Cosecretion of amylin with insulin (in an approximate 1:100 molar ratio) from the β-cells of pancreatic islets has been recognized for a long time. Thus, insulinopenic patients with type 1 diabetes mellitus (T1DM) do not secrete amylin after food ingestion. Secretion of amylin in those with type 2 diabetes mellitus (T2DM) mimics that of insulin: “too little, too late”. Clinicians had attempted to correct only hypoinsulinemia because the molecular structure of amylin and its mechanism of action (it exerts its effects by specific binding to its cognate receptor in the central nervous system) had not been elucidated until the late 1990s. It was not until March 2005 that the administration of pramlintide (Symlin®; Amylin Pharmaceuticals Inc., San Diego, CA, USA) was approved by the US Food and Drug Administration (FDA) (www.fda.gov) for use in the US in patients with T1DM and T2DM, in conjunction with administration of prandial insulin to improve postprandial glycemic control. Figure 1 shows the amino acid sequences of native human amylin and synthetic pramlintide, with three amino acid differences in their primary structure. Native human amylin is poorly soluble and self-aggregating, which make it unsuitable for clinical use. The amino acid modifications of pramlintide acetate make it a stable bioactive peptide analog useful for clinical purposes.
Preapproval studies of pramlintide in both T1DM and T2DM as an adjunct to meal-time insulin injections have been summarized in several review articles.[12, 13] The results of typical experiments in T1DM and T2DM patients from these studies are summarized in Tables 1 and 2.[14-22] It is evident that the major effect of pramlintide is on the reduction of postprandial glucose excursions.[18, 23-25] That short-term effect was amply demonstrated in the initial placebo-controlled studies in patients with both T1DM and T2DM. The critical question for practitioners is whether these acute effects will translate into long-term glycemic benefits. A brief overview of some of the clinical trials is given below.
Table 1. Selected studies of pramlintide in patients with type 1 diabetes
|Study||No. patients||Study duration (weeks)||Pramlintide (μg)||ΔHbA1c (%)||ΔWeight (kg)|
|Buse et al.||586||26||60, t.i.d.||–0.20||–1.6|
|90, b.i.d.||–0.10 (ns)||–0.7|
|90, t.i.d.||–0.10 (ns)||–1.6|
|Fineman et al.||479||52||60, t.i.d.||–0.40||–1.4|
|Hollander et al.||480||52||30, q.i.d.||–0.39||–0.5|
|Hollander et al.||479||52||60, t.i.d.||–0.29||–0.4|
Table 2. Studies of pramlintide in patients with type 2 diabetes
|Study||No. patients||Study duration (weeks)||Pramlintide (μg)||ΔHbA1c (%)||ΔWeight (kg)|
|Gottlieb et al.||203||4||30, q.i.d.||–0.53||–0.36|
|Gottlieb et al.||499||26||90, b.i.d.||–0.30||–0.80|
|Kolterman et al.||538||52||30, t.i.d.||–0.30||–0.30|
|Hollander et al.||656||52||60, t.i.d.||–0.35||–0.50|
|Thompson et al.||498||26||120, b.i.d.||–0.41||–1.80|
Patients with T2DM (baseline HbA1c 9.1 ± 1.2%) were enrolled in a 52-week double-blind placebo-controlled parallel-group study and were treated with insulin (alone or in combination with sulfonylureas and/or metformin) and randomized to receive additional preprandial subcutaneous injections of either placebo or pramlintide (60 μg, t.i.d.; 90 μg, b.i.d.; or 120 μg, b.i.d.). Treatment with pramlintide 120 μg, b.i.d., resulted in a reduction from the rather high baseline in HbA1c by 0.68% and 0.62% at Weeks 26 and 52, respectively, which was significantly greater than that seen with placebo. The glycemic improvement with pramlintide 120 μg, b.i.d., was accompanied by a mean weight loss of –1.4 kg (compared with a gain of +0.7 kg with placebo) at Week 52 (P < 0.05) and occurred without an overall increase in the rate of severe hypoglycemia.
Reductions in postprandial glucose excursions by pramlintide also translated into sustained reductions in HbA1c in patients with T1DM. In a study of subjects with T1DM, pramlintide treatment led to a mean reduction in HbA1c of 0.67% from baseline (mean 8.7%) to Week 13, which was significantly greater than the reduction in HbA1c in patients randomized to receive placebo (0.16%). A significant placebo-corrected treatment difference was sustained through Week 52. The rate of severe hypoglycemia was not increased by pramlintide. In another study of patients with T1DM, the addition of pramlintide 60 μg, t.i.d. or q.i.d., to insulin led to a small but significant reduction in HbA1c (from the high baseline of 8.9 ± 1.0%) of 0.29% and 0.34%, respectively, over 52 weeks. This was seen without an increase in insulin doses in the pramlintide-treated subjects. Within the subset of patients with an entry HbA1c between 7.0% and 8.5% (∼28% of all patients enrolled in three long-term studies), 196 were treated with placebo + insulin (baseline HbA1c 7.9 ± 0.4%, body weight 76.0 ± 14.3 kg) and 281 were treated with pramlintide + insulin (baseline HbA1c 7.9 ± 0.4%, body weight 75.4 ± 13.1 kg). The addition of pramlintide resulted in significant reductions in HbA1c and body weight from baseline to Week 26 (0.3% and 1.8 kg, placebo-corrected treatment differences, respectively; both P ≤ 0.0009). These changes occurred without an increase in the overall risk of severe hypoglycemia.
