Intended for healthcare professionals


Thiazolidinediones for type 2 diabetes

BMJ 2000; 321 doi: (Published 29 July 2000) Cite this as: BMJ 2000;321:252

New agents reduce insulin resistance but need long term clinical trials

  1. Andrew J Krentz (a.j.krentz{at}, consultant in diabetes and endocrinology,
  2. Clifford J Bailey (c.j.bailey{at}, head of diabetes research,
  3. Arne Melander (arne.melander{at}, head of NEPI Foundation
  1. Southampton General Hospital, Southampton SO16 6YD
  2. Department of Pharmaceutical Sciences, University of Aston, Birmingham B4 7ET
  3. Medical Research Centre, Malmö University Hospital SE-205 02, Sweden

    Insulin resistance, or more appropriately the reduced action of insulin, is a prominent defect in type 2 diabetes.1 It is commonly present in people before diabetes has developed and has even been observed in euglycaemic relatives of patients with type 2 diabetes.2 It has been proposed that the reduced action of insulin is fundamental to the cardiovascular risk factors that are part of the syndrome of insulin resistance.3 Avoidance of obesity and adequate levels of physical activity are non-pharmacological cornerstones of the fight against insulin resistance. Before the introduction of troglitazone in 1997 metformin was the only drug able to sensitise target tissues (skeletal muscle, adipose tissue, and the liver) to insulin. Troglitazone was the first of a new class of drugs with direct insulin sensitising actions—the thiazolidinediones (also known as glitazones).4 Troglitazone has now been superseded by more potent agents, rosiglitazone and pioglitazone.

    Thiazolidinediones activate nuclear peroxisome proliferator activated receptor γ (PPAR-γ), which is expressed predominantly in adipose tissue. 4 5 Insulin action is improved through the increased transcription of genes in adipocyte differentiation and lipid and glucose metabolism. Insulin resistance is reduced when assessed with techniques such as the glucose clamp.1 Blood glucose concentrations are reduced in concert with a fall in circulating insulin concentration. In addition, thiazolidinediones maintain the insulin content of β cells of the pancreas in animal models.4 Improvements in glucose metabolism may be partly attributable to a reduction in the concentrations of circulating non-esterified fatty acids and reduced activity of the glucose-fatty acid (Randle) cycle.4 Patients with insulin resistance often have elevated serum concentrations of triglyceride with low concentrations of high density lipoprotein cholesterol, and this dyslipidaemia contributes to their increased risk of atherosclerotic cardiovascular disease.3 Thiazolidinediones increase the concentration of high density lipoprotein cholesterol, and rosiglitazone protects against endothelial dysfunction and lowers blood pressure in insulin resistant and hypertensive rats.4 Although weight gain is common with thiazolidinediones, reports in humans of a redistribution away from visceral adiposity with troglitazone are of interest; this depot is closely linked with the syndrome of insulin resistance.6 Thus, thiazolidinediones might reduce cardiovascular risk.

    Troglitazone was available in Britain for a few weeks in 1997 before its distributor (GlaxoWellcome) withdrew the drug in response to reports from Japan and the United States of severe and unpredictable hepatoxicity. In the United States troglitazone was withdrawn in March 2000, when the Food and Drug Administration had received reports of 61 deaths from hepatic failure and seven liver transplants associated with the drug.7 Troglitazone remains available in Japan and several other countries.

    Last week saw the arrival of rosiglitazone in Europe. This drug, together with pioglitazone, has been available in the United States (and elsewhere) since 1999. Pioglitazone is expected to be launched in Europe at the end of the year. Both drugs have been granted limited indications in defined circumstances: in combination with metformin in obese patients with insufficient glycaemic control and in combination with sulphonylureas if metformin is either not tolerated or contraindicated (such as in renal impairment).8 These stipulations contrast with the situation in the United States, where both drugs are licensed for use as monotherapy (when non-pharmacological measures have failed). Clinical trials show that combination therapy using a thiazolidinedione with metformin (the main action of which is to reduce glucose production by the liver) or a sulphonylurea (to increase endogenous insulin secretion) is particularly effective in lowering glucose concentrations.4 Substantial reductions in insulin doses have been reported when thiazolidinediones are used in combination with insulin.4 However, an increased incidence of cardiac failure was seen in clinical trials in which rosiglitazone was used in combination with insulin.9 Combination therapy of rosiglitazone with insulin is therefore contraindicated (although pioglitazone is approved for use in combination with insulin treatment in the United States).

    Reassuringly, extensive use of rosiglitazone and pioglitazone has produced little evidence that it has caused hepatic impairment. However, caution dictates that these drugs are contraindicated in patients with hepatic impairment or if pretreatment concentrations of alanine aminotransferase are raised more than 2.5 times the upper limit of normal.9 Cardiac failure of any degree (past or present) is a contraindication, and patients with reduced cardiac reserve should be monitored closely.9 There is also a risk of pregnancy in anovulatory women with insulin resistance. Pioglitazone induces cytochrome P450 isoform CYP3A4, raising the possibility of drug interactions, such as with oral contraceptives.

    The United Kingdom prospective diabetes study shows that better glycaemic control reduces the risk of microvascular complications.10 In addition, the trial exposes the need for additional drugs that are effective against diabetes.11 The thiazolidinediones have had a faltering start. An appraisal by the National Institute for Clinical Excellence (NICE) in the United Kingdom is planned. However, over a million patients have now been given these drugs, and the continuing paucity of publications in peer reviewed journals is a concern.12 The rising global incidence of type 2 diabetes13 suggests that these drugs could have an important impact on diabetes care.


    Competing interests: AJK has provided ac hoc consultancy to GlaxoWellcome and Takeda (distributors of troglitazone and pioglitazone, respectively). CJB has provided ad hoc consultancy for SmithKline Beecham (distributor of rosiglitazone) and GlaxoWellcome.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    View Abstract