Pathophysiology/Complications ORIGINAL ARTICLE

Colesevelam Improves Oral but Not Intravenous Glucose Tolerance by a Mechanism Independent of Insulin Sensitivity and b-Cell Function

ANNA L. MARINA, MD BRENDA K. MONTGOMERY, RN, MS, CDE currently under investigation with studies KRISTINA M. UTZSCHNEIDER, MD SANTICA M. MARCOVINA, PHD, SCD reporting conflicting results (6). There is a LORENA A. WRIGHT, MD STEVEN E. KAHN, MB, CHB complex interaction between FXR and LXR, which often have counterbalancing effects (7). LXR has been described as a OBJECTIVEd To determine the mechanism by which the bile acid sequestrant colesevelam glucose sensor (8), capable of improving improves glycemic control. glucose tolerance by promoting glucose RESEARCH DESIGN AND METHODSdWe performed a frequently sampled intrave- utilization and triglyceride synthesis and nous glucose tolerance test (FSIGT) with minimal model analysis and a meal tolerance test (MTT) inhibiting gluconeogenesis (9,10). We in 20 subjects with impaired fasting glucose (11 men, 9 women; mean age 60.7 6 1.9 years, BMI hypothesized that independent of the 29.4 6 0.9 kg/m2) in a single-blind study after 2 weeks of placebo treatment and 8 weeks of precise mechanism of the effects of bile colesevelam 3.75 g daily. From these tests, insulin sensitivity, b-cell function, and glucose acid sequestrants on glucose, if the FXR- tolerance were determined, along with gastrointestinal peptide levels during the MTT. LXR hypothesis is correct, treatment of humans with colesevelam would result RESULTSdFasting plasma glucose and HbA1c decreased with colesevelam (from 5.9 6 0.1 to 5.7 6 0.1 mmol/L, P , 0.05, and from 5.86 6 0.06 to 5.76 6 0.06%, P = 0.01, respectively), but in improvement in insulin sensitivity. fasting insulin did not change. Colesevelam had no effect on any FSIGT measures. In contrast, the While animal studies show improvement MTT incremental area under the curve (iAUC) for both glucose (from 249.3 6 28.5 to 198.8 6 23.6 in insulin sensitivity during treatment mmol/L z min, P , 0.01) and insulin (from 20,130 [13,542–35,292] to 13,086 [9,804–21,138] with bile acid sequestrants (11,12), such pmol/L z min, P , 0.05) decreased with colesevelam. However, the ratio of iAUC insulin to iAUC glucose an effect has not been clearly demon- was not changed. iAUC for cholecystokinin (CCK) increased (from 43.2 [0–130.1] to 127.1 [47.2– strated in humans (13). 295.2] pmol/L z min, P , 0.01), while iAUC for fibroblast growth factor 19 decreased (from 11,185 – – z P , The ability of bile acid sequestrants to [1,346 17,661] to 2,093 [673 6,707] pg/mL min, 0.01) with colesevelam. However, iAUC for lower blood glucose also has been linked glucagon, glucose-dependent insulinotropic peptide, and glucagon-like peptide 1 did not change. to their possible effect on intestinal glucagon- CONCLUSIONSdColesevelam improves oral but not intravenous glucose tolerance without like peptide 1 (GLP-1) secretion and, in changing insulin sensitivity, b-cell function, or incretins. This effect may be at least partially some studies, peptide Tyr-Tyr (PYY) release explained by the colesevelam-induced increase in CCK. (11,12,14). It has been suggested that se- questration of bile acids may interfere with – Diabetes Care 35:1119 1125, 2012 free fatty acid absorption in the proximal small intestine, resulting in increased free olesevelam is a bile acid sequestrant other bile acid sequestrants is mediated fatty acid delivery to the ileum and, con- Cthat is used for the treatment of hy- by the nuclear receptors farnesoid X re- sequently, enhanced GLP-1 secretion by percholesterolemia. More recent, it ceptor (FXR) and liver X receptor (LXR) the ileal L-cells (11). Furthermore, the in- has beenapprovedforuse inpatients with (3). Activation of FXR by bile acids leads creased levels of bile acids in the intestinal type 2 diabetes because it improves gly- to a negative feedback inhibition of bile lumen during treatment with bile acid se- cemic control, with decreases in HbA1c acid biosynthesis and secretion, in part questrants could also stimulate GLP-1 re- of ;0.5% compared with placebo when via increased expression of fibroblast lease via the G-protein–coupled receptor used in combination with metformin, sul- growth factor (FGF)-19 by enterocytes re- TGR5 (15). We hypothesized that if the fonylurea, or insulin (1–4). However, the sulting in diminished CYP7A1 expression glucose-lowering effect of colesevelam mechanism(s) by which colesevelam im- in the liver (5,6). Binding of bile acids by was related to increased incretin release, proves glucose tolerance is unknown. bile acid sequestrants reverses these ef- such an increase would be associated It has been suggested that the glu- fects. FXR appears to directly affect glu- with an improvement in islet (b- and/or cose-lowering effect of colesevelam and cose metabolism, but its specificroleis a-cell) function with