Effect of glyburide on hepatic glucose metabolism

Effect of glyburide on hepatic glucose metabolism

Effect of Glyburide Metabolism on Hepatic Glucose OWEN P. MCGUINNESS, PH.D.,ALAND. CHERRINGTON, PH.D.Nashvi//e, Glyburide, along with the other sec...

1MB Sizes 1 Downloads 15 Views

Effect of Glyburide Metabolism

on Hepatic Glucose

OWEN P. MCGUINNESS, PH.D.,ALAND. CHERRINGTON, PH.D.Nashvi//e,

Glyburide, along with the other second-generation oral hypoglycemic agent glipizide, has been used as adjunctive therapy for the treatment of non-insulin-dependent diabetes mellitus. After glyburide therapy, basal glycemia and glucose response to a meal are greatly improved. The mechanism for the glucose-lowering effect of the drug remains controversial. Glyburide is generally thought to exert two major actions: stimulation of pancreatic insulin secretion and enhancement of insulin action in hepatic and extrahepatic tissues. Studies in patients with noninsulin-dependent diabetes mellitus indicate that the action of glyburide on the liver plays a central role in decreasing glucose. With short-term therapy, glyburide decreases hepatic glucose production by elevating pancreatic insulin secretion. However, this increase in pancreatic insulin secretion is not sustained as therapy is continued, suggesting that glyburide then acts directly on the liver. The mechanism for the improvement in hepatic glucose metabolism after long-term treatment is not known. In u&o, glyburide has been shown to inhibit gluconeogenesis as well as glycogenolysis and to enhance hepatic glucose uptake, thus providing possible explanations for the action of the drug on the liver.

From the Department of Molecular Physiology and Biophysics, Vanderbilt University School of Medicine, Nashville, Tennessee. Requests for reprints should be addressed to Alan D. Cherriygton, Ph.D., Department of Molecular Physiplogy and Biophysics, $$zrbilt University School of Medicine, 702 Light Hall, Nashvrlle, Tennessee 37232.

2A-26s

August 20, 1990

The American Journal of Medicine

Tennessee

ral hypoglycemic drugs and diet have been the 0 mainstays in the management of non-insulindependent diabetes mellitus (NIDDM). The mechanisms by which oral hypoglycemic agents improve glucose control in NIDDM, however, are not completely understood. Glyburide (glibenclamide), a secondgeneration sulfonylurea, has been used extensively in the treatment of NIDDM for a number of years. Initial studies suggested that its major site of action is the pancreas. After short-term glyburide therapy (1 day to 2 weeks), basal insulin levels are elevated and the insulin response to a meal is increased, leading to marked reductions in both preprandial and postprandial glycemia. After more prolonged therapy (6 months to 2 years), basal insulin levels and the insulin response to a meal are similar to, or only slightly greater than, those observed before treatment. Nevertheless, the associated glucose levels remain markedly improved, suggesting that glyburide also acts at an extrapancreatic site. In principle, the drug could act by decreasing glucose production and increasing glucose utilization. Because insulin regulates both of these processes, it has been suggested that glyburide may cause an increase in insulin sensitivity and/or responsiveness in the liver and the periphery (muscle and adipose tissue). Many studies have confirmed that glyburide therapy increases peripheral insulin action. In general, glyburide is thought to do so, not by affecting peripheral insulin sensitivity (ED&, but rather by increasing the responsiveness of fat and muscle to insulin. However, improvement in peripheral insulin action alone cannot explain the enhancement in glucose control observed with glyburide therapy. In the basal state, the majority of glucose utilization occurs in insulin-independent tissues (e.g., the brain and the hemopoietic system), not in insulinsensitive muscle and adipose tissue. An improvement in peripheral insulin responsiveness would then have minimal impact on the fasting hyperglycemia of patients with NIDDM, who characteristically have elevated hepatic glucose production (HGP) and impaired glucose removal. Suppression of HGP, on the other hand, could lead to a marked improvement in fasting glycemia. An understanding of the role that the hepatic effects of glyburide play in the overall action of the drug on glucose homeostasis requires direct examination of its effects on glucose production and uptake by the liver. In this regard, it becomes important to examine the known regulators of hepatic glucose output in both normal subjects and patients with NIDDM. In addition, the validity of the methods used to assess hepatic insulin action in z&o must be evaluated. INSULIN AND GLUCOSE PRODUCTION IN NORMAL PERSONS Insulin is known to be a potent inhibitor of HGP in humans as well as in a variety of other animal species.

Volume 89 (suppl 2A)

SYMPOSIUMON GLYBURIOE/ McGUlNNESSand CHERRINGTON

Studies conducted in the past decade using tracer and arteriovenous difference techniques have attempted to characterize as precisely as possible the relationship between insulin and glucose. Under euglycemic conditions, the infusion of even a small amount of insulin through a peripheral vein has been shown to cause tracer-determined glucose production (RJ to decrease rapidly and substantially (Table I). The conclusion drawn from these findings is that insulin levels of approximately 30 pU/mL are half maximal for R, suppression [1,2]. However, the accuracy of this estimate must be questioned. Overestimation of the effectiveness of insulin can be attributed to three factors: (1) After low rates of infusion of insulin, there is a less complete suppression of endogenous insulin secretion than when higher infusion rates are used, leading to a more dramatic underestimation of portal insulin concentrations calculated from arterial insulin levels. If portal insulin levels had been measured directly and used in the data analysis, the dose-response curves would have been shifted to the right (Figure 1). (2) The infusion of insulin results in a suppression of glucagon secretion [3-51. Furthermore, subtle variations in glucagon can have significant effects on R, even under hyperinsulinemic conditions [6]. Therefore, the reductions in glucagon that occur during insulin infusion augment the ability of insulin to reduce R,, thus increasing the apparent sensitivity of the liver to insulin. Had the glucagon level been fixed at a basal value in the studies discussed previously, the dose-response curve would again have been shifted to the right. (3) The use of [3H]3-glucose to estimate endogenous R, during a euglycemichyperinsulinemic clamp results in additional inaccuracies [‘7-121. The exact cause of such inaccuracies is debatable (e.g., method of modeling, tracer contaminant, isotope discrimination), but two things are clear: (1) Total R, is underestimated using this approach, particularly during non-steady-state conditions; and (2) in turn, the time required for the effect of insulin on the liver to stabilize is also underestimated. The use of purified tracers and more sophisticated models for non-steady-state data analysis (which is now possi-

TABLEI Effect of Hyperinsulinemia (peripheral delivery) on R, under Euglycemic Conditions in hlale Subjects Who Fasted Overnight InsulinInfusionRate (mu/kg/minute)

Arterial Insulin &t/mL)

R, (mg/kg/minute)

Rizza eta/[11

DeFronzo eta/[2]*

652 10 zi

i:!

