|Home | About | Journals | Submit | Contact Us | Français|
Colesevelam hydrochloride (COL), a bile acid sequestrant indicated as an adjunct to diet and exercise for reducing low-density lipoprotein cholesterol (LDL-C) in patients with primary hypercholesterolemia, was shown in a pilot study to reduce both glycated hemoglobin (HbA1c) and LDL-C in patients with type 2 diabetes mellitus (T2DM). Three double-blind, placebo-controlled trials in T2DM have now independently confirmed the HbA1c and LDL-C reductions with COL. In each of the primary studies, a significant mean treatment difference in HbA1c (−0.54%, −0.50%, and −0.54%) and LDL-C (−15.9%, −12.8%, and −16.7%) resulted from the addition of 3.75 grams/day of COL to existing metformin, insulin, or sulfonylurea-based therapy, respectively, in patients with T2DM inadequately controlled on their current antidiabetic regimen. Here we report the results of a pooled analysis of data for the 1018 patients included in the three primary studies. By study end, HbA1c, fasting plasma glucose (FPG), LDL-C, total cholesterol (TC), non-high-density lipoprotein cholesterol (HDL-C), apolipoprotein B (ApoB), and high-sensitivity C-reactive protein (hsCRP) were significantly reduced with COL versus placebo. Triglyceride (TG) and ApoA-I were significantly increased in the COL group relative to placebo. HDL-C did not change in either group, and the between-group treatment difference was small and not significant. Results of this pooled analysis are consistent with results reported previously in each of the primary COL studies and indicate that the HbA1c and LDL-C-lowering effects of COL are consistent, occurring regardless of whether COL is added to metformin, insulin, or sulfonylurea-based therapy. In conclusion, COL represents a novel therapeutic option by significantly lowering both LDL-C and HbA1c in patients with T2DM, two important treatment goals to forestall vascular complications.
Colesevelam HCl (COL), a bile acid sequestrant indicated as an adjunct to diet and exercise for reducing low-density lipoprotein cholesterol (LDL-C) in patients with primary hypercholesterolemia, was shown in a pilot study to reduce both LDL-C and glycated hemoglobin (HbA1c) in patients with type 2 diabetes.1 Three double-blind, placebo-controlled pivotal trials have now independently confirmed the LDL-C and HbA1c reductions with COL in patients with type 2 diabetes.2–4 In each of the primary studies, a significant mean treatment difference in LDL-C (15.9%, 12.8%, and 16.7%) and HbA1c (0.54%, 0.50%, and 0.54%) resulted from the addition of 3.75 grams/day of COL to existing metformin, insulin, or sulfonylurea-based therapy, respectively, in patients with inadequately controlled type 2 diabetes (T2DM) (HbA1c 7.5–9.5%, inclusive) on their current antidiabetes regimen. The results of the pivotal trials led to the recent Food and Drug Administration approval of a new indication for COL as an adjunct to diet and exercise for improving glycemic control in adult patients with T2DM. In this report, we present the pooled data of all three studies.
