ORIGINAL ARTICLES

Pramlintide: A Human Amylin AnalogueReduced Postprandial Plasma Glucose,Insulin, and C-peptide Concentrationsin Patients with Type 2 DiabetesR.G. Thompson1, A. Gottlieb1, K. Organ1, J. Koda1, J. Kisicki2, O.G. Kolterman1

1Amylin Pharmaceuticals, Inc., 9373 Towne Centre Drive, SanDiego, California, USA2Harris Laboratories, Inc., 624 Peach Street, Lincoln, Nebraska, USA

In order to determine the influence of a 5 h infusion of pramlintide compared to placeboon postprandial glucose, lactate, insulin, and C-peptide concentrations in patients withType 2 diabetes, a single-blind, randomized, cross-over study was conducted in 24 patients;12 treated with exogenous insulin and 12 managed with diet and/or oral hypoglycaemicagents. One hour after initiation of infusion, patients consumed a Sustacal test meal.The protocol was repeated on the following day with each patient receiving the alternatestudy medication. Pramlintide infusion in the insulin-treated patients resulted in statisticallysignificant reductions in mean glucose, insulin, C-peptide, and lactate concentrationsduring the 4-h period after the Sustacal test meal. Pramlintide infusion also resulted insignificant reductions of mean insulin, C-peptide, and lactate concentrations, but notglucose concentrations, in the patients treated with diet and/or oral hypoglycaemic agents.Within this latter group, reduction in postprandial glucose concentrations in individualpatients correlated with glycated haemoglobin values. These results suggest that adminis-tration of pramlintide may improve glycaemic control in patients with Type 2 diabetestreated with insulin or poorly controlled on diet and/or oral hypoglycaemic agents. 1997by John Wiley & Sons, Ltd.

Diabet. Med. 14: 547–555

No of Figures: 6. No of Tables: 1. No of Refs: 25

KEY WORDS Type 2 diabetes Glucose C-peptide Amylin Pramlintide

Received 2 October 1996; revised 7 February 1997; accepted 20 February 1997

plasma glucose after both mixed meals and standardizedIntroductionliquid Sustacal test meals in Type 1 diabetes.6 Morerecently in Type 1 diabetes, reductions in mean 24 hAmylin is markedly reduced or undetectable in plasma

from patients with Type 1 (insulin-dependent) diabetes.1 and postprandial glucose concentrations were achievedat doses of pramlintide that produced plasma concen-This peptide hormone is secreted by pancreatic b-cells

in conjunction with insulin in response to nutrient trations of pramlintide similar to circulating amylinconcentrations observed in control subjects.7stimuli2 and has been shown in animals to modulate

gastric emptying and reduce postprandial glucose concen- Amylin concentrations, both fasting and stimulated,vary in patients with Type 2 (non-insulin dependent)trations.3 Pramlintide (25, 28, 29 pro-human amylin) is

a synthetic analogue of human amylin that has the diabetes depending on disease progression and degreeof obesity. This is similar to the variation in insulindesired biological activities of human amylin but is

substantially different from and better than human amylin concentrations, but the relationship between the two hasnot been defined. Fasting amylin concentrations inin that it does not self-aggregate. Intravenous infusions

of pramlintide have been shown to modulate gastric patients with Type 2 diabetes not on insulin werereported to be higher than in patients on insulin therapy.8emptying of both liquids and solids4 and also reduce

postprandial glucose excursions in patients with Type 1 Stimulated amylin concentrations were higher in obesecompared with non-obese patients with Type 2 diabetes,8diabetes.5 Self-administered subcutaneous injections of

pramlintide have resulted in significant reductions of and were significantly lower in subjects with overt Type2 diabetes compared with a group of subjects withnormal glucose tolerance matched for obesity.9 Therefore

Abbreviations: AUC area under the curve amylin concentrations are abnormal in a subset of* Correspondence to: Dr Robert G. Thompson, Amylin Pharmaceutical,

patients with Type 2 diabetes and lower than in subjectsInc., 9373 Towne Centre Drive, San Diego, CA 92121, USASponsors: Amylin Pharmaceuticals, Inc.; Harris Laboratories, Inc. with normal glucose tolerance. A randomized, single-

547CCC 0742–3071/97/070547–09$17.50 1997 by John Wiley & Sons, Ltd. DIABETIC MEDICINE, 1997; 14: 547–555

