differences in effect of insulin detemir and nph insulin ...epubs.surrey.ac.uk › 855527 › 1 ›...
TRANSCRIPT
Differences in Effect of Insulin Detemir and NPH Insulin on Glucose and Lipid Metabolism
by
Sara Victoria Margaret Hordern MBBS BSc MRCP
Submitted for the degree of Doctor of Medicine
Postgraduate Medical School University of Surrey
August 2004
Sara Victoria Margaret Hordern 2004
ProQuest N um ber: 27605345
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a com p le te manuscript and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
uestProQuest 27605345
Published by ProQuest LLO (2019). Copyright of the Dissertation is held by the Author.
All rights reserved.This work is protected against unauthorized copying under Title 17, United States C ode
Microform Edition © ProQuest LLO.
ProQuest LLO.789 East Eisenhower Parkway
P.Q. Box 1346 Ann Arbor, Ml 48106- 1346
Summary
Insulin detemir has a 14 C fatty acid chain, which associates with free fatty acid binding
sites on albumin. Capillary endothelial cells in adipose tissue and muscle may limit transfer
of insulin detemir from the circulation into the extravascular extracellular space. In the
liver, the open sinusoids may expose hepatocytes to insulin detemir enabling it to have
greater effect on the liver than on the periphery.
We investigated the effects of two doses of insulin detemir and one dose of NPH insulin on
hepatic glucose output (Ra), peripheral glucose uptake (Rd), glycerol Ra using stable isotope
techniques and on NEFA, IGF-1, IGFBP-1, C-peptide, and glucagon concentrations in 7
healthy volunteers, 6 volunteers with type 1 diabetes and 6 volunteers with type 2 diabetes
during a 16h euglycaemic clamp.
In healthy volunteers we demonstrated that 9nmol/kg insulin detemir had greater effect on
suppression of glucose Ra than an equipotent dose of NPH (0.3iu/kg or 1.8nmol/kg). In
subjects with type 1 and type 2 diabetes 18nmol/kg insulin detemir was equipotent to 0.6
iu/kg (3.6 nmol/kg) NPH, with a similar time action profile. In subjects with type 1 diabetes
we were unable to demonstrate differences in the effect of insulin detemir on glucose Ra but
demonstrated small but significant differences in effect on glucose Rd compared to the
equipotent dose of NPH. The suppression of NEFA concentrations by insulin detemir was
less than with an equipotent dose of NPH, both in healthy subjects and to a lesser extent in
those with type 1 diabetes. We were unable to demonstrate a statistical difference in effect
on glucose or lipid metabolism in subjects with type 2 diabetes. There were no differences
in the effect of these two insulins on glycerol Ra, IGF-1, IGFBP-1, glucagon or c-peptide
concentrations in any of our three subject groups. We demonstrated a dose response effect
for insulin detemir.
This study suggests that when equipotent doses of insulin detemir are compared to NPH
insulin, detemir has greater effect on the liver than peripheral tissues and thus the potential
to restore the normal insulin gradient.
Contents
Summary 1
Contents 2
Acknowledgements 6
List of Figures 7
List of Tables 9
Abbreviations 10
Chapter 1. Introduction
Diabetes 13
Definition of diabetes 13
Epidemiology 14
History of diabetes 15
The insulin molecule 20
Structure of the insulin molecule 20
Biosynthesis of insulin 21
The insulin receptor 22
Glucose metabolism 24
Endogenous glucose production 25
Glucose utilization 25
Glucose homeostasis 28
Fat and Linid Metabolism 32
The Endothelial Barrier 33
Henatic Insulinisation 36
Lone Actine Insulin 40
Background 40
Insulin detemir 43
Physical and Chemical Properties 43
Pharmacology 43
Formulation 45
Efficacy pharmacology 46
Pharmacokinetics and protein binding 46
Mitogenicity 48
Aims of this study 48
Chapter 2. Clinical Methodology
Introduction to the glucose damn technique 49
Isotope dilution 49
Calculations 51
Clamp Protocol 52
Preparation of Tracer solutions 56
Laboratory assessments 56
Glucose concentration and enrichment 56
Glycerol concentration and enrichment 57
Non-Esterified Fatty Acids (NEFA) 58
Insulin and insulin detemir, C-peptide and Glucagon59
IGF-1 59
IGFBP-1 60
Trial Design 61
Type of trial 61
Rationale for trial design 62
Rationale for dosage regime 62
Trial population 63
Number of subjects in each group 63
Inclusion Criteria 63
Exclusion criteria 64
Safety visit 65
Statistical analyses 66
Chapters. Results
Healthy Subiects 67
Subject demographics 67
Plasma glucose 67
Glucose infusion rate 68
Glucose metabolism 70
Glycerol metabolism 75
NEFA 77
IGF-1 78
IGFBP-1 79
Insulin and insulin detemir 80
C-peptide 81
Glucagon 82
Tvne 1 Subiects 83
Subject demographics 83
Plasma glucose 83
Glucose infusion rate 84
Glucose metabolism 86
Glycerol metabolism 90
NEFA 91
IGF-1 93
IGFBP-1 94
Insulin and insulin detemir 96
C-peptide 96
Glucagon 96
Type 2 subiects 97
Subject demographics 97
Plasma glucose 97
Glucose infusion rate 98
Glucose metabolism 100
Glycerol metabolism 103
NEFA 104
IGF-1 106
IGFBP-1 107
Insulin and insulin detemir 108
C-peptide 108
Glucagon 109
Combined Results 110
Plasma Glucose 110
Glucose Infusion Rate 111
Glucose Metabolism 112
Chapter 4. Discussion 116
Chapter 5. References and Bibliography 133
Acknowledgments
I would like to thank Sarah Hills for her assistance performing the clamps, but more
importantly Juliet Wright for clamping in extreme circumstances, for all her support while I
was entering data and doing the calculations, without her help, support and fantastic sense of
humour at the lowest times, this study would never have been completed.
I would like to thank Jo Bullen for helping me order the equipment I needed for the studies.
I would like to thank Margot Umpleby for all her help with the stable isotope data, for
teaching me how to do the calculations, for setting up the spreadsheets for the calculations
and for her support in the papers that we have written.
I would like to thank Fariba Shojaee-Moradie for teaching me how to clamp, for helping us
to do the first clamp and for her help along with Nicola Jackson, Premilla Croos and
William Jefferson in measuring glucose, glycerol and NEFA.
I would like to thank Gwen Wark for measuring IGF-land IGFBP-1 samples. I would also
like to thank Steven Halloran and Derek Teale in the biochemistry department at The Royal
Surrey County Hospital for their help with obtaining our YSI instrument and providing
samples for its calibration.
I would like to thank the late Ken McRae for all his help with the statistical analysis for this
study, and latterly Julie Amiss and David Lovell for their statistical help.
I would like to thank Guildford Pharmacology Unit for allowing us to use their centrifuge
during the clamps.
I would like to thank Lisbeth Jacobsen, Lars Endahl and Eberhard Draeger from Novo
Nordisk for their help in the study design.
I would like to thank Novo Nordisk and the Dunhill Medical Trust for sponsorship and
funding of the study.
I would like to thank Professor Gordon Ferns for being one of my educational supervisors.
