NPH or glargine insulin, with fasting blood glucose titrated to 100 â105 mg/dl. After 9 months, A1C levels were 7.1â. 7.2%, despite achievement of the fasting.
Reviews/Commentaries/ADA Statements P E R S P E C T I V E S
O N
T H E
N E W S
Achieving Glycemic Goals in Type 2 Diabetes ZACHARY T. BLOOMGARDEN, MD
T
his is the fourth in a series of articles on presentations at the American Diabetes Association’s 66th Scientific Sessions, Washington, DC, 9 –13 June 2006, addressing aspects of the treatment of type 2 diabetes. A pilot project is underway to offer the Perspectives on the News commentaries as a monthly Web-based CME activity. Please access www.diabetes.procampus. net to view our initial efforts. We look forward to your comments. Achieving glycemic goals in type 2 diabetes There is increasing evidence that the majority of individuals with diabetes are not treated to currently recommended glycemic goals. Rubino et al. (abstract 324) analyzed a U.K. database of 31,289 individuals with type 2 diabetes, of whom 2,501 had not received insulin and had A1C ⬎8% despite treatment with two oral hypoglycemic agents (abstract numbers refer to the ADA Scientific Sessions, Diabetes 55 [Suppl. 1], 2006). After 1.8 years, 25% had started insulin; 4.9 years elapsed before half had started insulin, with similar lack of rapidity of treatment intensification among those diagnosed with retinopathy (18%) or neuropathy (7%). With A1C thresholds of 7 and 9%, half had started insulin at 6.3 and 4.2 years, respectively. Similarly, Nichols et al. (abstract 535) identified 4,365 individuals from the Kaiser Permanente Northwest pharmacy records starting sulfonylurea plus metformin, with a mean A1C 8.4%, and 79 and 51% of patients achieving A1C ⬍8% and ⬍7%, respectively. Sixty-six percent of the former group rose to ⬎8% (after a mean of 16 months), and 75% of the latter group rose to ⬎7% (mean of 11 months), with 73% not receiving insulin over 33 months of follow-up. Those whose A1C initially fell
below 7% ultimately receiving insulin had mean A1C 9.2% when this treatment was added, at a mean of 28 months. Riedel and Plauschinat (abstracts 552 and 553) assessed treatment patterns of 9,416 type 2 diabetic individuals initiated on oral agents (mean A1C 8.4%), the twothirds whose baseline A1C was ⱖ7% with mean A1C 9.5%. An additional treatment was added at 8 months, with mean A1C 8.5%, although half of those receiving an additional oral agent did not have repeat A1C testing, a factor which the authors speculate might contribute to delayed therapy intensification. Only 31% of those receiving an oral hypoglycemic agent achieved A1C ⬍7%, and 39% of these individuals subsequently had A1C rising to ⱖ7% over a mean of 18 months. Comparing 5,453 individuals started on metformin, 2,373 with a sulfonylurea, and 1,590 with a thiazolidinedione (TZD), treatment failure was more likely with a sulfonylurea than with metformin, with a trend to lowest treatment failure rates with a TZD. Considerations for use of insulin in treatment of type 2 diabetes In context of these findings, a symposium on approaches to treatment of type 2 diabetic individuals failing to achieve adequate glycemic control with oral agents gave a number of important insights. Matthew Riddle, Portland, Oregon, discussed the use of basal insulin, noting that a patient with an A1C of ⬃9% will have a 24-h glucose pattern with a relatively constant increase in glucose levels through the day from the pattern exhibited by a diabetic patient under better control, so that at least half of the patient’s “glycemic burden” will reflect increased basal glycemia. This person’s treatment, Riddle concluded, requires provision of basal insulin, lowering fasting glucose levels,
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Zachary T. Bloomgarden, MD, is a practicing endocrinologist in New York, New York, and is affiliated with the Division of Endocrinology, Mount Sinai School of Medicine, New York, New York. Abbreviations: ALT, alanine aminotransferase; CT, computed tomography; FFA, free fatty acid; HOMA, homeostasis model assessment; IL, interleukin; NAFLD, nonalcoholic fatty liver disease; PPAR, peroxisome proliferator–activated receptor; TZD, thiazolidinedione. DOI: 10.2337/dc07-zb01 © 2007 by the American Diabetes Association.