ORIGINAL ARTICLESblinded, cross-over study using a 5-h infusion of either Blood samples were obtained every 15 min throughout

the study from an indwelling catheter. The patientsplacebo or pramlintide was conducted in 24 patientswith Type 2 diabetes to examine the influence of consumed their usual lunch after completion of the

infusion of study medication. This protocol was repeatedpramlintide on plasma glucose, insulin, C-peptide, andlactate concentrations. on the following day with each patient receiving the

alternate study medication.Patients and Methods

Laboratory MethodsPatients

Glucose concentrations were assayed on a TechniconAxon Chemistry Analyzer (Miles Inc., Tarrytown, NY)Recruitment qualifications included patients with a docu-

mented history of Type 2 diabetes for not less than 2 using the hexokinase method.10 Amylin concentrationswere measured using two separate monoclonal based,and not more than 10 years’ duration. All participants

had normal baseline clinical and laboratory parameters sensitive two-site immunofluorescent assays for amylinand amylin-like peptides in human plasma (for details,(except for plasma glucose), stable weight within 30 %

of their desirable weight (Metropolitan Life Tables), no see Percy et al.11). These assays differ only in theircapture antibodies. One assay employs monoclonalclinically significant history of renal or other chronic

diseases, and no evidence of unstable diabetic retino- antibody F002-2 which has been shown to bind to bothhuman amylin and high molecular weight (HMW) amylin-pathy. Patients were recruited into two separate groups

based upon their current diabetes therapy. In one group, like-peptides.12 The HMW amylin-like peptides havebeen reported to consist of two distinct peaks on reverse12 patients were using exogenous insulin (INS) as well

as diet while in the other group, 12 patients were using phase HPLC which are distinct from the amylin peakand have higher molecular weights.12,13 The other assaydiet and/or hypoglycaemic agents (D/OHA) but not

exogenous insulin. The study was approved by the local employs monoclonal antibody F024-4.4 which is specificfor the intact amylin molecule as it does not bind to theInstitutional Review Board and all patients signed an

informed consent. additional peaks of amylin in human plasma.12 Theassays had an average intra-assay CV of less than 10 %and an average inter-assay CV of less than 15 %.Infusion Study

Plasma pramlintide concentrations were measured intriplicate in a specific two-site immunoradiometric assayPatients were admitted to an inpatient unit for 1 day of

acclimatization and 2 days of study. Pramlintide and that employs two monoclonal antibodies.14 To performthe immunoassay, the monoclonal capture antibodyplacebo (50 mg mannitol per ml and 0.2 mg Tween 80

per ml) were administered according to a randomized, F024-4.4 was bound to the solid phase (Immulon II 96-well plates) in carbonate buffer overnight. After thebalanced, single-blind, cross-over design. The randomiz-

ation schedule, prepared by Harris Laboratories Inc., incubation, the plate was washed with phosphate-buffered saline (PBS)/Tween and blocked with 1 % non-randomized the 24 patients to one of four groups:

insulin-treated receiving pramlintide followed by placebo, fat dry milk. After another wash, sample or standard isadded and the plate was incubated again. The plate wasinsulin-treated receiving placebo followed by pramlintide,

non-insulin-treated receiving pramlintide followed by washed again and iodinated monoclonal antibody F025-27, which binds to the amidated C-terminal region ofplacebo, and non-insulin-treated receiving placebo fol-

lowed by pramlintide. Six patients were randomized to pramlintide, was added and incubated. After a finalwash, the strips were counted well-by-well.each group.

After an overnight fast, intravenous lines were inserted Plasma insulin concentrations were measured with theIncstar Radioimmunoassay Kit (Stillwater, MN) usingin the antecubital fossae of both arms and maintained

for 6.5 h on each study day. One line was used for iodinated tracer and second antibody precipitation.15 C-peptide assays were completed by Endocrine Sciencesstudy drug administration and the other for blood

sampling. Thirty minutes after initiation of sampling, a (Calabasas, California). Samples were pre-treated withPEG to eliminate potential interference fromconstant infusion of placebo or pramlintide at a dose of

