariate model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea (mean daily dosage in quartiles), number of oral antidiabetic agents, glinides, nephropathy, neuropathy, chronic liver disease, statins, retinopathy, calcium channel blockers, ACE inhibitors, peripheral vascular disease, depression, beta-blockers, aspirin, chronic kidney disease, chronic lung disease, cerebrovascular disease.

Nonuse6,82227,250ReferenceReference6,12726,267ReferenceReference
Any use3781,4621.03 (0.92-1.17)1.04 (0.91-1.20)1,0734,4450.95 (0.88-1.03)0.86 (0.76-0.96)
Cumulative dosage        
 < 120 DDD1705921.15 (0.96-1.37)1.15 (0.95-1.39)4341,7850.96 (0.86-1.07)0.89 (0.77-1.03)
 ≥ 120 DDD2808700.96 (0.82-1.12)0.97 (0.82-1.16)6392,6600.94 (0.86-1.04)0.83 (0.73-0.95)
Cumulative duration        
 ≤1 year2789811.13 (0.99-1.30)1.15 (0.98-1.34)6782,6710.99 (0.91-1.09)0.91 (0.80-1.04)
 1-2 years602670.90 (0.68-1.19)0.82 (0.61-1.11)2209500.91 (0.78-1.05)0.78 (0.65-0.94)
 2-3 years261330.79 (0.51-1.20)0.86 (0.55-1.33)994760.81 (0.65-1.02)0.69 (0.54-0.88)
 ≥3 years14810.69 (0.39-1.22)0.77 (0.43-1.39)763280.91 (0.70-1.17)0.83 (0.63-1.10)

The cancer risk associated with antidiabetic therapies other than TZD are also reported (Supporting Table F). In general, short-acting human insulin, sulfonylureas, and glinides were significantly associated with an increased risk for liver, colorectal, and lung cancer. In contrast, metformin was associated with a decreased risk for liver cancer.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Consistent with previous in vitro studies on TZDs which showed antiproliferation and prodifferentiation effects, our data have provided an association between the clinical use of TZDs and a reduced risk for several cancer incidences, in particular liver cancer. The association became stronger when the duration of TZD use was longer and the dosage was higher. Rosiglitazone, but not pioglitazone, was associated with a significantly reduced risk for colorectal cancer. No association between both TZDs and lung and bladder cancer was observed.

Previous reports on the association between TZD use and cancer incidence have been inconsistent. The report from the data obtained from the Veterans Integrated Services Network 16 (VISN 16) cohort of 87,678 individuals showed a 33% reduction in lung cancer risk among TZD users compared with nonusers (relative risk: 0.65, 95% CI: 0.51-0.87). However, as rosiglitazone and pioglitazone were combined, the risk reduction for colorectal cancer did not reach statistical significance. 18 In contrast, the present study results did not show a decreased risk for lung cancer. Although numerous in vitro studies support the protective effect of TZDs in lung cancer, the specific tissue or type of cancer and its stage might contribute to the efficacy or failure of TZDs as antineoplastic agents. 19, 20, 24-29 Because the risk factors, genetic expressions, and pharmaceutical responses of lung cancer of the Taiwanese differ significantly from those in the Western countries, there might also be a differential response to TZDs. 30

On the contrary, our analysis showed a protective effect of rosiglitazone on colorectal cancers, which was not evident in the VISN 16 cohort. In animal studies, PPAR-γ agonists inhibited tumor growth and colon carcinogenesis through induction of apoptosis and suppression of the cell cycle. 31-34 The current study, to the best of our knowledge, provides the first evidence that rosiglitazone but not pioglitazone might reduce the risk of colorectal cancer.