174
and consequently downward shifting the overall pattern of hyperglycemia. Less optimal, from this perspective, would be a premixed insulin, while neither prandial insulin nor exenatide, which lowers postprandial glycemia in particular, would be appropriate alone in this formulation. Portal insulin is the main regulator of glucose production, but systemic insulin, by suppressing free fatty acid (FFA) release, further indirectly reduces glucose production. Injected insulin augments both portal and systemic insulin, effectively suppressing basal glucose overproduction. Further, Riddle stated that replacing basal insulin reduces hypertriglyceridemia, improves HDL cholesterol, improves vasodilation, suppresses inflammatory markers, improves myocardial metabolism under stress, and reduces mortality in the intensive care unit. Endotheliumdependent vasodilation improves after starting basal insulin (1) and plasma markers of inflammation are suppressed, particularly in individuals with both diabetes and metabolic syndrome. In individuals having coronary artery bypass graft surgery, infusion of glucose, insulin, and potassium decreases plasma FFA and -hydroxybutyrate, improving myocardial function (2), a finding that Riddle suggested could be extrapolated to longer-term effects of basal insulin administration. Riddle described a trial comparing the basal insulins glargine and NPH, terming it a “proof of principle” study of the concept that aggressive algorithmbased approaches can effectively lower glycemia (3). Enrolled type 2 diabetic individuals inadequately controlled on oral agents were set a 100 mg/dl fasting glucose goal. Of 367 individuals randomized to glargine and 389 to NPH, 90% received a sulfonylurea, 83% metformin, and 10% a TZD. Enrolled individuals received 2, 4, 6, or 8 units insulin at bedtime, initially based on fasting glucose 100 –120, 120 – 140, 140 –180, or ⬎180 mg/dl, respectively, with titration every 2 days based on the mean glucose using the same schedule of increments. A glycemic plateau began to be seen between 8 and 12 weeks. Little severe hypoglycemia was seen, but mildto-moderate hypoglycemia did occur, with nocturnal glucose (documented DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
Bloomgarden ⱕ72 mg/dl) in 27% of those receiving insulin glargine and in 33% of those receiving NPH. Riddle commented that postprandial glucose increments limit achievement of A1C ⬍7% with this approach, citing a similarly designed Finnish study of metformin-treated individuals randomized to the addition of NPH or glargine insulin, with fasting blood glucose titrated to 100 –105 mg/dl. After 9 months, A1C levels were 7.1– 7.2%, despite achievement of the fasting glucose goal, in association with elevated glucose levels after breakfast and dinner (4). Potential approaches for further normalization of glycemia would include switching to multiple doses of a premixed insulin, adding exenatide, or adding premeal rapid-acting insulin, with Riddle suggesting the latter to be the best approach. He reviewed a study comparing lispro 75/25 with human NPH/regular 70/30, without oral agents, in which A1C averaged 8% and fasting glucose exceeded 150 mg/dl. Hypoglycemia was particularly seen before lunch, suggesting that the premixed insulin regimens may not be desirable for achieving close glycemic control. In a study comparing premixed 70/30 twice daily without oral agents with glargine plus oral agents, the latter led to better glycemia, although with waning of glycemic control in the late afternoon (5). Although a comparison of metformin combined with aspart 70/30 twice daily versus glargine once daily showed the former to be associated with improved glycemia, there was a glucose nadir with the premixed insulin in the late morning, and this group had greater weight gain (6); Riddle interpreted to suggest that this approach results in “defensive eating” at lunch. Riddle noted that both insulin detemir and NPH twice daily with oral agents should be considered, suggesting that these approaches lead to similar fasting glucose and somewhat better A1C than that seen with either glargine or NPH once daily (7). He concluded that using treat-to-target basal insulin is a promising initial approach but may not suffice for achieving optimal glycemia, that premixed insulin also has undesirable characteristics, that exenatide plus basal insulin is promising, and that the use of a basal-bolus insulin approach may be ideal, although one must take into account the intensive physician-patient interaction required for such treatment. Food intake varies greatly, he noted, suggesting that a potential approach is to titrate to optimal fasting glucose with basal DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
insulin and then add a rapid-acting analog before the main meal of the day, titrating to a subsequent preprandial (or bedtime) glucose of ⬃120 mg/dl. He suggested that such an approach allows the type 2 diabetes patient with A1C 9% on oral agents to achieve a fasting glucose of 105 mg/dl and an A1C of 6.6%. Philip Raskin, Dallas, Texas, gave a somewhat different view of approaches with mixed insulin, suggesting that this is a clinically useful approach. He reviewed the treatment goals of A1C ⬍6.5 to 7%, noting that because of progressive -cell failure, oral hypoglycemic agents are not completely effective for many patients, leading to the need for insulin treatment. Raskin suggested that the “right way to treat diabetes is combination therapy with insulin and oral agents” to improve insulin sensitivity. He described a study of 43 individuals with insulin-treated type 2 diabetes receiving metformin or placebo and ⬎50 units insulin/day, with an A1C goal ⬍5.6% (8). With insulin alone, A1C decreased from 9.1 to 7.5%, while those patients receiving insulin with metformin showed a fall from 9 to 6.5%. A1C ⬍7% was achieved by 41 versus 59%, with 54% of those receiving combination treatment reaching an A1C ⬍6.5%. Furthermore, the metformin plus insulin group required a less complicated treatment schedule, with many of these patients controlled with human 70/30 insulin twice daily. Raskin pointed out that there was no weight gain in the metformin group, while those receiving insulin alone gained 3 kg. In a study of 28 individuals receiving insulin with metformin, or insulin with troglitazone, and subsequently given both sensitizers with continued insulin administration, Raskin showed evidence suggesting that the TZD combination with insulin more effectively lowered glucose levels than that with metformin, with administration of both sensitizers with insulin leading to still better A1C levels (9). He noted that although the TZD was more potent in glucose lowering, the approach best minimizing weight gain was to first give insulin with metformin and then to add troglitazone. Insulin is, of course, needed for individuals presenting with major hyperglycemia or for individuals with hyperglycemia despite lifestyle and maximal oral agent treatment. Raskin advised giving two daily injections of 70/30 insulin, usually with metformin, with subsequent adjustment for insulin dose/type/number of injections and the addition of a TZD to be an