0.25 0.5 1.0 5.0 10.0

0.0 2.3 0.7 0.3 0.1 0.0 0.0

101 428 1189

*Data adapted with permission from [2] and the American Diabetes Association.

ble) would again have caused a shift to the right in the dose-response curve relating insulin to R, suppression. Offsetting the above problems, which have caused an overestimation of the sensitivity of the liver to insulin, is a factor which led to an underestimation of the action of insulin. To construct a complete doseresponse curve relating insulin to glucose production, it is necessary to determine the ability of basal levels of insulin to initially inhibit R, and of increments in insulin to further inhibit it. The ability of basal insulin to inhibit R, has been determined in humans 1131and in dogs [14]. In general, R, doubles within 30 minutes of selective insulin removal (i.e., that brought about by somatostatin infusion and basal glucagon replacement). Based on these data, the inhibitory action of insulin has probably been underestimated because it takes a prolonged period for the effects of insulin deficiency to become fully manifest and because the ensuing hyperglycemia blunts the increase in R,. Nevertheless, because the increase in R, that occurs in response to acute insulin deficiency exceeds the basal R, (and therefore the maximal insulin-induced suppres-

r

.. \ . . ..

\ L. *. ... . \ . ... \ ‘. . . .. . \ \

\

hepatic insulin sensitivity decreased .... .... normal increased . ...

....

\

Figure 1. Dose-dependent

insulin suppression of HGP. Estimation of hepatic insulin sensitivity and the potential alterations in hepatic insulin sensitivity after glyburide treatment.

....

- - -

*.

INSULINb~l$ENTRATION m August 20, 1990

The American Journal of Medicine

Volume 89 (suppl 2A)

21-27s

SYMPOSIUMON GLYBURIDEIMcGUlNNESSand CHERRINGTON

sion of R,), the EDS0 for insulin action on R, must be below the portal insulin level observed after an overnight fast (20 to 25 pU/mL). In summary, the liver of a normal person responds to small changes in portal insulin by altering its rate of glucose production. The precise relationship between portal insulin levels and glucose production, however, has not been established. Most studies performed to date have been flawed in that they failed to consider the entire range of insulin levels and take into account portal insulin levels, and to fix the basal glucagon concentration. In addition, these studies used a tracer technique that underestimated R,. Nevertheless, since the removal of insulin causes a slightly larger increase in R, than the maximal insulin-induced inhibition of R,, the EDS0 for the response must be slightly below the basal portal insulin level. This explains why small changes in insulin can alter R, (glucose production) dramatically.

INSULINAND GLUCOSEPRODUCTION IN NIDDM Elevated endogenous glucose production rates and reduced glucose clearance are characteristic of NIDDM. The consequences of these defects are fasting hyperglycemia and excessive postprandial blood glucose excursions. The relative contributions of impaired hepatic and peripheral insulin sensitivities to the pathogenesis of hyperglycemia in NIDDM have recently been evaluated [15]. In diabetic persons who had plasma glucose concentrations of less than 140 mg/dL after an overnight fast (10 to 12 hours), HGP rates were similar to those observed in normal control subjects (1.85 + 0.03 versus 1.84 + 0.02 mg/ kg/minute). However, basal plasma insulin and plasma glucose concentrations were significantly higher in patients with NIDDM? indicating relative hepatic resistance. The metabolic clearance rate of glucose was significantly reduced in patients with NIDDM compared with normal subjects (1.56 f 0.03 versus 2.00 & 0.03 ml/kg/minute). Thus, the investigators concluded that in subjects with mild fasting hyperglycemia, the major defect is in peripheral glucose uptake. As the fasting plasma glucose concentration increased above 140 mg/dL, a progressive increase in hepatic glucose production was observed. R, increased to approximately 3.4 mgikg/minute in patients with a fasting glycemia of approximately 300 mg/dL. Because the metabolic clearance rate of glucose decreased by only an additional 26 percent as the prevailing glycemia increased above 180 mg/dL, the progressive increase in hepatic glucose output would appear to be the major factor contributing to fasting hyperglycemia in severely hyperglycemic persons. The marked increase in glucose production that occurred in association with the increase in basal glycemia may be explained by a failure of the plasma insulin levels to remain elevated or by increased hepatic insulin resistance, or both. The concept that the liver is a major site of insulin resistance in severely hyperglycemic patients with NIDDM was further supported by Campbell et al [16]. They concluded that, in patients with NIDDM, the peripheral tissues and the liver were equally resistant to insulin action. They suggested, however, that elevated HGP is the major cause of basal hyperglycemia in such patients. As already discussed, this conclusion is logical since only 25 percent of the total rate of glu2A-28s

August 20, 1990

The American Journal of Medicine

cose utilization occurs in insulin-sensitive tissues. A 50 percent decrease in glucose removal by insulin-sensitive tissues would increase the plasma glucose level by only 13 percent. Therefore, hepatic insulin resistance and the consequent increase in HGP are the main causes for the additional increase in fasting glycemia (~140 mg/dL) in patients with NIDDM. Not all studies, however, support the concept that hepatic overproduction of glucose is a major cause of severe fasting hyperglycemia in NIDDM. Chen et al [17] concluded that a defect in peripheral glucose disposal must be the major cause for the expanded glucose pool in patients with NIDDM. They observed that after a X&hour fast, glucose production was elevated in patients with NIDDM. After a 1Bhour fast, glucose production was similar in normal subjects and in patients with NIDDM with severe hyperglycemia (approximately 300 mg/dL). However, because glucose production was inappropriately high for the prevailing glycemia in the NIDDM patients, it is likely that the liver was insulin resistant. Measuring endogenous glucose production in patients with NIDDM is not without its difficulties. The first problem is that, in contrast to normal subjects, basal glycemia and HGP are not in a steady state after a 12-hour fast. Glauber et al [18] found that during prolonged studies (10.5 hours) glycemia and glucose production decreased by 34 and 40 percent, respectively. When the prevailing glucose concentration was kept constant by an exogenous glucose infusion, glucose production was decreased to an even greater extent (by approximately 65 percent; see Figure 2). The second problem is that the typical 2-hour tracer equilibration period used in normal subjects is not adequate in patients with NIDDM because of the prevailing hyperglycemia and the low fractional turnover of glucose. Chen et al [19] concluded that at least 4 hours of isotope infusion are necessary for accurate basal glucose production measurements; otherwise, falsely high values may be obtained. Because of the non-steady-state kinetics, expanded size, and low fractional turnover rate of the plasma glucose pool in NIDDM patients, calculating the relative contribution of increased HGP and impaired peripheral glucose utilization is difficult. As a result of the slow kinetics, elevated HGP in the morning, which normalizes in the afternoon, may explain the fasting hyperglycemia in both the morning and afternoon. Consequently, directly calculating the relative contributions of increased HGP and impaired glucose utilization to the fasting hyperglycemia at one point in time (i.e., in the morning) will yield incorrect results. One possible solution is to calculate their relative contributions over a 24-hour period, which may provide a more realistic assessment of their relative importance. There is general agreement that peripheral and hepatic insulin action are decreased in NIDDM; however, the magnitude of hepatic insulin resistance and the relative contributions of peripheral and hepatic insulin resistance to basal hyperglycemia are still controversial. This controversy may in part be due to the heterogeneity of the NIDDM population and to limitations in the methods used to assess hepatic insulin action. However, irrespective of the method used, hepatic overproduction of glucose (an absolute increase in glucose production or an inappropriately