Finally I would like to thank David Russell-Jones for all his invaluable support throughout.
For the concept and design of the study, for helping with arranging the funding, for keeping
our morale high during clamping, and for his persistence in encouraging me to finish writing
up.
List of figures
1. Schematic representation of the three dimensional structure of insulin 20
2. The glycolysis pathway 27
3. Schematic representation of albumin binding of insulin detemir, and potential
mechanism for differences in effect on the liver and periphery. 36
4. Schematic representation of the Physiological Insulin Production. 37
5. Schematic Representaion of Pharmacological Replacement of Insulin. 37
6. Schematic structure of insulin detemir. 42
7. Structure of insulin detemir dimer. 44
8. Di-hexamerisation of insulin detemir. 45
9. Distinct binding sites on albumin. 47
10. Graph of healthy subjects mean plasma glucose. 68
11. Graph of healthy subjects mean glucose infusion rates 69
12. Graph of healthy subjects mean glucose rate of appearance. 70
13. Graph of healthy subjects mean glucose rate of disappearance. 71
14. Graph of healthy subjects mean glucose Ra and Rd. 71
15. Bar charts showing the relative proportion of glucose lowering effect from effect on
glucose Ra and glucose Rd of each insulin. 72
16. Pie chart showing the overall mean proportion of effect of each insulin on glucose
Ra and Rd. 73
17. Graph of Healthy Subjects Glucose TraceriTracee ratio during the clamp studies 75
18. Graph of healthy subjects mean rate of appearance of glycerol. 76
19. Graph of healthy subjects mean plasma NEFA concentrations. 78
20. Graph of healthy subjects mean plasma IGF-1 concentrations. 79
21. Graph of healthy subjects mean plasma IGFBP-1 concentrations. 80
22. Graph of healthy subjects mean plasma insulin and insulin detemir concentrations.
81
23. Graph of healthy subjects mean plasma c-peptide concentrations. 82
24. Graph of Type 1 Subjects Mean Plasma Glucose. 84
25. Graph of Type 1 Subjects Mean Glucose Infusion Rates. 85
26. Graph of Type 1 Subjects Mean Glucose Rate of appearance. 86
27. Graph of Type 1 Subjects Mean Glucose Rate of Disappearance. 87
28. Graph of Type 1 Subjects Mean Glucose Ra and Rd 87
29. Graph of Type 1 Subjects Mean Glucose Ra and Rd Shown as change from 60
minutes. 89
30. Graph of Type 1 Subjects Mean Glucose Tracer:Tracee Ratio. 89
31. Graph of Type 1 Subjects Mean Gycerol Rate of Appearance. 91
32. Graph of Type 1 Subjects Mean Plasma Non Esterifred Fatty Acid Concentrations.92
33. Graph of Type 1 Subjects Mean Plasma Insulin Like Growth Factor-1
Concentrations. 94
34. Graph of Type 1 Subjects Mean Plasma Insulin Like Growth Factor Binding Protein-
1 Concentrations. 95
35. Graph of Type 2 Subjects Mean Plasma Glucose Concentrations. 98
36. Graph of Type 2 Subjects Mean Glucose Infusion Rates. 99
37. Graph of Type 2 Subjects Mean Glucose Rate of Appearance. 100
38. Graph of Type 2 Subjects Mean Glucose Rate of Disappearance. 101
39. Graph of Type 2 Subjects Mean Glucose Ra and Rd. 101
40. Graph of Type 2 Subjects Mean Glucose Ra and Rd Expressed as a change from 60
minutes. 102
41. Graph of Type 2 Subjects Mean Plasma Glucose Tracer:Tracee Ratio. 102
42. Graph of Type 2 Subjects Mean Glycerol Rate of Appearance. 104
43. Graph of Type 2 Subjects Mean Plasma Non Esterifred Fatty Acid Concentrations.
105
44. Graph of Type 2 Subjects Mean Plasma Insulin Like Growth Factor-1
Concentrations. 106
45. Graph of Type 2 Subjects Mean Plasma Insulin Like Growth Factor Binding Protein-
1 Concentrations. 108
46. Graph of type 2 Subjects Mean Plasma C-Peptide Concentrations. 109
47. Graph of Mean Glucose Infusion Rates for Each Subject Group. 110
48. Graph of Mean Glucose Infusion Rates for Each Insulin from the Combined Data
from each Subject Group. 112
49. Graph of Mean Glucose Ra and Rd for each insulin from the combined data from
each subject group. 113
50. Bar charts showing the relative proportion of glucose lowering effect from effect on
glucose Ra and glucose Rd of each insulin using data combined from the three
subject groups. 114
51. Pie chart showing the overall mean proportion of effect of each insulin on glucose
Ra and Rd using data combined from the three subject groups. 115
52. Diagram of the Physiological Plasma Insulin Secretion Profile. 117
List of tables
1. Blood Sampling Times 55
2. Healthy Subjects Plasma Glucose and Glucose Infusion Rate. 69
3. Healthy Subjects Glucose Metabolism 74
4. Healthy Subjects Glycerol Metabolism. 76
5. Healthy Subjects NEFA results. 77
6. Healthy Subjects IGF-1 results. 78
7. Healthy Subjects IGFBP-1 results. 80
8. Type 1 Subjects Plasma Glucose and Glucose Infusion Rate. 85
9. Type 1 Subjects Glucose Metabolism. 90
10. Type 1 Subjects Glycerol Metabolism. 91
11. Type 1 Subjects NEFA results. 92
12. Type 1 Subjects IGF-1 Results. 93
13. Type 1 Subjects IGFBP-1 Results. 95
14. Type 2 Subjects Plasma Glucose and Glucose Infusion Rate. 99
15. Type 2 Subjects Glucose Metabolism. 103
16. Type 2 Subjects Glycerol Metabolism. 104
17. Type 2 Subj ects NEFA Results. 105
18. Type 2 Subjects IGF-1 Results. 106
19. Type 2 Subjects IGFBP-1 Results. 107
20. Combined Glucose and Glucose Infusion Rate Data. I l l
21. Combined Glucose Metabolism Data. 113
Abbreviations used in thesis
Degrees Celsius
ACTH Adrenocortrophic Hormone
AD Ano Domini
ADA American Diabetes Association
ANOVA Analysis of variance
AOC Area over curve
APE Atoms percent excess
ATP Adenosine tri-phosphate
AUC Area under curve
a-v arterio-venous
BC Before Christ
BMI Body Mass Index
cAMP cyclic adenosinemono-phosphate
cDNA complimentary deoxyribonucleic acid
Cl confidence interval
Cmax Maximum concentration
C-Peptide Connecting peptide
CV Coefficent of variation
Da Daltons
DCCT Diabetes control and complications trial
DNA deoxyribonucleic acid
g grams
GH Growth Hormone
GIR Glucose Infiision Rate
GLUT Glucose transporter
GTP Guanine Tri-Phosphate
H^ Hydrogen ion
HbAic Glycosylated Haemoglobin
10
HD High dose detemir (18nmol/kg)
HDL High density lipoprotein
IDL Intermediate density lipoprotein
lEMA Immunoenzymometric assay
IGF-1 Insulin like growth factor-1
IGFBP-1 Insulin like growth factor binding protein-1
IGF-II Insulin like growth factor-II
IRS-1 Insulin receptor substrate
IRTK Insulin receptor tyrosine kinase
lU International Units
kDa Kilo Daltons
Kg Kilogram
1 Litre
lb pounds
LD Low dose detemir (9nmol/kg)
LDL Low density lipoprotein
m metre
mg milligram
min minutes
ml millilitres
mmol millimol
mRNAmessenger ribonucleic acid
MTBSTFA N-Methyl-N-(t-butyldimethylsilyl)trifluoroacetamide
NEFA Non esterified fatty acid
nm nanometre
NMR Nuclear magnetic resonance
NPH Neutral Protamine Hagedom
OGTT Oral glucose tolerance test
pH pondus hydrogenii (potential hydrogen)
PEPCK Phosphoenolpyruvate Carboxykinase
Ra Rate of appearance
11
Rd Rate of Disappearance
RER Rough endoplasmic reticulum
rpm revolutions per minute
SD Standard deviation
SEM Standard Error of the Mean
SH Src Homology
SIM Selected ion monitoring
Src Sarcoma
SSA Sulphosalicylic acid
Tmax Time to reach maximum concentration
UK United Kingdom
UKPDS United Kingdom prospective diabetes study
USA United States of America
VEI Vesicle encapsulated insulin
VLDL Very low density lipoprotein
WHO World Health Organisation
a Alpha
p Beta
p micro
12
Chapter 1
Introduction
Diabetes
Definition
Diabetes mellitus is a state of chronic hyperglycaemia, classically associated with symptoms
of excessive thirst, increased urine volume and weight loss. It results from the insufficiency
of or diminished effectiveness of endogenous insulin.