meals. Thus, the primary objective of this ccccccccccccccccccccccccccccccccccccccccccccccccc study was to determine whether the From the Division of Metabolism, Endocrinology, and Nutrition, Department of Medicine, VA Puget Sound glucose-lowering properties of colesevelam Health Care System and University of Washington, Seattle, Washington. are the result of improvements in in- Corresponding author: Steven E. Kahn, [email protected]. sulin sensitivity and/or b-anda-cell Received 20 October 2011 and accepted 4 February 2012. function. Furthermore, we wished to DOI: 10.2337/dc11-2050. Clinical trial reg. no. NCT00990184, clinicaltrials.gov. © 2012 by the American Diabetes Association. Readers may use this article as long as the work is properly determine whether any improvements cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/ could be attributed to changes in the re- licenses/by-nc-nd/3.0/ for details. lease of incretins or other gastrointestinal care.diabetesjournals.org DIABETES CARE, VOLUME 35, MAY 2012 1119 Colesevelam and glucose metabolism peptides. We chose to study subjects with within 3 months; or use of HIV protease the end of the study. Reported adverse impaired fasting glucose (IFG) because inhibitors, warfarin, phenytoin, or any events were recorded, and laboratory safety they are at high risk of developing type 2 investigational drug within 30 days. Other assessments (complete blood count, elec- diabetes and because with their mild im- exclusion criteria were triglycerides .5.6 trolytes, plasma creatinine, liver function pairment in glucose metabolism, it is pos- mmol/L, uncontrolled hypothyroidism, tests, lactate dehydrogenase, creatine sible to use a number of sensitive methods clinical hepatic disease or liver function phosphokinase, lipid panel, and urinaly- to quantify changes in several important tests greater than two times the upper limit sis) were performed on days 14, 42, and parameters that regulate glucose without of normal, and history of major gastroin- 70. A standard 12-lead electrocardiogram excess concern of the deleterious effects of testinal surgery (gastrectomy, gastroenter- was recorded on days 14 and 70. glucose per se. Thus, we performed insu- ostomy, and bowel resection), dysphagia, lin-modified frequently sampled intrave- swallowing disorders, intestinal motility Assays nous glucose tolerance tests (FSIGTs) to disorder, or pancreatitis. Plasma glucose was measured using the quantify insulin sensitivity and b-cell func- hexokinase method (Roche Diagnostics, tion (16) and standardized meal tolerance Study procedures Indianapolis, IN), lipids were measured by tests (MTTs) to evaluate postprandial glu- Interventions. Subjects were provided enzymatic methods (Roche Diagnostics), cose, insulin, glucagon, and incretin re- with blinded active medication (colesevelam) and dextran sulfate precipitation for HDL sponses, all before and after treatment or matching placebo. Medication was cholesterol. Insulin and C-peptide levels with colesevelam. Levels of gastrointestinal takenwiththeeveningmeal,withpla- were measured using two-site immunoen- peptides were also measured immediately cebo being consumed for 2 weeks prior zymatic assays (Tosoh Bioscience, San before and during the standardized meal, to the performance of the first set of out- Francisco, CA). Radioimmunoassays were when they were expected to produce their come assessments. Thereafter, subjects used to measure plasma levels of pro- physiological effects. started taking 3.75 g colesevelam once insulin (Millipore, St. Charles, MO; HPI- daily with the evening meal and did so for 15K, minimum detection limit 2 pmol/L, RESEARCH DESIGN AND the next 8 weeks, at the end of which they intra-assay coefficient of variation [CV] 1.5– METHODSdThe study used a single- underwent a second series of outcome as- 6.9%, interassay CV 1.5–10.1%), glucagon blind, single-treatment design with a sessments. Medication compliance was as- (Millipore; GL-32K, minimum detection 2-week placebo run-in phase followed sessed by counting the number of unused limit 20 ng/L, intra-assayCV4–6.8%, inter- by 8 weeks of treatment with unbranded tablets returned on days 14, 42, and 70. assay CV 7.3–13.5%), total PYY (Millipore; active colesevelam hydrochloride (3.75 g No subjects were excluded based on the PYYT-66HK, minimum detection limit daily) with the evening meal. Subjects protocol requirement that subjects take 10 pg/mL, intra-assay CV 2.9–9.4%, in- were blinded to treatment throughout the .80% of the prescribed medication dur- terassay CV 5.5–8.5%), and sulfated cho- study, having been told that any time ing the placebo period (first 14 days). lecystokinin (CCK) (ALPCO, Salem, NH; during the study they may receive active Subjects were asked to maintain prior ex- 13-CCK-HU-R100, minimum detection medication or placebo. An insulin-modified ercise and dietary habits throughout the limit 0.3 pmol/L, intra-assay