Volume 89 (suppl 2A)

SYMPOSIUM

250T

qlucoss

.

saline

ON GLYBURlDEi

and CHERRINGTON

infused

infused

J

~A-,-A

saline infused

IA-A-,,

** . .A-._

glucose

Figure 2. Effects of fasting and an isoglycemic clamp on plasma glucose and glucose turnover in patients with NIDDM. Data are expressed as mean. Adapted with permission from [18] and the American Diabetes Association.

McGUlNNESS

. . - -A. f

--A---.

infused’

04

LO 0

100

t

200

300

400

500

600

TIME bin)

07:oo

high glucose production for the prevailing glycemia) plays a central role in contributing to the alterations in glucose metabolism in patients with NIDDM.

INSULIN,GLYCOGENOLYSIS, AND GLUCONEOGENESIS Glucose is produced by the liver by two processes: glycogenolysis and gluconeogenesis. It is of interest to compare the ability of insulin to inhibit these two processes in viwo. Early studies 120-221 suggesting an inhibitory effect of insulin on gluconeogenesis, while prophetic, were seriously flawed and cannot be used as proof for such an action of the hormone. Studies conducted a decade ago by Chiasson et al [3,4] in both humans and dogs were the first to prove definitively that insulin could inhibit gluconeogenesis in viva. Insulin levels of approximately 200 PTJlmL (in the portal vein) were required to obtain 80 percent suppression of basal gluconeogenesis. Based on these findings, one might conclude that gluconeogenesis is less sensitive to insulin than is glycogenolysis. Before reaching such a conclusion, however, it is necessary to consider the effect of removal of basal insulin on the gluconeogenic rate. Selective removal of insulin (achieved using a somatostatin infusion combined with intraportal glucagon replacement) was associated with a threefold increase in the gluconeogenic rate as determined by using tracer ([14C]alanine) techniques [14]. Because the glucose level increased in that study, and because hyperglycemia is inhibitory to gluconeogenesis [23], it is likely that the inhibitory action of insulin per se was somewhat underestimated. Considered together, these experiments show that insulin exerts a potent inhibitory effect on gluconeogenesis. In fact, the ED,, for inhibition must lie considerably below the basal portal insulin level. Such a conclusion would explain why small increases in insulin above baseline do not reduce gluconeogenesis appreciably, whereas small reductions in insulin can cause marked increases in the gluconeogenic rate. With regard to the relative sensitivities of glycogenolysis and gluconeogenesis, gluconeogenesis is more sensitive to inhibition with insulin than is glycogenolysis. The latter point is particularly relevant to eating and to liver glycogen repletion, since it is now believed that gluconeogenesis remains active in the postprandial period when portal insulin levels are elevated and glycogenolysis is inhibited.

In NIDDM, the relative contributions of glycogenolysis and gluconeogenesis to the elevated endogenous HGP and their relative sensitivities to suppression by insulin have not been extensively examined. A few studies have attempted to determine the relative contributions of glycogenolysis and gluconeogenesis to total glucose production in patients with NIDDM. Total gluconeogenic precursor uptake by the splanchnit bed and the production of [14Clglucose from [14C]alanine in NIDDM [24] were found to be normal, suggesting that gluconeogenesis in such patients is not elevated. Recent data obtained by Consoli et al [25] suggested that fasting hyperglycemia in NIDDM is secondary to increased $luconeogenesis l-251. To measure gluconeogenesis, [’ Clacetate was infused to introduce label into the gluconeogenic pathway and plasma beta-hydroxybutyrate-specific activity was assessed as a means of estimating mitochondrial acetyl-coenzyme A (CoA)-specific activity. When this new method was used, gluconeogenesis was increased threefold in NIDDM patients (2.3 -+ 0.3 versus 0.7 rt 0.01 mg/kg/min for NIDDM patients and normal subjects, respectively). In contrast, hepatic glycogenolysis was not significantly altered in NIDDM. Although these results suggest that gluconeogenesis is altered in NIDDM patients, the methodology has come under considerable criticism, making interpretation of the results difficult. One criticism is that the plasma betahydroxybutyrate-specific activity may not adequately reflect liver acetyl-CoA-specific activity but may also reflect the acetyl-CoA pool in other tissues (26). Another criticism is that liver acetyl-CoA is compartmentalized and nonhomogeneous. If that is the case, determining to what extent the plasma-specific activity reflects precursor-specific activity in the gluconeogenie compartment becomes difficult. The differing conclusions from the aforementioned studies may be related to the limitations in the different methods used to assess gluconeogenesis or the heterogeneity of the populations studied. Consequently, the extent-to which gluconeogenesis is elevated in NIDDM patients remains unclear.

INSULINAND LIVERGLUCOSEUPTAKE Physiologic hyperinsulinemia alone (i.e., that which is brought about in the presence of euglycemia) causes minimal net hepatic glucose uptake in dogs [51 and even less in humans 121. Similarly, hyperglycemia

August 20, 1990

The American Journal of Medicine

Volume 89 (suppl 2A)

2A-29s

SYMPOSIUMON GLYBURIDE/ McGUlNNESSand CHERRINGTON

ORAL GLUCOSE (100 gd I

l -

NIDDM

.-.

NORMAL

o-0

/

i

04

I

1

I

100

0-l

-60

-30

0

I

30

I

I

60 TIME

90 (min)

alone (i.e., that which is brought about in the presence of euinsulinemia) causes little or no net hepatic glucose uptake [2,27]. The combination of modest hyperinsulinemia and hyperglycemia, however, results in a small but significant amount of net hepatic glucose uptake in humans [2,27] and in dogs [28-321. Despite this observation, the combination of the two signals cannot explain the rate of net hepatic glucose uptake seen after feeding [27,33]. 2A-30s

August 20, 1990

The American Journal of Medicine

120

150

180

Figure 3. Plasma insulin and glucose concentrations and net splanchnic glucose balance in diabetic subjects after glucose ingestion. Data are expressed as mean. Adapted with permission from [36] and the American Diabetes Association.