Classification of diabetes (American Diabetes Association 1999)
I Type 1 diabetes (P-cell destruction, usually leading to absolute insulin deficiency)
A. Immune mediated
B. Idiopathic
II Type 2 diabetes (ranging from predominantly insulin resistance with relative insulin
deficiency to predominantly an insulin secretory defect with insulin resistance).
III Other specific types of diabetes
A. Genetic defects of P-cell function
B. Genetic defects of insulin action
C. Diseases of the exocrine pancreas
D. Endocrinopathies
E. Drug or chemical induced
F. Infections
G. Uncommon forms of immune mediated diabetes
H. Other genetic syndromes sometimes associated with diabetes.
IV Gestational diabetes
Criteria for the diagnosis of diabetes mellitus
I. Symptoms of diabetes plus casual plasma glucose concentration >200mg/dl
(11.1 mmol/1). Casual is defined as any time of day without regard to time since last meal.
The classic symptoms of diabetes include polyuria, polydipsia and unexplained weight loss.
13
Or
2. Fasting plasma glucose > 126mg/dl (7.0mmol/l). Fasting is defined as no calorie intake
for at least 8 hours.
Or
2-hour plasma glucose >200mg/dl (11.1 mmol/1) during an OGTT. The test should be
performed as described by WHO, using a glucose load containing the equivalent of 75mg
anhydrous glucose dissolved in water.
In the absence of unequivocal hyperglycaemia with acute metabolic decompensation, these
criteria should be confirmed by repeat testing on a different day.
Epidemiology
The world-wide incidence of type 1 diabetes varies greatly from as high as 36 per 100 000
per year in some Scandinavian countries while in countries such as Japan it is only 2 per 100
000 per year, Venezuela has the lowest recorded incidence of 0.1 per 100 000 per year
(Wareham, 2002). The disease is predominantly one of white Caucasian populations and is
relatively rare in both oriental and black populations (Bell and Hockaday, 1996). The
incidence of type 1 diabetes appears to be increasing. Increases have been reported in
Scotland of about 2 % per annum during the period 1984-1993 (Rangasami et al, 1997).
The overall pooled incidence in 37 countries was 3% per year, the increase was relatively
greater in populations with the lowest incidence. The incidence of type 1 diabetes also
varies with season, being highest in autumn and winter. Type 1 diabetes incidence varies by
age and sex, with in the UK the peak incidence in boys being at 12-13 years and the peak
incidence in girls being 9-12 years. The incidence is slightly higher in boys than girls (Ratio
1.08: 1). (Wareham, 2002)
Type 2 diabetes is far more common than type 1. In some populations of Polynesian-
Micronesian stock the prevalence may be as high as 40-50%, whether on affluent Pacific
islands (eg Nauru) or in reservations in Arizona (Pima Indians). The prevalence in the
United Kingdom was believed to be 2%. However recent studies have suggested values
14
ranging from 8 to 12% depending on ethnic mix (Bell and Hockaday, 1996). There are
currently believed to be 1 million people in the UK with undiagnosed diabetes (Diabetes
UK). The prevalence of type 2 diabetes is lowest in rural areas of developing countries, is
generally intermediate in developed countries and is highest in some ethnic groups who have
adopted Western lifestyle patterns. The populations with the highest prevalence (Pima
Indians and Nauruans) have a high prevalence of obesity. It is hypothesized that genetic
susceptibility to obesity in these populations would be disadvantageous in times of food
abundance, but would give evolutionary advantages at times when food is scarce, giving
rise to increased prevalence of the gene by natural selection. This “thrifty genotype”
hypothesis is supported by evidence of gene environment interaction - individuals who
migrate from low prevalence areas to the West are at increased risk of type 2 diabetes. The
prevalence of type 2 diabetes is also increasing with figures from the USA showing a 33%
increase in prevalence from 4.9% in 1990 to 6.5% in 1998 (Mokdad et al, 2000). This
increase in prevalence mirrors the increasing prevalence of obesity. The main aetiological
risk factors for type 2 diabetes are age, obesity, family history, physical inactivity and
dietary factors such as a high proportion of energy consumed as saturated fat and low intake
of finit and vegetables. (Wareham, 2002)
History of diabetes.
Early Hindu writings attributed to Susrata in 400BC, describe people with a mysterious and
deadly disease that caused increased thirst, enormous urine output and wasting away of the
body whose urine was attractive to black ants. (MacCracken, 1997, Owens, 1986). Areteus
of Cappadocia a contempory of Nero (30-90 AD) wrote “diabetes is a strange disease,
which fortunately is not very frequent. It consists o f the flesh and bones running together
into the urine.... The illness develops very slowly. Its final outcome is death... The patients
are tortured by an unquenchable thirst; they never cease drinking and urinating.... The
disease was called diabetes, as though it were a siphon because it converts the human body
into a pipe for the transflux o f liquid humors. Now since the patient goes on drinking and
urinating while only the smallest portion o f what he drinks is assimilated in the body, life
naturally cannot be preserved fo r very long, fo r a portion o f the flesh is also excreted
15
through the urine.” (Hoskins, 1941). The term diabetes (from Greek Dia, through and
baino, I flow and mellita, a bee) had actually been coined earlier by Apollonius of Memphis
and Demetrius of Apamea around 250 BC.
It wasn’t until 1674 that Thomas Willis re-described the presence of glycosuria and
concluded that this must be preceeded by presence of sweetness in the blood. In 1774
Matthew Dobson demonstrated that the sweetness of urine and serum was due to sugar. In
1849 Claude Bernard working on glucagon metabolism in the liver linked diabetes with
glycogen metabolism. In 1869 Paul Langerhans, then 22 described small bodies in the
pancreas in his thesis in Berlin. Unfortunately he was unable to continue his work,
developing tuberculosis, he had to leave Germany to live in Madeira, where he died at the
age of 40 (Andreani, 1999).
