American Diabetes Association (ADA): Annual Scientific Sessions 2012

PHILADELPHIA – Introduction of basal insulin glargine for patients with type 2 diabetes who were uncontrolled on metformin was associated with improved glycemic control compared with the addition of sitagliptin in a 24-week, multicenter, open-label trial of more than 500 patients.

However, glargine was associated with a greater degree of weight gain and significantly greater rates of hypoglycemia, although severe hypoglycemia rates were rare, Dr. Pablo J. Aschner reported.

The study, funded by Sanofi, is the first randomized trial to compare two glucose-lowering treatment options that are often considered as second-line add-on treatment to metformin, in real-world conditions. "The results of this study support the option of introduction of basal insulin in patients with type 2 diabetes inadequately controlled by metformin, with the potential for long-term benefits arising from the achievement of optimum glycemic control early in the course of the disease," said Dr. Aschner, an endocrinologist and clinical epidemiologist, and professor of endocrinology and director of research at Javeriana University Hospital, Bogotá, Colombia.

The findings were simultaneously published online (Lancet 2012 June 9 [doi:10.1016/S0140-6736(12)60439-5]).

In the Evaluation of Insulin Glargine Versus Sitagliptin in Insulin-Naive Patients (EASIE) study, 250 patients were randomized to glargine in doses starting at 0.2 units per kg of body weight injected at dinner or bedtime with a prefilled pen, and titrated up or down based on self-monitored blood glucose levels. The 265 patients in the sitagliptin group received a fixed oral dose of 100 mg once daily taken in the morning and no changes in dose were allowed during the trial. The patients, 480 of whom completed the trial (227 glargine/253 sitagliptin), were aged 35-70 years, diagnosed with type 2 diabetes for at least 6 months, and had hemoglobin A_1c values between 7% and 11% while on metformin therapy.

HbA_1c, 8.5% for both groups at baseline, was reduced to a significantly greater extent with insulin glargine than with sitagliptin throughout the study, with mean reductions of 1.72 versus 1.13 percentage points. At 24 weeks, more patients in the glargine group had achieved A_1c levels below 7% (68% versus 42%) compared with those in the sitagliptin group. The story was similar for HbA_1c levels of less than 6.5% (40% vs. 17%). Patients in the glargine group also had significantly greater reductions in self-monitored fasting plasma glucose, with an adjusted mean difference between groups of 41.2 mg/dL, Dr. Aschner reported.

Glargine doses increased throughout the study, from 0.19 units/kg at baseline to 0.45 units/kg at 12 weeks and then leveling off to 0.5 units/kg at 24 weeks, at which point patients were taking an average of 41.4 units/day.

Body weight increased by 0.44 kg in the insulin glargine group and decreased by 1.08 kg in the sitagliptin group, a significant difference. More patients in the glargine group had one or more episodes of symptomatic hypoglycemia (46% vs. 54%). Glargine was associated with 8.45 times more hypoglycemic events per patient-year than was sitagliptin (4.21 vs. 0.5 events/patient-year). Severe hypoglycemia was 3.4-fold more common with glargine, although it was rare in both groups (1% with glargine and less than 1% with sitagliptin). The incidence of other treatment-emergent adverse events did not differ between the groups.

In all, Dr. Aschner noted, these findings suggest that glycemic control can be achieved with an average supplementary dose of 0.5 units per kg of insulin glargine and a weight gain of less than 0.5 kg. The results of this study are in general agreement with previous results for both insulin glargine and sitagliptin, he added.

Intensive lowering of blood glucose should be accompanied by structured self-monitoring of blood glucose, especially in high-risk people with long disease duration, in an attempt to avoid hypoglycemia-related adverse effects such as cardiovascular disease and related death, Dr. Aschner advised.

"In view of the possible long-term benefits of intensive lowering of blood glucose and the superior efficacy of insulin to optimize glycemic control, strong arguments could be made to use insulin early when the dose is expected to be fairly low, with a reduced risk of hypoglycemia," he concluded.

This study was funded by Sanofi. Dr. Aschner has served on advisory boards for AstraZeneca, Eli Lilly & Co, GlaxoSmithKline, Janssen, Merck, Sharpe & Dohme, Novartis, and Sanofi and on speakers’ bureaus for AstraZeneca, Eli Lilly & Co, Merck, Sharpe & Dohme, Novartis, and Sanofi.

PHILADELPHIA – Introduction of basal insulin glargine for patients with type 2 diabetes who were uncontrolled on metformin was associated with improved glycemic control compared with the addition of sitagliptin in a 24-week, multicenter, open-label trial of more than 500 patients.

However, glargine was associated with a greater degree of weight gain and significantly greater rates of hypoglycemia, although severe hypoglycemia rates were rare, Dr. Pablo J. Aschner reported.

The study, funded by Sanofi, is the first randomized trial to compare two glucose-lowering treatment options that are often considered as second-line add-on treatment to metformin, in real-world conditions. "The results of this study support the option of introduction of basal insulin in patients with type 2 diabetes inadequately controlled by metformin, with the potential for long-term benefits arising from the achievement of optimum glycemic control early in the course of the disease," said Dr. Aschner, an endocrinologist and clinical epidemiologist, and professor of endocrinology and director of research at Javeriana University Hospital, Bogotá, Colombia.

The findings were simultaneously published online (Lancet 2012 June 9 [doi:10.1016/S0140-6736(12)60439-5]).

In the Evaluation of Insulin Glargine Versus Sitagliptin in Insulin-Naive Patients (EASIE) study, 250 patients were randomized to glargine in doses starting at 0.2 units per kg of body weight injected at dinner or bedtime with a prefilled pen, and titrated up or down based on self-monitored blood glucose levels. The 265 patients in the sitagliptin group received a fixed oral dose of 100 mg once daily taken in the morning and no changes in dose were allowed during the trial. The patients, 480 of whom completed the trial (227 glargine/253 sitagliptin), were aged 35-70 years, diagnosed with type 2 diabetes for at least 6 months, and had hemoglobin A_1c values between 7% and 11% while on metformin therapy.

HbA_1c, 8.5% for both groups at baseline, was reduced to a significantly greater extent with insulin glargine than with sitagliptin throughout the study, with mean reductions of 1.72 versus 1.13 percentage points. At 24 weeks, more patients in the glargine group had achieved A_1c levels below 7% (68% versus 42%) compared with those in the sitagliptin group. The story was similar for HbA_1c levels of less than 6.5% (40% vs. 17%). Patients in the glargine group also had significantly greater reductions in self-monitored fasting plasma glucose, with an adjusted mean difference between groups of 41.2 mg/dL, Dr. Aschner reported.

Glargine doses increased throughout the study, from 0.19 units/kg at baseline to 0.45 units/kg at 12 weeks and then leveling off to 0.5 units/kg at 24 weeks, at which point patients were taking an average of 41.4 units/day.

Body weight increased by 0.44 kg in the insulin glargine group and decreased by 1.08 kg in the sitagliptin group, a significant difference. More patients in the glargine group had one or more episodes of symptomatic hypoglycemia (46% vs. 54%). Glargine was associated with 8.45 times more hypoglycemic events per patient-year than was sitagliptin (4.21 vs. 0.5 events/patient-year). Severe hypoglycemia was 3.4-fold more common with glargine, although it was rare in both groups (1% with glargine and less than 1% with sitagliptin). The incidence of other treatment-emergent adverse events did not differ between the groups.

In all, Dr. Aschner noted, these findings suggest that glycemic control can be achieved with an average supplementary dose of 0.5 units per kg of insulin glargine and a weight gain of less than 0.5 kg. The results of this study are in general agreement with previous results for both insulin glargine and sitagliptin, he added.

Intensive lowering of blood glucose should be accompanied by structured self-monitoring of blood glucose, especially in high-risk people with long disease duration, in an attempt to avoid hypoglycemia-related adverse effects such as cardiovascular disease and related death, Dr. Aschner advised.

"In view of the possible long-term benefits of intensive lowering of blood glucose and the superior efficacy of insulin to optimize glycemic control, strong arguments could be made to use insulin early when the dose is expected to be fairly low, with a reduced risk of hypoglycemia," he concluded.

This study was funded by Sanofi. Dr. Aschner has served on advisory boards for AstraZeneca, Eli Lilly & Co, GlaxoSmithKline, Janssen, Merck, Sharpe & Dohme, Novartis, and Sanofi and on speakers’ bureaus for AstraZeneca, Eli Lilly & Co, Merck, Sharpe & Dohme, Novartis, and Sanofi.

PHILADELPHIA – Introduction of basal insulin glargine for patients with type 2 diabetes who were uncontrolled on metformin was associated with improved glycemic control compared with the addition of sitagliptin in a 24-week, multicenter, open-label trial of more than 500 patients.

However, glargine was associated with a greater degree of weight gain and significantly greater rates of hypoglycemia, although severe hypoglycemia rates were rare, Dr. Pablo J. Aschner reported.

The study, funded by Sanofi, is the first randomized trial to compare two glucose-lowering treatment options that are often considered as second-line add-on treatment to metformin, in real-world conditions. "The results of this study support the option of introduction of basal insulin in patients with type 2 diabetes inadequately controlled by metformin, with the potential for long-term benefits arising from the achievement of optimum glycemic control early in the course of the disease," said Dr. Aschner, an endocrinologist and clinical epidemiologist, and professor of endocrinology and director of research at Javeriana University Hospital, Bogotá, Colombia.

The findings were simultaneously published online (Lancet 2012 June 9 [doi:10.1016/S0140-6736(12)60439-5]).

In the Evaluation of Insulin Glargine Versus Sitagliptin in Insulin-Naive Patients (EASIE) study, 250 patients were randomized to glargine in doses starting at 0.2 units per kg of body weight injected at dinner or bedtime with a prefilled pen, and titrated up or down based on self-monitored blood glucose levels. The 265 patients in the sitagliptin group received a fixed oral dose of 100 mg once daily taken in the morning and no changes in dose were allowed during the trial. The patients, 480 of whom completed the trial (227 glargine/253 sitagliptin), were aged 35-70 years, diagnosed with type 2 diabetes for at least 6 months, and had hemoglobin A_1c values between 7% and 11% while on metformin therapy.

HbA_1c, 8.5% for both groups at baseline, was reduced to a significantly greater extent with insulin glargine than with sitagliptin throughout the study, with mean reductions of 1.72 versus 1.13 percentage points. At 24 weeks, more patients in the glargine group had achieved A_1c levels below 7% (68% versus 42%) compared with those in the sitagliptin group. The story was similar for HbA_1c levels of less than 6.5% (40% vs. 17%). Patients in the glargine group also had significantly greater reductions in self-monitored fasting plasma glucose, with an adjusted mean difference between groups of 41.2 mg/dL, Dr. Aschner reported.

Glargine doses increased throughout the study, from 0.19 units/kg at baseline to 0.45 units/kg at 12 weeks and then leveling off to 0.5 units/kg at 24 weeks, at which point patients were taking an average of 41.4 units/day.

Body weight increased by 0.44 kg in the insulin glargine group and decreased by 1.08 kg in the sitagliptin group, a significant difference. More patients in the glargine group had one or more episodes of symptomatic hypoglycemia (46% vs. 54%). Glargine was associated with 8.45 times more hypoglycemic events per patient-year than was sitagliptin (4.21 vs. 0.5 events/patient-year). Severe hypoglycemia was 3.4-fold more common with glargine, although it was rare in both groups (1% with glargine and less than 1% with sitagliptin). The incidence of other treatment-emergent adverse events did not differ between the groups.

In all, Dr. Aschner noted, these findings suggest that glycemic control can be achieved with an average supplementary dose of 0.5 units per kg of insulin glargine and a weight gain of less than 0.5 kg. The results of this study are in general agreement with previous results for both insulin glargine and sitagliptin, he added.

Intensive lowering of blood glucose should be accompanied by structured self-monitoring of blood glucose, especially in high-risk people with long disease duration, in an attempt to avoid hypoglycemia-related adverse effects such as cardiovascular disease and related death, Dr. Aschner advised.

"In view of the possible long-term benefits of intensive lowering of blood glucose and the superior efficacy of insulin to optimize glycemic control, strong arguments could be made to use insulin early when the dose is expected to be fairly low, with a reduced risk of hypoglycemia," he concluded.

This study was funded by Sanofi. Dr. Aschner has served on advisory boards for AstraZeneca, Eli Lilly & Co, GlaxoSmithKline, Janssen, Merck, Sharpe & Dohme, Novartis, and Sanofi and on speakers’ bureaus for AstraZeneca, Eli Lilly & Co, Merck, Sharpe & Dohme, Novartis, and Sanofi.

PHILADELPHIA – Daily insulin glargine for patients with diabetes appears to have no effect on the occurrence of breast, prostate, colorectal, lung, pancreatic, or any other type of cancer.

Studies of three large administrative databases agree: Compared with other forms of insulin, glargine (Lantus) confers absolutely no increase in the risk of any of these cancers.