100 mg h−1 was started and maintained for 5.0 h. The insulin/proinsulin antibodies. The detection antibodyused was developed against the 31 amino acid sequencepatients’ usual morning dose of insulin or oral hypoglyca-

emic agent was administered 30 min after the start of of human C-peptide. Bound and unbound fractions wereseparated by the second antibody-PEG method. Sensitivitythe infusion of study drug. Thirty minutes later (1 h after

initiation of the infusion), the patients consumed a of the assay was 0.05 ng ml−1.Lactate was measured on a Technicon Axon ChemistrySustacal test meal (7 kcal kg−1 body weight). Sustacal

is a mixture of carbohydrate (sugar and corn syrup), fat Analyzer (Miles Inc., Tarrytown, NY) using lactateoxidase. The hydrogen peroxide produced by this reaction(partially hydrogenated soy oil), and protein (sodium and

calcium caseinate and soy protein isolate) providing was coupled to a chromogen using peroxidase and thereaction was monitored at 540 nm.16 Whole blood55 %, 21 %, and 24 % of total calories, respectively.

548 R.G. THOMPSON ET AL.

ORIGINAL ARTICLESTable 1. Demographic characteristicssamples containing EDTA were haemolysed for the

HbA1c assay. HbA1c levels were measured by CorningAll patients Insulin Diet/OHASciCor Laboratories (Indianapolis, IN) using a Biorad

Variant HPLC system, with an assay range of 3.6 %Number 24 12 12

to 17.8 %. Age (yr)a 45.5 ± 1.5 46.8 ± 2.0 44.3 ± 2.3Height (cm)a 175.6 ± 2.1 176.4 ± 2.8 174.9 ± 2.6Weight (kg)a 87.5 ± 2.3 86.4 ± 2.7 88.6 ± 3.7Statistical AnalysesBMI (kg m−2)a 28.9 ± 0.6 28.4 ± 1.0 29.4 ± 0.7Duration of diabetes (yr)a 5.3 ± 0.4 5.0 ± 0.5 5.5 ± 0.6The glucose, lactate, insulin, and C-peptide profilesFasting glucose 8.9 ± 0.6 9.4 ± 0.9 8.5 ± 0.9

following the Sustacal meal (hours 1–5 of the profile) (mmol l−1)awere used to determine the effect of pramlintide compared HbA1c(%)a 8.15 ± 0.44 8.75 ± 0.43 7.55 ± 0.74

Sexto placebo in subjects with Type 2 diabetes mellitus.Male (n, %) 14 58.3 7 58.3 7 58.3Data were analysed for all patients combined (allFemale (n, %) 10 41.7 5 41.7 5 41.7patients), the INS group and the D/OHA group. The

Treatmentincremental plasma glucose AUC was defined as the Insulin onlyb 10 10 0area under the curve corrected for the 1 h time point Insulinb + OHAc 2 2 0

Diet + OHAc 7 0 7(i.e. plasma glucose concentration at the time of ingestionDiet only 5 0 5of the Sustacal meal subtracted from all subsequent

glucose concentrations in the profile). The mean plasmaaData represent mean ± SEM.glucose concentration was defined as AUC/duration. The bInsulin included regular/soluble, NPH, Ultralente, Lente, Humulin

incremental plasma glucose Cmax was defined as the R, Humulin N, and Humulin 70/30.cOral hypoglycemic agents included Glucotrol, Micronase, Glyn-maximum glucose concentration corrected for the 1 hase, and Diabeta.time point (i.e. plasma glucose concentration at the time

of ingestion of the Sustacal meal subtracted from allsubsequent glucose concentrations in the profile).

Cross-over analysis methods were performed to testfor treatment effects, period effects, and sequence effectsfor glucose, lactate, insulin and C-peptide.17 All statisticaltests were conducted as two-tailed tests with a criticalprobability level of 0.05 for declaring statistical signifi-cance. A Pearson correlation analysis between HbA1c

and the difference in the incremental plasma glucoseAUC (placebo incremental AUC minus pramlintideincremental AUC) was also performed.

All statistical analyses were performed on the entireevaluable patient population as well as for each subgroupof evaluable patients (based on current insulin use)individually. All statistical analyses were performed usingthe Statistical Analysis System (SAS),18 version 6.07 ona UNIX platform. Pharmacokinetic parameters (e.g. zero-hour AUC, and zero-hour Cmax) were calculated usingBIOPAK, version 2.0.