This was a past-hoc analysis of the pooled data from all patients included in the intent-to-treat (ITT) populations in the three pivotal COL trials (n=1018). This included 301 patients from the COL+metformin trial, 280 from the COL+insulin trial, and 437 from the COL+sulfonylurea trial. In the pivotal trials, COL or placebo was added on to established antidiabetes therapy. As such, the study medication represented a second-line antidiabetic therapy (if added to existing antidiabetes monotherapy) or a third-line or fourth-line agent, etc. (if added to existing antidiabetes combination therapy). Analyses of the glycemic and lipid effects of COL were stratified by background antidiabetes therapy to determine if a differential response to COL occurred, depending on whether this agent was added to existing antidiabetes monotherapy or combination therapy. As such, the current analyses evaluated the effect of COL compared with placebo in: (1) the total ITT population across all three pivotal trials (n=1018); (2) those who received study medication as a second-line agent (n=414); and (3) those who received study medication as a third-line, fourth-line agent, etc. (n=604). In addition, analyses were conducted to determine the effects of COL versus placebo treatment in patients who were on concomitant statin treatment at baseline of the pivotal trials, as well as stratified by age, race, and gender. The details with respect to study design analyses, adverse reactions, etc., are presented in the three original reports.2–4
As shown in Table 1, the demographic and baseline characteristics for the placebo and COL groups were comparable. By end of study, mean HbA1c was significantly reduced with COL versus placebo treatment (between-group treatment difference −0.54%; P<0.0001; Fig. 1). Also, mean fasting plasma glucose (FPG) was significantly reduced with COL versus placebo (treatment difference −15.1mg/dL; P<0.0001; Fig. 2). COL therapy resulted in a significant reduction in both total and LDL-C compared with placebo treatment (treatment difference −5.15% and −15.3%, respectively; P<0.0001; Fig. 3). By study end, median triglyceride (TG) was significantly increased in the COL group relative to placebo (treatment difference 15.0%; P<0.0001). However, at study end, mean non-high-density lipoprotein cholesterol (HDL-C) and apolipoprotein B (ApoB) levels were significantly reduced with COL versus placebo (treatment difference −6.80% and −6.6%, respectively; P<0.0001). There was no significant effect on mean HDL-C levels between the two groups at study end (treatment difference 0.02%), but mean apoA-I levels increased significantly in the COL group relative to placebo (treatment difference 2.8%; P<0.0001). Median levels of high-sensitivity C-reactive protein (hsCRP) were significantly reduced with COL relative to placebo treatment (treatment difference −0.4mg/L; P=0.0009).
With respect to the subgroup analyses, the effect seen for the total group was also manifest in the monotherapy group with regard to significant reductions in hemoglobin HbA1c, FPG, and LDL-C (Figs. 1–3). Median TG levels were significantly increased in the COL group relative to placebo (treatment difference 12.3%; P=0.0013). There was a significant treatment difference in total cholesterol (TC), non-HDL-C, and ApoB levels (treatment difference −4.84%, −6.17%, and −4.98%; P=0.003, P=0.005, and P=0.009, respectively). The reduction in hsCRP was not significant (treatment difference −0.2mg/L).
In the combination treatment group, as shown for the total cohort at study end, there was significant reduction in mean HbA1c, FPG, and LDL-C levels (P<0.0001; Figs. 1–3). Although there was a significant increase in median TG levels of COL treatment compared to placebo at study end (treatment difference 16.6%; P<0.0001), TC, non-HDL-C, and ApoB levels were reduced significantly (treatment difference −5.4%, −7.2%, and −7.7%, respectively; P<0.0001 for all). ApoA-I levels increased significantly with COL versus placebo treatment (treatment difference 3.4%; P<0.0001). By study end, median hsCRP was significantly reduced with COL treatment compare to placebo (treatment difference −0.5mg/L; P=0.0027).
COL therapy resulted in a significant reduction in HbA1c (−0.45%; P<0.0001) and LDL-C (−15.6%; P<0.0001) in patients on statin therapy at baseline.
The benefit with regard to reduction in HbA1c and LDL-C was manifest in both patients under 65 years and those ≥65 years (P<0.0001). When patients were stratified by race, a significant treatment difference in least squares (LS) mean HbA1c was reported between the COL and placebo in all groups (−0.48% [Caucasians], −0.54% [Hispanics], and −0.77% [African Americans]; P<0.001 for all). The treatment difference in LS mean percent LDL-C was −16.2% for Caucasians, −11.3% for Hispanics, and −19.6% for African Americans (P<0.0001 for all). The reduction in LDL-C and HbA1c was manifest in both male and female subgroups.