optimal approach. With this treatment in his clinic, A1C decreased from 10.6 to 5.8% with what he characterized as modest weight gain. Raskin recommended initiating insulin at 0.4 – 0.7 units 䡠 kg⫺1 䡠 day⫺1 and continuing metformin, and often TZDs, but discontinuing secretagogues and ␣-glucosidase inhibitors. He recommended starting with either bedtime glargine or NPH, or two daily injections of mixed insulin, changing the insulin dosing and type as required. He referred to one of the studies reviewed by Riddle, of 233 individuals with type 2 diabetes having A1C ⱖ8% and BMI ⱕ40 kg/m2, receiving at least 1,500 mg metformin daily, many also receiving 30 mg pioglitazone daily (6). He showed that those receiving the mixed insulin aspart 70/30 had a 2.8% reduction in A1C, while those receiving glargine had a 2.4% reduction. Those whose baseline A1C was at least 8.5% showed a 3.1 versus 2.6% reduction, while those with baseline A1C ⬍8% had a decrement of 1.4% regardless of whether glargine or aspart 70/30 was administered. The mixed insulin was particularly effective in reducing postprandial glucose levels, although Raskin acknowledged that more weight was gained, and hypoglycemia occurred more frequently with mixed insulin treatment. A similar study showed superior glucose lowering with premixed lispro 75/25 insulin twice daily to that with glargine in metformin-treated individuals (10). Raskin also reviewed a study in which aspart 70/30 was used starting with one dose, then advancing to two doses, and finally to three doses daily, as required for glycemic control, showing that 41, 70, and 77% achieved A1C ⬍7%, respectively, and that 21, 52, and 60% achieved A1C ⬍6.5%, respectively (11), leading to his conclusion that “two shots are better than one and three shots are better than two.” Raskin stressed that insulin should not be used as a threat and that “it’s not their fault” if a patient requires insulin; thus, one should address the understandable anxiety and sense of personal failure experienced by many individuals with diabetes who do require this treatment. Fear of hypoglycemia, fear of injections, and weight gain are additional issues, but one should emphasize the benefit of improved glycemia in reducing the risk of complications. He suggested that “70/30 is more than just a compromise. It is safe, and very effective.” Robert Ratner, Washington, DC, began a discussion of the potential benefit of 175
Perspectives on the News exenatide by agreeing, “Insulin is a good drug. It works.” There are, however, issues. Many individuals are often psychologically resistant to this treatment, and many physicians feel that initiation of insulin treatment requires overly great effort. There are issues of weight gain, hypoglycemia, inconvenience, and cost, both of the treatment itself and of the more intensive glucose monitoring required. Ratner discussed the phenomenon of physician delay in initiating or intensifying pharmacologic treatment in individuals with diabetes. There is a 3-year delay before initiation of an oral agent treatment, with a mean A1C of 8.8%, followed by a delay of ⱖ3 years to achieve an A1C of 9.4% before addition of a second oral agent. Again, physicians typically wait for the A1C to approach 9% before adding insulin, resulting in a further 3-year delay. Intensification of treatment with oral agents is usually only of modest benefit. In an observational report of triple oral therapy, A1C stabilized at 8% (12). A randomized controlled trial of triple oral agent therapy reported a reduction in A1C from 9.7 to 8.5% (13), leading Ratner to comment, “starting triple oral agent therapy above nine isn’t going to get you where you want to go.” At lower levels of glycemia, addition of pioglitazone to a sulfonylurea plus metformin allows glycemic control comparable with that of adding an evening dose of NPH (14). Similarly, addition of the sulfonylurea glimepiride in individuals receiving metformin with a TZD decreases A1C (15). The question, Ratner asked, is whether it is “worth it” to add a TZD in metformin-sulfonylurea failures, rather than adding glargine, and it appears to have been answered in the affirmative. He reviewed a comparison of the two approaches showing similar A1C reduction, although neither led to a fall in mean A1C below 7% (16). Similarly, a study comparing the combination of metformin with premixed insulin versus triple oral agent therapy showed that “insulin brought the A1C down more quickly but at six months . . . it was no better” (17). He concluded that “the barriers to insulin therapy may actually be reasonable,” but that “the difficulty is the natural history of the disease.” He cited an observational study showing that half of triple oral agent–treated individuals fail to maintain glycemic control over a 6-year period (18), leading Ratner to question, “Can we get any better?” He reviewed a study of exenatide versus placebo in type 2 dia176