When glucose is given orally to humans [27] or to dogs [31-331, the maximal rate of net hepatic glucose uptake approaches 6.0 mg/kg/minute. If the same glucose and insulin levels are created through infusion of glucose into a peripheral vein, the response of the liver varies from 20 to 50 percent of that seen after feeding. Such findings led DeFronzo et aE [271 to suggest that a “gut factor”’ was involved in liver signaling during the normal response to an oral glucose load.

Volume 89 (suppl 2A)

SYMPOSIUM

Studies by several groups, however, have shown that portal and oral glucose loading augment net glucose uptake in a similar manner, and that both result in greater glucose uptake than is observed with peripheral glucose loading [28-321. These findings suggest that some aspect of portal glucose delivery might result in an insulin-like effect on the liver. Adkins et al [34] tested this hypothesis directly by fixing both portal and arterial insulin and glucagon levels at basal values and doubling the glucose load to the liver using either a portal or a peripheral glucose infusion. In the presence of the peripheral glucose infusion, net hepatic glucose output was abolished but net hepatic glucose uptake did not occur. In the presence of the portal glucose infusion, however, the liver took up significant amounts of glucose (1.4 mg/kg/minute). When the experiment was repeated in the presence of a fourfold increase in insulin [29], net hepatic glucose uptake was 1.4 mg/kg/minute in the presence of the peripheral glucose infusion and 3.5 mg/kg/minute in the presence of the portal glucose infusion. When the glucose and insulin levels seen after oral glucose consumption and the arterial-portal glucose gradient present during feeding were combined, the rate of net hepatic glucose uptake was indistinguishable from that seen after oral glucose loading [35]. Under conditions normally seen postprandially, the effect of the “portal signal” appears to be as great as the effect of the elevation in insulin. The nature of this new regulatory mechanism and its relationship to insulin remain to be established. Very few studies have examined the role of the liver in postprandial glucose disposition in patients with NIDDM, primarily because of the invasive nature of measuring splanchnic glucose balance. Felig et al [361, using the arteriovenous difference technique, examined splanchnic glucose uptake after an oral glucose load. Splanchnic glucose output was 33 percent greater in patients with NIDDM compared with controls (Figure 3), leading them to conclude that glucose extraction by the liver was decreased in these patients. This conclusion was not necessarily warranted, because it is possible that HGP may not have been inhibited equally in the two groups during the oral glucose load. In addition, the investigators assumed that 100 percent of the glucose load was absorbed as glucose during the 180-minute study. Evidence obtained subsequently in normal humans suggests that only 73 percent of such a glucose load would be absorbed as glucose [37]. Firth et al [38] concluded that the postprandial hyperglycemia in NIDDM was due to an impairment in the suppression of endogenous glucose production and an impairment in the stimulation of peripheral and splanchnic glucose uptake. Since the insulin response after a meal in the diabetic group was blunted, the investigators were unable to determine if the impairment was due to hepatic insulin resistance. Sacca et al [39] examined the role of splanchnic tissues in the development of impaired glucose tolerance. During an intravenous glucose infusion, splanchnic uptake was increased in the insulin-intolerant group and peripheral glucose uptake was decreased by 3’7 percent. The larger splanchnic uptake of glucose was explained by the much larger glycemic and insulin excursions seen in the intolerant group. It was not clear whether hepatic insulin sensitivity was altered. In summary, splanchnic handling of an oral glucose

ON GLYBURIDEI

MeGUiNNESS

and CHERRINGTON

load is abnormal in patients with NIDDM. The evidence suggests that the liver plays a central role in the impaired response to a meal by failing to extract an adequate amount of glucose and by failing to appropriately suppress its rate of endogenous glucose production. To what extent hepatic insulin resistance contributes to abnormal splanchnic handling of glucose in the diabetic patient remains unclear. EFFECT OF GLYBURIDE ON THE LIVER Normal Subjects, Glucose-Intolerant Subjects, and Patients with Insulin-Dependent Diabetes Mellitus

The short-term administration of glyburide (5 mg) to nondiabetic subjects 15 minutes before breakfast led to a sustained increase in pancreatic insulin secretion (increased 40 percent for approximately 7 hours) and a brief period of hypoglycemia (~35 mg/dL approximately 60 minutes after the meal) (Figure 4 [40]). When glyburide (2.5 mg/day) was given for 6 weeks, the levels of insulin and glucose apparent after an overnight fast were both normal, as was hepatic glucose production (Table II) 1411. The response to an oral glucose tolerance test, however, was improved without a concomitant increase in the pancreatic insulin response, suggesting that insulin sensitivity had been improved (Figure 5) 141). This was confirmed using the euglycemic-hyperinsulinemic clamp technique. Significant increases in insulin action of 8 and 16 percent were observed after 3 days and 6 weeks of glyburide administration, respectively 1411. In patients with borderline diabetes, the effects of glyburide are similar to those in normal subjects. In patients with mild diabetes treated with glyburide in a double-blind study for 2 years (4 mg per day), no detectable changes in basal insulin or glucose were measured, but the response to an oral glucose tolerance test was slightly improved [42]. Studies have also examined the response of patients with insulin dependent diabetes mellitus to glyburide. Although treatment of insulin-dependent diabetes mellitus with glyburide leads to a slight improvement in whole-body insulin sensitivity, it does not seem to improve glucose control (no change in glycosylated hemoglobin) or to decrease insulin requirements [43]. In patients who are not insulin-resistant or are only mildly so, glyburide produces an increase in postprandial pancreatic insulin secretion and hypoglycemia. After long-term treatment with glyburide, a slight improvement in glucose tolerance persists and is probably due to subtle improvements in hepatic insulin action or to undetectable increases in pancreatic insulin secretion, or both. This small response to glyburide may in part be due to the low doses used. The effects of the drug on HGP and uptake are unclear. No studies have examined the effect of glyburide on glycogenolysis, gluconeogenesis, or hepatic glucose uptake in these populations. NIDDM

After treatment of NIDDM patients with glyburide (8 to 11 mg per day) for 2 and 6 months, fasting hyperglycemia and the glucose responses to an oral glucose tolerance test (OGTT) were improved [44]. After 2 months of therapy! basal insulin levels were elevated when compared with levels before treatment (22 +- 8 versus 12 ? 4 pU/mL), and the peak insulin response to an OGTT was markedly increased (97 ? 33 versus

August 20, 1990

The American Journal of Medicine

Volume 89 (suppl 2A)

2A-31s

SYMPOSIUM

ON GLYBURlDE/

McGUlNNESS

Glyburide

and CHERRINGTON

(5 mg

p.o.)

control glyburide

J T

Glyburide

, 1(

o-o 0-e

!.

isocaloric meals I

(5 mg p.o.)

isocaloric

Z0,

7

9

11

13

15

17

19

TIME (hours) TABLE II Effect of Glyburide Administration on Glucose Metabolism in Normal Subjects and in Patients with NIDDM Normal

Glucose (mg/dL) Insulin fj.dJ/mL) Glucose production (mg/kg/minute)

NIDDM

Baseline

After Therapy*

Baseline

Aftef Therapy-t

79*2 IOf 1

7924 1321

1;; i ;9

1;; ‘: ;6

1.9 + 0.1

1.9 + 0.1

2.4l 0.3 1.7ii 0.2

Normal subjects were given glyburide (2.5 + 1 mg per day) for 6 weeks. ‘atients with NIDDM were given glyburide (15 + 2 mg per day) for 3 months. ata adapted with permission from [41] and the American Diabetes Association.