Since 2009, when four articles simultaneously raised the specter of a glargine-cancer connection, researchers, physicians, and patients have struggled with concerns that the compound could impart its own health risks. These questions must be answered as definitively as possible, Peter Boyle, Ph.D., said at the annual scientific sessions of the American Diabetes Association.

Not doing so could have devastating consequences.

"None of our results support this hypothesis that short-term use of glargine is associated with any form of cancer."

"The big worry is that by 2030, we will have more than 500 million people living with diabetes. Many of them will require insulin. Without it, they will die unnecessarily prematurely. If we don’t take a more regulated approach [into researching the cancer association], and use a lot of common sense while doing so, there could be a shortage of a drug that we absolutely need to treat this very common disease."

Dr. Boyle is the primary investigator of the Northern European Study of Insulin and Cancer, the largest-yet registry study of glargine and NPH insulin, and their relation to cancer. His study, combined with the two others presented at the same session, comprised more than 615,000 patients, with well over 1.5 million patient/years of follow-up data.

Dr. Boyle examined the glargine-cancer connection in 447,821 patients with diabetes who used daily insulin. This study had the longest follow-up of the trio – 3 years for patients using glargine and 3.5 years for those using other types of insulin.

It was conducted in five countries: Scotland, Norway, Sweden, Denmark, and Finland. This study also had the longest follow-up time – 3 years for patients using glargine and 3.5 years for those using other insulins. It compared cancer incidence in two groups: patients taking glargine or other insulin for the first time, and patients who switched from another insulin to glargine.

Over the study period, 17,800 new cases of cancer developed, said Dr. Boyle, president of the International Prevention Research Institute, Lyon, France. "In none of the individual cancers, nor in the overall analysis, did we see any evidence at all of an increased risk of cancer associated with the use of glargine, compared with other forms of insulin. Even though we saw some inconsistencies [in cancer incidence] across the different countries, none of the risk analyses came anywhere close to meeting statistical significance."

The overall risk for colorectal cancer in patients taking glargine, compared with other forms of insulin, was 0.86 – a finding consistent over all the study centers.

Findings for prostate cancer were less consistent, but this was most likely because some of the countries in the study conduct many prostate cancer screenings, "while in other countries, PSA testing is virtually verboten," Dr. Boyle said. Nevertheless, the overall risk for prostate cancer in glargine vs. other insulins was 1.11 – still not statistically significant.

Breast cancers showed a similar finding, with an overall relative risk of 1.12 for glargine, compared with other forms of insulin.

A predefined exploratory analysis examined risks associated with lung and pancreatic cancer. For glargine vs. other forms of insulin, the overall relative risk for lung cancer was 0.97; for pancreatic cancer, the relative risk was 0.99.

"There is just nothing there at all," Dr. Boyle said.

"The big worry is that by 2030, we will have more than 500 million people living with diabetes."

The second portion of the analysis examined cancer risk in light of exposure length. Breast cancer was the only type that showed any signal, and that was a confusing one, he said. While in most countries, the risk declined with time, "In the Scottish data, increasing use was associated with increasing hazard. This really is terribly inconsistent, and we need to look into it in greater detail."

However, he said, the contradictory finding perfectly illustrates the trouble with using large surveillance databases to make any kind of prediction about drug/adverse event associations.

"Quite frankly, these administrative databases were not designed for epidemiologic studies. We’ve got to do our best to come up with real findings that mean something rather than just torturing the data until something shows up. Critical bits of information are missing from these pictures. We don’t know why a patient might change from one therapy to another, for instance. There are a lot of questions like that, and these are the first things that a clinician needs to know."

Dr. Laurel Habel, a senior researcher at Kaiser Permanente, presented similar findings of no increased cancer risk from glargine exposure. Her study was based on Kaiser’s northern and southern California claims and coding databases. The study group comprised 115,000 patients with diabetes – 100,000 took NPH insulin, 27,000 took glargine, and 12,000 took both. The median follow-up time was 2.3 years for glargine and 3.6 years for NPH.

"We found no evidence of increased prostate or colorectal cancer for glargine users, whether they were new users or whether they had switched to it," she said.

In a time exposure analysis of breast cancer risk, those who had been on glargine at baseline had a "suggestion of a very modest increase in risk, of 1.6." after 24 months. Curiously, she said, there was no risk increase for patients who had been on glargine for at least 2 years after switching from another insulin.

"We can’t think of any biological reason why we would only see this in the new users, so it might be a case of confounding," she said. "None of our results support this hypothesis that short-term use of glargine is associated with any form of cancer."

For prostate and colorectal cancers, there were no significant associations in any of the analyses, or in the time-dependent comparisons

Dr. Til Stürmer of the University of North Carolina at Chapel Hill presented the smallest study, but had results entirely consistent with those of his colleagues. Again, the comparison group of 52,553 patients was drawn from a large healthcare database. Median follow-up was 1.5 years for both the glargine and other insulin groups.

In looking at the relative risks of breast, prostate, colon. and all cancer, Dr. Stürmer found no significant associations with either new glargine users or those who switched medications. Compared with other insulins, glargine had a relative risk of 1.1 for breast cancer, 1.2 for prostate cancer, 0.9 for colon cancer, and 1.1 for all forms of cancer.

"In a subgroup looking at breast cancer stratified by time [0-6 months, 6-12 months, 12-24 months and longer than 24 months] the hazard ratios jumped around the null, but we only had 14 cancers in 4,000 person/years," he added.

All of the registry studies received support from Sanofi. Other than this funding, none of the presenters had any financial declarations relevant to their studies.

PHILADELPHIA – Daily insulin glargine for patients with diabetes appears to have no effect on the occurrence of breast, prostate, colorectal, lung, pancreatic, or any other type of cancer.

Studies of three large administrative databases agree: Compared with other forms of insulin, glargine (Lantus) confers absolutely no increase in the risk of any of these cancers.

Since 2009, when four articles simultaneously raised the specter of a glargine-cancer connection, researchers, physicians, and patients have struggled with concerns that the compound could impart its own health risks. These questions must be answered as definitively as possible, Peter Boyle, Ph.D., said at the annual scientific sessions of the American Diabetes Association.

Not doing so could have devastating consequences.

"None of our results support this hypothesis that short-term use of glargine is associated with any form of cancer."

"The big worry is that by 2030, we will have more than 500 million people living with diabetes. Many of them will require insulin. Without it, they will die unnecessarily prematurely. If we don’t take a more regulated approach [into researching the cancer association], and use a lot of common sense while doing so, there could be a shortage of a drug that we absolutely need to treat this very common disease."

Dr. Boyle is the primary investigator of the Northern European Study of Insulin and Cancer, the largest-yet registry study of glargine and NPH insulin, and their relation to cancer. His study, combined with the two others presented at the same session, comprised more than 615,000 patients, with well over 1.5 million patient/years of follow-up data.

Dr. Boyle examined the glargine-cancer connection in 447,821 patients with diabetes who used daily insulin. This study had the longest follow-up of the trio – 3 years for patients using glargine and 3.5 years for those using other types of insulin.

It was conducted in five countries: Scotland, Norway, Sweden, Denmark, and Finland. This study also had the longest follow-up time – 3 years for patients using glargine and 3.5 years for those using other insulins. It compared cancer incidence in two groups: patients taking glargine or other insulin for the first time, and patients who switched from another insulin to glargine.

Over the study period, 17,800 new cases of cancer developed, said Dr. Boyle, president of the International Prevention Research Institute, Lyon, France. "In none of the individual cancers, nor in the overall analysis, did we see any evidence at all of an increased risk of cancer associated with the use of glargine, compared with other forms of insulin. Even though we saw some inconsistencies [in cancer incidence] across the different countries, none of the risk analyses came anywhere close to meeting statistical significance."

The overall risk for colorectal cancer in patients taking glargine, compared with other forms of insulin, was 0.86 – a finding consistent over all the study centers.

Findings for prostate cancer were less consistent, but this was most likely because some of the countries in the study conduct many prostate cancer screenings, "while in other countries, PSA testing is virtually verboten," Dr. Boyle said. Nevertheless, the overall risk for prostate cancer in glargine vs. other insulins was 1.11 – still not statistically significant.

Breast cancers showed a similar finding, with an overall relative risk of 1.12 for glargine, compared with other forms of insulin.

A predefined exploratory analysis examined risks associated with lung and pancreatic cancer. For glargine vs. other forms of insulin, the overall relative risk for lung cancer was 0.97; for pancreatic cancer, the relative risk was 0.99.

"There is just nothing there at all," Dr. Boyle said.

"The big worry is that by 2030, we will have more than 500 million people living with diabetes."

The second portion of the analysis examined cancer risk in light of exposure length. Breast cancer was the only type that showed any signal, and that was a confusing one, he said. While in most countries, the risk declined with time, "In the Scottish data, increasing use was associated with increasing hazard. This really is terribly inconsistent, and we need to look into it in greater detail."

However, he said, the contradictory finding perfectly illustrates the trouble with using large surveillance databases to make any kind of prediction about drug/adverse event associations.

"Quite frankly, these administrative databases were not designed for epidemiologic studies. We’ve got to do our best to come up with real findings that mean something rather than just torturing the data until something shows up. Critical bits of information are missing from these pictures. We don’t know why a patient might change from one therapy to another, for instance. There are a lot of questions like that, and these are the first things that a clinician needs to know."

Dr. Laurel Habel, a senior researcher at Kaiser Permanente, presented similar findings of no increased cancer risk from glargine exposure. Her study was based on Kaiser’s northern and southern California claims and coding databases. The study group comprised 115,000 patients with diabetes – 100,000 took NPH insulin, 27,000 took glargine, and 12,000 took both. The median follow-up time was 2.3 years for glargine and 3.6 years for NPH.

"We found no evidence of increased prostate or colorectal cancer for glargine users, whether they were new users or whether they had switched to it," she said.

In a time exposure analysis of breast cancer risk, those who had been on glargine at baseline had a "suggestion of a very modest increase in risk, of 1.6." after 24 months. Curiously, she said, there was no risk increase for patients who had been on glargine for at least 2 years after switching from another insulin.

"We can’t think of any biological reason why we would only see this in the new users, so it might be a case of confounding," she said. "None of our results support this hypothesis that short-term use of glargine is associated with any form of cancer."

For prostate and colorectal cancers, there were no significant associations in any of the analyses, or in the time-dependent comparisons

Dr. Til Stürmer of the University of North Carolina at Chapel Hill presented the smallest study, but had results entirely consistent with those of his colleagues. Again, the comparison group of 52,553 patients was drawn from a large healthcare database. Median follow-up was 1.5 years for both the glargine and other insulin groups.

In looking at the relative risks of breast, prostate, colon. and all cancer, Dr. Stürmer found no significant associations with either new glargine users or those who switched medications. Compared with other insulins, glargine had a relative risk of 1.1 for breast cancer, 1.2 for prostate cancer, 0.9 for colon cancer, and 1.1 for all forms of cancer.

"In a subgroup looking at breast cancer stratified by time [0-6 months, 6-12 months, 12-24 months and longer than 24 months] the hazard ratios jumped around the null, but we only had 14 cancers in 4,000 person/years," he added.

All of the registry studies received support from Sanofi. Other than this funding, none of the presenters had any financial declarations relevant to their studies.

PHILADELPHIA – Daily insulin glargine for patients with diabetes appears to have no effect on the occurrence of breast, prostate, colorectal, lung, pancreatic, or any other type of cancer.

Studies of three large administrative databases agree: Compared with other forms of insulin, glargine (Lantus) confers absolutely no increase in the risk of any of these cancers.

Since 2009, when four articles simultaneously raised the specter of a glargine-cancer connection, researchers, physicians, and patients have struggled with concerns that the compound could impart its own health risks. These questions must be answered as definitively as possible, Peter Boyle, Ph.D., said at the annual scientific sessions of the American Diabetes Association.

Not doing so could have devastating consequences.

"None of our results support this hypothesis that short-term use of glargine is associated with any form of cancer."

"The big worry is that by 2030, we will have more than 500 million people living with diabetes. Many of them will require insulin. Without it, they will die unnecessarily prematurely. If we don’t take a more regulated approach [into researching the cancer association], and use a lot of common sense while doing so, there could be a shortage of a drug that we absolutely need to treat this very common disease."

Dr. Boyle is the primary investigator of the Northern European Study of Insulin and Cancer, the largest-yet registry study of glargine and NPH insulin, and their relation to cancer. His study, combined with the two others presented at the same session, comprised more than 615,000 patients, with well over 1.5 million patient/years of follow-up data.

Dr. Boyle examined the glargine-cancer connection in 447,821 patients with diabetes who used daily insulin. This study had the longest follow-up of the trio – 3 years for patients using glargine and 3.5 years for those using other types of insulin.

It was conducted in five countries: Scotland, Norway, Sweden, Denmark, and Finland. This study also had the longest follow-up time – 3 years for patients using glargine and 3.5 years for those using other insulins. It compared cancer incidence in two groups: patients taking glargine or other insulin for the first time, and patients who switched from another insulin to glargine.