In the late 19 Century Minkowski and von Mehring were interested in the study of the
pancreas in relation to digestion. However it was a kennel worker who was caring for the
experimental animals of the two doctors who noticed that their urine was attracting great
swarms of flies, he drew the doctor’s attention to this whereupon it was discovered the urine
had large amounts of sugar (Minkowski later denied this version of events). Thus
experimental diabetes had been produced. Minkowski recognised the significance of the
discovery and tried to make an extract of dog’s pancreas that would have a beneficial effect
on the experimental diabetes he had produced but without success. Caparelli however
anticipated him and discovered a simple saline extract of pancreas injected directly into the
abdominal cavity of dogs had a marked lessening or even disappearance of glucose in the
urine.
Clinicians immediately attempted to use the knowledge, some administered fresh pancreas
by mouth, some gave juice expressed from the gland, and some administered extracts
rectally. Bormann administered roast pancreas by mouth, later infusions by rectum and
finally subcutaneous injections of raw pancreas extract. One of his patients who received
the latter treatment gained eight pounds in six weeks. Others were finding similar results
however these pancreatic extract contained digestive enzymes and patients were unable to
16
tolerate the injections for long (Hoskins, 1941). In 1898 Blumenthal squeezed the juice
from raw pancreas and treated it with alcohol to remove the protein. When administered
intravenously it killed his experimental animals, and when injected subcutaneously caused
ulcers however he was able to show that subcutaneous injection of his preparation increased
glucose utilization by 40%. Blumenthal was actually using a crude preparation o f insulin,
and had his side effects been less severe he may have enjoyed some of the glory given to
workers a quarter of a century later.
The next ten years revealed little hope for the treatment of diabetes, indeed “The Household
Physician” “A family guide to the preservation of health and to the domestic treatment of
ailments and disease” (McGregor-Robertson, 1898) wrote that the chief treatment consisted
of regulation of the diet. It gives two tables of permitted and not permitted foods and drinks
and also recommended that the patient “take regular moderate exercise, flannels should be
worn, and warm baths taken frequently”. Of pharmacological approaches to treatment its
author wrote, “many medicines have been tried, but none have been very successful. Opium
and ergot are probably the best, and may be taken in pill containing % grain powdered
opium and 2 grains ergotine, one thrice daily. In many cases the use o f opium ought be
pushed but not without medical supervision.” In 1913 Frederick Allen wrote “all
authorities are agreed upon the failure ofpancreatic opotherapy in diabetes ...injections o f
pancreatic preparations have both proved useless and harmful. The failure began with
Minkowski and has continued to the present without interruption... The negative reports
have been numerous and trustworthy. ” (Allen, 1913).
Frederick Madison Allen (bom in 1876) working at Harvard Medical school, carried out
extensive work on diabetes and carbohydrate metabolism. He concluded that while diabetes
was mostly a problem of carbohydrate metabolism, it was not just the carbohydatres but the
proteins and fats as well. Diets that involved cutting back sharply on carbohydrates and
increasing proteins and fats achieved little and if anything caused a deterioration in the
condition with a higher rate of acidosis and death resulting from the metabolism of fats.
Described as a stem, cold, tireless scientist (Bliss 1982) Allen had a ward of diabetic
patients at the Rockefeller Institute in New York. Here he applied his diets to an assortment
17
of diabetic patients restricting some to as little as 400 calories per day. He published his
findings of 76 of the 100 patients he treated in 1919 (Total dietary regulation in the
treatment of diabetes), some patients survived for a few years others only for a few months.
Allen’s diets were generally thought to be the best therapy at the time and were adopted by
medical schools and up to date practitioners.
During this time many were dying from diabetes, but Eugene Opie (an American at Johns
Hopkins University in Baltimore) in 1901 described that in a high proportion of autopsies
destructive changes could be seen in the pancreatic “islands of Langerhans”.
In 1908 Georg Ludwig Zuelzer a young German physician collected pancreas glands from
animals at the height of digestion. He pressed out the juice and treated it with alcohol to
precipitate the proteins. The material was then freed from the alcohol and injected into
dogs. He managed to reduce the urine sugar output in these dogs with the substance he
called “acomatrol”. He treated 8 patients and successfully reduced the amount of sugar in
their urine. Indeed he injected acomatrol into a fifty year old diabetic in a private clinic in
Berlin. The patient was comatose, but after receiving the extract his condition improved,
only to deteriorate and die when supplies of Zuelzer’s acomatrol dried up (MacCracken,
1997). His extract needed purifyiiig further to decrease the side effects of vomiting, high
fevers and convulsions that his patients were suffering. Zuelzer continued to attempt to
make a pancreas preparation suitable for the treatment o f diabetes. He visited many
pharmaceutical companies to try to gain backing without success. Finally in 1911 he was
given funding by one, however he had little success and in 1914 his hospital was taken over
by the military. The presence of digestive enzymes in the pancreatic products was causing
great difficulties. Several others including Scott, Starling, Knowlten and Macleod continued
research along the lines of extracting and purifying the pancreatic juices, but their work was
also interrupted by the outbreak of World War 1.
In 1921 Banting (an orthopaedic surgeon) returned from the war, and reading an article by
Barron written in 1920 describing the old experiment of ligating the pancreatic ducts which
results in atrophy of the exocrine pancreas leaving the islets intact, it occurred to him that
18
the troublesome digestive enzymes may be circumvented using this technique. Under
Professor Macleod at the University of Toronto he was given the assistance of Best, a
medical student. They started ligating ducts in the pancreas of dogs and after the pancreas
had degenerated injected extracts of it into depancreatised dogs. They recorded a lowering
of the urine sugar output, however they soon ran out of pancreatic tissue. At this point they
too were confronted with the problem of preparation and refinement of the extracts of
pancreatic tissue. Collip was then added to the group and brought to bear the experience
gained in several years of devotion to this type of chemistry. They moved on to using the
foetal calf pancreas to avoid the activity of trypsin, and then onto ox pancreas, with which
they discovered that if the whole gland was immediately placed in acid aqueous alcohol
yielded adequate amounts of insulin. The use of alcohol and acid in the extraction process
circumvented the enzyme action. The resulting extract in concentrated form was first given
to the investigators themselves.
In January 1922 its first use in patients was begun. Leonard Thompson a 14 year old
weighing only 641b was the first patient to be given insulin. He received 7.5 ml in each
buttock of “thick brown muck”. Although initially his blood sugar and glycosuria fell,
unfortunately he developed abscesses at the injection sites and became more unwell. The
injections were stopped. Six weeks later he was given Collip’s more refined extract, his
blood sugar fell and he quickly began to gain weight. Collip’s extract was then given to a
number of patients with diabetes at Toronto General hospital, they responded well. During
the summer of 1922 insulin began being produced commercially by Eli Lilly in Indianapolis,
USA. The first insulin complementary deoxyribonucleic acid (cDNA) was cloned and
sequenced from rat islet messenger ribonucleic acid (mRNA) by Ullrich et al., (1977). The
human insulin gene was later fully characterised (Bell et al., 1980). Owerbach et al., 1981,
located the human insulin gene to the short arm of chromosome 11. Recombinant DNA
techniques have allowed biosynthetic human insulin to be produced by Escherichia coli
(Keefer et al., 1981).