In summary, the 26–52-week trials in patients with T1DM showed HbA1c lowering in the range 0.3%–0.6% from the rather high baseline of approximately 8.9%. The doses tested ranged from 30 to 90 μg, t.i.d. There was a mean weight loss of 0.9–1.4 kg. Hypoglycemia was rare and its incidence decreased with time. Similarly, studies of pramlintide in patients with T2DM lasted 26 or 52 weeks and resulted in a decline in HbA1c of 0.4%–0.7% from a baseline of approximately 9.1%. In general, weight loss was in the range 0.9–1.4 kg.
After approval of pramlintide in the US, several potentially relevant insights came from results of studies of subcutaneous continuous infusion of pramlintide in patients with T1DM and in combination with basal insulin in subjects with T2DM. For example, in a 24-h study, 13 adolescents with T1DM participated in a randomized, controlled crossover design trial that compared subcutaneous insulin therapy with insulin and pramlintide infusion. Pramlintide was delivered by a separate pump. The addition of pramlintide led to a 26% reduction in postprandial hyperglycemia and reduced postprandial glucagon concentrations. In a longer 16-week open-label study in which 11 patients with T1DM were enrolled (mean age 39 years, baseline HbA1c 8.2%, body mass index [BMI] 29.7 kg/m2), pramlintide was delivered in a continuous basal-bolus subcutaneous infusion. Subjects continued on their long-standing subcutaneous continuous insulin infusion therapy via a different pump. After 16 weeks HbA1c declined to a mean value of 7.8%, fasting plasma glucose (FPG) had decreased from 198 to 136 mg/dL, and weight was reduced by 0.5 kg. On average, patients reduced their bolus insulin by 20% by the end of the study while basal insulin was unchanged. There were no episodes of severe hypoglycemia documented.
Among patients with T2DM, three studies will be singled out. First, 212 patients with poorly controlled diabetes (baseline mean HbA1c 8.5%) on treatment with insulin glargine with or without concomitant oral hypoglycemic agents (OHA) were randomized in a double-blind fashion to either pramlintide (60 or 120 μg, b.i.d. or t.i.d.) or placebo for 16 weeks. The primary endpoint was a composite score of achieving HbA1c ≤7%, HbA1c reduction of ≥0.5%, mean daily postprandial glucose increments of ≤40 mg/dL, no increase in body weight, and no severe hypoglycemia. This composite endpoint was achieved by more patients in the pramlintide group (25% vs 7%); HbA1c was reduced by 0.7% and the subjects lost, on average, 1.6 kg. All these parameters were significantly better than those achieved by patients randomized to the placebo arm of the study. Second, 113 patients with uncontrolled T2DM (baseline HbA1c 8.2%) on basal insulin and OHA were randomized for 24 weeks to either meal-time pramlintide (120 μg, t.i.d.) or to titrated rapid-acting insulin analogs. The primary endpoint of that study was the proportion of subjects reaching HbA1c ≤7% without weight gain or severe hypoglycemia, and occurred in more patients on pramlintide than on prandial insulin (30% vs 11%, respectively). Both therapies achieved the same improvements in HbA1c and FPG, but patients on insulin gained weight and had more mild-to-moderate hypoglycemia, whereas more patients in the pramlintide group complained of nausea (21%). Third, a retrospective chart review of clinical and laboratory data over 24 weeks of pramlintide therapy of 92 insulin-treated patients with T2DM was conducted and showed significant declines in HbA1c (from 8.3% to 7.9%), weight (from 104.4 to 103.2 kg) and low-density lipoprotein–cholesterol levels. This improvement occurred despite reduced prandial insulin and adjustment of basal insulin and OHA.
In a interesting post-hoc analysis presented at the 2012 Scientific Session of the American Diabetes Association (ADA), data were gathered from seven-point glucose profiles in patients with T2DM on pramlintide. The proportion of pre- and postprandial readings that fell above, below, or within glycemic targets in an open-label, 6-mont clinical practice trial showed a shift to more favorable glycemia (more euglycemia, less hyperglycemia) by Week 4 and was sustained through the end of the study. The percentage of measurements within the ADA-defined “euglycemic” range (i.e. preprandial 70–130 mg/dL, postprandial 70–180 mg/dL) increased from 37.2 ± 2.6% at baseline to 54.6 ± 3.1% during the final time period (4–6 months; P < 0.0001).
Adverse events seen most commonly in patients with T1DM and T2DM injected with pramlintide have been gastrointestinal in nature. Specifically, for those with T1DM, the adverse events were nausea (initially reported by 37%–48% patients in different studies), anorexia (0%–17%) and vomiting (7%–11%).[14-22] In patients with T2DM, nausea was reported initially by approximately 30% patients, headache by 5%–13%, anorexia by 0%–9%, and vomiting by 7%–8%.[14-22] These events were usually mild-to-moderate in nature, dose dependent, and dissipated over time.
Clinical recommendation for use of pramlintide in T1DM and T2DM is to use it (subcutaneously via a pen device) only alongside prandial injections of rapid-acting insulin for meals containing at least 30 g carbohydrates (www.fda.gov). Given the propensity to cause nausea, the dose of pramlintide is titrated gradually. Patients with T1DM are more sensitive to its action and start at a dose of 15 μg, t.i.d., pramlintide and increase it every 3 days (provided there is no nausea) by increments of 15 μg until they reach the maximum recommended dose of 60 μg, t.i.d. Patients with T2DM start at a dose of 60 μg, t.i.d., pramlintide and increase the dose to 120 μg, t.i.d., after 3 days. It is also recommended to initially decrease the dose of prandial insulin (not basal insulin) by 25–50% to minimize chances of hypoglycemia when pramlintide is added. This recommendation is usually relevant for those insulin-sensitive patients with T1DM who are relatively closer to glycemic targets. Patients with T2DM who are poorly controlled typically do not need prandial insulin decreases. Of course, the dose of insulin is uptitrated once a stable dose of pramlintide is reached.