Over the study period, 17,800 new cases of cancer developed, said Dr. Boyle, president of the International Prevention Research Institute, Lyon, France. "In none of the individual cancers, nor in the overall analysis, did we see any evidence at all of an increased risk of cancer associated with the use of glargine, compared with other forms of insulin. Even though we saw some inconsistencies [in cancer incidence] across the different countries, none of the risk analyses came anywhere close to meeting statistical significance."

The overall risk for colorectal cancer in patients taking glargine, compared with other forms of insulin, was 0.86 – a finding consistent over all the study centers.

Findings for prostate cancer were less consistent, but this was most likely because some of the countries in the study conduct many prostate cancer screenings, "while in other countries, PSA testing is virtually verboten," Dr. Boyle said. Nevertheless, the overall risk for prostate cancer in glargine vs. other insulins was 1.11 – still not statistically significant.

Breast cancers showed a similar finding, with an overall relative risk of 1.12 for glargine, compared with other forms of insulin.

A predefined exploratory analysis examined risks associated with lung and pancreatic cancer. For glargine vs. other forms of insulin, the overall relative risk for lung cancer was 0.97; for pancreatic cancer, the relative risk was 0.99.

"There is just nothing there at all," Dr. Boyle said.

"The big worry is that by 2030, we will have more than 500 million people living with diabetes."

The second portion of the analysis examined cancer risk in light of exposure length. Breast cancer was the only type that showed any signal, and that was a confusing one, he said. While in most countries, the risk declined with time, "In the Scottish data, increasing use was associated with increasing hazard. This really is terribly inconsistent, and we need to look into it in greater detail."

However, he said, the contradictory finding perfectly illustrates the trouble with using large surveillance databases to make any kind of prediction about drug/adverse event associations.

"Quite frankly, these administrative databases were not designed for epidemiologic studies. We’ve got to do our best to come up with real findings that mean something rather than just torturing the data until something shows up. Critical bits of information are missing from these pictures. We don’t know why a patient might change from one therapy to another, for instance. There are a lot of questions like that, and these are the first things that a clinician needs to know."

Dr. Laurel Habel, a senior researcher at Kaiser Permanente, presented similar findings of no increased cancer risk from glargine exposure. Her study was based on Kaiser’s northern and southern California claims and coding databases. The study group comprised 115,000 patients with diabetes – 100,000 took NPH insulin, 27,000 took glargine, and 12,000 took both. The median follow-up time was 2.3 years for glargine and 3.6 years for NPH.

"We found no evidence of increased prostate or colorectal cancer for glargine users, whether they were new users or whether they had switched to it," she said.

In a time exposure analysis of breast cancer risk, those who had been on glargine at baseline had a "suggestion of a very modest increase in risk, of 1.6." after 24 months. Curiously, she said, there was no risk increase for patients who had been on glargine for at least 2 years after switching from another insulin.

"We can’t think of any biological reason why we would only see this in the new users, so it might be a case of confounding," she said. "None of our results support this hypothesis that short-term use of glargine is associated with any form of cancer."

For prostate and colorectal cancers, there were no significant associations in any of the analyses, or in the time-dependent comparisons

Dr. Til Stürmer of the University of North Carolina at Chapel Hill presented the smallest study, but had results entirely consistent with those of his colleagues. Again, the comparison group of 52,553 patients was drawn from a large healthcare database. Median follow-up was 1.5 years for both the glargine and other insulin groups.

In looking at the relative risks of breast, prostate, colon. and all cancer, Dr. Stürmer found no significant associations with either new glargine users or those who switched medications. Compared with other insulins, glargine had a relative risk of 1.1 for breast cancer, 1.2 for prostate cancer, 0.9 for colon cancer, and 1.1 for all forms of cancer.

"In a subgroup looking at breast cancer stratified by time [0-6 months, 6-12 months, 12-24 months and longer than 24 months] the hazard ratios jumped around the null, but we only had 14 cancers in 4,000 person/years," he added.

All of the registry studies received support from Sanofi. Other than this funding, none of the presenters had any financial declarations relevant to their studies.

PHILADELPHIA – The prevalence of type 2 diabetes, once considered an almost exclusively adult-onset disease, increased by 21% from 2001 to 2009 among Americans younger than 20 years.

The prevalence of type 1 diabetes increased by 23% over the same time period, Dr. Dana Dabelea reported at the annual meeting of the American Diabetes Association.

"Every year, we see about 15,000 new cases of type 1 and 3,700 new cases of type 2 diabetes, and the numbers of children living with these diseases are increasing," she said at a press briefing.

Although type 1 diabetes remains "an extremely rare" and very dangerous condition, evidence continues to emerge that type 2 diabetes is anything but benign, said Dr. Dabelea, a professor of epidemiology at the University of Colorado at Denver, Aurora.

"It’s believed that this increase we’re seeing is making type 2 more and more a pediatric condition," she said. "In youth, this disease is associated with early signs and symptoms of chronic complications that may increase lifelong comorbidities, reduce quality life and life expectancy, and increase health care costs."

Dr. Dabelea and her coinvestigators discussed new data from the SEARCH for Diabetes in Youth study. The registry study was launched in 2000, and is expected to continue at least through 2015. It’s being conducted in five centers in the United States.

The primary aim is to track the prevalence of new diabetes cases, providing the first national picture of the rising epidemic. Subanalyses look at the diseases’ complications and comorbidities, associations between diabetes and lifestyle, and the impact of nutrition on the diseases.

The new data suggest that from 2001 to 2009, almost 189,000 U.S. residents aged younger than 20 years had diabetes (about 168,000 with type 1 and about 19,000 with type 2).

The prevalence of both types of diabetes has increased in both boys and girls and across all races and ethnicities. The prevalence of type 1 in the study population rose from 1.55 per 1,000 in 2001 to 2.03 per 1,000 in 2009, a statistically significant difference. The prevalence of type 2 during that period rose significantly from 2.9 per 10,000 to 3.6 per 10,000.

"Although the proportion was highest in Native Americans and blacks, the increase was also observed in whites and Hispanics," Dr. Dabelea said. "This may suggest that pediatric type 2 diabetes is plateauing in the traditionally high-risk groups, but still increasing in other groups."

SEARCH investigators presented a number of substudy findings at the meeting, including the following:

• Young people with either form of the disease may show signs of impending kidney damage. In a group of 5,000 patients with a 4-year diabetes history, up to 17% had albuminuria. The numbers were highest in children with insulin resistance but no diabetes autoantibodies (17%), and lowest (8%) in those with diabetes autoantibodies who were insulin sensitive.

• Television time seems directly tied to blood glucose and lipid levels in both types of diabetes. In a cohort of 1,400 SEARCH patients aged younger than 10 years, HbA_1c increased as TV time increased. Triglyceride levels were significantly higher in children who watched TV for 3 or more hours each day than in those who watched less.

• Peripheral neuropathy strikes children with diabetes at about the same rate as it does adults. "Our findings suggest that children, adolescents, and young adults with diabetes are not only at risk for diabetic peripheral neuropathy, but that many already show measurable signs of it," the authors noted.

• Among a subset of 222 SEARCH subjects, there were early indications that cardiovascular parasympathetic nerves were impaired, perhaps predisposing children with diabetes to future cardiovascular disease.

SEARCH is being sponsored by the Centers for Disease Control and Prevention and the National Institutes of Health. Dr. Dabelea did not have any financial disclosures.

PHILADELPHIA – The prevalence of type 2 diabetes, once considered an almost exclusively adult-onset disease, increased by 21% from 2001 to 2009 among Americans younger than 20 years.

The prevalence of type 1 diabetes increased by 23% over the same time period, Dr. Dana Dabelea reported at the annual meeting of the American Diabetes Association.

"Every year, we see about 15,000 new cases of type 1 and 3,700 new cases of type 2 diabetes, and the numbers of children living with these diseases are increasing," she said at a press briefing.

Although type 1 diabetes remains "an extremely rare" and very dangerous condition, evidence continues to emerge that type 2 diabetes is anything but benign, said Dr. Dabelea, a professor of epidemiology at the University of Colorado at Denver, Aurora.

"It’s believed that this increase we’re seeing is making type 2 more and more a pediatric condition," she said. "In youth, this disease is associated with early signs and symptoms of chronic complications that may increase lifelong comorbidities, reduce quality life and life expectancy, and increase health care costs."

Dr. Dabelea and her coinvestigators discussed new data from the SEARCH for Diabetes in Youth study. The registry study was launched in 2000, and is expected to continue at least through 2015. It’s being conducted in five centers in the United States.

The primary aim is to track the prevalence of new diabetes cases, providing the first national picture of the rising epidemic. Subanalyses look at the diseases’ complications and comorbidities, associations between diabetes and lifestyle, and the impact of nutrition on the diseases.

The new data suggest that from 2001 to 2009, almost 189,000 U.S. residents aged younger than 20 years had diabetes (about 168,000 with type 1 and about 19,000 with type 2).

The prevalence of both types of diabetes has increased in both boys and girls and across all races and ethnicities. The prevalence of type 1 in the study population rose from 1.55 per 1,000 in 2001 to 2.03 per 1,000 in 2009, a statistically significant difference. The prevalence of type 2 during that period rose significantly from 2.9 per 10,000 to 3.6 per 10,000.

"Although the proportion was highest in Native Americans and blacks, the increase was also observed in whites and Hispanics," Dr. Dabelea said. "This may suggest that pediatric type 2 diabetes is plateauing in the traditionally high-risk groups, but still increasing in other groups."

SEARCH investigators presented a number of substudy findings at the meeting, including the following:

• Young people with either form of the disease may show signs of impending kidney damage. In a group of 5,000 patients with a 4-year diabetes history, up to 17% had albuminuria. The numbers were highest in children with insulin resistance but no diabetes autoantibodies (17%), and lowest (8%) in those with diabetes autoantibodies who were insulin sensitive.

• Television time seems directly tied to blood glucose and lipid levels in both types of diabetes. In a cohort of 1,400 SEARCH patients aged younger than 10 years, HbA_1c increased as TV time increased. Triglyceride levels were significantly higher in children who watched TV for 3 or more hours each day than in those who watched less.

• Peripheral neuropathy strikes children with diabetes at about the same rate as it does adults. "Our findings suggest that children, adolescents, and young adults with diabetes are not only at risk for diabetic peripheral neuropathy, but that many already show measurable signs of it," the authors noted.

• Among a subset of 222 SEARCH subjects, there were early indications that cardiovascular parasympathetic nerves were impaired, perhaps predisposing children with diabetes to future cardiovascular disease.

SEARCH is being sponsored by the Centers for Disease Control and Prevention and the National Institutes of Health. Dr. Dabelea did not have any financial disclosures.

PHILADELPHIA – The prevalence of type 2 diabetes, once considered an almost exclusively adult-onset disease, increased by 21% from 2001 to 2009 among Americans younger than 20 years.

The prevalence of type 1 diabetes increased by 23% over the same time period, Dr. Dana Dabelea reported at the annual meeting of the American Diabetes Association.

"Every year, we see about 15,000 new cases of type 1 and 3,700 new cases of type 2 diabetes, and the numbers of children living with these diseases are increasing," she said at a press briefing.

Although type 1 diabetes remains "an extremely rare" and very dangerous condition, evidence continues to emerge that type 2 diabetes is anything but benign, said Dr. Dabelea, a professor of epidemiology at the University of Colorado at Denver, Aurora.

"It’s believed that this increase we’re seeing is making type 2 more and more a pediatric condition," she said. "In youth, this disease is associated with early signs and symptoms of chronic complications that may increase lifelong comorbidities, reduce quality life and life expectancy, and increase health care costs."

Dr. Dabelea and her coinvestigators discussed new data from the SEARCH for Diabetes in Youth study. The registry study was launched in 2000, and is expected to continue at least through 2015. It’s being conducted in five centers in the United States.

The primary aim is to track the prevalence of new diabetes cases, providing the first national picture of the rising epidemic. Subanalyses look at the diseases’ complications and comorbidities, associations between diabetes and lifestyle, and the impact of nutrition on the diseases.

The new data suggest that from 2001 to 2009, almost 189,000 U.S. residents aged younger than 20 years had diabetes (about 168,000 with type 1 and about 19,000 with type 2).

The prevalence of both types of diabetes has increased in both boys and girls and across all races and ethnicities. The prevalence of type 1 in the study population rose from 1.55 per 1,000 in 2001 to 2.03 per 1,000 in 2009, a statistically significant difference. The prevalence of type 2 during that period rose significantly from 2.9 per 10,000 to 3.6 per 10,000.

"Although the proportion was highest in Native Americans and blacks, the increase was also observed in whites and Hispanics," Dr. Dabelea said. "This may suggest that pediatric type 2 diabetes is plateauing in the traditionally high-risk groups, but still increasing in other groups."

SEARCH investigators presented a number of substudy findings at the meeting, including the following:

• Young people with either form of the disease may show signs of impending kidney damage. In a group of 5,000 patients with a 4-year diabetes history, up to 17% had albuminuria. The numbers were highest in children with insulin resistance but no diabetes autoantibodies (17%), and lowest (8%) in those with diabetes autoantibodies who were insulin sensitive.