19
The Insulin molecule
Structure of the insulin molecule.
Sanger and co-workers determined the complete amino acid sequence of insulin and its
primary structure in 1955 (Ryle et al 1955). Schlichchtkrull (1956) demonstrated that the
rhombohedral crystals of insulin contain two zinc atoms and consisted of six molecules of
insulin. Extensive crystallographic studies of the three dimensional model of the insulin
molecule demonstrated that the hexamer of insulin with a molecular weight of 36000 Da (at
neutral pH in the presence of zinc) consisted of three slightly asymmetric dimers each with a
molecular weight of 12000Da (Adams et al 1969, Blundell et al 1972, Hodgkin, 1972) (see
figure 1).
Figure 1 (after Hodgkin, 1972)
Schematic representation of the three dimensional structure of insulin
20
The secondary structure of the monomers includes three a-helical regions (Blundell et al
1971 & 1972), of which two exist on the A chain and one on the B chain. The folding of the
B chain around the A chain core is also brought about with disulphide bridges between
cysteine residues. The globular structure achieved in this way provides the molecule with a
hydrophobic core which is essential for its structural stability, all polar residues are on the
surface and the non polar residues, mainly on the B chain, are buried during the formation of
the dimers and the 2 zinc-insulin hexamers. Insulin aggregation in solution depends on
concentration, pH, ionic strength and the presence of divalent cations and anions (Blundell
et al., 1972, Bradbury and Brown, 1977). X-ray studies demonstrate that most mammalian
and fish insulins form similar hexamers, (except the hagfish, guinea pig and coypu (Cutfield
et al., 1974, Peterson et al., 1974, Wood et a l, 1975).
Biosynthesis of insulin
Insulin is first synthesized as preproinsulin (molecular weight 11 500Da) on the rough
endoplasmic reticulum (RER) in the P-cells of the islets of Langerhans in the pancreas
(Patzelt et a l, 1978, Steiner et a l ,1974). The islets of Langerhans comprise 1-3% of the
pancreatic mass and are probably derived from the ectoderm (Bell and Hockaday, 1996).
The N terminus of preproinsulin has a hydrophobic extension of 23 amino acid residues
which directs release of the nascent polypeptide chain to the cisternal space of the RER
(Chan et a l, 1976). The signal sequence is rapidly removed by proteolytic cleavage and the
resulting proinsulin molecule (molecular weight 9390Da and 86 amino acid residues) is
transferred via the vesicles of the RER to the Golgi apparatus. This transfer is an energy
dependent process (Howell and Tyhurst, 1972) and by pulse labelling technique has been
shown to occur in about 15 minutes.
In the Golgi apparatus proinsulin is packed into clathrin coated trans-Golgi vesicles
containing an ATP-dependent proton pump (Orci et a l, 1986 & 1984). As the vesicles
mature, H+ ions are transported inwards, via the proton pump, causing a reduction in pH and
thus increasing the activity of the proteolytic enzymes and ensuring the conversion of
21
proinsulin to insulin within maturing vesicles. This process takes 60-90 minutes (Orci et al.,
1973)
The connecting peptide (C-peptide) is joined to the insulin molecule by basic amino acids;
two arginines at the amino end and a lysine and arginine at the carboxyl end. These are the
points of trypsin and carboxypeptidase B-like enzymatic cleavage which convert proinsulin
into insulin, consequently releasing these basic residues (Orci et al, 1986, Steiner and Oyer,
1967, Chance et al, 1968).
The conversion of proinsulin to insulin takes place in non-coated mature secretory granules
(Orci et al., 1984) and the removal of the C-peptide chain decreases its solubility. The
resulting insulin is crystallized with 2 zinc ions per 6 insulin molecules and is stored in
equimolar ratio with C-peptide in secretory granules where it is available for secretion
through exocytosis (Orci et al., 1984). A small amount of proinsulin co-crystallizes with
insulin (Steiner, 1973) and constitutes approximately 5% of insulin secretion.
Insulin secretion is stimulated primarily by a rise in glucose concentration. Glucose is
metabolised through the glycolytic pathway generating ATP within p-cells (Randle, 1993).
The increase in intracellular ATP concentration (Ashcroft et al., 1984) or ATP/ADP ratio
(Ashcroft and Ashcroft, 1990) leads to closure of the ATP sensitive potassium channels and
depolarisation of the p-cell membrane thus causing the entry of calcium through calcium
channels (Sussman et al., 1987). Calcium then activates Ca-calmodulin dependent protein
kinase which in turn leads to phosphorylation of specific protein substrates and the release
of insulin from the P cell.
The Insulin Receptor
The insulin receptor consists of a membrane spanning glycoprotein complex lodged in the
plasma membranes of cells, with an apparent molecular weight of 450 kDa and consisting of
two distinct glycosylated subunits (Yip et al, 1978 and 1980). There are two a-subunits and
two p-subunits linked covalently by disulphide bonds. The two a-subunits are located at the
22
extracellular face of the plasma membrane and contain the insulin binding domain. The two
p-subunits are held together by disulphide bonds, they consist of three regions: an
extracellular domain of 194 amino acid residues with a segment rich in cysteine residues and
forming disulphide bridges with the a-subunits, a hydrophobic transmembrane domain of 23
amino acid residues, and an intracellular domain of 403 amino acid residues that contains a
catalytic site and several sites of autophosphorylation (Kasuga et al, 1982 a and b, White et
al, 1985).
Following binding of insulin to its receptor a series of events is triggered: internalisation of
insulin, processing of insulin and/or the insulin receptor comlex and initiation of a wide
spectrum of metabolic processes. Experimental evidence suggests that insulin binds to its
receptor with high affinity in a ratio of 1:1 at physiological concentrations. Carpentier et al
(1979 a & b) using radio-iodinated (^ I) insulin demonstrated that the insulin receptor
complex is internalised (endocytosis), generating coated vesicles containing the complex in
the subplasmalemmal cytoplasm. The acidic pH within the endosomes causes dissociation
of insulin from the receptor (Farquhar, 1983; Tyko and Maxfield, 1982). Insulin is then
degraded within the secondary lysosomes. The internalised receptor (Fehlmann et al 1982 a
& b; Juul et al 1986) is either recycled to the cell surface from endosomes (Marshall et al
1981; Carpentier et al, 1986) or degraded (Heindreich et al, 1983; Berhanu et al, 1982).
Insulin affects many cellular systems including: transport of molecules across the plasma
membrane, levels of cyclic nucleotides, activities of key enzymes in intermediaiy
metabolism, rates of protein synthesis and degradation, rates of DNA and RNA synthesis,
including specific gene expression and cellular growth and differentiation. However, it was
not understood how the rapid effects of insulin (observed within seconds) could be
explained by the process of internalisation (Lamer et al., 1981).
Kasuga et al., (1982a) found that the addition of insulin to cells led to rapid phosphorylation
of the P-subunit on serine and threonine residues. Subsequently, the (3-subunit was shown to
be phosphorylated on tryrosine residues in cell free systems which led to the idea of tyrosine
kinase activity associated with the insulin receptor itself (Kasuga et al., 1982b).