42 + 15 pU/mL). However, after 6 months, the basal insulin level (11 + 3 pU/mL) and the insulin response to an OGTT (peak of 54 * 14 PUlmL) had returned to pretreatment levels. Despite this, basal glycemia (211 + 16, 146 + 41, and 136 & 19 mg/dL before and after 2 and 6 months of therapy, respectively) and the response to an OGTT remained markedly improved. Kolterman et al [45] examined the impact of glyburide therapy (3 months; 24 mg per day) on HGP in NIDDM patients after an overnight fast. They observed, as other investigators had, that after glybu2&32S

August 20,

1990 The American Journal of Medicine

1 21

Figure 4. Venous blood glucose and insulin concentrations in normal subjects after the shortterm administration of glyburide. Data are expressed as mean. Adapted with permission from

[401.

ride therapy, *basal insulin levels and insulin response to a meal were increased. They also observed that some subjects were “good” responders (group I; Figure 6) and others were “poor” responders (group II; Figure 7). Glyburide treatment caused the basal glucose level in the “good” responders to decrease 128 mg/dL! whereas the glucose level decreased only 48 mg/dL m the poor responders. The basal insulin levels were similar before treatment (20 -+ 3 and 29 * ‘i-pU/mL in groups I and II, respectively) and tended to increase after glyburide therapy (29 ? 10 and 41 t 21 PUlmL in groups I and II, respectively). The good responders had a 48 percent reduction in the glucose response and a 50 percent increase in the insulin response to a meal. In the poor responders, glyburide induced only a 26 percent decrease in the glucose excursion and a larger increase (66 percent) in the insulin response to a meal. When therapy was extended to 18 months, the insulin response to a meal in the good responders decreased relative to that seen at 3 months, without worsening of the glycemic excursion. These data support the concept that glyburide has an extrapancreatic effect. Basal (overnight fasted) HGP in the good responders had decreased 36 and 28 percent after 3 and 18 months of glyburide therapy, respectively, which accounts for at least 50 percent of the 49 percent de-

Volume 89 (suppl 2A)

SYMPOSIUMON GLYBURIDEI McGUlNNESSand CHERRINGTON glucose

(1 g/kg; p.o.)

180-

.. Y ,o

6

l . . . . .,,

60--

80

\ .,

O-k,‘.

. 0..

*

[email protected]

< 180

I 210

T

Figure 5. Arterial plasma glucose and insulin concentrations after the administration of an oral glucose load to normal subjects before and after treatment with glyburide for 6 weeks. Data are expressed as mean. Adapted with permission from [41] and the American Diabetes Association

0-i 0

30

6 60

SO

TIME

crease in the basal glucose level [45]. On the other hand, in the poor responders basal HGP decreased 3 percent and increased 15 percent at 3 and 18 months, respectively, indicating that the 26 percent decrease in glucose was caused by an increase in peripheral glucose removal. A strong positive correlation between basal glucose production and basal glycemia was observed. A reduction in hepatic glucose output after glyburide therapy (15 mg per day; 3 months) and the correlation between basal HGP and basal glycemia were confirmed by Simonson et al, as shown in Table II [411. However, their work suggested that nearly all of the decrease m the glucose concentration was the result of a reduction in glucose production; plasma glucose decreased 28 percent and hepatic glucose output 26 percent [41]. The enhancement of peripheral insulin sensitivity noted in both of these studies accounted for a relatively minor portion of the improvement in basal glycemia. The greater role of the liver in improving glycemia may be related to the difference in severity of the diabetes in the two studies. Subjects in the Kolterman study [45] were more hyperglycemic before therapy (264 + 17 versus 198 + 19 mg/dL) than those in the Simonson study [41]. As already mentioned, the importance of the liver in the regulation of plasma glucose levels is not unexpected, because insulin-sensitive tissues account for

120

I 150

(min)

only a small portion of the glucose utilized in the basal state. Complete impairment of peripheral insulin action can potentially increase the plasma glucose concentration 20 to 25 percent in the presence of normal glucose production. The peripheral tissues, however, play an ever-increasing role in determining the prevailing glycemia as the rate of basal hepatic glucose production increases. This is because the Michaelis constant of brain glucose transport is low. Consequently, as the glycemia increases, its relative contribution to the total glucose disposal decreases. Therefore, in NIDDM patients, whose glucose production is 30 percent above normal, the extra glucose must be consumed for the most part by insulin-sensitive tissues or, if the capacity of these tissues to utilize glucose is exceeded, expelled in the urine. It remains unclear ,whether the improvement in HGP after glyburide therapy is the result of a direct effect of glyburide on the liver or is secondary to the improvement in pancreatic insulin secretion. Patients with NIDDM are sensitive to the actions of glyburide. After brief periods of glyburide therapy, the major effect is increased pancreatic insulin secretion. However, after prolonged therapy, basal insulin levels and the insulin secretory response return to pretreatment levels, but the glucose-lowering effect of glyburide persists. The decrease in basal glycemia

August 20, 1990

The American Journal of Medicine

Volume 89 (suppl 2A)

2A-33s

SYMPOSIUMON GLYBURIDE/McGUlNNESS and CHERRINGTON

“GOOD” RESPONDERS

pre-glyburide

O-O

post-glyburide

o-o

T

meal

O7

I 11

13

TIME (hours)

after prolonged glyburide therapy is primarily due to a reduction in HGP, which may be: (1) a direct effect of glyburide on the liver; (2) secondary to the improvement in glucose control after glyburide therapy; or (3) the result of an improvement in hepatic insulin sensitivity. The role of changes in gluconeogenesis and glycogenolysis in response to glyburide is unknown. The improvement in the response to an OGTT suggests that glyburide may enhance hepatic glucose uptake, although direct evidence for this has not been obtained.

