HOME HELP SUBSCRIBE ARCHIVE SEARCH TABLE OF CONTENTS FEEDBACK EDITORIAL BOARD ABOUT DOC NEWS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Services Email this article to a friend Similar articles in this journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hirsch, I. B. Search for Related Content PubMed Articles by Hirsch, I. B. Social Bookmarking                What's this?

Agents for Change

Irl B. Hirsch, MD

The introduction of troglitazone (Rezulin, Parke-Davis) in 1997 was highlyanticipated among those in the diabetes community. For the first time, itappeared we would have an agent that attacked the fundamental problem in type2 diabetes: insulin resistance. It was reasoned that agents that bind tonuclear receptors in fat cells and modulate the transcription of genesinvolved in the pathways of insulin action should be the basis of thetreatment of type 2 diabetes, impaired glucose tolerance, and polycysticovarian syndrome (PCOS).

Hardly a panacea, however, the 8-year history of thiazolidinediones hasbeen cluttered with success, controversy, and disappointment—includingdeaths from liver failure.¹ Clearly, our understanding of theseagents is still evolving.

Initially troglitazone, and later rosiglitazone and then pioglitazone, wereapproved for the treatment of type 2 diabetes as either monotherapy orcombination therapy with other oral agents or with insulin. As a group, theseagents are "moderately effective" in lowering plasma glucoselevels.² Review of the literature shows that, in general, theseagents lower glycated hemoglobin (A1C) by about 1–1.5%.^2,3This is intermediate to the weakest agents (neteglinide and acarbose) and thestronger agents (glimeperide, glyburide, and metformin).^4,5 Sincethere is variability among studies in terms of anti-hyperglycemic effects,even if one were to argue that the effects of all of the agents are about thesame,⁶ there is the important issue of cost. According todrugstore.com(accessed 12/23/04), one month of generic glyburide 10 mg daily would cost$24, metformin 1 gram twice daily would cost $56, rosiglitazone 8 mg oncedaily would cost $144, and pioglitazone 45 mg daily would cost $165. Thequestion I have been asking, as have many pharmacy and therapeutics committeeswhere these drugs are either not on formulary(www.ghc.org/pharmacy/index.jhtml)or have restricted use(www.vapbm.org/natform/NATFORM010-04.xls),is quite simple: Is the higher cost of these agents worth the six- tosevenfold increase compared with generic glyburide?

This is an important question; however, evolving data suggests that theseagents may have more to offer than the modest glycemic response. These agentshave been shown to reduce a variety of surrogate end points of cardiovasculardisease. These include urinary albumin in addition to key cytokines (such asinterleukin 6 and tumor necrosis factor alpha), acute phase protein (such asC-reactive protein [CRP]), pro-coagulation factors (such as von Willebrandfactor and plasminogen-activator inhibitor type 1), adhesion molecules, andfree fatty acids.^2,7 It is also clear that these agents result inan increase in adiponectin and LDL particle size in individuals with orwithout diabetes,^7,8 although a recent head-on-head trial ofpioglitazone versus rosiglitazone revealed an advantage with theformer.⁷ Taken together, one would expect these effects to resultin a reduction of cardiovascular events.

However, as we are frequently reminded, surrogate outcomes may be extremelymisleading.^10,11 Nevertheless, Choi and colleagues recentlyreported that when 95 patients were randomized to either rosiglitazone orcontrol to achieve a similar level of glycemic control (A1C of 7.2%), thosereceiving the thiazolidinedione had significantly less restenosis at the endof 6 months.¹² This is not a trivial point, since many havereported coronary artery stent restenosis in type 2 diabetes to be above 50%.Choi reported stent restenosis to decrease from 38.2% to 17.6%.¹⁰Not surprisingly, CRP levels, a surrogate marker of inflammation, were alsoreduced.

Finally! A study with a clinically meaningful end point suggesting thatindeed, by reducing inflammation (note: not blood glucose in thisstudy) the progression of atherosclerosis may be reduced. Does this studyfinally end all of the debate about the use of these agents for those withtype 2 diabetes? I would suggest not.

First, there is a significant risk of congestive heart failure withthiazolidinediones,^13,14 to the point that the American DiabetesAssociation and the American Heart Association published a consensus statementwith specific recommendations about the use and monitoring of patients withsymptomatic heart disease, heart failure, and edema.¹⁵ Second,weight gain with these agents can be profound. One recent report noted that16% of patients gained 5.1 to 10 kg, while 2% gained more than 10kg.¹⁶ Unfortunately, there is no way to predict who will gain themost weight prior to initiating the drug. Although the fat gain does not comefrom the more metabolically active visceral fat,¹⁷ I can't believethat adding more than 5 kg of inactive fat can be a good thing.

More recently, there are data to suggest that these agents may promoteosteoporosis.¹⁸ In general, osteoporosis is not a major publichealth problem in this population, since obesity appears to protect againstosteoporosis. Still, as with many previous drugs—going back beforephenformin to the more recent escapade with rofecoxib (Vioxx, Merck)—itis not unusual for a signal to be picked up long after a drug is released.

So what is the conclusion? My take is that we have a lot more to learnabout thiazolidinediones. There may be tremendous benefits from using theseagents in the prevention of type 2 diabetes. The DREAM study, forexample, is assessing how rosiglitazone may compare to ramipril in preventingdiabetes in a high-risk population. Furthermore, we need confirmation abouthow thiazolidinediones may retard atherogenesis. The PROactive Study isassessing this in more than 5,000 patients.¹⁹ Ideally, we need moreinformation as to who is at highest risk to develop edema and weight gain sothese agents can be avoided in that population.

It may very well be that these agents would be better used early in thecourse of diabetes—even before the disease develops—as opposed tohow many use them now: to avoid insulin therapy. There are also provocativedata suggesting these agents would be well suited for other disease processes,such as ulcerative colitis, psoriasis, multiple sclerosis, andasthma.²⁰

Finally, two ongoing studies may be critical in teaching us how to best usethese drugs. The first is A Diabetes Outcome Progression Trial (ADOPT), whichwill compare the effectiveness of rosiglitazone, metformin, and glyburide inearly type 2 diabetes.²¹ The other trial to note is the Action toControl Cardiovascular Risk in Diabetes (ACCORD), in which rosiglitazone isbeing used to assist in treating A1C levels to a target of 6.0%.

For now, I believe we still need to be cautious. While we finally have datasuggesting thiazolidinediones may retard atherosclerosis, adverse effects andexpense preclude their use in many patients. Hopefully, in the near future wewill have answers to many of today's questions about this interesting class ofdrugs.

Footnotes

References are available athttp://docnews.diabetesjournals.org/misc/FebReferences05.pdf.

Irl B. Hirsch, MD, is Editor-in-Chief of DOC News, and medicaldirector of the diabetes care center at University of Washington MedicalCenter.

CiteULike    Del.icio.us    Digg    Reddit    Technorati    What's this?

Services Email this article to a friend Similar articles in this journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hirsch, I. B. Search for Related Content PubMed Articles by Hirsch, I. B. Social Bookmarking                What's this?