23
It was then shown that immediately after insulin binds to the receptor a cytoplasmic protein,
is phosphorylated on tyrosine reidues (White et a l, 1985). This protein which is now
referred to as insulin receptor substrate 1 (IRS-1) has been purified from rat liver by affinity
chromatography and the cDNA of this protein has been isolated (Rosenberg et a l, 1991; Sun
et a l, 1991). This protein has 10 tyrosine phosphorylation sites. After phosphorylation, 6 of
these tyrosine-containing regions can interact with other proteins which have Src homology
(SH2) domains (Koch et a l, 1991). One of the cytoplasmic proteins which has the SH2
domain is phosphotidylinositol 3’-kinase. It has been demonstrated that
phosphotidylinositol 3’-kinase is associated with the IRS-1 (Sun et a l, 1991). Several other
signalling proteins containing SH2 domains eg phosphotyrosine phosphatase Syp (Kuhne et
a l, 1993) and growth factor receptor bound-2 (grb-2) (Baltensperger et a l, 1993) have been
shown to couple with IRS-1. This appears to connect the insulin receptor to a signalling
chain which involves GTP-binding proteins such as ras which interact with a cascade of
serine/threonine specific kinases (MAP-kinase cascade, mitten-activated protein
kinase)(Gammeltoft and van Obberghen, 1986).
In vitro studies with adipocytes treated with a tyrosine phosphatase inhibitor (okadaic acid)
demonstrated that IRS-1 was phosphorylated on serine and threonine residues, but in these
cells the stimulation by insulin of glucose transport was inhibited (Tanti et al, 1994).
Further studies involving microinjection of anti-IRS-1 antibody into rat fibroblasts have
suggested that IRS-1 initiates mitogenic activity of IGF-1 and insulin receptors (Rose et al,
1994). These findings have led to the suggestion that IRS-1 has a major role in insulin
mediated metabolic and mitogenic signalling and glucose transport.
Glucose Metabolism
Usually when the concentration of glucose is elevated (fed state) and other nutrients are
adequate, the liver, muscle and adipose tissues are geared to anabolic process and storage.
When dietary glucose is not available, ie in the catabolic state, glycogen stores from the liver
and fat and protein stores are mobilized to provide an alternative energy source.
24
Endogenous Glucose Production
Circulating plasma glucose, in the fed state, is mainly derived from the splanchnic bed.
However, during a short fast, liver alone, and in a prolonged fast both liver and kidney
(Flock et al., 1969) are geared towards glucose production. These two tissues contain
glucose 6 phosphatase activity which catalyses the conversion of glucose 6 phosphate to
glucose. Glucose can then be transported out of the cell by glucose transporters.
Glucose from the liver is produced by two processes: glycogenolysis and gluconeogenesis.
Initially glycogen stores in the liver are mobilized. Studies of serial liver biopsies in human
subjects (Nilson and Hultman, 1973) and nuclear magnetic resonance (NMR) studies to
measure the rate of glycogen depletion from the liver (Rothman et al., 1991) have suggested
that after an overnight fast, glycogenolysis contributes to 40-50% to the hepatic glucose
production. After depletion of glycogen stores (20 hours of fast) gluconeogenesis maintains
hepatic glucose output (Consoli et al., 1987). A variety of substrates are used in this
process. In muscle, during a prolonged fast glycogen stores and protein are released as
lactate and alanine which are then transported and used in the liver for glucose production.
Glycerol from adipose tissue during a prolonged fast is another gluconeogenic source and
provides 50% of the hepatic glucose output (Bortz et al, 1972 and Nurjhan et al, 1986).
Insulin affects glucose homeostasis by inhibiting hepatic glucose output and stimulating
glucose disposal.
Glucose Utilization
In most tissues glucose is taken up via transport systems that display saturation kinetics.
The intracellular glucose concentration is normally lower than the extracellular glucose
concentration, and glucose is transported down its concentration gradient by facilitative
diffusion via glucose transporters across the cell membrane which is thought to be the rate
limiting step (Britton, 1964) in glucose disposal.
25
Glucose taken up by the cells is either stored as glycogen or lipid, oxidised to CO2 and H2O
or converted to lactate and alanine. The Cori (lactate) and Glucose-alanine cycles account
for 20-25% of glucose utilization (Felig, 1973).
Various factors can influence the fate of circulating glucose, including the degree of fasting,
the availability of substrates such as free fatty acids and various endocrine factors. The
tissues involved in glucose utilization include splanchnic tissues, muscle, brain and adipose
tissue. Brain and muscle are the major tissues for glucose oxidation whereas splanchnic
tissues, adipose tissues and under certain circumstances muscle are geared for storage.
Both glucose production and utilization are affected by insulin and glucagon. The
glycolysis pathway where glucose is broken down to pyruvate and lactate is illustrated in fig
1. The steps where insulin or glucagon has an effect by causing phosphorylation or
dephosphorylation of the enzyme has been shown.
26
GLUCOSE
/GLYCOGEN
^ ^ glycogen
phosphoryase -GLUCOSE-1- PHOSPHATE -----
GLUT ' GLUCOSE* * c T I glucokinaseÎ 1Glucose-6-
- phosphstase | T hexokinase +
GLUCOSE-6-PHOSPHATEGlucose phosphate isomerase
Loss of 2- H
FRUCTOSE-6-PHOSPHATE
î l 6-phosphofructokinase 4"fiructose 1,6-
- biphosphatase
FRUCTOSE-l,6-PHOSPHATE
aldolase
DIHYDROXYACTONEPHOSPHATE T
Loss of 3-^H t iî i
+ Stimulation by Insulin - Inhibition by Insulin * * Stimulation by glucagon
OXALOCETATE ^
^ GLYCERALDEHYDE-3-PHOSPHATEglyceraldehydes-phosphatedehydrogenase
3-PHOSPHOGYCEROYLPHOSPHATEphosphoglycerate
3-PHOSPHOGLYCERATE
^ ^ phosphoglyceromutase
2-PHOSPHOGLYCERATE
enolase
PHOSPHOENOLPYRUVATE (PEP)
Pyruvatekinase +
PYRUVATEpryruvate carboxylase
Loss of 6- H
lactatedehydrogenase
LACTATE
Pyruvatedehydrogenase
ACETYL-CoA
Figure [ 2]- The glycolysis pathway showing the effects of insulin and glucagon on the enzymes involved and the steps at which 2-[^H], 3-[^H] and 6-[^H]glucose lose their label. PEPCK (phosphoenolpymvate carboxykinase) (Umpleby and Sonksen, 1987). GLUT is representing glucose transporters.
27
Glucose Homeostasis
The regulation of plasma glucose levels is of central importance. Tissues such as muscle
and liver are able to metabolise substrates other than glucose, however the brain almost
exclusively utilizes glucose as a source of energy. Some researchers have demonstrated that
the brain is able to metabolise ketones and lactate (Veneman et al, 1994, Maran et al, 1994,
Amiel et al, 1991). The brain has high energy requirements estimated at approximately
Img/kg/min glucose corresponding to about lOOg glucose per 24 hours, however it has low
energy reserves (enough for only a few minutes), and is unable to synthesize glucose (Bolli,
1998). The energy utilised is required to maintain the ionic gradient across nerve
membranes critical for maintenance of brain activity, and as a result of brainstem functions
critical in maintaining life. The brain is therefore mainly dependent on the maintenance of
a constant supply of glucose from the blood.
