EASD 2016

MUNICH – Brain changes suggestive of cerebral microvascular dysfunction are already apparent in patients with prediabetes.

The changes – increased white matter hyperintensities and decreased white matter volume – are even more pronounced in subjects with type 2 diabetes, Marnix van Agtmaal, MD, said at the annual meeting of the European Association for the Study of Diabetes. Patients with frank diabetes also showed an increase in intracranial cerebrospinal fluid – a correlate of the decrease in brain volume, said Dr. van Agtmaal of Maastricht (the Netherlands) University Medical Center.

The changes are probably caused by diabetes-related endothelial dysfunction, he said.

“The brain is highly dependent on properly functioning microcirculation. This is critical, since the brain has high energy demand and no energy reserve. In prediabetes and type 2 diabetes, microvascular endothelial dysfunction occurs. This leads to cerebral hypoperfusion, which in turns causes chronic ischemia. This contributes to small vessel disease leading to brain atrophy and, eventually, cognitive decline and dementia.”

The 2,251 subjects in the analysis were drawn from the Maastricht study, an ongoing observational study of people with type 2 diabetes.

Among the group, 350 had prediabetes, defined as impaired fasting glucose, impaired glucose tolerance, or a combination of the two. Type 2 diabetes was present in 528. The rest had healthy glucose metabolism.

As the cohort progressed from healthy glucose metabolism to prediabetes and then diabetes, they became older (aged 58 years in the healthy group vs. 62 years in the diabetes group), heavier, and displayed worsening cardiovascular risk factors, with increasing systolic blood pressure and progressively poorer lipid profiles. Kidney function was preserved in all patients, however.

The groups were not balanced in terms of sex: 56% of those with healthy glucose metabolism were women, compared with 47% of those with prediabetes and 31% of those with type 2 diabetes.

Dr. van Agtmaal and his colleagues examined white matter hyperintensities, white matter volume, gray matter volume, and intracranial CSF. They conducted three linear regression models: a crude unadjusted model, a partially adjusted model that controlled for age, sex, and intracranial volume; and a fully adjusted model that controlled for those factors, plus systolic blood pressure, lipids, smoking, kidney function, and education.

There was a clear linear association between white matter hyperintensity (WMH) volume and healthy glucose metabolism, prediabetes, and type 2 diabetes. In the crude analysis, the healthy subjects carried about 0.75 mL of WMH. Prediabetic subjects carried about 1.25 mL, and those with diabetes, about 2 mL.

In both the partially and fully adjusted models, this relationship was somewhat attenuated, but it remained significant for both prediabetes and diabetes.

The crude model also found that both diabetes groups had significantly lower white matter volume than did the healthy subjects. In healthy subjects, the mean volume was about 480 mL. This was about 467 mL in those with prediabetes and 466 mL in those with type 2 diabetes. Again, the partially and fully adjusted models slightly attenuated the relationship, but it remained significant in both disease states.

The crude model showed that gray matter was decreased in both prediabetes and type 2 diabetes. In healthy subjects, total gray matter was about 667 mL. In those with prediabetes, it was about 655 mL, and in those with type 2 diabetes, about 645 mL. However, the significant associations disappeared for both diabetes and prediabetes in both adjusted models.

Intracranial CSF was also different among the three groups in the crude model. In the healthy subjects, the total intracranial CSF volume was about 248 mL. In those with prediabetes, it was about 255 mL, and in those with type 2 diabetes, about 270 mL.

The association with prediabetes disappeared in the fully adjusted model – but for type 2 diabetes, it remained strongly significant.

Dr. van Agtmaal has not correlated the imaging findings with any cognitive testing on these subjects but said that study is coming.

“Further analysis will also look at cognitive decline and the development of dementia in the group,” he said. “We also intend to look at associations with other outcomes of cerebral dysfunction, including depression.”

Dr. van Agtmaal had no financial disclosures.

msullivan@frontlinemedcom.com

MUNICH – Brain changes suggestive of cerebral microvascular dysfunction are already apparent in patients with prediabetes.