IN VITROEFFECTSOF GLYBURIDE Although evidence for a direct effect of glyburide on the liver in humans is limited, a considerable amount of in vitro data indicates that glyburide has a direct effect on hepatic glucose metabolism (glycogenolysis, gluconeogenesis, and glycogen synthesis). In perfused livers from fed rats, the short-term administration of glyburide (1.6 pg/mL) enhances the ability of insulin to inhibit glucagon-stimulated HGP [46]. In the presence of basal (1 x ; 10 $J/mL) insulin, glyburide inhibited glucagon-stimulated HGP by 40 percent (Figure 8). This decrease was paralleled by an inhibition of phosphorylase activation. This finding suggests that glyburide inhibited glycogenolysis (Figure 4). In the absence of insulin, glyburide did not significantly in2A-34S

I 15

Figure 6. Plasmaarterial glucoseand insulin concentrationsin patients with NIDDM who were “good” respondersbefore and after 3 monthsof glyburidetherapy.Dataare expressed asmean.Adaptedwith permissionfrom [45]and the AmericanDiabetesAssociation.

hibit glucagon-stimulated HGP, suggesting that glyburide requires the presence of insulin to exert its action. In the presence of markedly elevated insulin levels, however, glyburide did not enhance the inhibition by insulin of glucagon-stimulated HGP. These data provide evidence that glyburide enhances hepatic insulin sensitivity without altering hepatic insulin responsiveness. Gluconeogenesis can be inhibited by glyburide in vitro. In liver homogenates obtained from alloxantreated diabetic rats that were exposed to pharmacologic levels of glyburide (200 pg/mL), an inhibition of phosphoenolpyruvate carboxykinase was observed [47]. The inhibition of pyruvate carboxylase by glyburide (1 pg/mL) in the absence of insulin has been observed in hepatic mitochondria isolated from fasted rats [48]. This inhibition was associated with a reduction in mitochondrial adenosine triphosphate. This finding suggests a mild uncoupling of oxidative phosphorylation [49]. In addition to inhibiting the gluconeogenic enzymes, glyburide (5 pg/mL) also inhibits fat oxidation [50]. This might potentially reduce the available energy supply to the liver, thereby causing an inhibition of gluconeogenesis. The inhibition of fat oxidation by glyburide is caused by an inhibition of longchain fatty acid oxidation mediated by an inhibition of carnitine palmitoyltransferase I [50].

August20, 1990 The AmericanJournalof Medicine Volume89 (SIJPPl2A]

SYMPOSIUM

“POOR”

ON GLYBURIDE / McGUlNNESS

RESPONDERS

O------a

/

-\

0/ 0/

and CHERRINGTON

l ---. <\o/;/b ‘\/ pre-glyburide post-glyburide

T

O-O l

-

l

T

meal

meal

50

00

50

Figure 7. Plasma arterial glucose and insulin concentrations in patients with NIDDM who were “poor” responders before and after 3 months of glyburide therapy. Data are expressed as mean. Adapted with permission from [45] and the American Diabetes Association.

0 9

11

13

15

TIME (hours)

In the absence of insulin, glyburide has also been shown to activate the glycolytic pathway in the liver by stimulating phosphofructokinase and pyruvate kinase. In human liver slices, glyburide (1 pg/mL) reduced the inhibitory effect of glucagon and epinephrine on pyruvate kinase 77 and 46 percent, respectively [49]. It also reduced the inhibitory effect of glucagon and epinephrine on phosphofructokinase 99 and 70 percent, respectively. This inhibition can be explained by the finding that fructose-2,6-bisphosphate in the liver is increased after glyburide (0.5 pg/mL) administration [51]. Fructose-2,6-bisphosphate is a potent activator of phosphofructokinase. The stimulation of the glycolytic pathway will lead to a decrease in net gluconeogenesis. Fleig et al [52] demonstrated that, when primary cultured hepatocytes were exposed to glyburide for 24 hours in the presence of insulin, there was an enhancement of insulin-stimulated glycogen synthesis (Figure 9). This stimulation occurred without a change in insulin binding, suggesting that the effect of glyburide was mediated by a postreceptor mechanism. In the absence of insulin, no effect of glyburide was observed. Glyburide has also been shown to enhance other insulin-sensitive processes, such as lipogenesis [53,54] and amino acid uptake [54]. Initially it was thought that glyburide increases

hepatic insulin sensitivity by increasing the number of hepatic insulin receptors. This hypothesis was based on the observation that, in cultured human fibroblasts, glyburide inhibited an insulin-induced decrease in insulin binding [55]. In addition, Beck-Neilson et al [56] observed an increase in insulin binding to monocytes after treatment of obese diabetic patients with glyburide. However, subsequent studies in vitro have not confirmed these findings [52,57]. Since the evidence suggests that the effect of glyburide in vitro is not simply the result of a change in the insulin receptor and in some cases does not require the presence of insulin, glyburide must exert a primary effect at the postreceptor level. The mechanism by which glyburide might exert its postreceptor effect may involve an inhibition of cyclic adenosine monophosphate-dependent protein kinase [58]. Glyburide was found to bind noncompetitively to the catalytic subunit of the enzyme and thus decrease the potencies of hormones such as glucagon and epinephrine to catalyze phosphorylation. Glyburide is effective in lowering the prevailing glycemia observed in patients with NIDDM. The mechanism by which it lowers glycemia is not completely understood. The ability of the drug to decrease HGP, however, plays a central role in the improvement of glycemia. After short-term therapy, glyburide re-

August 20, 1990

The American Journal of Medicine

Volume 89 (suppl 2A)

2A-35s

SYMPOSIUMON GLYBURlDE/McGUlNNESSand CHERRINGTON

i

saline glyburide

1x

4x

INSULIN CONCENTFWlON

II m

Figure 8. Top, changein hepaticglucoseproductionafter exposure(for 30 min utes)of perfusedrat liverto 8x basalglucagon(88 pg/mL)in the presenceof 0, 1, 4, and 24x basalinsulinconcentrations(lx basal= 10 yUlmL) in the presence or absenceof glyburide(1.6 &ml). Bottom, percentchangein phosphorylasea activityafter exposure(for 30 minutes)of perfusedrat liver to 8x basalglucagon concentration(88 pg/mL)in the presenceof 0, I, 4, and 24x basalinsulinconcentrations(1x = 10 $J/mL). *p CO.05versussalinecontrolat correspondinginsulin concentrations.Data are expressedas meanf S.E.M.Adapted with permission from [46] and the AmericanDiabetesAssociation.

24X

lx= times basal)

T

‘\.

giyburide \

b

/ control

b

i

1’0

INSULIN CONCENTRATION 2A-36s

lb0

lo-00 (@/ml)

August20, 1990 The AmericanJournal of Medicine Volume89 (suppl2A)

1 oboe

Figure 9. Insulinstimulationof hepaticglycogen synthesisin cultured hepatocytespreincubated in the absence(top) or presence(bottom) of insulin(560pUlmL) withor withoutglyburide(2 &ml) for 24 hours. Data are expressedas mean. DNA= deoxyribonucleicacid. Adapted with permissionfrom 1521andthe AmericanDia betes Association.