betic individuals receiving a sulfonylurea plus metformin (19). Those receiving exenatide had an initial A1C of 8.5%, decreasing 0.5–1.0%, and a body weight of ⬃100 kg, decreasing by 2 kg. Hypoglycemia did occur, although with progressively decreasing frequency over time as the sulfonylurea dose was reduced, suggesting a strategy of decreasing sulfonylurea to the minimal effective dose on initiation of exenatide treatment. Ratner noted that ongoing studies will address the use of exenatide with other combinations. Asking why not use insulin only, Ratner emphasized that “we still have a problem” with hypoglycemia. In a study comparing glargine with exenatide in type 2 diabetic individuals receiving oral agents, there was similar achievement of A1C ⬍7%, but a 2-kg weight gain versus a 2- to 2.5-kg loss of weight, leading Ratner to suggest that the combination of less weight gain and less hypoglycemia makes exenatide “the logical next step,” offering a simpler, equally durable, safe approach to that with insulin, while leading to minimal hypoglycemia and to weight loss. He concluded, “This is the last great hope of changing the natural history of type 2 diabetes.” Aspects of insulin resistance Hayashi et al. (abstract 299) followed 407 Japanese Americans not receiving diabetes treatment over 10 –11 years, finding that computed tomography (CT)scan measurement of intra-abdominal fat area showed significant correlation with subsequent insulin resistance based on homeostasis model assessment (HOMA-IR, while subcutaneous abdominal, thoracic, and thigh fat areas did not correlate in multivariate analysis. Similarly, Fox et al. (abstract 300) reported greater association of CT volumetric measures of visceral than subcutaneous adipose tissue with diabetes risk and significant effect of visceral adiposity after adjusting for BMI and waist circumference in women but not in men among 1,189 participants from the Framingham Offspring cohort. Of course, a great deal more than total fat measures may be gleaned from imaging studies, with Matsuda et al. (abstract 940) finding that pancreatic CT density was 43% lower among 69 type 2 diabetic individuals receiving insulin than among 63 diet only– treated patients, suggesting pancreatic fat accumulation contributing to -cell damage. Li et al. (abstract 931) analyzed 7,922 and 2,933 nondiabetic adults from Na-
tional Health and Nutrition Examination Surveys in 1988 –1994 and in 1999 – 2002, respectively, finding a 20% increase in HOMA-IR, with a 32.3% increase in the prevalence of insulin resistance based on the 75th percentile of HOMA in the initial study. In an interesting report, Kolonics et al. (abstract 602) performed hyperinsulinemic-euglycemic clamp studies in 7 individuals with normal insulin sensitivity, 17 with glucose intolerance and insulin resistance, and 7 with type 2 diabetes. Mitochondrial number was 40 and 52% as great and nitric oxide production 8 and 28% as great in T lymphocytes isolated from the insulinresistant and diabetic groups, respectively (compared with the insulin-sensitive group), suggesting this to be a method that could be used in distinguishing individuals with and without insulin resistance. Type 2 diabetes treatment by cytokine modulation Several new approaches are being developed to improve glycemic treatment of type 2 diabetes. The proimflammatory cytokine interleukin (IL)-1 induces rodent and human -cell apoptosis in vitro, and type 2 diabetes is associated with increased islet IL-1 expression. Thomas R. Mandrup-Poulsen, Gentofte, Denmark, described a study of the effect of the IL-1 receptor antagonist Anakinra (Kineret) in 64 type 2 diabetic individuals. After 13 weeks, those patients with fasting glucose ⬎144 or A1C ⬎8% were also given insulin to intensify treatment. A1C decreased 0.4% with treatment, while increasing 0.1% in a placebo group, with the 120min postmeal plasma glucose increasing 1 mmol/l with placebo and decreasing 0.6 mmol/l with active treatment. Insulin sensitivity decreased in the placebo group, while remaining stable in those receiving the IL-1 antagonist, and the insulin secretory meal response decreased with placebo and increased with treatment, suggesting that the intervention had dual effects on insulin resistance and on insulin secretion. Tesauro et al. (abstract 74) found that the vasodilatory response to nitroprusside and to acetylcholine, which increases endogenous nitric oxide release from vascular endothelium, was impaired in 12 individuals with versus 12 without metabolic syndrome and that intraarterial administration of infliximab, a tumor necrosis factor-␣–neutralizing antibody, increased this vasodilatory effect, suggesting a mechanism of the endothelial dysfunction associated with insulin DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
Bloomgarden resistance. In an interesting study of immune insulin resistance, Nugaram et al. (abstract 605) treated an anti-nuclear antibody–positive 18-year-old man with anti-insulin receptor antibody–mediated diabetes, who failed to respond to metformin, pioglitazone, and up to 500 units daily insulin, with four weekly 375 mg/m2 intravenous doses of the chimeric murine/human monoclonal antibody against CD20 molecule on B lymphocytes Rituximab, allowing resolution of the syndrome. Morino et al. (abstract 6-LB) induced heat-shock protein-72, by combined mild electrical stimulation and hyperthermia in both high-fat fed and db/db mice, finding a 20% reduction in fasting glucose, a 38% decrease in fasting insulin, and a doubling of serum adiponectin and of uncoupling protein-1 mRNA expression in brown adipose tissue and with a 34 and 44% respective reduction in visceral and subcutaneous fat, reduction in adipocyte size, and improvement in fatty liver, suggesting a potential therapeutic approach for insulin-resistant states including type 2 diabetes and the metabolic syndrome. In a related study, Kolonics et al. (abstract 10-LB) studied BGP-15 (N-Gene Research Laboratories), a hydroxylamine derivative that increases heat-shock proteins and restores constitutive nitric oxide synthase activity in hyperglycemia, and was found to double insulin-stimulated glucose uptake in two animal models. In 42 nondiabetic individuals with insulin resistance given the agent for 28 days, whole-body glucose utilization similarly increased 1.6 –1.75 mg 䡠 kg⫺1 䡠 min⫺1. In vitro, nitric oxide synthase and mitochondrial function improved, suggesting this to be a candidate pharmacologic insulin sensitizer. Insulin sensitizers A number of studies addressed issues of insulin resistance and treatment approaches to improve insulin sensitivity. Von Eynatten et al. (abstract 13-LB) reported that adipocyte-specific fatty acid– binding protein, which is increased in obese rodents in adipocytes and macrophages and appears to protect against atherosclerosis in knockout models, was elevated in individuals with diabetes and coronary artery disease, correlating with reduced insulin sensitivity and with A1C, triglycerides, C-reactive protein, and low HDL cholesterol; this suggests that it may be a useful biomarker of insulin resistance. Rendell and McGettigan (abstract DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