The changes – increased white matter hyperintensities and decreased white matter volume – are even more pronounced in subjects with type 2 diabetes, Marnix van Agtmaal, MD, said at the annual meeting of the European Association for the Study of Diabetes. Patients with frank diabetes also showed an increase in intracranial cerebrospinal fluid – a correlate of the decrease in brain volume, said Dr. van Agtmaal of Maastricht (the Netherlands) University Medical Center.

The changes are probably caused by diabetes-related endothelial dysfunction, he said.

“The brain is highly dependent on properly functioning microcirculation. This is critical, since the brain has high energy demand and no energy reserve. In prediabetes and type 2 diabetes, microvascular endothelial dysfunction occurs. This leads to cerebral hypoperfusion, which in turns causes chronic ischemia. This contributes to small vessel disease leading to brain atrophy and, eventually, cognitive decline and dementia.”

The 2,251 subjects in the analysis were drawn from the Maastricht study, an ongoing observational study of people with type 2 diabetes.

Among the group, 350 had prediabetes, defined as impaired fasting glucose, impaired glucose tolerance, or a combination of the two. Type 2 diabetes was present in 528. The rest had healthy glucose metabolism.

As the cohort progressed from healthy glucose metabolism to prediabetes and then diabetes, they became older (aged 58 years in the healthy group vs. 62 years in the diabetes group), heavier, and displayed worsening cardiovascular risk factors, with increasing systolic blood pressure and progressively poorer lipid profiles. Kidney function was preserved in all patients, however.

The groups were not balanced in terms of sex: 56% of those with healthy glucose metabolism were women, compared with 47% of those with prediabetes and 31% of those with type 2 diabetes.

Dr. van Agtmaal and his colleagues examined white matter hyperintensities, white matter volume, gray matter volume, and intracranial CSF. They conducted three linear regression models: a crude unadjusted model, a partially adjusted model that controlled for age, sex, and intracranial volume; and a fully adjusted model that controlled for those factors, plus systolic blood pressure, lipids, smoking, kidney function, and education.

There was a clear linear association between white matter hyperintensity (WMH) volume and healthy glucose metabolism, prediabetes, and type 2 diabetes. In the crude analysis, the healthy subjects carried about 0.75 mL of WMH. Prediabetic subjects carried about 1.25 mL, and those with diabetes, about 2 mL.

In both the partially and fully adjusted models, this relationship was somewhat attenuated, but it remained significant for both prediabetes and diabetes.

The crude model also found that both diabetes groups had significantly lower white matter volume than did the healthy subjects. In healthy subjects, the mean volume was about 480 mL. This was about 467 mL in those with prediabetes and 466 mL in those with type 2 diabetes. Again, the partially and fully adjusted models slightly attenuated the relationship, but it remained significant in both disease states.

The crude model showed that gray matter was decreased in both prediabetes and type 2 diabetes. In healthy subjects, total gray matter was about 667 mL. In those with prediabetes, it was about 655 mL, and in those with type 2 diabetes, about 645 mL. However, the significant associations disappeared for both diabetes and prediabetes in both adjusted models.

Intracranial CSF was also different among the three groups in the crude model. In the healthy subjects, the total intracranial CSF volume was about 248 mL. In those with prediabetes, it was about 255 mL, and in those with type 2 diabetes, about 270 mL.

The association with prediabetes disappeared in the fully adjusted model – but for type 2 diabetes, it remained strongly significant.

Dr. van Agtmaal has not correlated the imaging findings with any cognitive testing on these subjects but said that study is coming.

“Further analysis will also look at cognitive decline and the development of dementia in the group,” he said. “We also intend to look at associations with other outcomes of cerebral dysfunction, including depression.”

Dr. van Agtmaal had no financial disclosures.

msullivan@frontlinemedcom.com

MUNICH – Brain changes suggestive of cerebral microvascular dysfunction are already apparent in patients with prediabetes.

The changes – increased white matter hyperintensities and decreased white matter volume – are even more pronounced in subjects with type 2 diabetes, Marnix van Agtmaal, MD, said at the annual meeting of the European Association for the Study of Diabetes. Patients with frank diabetes also showed an increase in intracranial cerebrospinal fluid – a correlate of the decrease in brain volume, said Dr. van Agtmaal of Maastricht (the Netherlands) University Medical Center.

The changes are probably caused by diabetes-related endothelial dysfunction, he said.