CV 2–5.5%, FSIGT and an MTT were performed on study. interassay CV 4.1–13.7%). Enzyme- days 14 and 70, before and at the end of FSIGT. An FSIGT was performed on linked immunosorbent assays were used treatment with colesevelam, respectively. days 14 and 70 after a 10-h fast. After to measure total glucose-dependent insuli- The study was registered at ClinicalTrials three basal blood samples were drawn, an notropic peptide (GIP) (Millipore; EZHGIP- .gov as NCT00990184. The VA Puget Sound intravenous glucose bolus (50% dextrose 54K, minimum detection limit 1.65 Health System Institutional Review Board at 11.4 g/m2 body surface area) was ad- pmol/L, intra-assay CV 3–8.8%, interas- approved the protocol, and written in- ministered during a 60-s period at time 0. say CV 1.8–6.1%), total GLP-1 (ALPCO; formed consent was obtained from all Eleven blood samples were collected dur- 43-GPTHU-E-01, minimum detection subjects prior to their participation in the ing the next 19 min, followed at 20 min by limit 0.6 pmol/L, intra-assay CV 3.7–4.7%, study. An independent data safety moni- the commencement of an insulin infusion interassay CV 6.2–9.5%), FGF-19 (R&D toring committee oversaw the performance (0.03 units/kg) administered during a 5-min Systems, Minneapolis, MN; DF1900, min- of the study. period. Subsequently, 21 blood samples imum detection limit 0.53–3.35 pg/mL, in- Eligible subjects were males and females were collected up until 240 min after the tra-assay CV 3.6–6.4%, interassay CV (postmenopausal, surgically sterile, or using start of glucose administration. 4.5–5.5%), and FGF-21 (R&D Systems; double-barrier method of contraception), MTT. At 30 min after completion of the minimum detection limit 1.61–8.69 pg/ aged 18–75 years, with a fasting plasma glu- FSIGT, subjects were given a standard- mL, intra-assay CV 2.9–3.9%, interassay cose of 5.6–6.4 mmol/L (100–115 mg/dL) ized liquid meal consisting of a can of CV 5.2–10.9%). When the result was be- at screening (average of two measurements Resource2.0(237mL;480kcal;20gpro- low the minimum detection limit for the during screening) and an HbA1c ,6.5%. tein, 52 g carbohydrate, 21 g fat, with added assay, the minimal detectable concentra- They also had to be otherwise in good minerals and vitamins). A dose of placebo tion was used for the analyses. health as determined by medical history, or colesevelam was taken at the beginning physical examination, electrocardiogram, of the standardized meal. Blood samples Calculations and laboratory tests at screening. were obtained at 5 and 1 min before the A number of measures were calculated Subjects were excluded if they had a meal and every 30 min for 120 min after from the FSIGT. The insulin sensitivity history of diabetes or treatment with glucose- starting the meal. index (SI) was determined from the glu- lowering agents, except for insulin during Safety assessments. A medical history, cose and insulin data using Bergman’smin- pregnancy; use of chronic oral or parenteral vital signs, and physical examination were imal model (16). The acute insulin, acute corticosteroids or bile acid sequestrants performed prior to randomization and at C-peptide, and acute proinsulin responses

1120 DIABETES CARE, VOLUME 35, MAY 2012 care.diabetesjournals.org Marina and Associates to glucose (AIRg, ACRg, and APIRg, re- fasting plasma glucose was 6.0 6 0.2 As illustrated in Fig. 1, glucose and spectively) were calculated as the mean in- mmol/L (range 5.6–6.4). All subjects were insulin concentrations during the FSIGT cremental responses above basal from compliant with medication usage, taking did not differ at the end of the placebo time 0 to 10 min and the glucose disap- .80% of both placebo and colesevelam run-in and colesevelam treatment periods. K pearance constant ( g) as the slope of the study medication. In general, colesevelam Thus, as listed in Table 1, SI, AIRg, ACRg, regression line relating the natural log of the was well tolerated, with 35% of subjects and APIRg did not change with treatment. glucose concentration from 10 to 19 min. reporting constipation while on the med- Furthermore, the disposition index also The disposition index was computed as the ication. One subject had an acute episode was not altered with colesevelam treat- product of SI and AIRg and provides a of cholelithiasis while taking colesevelam ment, and intravenous glucose tolerance, measure of b-cell function (17). but was not excluded from the study. There determined as Kg, did not change. From the MTT, incremental areas were no other serious adverse events. In contrastto what wasobserved with under the curves (iAUCs) for glucose, in- the FSIGT, changes in glucose, insulin, sulin, C-peptide, and glucagon were calcu- Effect of colesevelam on body and C-peptide profiles were observed lated using the trapezoidal method. In a anthropometry and lipids during the MTT (Fig. 2). Glucose toler- similar manner, iAUCs were