Results Figure 1. Plasma glucose concentrations (mean ± SEM) afterSustacal test meal during placebo and pramlintide infusionsin patients with Type 2 diabetes. Results are shown for allPatientscombined patients

The demographics of the 24 patients enrolled in thestudy are summarized in Table 1. The INS and D/OHA compared to 12.0 ± 0.8 mmol l−1 during placebo infusion

(p = 0.0012). The incremental plasma glucose AUC ingroups were similar except for type of treatment.all patients for the 4 h after Sustacal ingestion wassignificantly reduced to 2.9 ± 1.6 mmol l−1.h during pram-Glucoselintide infusion compared to 11.3 ± 1.9 mmol l−1.h duringplacebo infusion (p = 0.0004). The reduction of incremen-The plasma glucose profiles following the Sustacal

meal in all patients are illustrated in Figure 1. The mean tal plasma glucose Cmax, 5.1 ± 0.6 mmol l−1.h duringplacebo infusion compared to 2.3 ± 0.5 mmol l−1.h duringplasma glucose concentration in all patients for the 4 h

after Sustacal ingestion was significantly reduced to pramlintide infusion, also was statistically significant inall patients (p 0.0001).9.8 ± 5.7 (SEM) mmol l−1 during pramlintide infusion

549PRAMLINTIDE AND GLUCOSE IN NIDDM

ORIGINAL ARTICLESThe plasma glucose profiles following the Sustacal glucose AUC is shown for each patient with patients

rank-ordered by entry HbA1c. This results in a clusteringmeal in the INS group are illustrated in Figure 2. Themean plasma glucose concentration in the INS group of pramlintide responders (reduction in plasma glucose

concentrations from placebo to pramlintide) within thefor the 4 h after Sustacal ingestion was significantlyreduced to 10.2 ± 0.8 mmol l−1 during pramlintide HbA1c values higher than 8.0 %. This subgroup of 5

patients exhibited a statistically significant reduction ininfusion compared to 13.7 ± 0.9 mmol l−1 during placeboinfusion (p = 0.0031). The incremental plasma glucose the incremental glucose AUC for the 4 h after the

Sustacal meal (p = 0.032) and the incremental plasmaAUC in the INS group for the 4 h after Sustacal ingestionwas significantly reduced to 2.8 ± 2.9 mmol l−1.h during glucose Cmax (p = 0.024) during pramlintide infusion

compared to placebo infusion, in a manner that paralleledpramlintide infusion compared to 15.8 ± 2.9 mmol l−1.hduring placebo infusion (p = 0.0012). The reduction of the INS group.

Plasma glucose data were further analysed by calculat-incremental plasma glucose Cmax, 6.6 ± 0.9 mmol l−1.hduring placebo infusion compared to 2.3 ± 0.8 mmol l−1.h ing incremental AUC values from glucose concentrations

over the first 2 h following the Sustacal meal (hoursduring pramlintide infusion, also was statistically signifi-cant in the INS group (p = 0.0007). 1–3). The incremental plasma glucose AUC in all patients

for the first 2 h after Sustacal ingestion was significantlyThe plasma glucose profiles following the Sustacalmeal in the D/OHA group are illustrated in Figure 2. reduced to 1.6 ± 0.6 mmol l−1.h during pramlintide

infusion compared to 5.9 ± 0.7 mmol l−1.h during placeboThe mean plasma glucose concentration in the D/OHAgroup for the 4 h after Sustacal ingestion decreased infusion (p = 0.0001). The incremental plasma glucose

AUC in the INS group for the first 2 h after Sustacalfrom 10.2 ± 1.1 mmol l−1 during placebo infusion to9.5 ± 0.9 mmol l−1 during pramlintide infusion, showing ingestion also was significantly reduced to

1.4 ± 0.9 mmol l−1.h during pramlintide infusion com-a trend toward statistical significance (p = 0.15). Similarly,the incremental plasma glucose AUC in the D/OHA pared to 7.7 ± 1.1 mmol l−1.h during placebo infusion

(p = 0.0003). In addition, the incremental plasma glu-group for the 4 h after Sustacal ingestion decreasedfrom 6.9 ± 1.6 mmol l−1.h during placebo infusion to cose AUC in the D/OHA group for the first 2 h

after Sustacal ingestion was significantly reduced to3.0 ± 1.5 mmol l−1.h during pramlintide infusion, showinga trend toward statistical significance (p =0.12). However, 1.9 ± 0.7 mmol l−1.h during pramlintide infusion com-

pared to 4.0 ± 0.6 mmol l−1.h during placebo infusionthe reduction of incremental plasma glucose Cmax,3.7 ± 0.4 mmol l−1.h during placebo infusion compared (p = 0.0322).to 2.2 ± 0.6 mmol l−1.h during pramlintide infusion, wasstatistically significant in the D/OHA group (p = 0.038). Lactate