606) studied 568 type 2 diabetic men, 44% with testosterone ⬍300 mg/dl and 54% with testosterone ⱖ300 mg/dl, the latter having 50% higher HOMA of insulin sensitivity, leading the authors to speculate that androgen deficiency might contribute to insulin resistance, and suggesting that testosterone might be considered another potential biomarker. In an interesting study that may relate to the development of heart failure with TZD treatment, Son et al. (abstract 700) studied a mouse model with increased heart-specific PPAR (peroxisome proliferator–activated receptor)␥1 expression, showing no effect on systemic metabolic parameters (weight, glucose, triglyceride, or fatty acid levels) at 8 months but increased cardiac PPAR␥ target genes, including lipoprotein lipase and carnitine palmitoyltransferase-1, with increases in cardiac weight and triglyceride levels in association with development of dilated cardiomyopathy. There remain, however, intriguing studies suggesting an antiatherosclerotic benefit of these agents. Kubota et al. (abstract 138) studied mice with and without deletion of the adiponectin gene, showing that neointimal formation in response to arterial injury was increased in mice not expressing adiponectin and that pioglitazone reduced the neointimal response in both sets of animals, decreasing both inflammation and the proliferation of vascular smooth muscle, although with a delay in those lacking the ability to produce adiponectin in response to TZD. Katayama et al. (abstract 705) randomized 28 nondiabetic individuals with metabolic syndrome who had coronary stenting to 30 mg pioglitazone daily versus placebo, showing reduction in intimal area on intravascular ultrasound at 6 months, with 0 versus 31% developing restenosis. Zhao et al. (abstract 449), noting that glycosphingolipids can modulate the activity of the insulin receptor, studied the glucosylceramide synthase inhibitor Genz-123346, noting improved glycemia and preservation of insulin secretory capacity in type 2 diabetic rodent models, with increased muscle insulin signaling. Dhalla et al. (abstract 458) administered CVT-3619, a partial A1 adenosine receptor agonist that inhibits adipose tissue lipolysis and lowers circulating FFA levels, to Zucker diabetic fatty rats and showed a 68% reduction in FFA levels, with reduction in fasting glucose from 263 to 140 mg/dl, insulin from 3.8 to 1.3 ng/ml, and triglyceride from 309 to 89 mg/dl, with-
out effect on pulse or blood pressure. Donner (abstract 461) administered 15 g of the bulk hexose sweetener D-tagatose three times daily with meals to eight type 2 diabetic individuals for 14 months, finding a 4-kg weight loss and glycohemoglobin falling from 11.2 to 9.5%. Kraegen et al. (abstract 603) administered berberine, a natural plant product used in traditional Chinese medicine, in rodent type 2 diabetic models, showing improvement in glucose and in insulin sensitivity. In myocytes, AMPK phosphorylation doubled with increased GLUT4 translocation. Yin et al. (abstract 584) incubated myocytes and adipocytes with berberine, showing an increase in glucose uptake associated with induction of GLUT4 protein. Yamada et al. (abstract 581) described studies with a non-TZD insulin sensitizer, K-111 (Kowa Company). Of 80 previously untreated type 2 diabetic individuals receiving 0, 5, 10, or 20 mg daily, fasting glucose decreased 22 mg/dl at the highest dosage, with reduction in triglyceride at all doses, perhaps reflecting the PPAR␣ agonist activity of the compound. Nieuwdorp et al. (abstract 701) compared effects of metformin alone, fenofibrate alone, and the combination in 681 individuals, of whom 369 had fasting glucose ⱖ126 mg/dl, finding trends to improvement in insulin sensitivity (based on fasting glucose and insulin measurement) and reduction in fasting and postload glucose with the combination. Takata et al. (abstract 14-LB) studied LDL receptor knockout rats, finding that angiotensin infusion increased macrophages and vascular smooth muscle cell PPAR␦ expression in atherosclerotic lesions but that the PPAR␦ agonist GW0724 reduced angiotensin-induced atherosclerosis by two-thirds and decreased expression of inflammatory markers including osteopontin, tumor necrosis factor-␣, and IL-6 in the lesions, also increasing HDL cholesterol and reducing triglycerides and FFAs. Guha et al. (abstract 55-LB) studied a different PPAR␦ agonist, KD3010, in high-fat fed mice, finding decreased visceral adiposity, adipocyte hypertrophy, hepatic steatosis, triglycerides, and FFA levels, along with improvement in insulin sensitivity and attenuation of weight gain. In a clinical study, Schaefer et al. (abstract 479) administered a non-TZD insulin sensitizer with PPAR␣ activity, K-111 (KOWA Pharmaceuticals), to 199 type 2 diabetic individuals at doses of 0, 5, 10, 20, and 40 mg daily, showing improvement in glyce177
Perspectives on the News mia at daily doses ⱖ10 mg, with reductions in insulin, C-peptide, triglyceride, and cholesterol without change in body weight. TZDs A number of presentations pertained to TZD treatment. Triscari et al. (abstracts 570 and 571) studied the new agent, rivoglitazone. In 50 healthy subjects, adiponectin was unchanged with placebo but increased 1.7-, 3.3-, 5.3-, and 4.6fold at 14 days in those receiving 1-, 2.5-, 5-, and 10-mg doses, without change in insulin or glucose levels. In 426 type 2 diabetic patients, placebo-adjusted fasting glucose levels decreased 13, 39, and 57 mg/dl after a 6-week period of treatment with 0.5, 2, and 5 mg rivoglitazone, while decreasing 19 mg/dl in patients receiving 30 mg pioglitazone daily. Adiponectin increased 5.1, 16.8, and 26.0 g/ml with the three rivoglitazone dosages and 7.6 g/ml with pioglitazone. Scho¨ndorf et al. (abstract 609) studied 675 individuals treated with pioglitazone, finding doses of 15, 30, and 45 mg daily to be associated with 2-, 2.5-, and 3-fold respective increases in adiponectin. Kang et al. (abstract 499) found that polymorphisms of the gene for perilipin, an adipocyte-specific phosphoprotein associated with the periphery of lipid storage droplets, required for lipid storage and fatty acid release (20), correlated with weight gain in 160 type 2 diabetic