calculated for As listed in Table 1, after 8 weeks of treat- ance during the MTT, calculated as the the gastrointestinal peptides GLP-1, GIP, ment with colesevelam, weight did not iAUC for glucose, improved with colesevelam PYY, and CCK, as well as for FGF-19 and change. While subjects were not required treatment (from 249.3 6 28.5 to FGF-21. to have lipid abnormalities for entry into 198.8 6 23.6 mmol/L z min, P , 0.01). the study, the changes in plasma lipid Furthermore, colesevelam administration Statistical analysis levels with colesevelam treatment were was associated with decreases in the iAUC Sample size was determined for AIRg consistent with the known effect of the for both insulin (from 20,130 [13,542– basedonaone-sample,two-sidedt test medication to decrease total and LDL 35,292] to 13,086 [9,804–21,138] at a significance level of 5% for compari- cholesterol, further supporting that sub- pmol/L z min, P , 0.05) and C-peptide son between the value at the end of the jects were compliant with the use of (from 122.8 [91.9–194.3] to 95.4 [75.3– placebo run-in period and the value after colesevelam. 140.2] nmol/L z min, P , 0.05). The re- 8 weeks of colesevelam treatment. The sult of the parallel changes in glucose and use of 20 subjects was calculated to pro- Effect of colesevelam on glucose the two b-cell peptides meant that the ra- vide 85% power to detect a clinically rel- tolerance, insulin sensitivity, and tios of each peptide to glucose were not evant 30% change from baseline in AIRg, islet function different for insulin (ratio of iAUC insulin assuming an intrasubject CV of 0.3 (18). Treatment with colesevelam improved to iAUC glucose: 125.4 6 16.2 to Paired-sample t tests were used to glucose levels, which were quantified as 109.2 6 16.2 [pmol/L]/[mmol/L], P = compare the means of normally distrib- significant decreases in fasting plasma 0.08) or C-peptide (ratio of iAUC C-pep- uted variables before and at the end of glucose and HbA1c (Table 1). Fasting plasma tide to iAUC glucose: 0.66 6 0.06 to treatment with colesevelam. Relation- insulin, proinsulin, and C-peptide concen- 0.60 6 0.06 [nmol/L]/[mmol/L], P = ships between variables were examined trations did not change with colesevelam 0.18), indicating that b-cell function had using linear regression. Those variables administration (Table 1). not changed. that were not normally distributed were log-transformed to achieve normal distri- bution or analyzed using the Wilcoxon Table 1dWeight, fasting plasma lipids, and measures of b-cell function, insulin signed rank test. Unless otherwise speci- sensitivity, and glucose tolerance before and at the end of treatment with colesevelam fied, all data are mean 6 SEM for nor- mally distributed variables and median Day 14 Day 70 (interquartile range) for nonnormally dis- (before treatment) (end of treatment) P value tributed data. Statistical analyses were performed with SPSS, version 13.0, with Weight (kg) 86.3 6 2.9 86.8 6 3.0 0.53 P , 0.05 considered significant. Total cholesterol (mmol/L) 4.59 6 0.19 4.06 6 0.19 ,0.001 LDL cholesterol (mmol/L) 2.79 6 0.18 2.25 6 0.16 ,0.001 RESULTS HDL cholesterol (mmol/L) 1.02 (0.87–1.30) 0.97 (0.90–1.32) 0.89 VLDL cholesterol (mmol/L) 0.57 (0.45–0.91) 0.65 (0.48–0.82) 0.46 Subject characteristics and Triglycerides (mmol/L) 1.23 (0.97–1.96) 1.40 (1.03–1.78) 0.46 disposition Fasting plasma glucose (mmol/L) 5.9 6 0.1 5.7 6 0.1 0.01 6 6 A total of 21 subjects met eligibility HbA1c (%) 5.86 0.06 5.76 0.06 0.01 criteria and were enrolled in the study. Fasting insulin (pmol/L) 56 (26–92) 43 (25–83) 0.28 Of these, 20 (11 men, 9 women) completed Fasting proinsulin (pmol/L) 11.6 (9.4–16.4) 11.5 (8.2–16.5) 0.68 the study. A male subject was withdrawn Fasting C-peptide (nmol/L) 0.8 (0.5–1.2) 0.7 (0.5–1.2) 0.42 3 25 21 z 21 6 6 during the placebo run-in phase because SI ( 10 min [pmol/L] )4.80.5 4.6 0.6 0.76 of an acute cerebrovascular accident. AIRg (pmol/L z min) 1,752 (564–3,306) 1,866 (678–2,670) 0.12 At randomization, subjects were ACRg (nmol/L z min) 5.8 (2.5–9.3) 6.7 (3.1–9.1) 0.16 60.7 6 8.7 years (mean 6 SD; range 40– APIRg (pmol/L z min) 45.2 (24.1–92.6) 44.6 (15.0–84.4) 0.15 2 75) and had a BMI of 29.4 6 4.2 kg/m2 Disposition index (3 10 4)569(290–1,235) 639 (325–1,012) 0.69 K z 21 6 6 (range 23.5–36), consistent with the sub- g (% min ) 1.26 0.08 1.32 0.10 0.55 jects on average being overweight. Their Data are mean 6 SEM or median (interquartile range). care.diabetesjournals.org DIABETES CARE, VOLUME 35, MAY 2012 1121 Colesevelam and glucose metabolism

4,610 6 404 to 4,209 6 451 pmol/L z min, P = 0.33, respectively). In a similar manner, there was no effect of colesevelam on the basal or postprandial levels of PYY (from 2,816 [1,108–4,323] to 2,271 [1,079– 4,373] pg/mL z min, P =0.46). The changes in CCK during the MTT were not correlated with the changes in glucose or the islet peptides insulin, C-peptide, or glucagon.