• Television time seems directly tied to blood glucose and lipid levels in both types of diabetes. In a cohort of 1,400 SEARCH patients aged younger than 10 years, HbA_1c increased as TV time increased. Triglyceride levels were significantly higher in children who watched TV for 3 or more hours each day than in those who watched less.

• Peripheral neuropathy strikes children with diabetes at about the same rate as it does adults. "Our findings suggest that children, adolescents, and young adults with diabetes are not only at risk for diabetic peripheral neuropathy, but that many already show measurable signs of it," the authors noted.

• Among a subset of 222 SEARCH subjects, there were early indications that cardiovascular parasympathetic nerves were impaired, perhaps predisposing children with diabetes to future cardiovascular disease.

SEARCH is being sponsored by the Centers for Disease Control and Prevention and the National Institutes of Health. Dr. Dabelea did not have any financial disclosures.

Aggressive blood pressure lowering, and the degree of a patient’s adiposity, appear unrelated to cardiovascular disease outcomes in individuals with diabetes, according to a new analysis of data from the large, randomized controlled trial, Action to Control Cardiovascular Risk in Diabetes, presented at the annual meeting of the American Diabetes Association in Philadelphia.

The subanalysis follows on findings from the original blood-pressure study (n = 4,733) that showed no difference in cardiovascular disease (CVD) risk between people with diabetes mellitus whose blood pressure was treated to less than 120 mm Hg, compared with those with the standard treatment goal of less than 140 mm Hg. The primary measure of CVD outcomes looked at a composite of nonfatal MI, nonfatal stroke, or CVD death (N. Engl. J. Med. 2010;362:1575-85).

That original study, however, did not control for waist-to-height ratios. Presumably, a waist-to-height ratio of less than 50% would be associated with less CVD risk, compared with a ratio higher than 50%, postulated Dr. Joshua I. Barzilay of Emory University, Atlanta, and his colleagues.

His team analyzed data from the 4,687 people in the ACCORD blood-pressure cohort for whom waist-to-height data were available. All subjects had diabetes and hypertension; 47.7% were women; the subjects’ mean age was 62.2 years; and the mean follow-up was 4.7 years. Intensive lowering of blood pressure did not improve combined CVD risk, compared with standard treatment, they found (Diabetes Care 2012;35:1401-5).

Although the investigators found that controlling for the waist-to-height quartile did not affect risk for the combined outcome measure as they had hypothesized it would, one secondary end point in the study (CVD death) was found to be significantly related to the waist-to-height quartile (hazard ratio, 2.32; 95% confidence interval, 1.40–3.83; P = .0009 comparing the heaviest to lightest quartiles). Neither of the two other secondary outcome measures – fatal or nonfatal stroke and nonfatal MI – was affected.

The association between central obesity and increased risk of CVD death (but not MI or stroke) was "surprising, because morbidity outcomes share similar pathomechanisms with mortality outcomes." However, they wrote, the finding was in keeping with some other published studies. Increased "levels of inflammation factors are more strongly associated with fatal than with nonfatal CVD events. Greater degrees of central obesity are related to increased inflammation levels," they wrote.

The study is limited by its post hoc design, modest number of outcomes measured, and short follow-up. "The power to detect significant differences by the degree of central obesity was limited. With longer follow-up and more events, statistical significance could possibly be achieved for several outcomes."

The subanalysis was funded by the National Heart, Lung, and Blood Institute, and other federal agencies. Aside from study medicines and equipment, which were donated by various manufacturers, no conflicts of interest were reported.

Aggressive blood pressure lowering, and the degree of a patient’s adiposity, appear unrelated to cardiovascular disease outcomes in individuals with diabetes, according to a new analysis of data from the large, randomized controlled trial, Action to Control Cardiovascular Risk in Diabetes, presented at the annual meeting of the American Diabetes Association in Philadelphia.

The subanalysis follows on findings from the original blood-pressure study (n = 4,733) that showed no difference in cardiovascular disease (CVD) risk between people with diabetes mellitus whose blood pressure was treated to less than 120 mm Hg, compared with those with the standard treatment goal of less than 140 mm Hg. The primary measure of CVD outcomes looked at a composite of nonfatal MI, nonfatal stroke, or CVD death (N. Engl. J. Med. 2010;362:1575-85).

That original study, however, did not control for waist-to-height ratios. Presumably, a waist-to-height ratio of less than 50% would be associated with less CVD risk, compared with a ratio higher than 50%, postulated Dr. Joshua I. Barzilay of Emory University, Atlanta, and his colleagues.

His team analyzed data from the 4,687 people in the ACCORD blood-pressure cohort for whom waist-to-height data were available. All subjects had diabetes and hypertension; 47.7% were women; the subjects’ mean age was 62.2 years; and the mean follow-up was 4.7 years. Intensive lowering of blood pressure did not improve combined CVD risk, compared with standard treatment, they found (Diabetes Care 2012;35:1401-5).

Although the investigators found that controlling for the waist-to-height quartile did not affect risk for the combined outcome measure as they had hypothesized it would, one secondary end point in the study (CVD death) was found to be significantly related to the waist-to-height quartile (hazard ratio, 2.32; 95% confidence interval, 1.40–3.83; P = .0009 comparing the heaviest to lightest quartiles). Neither of the two other secondary outcome measures – fatal or nonfatal stroke and nonfatal MI – was affected.

The association between central obesity and increased risk of CVD death (but not MI or stroke) was "surprising, because morbidity outcomes share similar pathomechanisms with mortality outcomes." However, they wrote, the finding was in keeping with some other published studies. Increased "levels of inflammation factors are more strongly associated with fatal than with nonfatal CVD events. Greater degrees of central obesity are related to increased inflammation levels," they wrote.

The study is limited by its post hoc design, modest number of outcomes measured, and short follow-up. "The power to detect significant differences by the degree of central obesity was limited. With longer follow-up and more events, statistical significance could possibly be achieved for several outcomes."

The subanalysis was funded by the National Heart, Lung, and Blood Institute, and other federal agencies. Aside from study medicines and equipment, which were donated by various manufacturers, no conflicts of interest were reported.

Aggressive blood pressure lowering, and the degree of a patient’s adiposity, appear unrelated to cardiovascular disease outcomes in individuals with diabetes, according to a new analysis of data from the large, randomized controlled trial, Action to Control Cardiovascular Risk in Diabetes, presented at the annual meeting of the American Diabetes Association in Philadelphia.

The subanalysis follows on findings from the original blood-pressure study (n = 4,733) that showed no difference in cardiovascular disease (CVD) risk between people with diabetes mellitus whose blood pressure was treated to less than 120 mm Hg, compared with those with the standard treatment goal of less than 140 mm Hg. The primary measure of CVD outcomes looked at a composite of nonfatal MI, nonfatal stroke, or CVD death (N. Engl. J. Med. 2010;362:1575-85).

That original study, however, did not control for waist-to-height ratios. Presumably, a waist-to-height ratio of less than 50% would be associated with less CVD risk, compared with a ratio higher than 50%, postulated Dr. Joshua I. Barzilay of Emory University, Atlanta, and his colleagues.

His team analyzed data from the 4,687 people in the ACCORD blood-pressure cohort for whom waist-to-height data were available. All subjects had diabetes and hypertension; 47.7% were women; the subjects’ mean age was 62.2 years; and the mean follow-up was 4.7 years. Intensive lowering of blood pressure did not improve combined CVD risk, compared with standard treatment, they found (Diabetes Care 2012;35:1401-5).

Although the investigators found that controlling for the waist-to-height quartile did not affect risk for the combined outcome measure as they had hypothesized it would, one secondary end point in the study (CVD death) was found to be significantly related to the waist-to-height quartile (hazard ratio, 2.32; 95% confidence interval, 1.40–3.83; P = .0009 comparing the heaviest to lightest quartiles). Neither of the two other secondary outcome measures – fatal or nonfatal stroke and nonfatal MI – was affected.

The association between central obesity and increased risk of CVD death (but not MI or stroke) was "surprising, because morbidity outcomes share similar pathomechanisms with mortality outcomes." However, they wrote, the finding was in keeping with some other published studies. Increased "levels of inflammation factors are more strongly associated with fatal than with nonfatal CVD events. Greater degrees of central obesity are related to increased inflammation levels," they wrote.

The study is limited by its post hoc design, modest number of outcomes measured, and short follow-up. "The power to detect significant differences by the degree of central obesity was limited. With longer follow-up and more events, statistical significance could possibly be achieved for several outcomes."

The subanalysis was funded by the National Heart, Lung, and Blood Institute, and other federal agencies. Aside from study medicines and equipment, which were donated by various manufacturers, no conflicts of interest were reported.

PHILADELPHIA – Getting a firm, early grip on blood glucose confers the best chance for durable glycemic control in children with type 2 diabetes.

Any rapid increase in hemoglobin A_1c after initial stabilization probably signals the need to quickly gear up treatment, Dr. Sonia Caprio said at the annual scientific sessions of the American Diabetes Association.

"The best single predictor of treatment failure – regardless of therapy – is baseline [hemoglobin A_1c], even if it’s in the nondiabetic range" after stabilization, said Dr. Caprio, a professor of pediatric endocrinology at Yale University in New Haven, Conn. "A rapidly rising HbA_1c, even if it’s still in the normal range, is associated with failure, and suggests the need for more intense treatment."

These new findings from the TODAY (Treatment Options for Type 2 Diabetes in Adolescents and Youth) study mean that physicians may need new treatment markers for diabetes regimens in youth, she said. "We need to identify specific thresholds for intensifying therapy, perhaps at diagnosis, perhaps after a certain period of treatment, or when the patient reaches a certain HbA_1c threshold."

Dr. Caprio joined other TODAY investigators in a lively discussion of the study. TODAY’s primary end points were released in April; the ADA presentation not only broke down predictors of treatment failure, but pointed up a disturbing scenario of the medical comorbidities that already haunt children with type 2 diabetes.

Hypertension, enlarged hearts, retinopathy, dyslipidemia, and incipient kidney dysfunction were all common among the 699 children who participated in the study, said Dr. Neil White of Washington University, St. Louis. The problems occurred early and frequently.

"Within 7-8 months after diagnosis, 14% already had nonproliferative retinopathy. That number is similar to the amount found in the Diabetes Prevention Program study, for a population that was an average of 55 years old. These subjects are getting the signs of disease during childhood that older adults have."

TODAY examined three different treatment regimens among 699 children with newly diagnosed type 2 diabetes: metformin alone, metformin plus lifestyle modifications, and metformin plus rosiglitazone (N. Engl. J. Med. 2012 April 29 [doi:10.1056/NEJMoa1109333]).

The 60-month trial started all patients aged 10-17 years on metformin 2,000 mg/day. Randomization occurred after HbA_1c stabilized at 8% or lower with metformin therapy. The primary end point was time to the failure of glycemic control (defined as an HbA_1c of at least 8% for 6 months, or sustained metabolic decompensation that required insulin treatment).

Overall, nearly half of the cohort (46%) failed to maintain glycemic control; the median time to failure was 11 months. However, compared with the combination therapy group, significantly more of those taking metformin alone failed glycemic control (52% vs. 39%). The failure rate in the lifestyle intervention group was 47% – not significantly different from that of the metformin mono- or combination therapy groups.

The new analyses examined predictors of response, and looked at medical comorbidities among the patients. A seemingly innocuous increase in HbA_1c during the initial 6-month stabilization phase was the strongest predictor of whether a child would achieve a durable response to the long-term therapy, said Dr. Kenneth Copeland, a TODAY investigator and the Milburn Endowed Chair of Pediatrics at the University of Oklahoma Health Sciences Center, Oklahoma City.

In a univariate analysis, a number of baseline factors were associated with failure, including race (with blacks doing significantly worse than whites); the presence of depression; lower household income; lower insulinogenic index, and lower HbA_1c.

The baseline difference in HbA_1c was statistically significant, but still apparently small, he said. After initial stabilization, those who maintained and failed control still had blood glucose levels in the nondiabetic range (5.7% vs. 6.4%)

But in a multivariate model, only baseline HbA_1c strongly predicted response to long-term therapy. Those who maintained glucose control had very moderate HbA_1c increases, from a median of 6.4% to about 7% by the end of the study. Those who failed therapy had a much steeper increase, rising from 5.7% at baseline to about 7.3% at the end.

"Baseline HbA_1c appears to be the best predictor of failure, regardless of whether it is still in the nondiabetic range. A rapidly rising level – even in the normal range – is associated with incipient failure and may suggest the need to intensify treatment early."

The increases in blood sugar that arose during stabilization and treatment probably signaled beta-cell decline in an increasingly stressed pancreas, Dr. Caprio said. The analysis used two algorithms to assess beta-cell function: the insulinogenic index (the difference in blood glucose before and after insulin administration) and the oral disposition index (a measure of beta-cell function relative to insulin). The baseline for this analysis was the point of randomization, when glucose was stabilized at 8% or lower after 6 months of treatment.