The change in incremental glucose AUC withinthe D/OHA group following Sustacal ingestion was In all patients, the mean plasma lactate concentration

for the 4 h after Sustacal ingestion was significantlypositively correlated with glucose control at entry to thestudy as measured by HbA1c (R2 = 0.56, p = 0.005). This reduced to 1.13 ± 0.08 mg dl−1 during pramlintide

infusion compared to 1.53 ± 0.09 mg dl−1 during placeborelationship is demonstrated in Figure 3, where thechange between treatment periods in the incremental infusion (p = 0.0001). The mean plasma lactate concen-

Figure 2. Plasma glucose concentrations (mean ± SEM) after Sustacal test meal during placebo and pramlintide infusions inpatients with Type 2 diabetes. Results are shown for the two patient groups divided according to treatment

550 R.G. THOMPSON ET AL.

ORIGINAL ARTICLES

Figure 3. Change (placebo–pramlintide) in zero-hour plasma glucose AUC after Sustacal test meal in patients with Type 2diabetes who were treated with diet and/or oral hypoglycaemic agents but not insulin. Patients have been rank-ordered based onHbA1c results from the screening visit

trations after the Sustacal meal in the INS and D/OHA insulin concentration after the Sustacal meal in theINS group during placebo infusion wasgroups are shown in Figure 4. The mean plasma lactate

concentration in the INS group over this 4-h time period 531.3 ± 111.1 pmol l−1 compared to343.1 ± 82.6 pmol l−1 during pramlintide infusionwas significantly reduced to 1.18 ± 0.11 mg dl−1 during

pramlintide infusion compared to 1.59 ± 1.14 mg dl−1 (p = 0.022). In the INS group these serum insulin concen-trations reflect a combination of the identical doses ofduring placebo infusion (p = 0.0035). The mean plasma

lactate concentration in the D/OHA group over this 4-h exogenous insulin administered on both treatment daysplus residual endogenous insulin secretion. In the D/OHAtime period also was significantly reduced, to

1.08 ± 0.12 mg dl−1 during pramlintide infusion com- group, the mean serum insulin concentrations after theSustacal meal, 375.0 ± 63.2 pmol l−1 during placebopared to 1.49 ± 0.34 mg dl−1 placebo infusion

(p = 0.0025). infusion and 189.6 ± 58.3 pmol l−1 during pramlintideinfusion (p = 0.0001), represent only endogenous insulin.

InsulinC-peptide

In all patients, the mean serum insulin concentrationafter Sustacal ingestion was significantly reduced to The reduction of serum C-peptide concentrations during

pramlintide infusion compared to placebo infusion paral-266.7 ± 52.1 pmol l−1 during pramlintide infusion com-pared to 467.4 ± 63.9 pmol l−1 during placebo infusion leled the reduction of serum insulin concentrations. In

all patients, the mean C-peptide concentration after the(p = 0.0001). Pramlintide infusion also resulted in asignificant reduction of serum insulin concentrations after Sustacal meal was significantly reduced to

0.37 ± 0.06 nmol l−1 during pramlintide infusion com-Sustacal ingestion compared to placebo in the INS andD/OHA groups, as shown in Figure 5. The mean serum pared to 0.74 ± 0.08 nmol l−1 during placebo infusion

551PRAMLINTIDE AND GLUCOSE IN NIDDM

ORIGINAL ARTICLES

Figure 4. Plasma lactate concentrations (mean ± SEM) after Sustacal test meal during placebo and pramlintide infusions in patientswith Type 2 diabetes. Results are shown for the two patient groups divided according to treatment

Figure 5. Serum insulin concentrations (mean ± SEM) after Sustacal test meal during placebo and pramlintide infusions in patientswith Type 2 diabetes. Results are shown for the two patient groups divided according to treatment