CONCLUSIONSdWe found that colesevelam significantly improved glu- Figure 1dPlasma glucose (A)andinsulin(B) levels during the FSIGT.-and solid line, before cose metabolism in subjects with IFG as treatment with colesevelam (day 14); ▫ and dashed line, end of treatment with colesevelam (day 70). demonstrated by reductions in both the fasting plasma glucose and HbA1c.These beneficial changes in glucose control did On the basis of the glucagon profile, treatment (from 11,185 [1,346–17,661] – z P , not appear to be the result of improve- colesevelam treatment also had no effect to 2,093 [673 6,707] pg/mL min, ments in insulin sensitivity or b-cell func- on a-cell function. The iAUC for glucagon 0.01), while that for FGF-21 did not 2 6 6 tion because we observed no change in SI was similar before and at the end of ther- change (from 1,157 948 to 175 using the minimal model of glucose kinet- apy (617 6 133 vs. 434 6 118 ng/L z min, 980 pg/mL z min, P =0.30). P ics or AIRg, respectively. In keeping with = 0.23) (Fig. 2). Treatment with colesevelam was as- the lack of changes in these two important sociated with increases in CCK during the determinants of intravenous glucose tol- Effect of colesevelam on the release MTT at 60, 90, and 120 min, the result erance (17), we observed no change in of gastrointestinal tract–related being a significant increase in iAUC this measure determined as Kg. Of great proteins (from 27.5 [2.4–109.2] to 200.3 [40.3– fi z P interest was the nding that in contrast to The effect of colesevelam on various 300] pmol/L min, = 0.001). Basal MTT what we observed for intravenous glucose gastrointestinal peptides is illustrated in CCK concentrations did not differ before tolerance, oral glucose tolerance did im- Fig. 3. and after treatment. During the MTT, prove with colesevelam administration. Colesevelam was associated with dif- iAUC for incretin peptides GLP-1 and fi Again, on the basis of the fasting insulin ferences in the MTT pro le of FGF-19 but GIP was not affected by colesevelam concentrations and the insulin and not FGF-21. The iAUC for FGF-19 was therapy (from 133 [38–172] to 98 [26– fi z P C-peptide responses after meal ingestion, signi cantly lower with colesevelam 166] pmol/L min, =0.25,andfrom insulin sensitivity and b-cell function were not changed by colesevelam treat- ment. The lack of change in the incretins GLP-1 and GIP is compatible with our observation of a lack of change in b-cell function. Thus, it appears that the improve- ment in glucose metabolism is probably independent of an effect of insulin. By what mechanism does colesevelam affect glucose tolerance independent of insulin sensitivity or b-cell function? Our observation that colesevelam treatment is associated with an elevation in CCK levels after the meal may provide an explana- tion. CCK is produced by the enterochro- maffin I cells of the proximal small intestine and has numerous physiologic effects, including decreasing food intake, slowing gastric emptying, and stimulating pancreatic exocrine and endocrine secre- tion, bile release, and intestinal motility (19). Given that our subjects did not lose weight and we observed no improve- ment in b-cell function, one hypothesis is that colesevelam-induced increases in CCK could improve oral glucose toler- Figure 2dPlasma glucose (A), insulin (B), C-peptide (C), and glucagon (D) levels during the ance in part by a delay in gastric empty- MTT. - and solid line, before treatment with colesevelam (day 14); ▫ and dashed line, end of ing. This hypothesis is supported by treatment with colesevelam (day 70). studies showing that concomitant CCK

1122 DIABETES CARE, VOLUME 35, MAY 2012 care.diabetesjournals.org Marina and Associates

Bile acid sequestrants have been shown to mildly increase basal CCK levels during the first days of treatment. This effect subsequently wears off, which may be related to the adaptation of CCK receptors (23,24) and may explain why we did not see an increase in basal CCK levels with colesevelam. In keeping with our observation of increased CCK levels after meal ingestion with colesevelam treatment, the cholestyramine-induced increase in plasma CCK levels persisted after the effect on the basal levels could no longer be detected (24). This effect of bile acid sequestrants to alter CCK levels appears to be mediated by an effect of bile acids (25–28). We did not observe an effect of colesevelam to change postprandial glu- cagon levels, which is in keeping with the findings of others (14,29). Furthermore, we observed no increase in GLP-1, GIP, and PYY release. The lack of an increase in GLP-1 in the peripheral circulation is con- trary to what has been reported by others (11,12,14). The reason we failed to repli- cate this findingisnotreadilyapparent, but the lack of improvements in b-and a-cell function is consistent with our ob- servation of unchanged GLP-1 levels. However, the effect of treatment to de- crease FGF-19 levels is compatible with the occurrence of other enteric effects of colesevelam. Figure 3dPlasma total GLP-1 (A), total GIP (B), PYY (C), CCK (D), FGF-19 (E), and FGF-21 Some animal studies suggest that the (F) levels during the MTT.