individuals treated with rosiglitazone for 24 weeks; those with the AA genotype of the 1148G⬎A polymorphism did not show weight gain, while GG and GA genotypes gained 1.3 and 0.9 kg, respectively. Another genetic polymorphism, the Pro12Ala variant of the PPAR␥2 gene, is associated with decreased PPAR␥ activity. Of 326 type 2 diabetic individuals, Vestergaard et al. (abstract 321) showed that 78 with the Pro12Ala genotype had less benefit from treatment with pioglitazone than the 248 individuals with the Pro12Pro genotype. Such assessment may allow better selection of candidates for a TZD. Uemura et al. (abstract 573) measured body water and fat mass in 47 type 2 diabetic individuals treated with 45 mg pioglitazone versus placebo for 8 weeks, showing that significant weight gain and increase in body fat could be demonstrated at 4 weeks but that increase in total body water began at 2 weeks, a dissociation between the two effects. Miyazaki et al. (abstract 529) treated 20 type 2 diabetic individuals with 8 mg rosiglitazone 178
daily for 16 weeks and found a reduction in fasting glucose from 185 to 139 mg/dl, in A1C from 8.5 to 7.1%, and in FFA from 789 to 656 Eq/l, with increases in adiponectin from 6.2 to 18.2 g/ml and a 33% increase in insulin-mediated total body glucose disposal; this correlated with both improved insulin-stimulated insulin receptor substrate-1 tyrosine phosphorylation and with adiponectin. Krzyzanowska et al. (abstract 508) treated eight healthy men with 8 mg rosiglitazone daily versus placebo for 21 days, with adiponectin increasing from 5.9 to 14.3 g/ml and FFAs decreasing from 377 to 188 mol/l, while no effect on either parameter was seen in eight men given placebo. With infusion of heparin and a lipid emulsion on day 21, FFA levels increased to a lesser extent and adiponectin further increased to 18.1 g/ml with rosiglitazone. Insulin sensitivity decreased similarly in both groups, however, suggesting that the TZD-induced increase in adiponectin may not explain the insulin sensitizing effects of these agents. TZDs may allow more sustained glycemic control than seen with sulfonylureas. Spanheimer et al. (abstract 320) treated 2,120 individuals with pioglitazone versus glyburide (with metformin if needed) for 3 years, with insulin added for A1C levels ⬎7.5%. Baseline A1C was 9.5%, and glycemic control was comparable in the two groups for the 1st year of study, but a significantly greater decline was seen in A1C in the pioglitazone group from 72 to 156 weeks by 2–2.5 versus 1.5–2% in those randomized to glyburide. Markolf et al. (abstract 604) randomized 500 metformin-treated type 2 diabetic individuals to 30 mg pioglitazone versus 3.5 mg glyburide daily, finding that A1C decreased from a baseline of 8.5% by 1% vs. 0.6%, with 22 vs. 55% requiring insulin over the subsequent 3.5 years for annual progression rates to insulin of 7 vs. 16%. Oerter et al. (abstract 539) reported a 96-week study of pioglitazone versus glyburide in 173 versus 206 metformin-treated individuals in 58 “routine care” outpatient clinics. A1C decreased from a baseline of 7.8% by 0.8 vs. 0.6%, and fasting glucose from a baseline of 164 mg/dl by 18 versus 9 mg/dl, with medication costs of 1,436 versus 687, but total treatment costs of 2,448 versus 2,163, as the thiazolidinedione was associated with lower admission rates. The potency of TZDs is not, however, sufficient for all individuals with poor glycemic control. Davidson et al. (abstract 569)
reported an observational study suggesting similar effects of adding rosiglitazone and pioglitazone to a maximal-tolerated dose of metformin plus sulfonylurea treatment. At 4 months, mean A1C fell from 9.3 to 7.5 with rosiglitazone and from 9.5 to 7.4 with pioglitazone, including 65 and 62%, respectively, achieving an A1C ⱕ7.5%. Among responders, only 61 and 62%, respectively, maintained an A1c ⱕ7.5% at 1 year. Kupfer et al. (abstract 509) reported a safety analysis of a 3-year study of pioglitazone versus glyburide in 1,051 versus 1,046 individuals, with metformin and then insulin added if needed, finding that 1.2 versus 0% discontinued treatment because of edema and 0.7 versus 0.2% because of weight gain; however, 0.2 versus 1.9% discontinued because of hypoglycemia. Myocardial infarction occurred in 6 versus 11, congestive heart failure in 13 versus 12, edema in 84 versus 36, and hypoglycemia in 40 versus 119, respectively. New type 2 diabetes treatment approaches Bile acid sequestrants may lower glucose as well as LDL cholesterol levels, in individuals with diabetes. Kalin et al. (abstract 498), Schwarz et al. (abstract 563), and Zieve et al. (abstract 589) administered 3.75 g colesevelam daily versus placebo to 59 type 2 diabetic individuals, finding a placebo-adjusted reduction in fasting glucose of 23 and 18 mg/dl at 4 and 8 weeks, respectively. At 12 weeks, fasting glucose had fallen from 170 to 165 mg/dl, and the glucose level 1 h following a sucrosecontaining meal (Ensure) decreased from 269 to 251 mg/dl with colesevelam, while with placebo fasting glucose increased from 188 to 190 mg/dl and postprandial glucose increased from 285 to 288 mg/dl. There was a 0.2% reduction in A1C versus a 0.3% increase among individuals receiving placebo from the 8% baseline. LDL cholesterol fell from 123 to 108 mg/dl with colesevelam, whereas it increased from 120 to 122 with placebo. Kawabata et al. (abstract 502) administered the similar anion-exchange resin colestimide, 3 g daily, to 27 type 2 diabetic individuals, finding a decrease in A1C from 7.7 to 6.8%, in fasting glucose from 164 to 152 mg/dl, and LDL cholesterol from 130 to 103 mg/dl. Triscari et al. (abstract 444) administered CS-917, a fructose 1,6-bisphosphatase inhibitor. Fasting glucose decreased in 39 type 2 diabetic patients DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