Intricate mechanisms to protect the supply of glucose to the brain have therefore evolved.
Only the liver and the kidney are able to release glucose. These organs contain glucose-6 -
phosphatase which liberates glucose into the bloodstream (Bolli and Fanelli, 1999). The
liver and the kidney do not contribute equally to postabsorptive glucose homeostasis. It is
generally believed that the liver makes the major quantitative contribution during this state,
whereas the kidney has a minor role, except during prolonged starvation, when it is
estimated that the kidney may contribute almost half of glucose production (Bolli 1998,
Cahill et al,1966,Cahill 1970).
In normal subjects the plasma (extracellular) glucose concentration remains between 3.5 and
7 mmol/1 despite intermittent glucose loads entering the body from the gastrointestinal tract
following meals. The renal tubule absorbs all glucose from the glomerular filtrate up to
concentrations of around lOmmol/1, so that in the non-diabetic individual with a normal
renal threshold, urine is glucose free. Blood glucose is sensed in the brain (Frohman and
Nagai, 1976, Ritter et al, 1981, Borg et al, 1995), pancreas (Taborsky et al, 1998) and liver
(Connolly et al, 1996), each of these organs is involved in the counter-regulation response to
low blood glucose in healthy subjects.
28
The release of insulin, glucagon, growth hormone, catecholamines, glucocorticoids and
insulin like growth-factor 1 (IGF-1) have all been shown to be sensitive to changes in
glucose concentration (Garber et ah, 1976, DeFronzo et ah, 1977, Cryer, 1991, Bolli and
Fanelli, 1999).
Both insulin (secreted from the (3-cells of the pancreas) and glucagon (secreted from the a
cells) play an important role in the regulation of glucose homeostasis. A fall in blood sugar '
concentration that may occur in the postabsorptive state leads initially to a reduction in
insulin secretion, by direct action of glucose on p-cells of the islets of Langerhans (Cryer
1997, Bolli, 1998). This leads to a reduction in the plasma insulin concentration in the
portal circulation, and therefore a reduction in the suppression of hepatic glucose output (or
a rise in hepatic glucose output). If the blood glucose continues to drop counter-regulatory
hormones are stimulated and released, oc-cells of the islets of Langehans in the pancreas are
stimulated to release glucagon, this is reversed in hyperglycaemia (Matschinksy and
Bedoya, 1989). Like insulin, glucagon is also released into the portal circulation, it has no
effect on peripheral glucose uptake, but increases hepatic glucose production. The increase
in hepatic glucose output is early and sustained probably because of the sustained
stimulation of hepatic gluconeogenesis, the main mechanism of increased hepatic glucose
output. The kidney does not appear to be affected by glucagon stimulation.
However, hyperglycaemia per se (with suppression of insulin and glucagons secretion) has
an inhibitory effect on hepatic glucose production in man (Liljenquist et al., 1979). In
addition, during hypoglycaemia induced by insulin in humans, Bolli et al (1985) found that
when glucose concentrations decreased below 2.5mmol/l there was a residual hepatic
glucose output independent of neurohormonal influences or an autoregulatory process by the
liver. In dogs also, it has been shown that this residual hepatic glucose output is maintained
by glycogenolysis and gluconeogenesis after a prolonged hypoglycaemia induced by insulin
infusion (Tyler Frizzell et al., 1988).
29
ACTH is released from the pituitary and stimulates a rise in cortisol several hours after
hypoglycaemia. Short term infusion of cortisol (the main glucocorticoid in man) affects
glucose homeostasis by increasing hepatic glucose output and decreasing peripheral glucose
disposal (Rizza et al, 1982). However studies of patients with Cushing’s syndrome (ie with
long term effects of excess glucocorticoids) using an intravenous bolus of D-[3-^H] glucose
have shown that the hyperglycaemia sometimes found in these patients is primarily due to
decreased glucose disposal rather than increased glucose production (Bowes et al., 1991).
Furthermore, both basal (Bowes et al., 1991) and glucose stimulated insulin secretion
(Kitabchi et al., 1973) are increased in the presence of excess glucocorticoids. In contrast, in
the absence of glucocorticoids, insulin secretion is suppressed, for example following
adrenalectomy in man, although the absence of the catecholamines (adrenaline and
noradrenaline) may contribute to the effect of this example (Lenzen and Bailey 1984;
Marker et al., 1991).
Adrenaline is released from the adrenal medulla as a result of stimulation of the sympathetic
nervous system (Bolli and Fanelli, 1999). Catecholamines exhibit multiorgan effects.
Adrenaline reduces secretion of insulin from the pancreas, lowering portal plasma insulin
concentration further, and thus reducing the suppression of hepatic glucose output. It also
increases hepatic and renal glucose output, probably secondary to increased
gluconeogenesis. Adrenaline reduces peripheral glucose uptake and stimulates lipolysis.
The stimulation of lipolysis contributes to counter regulation via non esterified fatty acids
which stimulate hepatic gluconeogenesis and suppress glucose oxidation and use. The effect
of free fatty acids on the production of glucose is greater than that on the use of glucose.
Noradrenaline concentration in plasma rises less, but it acts though the same pathways as
adrenaline with similar effects. It also inhibits insulin secretion and stimulates glucagon
synthesis.
Growth hormone is released from the pituitary as a result of hypoglycaemia, it also increases
hepatic glucose output and suppresses peripheral glucose uptake. It stimulates lipolysis and
therefore increases plasma concentration of non esterified fatty acids and glycerol.
30
Hepatic autoregulation occurs when hypoglycaemia is severe. Investigators have
demonstrated that endogenous glucose production increases despite complete neurohumoral
counter regulatory blockade. (As reviewed by Bolli 1998 and Bolli & Fanelli 1999).
A large part o f the counter regulation hormones’ effect, at least the ones that possess
peripheral effects are mediated by the common pathway of increased lipolysis and greater
plasma levels of free fatty acids. As described above these fatty acids exert their counter
regulatory effect mainly at the liver. Thus all counter regulatory responses to
hypoglycaemia increase hepatic glucose output either directly alone or in combination with
an indirect effect by the stimulation of lipolysis, some also decrease peripheral glucose
uptake. In normal physiology these defence mechanisms are appropriate as insulin is
released into the portal circulation and has its first effects at the liver, and the liver plays a
vital role in the production of glucose to support brain metabolism.
In insulin dependent diabetes no insulin is released by the pancreas in most cases exogenous
insulin is administered by subcutaneously delivery, it circulates peripherally first, and is
controlled by its pharmacokinetics, with peak actions occurring 1 to 8 hours after injection
(Bolli, 1998). The insulin profile does not resemble physiological insulin profiles of a
normal 24 hour day, and is unable to respond to changing insulin need (Amiel, 1998, Cryer,
2003). In this situation the liver is relatively underinsulinised and peripheral tissue over
insulinised. In the event of hypoglycaemia, the first line of counter regulation (that is a
decrease in insulin release) is not possible, insulin levels remain inappropriately high
suppressing hepatic glucose output, increasing peripheral glucose uptake and inhibiting
lipolysis. Further counter regulation has evolved to exert itself primarily at the liver,
suppression of hepatic glucose output is thus easily overcome since in these circumstances
the liver is relatively underinsulinsed, however the peripheral over insulinisation presents a
more difficult obstacle, and therefore it could be argued that greater counter regulation has
to occur to prevent a further fall in plasma glucose and maintain the brain’s glucose supply.