-and solid line, before treatment with colesevelam (day 14); ▫ and beneficial effect of bile acid sequestrants dashed line, end of treatment with colesevelam (day 70). on glucose metabolism is related to an im- provement in insulin sensitivity (11,12). The data in humans is not extensive but infusion, which achieved physiological Although a decrease in MTT glucose would support our observations of no postprandial plasma CCK concentra- after treatment with colesevelam could be change in insulin sensitivity or fasting in- tions, and oral glucose administration de- related to a CCK-induced delay of gastric sulin after 8 weeks of treatment. A study layed gastric emptying and significantly emptying, this would not explain the using the euglycemic hyperinsulinemic reduced postprandial hyperglycemia and decrease in the fasting glucose level. clamp showed no improvement in periph- plasma insulin levels in healthy subjects Thus, it is likely that some other mecha- eral insulin sensitivity, although hepatic in- (20). In the same study, this effect was not nism is operative. Recent work in rats sulin sensitivity was not evaluated (13). observed when glucose was administered provides another possible explanation Furthermore, in this same study, coadmin- intraduodenally, confirming that CCK for a CCK effect applicable to our study. istration of the first dose of colesevelam lowers postprandial blood glucose by Intraduodenal administration of CCK in with a standard meal had no effect on slowing gastric motility. However, the the basal state decreases hepatic glucose postprandial glucose levels compared data supporting that colesevelam slows production by stimulating duodenal with baseline or placebo, suggesting that gastric emptying are limited. One study CCK-A receptors, with the signal trans- colesevelam does not impair glucose ab- that examines this issue suggests trends mitted to the nucleus of the solitary tract sorption. In our opinion, whether bile acid toward a delay of gastric emptying with in the hindbrain and then to the liver; this sequestrants affect hepatic insulin sensi- colesevelam (21). Because we did not mea- occurs without any spillover so that cir- tivity in humans is unclear and also de- sure gastric emptying or rate of appearance culating CCK levels did not change (22). serves further investigation. of glucose, we cannot say whether CCK had It is possible that the increase in plasma The finding of a change in glucose any of these effects in our study. Thus, a CCK after meal ingestion was associated tolerance after meal ingestion but not more definitive assessment of this issue with increased duodenal CCK levels and a with intravenous glucose administration deserves further study, and consideration centrally mediated effect to enhance sup- was quite unexpected. This observation should be given to comparing colesevelam’s pression of hepatic glucose output during does suggest that when an intervention effects on solid and liquid meals. the meal. known to alter glucose tolerance is also care.diabetesjournals.org DIABETES CARE, VOLUME 35, MAY 2012 1123 Colesevelam and glucose metabolism known to have an effect via the gastroin- d 7. Kalaany NY, Mangelsdorf DJ. LXRS and Acknowledgments This study was sup- FXR: the yin and yang of cholesterol and testinal tract, it should not be assumed ported in part by the U.S. Department of that changes in insulin sensitivity and fat metabolism. Annu Rev Physiol 2006; Veterans Affairs, National Institutes of Health 68:159–191 b-cell function will be discernable, and Grants DK-007247 and DK-017047. especially on intravenous testing. In fact, 8. Mitro N, Mak PA, Vargas L, et al. The The study was also supported by an nuclear receptor LXR is a glucose sensor. if we had performed only intravenous investigator-initiated proposal sponsored by – fl Nature 2007;445:219 