SYMPOSIUM

duces HGP by elevating pancreatic insulin secretion. Data on long-term therapy with glyburide suggest, however, that it also exerts a direct effect on the liver. In vitro studies have shown that the drug has the capacity to inhibit gluconeogenesis and glycogenolysis directly, but in vivo studies have not addressed this issue. It remains possible that the improved glucose control associated with glyburide therapy (and the augmentation of insulin release) may have secondary effects on hepatic glucose metabolism, leading to a suppression of glucose production. There is also in vitro evidence that glyburide can enhance postprandial hepatic glucose uptake, thus contributing to the ability of the drug to improve the response to an OGTT. Direct evidence for such an effect in viva is, again, lacking.

REFERENCES 1. Rizza RA, Mandarin0 LJ, Gerich JE: Dose-response characteristics for effects of insulin on production and utilization of glucose in man. Am J Physiol 1981; 240: E630-E635. 2. DeFronzo RA, Ferrannini E, Hendler R, Felig P, Wahren J: Regulation of splanchnic and peripheral glucose uptake by insulin and hyperglycemia in man. Diabetes 1983; 32: 3545. 3. Chiasson JL, Liljenquist JE, Finger FE, Lacy Ww: Differential sensitivity of glycogenolysis and gluconeogenesis to insulin infusions in dogs. Diabetes 1976; 25: 283-287. 4. Chiasson JL, Atkinson RL, Cherrington A: Insulin regulation of glucogenesis in man. Diabetes 1979; 2: 380-385. 5. McGuinness OP, Myers SR, Neal D, Cherrington AD: Chronic hyperinsulinemia decreases insulin action but not insulin sensitivity. Metabolism 1990; 39: 1-5. 6. Myers SR, Adkins-Marshall BA, Williams PE, Stinsen RI, Cherrington AD: Subtle changes in glucagon during the euglycemic clamp can markedly alter the apparent sensitivity of the liver to insulin (abstr). Diabetologia 1988; 31: 525A. 7. Finegood DT, Bergman RN, Vranic M: Modelling error and apparent isotope discrimination confound estimation of endogenous glucose production during euglycemic glucose clamps. Diabetes 1988; 37: 1025-1029. 8. McMahon MM, Schwenk WF, Haymond MW, Rizza RA: Underestimation of glucose turnover measured with [6-3H]. and [6,6,2He]- but not [6-t%] glucose during hyperinsulinemia in humans. Diabetes 1989; 38: 97-102. 9. Argoud GM, Schade DS, Eaton RP: Underestimation of hepatic glucose production by radioactive and stable tracers. Am J Physiol 1987; 252: E606-E700. 10. Yki-Jarvinen H, Consoli A, Nurjhan N, Young AA, Gerich JE: Mechanism for underestimation of isotopically determined glucose disposal. Diabetes 1989; 38: 744-748. 11. DeBodo RC, Steele R, Altszuler N, Dunn A, Bishop JS: On the hormonal regulation of carbohydrate metabolism: studies with %glucose. Prog Horm Res 1963; 19: 445-449. 12. Mari A, Cobelli C, Cherrington AD, McGuinness OP: A model for the study of glucose kinetics in the nonsteady state IN IFAC modelling and control in biomedical systems. New York: Pergamon, 1988; 357-358. 13. Sherwin RS, Tamborline W, Hendler R, Sacca L, DeFronzo RA, Felig P: Influence of glucagon replacement on the hyperglycemic and hyperketonemic response to prolonged somatostatin infusion in normal man. J Clin Endocrinol Metab 19R 45: 1104-1109. 14. Cherrington AD, Lacy WW, Chiasson JL: Effect of glucagon on glucose production during insulin deficiency in the dog. J Clin Invest 1978; 62: 664-669. 15. DeFronzo RA, Ferrannini E, Simonson DC: Fasting hyperglycaemia in non-insulin dependent diabetes mellitus: contribution of excessive hepatic glucose production and impaired tissue glucose uptake. Metabolism 1989; 38: 387-391. 16. Campbell PJ, Mandarin0 LJ, Gerich JE: Quantification of the relative impairment in action of insulin on hepatic glucose production and peripheral glucose uptake in noninsulindependent diabetes mellitus. Metabolism 1988; 37: 15-21. 17. Chen Y-DI, Jeng C-Y, Hellenbeck CB, Wu M-S, Reaven GM: Relationship between plasma glucose and insulin concentration, glucose production and glucose disposal in normal subjects and patients with non-insulin-dependent diabetes. J Clin Invest 1988; 82: 21-25. 18. Glauber M, Wallace P, Brechtel G: Effects of fasting on plasma glucose and prolonged tracer measurement of hepatic glucose output in NIDDM. Diabetes 1987; 36: 1187-1190. 19. Chen Y-DI, Swislocki ALM, Jeng C-Y,Juang JH, Reaven GM: Effect of time on measurement of hepatic glucose production. J Clin Endocrinol Metab 1988; 67: 1084T1089. 20. Nadkarni GB, Chitnis KE: Effect of insulin on gluconeogenesis from glycine2C14. Arch Biochem Biophys 1963; 101: 466-470. 21. Friedman B, Goodman EJ, Jr, Weinhouse S: Effects of insulin and fatty acids on gluconeogenesis in the rat. J Biol Chem 1967; 242: 3620-3624. 22 DeMeutter RC, Shreeve Ww: Conversion of DL-lactate-2-% or 3-% or pyruvate-2-‘% to blood glucose in humans: effects of diabetes, insulin, tolbutamide, and glucose load. J Clin Invest 1963; 42: 525-529. 23. Shulman GI, Lacy WW, Liijenquist JE, Keller U, Williams PE, Cherrington AD: Effect of glucose, independent of changes in insulin and glucagon secretion, on alanine metabolism in the conscious dog. J Clin Invest 1980; 65: 496-501. 24. Chochinov R, Bowen HF, Moorehouse JA: Circulating alanine disposal in diabetes mellhus. Diabetes 1978; 270: 987-999. 25. Consoli A, Nurjhan N, Capani F, Gerich JE: Predominant role of gluconeogenesis in increased hepatic glucose production in NIDDM. Diabetes 1989; 38: 550-557. 26. Kosugi K, Scofield R, Chandramouli V, Kumaran K, Schumann W, Landau B: Pathways