It is initially surprising that both pioglitazone and rosiglitazone are associated with a reduced risk for liver cancer. Hepatocellular carcinoma, one of the most incident, prevalent, and lethal malignancies in Taiwan, is regarded as a late-stage sequel of chronic infection of hepatitis B and C. 35, 36 With only a few exceptions, the development of hepatocellular carcinoma almost exclusively follows the sequence of chronic hepatic inflammation, cirrhosis of the liver, repair and regeneration of hepatic cells, and then carcinogenesis. 37 This might explain the finding that risk reduction was more evident in the patients with chronic liver disease. Despite the concern that physicians may preferentially prescribe TZDs to patients with better liver function (i.e., confounding by contraindication), the present study showed that among patients receiving rosiglitazone and/or pioglitazone there was a significantly higher proportion of patients with chronic liver disease as compared with those receiving sulfonylureas, metformin, or diet therapy. This finding suggests that in Taiwan diabetic patients with abnormal liver function, on the contrary, are more likely to have ever received TZDs. One explanation of this contradiction was that other antidiabetic medications, such as sulfonylurea and metformin, might be associated with more frequent adverse effects among these patients. Although the possibility of residual confounding by contraindication among those with abnormal liver function tests cannot be excluded, the observed protective effects of rosiglitazone and pioglitazone were less likely due to physicians' reluctance to prescribe rosiglitazone and/or pioglitazone to patients with chronic liver disease. Currently, there are many clinical investigations concerning antiviral therapy and interferon-α in the treatment of chronic hepatitis, aiming to stop the progression to cirrhosis and hepatocellular carcinoma. 38-40 TZDs may be considered a new component in the combination therapy because the protective effect is most prominent in the patients with chronic liver disease.

The present study also demonstrated that use of insulin, sulfonylurea, and glinides also increased the risk of cancer. The finding that both insulin and oral insulin secretagogues confer an increased risk suggests that an increasing insulin level plays an important role in carcinogenesis. 41 Insulin sensitizers (metformin and TZDs) do not increase insulin concentrations and, theoretically, may not influence the risk of cancer occurrence. The finding that metformin was associated with a decreased risk for liver cancer was comparable to the results in previous reports. 42 Further studies are warranted to elucidate the potential role of metformin to reduce the cancer risk among diabetic patients.

The strength of the current study includes that, on a national scale, there are far more cancer cases compared to previous epidemiological studies. As rosiglitazone and pioglitazone entered Taiwan's market in 2000 and 2001, respectively, diabetic patients in this study were all new-users to these two drugs and hence allowed us to capture all cancer occurrences following TZD treatment initiation. 43 Furthermore, this case-control study was designed to be nested within a clearly defined diabetic cohort. Each diabetic patient was followed from cohort entry (date of diabetes diagnosis for newly diagnosed patients and January 1 2000 for prevalent diabetes) to the earliest of cancer diagnosis, death, or December 31 2007. The cumulative dosage of TZDs and other antidiabetic therapy during the follow-up period was calculated and drug exposure experiences were compared between cancer cases and controls selected by risk-set sampling matched on age, sex, and follow-up time. This analysis accounted for the individual and time-varying nature of different glucose-lowering treatments and eliminated the bias introduced by defining treatment groups based on follow-up information.

There are also limitations in this study. First, we did not have the information of lifestyle risk factors or family history of cancer; there might be residual confounding by duration or severity of diabetes, as well as by obesity, smoking, and physical inactivity. Due to lack of data about patients' level of glycemic control, we could not examine whether a better glucose-lowering effect by TZDs as compared with nonuse may explain the association with a reduced cancer risk. Second, as our average cumulative treatment duration of TZDs was relatively short, we were not able to examine the long-term effect of TZDs on cancer occurrence. Third, we observed differential associations between pioglitazone and rosiglitazone with specific sites of cancer. Despite numerous in vitro and animal studies support the protective effects of TZDs, we are not able to identify the exact underlying physiological pathways that result in a reduced cancer risk and that differentiate pioglitazone and rosiglitazone. Fourth, one of the most recent studies included 193,099 patients in the Kaiser Permanente Northern California diabetes registry who were ≥40 years of age demonstrated that short-term use of pioglitazone was not associated with an increased incidence of bladder cancer (hazard ratio [HR] 1.2, 95% CI 0.9-1.5), but use for more than 2 years was weakly associated with increased risk (HR: 1.4, 95% CI: 1.03-2.0). Our results did not show a significant association despite a tendency to an increased risk. Due to the limited case number in bladder cancer, we could not exclude the possibility that a prolonged use of pioglitazone might potentially increase the risk for bladder cancer. 44 Fifth, PPAR-γ is one member of the nuclear receptor superfamily that contains in excess of 80 described receptors. Once activated, PPAR-γ will preferentially bind with retinoid X receptor α and signal antiproliferative, antiangiogenic, and prodifferentiation pathways in several tissue types, thus making it a highly useful target for down-regulation of carcinogenesis. 13 Rosiglitazone has PPAR-γ activity but pioglitazone has both PPAR-α and PPAR-γ activities. The mediation of cancer initiation and progression through dependent and independent pathways may also differ between rosiglitazone and pioglitazone. 45 The differential selectivity in activating PPAR signaling pathways might explain the cancer risk of different sites, but the true mechanisms remain to be clarified. Finally, TZDs are contraindicated in patients with congestive heart failure. 46 Pioglitazone and rosiglitazone show different cardiovascular safety profiles. 15, 47-50 We are not sure whether the reduced cancer risk could compensate for the potentially increased cardiovascular risk. The overall risks and benefits of TZD should be evaluated.

