Serving To Educate Primary Care Clinicians On Metabolic Issues
Evidence Based Medicine Header 1
Medical Education Header 2
Peer-to-peer Exchange Header 3
Metabolic Diseases Header 4
Back: Rough Seas for TZDs

Clinical Compass: Steering You in the Best Direction

Our Philosophy: The practice of medicine should be the artful application of judgment to science that produces therapeutic wisdom.

Clinical Compass is a commentary that will review clinical trials and events in medicine that impact patient care. After the review of each item, a summary conclusion ‘direction’ will be printed on the compass. Because many of the issues reviewed here may remain controversial, clinicians are encouraged to evaluate the full data for themselves.

Guest Commentator
Louis Kuritzky, MD
Clinical Assistant Professor, Department of Family Medicine
University of Florida
Gainsville, FL 

Rough Seas for Thiazolidinediones (TZDs) 

Setting: What is happening on the TZD voyage?
Heavy hopes have been pinned upon the TZDs as foundation medications for type 2 diabetes (DM2). Physiologic attributes of TZDs hold huge intellectual appeal. In addition to lowering glucose, TZDs improve one of the fundamental defects in diabetes—insulin resistance. Additionally, abdominal obesity, a marker for metabolic risk, is favorably affected by TZD treatment.

The PROactive trial [Dormandy JA, et al. Lancet. 2005;366:1279-1289], the largest randomized, controlled trial of TZDs (pioglitazone) on secondary prevention of macrovascular endpoints in DM2, failed to meet its primary endpoint, which was a composite of all-cause mortality, nonfatal myocardial infarction (MI), stroke, acute coronary syndrome, endovascular or surgical intervention in the coronary or leg arteries, and above-the-ankle amputation. Nonetheless, results of secondary endpoints (the main secondary endpoint was the composite of all-cause mortality, nonfatal MI, and stroke) suggest great promise. Since TZDs can be used in a complementary fashion with essentially all other available agents for DM2, and may even exert favorable effects upon lipids, the TZDs figure prominently in a disease process that most often requires polypharmacy. Well, what’s not to like? Obstacles Along the Voyage
At first, things seemed like smooth sailing for the TZDs. Although edema, hemodilution, and even frank heart failure were problematic from time to time, most patients tolerated the TZDs well. Weight gain, seen in a minority of patients, was perhaps less troublesome to most because it represented fluid accumulation, rather than increased fat mass. Then in June 2007, a disquieting meta-analysis by Nissen and Wolski [N Engl J Med. 2007;356:2457-2471] suggested that, based upon data from 42 clinical trials, the odds ratio for MI was 43% greater and for death 64% greater than persons NOT receiving TZDs. Although the meta-analysis focused solely on rosiglitazone, concern arose over whether the observed negative odds ratios were also pertinent to pioglitazone. Yet, pioglitazone is not the twin sister of rosiglitazone; for example, pioglitazone has a much more favorable lipid profile. Further, even though the PROactive trial did not prove its primary endpoint, there was a trend towards favorable results, and certainly no suggestion of an untoward cardiovascular (CV) risk profile with pioglitazone. So is this the time to turn the ship around and head home? 

The Next Lighthouse
Might a peek at some interim data from a very large rosiglitazone trial provide reassurance to those who were ready to jump ship from TZDs based solely upon damaging rosiglitazone data? Because the findings by Nissen and Wolski created quite a storm, additional data points might either send rosiglitazone to Davey Jones’ locker, or possibly buoy it up.

The RECORD trial (Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes) [forgive the spelling of ‘Glycaemia’… you know those Brits… left side of the highway and all…] is a randomized, multicenter, open-label, non-inferiority trial initiated to follow DM2 patients (n=4,447) for at least 6 years. The objective is to determine whether the addition of rosiglitazone in persons who had not achieved adequate glycemic control on metformin plus a sulfonylurea would have a more favorable risk than the addition of placebo on hospitalization or death from CV disease (the primary endpoint).

The RECORD trial (Home PD, et al. N Engl J Med. 2007;357:28-38) has not yet been completed. An interim analysis (mean follow-up 3.75 years) shows that there was no statistically significant difference in the incidence of MI, CV death, or all-cause mortality between those who received rosiglitazone and those who received placebo. Since the mean follow-up approaches 4 years, one would think that if rosiglitazone produced a meaningful CV risk, such a large population might demonstrate that. But one of the criticisms of this interim analysis is the very low event rate of 4.5 events per 1000 person-years, which is close to the normal population event rate and only 40% of other CV studies in patients with diabetes.

The design of the RECORD trial also has been questioned since the primary endpoint is a composite endpoint that tends to push the statistics toward even. Not surprisingly, the interim analysis shows that rosiglitazone WAS associated with more heart failure...but then, we were already aware of that.

But the saga goes on. Two NEW meta-analyses appeared in the September 12, 2007, issue of JAMA. The meta-analysis of pioglitazone [Lincoff AM, et al. JAMA. 2007;298:1180-1188] suggests reassuring news. On the other hand, the rosiglitazone meta-analysis [Singh S, et al. JAMA. 2007;298:1189-1195], which is based on studies of 12 months’ minimum duration and suggests a  42% increased risk of MI, again casts storm clouds over the path of rosiglitazone.

Which Way To Steer
It is reassuring that an interim analysis of a large trial (RECORD) specifically designed to discern CV effects of rosiglitazone does not show any negative impact. Of course, there remains the possibility that the numbers will look different at the conclusion of the trial… we’ll have to wait and see. The ultimate solution would be a large head-to-head prospective trial of rosiglitazone vs. pioglitazone vs. placebo. With the favorable headwinds enjoyed by pioglitazone, the manufacturers of pioglitazone have little to gain, so why would they fund such a trial? In the meantime, it appears prudent to optimize CV risk reduction in our DM2 patients, whether they are using a TZD or not. Patients (or clinicians) who have specific concerns about the CV risk of rosiglitazone might be more comfortable switching to pioglitazone, which has a more favorable lipid profile than rosiglitazone. Some clinicians have chosen to view the accumulated data about rosiglitazone as sufficient to discontinue its use entirely. Others, taking a position seemingly like the FDA, who after reviewing all of the data has chosen NOT to limit the use of rosiglitazone or add any warnings about CV mortality (only heart failure risks) to the prescribing information, may feel that the very small absolute risk suggested by these meta-analyses is insufficient to motivate switching from rosiglitazone.  There appears to be little lost by consideration of a switch from rosiglitazone to pioglitazone in those who are willing to do so. Until further information is available, the authors suggest: Change sails if feasible (that is, offer a switch from rosiglitazone to pioglitazone for those who are willing).Rough Seas Image