ON GLYBURIDEI

McGUlNNESS

and CHERRINGTON

of acetone’s metabolism rn the rat. J Biol Chem 1986; 261: 3952-3967. 27. DeFronzo RA, Ferrannini E, Hendler R, Wahren J, Felig P: Influence of hyperinsulinemia, hyperglycemia, and the route of glucose administration on splanchnic glucose exchange. Proc Nat1 Acad Sci USA 1978; 85: 5173-5178. 28. Cherrington AD, Williams PE, Abou Mourad N, Lacy WW, Liljenquist JE: Insulin as a mediator of hepatic glucose uptake in the conscious dog. Am J Physiol 1982; 242: E97ElOl. 29. Adkins-Marshall BA, Myers SR, Hendrick GK, et at Interaction between insulin and the route of glucose delivery in the regulation of net hepatic glucose uptake in the conscious dog. Diabetes 1990; 39: 87-95. 30. Bergman RN, Beir JR, Hourigan PM: lntraportal glucose infusion matched to oral glucose absorption. Diabetes 1982; 31: 27-33. 31. lshida T, Chap Z, Chou Ki, eta/: Differential effects of oral, peripheral intravenous, and intraportal glucose on hepatic glucose uptake and insulin and glucagon extraction in conscious dogs. J Clin Invest 1983; 72: 590-598. 32. Barrett EJ, Ferrannini E, Gusberg R, Bevilacqua S, DeFronzo RA: Hepatic and extrahepatic splanchnic glucose metabolism in the postabsorptive and glucose fed dog. Metabolism 1985; 34: 410-417. 33. Abumrad NN, Cherrington AD, Williams PE, Lacy WW, Rabin D: Absorption and disposition of a glucose load in the conscious dog. Am J Physiol 1982; 242: E398-E402. 34. Adkins BA, Myers SR, Hendrick GK, Stevenson RW, Williams PE, Cherrington AD: Im. portance of the route of Intravenous glucose delivery on hepatic glucose balance in the conscious dog. J Clin Invest 1987; 79: 557-563. 35. Myers S, McGuinness 0, Neal D, Cherrington AD: lntraportal glucose delivery alters the relationship between net hepatic glucose uptake and the insulin concentration. J Clin Invest, in press. 36. Felig P, Wahren J, Hendler R: Influence of maturity onset diabetes on splanchnic balance after oral glucose ingestion. Diabetes 1978; 27: 121-126. 37. Ferrannini E, Bjorkman 0, Reichard GA, et at The disposal of an oral glucose load in healthy subjects. A qualitative study. Diabetes 1985; 34: 580-588. 38. Firth RG, Bell PM, Marsh HM, Hansen I, Rizza RA: Postprandial hyperglycemia in patients with noninsulin dependent diabetes. Role of hepatic and extrahepatic tissues, J Clin Invest 1986; 77: 1525-1532. 39. Sacca L, Orofino G, Petrone A, Vigorito C: Differential roles of splanchnic and peripheral tissues in the pathogenesis of impaired glucose tolerance. J Clin Invest 1984; 73: 1683-1687. 40. Owens DR, Seldrup J, Wragg KG, Sherry KJ, Biggs PI: Repeated blood glucose and plasma insulin levels in normal volunteer subjects after receiving isocaloric meals before and after chlorpropamide and glibenclamide. Horm Metab Res 19n 9: 347-351. 41. Simonson DC, Ferrannini E, Bevilacqua S, eta/: Mechanism of improvement in glucose metabolism after chronic glyburide therapy. Diabetes 1984, 33: 838-845. 42. Papoz L, Job D, Eschwege E, ef al: Effect of oral hypoglycemic drugs on glucose tolerance and insulin secretion in borderline diabetic patients. Diabetologia 1978; 15: 373-380. 43. Pernet A. Trimble ER. Kuntschen F. Assal J. Hahn C. Renold AE: Sulfonvlureas in insulin dependent (type II) diabetes: evidence’for an extrapancreatic effect in vi&. J Clin Endocri. nol Metab 1985; 61: 247-251. 44. Feldman JM, Lebovitz HE: Endocrine and metabolic effects of glibenclamide. Evidence for extrapancreatic mechanism of action. Diabetes 1971; 20: 745-755. 45. Kolterman OG, Gray RS, Shapiro G, Scarlett JA, Griffin J, Olefsky JM: The acute and chronic effects of sulfonylurea therapy in type II diabetic subjects. Diabetes 1984; 33: 346-354. 46. McGuinness OP, Green DR, Cherrington AD: Glyburide sensitizes perfused rat liver to insulin induced suppression of glucose output. Diabetes 1987; 36: 472-476. 47. Foy JM, Standing VF: The effect of glibenclamide on two enzymes important in gluco. neogenesis. Arzneimittelforschung 1974; 24: 1279-1282. 48. White CW, Rashed HM, Pate1 TB: Sulfonylureas inhibit metabolic flux through liver pyruvate carboxylase reaction. J Pharmacol Exp Ther 1988; 246: 971-974. 49. Belfiore F, Rabuazzo AM, lanneilo S, Campione R, Castorina S, Urzi F: Extra-pancreatic action of glibenclamide in man: reduction in vitro of the inhibitory effect of glucagon and epinephrine on the hepatic key glycolytic enzymes phosphofructokinase (PFK) and pyru vate kinase (PK). Eur J Clin Invest 1989; 19: 367-371. 50. Pate1 T: Effect of sulfonylureas on hepatic fatty acid oxidation. Am J Physiol 1986; 251: E241-E246. 51. Hatao K, Kaku K, Matsuda M, Tsuchiya M, Kaneko T: Sulfonylurea stimulates fructose2,6-bisphosphate formation in proportion to its hypoglycemic action. Diabetes Res Commun 1985; 1: 49-53. 52. Fleig WE, Noether-Rerg G, Fussgaenger R, Ditschuneit H: Modulation by a sulfonylurea of insulin dependent glycogenesis, but not insulin binding in cultured rat hepatocytes. Evidence for a post receptor mechanism of action. Diabetes 1984; 33: 285-290. 53. Amatruda JM, Salhanick Al, Chang CL: Hepatic insulin resistance in non-insulin dependent diabetes mellitus and the effects of a sulfonylurea in potentiating insulin action. Diabetes Care 1984; 7 (suppl 1): 47-53. 54. Caro J, ltoop 0, Sinha MK: Glyburide but not ciglitazone enhances insulin action in the liver independent of insulin receptor kinase activation. Metabolism 1989; 38: 606-611. 55. Prince MJ, Olefsky JM: Direct effect of a sulfonylurea to increase human fibroblast insulin receptors. J Clin Invest 1980; 66: 608-611. 56. Beck-Neilson H, Pederson 0, Lindskov HO: Increased insulin sensitivity and cellular insulin binding in obese diabetics following treatment with glibenclamide. Acta Endocrinol 1979; 90: 451-462. ;;-;;l/in JR: Dual actions of sulfonylureas and glyburide. Am J Med 1985; 79 (suppl 38): 58. Okuno S, lnaba M, Nishizawa Y, lnoue A, Morii H: Effect of tolbutamide and glyburide on CAMP-dependent protein kinase activity in rat liver cytosol. Diabetes 1988; 37: 857861.

August 20, 1990

The American Journal of Medicine

Volume 89 (suppl 2A)

2A-37s