IRIS Stroke

From Wiki Journal Club
Jump to: navigation, search
Inzucchi SE, et al. "Pioglitazone Prevents Diabetes in Patients With Insulin Resistance and Cerebrovascular Disease". Diabetes Care. 2016. 39(10):1684-1692.

Clinical Question

In insulin-resistant nondiabetic patients with history of cardiovascular disease, does Pioglitazone help prevent diabetes?

Bottom Line

Pioglitazone use in non-diabetic patients with a history of an ischemic event showed a 2.8% absolute reduced risk of subsequent ischemic events and 3.9% absolute reduced risk of progression to diabetes.

Major Points

Pioglitazone is a thiazolidinedione that has been historically used for treatment of type 2 diabetes mellitus by working to increase insulin sensitivity. Currently the only accepted method of prevention of diabetes is lifestyle modifications including diet and exercise.

Recently the IRIS (Insulin Resistance Intervention after Stroke) trial demonstrated the usefulness of pioglitazone in reducing the risk of ischemic events in nondiabetic patients with insulin resistance who had a recent ischemic event by 24%. A prespecified secondary outcome of the IRIS trial was the prevention of diabetes. The study found that pioglitazone reduced the risk of diabetes by 52%. Pioglitazone is now the first hyperglycemic agent to be shown to prevent both diabetes and secondary cerebrovascular events in insulin-resistant nondiabetic patients.


There are currently no guidelines or recommendations to use pioglitazone to reduce the progression of diabetes or prevent ischemic events.


  • Multicenter, double-blind, parallel-group, randomized, controlled trial
  • N=3,895
    • Pioglitazone/therapy (n=1,939)
    • Placebo (n=1937)
  • Setting: 179 centers in seven countries (Australia, Canada, Germany, Israel, Italy, United Kingdom, and United States)
  • Enrollment: February 2005 - January 2013
  • Mean follow-up: 4.8 years
  • Analysis: Intention-to-treat
  • Primary outcome: Diagnosis of diabetes (two fasting plasma glucoses (FPG) ≥ 126mg/dL, two random plasma glucose ≥200 mg/dL, or the presence of compelling indications of hyperglycemia)


Inclusion Criteria

  • Over the age of 40
  • Qualifying ischemic stroke or TIA within 6 months
  • HOMA-IR (Homeostasis Model Assessment of Insulin Resistance) score > 3.0 tested >14 days after the ischemic event

Exclusion Criteria

  • Past diagnosis of diabetes with use of antihyperglycemic therapy
  • HbA1C ≥ 53 mmol/mol (7.0%) at the screening visit (repeated and confirmed)
  • FPG ≥ 126 mg/dL at the screening visit (repeated and confirmed)
  • Previous diagnosed heart failure, bladder cancer, moderate to severe dependent pitting edema, high risk of bladder cancer, irreversible medical condition with predicted survival <4 years, and oral corticosteroid use

Baseline Characteristics

  • Mean age: 63.5
  • 65% male, 35% female
  • Mean BMI: 30.0 kg/m2
  • Mean FPG: 98.2 mg/dL
  • Mean HbA1c: 5.8%
  • Insulin: 22.4 µIU/mL
  • HOMA-IR: 5.4


  • Randomized to pioglitazone 15mg or placebo in a 1:1 ratio
  • Pioglitazone group titrated 30 mg at the 4th week and 45mg at the 8th week
  • Patients were contacted every 2 weeks during the first 3 months to be assessed for adherence, adverse events, or outcome events
  • Patients were then contacted quarterly including an annual physical exam


Comparisons are pioglitazone vs. placebo

Primary Outcomes

Progression to diabetes
3.8% vs 7.7% (HR 0.48; 95% CI 0.33-0.69; P<0.0001; NNT=49)

Other Outcomes

Decreased by 24% from baseline in the pioglitazone group
Increased by 7% from baseline in the placebo group

Subgroup Analysis

Progression to diabetes, IFG (impaired fasting glucose) at baseline (FPG ≥100mg/dL)
6.5% vs 15.0% (NNT=11)
Progression to diabetes, normal FPG at baseline (FPG ≥100mg/dL)
1.8% vs 2.6% (NNT=125)

Adverse Events

Presented in the order pioglitazone, placebo; p-value

  • Weight gain (lbs): avg 5.8 gain, 1.2 loss; 0.001
  • Edema (%): 35.6, 24.9; 0.001 (NNH=9)
  • Bone fractures requiring hospitalization or surgery (%) 5.1, 3.2; 0.003 (NNH=52)


  • Some of the patients who were included in the study did have diabetes as defined by the ADA of HbA1c ≥6.5%. At baseline, 6.3% of the patients had an HbA1C between 6.5-7.0%. Therefore, the study is not universal to all non-diabetic patients especially because those with higher baseline HbA1c and FBG had better outcomes.
  • Adherence was reported as “not ideal” in this trial. There were many adverse events that caused people to become non-adherent to the intervention. They also state that if adherence would have been higher the results may have shown greater risk reduction but also more side effects. This trial used an intention-to-treat analysis so they included the date from these patients even if they knew they were not adherent. This model shows data that is more applicable to a real life situation because real patients will stop taking therapy if they do not like the adverse events. Because they included this data the study does not show the most accurate treatment effect.


U.S. National Institute for Neurological Disorders and Stroke of the National Institutes of Health. There are no conflicts to disclose.

Further Reading

  • IRIS Trial Kernan WN et al. “Pioglitazone after Ischemic Stroke or Transient Ischemic Attack”. N Engl J Med. 2016 April. 375: 1321-1331. DOI: 10.1056/NEJMoa1506930