“The brain is highly dependent on properly functioning microcirculation. This is critical, since the brain has high energy demand and no energy reserve. In prediabetes and type 2 diabetes, microvascular endothelial dysfunction occurs. This leads to cerebral hypoperfusion, which in turns causes chronic ischemia. This contributes to small vessel disease leading to brain atrophy and, eventually, cognitive decline and dementia.”

The 2,251 subjects in the analysis were drawn from the Maastricht study, an ongoing observational study of people with type 2 diabetes.

Among the group, 350 had prediabetes, defined as impaired fasting glucose, impaired glucose tolerance, or a combination of the two. Type 2 diabetes was present in 528. The rest had healthy glucose metabolism.

As the cohort progressed from healthy glucose metabolism to prediabetes and then diabetes, they became older (aged 58 years in the healthy group vs. 62 years in the diabetes group), heavier, and displayed worsening cardiovascular risk factors, with increasing systolic blood pressure and progressively poorer lipid profiles. Kidney function was preserved in all patients, however.

The groups were not balanced in terms of sex: 56% of those with healthy glucose metabolism were women, compared with 47% of those with prediabetes and 31% of those with type 2 diabetes.

Dr. van Agtmaal and his colleagues examined white matter hyperintensities, white matter volume, gray matter volume, and intracranial CSF. They conducted three linear regression models: a crude unadjusted model, a partially adjusted model that controlled for age, sex, and intracranial volume; and a fully adjusted model that controlled for those factors, plus systolic blood pressure, lipids, smoking, kidney function, and education.

There was a clear linear association between white matter hyperintensity (WMH) volume and healthy glucose metabolism, prediabetes, and type 2 diabetes. In the crude analysis, the healthy subjects carried about 0.75 mL of WMH. Prediabetic subjects carried about 1.25 mL, and those with diabetes, about 2 mL.

In both the partially and fully adjusted models, this relationship was somewhat attenuated, but it remained significant for both prediabetes and diabetes.

The crude model also found that both diabetes groups had significantly lower white matter volume than did the healthy subjects. In healthy subjects, the mean volume was about 480 mL. This was about 467 mL in those with prediabetes and 466 mL in those with type 2 diabetes. Again, the partially and fully adjusted models slightly attenuated the relationship, but it remained significant in both disease states.

The crude model showed that gray matter was decreased in both prediabetes and type 2 diabetes. In healthy subjects, total gray matter was about 667 mL. In those with prediabetes, it was about 655 mL, and in those with type 2 diabetes, about 645 mL. However, the significant associations disappeared for both diabetes and prediabetes in both adjusted models.

Intracranial CSF was also different among the three groups in the crude model. In the healthy subjects, the total intracranial CSF volume was about 248 mL. In those with prediabetes, it was about 255 mL, and in those with type 2 diabetes, about 270 mL.

The association with prediabetes disappeared in the fully adjusted model – but for type 2 diabetes, it remained strongly significant.

Dr. van Agtmaal has not correlated the imaging findings with any cognitive testing on these subjects but said that study is coming.

“Further analysis will also look at cognitive decline and the development of dementia in the group,” he said. “We also intend to look at associations with other outcomes of cerebral dysfunction, including depression.”

Dr. van Agtmaal had no financial disclosures.

msullivan@frontlinemedcom.com

MUNICH – A four-component imaging/biomarker screen was highly accurate for identifying silent myocardial infarction among asymptomatic patients with type 2 diabetes.

The screen is far more accurate than the current standards of invasive imaging or only looking for pathologic Q waves, Peter Swoboda, MD, said at the annual meeting of the European Association for the Study of Diabetes.

“By combining these four factors we came up with a tool that has a diagnostic area under the curve [AUC] of 0.85,” said Dr. Swoboda of Leeds (England) University. “This is far better than the 0.58 AUC that we have with Q waves only – a sensitivity of just 25%.”

The study was published online in June in the Journal of Clinical Endocrinology and Metabolism (JCEM 2016. doi: 10.1210/jc.2016-1318).

The screen employs noninvasive imaging and biomarkers to tap multiple clinical hallmarks of silent MI. The components are:

• electrocardiogram.

• echocardiography.

• biomarker assessment.

• cardiac magnetic resonance imaging, focusing on left ventricular ejection fraction and late gadolinium enhancement.