"At 6 months, we saw a dramatic difference between the rosiglitazone/metformin arm and the other arms," with a 20% increase in insulin sensitivity compared with no change in the other groups. Thereafter, however, insulin sensitivity started to decrease in the rosiglitazone group, until it plateaued.

By the end of the study, beta-cell function was also similar between the groups, but those in the rosiglitazone group showed a slower decline during the treatment period. This indicated that rosiglitazone was able to stabilize function for at least a period of time. "In contrast, we saw a dramatic fall-off in the beta-cell function in the other two groups, although by 24 months there were no differences among any of them."

"The favorable changes relative to insulin sensitivity were responsible for the observed reduced failure rate in the rosiglitazone group; telling is that this profound degree of insulin resistance in obese youth with type 2 diabetes is unresponsive to the effect of metformin, alone or with lifestyle therapy," Dr. Caprio said.

Because metformin failed those children, the better alternative appears to be more aggressive treatment early in the disease. "The introduction of such aggressive therapy early in the course of the disease appears to slow the decline of beta-cell adaptation and improve the chance or durable control."

Children in TODAY were plagued with much the same medical issues as are adults with type 2 diabetes, said Dr. Neil White, chairman of the TODAY data safety and monitoring committee.

The subjects had a very short duration of diabetes, having been diagnosed 5-7 months before the metformin stabilization period. At that point, 34% already had hypertension. Abnormal LDL cholesterol levels were present in 10%, and high triglycerides in 54%. Some 20% had microalbuminuria, suggesting possible progression to kidney disease.

All of the children, regardless of their treatment group or outcome, showed somewhat abnormal left ventricular and right auricular measurements, noted Dr. White, who is also director of the pediatric clinical research unit at Washington University.

Echocardiography showed that the children had a left auricular diameter of up to 2.4 cm/m height, greater than the population median. All also had a left ventricular mass greater than the population median of 30 g/m², with some reaching almost 45 g/m².

"These youngsters had big hearts that are going to predispose them to heart disease in the future," Dr. White said.

At the end of the study, most children (523) underwent fundus photography to evaluate retinopathy. About 14% of them had some degree of nonproliferative retinopathy, putting the young group’s prevalence squarely in line with prevalence in adults with type 2 diabetes.

"This profound insulin resistance was a culprit ... against which metformin had no effect at all," Dr. Caprio said. "Even with the two best drugs we had at the time of this study – metformin and rosiglitazone – we were not able to improve outcomes in 50% of these kids."

Dr. Copeland agreed.

"The finding that 50% of our children failed therapy indicates that this is an incredibly tough disease to treat by any modality we have. Our target as a society needs to be ‘How do we prevent diabetes in children?’ because once they get it, it’s going to be really tough to intervene effectively."

TODAY was funded by the National Institute of Diabetes and Digestive and Kidney Diseases. All of the presenters disclosed financial relationships with Daiichi Sankyo, but said these were not relevant to the TODAY study.

PHILADELPHIA – Getting a firm, early grip on blood glucose confers the best chance for durable glycemic control in children with type 2 diabetes.

Any rapid increase in hemoglobin A_1c after initial stabilization probably signals the need to quickly gear up treatment, Dr. Sonia Caprio said at the annual scientific sessions of the American Diabetes Association.

"The best single predictor of treatment failure – regardless of therapy – is baseline [hemoglobin A_1c], even if it’s in the nondiabetic range" after stabilization, said Dr. Caprio, a professor of pediatric endocrinology at Yale University in New Haven, Conn. "A rapidly rising HbA_1c, even if it’s still in the normal range, is associated with failure, and suggests the need for more intense treatment."

These new findings from the TODAY (Treatment Options for Type 2 Diabetes in Adolescents and Youth) study mean that physicians may need new treatment markers for diabetes regimens in youth, she said. "We need to identify specific thresholds for intensifying therapy, perhaps at diagnosis, perhaps after a certain period of treatment, or when the patient reaches a certain HbA_1c threshold."

Dr. Caprio joined other TODAY investigators in a lively discussion of the study. TODAY’s primary end points were released in April; the ADA presentation not only broke down predictors of treatment failure, but pointed up a disturbing scenario of the medical comorbidities that already haunt children with type 2 diabetes.

Hypertension, enlarged hearts, retinopathy, dyslipidemia, and incipient kidney dysfunction were all common among the 699 children who participated in the study, said Dr. Neil White of Washington University, St. Louis. The problems occurred early and frequently.

"Within 7-8 months after diagnosis, 14% already had nonproliferative retinopathy. That number is similar to the amount found in the Diabetes Prevention Program study, for a population that was an average of 55 years old. These subjects are getting the signs of disease during childhood that older adults have."

TODAY examined three different treatment regimens among 699 children with newly diagnosed type 2 diabetes: metformin alone, metformin plus lifestyle modifications, and metformin plus rosiglitazone (N. Engl. J. Med. 2012 April 29 [doi:10.1056/NEJMoa1109333]).

The 60-month trial started all patients aged 10-17 years on metformin 2,000 mg/day. Randomization occurred after HbA_1c stabilized at 8% or lower with metformin therapy. The primary end point was time to the failure of glycemic control (defined as an HbA_1c of at least 8% for 6 months, or sustained metabolic decompensation that required insulin treatment).

Overall, nearly half of the cohort (46%) failed to maintain glycemic control; the median time to failure was 11 months. However, compared with the combination therapy group, significantly more of those taking metformin alone failed glycemic control (52% vs. 39%). The failure rate in the lifestyle intervention group was 47% – not significantly different from that of the metformin mono- or combination therapy groups.

The new analyses examined predictors of response, and looked at medical comorbidities among the patients. A seemingly innocuous increase in HbA_1c during the initial 6-month stabilization phase was the strongest predictor of whether a child would achieve a durable response to the long-term therapy, said Dr. Kenneth Copeland, a TODAY investigator and the Milburn Endowed Chair of Pediatrics at the University of Oklahoma Health Sciences Center, Oklahoma City.

In a univariate analysis, a number of baseline factors were associated with failure, including race (with blacks doing significantly worse than whites); the presence of depression; lower household income; lower insulinogenic index, and lower HbA_1c.

The baseline difference in HbA_1c was statistically significant, but still apparently small, he said. After initial stabilization, those who maintained and failed control still had blood glucose levels in the nondiabetic range (5.7% vs. 6.4%)

But in a multivariate model, only baseline HbA_1c strongly predicted response to long-term therapy. Those who maintained glucose control had very moderate HbA_1c increases, from a median of 6.4% to about 7% by the end of the study. Those who failed therapy had a much steeper increase, rising from 5.7% at baseline to about 7.3% at the end.

"Baseline HbA_1c appears to be the best predictor of failure, regardless of whether it is still in the nondiabetic range. A rapidly rising level – even in the normal range – is associated with incipient failure and may suggest the need to intensify treatment early."

The increases in blood sugar that arose during stabilization and treatment probably signaled beta-cell decline in an increasingly stressed pancreas, Dr. Caprio said. The analysis used two algorithms to assess beta-cell function: the insulinogenic index (the difference in blood glucose before and after insulin administration) and the oral disposition index (a measure of beta-cell function relative to insulin). The baseline for this analysis was the point of randomization, when glucose was stabilized at 8% or lower after 6 months of treatment.

"At 6 months, we saw a dramatic difference between the rosiglitazone/metformin arm and the other arms," with a 20% increase in insulin sensitivity compared with no change in the other groups. Thereafter, however, insulin sensitivity started to decrease in the rosiglitazone group, until it plateaued.

By the end of the study, beta-cell function was also similar between the groups, but those in the rosiglitazone group showed a slower decline during the treatment period. This indicated that rosiglitazone was able to stabilize function for at least a period of time. "In contrast, we saw a dramatic fall-off in the beta-cell function in the other two groups, although by 24 months there were no differences among any of them."

"The favorable changes relative to insulin sensitivity were responsible for the observed reduced failure rate in the rosiglitazone group; telling is that this profound degree of insulin resistance in obese youth with type 2 diabetes is unresponsive to the effect of metformin, alone or with lifestyle therapy," Dr. Caprio said.

Because metformin failed those children, the better alternative appears to be more aggressive treatment early in the disease. "The introduction of such aggressive therapy early in the course of the disease appears to slow the decline of beta-cell adaptation and improve the chance or durable control."

Children in TODAY were plagued with much the same medical issues as are adults with type 2 diabetes, said Dr. Neil White, chairman of the TODAY data safety and monitoring committee.

The subjects had a very short duration of diabetes, having been diagnosed 5-7 months before the metformin stabilization period. At that point, 34% already had hypertension. Abnormal LDL cholesterol levels were present in 10%, and high triglycerides in 54%. Some 20% had microalbuminuria, suggesting possible progression to kidney disease.

All of the children, regardless of their treatment group or outcome, showed somewhat abnormal left ventricular and right auricular measurements, noted Dr. White, who is also director of the pediatric clinical research unit at Washington University.

Echocardiography showed that the children had a left auricular diameter of up to 2.4 cm/m height, greater than the population median. All also had a left ventricular mass greater than the population median of 30 g/m², with some reaching almost 45 g/m².

"These youngsters had big hearts that are going to predispose them to heart disease in the future," Dr. White said.

At the end of the study, most children (523) underwent fundus photography to evaluate retinopathy. About 14% of them had some degree of nonproliferative retinopathy, putting the young group’s prevalence squarely in line with prevalence in adults with type 2 diabetes.

"This profound insulin resistance was a culprit ... against which metformin had no effect at all," Dr. Caprio said. "Even with the two best drugs we had at the time of this study – metformin and rosiglitazone – we were not able to improve outcomes in 50% of these kids."

Dr. Copeland agreed.

"The finding that 50% of our children failed therapy indicates that this is an incredibly tough disease to treat by any modality we have. Our target as a society needs to be ‘How do we prevent diabetes in children?’ because once they get it, it’s going to be really tough to intervene effectively."

TODAY was funded by the National Institute of Diabetes and Digestive and Kidney Diseases. All of the presenters disclosed financial relationships with Daiichi Sankyo, but said these were not relevant to the TODAY study.

PHILADELPHIA – Getting a firm, early grip on blood glucose confers the best chance for durable glycemic control in children with type 2 diabetes.

Any rapid increase in hemoglobin A_1c after initial stabilization probably signals the need to quickly gear up treatment, Dr. Sonia Caprio said at the annual scientific sessions of the American Diabetes Association.

"The best single predictor of treatment failure – regardless of therapy – is baseline [hemoglobin A_1c], even if it’s in the nondiabetic range" after stabilization, said Dr. Caprio, a professor of pediatric endocrinology at Yale University in New Haven, Conn. "A rapidly rising HbA_1c, even if it’s still in the normal range, is associated with failure, and suggests the need for more intense treatment."

These new findings from the TODAY (Treatment Options for Type 2 Diabetes in Adolescents and Youth) study mean that physicians may need new treatment markers for diabetes regimens in youth, she said. "We need to identify specific thresholds for intensifying therapy, perhaps at diagnosis, perhaps after a certain period of treatment, or when the patient reaches a certain HbA_1c threshold."

Dr. Caprio joined other TODAY investigators in a lively discussion of the study. TODAY’s primary end points were released in April; the ADA presentation not only broke down predictors of treatment failure, but pointed up a disturbing scenario of the medical comorbidities that already haunt children with type 2 diabetes.

Hypertension, enlarged hearts, retinopathy, dyslipidemia, and incipient kidney dysfunction were all common among the 699 children who participated in the study, said Dr. Neil White of Washington University, St. Louis. The problems occurred early and frequently.

"Within 7-8 months after diagnosis, 14% already had nonproliferative retinopathy. That number is similar to the amount found in the Diabetes Prevention Program study, for a population that was an average of 55 years old. These subjects are getting the signs of disease during childhood that older adults have."

TODAY examined three different treatment regimens among 699 children with newly diagnosed type 2 diabetes: metformin alone, metformin plus lifestyle modifications, and metformin plus rosiglitazone (N. Engl. J. Med. 2012 April 29 [doi:10.1056/NEJMoa1109333]).

The 60-month trial started all patients aged 10-17 years on metformin 2,000 mg/day. Randomization occurred after HbA_1c stabilized at 8% or lower with metformin therapy. The primary end point was time to the failure of glycemic control (defined as an HbA_1c of at least 8% for 6 months, or sustained metabolic decompensation that required insulin treatment).

Overall, nearly half of the cohort (46%) failed to maintain glycemic control; the median time to failure was 11 months. However, compared with the combination therapy group, significantly more of those taking metformin alone failed glycemic control (52% vs. 39%). The failure rate in the lifestyle intervention group was 47% – not significantly different from that of the metformin mono- or combination therapy groups.

The new analyses examined predictors of response, and looked at medical comorbidities among the patients. A seemingly innocuous increase in HbA_1c during the initial 6-month stabilization phase was the strongest predictor of whether a child would achieve a durable response to the long-term therapy, said Dr. Kenneth Copeland, a TODAY investigator and the Milburn Endowed Chair of Pediatrics at the University of Oklahoma Health Sciences Center, Oklahoma City.