Bloomgarden treated with 50, 100, 200, and 400 mg daily for 14 days, although elevation in lactic acid levels occurred at the 400-mg dose. Okuno et al. (abstract 540) and Yoshida et al. (abstract 585) reported that CS-917, but not metformin, reduced gluconeogenesis from lactate and pyruvate in rat hepatocytes, with in vivo tritiated glucose turnover studies showing the agent to suppress glucose production ⬎75%, while metformin increased glucose utilization in association with an increase in glucose production. In Zucker diabetic fatty rats, metformin had a greater effect on glucose levels in the fed state, while CS-917 led to greater reduction in fasting glucose, and the combination effectively reducing A1C. Van Poelje et al. (abstract 575) demonstrated that CS-917 reduced both renal and hepatic gluconeogenesis in vitro. In the perfused kidney, gluconeogenic precursors glycerol 3-phosphate and fructose 1,6-bisphosphate were increased 4.1- and 5.8-fold, respectively, while fructose 6-phosphate and glucose levels fell ⬎50%. Nonalcoholic steatohepatitis An important therapeutic challenge will be the development of approaches to treatment for nonalcoholic fatty liver disease (NAFLD) (21). Guo et al. (abstract 904) studied 616 nondiabetic Hispanic Americans from 155 families with hypertensive probands, finding 54, 47, and 81% heritabilities of alkaline phosphatase, aspartate aminotransferase, and alanine aminotransferase (ALT), respectively, and showing significant genetic correlation with insulin sensitivity measured by hyperinsulinemic-euglycemic clamp, having correlation coefficients of ⫺0.35, ⫺0.57, and ⫺0.47, respectively. Tsuchiya et al. (abstract 572) randomized 39 individuals with NAFLD and type 2 diabetes or impaired glucose tolerance to 270 mg nateglinide or 0.6 mg voglibose daily, finding a similar reduction in glycemia, but an increase in liver CT density from 51 to 58 Hounsfield units with nateglinide, suggesting reduction in hepatic fat, without change in this parameter in the voglibose-treated group. There is growing evidence that TZDs may improve NAFLD. In analysis of 2097 individuals receiving pioglitazone versus glyburide for 36 weeks, Spanheimer et al. (abstract 322) found that 2.5 versus 7.2% had ALT levels ⬎1.5-fold above the upper limit of normal, with a mean 6 IU/l decrease versus a 2 IU/l increase in ALT levels. Zib et al. (abstract 75) randomized 51 DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
individuals with type 2 diabetes and baseline A1C 10.7% to either insulin alone or in combination with pioglitazone and found a 3% fall in A1C in both groups. Hepatic triglyceride levels and abdominal adiposity measured with proton magnetic resonance spectroscopy decreased 55% with combination treatment, while not changing with insulin alone. Ratziu et al. (abstract 84) reported a study of 63 biopsy-proven patients with nonalcoholic steatohepatitis randomized to 8 mg rosiglitazone daily versus placebo, with normalization of transaminase levels seen in 38 versus 6%, respectively. Repeat biopsy at 1 year showed a decrease in steatosis of at least 30% in 15 versus 5 patients, although this degree of improvement in steatosis was only seen in 1 of 9 rosiglitazone-treated individuals with type 2 diabetes, while occurring in 14 of 23 who did not have diabetes. Balas et al. (abstract 319) and Belfort et al. (abstract 439) reported that 48 individuals with type 2 diabetes or impaired glucose tolerance and biopsyproven nonalcoholic steatohepatitis randomized to 45 mg pioglitazone daily versus placebo for 6 months had improvement in transaminases and decreased by half in steatosis and inflammation. Both studies reported increases in insulin sensitivity to be associated with improvement in liver disease. PROactive analyses The PROactive (PROspective pioglitAzone Clinical Trial In macroVascular Events) study of 5,238 individuals with type 2 diabetes and cardiovascular disease treated with 45 mg pioglitazone daily versus placebo for a mean of 34.5 months was reported at the European Association for the Study of Diabetes meeting in September 2005 (22). A number of fascinating subanalyses of this study were presented at the American Diabetes Association meeting. Charbonnel and Scheen (abstract 448) discussed the long-term glycemic effects of pioglitazone versus placebo in the 1,314 individuals receiving metformin plus sulfonylurea at the beginning of the study. A1C decreased from baseline of 8% to 7.2% with pioglitazone, while decreasing to 7.8% in the placebo group, and 16% vs. 31% required additional insulin. Weight increased 4.1 kg versus decreasing 0.7 kg, and edema occurred in 29 versus 17%. The same authors (abstract 561) described characteristics of the 1,760 individuals in PROactive receiving insulin at study on-
set, finding insulin doses of 47 units daily in both groups at baseline and at study end 42 versus 55 units daily. A1C decreased 0.9% in individuals receiving pioglitazone, while decreasing 0.5% in those receiving placebo. Edema developed in 31 versus 18% and hypoglycemia in 41 versus 29%. Massi-Benedetti et al. (abstract 523) reported that of the 3,478 study participants not taking insulin at baseline, 183 versus 362 required insulin at study end. Increased frequency of edema and hypoglycemia with pioglitazone were again reported in this subset. Heine et al. (abstract 484) reported that 9.7 versus 10.7% of those receiving pioglitazone versus placebo had elevated ALT at baseline, with 5.9 versus 11.9% showing this abnormality at study end, suggesting an effect on NAFLD. Wilcox and Kupfer (abstract 317) presented further analysis of the effect of pioglitazone on major adverse cardiovascular events, showing that 13% of individuals receiving pioglitazone versus 16% of those receiving placebo had death, nonfatal nonsilent myocardial infarction, nonfatal stroke, or acute coronary syndrome and that 4 versus 5% had a myocardial infarction. Of 486 versus 498 individuals in the study who had a stroke before randomization, Wilcox et al. (abstract 579) reported in a prespecified analysis that stroke occurred during follow-up in 5.6% of those receiving pioglitazone versus 10.2% of those receiving placebo, a 47% reduction. Freemantle et al. (abstract 471) used a simulation model to replicate the ⬃10% cardiovascular disease benefit shown in the PROactive study, commenting that the study was underpowered, and that ⬎18,000 individuals would be required to demonstrate such an effect with statistical significance. Of course, as separation between the groups did not begin to be seen until ⬃18 months of the study, with mean follow-up of 34.5 months, a similar argument could be used to suggest that a 6-year study with the same number of participants would have demonstrated benefit. It is intriguing that if a principle end point similar to those described by Wilcox et al. had been chosen, the study would have been regarded as confirmedly positive.