The study cohort comprised 100 patients with type 2 diabetes without known heart disease and with no new cardiac symptoms. They underwent cardiac MRI, a 12-lead electrocardiogram, echocardiography, and serum biomarker assessment. Late gadolinium enhancement identified evidence of silent MI in 17 patients (17%).

There were few differences in the clinical characteristics of those who had experienced MI and those who had not, Dr. Swoboda noted. There were no differences at all in diabetes-related measures, including disease duration or hemoglobin A_1c levels. Blood pressures were similar. Patients with MI were significantly older (65 vs. 60 years).

In cardiac-specific measures, left ventricular ejection fraction was similar, as was left ventricular mass, end diastolic volume, and left atrial volume. There were however, other very important differences, Dr. Swoboda noted.

Imaging included a measure called “feature tracking analysis,” which measured the peak global longitudinal strain, systolic strain rate, and early and late diastolic strain rates during contraction. This analysis noted a significant difference in global longitudinal strain between the MI and non-MI groups.

Ventricular filling velocities as measured by the E/A ratio on ECG were also significantly different between the MI and non-MI groups (0.75 vs. 0.89, respectively). ECG also found pathologic Q waves in significantly more MI patients (24% vs. 7%).

Finally, the serum biomarker panel showed a very strong increase in B-type natriuretic peptide (NT-proBNP) among MI patients, compared with non-MI patients (105 vs. 52 ng/L). There were no significant differences in the other biomarkers, including C-reactive protein and high-sensitive cardiac troponin.

Dr. Swoboda and his team then compiled these findings into a composite measure, assigning them optimum cutoff measures:

• Age older than 62 years.

• E/A ratio 0.72 or lower.

• Global longitudinal strain of at least 18.4%.

• NT-proBNP more than 29 ng/L.

The system resulted in a diagnostic accuracy AUC of 0.85 – significantly better than any of the AUCs generated by the individual components. All patients who scored 0 or 1 were free of MI. Among the 28 with a score of 2, only three had experienced a silent MI. Among the 21 with a score of 3, seven had experienced a silent MI and 14 had not. Among the 16 with a score of 4, seven had experienced a silent MI and nine had not.

While Dr. Swoboda called the screening method “simple” during discussion, a colleague in the audience disagreed with that.

“A simple test is something like a blood test only, not an MRI. Not imaging. That is expensive and takes time,” said Naveed Sattar, MD, of the University of Glasgow, Scotland. “However, I do think your data add more to the evidence that BNP can be a really valuable marker of cardiovascular risk in patients with diabetes.”

Dr. Sattar recently examined the value of cardiac serum biomarkers in predicting cardiovascular disease and mortality in nearly 100,000 people without a history of heart disease. In these subjects, he wrote, “NT-proBNP assessment strongly predicted first-onset heart failure and augmented coronary heart disease and stroke prediction, suggesting that NT-proBNP concentration assessment could be used to integrate heart failure into cardiovascular disease primary prevention.”

The paper appeared online in Lancet Diabetes in September (Lancet Diab. 2016. doi: 10.1016/S2213-8587[16]30196-6).

Dr. Swoboda agreed that data continue to support the increased use of NT-proBNP as a marker of heart disease.

“I think that in the future, diabetes medicine is moving toward individualized patient care, based on individualized risk factors. The future of assessing asymptomatic cardiac patients might be a combination of BNP and MRI.”

Dr. Swoboda had no financial disclosures. Some of Dr. Sattar’s coauthors reported relationships with pharmaceutical companies.

msullivan@frontlinemedcom.com

MUNICH – A four-component imaging/biomarker screen was highly accurate for identifying silent myocardial infarction among asymptomatic patients with type 2 diabetes.

The screen is far more accurate than the current standards of invasive imaging or only looking for pathologic Q waves, Peter Swoboda, MD, said at the annual meeting of the European Association for the Study of Diabetes.

“By combining these four factors we came up with a tool that has a diagnostic area under the curve [AUC] of 0.85,” said Dr. Swoboda of Leeds (England) University. “This is far better than the 0.58 AUC that we have with Q waves only – a sensitivity of just 25%.”

The study was published online in June in the Journal of Clinical Endocrinology and Metabolism (JCEM 2016. doi: 10.1210/jc.2016-1318).