223 studies, we would not have observed the Daiichi Sankyo. No other potential con icts of 9. Laffitte BA, Chao LC, Li J, et al. Activation dissociation of glucose tolerance that was interest relevant to this article were reported. of liver X receptor improves glucose tol- critically dependent on the route of glu- A.L.M. researched data and wrote the man- erance through coordinate regulation of cose administration. We believe this dif- uscript. K.M.U., L.A.W., and S.E.K. researched glucose metabolism in liver and adipose ference in glucose tolerance was due to data, contributed to the discussion, and re- tissue. Proc Natl Acad Sci U S A 2003;100: viewed and edited the manuscript. B.K.M. and 5419–5424 the treatment and not related to the study S.M.M. reviewed and edited the manuscript. design in which the MTT followed the 10. Kim TH, Kim H, Park JM, et al. In- A.L.M. is the guarantor of this work and, as terrelationship between liver X receptor FSIGT. This conclusion is based on such, had full access to all the data in the study the fact that glucose levels at the end of alpha, sterol regulatory element-binding and takes responsibility for the integrity of the protein-1c, peroxisome proliferator- the FSIGT had reached a steady state and data and the accuracy of the data analysis. activated receptor gamma, and small at this time point did not differ between Parts of this study were presented in abstract fi heterodimer partner in the transcriptional placebo and active treatment. Lastly, form at the 71st Scienti c Sessions of the regulation of glucokinase gene expression whether the same outcome would apply American Diabetes Association, San Diego, – – in liver. J Biol Chem 2009;284:15071 in subjects who have normal glucose tol- California, 24 28 June 2011. 15083 erance or type 2 diabetes studied under The authors thank the study participants; 11. Shang Q, Saumoy M, Holst JJ, Salen G, Xu Morgan Arnold and Alison Emery, VA Puget G. Colesevelam improves insulin resis- similar conditions with colesevelam is not Sound Health Care System, for expert tech- known. tance in a diet-induced obesity (F-DIO) rat nical assistance; and John Brunzell, Alan model by increasing the release of GLP-1. We elected to study subjects with Chait, and William Hazzard, University of mild IFG so as to reduce any possible Am J Physiol Gastrointest Liver Physiol Washington,andthelateRobertKnoppfor 2010;298:G419–G424 effects of glucose toxicity on our outcome serving as members of the data safety monitor- 12. Chen L, McNulty J, Anderson D, et al. measures. Of interest, our observation ing committee. Cholestyramine reverses hyperglycemia of a lack of change in insulin sensitivity and enhances glucose-stimulated glucagon- and b-cell function despite glucose low- like peptide 1 release in Zucker diabetic ering with colesevelam suggests that there References fatty rats. J Pharmacol Exp Ther 2010;334: was no glucotoxic effect. Whether the 1. Zieve FJ, Kalin MF, Schwartz SL, Jones 164–170 change in CCK was a consequence of MR, Bailey WL. Results of the glucose- 13. Schwartz SL, Lai YL, Xu J, et al. The ef- the improvement in glucose tolerance or lowering effect of WelChol study (GLOWS): fect of colesevelam hydrochloride on vice versa cannot be answered defini- a randomized, double-blind, placebo- insulin sensitivity and secretion in pa- tively, but we favor the change in CCK controlled pilot study evaluating the effect tients with type 2 diabetes: a pilot study. Metab Syndr Relat Disord 2010;8:179– being the primary event. We also believe of colesevelam hydrochloride on glycemic control in subjects with type 2 diabetes. 188 the lack of change in insulin sensitivity – b Clin Ther 2007;29:74 83 14. Suzuki T, Oba K, Igari Y, et al. Colestimide and -cell function is not related to the 2. Fonseca VA, Rosenstock J, Wang AC, lowers plasma glucose levels and increases study design because we made quantita- Truitt KE, Jones MR. Colesevelam HCl plasma glucagon-like PEPTIDE-1 (7-36) tive measures and have previously dem- improves glycemic control and reduces levels in patients with type 2 diabetes mel- onstrated a treatment effect on glucose LDL cholesterol in patients with inad- litus complicated by hypercholesterolemia. metabolism using this approach (30). equately controlled type 2 diabetes on J Nihon Med Sch 2007;74:338–343 In conclusion, colesevelam improves sulfonylurea-based therapy. Diabetes Care 15. Katsuma S, Hirasawa A, Tsujimoto G. Bile fasting glucose and oral but not intrave- 