TODAY trial results show that rosiglitazone, but not lifestyle intervention, improves glycemic control when added to metformin monotherapy

The New England Journal of Medicine recently published the results of the Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) trial, which evaluated the ability of metformin monotherapy to provide sustained glycemic control compared to metformin in combination with the TZD rosiglitazone (GSK’s Avandia) or lifestyle intervention. The TODAY trial was initiated when reports suggested that up to 25% or more of new-onset diabetes cases in youth (depending on location) were due to type 2 diabetes. At the time of the initiation of TODAY, thiazolidinedione therapy was felt to be a valid treatment to reduce insulin resistance in this cohort, however, the findings of TODAY suggest more about more aggressive, dual therapy to than rosiglitazone itself.

The TODAY trial (ClinicalTrials.gov identifier: NCT00081328) entered 927 overweight youth (ages 10-17) recently diagnosed with type 2 diabetes into a run-in period of two to six months to attain an A1c level <8.0% using metformin alone. The 699 youth who completed the run-in were randomized 1:1:1 to metformin monotherapy (1000 mg twice daily; n=232) or to metformin in combination with rosiglitazone (4 mg twice daily n=233) or lifestyle intervention (n=234). The participants were followed for an average of 3.8 years (range of 2 - 6.5 years) and the primary endpoint was time to treatment failure, defined in two ways: 1) A1c level ≥ 8% for six months; or 2) the continued need for insulin three months after its initiation for metabolic decompensation or a second occurrence of decompensation within three months after the discontinuation of insulin.

Study results showed that metformin/rosiglitazone combination therapy provided a significantly lower rate of treatment failure (38.6%) than metformin alone (51.7%, p=0.006). The failure rate of metformin plus lifestyle intervention (46.6%) did not differ significantly from either of the other two groups, which was a significant disappointment and underscores that lifestyle modification may have been too late or insufficient. The reasons for treatment failure and the median time to failure (quite a short 11.5 months) did not differ significantly across groups. Furthermore, adjusting for sex, race/ethnic group, baseline BMI, or baseline A1c did not modify the results. Regarding weight, metformin plus lifestyle intervention provided greater weight loss than metformin monotherapy at six months, but not at 24 months. Insulin therapy was initiated at the time of confirmation of glycemic deterioration; however, data as to its effectiveness was not available in the NEJM article.

The study results were disappointing on a number of fronts: 1) metformin alone provided sustained glycemic control in only half the patients - the authors note that this result is unlikely due to poor adherence, since adherence was ≥ 80% during the first year of the study, when half the cases of treatment failure occurred; 2) lifestyle intervention provided limited improvements in weight that did not translate into prolonged glycemic control; and 3) rosiglitazone as an adjunct to metformin treatment increased weight but only slightly improved the durability of glycemic control. Use of Avandia is limited and it is uncertain whether its benefits extend to other drugs in the TZD or other classes. We believe the results underscore the need for stronger and earlier prevention for youth and that once the diagnosis of type 2 diabetes is made, more aggressive dual therapy, likely including insulin, should be initiated to avert glycemic deterioration. Overall, this result may suggest longer term that better treatments may be very useful, especially combination treatments that address multiple defects.

• BMI over time differed between the treatment groups, though BMI was not a determinant of treatment failure. The study measured changes in weight using “percentoverweight,” defined as [BMI - (BMI at the 50th percentile for age/sex)]/[BMI at the 50th percentile for age/sex]. At six months, metformin plus lifestyle intervention provided significantly greater improvements in average “percent weight” (-3.64 percentage points) than metformin monotherapy (-1.42 points) and metformin/rosiglitazone combination therapy (0.81 points). However, at 24 months, there was no significant difference in weight change between the metformin plus lifestyle intervention (-5.02 percentage points) and metformin monotherapy (-4.42 points) groups, though both treatments showed greater weight improvements compared to the metformin/rosiglitazone combination therapy (0.89 points). Furthermore, the proportion of subjects with a ≥7% reduction in “percent overweight” at six months did not differ between the metformin plus lifestyle intervention (31.2%) and metformin monotherapy (24.3%) groups, though both groups had a greater proportion of patients than the metformin/rosiglitazone combination therapy group (16.7%). In his editorial, David Allen, MD (University of Wisconsin School of Medicine, Madison, WI) noted that the subpar weight loss associated with lifestyle intervention was due less to the treatment’s lack of efficacy than to the poor uptake of lifestyle habits, likely influenced by today’s “sedentary, calorie-laden environment” (Allen, New England Journal of Medicine, 2012).

• During the first two years, the attendance rate for the lifestyle program was 75.2% and 53.6% of participants met the preplanned target of attending ≥75% of visits. However, neither changes in BMI nor the occurrence of treatment failure were influenced by whether or not participants met the attendance target. Even if lifestyle intervention did provide a therapeutic effect, the low percentage of subjects who met the attendance target underscores the difficulty of using a lifestyle program, since patients in “real-life” would likely have a harder time adhering to a similar weight-loss plan.

• Overall failure rates were significantly higher among boys (48.2%) than girls (44.3%, p=0.02). The metformin/rosiglitazone combination therapy had a significantly lower failure rate among girls (31.6%) than among boys (51.9%, p=0.03) and only among girls was the therapy significantly more effective than metformin alone or metformin plus lifestyle intervention.

• The occurrence of treatment failure was highest in non-Hispanic blacks (52.8%) compared to Hispanics (45.0%) and non-Hispanic whites (36.6%). The only treatment whose efficacy differed by racial group was metformin monotherapy, which was less effective in non- Hispanic blacks (66.2% failure rate) than in non-Hispanic whites (44.9%, p=0.01) and Hispanics (44.0%; p<0.001).

• Regarding secondary outcomes, there were no significant between-group differences in change from baseline to two years for lipid levels, blood pressure, albumin:creatinine ratio, insulin secretion, and insulin sensitivity. The change in fat mass from baseline differed significantly between the metformin plus lifestyle group and the metformin/rosiglitazone combination therapy group. Overall rates of cardiovascular (CV) risk factors at the end of the study were 33.8% for hypertension, 10.3% for high cholesterol level, 28.2% for hypertriglyceridemia, and 16.6% for microalbuminuria. These risk factors did not differ between treatment groups.

• Rates of serious adverse events differed between the metformin monotherapy group (18.1%), the metformin plus lifestyle intervention group (24.8%), and the metformin/rosiglitazone combination therapy group (14.6%), though 87% of these events were not related to the study treatment. Cases of hypoglycemia were very low for each group (1, 2, and 1 patient in each group, respectively). The trial found no effects of rosiglitazone on bone mineral content or fracture rates, though the authors acknowledge that this result should be “interpreted cautiously, given the limited sample size.”

• The ongoing Look AHEAD trial is investigating the effects of intensive lifestyle intervention (ILI) on obese adults aged 55-75 years with type 2 diabetes. The study (ClinialTrials.gov Identifier: NCT00017953) is scheduled to run from June 2011 to December 2014, and looks at a composite of cardiovascular (CV) outcomes as the primary endpoint. Patients in the ILI group experienced an averaged 8.6% weight loss after one year (compared to ~0.5% in the control group) and maintained an averaged 4.7% weight loss at the four-year mark (compared to 1.1% in the control group). While the study did not measure treatment failure, ILI provided a 0.36% decline in A1c, averaged over four years, compared to a 0.09% decline in the control group.

--by Nina Ran and Kelly Close