In conclusion, the results of this study show that both pioglitazone and rosiglitazone reduce the risk of incident liver cancer in type 2 diabetic patients. There is a better protection against cancer occurrence associated with a longer use and higher doses of TZDs. The association with individual sites of specific cancer differs between pioglitazone and rosiglitazone and the underlying mechanisms merit further investigations.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The corresponding authors have full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Chang C.H., Lin J.W.; Acquisition of data: Lai M.S.; Analysis and interpretation of data: Chang C.H., Lin J.W.; Drafting of the article: Lin J.W., Chang C.H.; Critical revision of the article for important intellectual content: Chuang L.M., Chan K.A.; Statistical analysis: Wu L.C.; Obtained funding: Lai M.S.; Study supervision: Lai M.S.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information
  • 1
    Nicolucci A. Epidemiological aspects of neoplasms in diabetes. Acta Diabetol 2010; 47: 87-95.
  • 2
    Giovannucci E, Harlan DM, Archer MC, Bergenstal RM, Gapstur SM, Habel LA, et al. Diabetes and cancer: a consensus report. Diabetes Care 2010; 33: 1674-1685.
  • 3
    Vigneri P, Frasca F, Sciacca L, Pandini G, Vigneri R. Diabetes and cancer. Endocr Relat Cancer 2009; 16: 1103-1123.
  • 4
    Emerging Risk Factors Collaboration. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011; 364: 829-841.
  • 5
    Liese AD, D'Agostino RBJr, Hamman RF, et al. The burden of diabetes mellitus among US youth: prevalence estimates from the SEARCH for Diabetes in Youth Study. Pediatrics 2006; 118: 1510-1518.
  • 6
    Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, 2009. CA Cancer J Clin 2009; 59: 225-249.
  • 7
    Ma J, Li H, Giovannucci E, Mucci L, Qiu W, Nguyen PL, Gaziano JM, et al. Prediagnostic body-mass index, plasma C-peptide concentration, and prostate cancer-specific mortality in men with prostate cancer: a long-term survival analysis. Lancet Oncol 2008; 9: 1039-1047.
  • 8
    Lee IM. Physical activity and cancer prevention—data from epidemiologic studies. Med Sci Sports Exerc 2003; 35: 1823-1827.
  • 9
    Kushi LH, Byers T, Doyle C, Bandera EV, McCullough M, McTiernan A, et al. American Cancer Society Guidelines on Nutrition and Physical Activity for cancer prevention: reducing the risk of cancer with healthy food choices and physical activity. CA Cancer J Clin 2006;56:254281.
  • 10
    Secretan B, Straif K, Baan R, Grosse Y, El Ghissassi F, Bouvard V, et al. A review of human carcinogens. Part E: tobacco, areca nut, alcohol, coal smoke, and salted fish. Lancet Oncol 2009; 10: 1033-1034.
  • 11
    Ohta K, Endo T, Haraguchi K, Hershman JM, Onaya T. Ligands for peroxisome proliferator-activated receptor gamma inhibit growth and induce apoptosis of human papillary thyroid carcinoma cells. J Clin Endocrinol Metab 2001; 86: 2170-2177.
  • 12
    Panigrahy D, Huang S, Kieran MW, Kaipainen A. PPARgamma as a therapeutic target for tumor angiogenesis and metastasis. Cancer Biol Ther 2005; 4: 687-693.
  • 13
    Ondrey F. Peroxisome proliferator-activated receptor gamma pathway targeting in carcinogenesis: implications for chemoprevention. Clin Cancer Res 2009; 15: 2-8.
  • 14
    FDA Drug Safety Communication: ongoing safety review of Actos (pioglitazone) and potential increased risk of bladder cancer after two years exposure. Washington, DC: Food and Drug Administration.
  • 15
    Dormandy J, Bhattacharya M, van Troostenburg de Bruyn AR. Safety and tolerability of pioglitazone in high-risk patients with type 2 diabetes: an overview of data from PROactive. Drug Saf 2009; 32: 187-202.