In a univariate analysis, a number of baseline factors were associated with failure, including race (with blacks doing significantly worse than whites); the presence of depression; lower household income; lower insulinogenic index, and lower HbA_1c.

The baseline difference in HbA_1c was statistically significant, but still apparently small, he said. After initial stabilization, those who maintained and failed control still had blood glucose levels in the nondiabetic range (5.7% vs. 6.4%)

But in a multivariate model, only baseline HbA_1c strongly predicted response to long-term therapy. Those who maintained glucose control had very moderate HbA_1c increases, from a median of 6.4% to about 7% by the end of the study. Those who failed therapy had a much steeper increase, rising from 5.7% at baseline to about 7.3% at the end.

"Baseline HbA_1c appears to be the best predictor of failure, regardless of whether it is still in the nondiabetic range. A rapidly rising level – even in the normal range – is associated with incipient failure and may suggest the need to intensify treatment early."

The increases in blood sugar that arose during stabilization and treatment probably signaled beta-cell decline in an increasingly stressed pancreas, Dr. Caprio said. The analysis used two algorithms to assess beta-cell function: the insulinogenic index (the difference in blood glucose before and after insulin administration) and the oral disposition index (a measure of beta-cell function relative to insulin). The baseline for this analysis was the point of randomization, when glucose was stabilized at 8% or lower after 6 months of treatment.

"At 6 months, we saw a dramatic difference between the rosiglitazone/metformin arm and the other arms," with a 20% increase in insulin sensitivity compared with no change in the other groups. Thereafter, however, insulin sensitivity started to decrease in the rosiglitazone group, until it plateaued.

By the end of the study, beta-cell function was also similar between the groups, but those in the rosiglitazone group showed a slower decline during the treatment period. This indicated that rosiglitazone was able to stabilize function for at least a period of time. "In contrast, we saw a dramatic fall-off in the beta-cell function in the other two groups, although by 24 months there were no differences among any of them."

"The favorable changes relative to insulin sensitivity were responsible for the observed reduced failure rate in the rosiglitazone group; telling is that this profound degree of insulin resistance in obese youth with type 2 diabetes is unresponsive to the effect of metformin, alone or with lifestyle therapy," Dr. Caprio said.

Because metformin failed those children, the better alternative appears to be more aggressive treatment early in the disease. "The introduction of such aggressive therapy early in the course of the disease appears to slow the decline of beta-cell adaptation and improve the chance or durable control."

Children in TODAY were plagued with much the same medical issues as are adults with type 2 diabetes, said Dr. Neil White, chairman of the TODAY data safety and monitoring committee.

The subjects had a very short duration of diabetes, having been diagnosed 5-7 months before the metformin stabilization period. At that point, 34% already had hypertension. Abnormal LDL cholesterol levels were present in 10%, and high triglycerides in 54%. Some 20% had microalbuminuria, suggesting possible progression to kidney disease.

All of the children, regardless of their treatment group or outcome, showed somewhat abnormal left ventricular and right auricular measurements, noted Dr. White, who is also director of the pediatric clinical research unit at Washington University.

Echocardiography showed that the children had a left auricular diameter of up to 2.4 cm/m height, greater than the population median. All also had a left ventricular mass greater than the population median of 30 g/m², with some reaching almost 45 g/m².

"These youngsters had big hearts that are going to predispose them to heart disease in the future," Dr. White said.

At the end of the study, most children (523) underwent fundus photography to evaluate retinopathy. About 14% of them had some degree of nonproliferative retinopathy, putting the young group’s prevalence squarely in line with prevalence in adults with type 2 diabetes.

"This profound insulin resistance was a culprit ... against which metformin had no effect at all," Dr. Caprio said. "Even with the two best drugs we had at the time of this study – metformin and rosiglitazone – we were not able to improve outcomes in 50% of these kids."

Dr. Copeland agreed.

"The finding that 50% of our children failed therapy indicates that this is an incredibly tough disease to treat by any modality we have. Our target as a society needs to be ‘How do we prevent diabetes in children?’ because once they get it, it’s going to be really tough to intervene effectively."

TODAY was funded by the National Institute of Diabetes and Digestive and Kidney Diseases. All of the presenters disclosed financial relationships with Daiichi Sankyo, but said these were not relevant to the TODAY study.

PHILADELPHIA – Exenatide twice daily prevented glycemic deterioration better than did glimepiride in type 2 diabetes patients who had metformin failure in an open label, randomized, controlled trial.

Common practice for patients with type 2 diabetes who are inadequately controlled by metformin has been to add a sulphonylurea to the treatment regimen. Although these agents have a rapid short-term effect and are low cost, they are associated with hypoglycemia and worsening glycemic control over time. Glucagonlike peptide–1 (GLP-1) receptor agonists, on the other hand, improve glycemic control with no increased risk of hypoglycemia, and are associated with weight loss and improvements in cardiovascular risk factors and can be used as alternative add-ons to metformin treatment, said Dr. Guntram Schernthaner at the annual scientific sessions of the American Diabetes Association.

"Up to now, EUREXA is the longest study undertaken with a GLP-1 receptor agonist, and could contribute substantially to decisions in clinical practice," said Dr. Schernthaner of Rudolfstiftung Hospital in Vienna.

In this open-label, randomized, controlled European trial, 1,029 patients from 128 centers in 14 countries were randomized to exenatide twice daily (starting at 5 mcg twice daily for 4 weeks, followed by 10 mcg twice daily for the remaining study period) or glimepiride per physician dosing practice. The intention-to-treat population comprised 490 patients in the exenatide group and 487 in the glimepiride group.

The study was simultaneously published in the Lancet (2012 June 9 [doi:10.1016/S0140-6736[12]60479-6]).

Patients were taking a median metformin dose of 2,000 mg/day. The mean exenatide dose was 17.4 mcg/day and the mean glimepiride dose was 2.0 mg/day. Average treatment time was about 2 years in both groups. Mean baseline hemoglobin A_1c was 7.5% in the exenatide group and 7.4% in those randomized to glimepiride.

The primary outcome was time to inadequate glycemic control (defined as an HbA_1c level greater than 9% after the first 3 months of treatment, or greater than 7% at two consecutive visits 3 months apart after the first 6 months). That end point was reached by 203 (41%) patients in the exenatide group, compared with 262 (54%) in the glimepiride group, a significant difference. These included 5 in the exenatide group and 7 in the glimepiride group who had an HbA_1c of more than 9%, and 198 vs. 255 with HbA_1c levels greater than 7% at two visits. The hazard ratio for inadequate glycemic control with exenatide, compared with glimepiride, was 0.75.

Median time to inadequate HbA_1c control was 180 weeks with exenatide vs. 142 weeks with glimepiride. The risk of treatment failure was greater with higher baseline HbA_1c, as was the reduction in risk with exenatide compared with glimepiride, Dr. Schernthaner reported.

Compared with those in the glimepiride group, significantly more patients in the exenatide group achieved HbA_1c values lower than 7% (31% vs. 45%, respectively) and 6.5% or less (18% vs. 29%). Fasting glucose levels were also significantly lower in the exenatide group after year 1, year 2, and year 3 of treatment. There was an overall treatment effect in favor of the exenatide group, with significant differences between groups in mean HbA_1c levels at years 2 and 3. Self-monitored postmeal excursions of blood glucose after meals were significantly lower in the exenatide group than in the glimepiride group.

Five patients in each treatment group died, but all deaths were from causes regarded as unrelated to the study treatment. Significantly more patients discontinued in the exenatide group than in the glimepiride group because of adverse events (49 vs. 17). However, discontinuations due to adverse events mostly gastrointestinal were significantly different between treatments only in the first 6 months of study, Dr. Schernthaner said.

Systolic blood pressure decreased in patients in the exenatide group but not in the glimepiride group, resulting in a significant, 3.1-mm Hg difference between groups from year 1 to year 3. Body weight fell from baseline to end point in the exenatide group by 3.32 kg, while it rose by 1.15 kg in the glimepiride group, also a significant difference.

The proportion of patients reporting hypoglycemia was lower in patients in the exenatide group than in those in the glimepiride group. For hypoglycemia of any type during the study, the estimated mean rate was 1.52 episodes per year in the exenatide group, compared with 5.32 episodes per year in the glimepiride group, he reported.

EUREXA was funded by Eli Lilly and Amylin Pharmaceuticals. Dr. Schernthaner disclosed that he has been a consultant for and received honoraria from Amgen, AstraZeneca, Bristol-Myers Squibb, Boehringer-Ingelheim, Eli Lilly, Merck, Novartis, Poxel, Roche, Sanofi-Aventis, Novo Nordisk, Servier, and Takeda. Dr. Madsbad has been a consultant for, adviser to, and/or speaker for Novartis Pharma, Novo Nordisk, Merck Sharp & Dohme, Sanofi-Aventis, AstraZeneca, Johnson & Johnson, Roche, Mannkind, Boehringer-Ingelheim, Zeeland, Lilly, Intarcia Therapeutics, Schering-Plough, and Bristol-Myers Squibb.

PHILADELPHIA – Exenatide twice daily prevented glycemic deterioration better than did glimepiride in type 2 diabetes patients who had metformin failure in an open label, randomized, controlled trial.

Common practice for patients with type 2 diabetes who are inadequately controlled by metformin has been to add a sulphonylurea to the treatment regimen. Although these agents have a rapid short-term effect and are low cost, they are associated with hypoglycemia and worsening glycemic control over time. Glucagonlike peptide–1 (GLP-1) receptor agonists, on the other hand, improve glycemic control with no increased risk of hypoglycemia, and are associated with weight loss and improvements in cardiovascular risk factors and can be used as alternative add-ons to metformin treatment, said Dr. Guntram Schernthaner at the annual scientific sessions of the American Diabetes Association.

"Up to now, EUREXA is the longest study undertaken with a GLP-1 receptor agonist, and could contribute substantially to decisions in clinical practice," said Dr. Schernthaner of Rudolfstiftung Hospital in Vienna.

In this open-label, randomized, controlled European trial, 1,029 patients from 128 centers in 14 countries were randomized to exenatide twice daily (starting at 5 mcg twice daily for 4 weeks, followed by 10 mcg twice daily for the remaining study period) or glimepiride per physician dosing practice. The intention-to-treat population comprised 490 patients in the exenatide group and 487 in the glimepiride group.

The study was simultaneously published in the Lancet (2012 June 9 [doi:10.1016/S0140-6736[12]60479-6]).

Patients were taking a median metformin dose of 2,000 mg/day. The mean exenatide dose was 17.4 mcg/day and the mean glimepiride dose was 2.0 mg/day. Average treatment time was about 2 years in both groups. Mean baseline hemoglobin A_1c was 7.5% in the exenatide group and 7.4% in those randomized to glimepiride.

The primary outcome was time to inadequate glycemic control (defined as an HbA_1c level greater than 9% after the first 3 months of treatment, or greater than 7% at two consecutive visits 3 months apart after the first 6 months). That end point was reached by 203 (41%) patients in the exenatide group, compared with 262 (54%) in the glimepiride group, a significant difference. These included 5 in the exenatide group and 7 in the glimepiride group who had an HbA_1c of more than 9%, and 198 vs. 255 with HbA_1c levels greater than 7% at two visits. The hazard ratio for inadequate glycemic control with exenatide, compared with glimepiride, was 0.75.

Median time to inadequate HbA_1c control was 180 weeks with exenatide vs. 142 weeks with glimepiride. The risk of treatment failure was greater with higher baseline HbA_1c, as was the reduction in risk with exenatide compared with glimepiride, Dr. Schernthaner reported.

Compared with those in the glimepiride group, significantly more patients in the exenatide group achieved HbA_1c values lower than 7% (31% vs. 45%, respectively) and 6.5% or less (18% vs. 29%). Fasting glucose levels were also significantly lower in the exenatide group after year 1, year 2, and year 3 of treatment. There was an overall treatment effect in favor of the exenatide group, with significant differences between groups in mean HbA_1c levels at years 2 and 3. Self-monitored postmeal excursions of blood glucose after meals were significantly lower in the exenatide group than in the glimepiride group.

Five patients in each treatment group died, but all deaths were from causes regarded as unrelated to the study treatment. Significantly more patients discontinued in the exenatide group than in the glimepiride group because of adverse events (49 vs. 17). However, discontinuations due to adverse events mostly gastrointestinal were significantly different between treatments only in the first 6 months of study, Dr. Schernthaner said.

Systolic blood pressure decreased in patients in the exenatide group but not in the glimepiride group, resulting in a significant, 3.1-mm Hg difference between groups from year 1 to year 3. Body weight fell from baseline to end point in the exenatide group by 3.32 kg, while it rose by 1.15 kg in the glimepiride group, also a significant difference.

The proportion of patients reporting hypoglycemia was lower in patients in the exenatide group than in those in the glimepiride group. For hypoglycemia of any type during the study, the estimated mean rate was 1.52 episodes per year in the exenatide group, compared with 5.32 episodes per year in the glimepiride group, he reported.