2008;31:1479–1484 acids promote glucagon-like peptide-1 nous glucose tolerance. Our data suggest 3. Bays HE, Goldberg RB, Truitt KE, Jones secretion through TGR5 in a murine en- MR. Colesevelam hydrochloride therapy teroendocrine cell line STC-1. Biochem that this effect of colesevelam is indepen- Biophys Res Commun 2005;329:386– dent of changes in insulin sensitivity, in patients with type 2 diabetes mellitus b treated with metformin: glucose and lipid 390 -cell function, and plasma incretins. effects. Arch Intern Med 2008;168:1975– 16. Bergman RN, Ider YZ, Bowden CR, Increased plasma CCK concentrations 1983 Cobelli C. Quantitative estimation of in- observed during treatment may be con- 4. Goldberg RB, Fonseca VA, Truitt KE, Jones sulin sensitivity. Am J Physiol 1979;236: tributing to postprandial glucose control MR. Efficacy and safety of colesevelam E667–E677 via a delay in gastric emptying. Further- in patients with type 2 diabetes mellitus 17. Kahn SE, Prigeon RL, McCulloch DK, more, given the reduction in fasting glu- and inadequate glycemic control receiving et al. Quantification of the relationship cose together with the decrease in the insulin-based therapy. Arch Intern Med between insulin sensitivity and beta-cell postprandial glucose excursion, the effect 2008;168:1531–1540 function in human subjects. Evidence for a 5. Holt JA, Luo G, Billin AN, et al. Definition hyperbolic function. Diabetes 1993;42: of colesevelam and other bile acid seques- 1663–1672 trants may be occurring through changes of a novel growth factor-dependent signal cascade for the suppression of bile acid bio- 18. Bardet S, Pasqual C, Maugendre D, Remy JP, in hepatic glucose production. Thus, fur- synthesis. Genes Dev 2003;17:1581–1591 Charbonnel B, Sai P. Inter and intra ther studies are needed to elucidate the 6. Lefebvre P, Cariou B, Lien F, Kuipers F, individual variability of acute insulin mechanism(s) by which altering bile Staels B. Role of bile acids and bile acid response during intravenous glucose tol- acid metabolism modulates glucose receptors in metabolic regulation. Physiol erance tests. Diabete Metab 1989;15:224– metabolism in humans. Rev 2009;89:147–191 232

1124 DIABETES CARE, VOLUME 35, MAY 2012 care.diabetesjournals.org Marina and Associates

19. Chandra R, Liddle RA. Neural and hor- 23. Kogire M, Gomez G, Uchida T, Ishizuka J, 27. Koide M, Okabayashi Y, Otsuki M. Role monal regulation of pancreatic secretion. Greeley GH Jr, Thompson JC. Chronic of endogenous bile on basal and post- Curr Opin Gastroenterol 2009;25:441– effect of oral cholestyramine, a bile salt prandial CCK release in humans. Dig Dis 446 sequestrant, and exogenous cholecysto- Sci 1993;38:1284–1290 20. Liddle RA, Rushakoff RJ, Morita ET, kinin on insulin release in rats. Pancreas 28. Otsuki M. Pathophysiological role of cho- Beccaria L, Carter JD, Goldfine ID. Physio- 1992;7:15–20 lecystokinin in humans. J Gastroenterol logical role for cholecystokinin in reducing 24. Koop I, Fellgiebel A, Koop H, Schafmayer Hepatol 2000;15(Suppl.):D71–D83 postprandial hyperglycemia in humans. A, Arnold R. Effect of cholestyramine on 29. Garg SK, Ritchie PJ, Moser EG, Snell- J Clin Invest 1988;81:1675–1681 plasma cholecystokinin and pancreatic Bergeon JK, Freson BJ, Hazenfield RM. 21. Odunsi-Shiyanbade ST, Camilleri M, polypeptide levels, and exocrine pancre- Effects of colesevelam on LDL-C, A1c and McKinzie S, et al. Effects of chenodeox- atic secretion. Eur J Clin Invest 1988;18: GLP-1 levels in patients with type 1 di- ycholate and a bile acid sequestrant, co- 517–523 abetes: a pilot randomized double-blind lesevelam, on intestinal transit and bowel 25. Liddle RA. Regulation of cholecystokinin trial. Diabetes Obes Metab 2011;13:137– function. Clin Gastroenterol Hepatol secretion by intraluminal releasing fac- 143 2010;8:159–165 tors. Am J Physiol 1995;269:G319–G327 30. Utzschneider KM, Tong J, Montgomery B, 22. Cheung GW, Kokorovic A, Lam CK, 26. Koop I, Schindler M, Bosshammer A, et al. The dipeptidyl peptidase-4 inhibitor ChariM,LamTK.Intestinalcholecysto- Scheibner J, Stange E, Koop H. Physiolog- vildagliptin improves beta-cell function kinin controls glucose production through ical control of cholecystokinin release and and insulin sensitivity in subjects with a neuronal network. Cell Metab 2009;10: pancreatic enzyme secretion by intraduo- impaired fasting glucose. Diabetes Care 99–109 denal bile acids. Gut 1996;39:661–667 2008;31:108–113

care.diabetesjournals.org DIABETES CARE, VOLUME 35, MAY 2012 1125