(p = 0.0001). The mean C-peptide concentrations after 9.0 ± 1.9 pmol l−1 in the INS group and14.1 ± 2.1 pmol l−1 in the D/OHA group. The plasmathe Sustacal meal in the INS and D/OHA groups are

shown in Figure 6. The mean C-peptide concentration amylin concentrations after Sustacal ingestion weresimilar for all patients, the INS group and the D/OHAafter Sustacal ingestion in the INS group was signifi-

cantly reduced during pramlintide infusion, group during the 4 h time period.0.20 ± 0.05 nmol l−1 compared to 0.50 ± 0.10 nmol l−1

during placebo infusion (p = 0.005). The mean C-peptide Pramlintideconcentration after Sustacal ingestion in the D/OHAgroup was also significantly reduced from The mean maximal concentration of pramlintide observed

during pramlintide infusion was 591.7 ± 25.2 pmol l−1.0.98 ± 0.08 nmol l−1 during placebo infusion to0.54 ± 0.08 nmol l−1 during the infusion of pramlintide(p = 0.0001). Gastrointestinal Side-effects

The most frequently reported side-effects were gastrointes-Human Amylintinal, with nausea being reported by 14 patients (58 %)during pramlintide infusion and 6 patients (25 %) duringThe mean fasting amylin concentration determined from

the F002-2 assay was 11.5 ± 1.5 pmol l−1 in all patients, placebo infusion. Five patients (21 %) vomited during or

552 R.G. THOMPSON ET AL.

ORIGINAL ARTICLES

Figure 6. Serum C-peptide concentrations (mean ± SEM) after Sustacal test meal during placebo and pramlintide infusions inpatients with Type 2 diabetes. Results are shown for the two patient groups divided according to treatment

following the pramlintide infusion (4 in the INS group this exploratory study in patients with Type 2 diabeteswas chosen because these patients were generally heavierand 1 in the D/OHA group) compared to 1 patient (4 %)

during or following the placebo infusion (in the D/OHA than patients with Type 1 diabetes and because it wasnot known whether patients with Type 2 diabetes wouldGroup). The vomiting occurred a minimum of 2 h after

ingestion of Sustacal. be more resistant to the effects of pramlintide thanpatients with Type 1 diabetes.

The 12 insulin-treated patients, who presumably hadCONCLUSIONSmore advanced b-cell failure, had a reduction inpostprandial glucose concentrations similar to that seenIn previous studies, both intravenous infusion and self-

administered subcutaneous injections of pramlintide have in patients with Type 1 disease.5 The response in thegroup treated with oral agents exhibited much greaterbeen shown to reduce the rise in plasma glucose

concentrations after either mixed meals or standardized inter-subject variability, with 5 of 12 showing a substantialreduction in plasma glucose concentrations during pram-Sustacal meals in amylin-deficient patients with Type

1 diabetes. The initial studies used doses of pramlintide lintide infusion while 7 had minimal or no response. Sixof these latter 7 patients had HbA1c values within thewhich resulted in pharmacological plasma pramlintide

concentrations,5 while more recent studies have normal range.Vomiting was more frequently reported during pramlin-employed subcutaneous administration of doses that

resulted in peak plasma pramlintide concentrations tide infusion than during placebo infusion, whichoccurred a minimum of 2 h after ingestion of Sustacal.similar to postprandial amylin concentrations in healthy

subjects.6,7 Pramlintide is at least equivalent in potency Therefore, to account for a possible influence of vomitingon the glucose lowering effect or pramlintide, zero-hourto human amylin when administered to rats. These

observations are consistent with a physiological role for glucose AUC values were recalculated from the plasmaglucose profiles over only the first 2 h following Sustacalamylin in modulating the rate of gastric emptying, the

rate limiting step for carbohydrate absorption.19 This ingestion. Zero-hour glucose AUC values remainedstatistically significantly reduced during pramlintideconcept is further strengthened by recent demonstrations

of accelerated gastric emptying in insulin-treated BB infusion compared to placebo in all patients and in theINS group. In addition, the zero-hour glucose AUCWistar diabetic rats, in which amylin administration

restored gastric emptying to normal.20 became statistically significantly reduced during pramlin-tide infusion compared to placebo in the D/OHA group.The present study was undertaken to evaluate whether

pramlintide administration would yield similar results in Thus, the vomiting does not explain the lower glucoseconcentrations during pramlintide infusion. Previouspatients with Type 2 diabetes. An exploratory study was

undertaken in 12 insulin-treated and 12 diet/sulfonylurea- studies in patients with Type 1 diabetes have shown thatthe side-effects of pramlintide including nausea are dosetreated patients employing a study design similar to that

employed in prior studies in patients with Type 1 related.6 It is probable that a lower dose of pramlintidewould reduce postprandial plasma glucose concen-diabetes.5 Pramlintide was infused at 100 mg h−1 over