EUREXA was funded by Eli Lilly and Amylin Pharmaceuticals. Dr. Schernthaner disclosed that he has been a consultant for and received honoraria from Amgen, AstraZeneca, Bristol-Myers Squibb, Boehringer-Ingelheim, Eli Lilly, Merck, Novartis, Poxel, Roche, Sanofi-Aventis, Novo Nordisk, Servier, and Takeda. Dr. Madsbad has been a consultant for, adviser to, and/or speaker for Novartis Pharma, Novo Nordisk, Merck Sharp & Dohme, Sanofi-Aventis, AstraZeneca, Johnson & Johnson, Roche, Mannkind, Boehringer-Ingelheim, Zeeland, Lilly, Intarcia Therapeutics, Schering-Plough, and Bristol-Myers Squibb.

PHILADELPHIA – Exenatide twice daily prevented glycemic deterioration better than did glimepiride in type 2 diabetes patients who had metformin failure in an open label, randomized, controlled trial.

Common practice for patients with type 2 diabetes who are inadequately controlled by metformin has been to add a sulphonylurea to the treatment regimen. Although these agents have a rapid short-term effect and are low cost, they are associated with hypoglycemia and worsening glycemic control over time. Glucagonlike peptide–1 (GLP-1) receptor agonists, on the other hand, improve glycemic control with no increased risk of hypoglycemia, and are associated with weight loss and improvements in cardiovascular risk factors and can be used as alternative add-ons to metformin treatment, said Dr. Guntram Schernthaner at the annual scientific sessions of the American Diabetes Association.

"Up to now, EUREXA is the longest study undertaken with a GLP-1 receptor agonist, and could contribute substantially to decisions in clinical practice," said Dr. Schernthaner of Rudolfstiftung Hospital in Vienna.

In this open-label, randomized, controlled European trial, 1,029 patients from 128 centers in 14 countries were randomized to exenatide twice daily (starting at 5 mcg twice daily for 4 weeks, followed by 10 mcg twice daily for the remaining study period) or glimepiride per physician dosing practice. The intention-to-treat population comprised 490 patients in the exenatide group and 487 in the glimepiride group.

The study was simultaneously published in the Lancet (2012 June 9 [doi:10.1016/S0140-6736[12]60479-6]).

Patients were taking a median metformin dose of 2,000 mg/day. The mean exenatide dose was 17.4 mcg/day and the mean glimepiride dose was 2.0 mg/day. Average treatment time was about 2 years in both groups. Mean baseline hemoglobin A_1c was 7.5% in the exenatide group and 7.4% in those randomized to glimepiride.

The primary outcome was time to inadequate glycemic control (defined as an HbA_1c level greater than 9% after the first 3 months of treatment, or greater than 7% at two consecutive visits 3 months apart after the first 6 months). That end point was reached by 203 (41%) patients in the exenatide group, compared with 262 (54%) in the glimepiride group, a significant difference. These included 5 in the exenatide group and 7 in the glimepiride group who had an HbA_1c of more than 9%, and 198 vs. 255 with HbA_1c levels greater than 7% at two visits. The hazard ratio for inadequate glycemic control with exenatide, compared with glimepiride, was 0.75.

Median time to inadequate HbA_1c control was 180 weeks with exenatide vs. 142 weeks with glimepiride. The risk of treatment failure was greater with higher baseline HbA_1c, as was the reduction in risk with exenatide compared with glimepiride, Dr. Schernthaner reported.

Compared with those in the glimepiride group, significantly more patients in the exenatide group achieved HbA_1c values lower than 7% (31% vs. 45%, respectively) and 6.5% or less (18% vs. 29%). Fasting glucose levels were also significantly lower in the exenatide group after year 1, year 2, and year 3 of treatment. There was an overall treatment effect in favor of the exenatide group, with significant differences between groups in mean HbA_1c levels at years 2 and 3. Self-monitored postmeal excursions of blood glucose after meals were significantly lower in the exenatide group than in the glimepiride group.

Five patients in each treatment group died, but all deaths were from causes regarded as unrelated to the study treatment. Significantly more patients discontinued in the exenatide group than in the glimepiride group because of adverse events (49 vs. 17). However, discontinuations due to adverse events mostly gastrointestinal were significantly different between treatments only in the first 6 months of study, Dr. Schernthaner said.

Systolic blood pressure decreased in patients in the exenatide group but not in the glimepiride group, resulting in a significant, 3.1-mm Hg difference between groups from year 1 to year 3. Body weight fell from baseline to end point in the exenatide group by 3.32 kg, while it rose by 1.15 kg in the glimepiride group, also a significant difference.

The proportion of patients reporting hypoglycemia was lower in patients in the exenatide group than in those in the glimepiride group. For hypoglycemia of any type during the study, the estimated mean rate was 1.52 episodes per year in the exenatide group, compared with 5.32 episodes per year in the glimepiride group, he reported.

EUREXA was funded by Eli Lilly and Amylin Pharmaceuticals. Dr. Schernthaner disclosed that he has been a consultant for and received honoraria from Amgen, AstraZeneca, Bristol-Myers Squibb, Boehringer-Ingelheim, Eli Lilly, Merck, Novartis, Poxel, Roche, Sanofi-Aventis, Novo Nordisk, Servier, and Takeda. Dr. Madsbad has been a consultant for, adviser to, and/or speaker for Novartis Pharma, Novo Nordisk, Merck Sharp & Dohme, Sanofi-Aventis, AstraZeneca, Johnson & Johnson, Roche, Mannkind, Boehringer-Ingelheim, Zeeland, Lilly, Intarcia Therapeutics, Schering-Plough, and Bristol-Myers Squibb.

Getting prediabetes patients to regain glucose control as early as possible provides the best shot at preventing disease progression, regardless of whether it is achieved through lifestyle interventions or drug therapy, according to the latest analysis from the landmark Diabetes Prevention Program.

In fact, even a transient reversion to normal glucose regulation was associated with a significantly reduced risk of diabetes among participants in the DPP, a major randomized trial comparing weight loss and dietary changes with drug therapy for preventing the onset of type 2 diabetes.

Compared with 1,096 participants in that multicenter trial who consistently had prediabetes, the risk of diabetes was 56% lower (hazard ratio, 0.44) among 894 individuals who had returned to normal glucose regulation during the intervention period, according to findings from the Diabetes Prevention Program Outcomes Study (DPPOS), an ongoing postintervention observational study, Dr. Leigh Perreault of the University of Colorado, Aurora, and her colleagues reported online in the June 9 Lancet.

"Diabetes risk reduction was strongly associated with the number of times normal glucose regulation was achieved. Specifically, diabetes risk was reduced 47% in DPPOS if normal glucose regulation was attained only once (HR, 0.53); 61% if it was reached twice (HR,0.39); and 67% if it was reached three times (HR, 0.33)," the investigators said (Lancet 2012 June 9 [doi:10.106/S0140-6736(12)60525-X]).

At baseline, all 3,234 participants were at risk for developing diabetes. They were randomized to receive lifestyle interventions, metformin treatment, or placebo, and were followed for a median of 3.2 years during the intervention.

The current analysis, at a median follow-up of 5.7 years, shows that the reduced risk for diabetes among those subjects who returned to normal glucose regulation during the intervention was not affected by group assignment, the investigators said.

Factors associated with increased diabetes risk during DPPOS included age younger than 45 years (HR, 1.47), and African American ethnic origin (HR, 1.77).

"Paradoxically, increased weight loss during DPP adversely affected diabetes risk (HR, 1.26) in DPPOS independent of previous treatment," the investigators also noted, adding that this was "probably because of the high rate of weight regain in DPPOS with associated adverse effects on diabetes risk."

High body mass index at the beginning of DPPOS also related to diabetes risk during DPPOS follow-up (HR, 1.14). Higher beta-cell function and insulin sensitivity were protective (HR, 0.80 and 0.83, respectively).

Those who consistently had prediabetes during DPP – despite intensive lifestyle interventions – had an increased risk of developing diabetes during DPPOS, compared with those who didn’t have prediabetes and who were in the placebo group (HR, 1.31).

"Although there is widespread consensus that diabetes prevention is crucially important, there is less agreement with respect to the particular intervention," the investigators noted. "Results from the present analysis would contend that the strategy is unimportant as long as the intervention is early (when someone has prediabetes) and can restore normal glucose regulation, even if transiently."

The findings also suggest that maintenance of prediabetes despite intensive lifestyle modification represents a high-risk state that may warrant additional interventions, she and her colleagues said.

"Together, these data serve as essential clinical information to support early and aggressive measures for long-term prevention of diabetes in people at risk," they concluded.

This study was funded by the National Institutes of Health. The authors reported having no relevant financial conflicts.

Getting prediabetes patients to regain glucose control as early as possible provides the best shot at preventing disease progression, regardless of whether it is achieved through lifestyle interventions or drug therapy, according to the latest analysis from the landmark Diabetes Prevention Program.

In fact, even a transient reversion to normal glucose regulation was associated with a significantly reduced risk of diabetes among participants in the DPP, a major randomized trial comparing weight loss and dietary changes with drug therapy for preventing the onset of type 2 diabetes.

Compared with 1,096 participants in that multicenter trial who consistently had prediabetes, the risk of diabetes was 56% lower (hazard ratio, 0.44) among 894 individuals who had returned to normal glucose regulation during the intervention period, according to findings from the Diabetes Prevention Program Outcomes Study (DPPOS), an ongoing postintervention observational study, Dr. Leigh Perreault of the University of Colorado, Aurora, and her colleagues reported online in the June 9 Lancet.

"Diabetes risk reduction was strongly associated with the number of times normal glucose regulation was achieved. Specifically, diabetes risk was reduced 47% in DPPOS if normal glucose regulation was attained only once (HR, 0.53); 61% if it was reached twice (HR,0.39); and 67% if it was reached three times (HR, 0.33)," the investigators said (Lancet 2012 June 9 [doi:10.106/S0140-6736(12)60525-X]).

At baseline, all 3,234 participants were at risk for developing diabetes. They were randomized to receive lifestyle interventions, metformin treatment, or placebo, and were followed for a median of 3.2 years during the intervention.

The current analysis, at a median follow-up of 5.7 years, shows that the reduced risk for diabetes among those subjects who returned to normal glucose regulation during the intervention was not affected by group assignment, the investigators said.

Factors associated with increased diabetes risk during DPPOS included age younger than 45 years (HR, 1.47), and African American ethnic origin (HR, 1.77).

"Paradoxically, increased weight loss during DPP adversely affected diabetes risk (HR, 1.26) in DPPOS independent of previous treatment," the investigators also noted, adding that this was "probably because of the high rate of weight regain in DPPOS with associated adverse effects on diabetes risk."

High body mass index at the beginning of DPPOS also related to diabetes risk during DPPOS follow-up (HR, 1.14). Higher beta-cell function and insulin sensitivity were protective (HR, 0.80 and 0.83, respectively).

Those who consistently had prediabetes during DPP – despite intensive lifestyle interventions – had an increased risk of developing diabetes during DPPOS, compared with those who didn’t have prediabetes and who were in the placebo group (HR, 1.31).

"Although there is widespread consensus that diabetes prevention is crucially important, there is less agreement with respect to the particular intervention," the investigators noted. "Results from the present analysis would contend that the strategy is unimportant as long as the intervention is early (when someone has prediabetes) and can restore normal glucose regulation, even if transiently."

The findings also suggest that maintenance of prediabetes despite intensive lifestyle modification represents a high-risk state that may warrant additional interventions, she and her colleagues said.

"Together, these data serve as essential clinical information to support early and aggressive measures for long-term prevention of diabetes in people at risk," they concluded.

This study was funded by the National Institutes of Health. The authors reported having no relevant financial conflicts.

Getting prediabetes patients to regain glucose control as early as possible provides the best shot at preventing disease progression, regardless of whether it is achieved through lifestyle interventions or drug therapy, according to the latest analysis from the landmark Diabetes Prevention Program.

In fact, even a transient reversion to normal glucose regulation was associated with a significantly reduced risk of diabetes among participants in the DPP, a major randomized trial comparing weight loss and dietary changes with drug therapy for preventing the onset of type 2 diabetes.

Compared with 1,096 participants in that multicenter trial who consistently had prediabetes, the risk of diabetes was 56% lower (hazard ratio, 0.44) among 894 individuals who had returned to normal glucose regulation during the intervention period, according to findings from the Diabetes Prevention Program Outcomes Study (DPPOS), an ongoing postintervention observational study, Dr. Leigh Perreault of the University of Colorado, Aurora, and her colleagues reported online in the June 9 Lancet.

"Diabetes risk reduction was strongly associated with the number of times normal glucose regulation was achieved. Specifically, diabetes risk was reduced 47% in DPPOS if normal glucose regulation was attained only once (HR, 0.53); 61% if it was reached twice (HR,0.39); and 67% if it was reached three times (HR, 0.33)," the investigators said (Lancet 2012 June 9 [doi:10.106/S0140-6736(12)60525-X]).