The screen employs noninvasive imaging and biomarkers to tap multiple clinical hallmarks of silent MI. The components are:

• electrocardiogram.

• echocardiography.

• biomarker assessment.

• cardiac magnetic resonance imaging, focusing on left ventricular ejection fraction and late gadolinium enhancement.

The study cohort comprised 100 patients with type 2 diabetes without known heart disease and with no new cardiac symptoms. They underwent cardiac MRI, a 12-lead electrocardiogram, echocardiography, and serum biomarker assessment. Late gadolinium enhancement identified evidence of silent MI in 17 patients (17%).

There were few differences in the clinical characteristics of those who had experienced MI and those who had not, Dr. Swoboda noted. There were no differences at all in diabetes-related measures, including disease duration or hemoglobin A_1c levels. Blood pressures were similar. Patients with MI were significantly older (65 vs. 60 years).

In cardiac-specific measures, left ventricular ejection fraction was similar, as was left ventricular mass, end diastolic volume, and left atrial volume. There were however, other very important differences, Dr. Swoboda noted.

Imaging included a measure called “feature tracking analysis,” which measured the peak global longitudinal strain, systolic strain rate, and early and late diastolic strain rates during contraction. This analysis noted a significant difference in global longitudinal strain between the MI and non-MI groups.

Ventricular filling velocities as measured by the E/A ratio on ECG were also significantly different between the MI and non-MI groups (0.75 vs. 0.89, respectively). ECG also found pathologic Q waves in significantly more MI patients (24% vs. 7%).

Finally, the serum biomarker panel showed a very strong increase in B-type natriuretic peptide (NT-proBNP) among MI patients, compared with non-MI patients (105 vs. 52 ng/L). There were no significant differences in the other biomarkers, including C-reactive protein and high-sensitive cardiac troponin.

Dr. Swoboda and his team then compiled these findings into a composite measure, assigning them optimum cutoff measures:

• Age older than 62 years.

• E/A ratio 0.72 or lower.

• Global longitudinal strain of at least 18.4%.

• NT-proBNP more than 29 ng/L.

The system resulted in a diagnostic accuracy AUC of 0.85 – significantly better than any of the AUCs generated by the individual components. All patients who scored 0 or 1 were free of MI. Among the 28 with a score of 2, only three had experienced a silent MI. Among the 21 with a score of 3, seven had experienced a silent MI and 14 had not. Among the 16 with a score of 4, seven had experienced a silent MI and nine had not.

While Dr. Swoboda called the screening method “simple” during discussion, a colleague in the audience disagreed with that.

“A simple test is something like a blood test only, not an MRI. Not imaging. That is expensive and takes time,” said Naveed Sattar, MD, of the University of Glasgow, Scotland. “However, I do think your data add more to the evidence that BNP can be a really valuable marker of cardiovascular risk in patients with diabetes.”

Dr. Sattar recently examined the value of cardiac serum biomarkers in predicting cardiovascular disease and mortality in nearly 100,000 people without a history of heart disease. In these subjects, he wrote, “NT-proBNP assessment strongly predicted first-onset heart failure and augmented coronary heart disease and stroke prediction, suggesting that NT-proBNP concentration assessment could be used to integrate heart failure into cardiovascular disease primary prevention.”

The paper appeared online in Lancet Diabetes in September (Lancet Diab. 2016. doi: 10.1016/S2213-8587[16]30196-6).

Dr. Swoboda agreed that data continue to support the increased use of NT-proBNP as a marker of heart disease.

“I think that in the future, diabetes medicine is moving toward individualized patient care, based on individualized risk factors. The future of assessing asymptomatic cardiac patients might be a combination of BNP and MRI.”

Dr. Swoboda had no financial disclosures. Some of Dr. Sattar’s coauthors reported relationships with pharmaceutical companies.

msullivan@frontlinemedcom.com

MUNICH – A four-component imaging/biomarker screen was highly accurate for identifying silent myocardial infarction among asymptomatic patients with type 2 diabetes.

The screen is far more accurate than the current standards of invasive imaging or only looking for pathologic Q waves, Peter Swoboda, MD, said at the annual meeting of the European Association for the Study of Diabetes.