5 h and the Sustacal meal adjusted to contain trations with reduced gastrointestinal side-effects.The effect of an agent that delays gastric emptying on7 kcal kg−1. The relatively high pramlintide dose used in

553PRAMLINTIDE AND GLUCOSE IN NIDDM

ORIGINAL ARTICLESpostprandial glucose levels was first demonstrated in seen in patients with Type 2 diabetes treated with diet

and/or sulphonylurea agents when HbA1c concentrationsnormal subjects by Liddle et al.21 and in subjects withType 2 diabetes by Phillips et al.22 Amylin replacement were greater than approximately 8 %. Additional studies

are required to assess the feasibility of achieving similartherapy employing pramlintide reduces postprandialglucose concentrations at least in part by modulating effects upon plasma glucose concentrations with subcut-

aneous administration.gastric emptying.20 Although serum insulin values aredifficult to interpret in insulin-treated patients, the resultssuggest that pramlintide reduced insulin secretion during

Acknowledgementsthe postprandial period. In addition to reducing postpran-dial insulin needs, pramlintide’s ability to modulate

This study was sponsored by Amylin Pharmaceuticalsgastric emptying leads to a more normal gastric emptyingInc. and Harris Laboratories Inc.rate, which was reported in a 1992 article to be

accelerated in patients with a recent diagnosis of Type2 diabetes.23 As stated by Horowitz and Fraser:24 ‘The

Referencescontribution of disordered gastric emptying to poorglycaemic control is unclear, but the demonstration that 1. Koda JE, Fineman M, Rink TJ, Daily GE, Muchmore DB,the rate of gastric emptying is a major factor in normal Linarelli LG. Amylin concentrations and glucose control.

Lancet 1992; 339 (8802): 1179–1180.blood glucose homeostasis suggests that this is likely to2. Hartter E, Svoboda T, Ludvik B, Schuller M, Lell B,be significant.’

Kuenburg E, et al. Basal and stimulated plasma levels ofHyperglycaemia has been shown to impede gastricpancreatic amylin indicate its co-secretion with insulin

emptying and is a confounding variable that has not in humans. Diabetologia 1991; 34: 52–54.been considered in many studies assessing gastric empty- 3. Brown K, Menius A, Smadefer E, Edwards J, James M.

The effects of amylin on change in plasma glucose anding rates. The contribution of accelerated gastric emptyinggastric emptying following an oral glucose load into the postprandial glycaemic surge observed in theseconscious dogs. Diabetes 1994; 43 (suppl 1): 172A.patients may be currently underestimated. The modu-

4. MacDonald I, King P, Kong MF, Stubbs T, Perkins A,lation of gastric emptying by an analogue of a human Moyses C, et al. Infusion of the human amylin analog,hormone and the subsequent achievement of improved AC137, delays gastric emptying in men with IDDM.

Diabetologia 1995; 38 (suppl 1): A32.glycaemic control represents an approach that differs5. Kolterman OG, Gottlieb A, Moyses C, Colburn W.from most therapies for Type 2 diabetes. The majority

Reduction of post-prandial hyperglycemia in subjects withof agents either under development or currently in useType I diabetes mellitus by intravenous infusion of AC137,

for treatment of Type 2 diabetes have focused on insulin a human amylin analogue. Diabetes Care 1995; 18:activity. This includes the development of rapid-acting 1179–1182.

6. Kolterman OG, Schwartz S, Corder C, Levy B, Klaffinsulin analogues and agents designed to increase insulinL, Peterson J, et al. Effect of 14 days’ subcutaneoussecretion and/or improve insulin activity. In contrast,administration of the human amylin analogue, pramlintidein this study, pramlintide produced a reduction of(AC137), on an intravenous insulin challenge and response

postprandial glucose concentrations while actually lower- to a standard liquid meal in patients with IDDM.ing plasma insulin concentrations. Diabetologia 1996; 39: 492–499.

7. Thompson RG, Peterson J, Gottlieb A, Mullane J. EffectsIn this study, when compared to placebo, pramlintideof pramlintide, an analog of human amylin, on plasmainfusion was associated with statistically significantlyglucose profiles in patients with IDDM: Results of areduced plasma lactate concentrations following themulticenter trial. Diabetes 1997; 46: 632–636.