  • 16
    Ramos-Nino ME, MacLean CD, Littenberg B. Association between cancer prevalence and use of thiazolidinediones: results from the Vermont Diabetes Information System. BMC Med 2007; 5: 17.
  • 17
    Monami M, Lamanna C, Marchionni N, Mannucci E. Rosiglitazone and risk of cancer: a meta-analysis of randomized clinical trials. Diabetes Care 2008; 31: 1455-1460.
  • 18
    Govindarajan R, Ratnasinghe L, Simmons DL, Siegel ER, Midathada MV, Kim L, et al. Thiazolidinediones and the risk of lung, prostate, and colon cancer in patients with diabetes. J Clin Oncol 2007; 25: 1476-1481.
  • 19
    Koro C, Barrett S, Qizilbash N. Cancer risks in thiazolidinedione users compared to other anti-diabetic agents. Pharmacoepidemiol Drug Saf 2007; 16: 485-492.
  • 20
    Lewis JD, Capra AM, Achacoso NS, Ferrara A, Levin TR, Quesenberry CPJr, et al. Thiazolidinedione therapy is not associated with increased colonic neoplasia risk in patients with diabetes mellitus. Gastroenterology 2008; 135: 1914-1923, e1.
  • 21
    Jick H, Garcia Rodriguez LA, Perez-Gutthann S.. Principles of epidemiological research on adverse and beneficial drug effects. Lancet 1998;352:17671770.
  • 22
    Essebag V, Platt RW, Abrahamowicz M, Pilote L.. Comparison of nested case-control and survival analysis methodologies for analysis of time-dependent exposure. BMC Med Res Methodol 2005;5:5.
  • 23
    Kramer JR, Davila JA, Miller ED, Richardson P, Giordano TP, El-Serag HB. The validity of viral hepatitis and chronic liver disease diagnoses in Veterans Affairs administrative databases. Aliment Pharmacol Ther 2008;27:274282.
  • 24
    Keshamouni VG, Reddy RC, Arenberg DA, Joel B, Thannickal VJ, Kalemkerian GP, et al. Peroxisome proliferator-activated receptor-gamma activation inhibits tumor progression in non-small-cell lung cancer. Oncogene 2004;23:100108.
  • 25
    Tsubouchi Y, Mukai S, Kawahito Y, Yamada R, Kohno M, Inoue K, et al. Meloxicam inhibits the growth of non-small cell lung cancer. Anticancer Res 2000;20:28672872.
  • 26
    Bren-Mattison Y, Van Putten V, Chan D, Winn R, Geraci MW, Nemenoff RA. Peroxisome proliferator-activated receptor-gamma (PPAR (gamma)) inhibits tumorigenesis by reversing the undifferentiated phenotype of metastatic non-small-cell lung cancer cells (NSCLC). Oncogene 2005; 24: 1412-1422.
  • 27
    Satoh T, Toyoda M, Hoshino H, Monden T, Yamada M, Shimizu H, et al. Activation of peroxisome proliferator-activated receptor-gamma stimulates the growth arrest and DNA-damage inducible 153 gene in non-small cell lung carcinoma cells. Oncogene 2002; 21: 2171-2180.
  • 28
    Han S, Ritzenthaler JD, Rivera HN, Roman J. Peroxisome proliferator-activated receptor-gamma ligands suppress fibronectin gene expression in human lung carcinoma cells: involvement of both CRE and Sp1. Am J Physiol Lung Cell Mol Physiol 2005; 289: L419-L428.
  • 29
    Russu WA. Thiazolidinedione anti-cancer activity: is inhibition of microtubule assembly implicated? Med Hypotheses 2007; 68: 343-346.
  • 30
    Chiang TA, Chen PH, Wu PF, Wang TN, Chang PY, Ko AM, et al. Important prognostic factors for the long-term survival of lung cancer subjects in Taiwan. BMC Cancer 2008;8:324.
  • 31
    Yoshizumi T, Ohta T, Ninomiya I, Terada I, Fushida S, Fujimura T, et al. Thiazolidinedione, a peroxisome proliferator-activated receptor-gamma ligand, inhibits growth and metastasis of HT-29 human colon cancer cells through differentiation-promoting effects. Int J Oncol 2004;25:631639.
  • 32
    Chintharlapalli S, Papineni S, Safe S. 1,1-Bis(3'-indolyl)-1-(p-substituted phenyl)methanes inhibit colon cancer cell and tumor growth through PPARgamma-dependent and PPARgamma-independent pathways. Mol Cancer Ther 2006; 5: 1362-1370.
