Evidence Table 16

Adverse events in comparative observational studies of TZDs vs other agents

Author, year (Quality rating)Study designComparisonStudy objectiveTime period coveredData sourceSample sizePopulation; baseline characteristicsOutcome measuresStatistical methodsResults
Hanefeld 2006 (POOR)Prospective cohortpioglitazone vs glibenclamideTo compare efficacy of add- on therapy with pioglitazone vs add-on therapy with glibenclamideNot reported75 primary care sites in Germany500
250 pio, 250 glibenclamide
Patients with type 2 diabetes insufficiently controlled with metformin alone in primary care practicesProgression to insulin therapyLog-rank testProgression to insulin:
Pio: 55/250 (22%)
Glibenclamide: 138/250 (55%)
Mean annual progression rates:
Pio: 6.6%
Glibenclamide: 16.4%
Hartung 2005Case-controlTZDs vs SU vs metformin vs metformin + SU vs insulin vs insulin + TZD vs alpha- glucosidase inhibitorOregon Medicaid claims data 1940 Hospital admission for heart failureAdjusted odds ratio (95% CI)
TZDs: 1.37 (0.98, 1.92)
SU: 0.95 (0.73, 1.24)
Metformin: 0.97 (0.72, 1.30)
Metformin + SU: 0.90 (0.60, 1.34)
Insulin: 1.25 (0.92, 1.69)
Insulin + TZDs: 1.35 (0.84, 2.18)
Alpha-glucosidase inhibitor: 0.82 (0.28, 2.18)
Johannes 2007 (GOOD)Retrospective cohortTZDs vs metformin plus sulfonylureaTo evaluate whether the risk of coronary heart disease differs among adult diabetic patients treated with TZDs and similar patients treated with combined oral metformin and SU therapyJanuary 1, 1999 to June 30, 2002US health care claims database, from 17 states25,140
12,570 metformin + SU initiators, matched to 12,570 TZD initiators
Incidence of acute MI, coronary revascularization, sudden deathCox Proportional Hazards regression, adjusted for age at study drug initiation, gender, year of cohort entry, insulin use, any oral antidiabetic medication use, claims indicators of smoking, and claims evidence of MI, angina, atherosclerosis, CR, congestive heart failure, hypertension, hyperlipidemia, and obesityAdjusted hazard ratio (95% CI)
TZDs: 1.02 (0.87, 1.20)
Metformin + SU (reference): 1.00
Kahler 2007 (FAIR)Retrospective cohortSU monotherapy vs metformin monotherapy vs metformin + SU vs TZD use vs no drugsTo evaluate the impact of several classes of oral antihyperglycemic therapy relative to SU monotherapy on all-cause mortality among a cohort of patietns with diabetes.October 1, 1998 to September 30, 2000Veterans Health Administration data from the Diabetes Epidemiology Cohort.39,721
19,053 SU monotherapy, 2,988 metformin monotherapy, 13,820 metformin + SU, 673 TZD users (alone or in combination with other oral agents), 3,185 no drugs
All-cause mortalityAdjusted odds ratio (95% CI)
SU (reference): 1.00
TZDs: 1.04 (0.75, 1.46)
Metformin: 0.87 (0.68, 1.10)
Metformin + SU: 0.92 (0.82, 1.05)
No drugs: 0.90 (0.74, 1.09)
Karter 2005 (FAIR)Retrospective cohortPioglitazone vs SU vs metformin vs insulinTo determine if short-term use of pioglitazone is associated with increased risk of admission to hospital because of heart failureOctober 1999 to November 2001Kaiser
Diabetes Registry
3556 pio, 5921 SU, 11,937 metformin, 2026 insulin
Hospital admission for heart failureAdjusted hazard ratio (95% CI)
Pioglitazone: 1.28 (0.85, 1.92)
Insulin: 1.56 (1.00, 2.45)
Metformin: 0.70 (0.49, 0.99)
SU (reference): 1.00
Koro 2007 (FAIR)Nested case-controlTZDs vs other anti- diabetic agentsTo evaluate the risk of breast, colon, and prostate cancers developing in patients exposed to TZDs compared with other anti- diabetic agentsJanuary 1, 1997 to December 31, 2004US Integrated
Services database
513 breast cancer cases, matched to 2557 controls 408 colon cancer cases, matched to 2027 controls 643 prostate cancer cases, matched to 3176 controls
Incidence of breast, colon, and prostate cancersConditional logistic regression. Adjusted odds ratio. Matching to control for age, sex,calendar time and years of recorded history in the database before the index dateAdjusted odds ratio (95% CI)
TZD use (mono- or combination therapy)
compared to other anti-diabetic agents
Breast cancer: 0.89 (0.68, 1.15)
Colon cancer: 1.03 (0.84, 1.32)
Prostate cancer: 1.04 (0.83, 1.31)
Masoudi 2005 (GOOD)Retrospective cohortTZDs vs metformin vs no insulin sensitizerTo provide information about the balance of risks and benefits relevant to recommendations for use of TZDs and metforminApril 1998 to March 1999, and July 2000 to June 2001Medicare16,417
2,226 TZD, 1,861 metformin, 12,069 no insulin sensitizer
Older patients with diabetes and heart failurePrimary: time to death due to all causesAdjusted for patient, provider, and hospital characteristics, sampling time frame, and differences in other medical treatment at dischargeAdjusted hazard ratio (95% CI) for all cause mortality:
TZD use: 0.87 (0.80, 0.94)
Metformin use: 0.86 (0.78, 0.97)
SU use: 0.99 (0.91, 1.08)
Insulin use: 0.96 (0.88, 1.05)
TZD + Metformin: 0.76 (0.58, 0.99)
McAfee 2007 (GOOD)Retrospective cohortrosiglitazone vs metformin vs sulfonylureaTo compare risk of MI and coronary revascularization in type 2 diabetic patients treated with rosiglitazone, metformin, or sulfonylureaJuly 1, 2000 through December 31, 2004Health insurance claim data from a managed care organization (Ingenix Research Database)26,931
(8977 rosiglitazone, 8977 metformin, 8977 sulfonylurea)
Patients with type 2 diabetesFirst occurrence of MI, coronary revascularization (CR), and composite of hospitalization for MI or CRKaplan Meier curves, Incidence rates and 95%
Adjusted for baseline covariates (age, sex, total cost, hyperlipidemia, and nitrate use)
Adjusted hazard ratio (95% CI) for composite outcome:
rosglitazone monotherapy vs metformin
monotherapy: 1.07 (0.85, 1.34)
rosiglitazone monotherapy vs SU monotherapy: 0.82 (0.67, 1.02)
rosiglitazone combined with insulin vs other oral antidiabetics combined with insulin: 0.88 (0.59, 1.32)
Rosiglitazone therapy vs all other non- rosiglitazone therapies: 0.93 (0.80, 1.10)
No increased risk of MI or coronary revascularization for rosiglitazone therapy vs other oral hypoglycemic therapies

From: Evidence Tables

Cover of Drug Class Review: Thiazolidinediones
Drug Class Review: Thiazolidinediones: Final Report Update 1 [Internet].
Norris SL, Carson S, Thakurta S, et al.
Portland (OR): Oregon Health & Science University; 2008 Aug.
Copyright © 2008, Oregon Health & Science University, Portland, Oregon.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.