Sustacal test meal in both insulin-treated and non- 8. Sanke T, Hanabusa T, Nakano Y, Oki C, Okai K,insulin-treated patients, in parallel with the reduced Nishimura S, et al. Plasma islet amyloid polypeptide

(amylin) levels and their responses to oral glucose in Typeplasma glucose concentrations. In studies with fasted2 (non-insulin-dependent) diabetic patients. Diabetologiarats,25 a dose-dependent increase in plasma lactate was1991; 34: 129–132.observed with pramlintide. Postprandial plasma lactate

9. Ludvik B, Lill B, Hartter E, Schnack C, Prager R. Decreaseconcentrations can be influenced by many factors, of stimulated amylin release precedes impairment ofincluding release from peripheral tissues, hepatic uptake, insulin secretion in Type II diabetes. Diabetes 1991; 40:

1615–1619.catecholamine concentrations, and exercise. The specific10. Slein MW, Cori GT, Cori CF. A comparative study ofinfluence of each factor on postprandial plasma lactate

hexokinase from yeast and animal tissues. J Biol Chemconcentrations, and any role for reduced lactate concen-1950; 186: 763–780.

trations in the reduction in plasma glucose concentrations 11. Percy AJ, Trainor D, Rittenhouse J, Janes S, Koda J.by pramlintide, remains to be established. Sensitive two-site enzyme immunoassays for amylin and

amylin-like peptides in human plasma. Clinical ChemistryIn summary, intravenous administration of the human1996; 42: 576–585.amylin analogue pramlintide to patients with Type 2

12. Fineman MS, Koda JE, Percy A, Wareham NJ, Hales CN.diabetes treated with insulin produced a significantElevation of high molecular weight amylin-like peptides

reduction in postprandial plasma glucose concentrations along with proinsulin and 32, 33 split proinsulin in newlyassociated with a concurrent reduction in plasma lactate diagnosed diabetes (Abstract) Diabetologia, 1994; 37

(suppl 1): A52.and serum insulin concentrations. Similar responses were

554 R.G. THOMPSON ET AL.

ORIGINAL ARTICLES13. Pittner RA, Albrandt K, Beaumont K, Gaeta L, Koda J, by subcutaneous injections of amylin. Diabetologia 1995;

38: 642–648.Moore C, et al. J Cellular Biochem 1994; 55 (suppl): 19–28.14. Percy AJ, Trainor D, Blase E, Redalieu E, Koda JE. 21. Liddle RA, Rushakoff RJ, Horita ET, Baccaria L, Carter

JD, Gordfine ID. Physiological role for cholecystokininDevelopment of a sensitive immunoassay for pramlintide,and its use in pharmacokinetic studies. J Clin Lig and in reducing postprandial hyperglycemia in humans. J Clin

Invest 1988; 81: 1675–1681.Assay 1997; 20: 32–39.15. Yalow RS, Berson S. Immunoassay of endogenous plasma 22. Phillips WT, Schwartz JG, McMahan CA. Reduced post-

prandial blood glucose levels in recently diagnosed non-insulin in man. J Clin Inv 1960; 39: 1157.16. Marbach E, Weil MH. Rapid enzymatic measurement of insulin-dependent diabetics secondary to pharmacologi-

cally induced delayed gastric emptying. Digestive Diseasesblood lactate and pyruvate. Clin Chem 1967; 13: 314.17. Fleiss JL. The Design and Analysis of Clinical Experiments. and Sciences 1993; 38: 51–58.

23. Phillips WT, Schwarz JG, McMahan CA. Rapid gastricNew York: Wiley, 1986: 263–290.18. SAS/STAT User’s Guide, 6th edn. Cary NC: SAS Institute emptying of an oral glucose solution in type 2 diabetic

patients. J Nucl Med 1992; 33: 1496–1500.Inc., 1990.19. Horowitz M, Edelbroek MAL, Wishart J, Straathof J. 24. Horowitz M, Fraser R. Disordered gastric motor function

in diabetes mellitus. Diabetologia 1994; 37: 543–551.Relationship between oral glucose tolerance and gastricemptying in normal healthy subjects. Diabetologia 1993; 25. Young AA, Vine W, Gedulin BR, Pittner R, Janes S, Gaeta

LSL, et al. Preclinical pharmacology of pramlintide in the36: 857–862.20. Young AA, Gedulin B, Vine W, Percy A, Rink, TJ. Gastric rat: comparisons with human and rat amylin. Drug

Development Research 1996; 37: 231–248.emptying is accelerated in diabetic BB rats and is slowed

555PRAMLINTIDE AND GLUCOSE IN NIDDM