  • 33
    Marin HE, Peraza MA, Billin AN, Willson TM, Ward JM, Kennett MJ, et al. Ligand activation of peroxisome proliferator-activated receptor beta inhibits colon carcinogenesis. Cancer Res 2006;66:43944401.
  • 34
    Takashima T, Fujiwara Y, Higuchi K, Arakawa T, Yano Y, Hasuma T, et al. PPAR-gamma ligands inhibit growth of human esophageal adenocarcinoma cells through induction of apoptosis, cell cycle arrest and reduction of ornithine decarboxylase activity. Int J Oncol 2001;19:465471.
  • 35
    Chen DS. Hepatocellular carcinoma in Taiwan. Hepatol Res 2007; 37( Suppl 2): S101-S105.
  • 36
    Chen CL, Yang HI, Yang WS, Liu CJ, Chen PJ, You SL, et al. Metabolic factors and risk of hepatocellular carcinoma by chronic hepatitis B/C infection: a follow-up study in Taiwan. Gastroenterology 2008; 135: 111-121.
  • 37
    Chemin I. Evaluation of a hepatitis B vaccination program in Taiwan: impact on hepatocellular carcinoma development. Future Oncol 2010; 6: 21-23.
  • 38
    Liu CH, Liang CC, Liu CJ, Tsai HB, Hung PH, Hsu SJ, et al. Pegylated interferon alpha-2a plus low-dose ribavirin for the retreatment of dialysis chronic hepatitis C patients who relapsed from prior interferon monotherapy. Gut 2009; 58: 314-316.
  • 39
    Liu CH, Liu CJ, Lin CL, Lin JW, Chen SI, Hung PH, et al. Pegylated interferon-alpha-2a plus ribavirin for treatment-naive Asian patients with hepatitis C virus genotype 1 infection: a multicenter, randomized controlled trial. Clin Infect Dis 2008; 47: 1260-1269.
  • 40
    Liu CH, Liang CC, Lin JW, Chen SI, Tsai HB, Chang CS, et al. Pegylated interferon alpha-2a versus standard interferon alpha-2a for treatment-naive dialysis patients with chronic hepatitis C: a randomised study. Gut 2008; 57: 525-530.
  • 41
    Mussig K, Staiger H, Kantartzis K, Fritsche A, Kanz L, Haring HU. Type 2 diabetes mellitus and risk of malignancy: is there a strategy to identify a subphenotype of patients with increased susceptibility to endogenous and exogenous hyperinsulinism? Diabet Med 2011; 28: 276286.
  • 42
    Li D. Metformin as an antitumor agent in cancer prevention and treatment. J Diabetes 2011; 3: 320-327.
  • 43
    Ray WA. Evaluating medication effects outside of clinical trials: new-user designs. Am J Epidemiol 2003; 158: 915-920.
  • 44
    Lewis JD, Ferrara A, Peng T, Hedderson M, Bilker WB, Quesenberry CPJr, et al. Risk of bladder cancer among diabetic patients treated with pioglitazone: interim report of a longitudinal cohort study. Diabetes Care 2011; 34: 916-922.
  • 45
    Nemenoff RA. Peroxisome proliferator-activated receptor-gamma in lung cancer: defining specific versus “off-target” effectors. J Thorac Oncol 2007; 2: 989-992.
  • 46
    Chaggar PS, Shaw SM, Williams SG. Review article: thiazolidinediones and heart failure. Diab Vasc Dis Res 2009; 6: 146-152.
  • 47
    Betteridge DJ. CHICAGO, PERISCOPE and PROactive: CV risk modification in diabetes with pioglitazone. Fundam Clin Pharmacol 2009; 23: 675-679.
  • 48
    Pratley RE. The PROactive Study: pioglitazone in the secondary prevention of macrovascular events in patients with type 2 diabetes. Curr Diab Rep 2006; 6: 45-46.
  • 49
    Bilous RW. Rosiglitazone and myocardial infarction: cause for concern or misleading meta-analysis? Diabet Med 2007;24:931933.
  • 50
    Diamond GA, Bax L, Kaul S. Uncertain effects of rosiglitazone on the risk for myocardial infarction and cardiovascular death. Ann Intern Med 2007; 147: 578-581.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_25509_sm_suppinfotabs.doc155KSupporting Information Tables

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.