“By combining these four factors we came up with a tool that has a diagnostic area under the curve [AUC] of 0.85,” said Dr. Swoboda of Leeds (England) University. “This is far better than the 0.58 AUC that we have with Q waves only – a sensitivity of just 25%.”

The study was published online in June in the Journal of Clinical Endocrinology and Metabolism (JCEM 2016. doi: 10.1210/jc.2016-1318).

The screen employs noninvasive imaging and biomarkers to tap multiple clinical hallmarks of silent MI. The components are:

• electrocardiogram.

• echocardiography.

• biomarker assessment.

• cardiac magnetic resonance imaging, focusing on left ventricular ejection fraction and late gadolinium enhancement.

The study cohort comprised 100 patients with type 2 diabetes without known heart disease and with no new cardiac symptoms. They underwent cardiac MRI, a 12-lead electrocardiogram, echocardiography, and serum biomarker assessment. Late gadolinium enhancement identified evidence of silent MI in 17 patients (17%).

There were few differences in the clinical characteristics of those who had experienced MI and those who had not, Dr. Swoboda noted. There were no differences at all in diabetes-related measures, including disease duration or hemoglobin A_1c levels. Blood pressures were similar. Patients with MI were significantly older (65 vs. 60 years).

In cardiac-specific measures, left ventricular ejection fraction was similar, as was left ventricular mass, end diastolic volume, and left atrial volume. There were however, other very important differences, Dr. Swoboda noted.

Imaging included a measure called “feature tracking analysis,” which measured the peak global longitudinal strain, systolic strain rate, and early and late diastolic strain rates during contraction. This analysis noted a significant difference in global longitudinal strain between the MI and non-MI groups.

Ventricular filling velocities as measured by the E/A ratio on ECG were also significantly different between the MI and non-MI groups (0.75 vs. 0.89, respectively). ECG also found pathologic Q waves in significantly more MI patients (24% vs. 7%).

Finally, the serum biomarker panel showed a very strong increase in B-type natriuretic peptide (NT-proBNP) among MI patients, compared with non-MI patients (105 vs. 52 ng/L). There were no significant differences in the other biomarkers, including C-reactive protein and high-sensitive cardiac troponin.

Dr. Swoboda and his team then compiled these findings into a composite measure, assigning them optimum cutoff measures:

• Age older than 62 years.

• E/A ratio 0.72 or lower.

• Global longitudinal strain of at least 18.4%.

• NT-proBNP more than 29 ng/L.

The system resulted in a diagnostic accuracy AUC of 0.85 – significantly better than any of the AUCs generated by the individual components. All patients who scored 0 or 1 were free of MI. Among the 28 with a score of 2, only three had experienced a silent MI. Among the 21 with a score of 3, seven had experienced a silent MI and 14 had not. Among the 16 with a score of 4, seven had experienced a silent MI and nine had not.

While Dr. Swoboda called the screening method “simple” during discussion, a colleague in the audience disagreed with that.

“A simple test is something like a blood test only, not an MRI. Not imaging. That is expensive and takes time,” said Naveed Sattar, MD, of the University of Glasgow, Scotland. “However, I do think your data add more to the evidence that BNP can be a really valuable marker of cardiovascular risk in patients with diabetes.”

Dr. Sattar recently examined the value of cardiac serum biomarkers in predicting cardiovascular disease and mortality in nearly 100,000 people without a history of heart disease. In these subjects, he wrote, “NT-proBNP assessment strongly predicted first-onset heart failure and augmented coronary heart disease and stroke prediction, suggesting that NT-proBNP concentration assessment could be used to integrate heart failure into cardiovascular disease primary prevention.”

The paper appeared online in Lancet Diabetes in September (Lancet Diab. 2016. doi: 10.1016/S2213-8587[16]30196-6).

Dr. Swoboda agreed that data continue to support the increased use of NT-proBNP as a marker of heart disease.

“I think that in the future, diabetes medicine is moving toward individualized patient care, based on individualized risk factors. The future of assessing asymptomatic cardiac patients might be a combination of BNP and MRI.”

Dr. Swoboda had no financial disclosures. Some of Dr. Sattar’s coauthors reported relationships with pharmaceutical companies.

msullivan@frontlinemedcom.com