At baseline, all 3,234 participants were at risk for developing diabetes. They were randomized to receive lifestyle interventions, metformin treatment, or placebo, and were followed for a median of 3.2 years during the intervention.

The current analysis, at a median follow-up of 5.7 years, shows that the reduced risk for diabetes among those subjects who returned to normal glucose regulation during the intervention was not affected by group assignment, the investigators said.

Factors associated with increased diabetes risk during DPPOS included age younger than 45 years (HR, 1.47), and African American ethnic origin (HR, 1.77).

"Paradoxically, increased weight loss during DPP adversely affected diabetes risk (HR, 1.26) in DPPOS independent of previous treatment," the investigators also noted, adding that this was "probably because of the high rate of weight regain in DPPOS with associated adverse effects on diabetes risk."

High body mass index at the beginning of DPPOS also related to diabetes risk during DPPOS follow-up (HR, 1.14). Higher beta-cell function and insulin sensitivity were protective (HR, 0.80 and 0.83, respectively).

Those who consistently had prediabetes during DPP – despite intensive lifestyle interventions – had an increased risk of developing diabetes during DPPOS, compared with those who didn’t have prediabetes and who were in the placebo group (HR, 1.31).

"Although there is widespread consensus that diabetes prevention is crucially important, there is less agreement with respect to the particular intervention," the investigators noted. "Results from the present analysis would contend that the strategy is unimportant as long as the intervention is early (when someone has prediabetes) and can restore normal glucose regulation, even if transiently."

The findings also suggest that maintenance of prediabetes despite intensive lifestyle modification represents a high-risk state that may warrant additional interventions, she and her colleagues said.

"Together, these data serve as essential clinical information to support early and aggressive measures for long-term prevention of diabetes in people at risk," they concluded.

This study was funded by the National Institutes of Health. The authors reported having no relevant financial conflicts.

Quality improvement strategies that aim to optimize systems of care have significant favorable effects on diabetes care outcomes, Andrea C. Tricco, Ph.D., of Li Ka Shing Knowledge Institute of St. Michael’s Hospital, Toronto, and her colleagues reported online June 9 in the Lancet.

While QI interventions targeting patients might be beneficial irrespective of baseline HbA_1c, those interventions targeting providers only seem beneficial when baseline HbA_1c is greater than 8.0%, according to the findings of their systematic review and meta-analysis, which together involved more than 123,000 patients.

The results were reported simultaneously at the annual meeting of the American Diabetes Association in Philadelphia.

Based on data from a total of 120 relevant trials, QI strategies reduced HbA_1c by a mean difference of 0.37%; and reduced LDL cholesterol by 0.10 mmol/L, systolic blood pressure by 3.13 mm Hg, and diastolic blood pressure by 1.55 mm Hg versus usual care over a median 12-month period, the authors found (Lancet 2012 June 9 [doi: 10.1016/S0140-6736(12)60480-2]).

Specifically, strategies targeting clinicians’ education and audit and feedback led to HbA_1c reductions of 0.33% and 0.44%, respectively, when baseline HbA_1c concentrations were greater than 8.0%, compared with no improvement when baseline HbA_1c was less than 8.0%.

Patient education seemed more effective than reminders when baseline HbA_1c was greater than 8.0%, but less effective when it was less than 8.0%.

Furthermore, QI strategies were associated with increases in aspirin use; use of any antihypertensive drugs; and screening for retinopathy, renal involvement, and foot complications over median follow-up periods ranging from 12 to 18 months. The QI strategies did not, however, have a significant effect on statin use, adequate control of hypertension, or smoking cessation rates over median follow-up periods ranging from 12 to 19 months.

In trials that enrolled patients with an HbA_1c of greater than 8.0%, declines of more than 0.5% were seen with certain QI strategies (specifically, those targeting team changes, case management, patient education, and promotion of self-management). In trials enrolling those with an HbA_1c of 8.0% or less, declines of that magnitude were seen for one QI strategy (facilitated relay).

Nonetheless, "after adjustment for median baseline HbA_1c values and effective sample size, the QI strategies were associated with significantly lower HbA_1c than usual care was," they said.

"We noted greater improvements in HbA_1c control for QI strategies targeting health systems and patients. In fact, all QI strategies were associated with significant changes in HbA_1c except for clinician education," they said.

The trials included in this review and meta-analysis were reported between July 2003 and July 2010, and were selected using 11 predefined QI strategies or financial incentives targeting health care professionals for the management of adult outpatients with diabetes.

"The QI strategies targeted health systems (e.g., team changes), professionals (e.g., professional reminders), or patients (e.g., promotion of self-management)," the investigators explained.

The findings, which represent an update of a previous review that assessed the effects of QI strategies on glycemic control, provide an assessment of the effects of QI strategies on a broader range of diabetic care.

The analysis was limited by the complexity of the QI strategies studied, the inability to control for all potential confounding factors, and the short duration of follow-up in many of the trials, the authors wrote.

Despite these limitations, the findings indicate that diabetes outcomes can be improved by QI interventions. Wide implementation of QI strategies could lead to important population benefits, as data have shown that a 1% reduction in mean HbA_1c results in 21% fewer deaths, 14% fewer myocardial infarctions, and a 37% decrease in microvascular complications at the population level, the investigators noted.

"Further research is needed to identify which interventions and combinations of QI strategies will optimally improve important outcomes in patients with diabetes at an acceptable cost to aid health-system planning," they concluded.

This study was funded by the Ontario Ministry of Health and Long-term Care and the Alberta Heritage Foundation for Medical Research Interdisciplinary Team Grants program. The authors reported having no conflicts of interest.

Quality improvement strategies that aim to optimize systems of care have significant favorable effects on diabetes care outcomes, Andrea C. Tricco, Ph.D., of Li Ka Shing Knowledge Institute of St. Michael’s Hospital, Toronto, and her colleagues reported online June 9 in the Lancet.

While QI interventions targeting patients might be beneficial irrespective of baseline HbA_1c, those interventions targeting providers only seem beneficial when baseline HbA_1c is greater than 8.0%, according to the findings of their systematic review and meta-analysis, which together involved more than 123,000 patients.

The results were reported simultaneously at the annual meeting of the American Diabetes Association in Philadelphia.

Based on data from a total of 120 relevant trials, QI strategies reduced HbA_1c by a mean difference of 0.37%; and reduced LDL cholesterol by 0.10 mmol/L, systolic blood pressure by 3.13 mm Hg, and diastolic blood pressure by 1.55 mm Hg versus usual care over a median 12-month period, the authors found (Lancet 2012 June 9 [doi: 10.1016/S0140-6736(12)60480-2]).

Specifically, strategies targeting clinicians’ education and audit and feedback led to HbA_1c reductions of 0.33% and 0.44%, respectively, when baseline HbA_1c concentrations were greater than 8.0%, compared with no improvement when baseline HbA_1c was less than 8.0%.

Patient education seemed more effective than reminders when baseline HbA_1c was greater than 8.0%, but less effective when it was less than 8.0%.

Furthermore, QI strategies were associated with increases in aspirin use; use of any antihypertensive drugs; and screening for retinopathy, renal involvement, and foot complications over median follow-up periods ranging from 12 to 18 months. The QI strategies did not, however, have a significant effect on statin use, adequate control of hypertension, or smoking cessation rates over median follow-up periods ranging from 12 to 19 months.

In trials that enrolled patients with an HbA_1c of greater than 8.0%, declines of more than 0.5% were seen with certain QI strategies (specifically, those targeting team changes, case management, patient education, and promotion of self-management). In trials enrolling those with an HbA_1c of 8.0% or less, declines of that magnitude were seen for one QI strategy (facilitated relay).

Nonetheless, "after adjustment for median baseline HbA_1c values and effective sample size, the QI strategies were associated with significantly lower HbA_1c than usual care was," they said.

"We noted greater improvements in HbA_1c control for QI strategies targeting health systems and patients. In fact, all QI strategies were associated with significant changes in HbA_1c except for clinician education," they said.

The trials included in this review and meta-analysis were reported between July 2003 and July 2010, and were selected using 11 predefined QI strategies or financial incentives targeting health care professionals for the management of adult outpatients with diabetes.

"The QI strategies targeted health systems (e.g., team changes), professionals (e.g., professional reminders), or patients (e.g., promotion of self-management)," the investigators explained.

The findings, which represent an update of a previous review that assessed the effects of QI strategies on glycemic control, provide an assessment of the effects of QI strategies on a broader range of diabetic care.

The analysis was limited by the complexity of the QI strategies studied, the inability to control for all potential confounding factors, and the short duration of follow-up in many of the trials, the authors wrote.

Despite these limitations, the findings indicate that diabetes outcomes can be improved by QI interventions. Wide implementation of QI strategies could lead to important population benefits, as data have shown that a 1% reduction in mean HbA_1c results in 21% fewer deaths, 14% fewer myocardial infarctions, and a 37% decrease in microvascular complications at the population level, the investigators noted.

"Further research is needed to identify which interventions and combinations of QI strategies will optimally improve important outcomes in patients with diabetes at an acceptable cost to aid health-system planning," they concluded.

This study was funded by the Ontario Ministry of Health and Long-term Care and the Alberta Heritage Foundation for Medical Research Interdisciplinary Team Grants program. The authors reported having no conflicts of interest.

Quality improvement strategies that aim to optimize systems of care have significant favorable effects on diabetes care outcomes, Andrea C. Tricco, Ph.D., of Li Ka Shing Knowledge Institute of St. Michael’s Hospital, Toronto, and her colleagues reported online June 9 in the Lancet.

While QI interventions targeting patients might be beneficial irrespective of baseline HbA_1c, those interventions targeting providers only seem beneficial when baseline HbA_1c is greater than 8.0%, according to the findings of their systematic review and meta-analysis, which together involved more than 123,000 patients.

The results were reported simultaneously at the annual meeting of the American Diabetes Association in Philadelphia.

Based on data from a total of 120 relevant trials, QI strategies reduced HbA_1c by a mean difference of 0.37%; and reduced LDL cholesterol by 0.10 mmol/L, systolic blood pressure by 3.13 mm Hg, and diastolic blood pressure by 1.55 mm Hg versus usual care over a median 12-month period, the authors found (Lancet 2012 June 9 [doi: 10.1016/S0140-6736(12)60480-2]).

Specifically, strategies targeting clinicians’ education and audit and feedback led to HbA_1c reductions of 0.33% and 0.44%, respectively, when baseline HbA_1c concentrations were greater than 8.0%, compared with no improvement when baseline HbA_1c was less than 8.0%.

Patient education seemed more effective than reminders when baseline HbA_1c was greater than 8.0%, but less effective when it was less than 8.0%.

Furthermore, QI strategies were associated with increases in aspirin use; use of any antihypertensive drugs; and screening for retinopathy, renal involvement, and foot complications over median follow-up periods ranging from 12 to 18 months. The QI strategies did not, however, have a significant effect on statin use, adequate control of hypertension, or smoking cessation rates over median follow-up periods ranging from 12 to 19 months.

In trials that enrolled patients with an HbA_1c of greater than 8.0%, declines of more than 0.5% were seen with certain QI strategies (specifically, those targeting team changes, case management, patient education, and promotion of self-management). In trials enrolling those with an HbA_1c of 8.0% or less, declines of that magnitude were seen for one QI strategy (facilitated relay).

Nonetheless, "after adjustment for median baseline HbA_1c values and effective sample size, the QI strategies were associated with significantly lower HbA_1c than usual care was," they said.

"We noted greater improvements in HbA_1c control for QI strategies targeting health systems and patients. In fact, all QI strategies were associated with significant changes in HbA_1c except for clinician education," they said.

The trials included in this review and meta-analysis were reported between July 2003 and July 2010, and were selected using 11 predefined QI strategies or financial incentives targeting health care professionals for the management of adult outpatients with diabetes.

"The QI strategies targeted health systems (e.g., team changes), professionals (e.g., professional reminders), or patients (e.g., promotion of self-management)," the investigators explained.

The findings, which represent an update of a previous review that assessed the effects of QI strategies on glycemic control, provide an assessment of the effects of QI strategies on a broader range of diabetic care.

The analysis was limited by the complexity of the QI strategies studied, the inability to control for all potential confounding factors, and the short duration of follow-up in many of the trials, the authors wrote.

Despite these limitations, the findings indicate that diabetes outcomes can be improved by QI interventions. Wide implementation of QI strategies could lead to important population benefits, as data have shown that a 1% reduction in mean HbA_1c results in 21% fewer deaths, 14% fewer myocardial infarctions, and a 37% decrease in microvascular complications at the population level, the investigators noted.

"Further research is needed to identify which interventions and combinations of QI strategies will optimally improve important outcomes in patients with diabetes at an acceptable cost to aid health-system planning," they concluded.

This study was funded by the Ontario Ministry of Health and Long-term Care and the Alberta Heritage Foundation for Medical Research Interdisciplinary Team Grants program. The authors reported having no conflicts of interest.