Overall, in all three studies, COL treatment was safe and well tolerated by the diabetic patients. In the substudy of patients on insulin therapy, the most frequently reported (incidence >1%) drug-related adverse events were constipation (6.8%), dyspepsia (3.4%), hypoglycemia (3.4%), flatulence (2.0%), and nausea (1.4%). The incidence of hypoglycemia in the placebo group was 5.7%. There was no significant weight gain noted for either group by study end. In the metformin-treated substudy, the most frequent adverse reaction that occurred in >5.0% of the subjects was constipation (6.9%) in the COL group. In the study, in which COL was added to sulfonylurea-based therapy, the most common adverse reactions that occurred in the COL group compared to placebo were constipation (7.9%), diarrhea (3.5%), dyspepsia (3.9%), nausea (3.9%) and vomiting (3.1%). The frequency of influenza (3.1%), upper respiratory tract infection (7.4%), and urinary tract infection (4.4%) was greater. Furthermore, frequency of hypoglycemia was more common (2.6%). No weight gain was noted in either group.
The major cause of morbidity and mortality in patients with T2DM is cardiovascular disease. It has previously been shown in both the Collaborative Atorvastatin Diabetes Study (CARDS)5 and Heart Protection Study6 that statin therapy results in significant reduction in LDL-C and cardiovascular events. However, statin therapy does not appear to have a benefit with regard to glycemia. Also, recently the UK Prospective Diabetes Study (UKPDS) Study7 has reported a legacy effect with good early glycemic control and reduction in future cardiovascular events. Thus, therapies targeting both LDL-C and glycemia become attractive strategies for forestalling vascular complications in patients with diabetes. In the Glucose Lowering Effect of Welchol Study (GLOWS) study, it was previously reported by Zieve et al. that COL therapy resulted in a reduction in both LDL-C and HbA1c.1 However, this study suffered from a small sample size of n=65. In the present analysis, which includes 1018 patients, the effects seen in the GLOWS study were confirmed in patients on metformin therapy, sulfonylurea therapy, or insulin therapy. In all three of these studies in monotherapy and combination thereapy, COL resulted in a significant reduction in both LDL-C and HbA1c. With regard to the side-effect profile in this pooled analysis, it appears that the major side effect that occurred in all three studies was gastrointestinal side effects, mainly constipation. Thus, this pooled analysis of large sample size confirms the efficacy of COL therapy as an attractive strategy in patients with diabetes who are not at the HbA1c goal of <7.0% or LDL-C goal of <100mg/dL. The advantage of the COL therapy is that it can be used safely in combination with the other antidiabetic therapies, as shown in these three studies in this pooled analysis.2–4 Furthermore, the efficacy of combination of COL with statin therapy is clearly safe and results in greater reduction in LDL and non-HDL-C.
It is important to appreciate in these three studies that in addition to reduction in LDL-C with COL therapy, there was significant reduction in non-HDL-C and ApoB levels in spite of the modest increase in plasma TGs because both LDL-C and non-HDL-C are targets for treatment.8 It is also possible in future guidelines that ApoB will also become a target for treatment. Whereas the reduction in LDL is related to bile acid sequestration induced by COL, the precise mechanism for the glucose-lowering effect remains to be elucidated. Numerous mechanisms have been proposed; the most attractive appears to be with respect to the interactions of bile acids with nuclear receptors. Bile acids are endogenous ligands of the farnesoid X receptor. This might be a potential mechanism for the reduction in glucose levels.9 The modest increase in TG of 15.0% should be interpreted in conjunction with the benefit of COL therapy in reducing LDL-C, non-HDL-C, and ApoB levels.
Recently, the Jupiter study10 showed the benefit of statin therapy in patients with elevated CRP and normal LDL-C. Also, there is much data suggesting that CRP could contribute to atherothrombosis.11 It has been previously been shown12 and confirmed in this pooled analysis that COL reduces CRP levels. In the future, CRP could emerge as a target for treatment; it is elevated in patients with diabetes. This would make COL therapy even more attractive because, in addition to reduction in LDL-C, ApoB, non-HDL-C, and HbA1c, it would help in reduction in CRP levels.
In conclusion, this pooled analysis confirms that COL therapy in patients with T2DM results in a significantly greater reduction in both LDL-C and HbA1c and should seriously be considered as a viable strategy in obtaining the goals recommended by American Diabetes Association in type 2 diabetic patients.
This study was supported by the National Institutes of Health (I.J.; K24 AT 00596). Editorial assistance for the preparation of this manuscript was provided by Karen Stauffer, Ph.D., and Manpreet Kaur.