References 1. Vehkavaara S, Yki-Jarvinen H: 3.5 years of insulin therapy with insulin glargine improves in vivo endothelial function in type 2 diabetes. Arterioscler Thromb Vasc 179
Perspectives on the News Biol 24:325–330, 2004 2. Szabo Z, Arnqvist H, Hakanson E, Jorfeldt L, Svedjeholm R: Effects of high-dose glucose-insulin-potassium on myocardial metabolism after coronary surgery in patients with type II diabetes. Clin Sci 101: 37– 43, 2001 3. Riddle MC, Rosenstock J, Gerich J, the Insulin Glargine 4002 Study Investigators: The treat-to-target trial: randomized addition of glargine or human NPH insulin to oral therapy of type 2 diabetic patients. Diabetes Care 26:3080 –3086, 2003 4. Yki-Jarvinen H, Kauppinen-Makelin R, Tiikkainen M, Vahatalo M, Virtamo H, Nikkila K, Tulokas T, Hulme S, Hardy K, McNulty S, Hanninen J, Levanen H, Lahdenpera S, Lehtonen R, Ryysy L: Insulin glargine or NPH combined with metformin in type 2 diabetes: the LANMET study. Diabetologia 49:442– 451, 2006 5. Janka HU, Plewe G, Riddle MC, KliebeFrisch C, Schweitzer MA, Yki-Jarvinen H: Comparison of basal insulin added to oral agents versus twice-daily premixed insulin as initial insulin therapy for type 2 diabetes. Diabetes Care 28:254 –259, 2005 6. Raskin P, Allen E, Hollander P, Lewin A, Gabbay RA, Hu P, Bode B, Garber A, the INITIATE Study Group: Initiating insulin therapy in type 2 diabetes: a comparison of biphasic and basal insulin analogs. Diabetes Care 28:260 –265, 2005 7. Hermansen K, Davies M, Derezinski T, Martinez Ravn G, Clauson P, Home P: A 26-week, randomized, parallel, treat-totarget trial comparing insulin detemir with NPH insulin as add-on therapy to oral glucose-lowering drugs in insulin-naive people with type 2 diabetes. Diabetes Care 29:1269 –1274, 2006 8. Aviles-Santa L, Sinding J, Raskin P: Effects of metformin in patients with poorly controlled, insulin-treated type 2 diabetes mellitus: a randomized, double-blind, placebo-controlled trial. Ann Intern Med 131:182–188, 1999 9. Strowig SM, Aviles-Santa ML, Raskin P:
180
10.
11.
12. 13.
14.
15.
16.
Improved glycemic control without weight gain using triple therapy in type 2 diabetes. Diabetes Care 27:1577–1583, 2004 Malone JK, Bai S, Campaigne BN, Reviriego J, Augendre-Ferrante B: Twicedaily pre-mixed insulin rather than basal insulin therapy alone results in better overall glycaemic control in patients with type 2 diabetes. Diabet Med 22:374 –381, 2005 Garber AJ, Wahlen J, Wahl T, Bressler P, Braceras R, Allen E, Jain R: Attainment of glycaemic goals in type 2 diabetes with once-, twice-, or thrice-daily dosing with biphasic insulin aspart 70/30 (The 1–2-3 study). Diabetes Obes Metab 8:58 – 66, 2006 Ovalle F, Bell DS: Triple oral antidiabetic therapy in type 2 diabetes mellitus. Endocr Pract 4:146 –147, 1998 Yale JF, Valiquett TR, Ghazzi MN, Owens-Grillo JK, Whitcomb RW, Foyt HL: The effect of a thiazolidinedione drug, troglitazone, on glycemia in patients with type 2 diabetes mellitus poorly controlled with sulfonylurea and metformin: a multicenter, randomized, double-blind, placebo-controlled trial. Ann Intern Med 134:737–745, 2001 Aljabri K, Kozak SE, Thompson DM: Addition of pioglitazone or bedtime insulin to maximal doses of sulfonylurea and metformin in type 2 diabetes patients with poor glucose control: a prospective, randomized trial. Am J Med 116:230 – 235, 2004 Roberts VL, Stewart J, Issa M, Lake B, Melis R: Triple therapy with glimepiride in patients with type 2 diabetes mellitus inadequately controlled by metformin and a thiazolidinedione: results of a 30-week, randomized, double-blind, placebo-controlled, parallel-group study. Clin Ther 27:1535–1547, 2005 Rosenstock J, Sugimoto D, Strange P, Stewart JA, Soltes-Rak E, Dailey G: Triple therapy in type 2 diabetes: insulin glargine or rosiglitazone added to combi-
17.
18.
19.
20.
21.
22.
nation therapy of sulfonylurea plus metformin in insulin-naive patients. Diabetes Care 29:554 –559, 2006 Schwartz S, Sievers R, Strange P, Lyness WH, Hollander P, the INS-2061 Study Team: Insulin 70/30 mix plus metformin versus triple oral therapy in the treatment of type 2 diabetes after failure of two oral drugs: efficacy, safety, and cost analysis. Diabetes Care 26:2238 –2243, 2003 Bell DS, Ovalle F: Long-term glycaemic efficacy and weight changes associated with thiazolidinediones when added at an advanced stage of type 2 diabetes. Diabetes Obes Metab 8:110 –115, 2006 Kendall DM, Riddle MC, Rosenstock J, Zhuang D, Kim DD, Fineman MS, Baron AD: Effects of exenatide (exendin-4) on glycemic control over 30 weeks in patients with type 2 diabetes treated with metformin and a sulfonylurea. Diabetes Care 28:1083–1091, 2005 Tansey JT, Sztalryd C, Hlavin EM, Kimmel AR, Londos C: The central role of perilipin a in lipid metabolism and adipocyte lipolysis. IUBMB Life 56:379 –385, 2004 Bloomgarden ZT: Update on the thiazolidinediones [article online]. Medscape. Available from http://www.medscape. com/viewarticle/536503. Accessed 26 October 2006 Dormandy JA, Charbonnel B, Eckland DJ, Erdmann E, Massi-Benedetti M, Moules IK, Skene AM, Tan MH, Lefebvre PJ, Murray GD, Standl E, Wilcox RG, Wilhelmsen L, Betteridge J, Birkeland K, Golay A, Heine RJ, Koranyi L, Laakso M, Mokan M, Norkus A, Pirags V, Podar T, Scheen A, Scherbaum W, Schernthaner G, Schmitz O, Skrha J, Smith U, Taton J, the PROactive Investigators: Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet 366:1279 –1289, 2005
DIABETES CARE, VOLUME 30, NUMBER 1, JANUARY 2007