This may cause later hyperglycaemia as a result of the sustained release of counter
regulatory hormones and their prolonged duration of action may be a factor in the overall
deterioration in glycaemic control seen in patients with frequent hypoglycaemia. Some have
31
proposed that intraperitoneal insulin would be a more physiological replacement (see hepatic
insulinisation page 36)
Fat and Lipid metabolism
Lipids in the plasma exist as free cholesterol, cholesterol ester, triglycerides (glycerol and
fatty acids) and phospholipids (triglycerides and a phosphate derived group).
Plasma lipids are transported as lipid protein complexes. These lipoproteins are classified as
chylomicrons, very low density lipoproteins (VLDL), intermediate density lipoproteins
(IDL), low density lipoproteins (LDL) and high density lipoprotein (HDL) according to their
densities. Insulin stimulates fatty acid synthesis from pyruvate and lactate (Denton et al.,
1975) by increasing the activity of acetyl CoA carboxylase and pyruvate dehydrogenase and
increases the estérification of fatty-acyl CoA and glycerol to form triglycerides. Acetyl CoA
carboxylase is inactivated by reversible phosphorylation which is catalysed by the enzyme
cyclic adenosinemono-phosphate (cAMP) dependent protein kinase. Hormones that
decrease the concentration of cAMP (for example insulin) will activate the enzyme whereas
those that elevate the concentration of cAMP (for example adrenaline or glucagon) will
cause the inactivation of acetyl CoA carboxylase (Brownsey and Denton, 1982; Holland et
al., 1985).
Insulin also inhibits lipolysis (Garratt et al., 1979) in adipose tissue by dephosphorylation of
the enzyme triacylglycerol lipase to an inactive form via lowering cAMP concentrations
hence decreasing the activity of the protein kinase. The activity of lipoprotein lipase which
hydrolyses chylomicrons and other lipoproteins to fatty acids and glycerol is increased by
insulin in a dose dependent manner. The increase in activity is caused by cleaving the
enzyme from the cell membrane via stimulation of a phosphatidylinositol-specific
phospholipase C (Chan et al., 1998).
32
When the concentrations of fatty acids and ketone bodies are elevated as during starvation or
exercise there is an increased rate of their oxidation. This reduces glucose oxidation and
utilization and increases the flux through gluconeogenesis (Randle et al., 1963). In contrast,
in the presence of an elevated glucose concentration lipid oxidation is suppressed due to
inhibition of lipolysis and increased fatty acid re-esteriflcation (Sugden et al., 1989).
The Endothelial Barrier
The capillaries are the sites of exchange of substances between flowing blood, extracellular
fluid and tissue cells. The endothelial barrier provides a membrane across which substances
move both actively and passively in response to pressure and concentration gradients. For
small molecules (for example water, gases and small ions) equilibration is rapid (Hill 1928)
but for macromolecules (for example yglobulins, albumin and probably insulin)the process
is slower (Renkin, 1964). In general, there are two continuous layers in the walls of the
capillaries: the endothelium and its underlying basement membrane or basal lamina. The
permeability of the capillaries is dependent on the porosity of these layers.
Endothelial cells of some capillaries have been shown to have a large proportion of small
vesicles and fenestrae (Wissig and Charonis, 1984). Vesicles have been shown to shuttle
fluid across endothelial cells (Palade, 1961). Fenestrae are openings throughout the
endothelial plasma membrane with a range of pore diameters around 30-80nm closed by a
delicate unilaminar diaphragm 2-4nm thick with a nodal swelling 15nm in diameter at its
centre (Clementi and Palade, 1969).
Endothélia of different types of capillaries differ in their permeability (Renkin 1964). The
tightest capillaries are those of the central nervous system, retina, iris, and larger peripheral
nerves. The endothelial cells of these capillaries have tight junctions, lack fenestrae and
have minimal vesicle activity (Kamovsky et al 1967). Capillaries of muscle and skin have
small openings in the junctions between endothelial cells or nonrestrictive gaps between
cells.
33
A bidirectional shuttle of pinocytic vesicles has been described for transport of
macromolecules across their walls independent of fluid volume movement (Garlick and
Renkin, 1970, and Renkin 1979). Data from studies with various sized dextrans,
polyvinylpyrrolidone and proteins ranging from l - 1 2 nm radius from skeletal muscle
demonstrated that solutes measuring 4-12nm in radius drained from lymphatics and did not
easily cross through the capillary wall (reviewed by Parker et al 1984). Wissig, 1979 and
Williams and Wissig, 1975 showed after intravenous injection of a variety of molecules with
peroxidase activity (size range 1900-40000 Da or molecular diameter 2-5nm) that the
smaller molecules leave the capillary lumen more rapidly and reach a higher concentration
in the pericapillary interstitium of the diaphragm muscle than do larger molecules. It was
shown that ferritin (an electron dense protein probe) with Mr 750 000 Da and molecular
diameter lln m was unable to leave the capillary lumen (Bums and Palade, 1968b). The
behaviour of these molecules in the size range 1,900-750 OOODa suggests that they use a
different diffusion pathway with some restrictions to enter the interstitium. Two structures
have been proposed to serve this pathway, 1 , the transendothelial vesicular channels and 2 ,
the cleft between endothelial cells (Wissig and Williams, 1975).
The most permeable capillaries are found in the liver. In the liver, sinusoids are lined by
highly fenestrated epithelial cells, between which there are large gaps. The sinusoids also
have no basal lamina (Wisse 1970). Because of the large gaps in the endothelial cell layer
and the absence of a continuous basal lamina, no significant barrier exists between blood
plasma in the sinusoid and the hepatocyte plasma membrane (Ross et al, 2003). This is
known as the “space of Disse” (the perisinusoidal space)(Gray, 1973, Reichen, 1999).
Measurement of circulating chylomicrons in the space of Disse in the rat liver revealed a
substantial filtration effect of endothelial fenestrations only on chylomicrons greater than
250nm in diameter (Naito and Wisse, 1978).
Early infusion studies with pork proinsulin in dogs (Tomkins et al 1981) and
hyperinsulinaemic-euglycaemic clamp studies with human proinsulin in healthy volunteers
(Revers et al, 1984, Glauber et al, 1986, Galloway et al 1992) suggested that proinsulin was
34
more active in the liver (glucose Ra) than in the periphery compared to (glucose Rd) insulin.
The reasons for the differential effect of proinsulin on glucose Ra and Rd could be attributed
to differences in local distribution or metabolism, to differences in binding or activation of
the receptors or to differences in the post-receptor cascade. Although proinsulin had 8-9%
biopotency of that of insulin, the postreceptor activities of insulin and proinsulin were
shown to be the same in cultured rat hepatocytes (Agius et al, 1986). Furthermore, Ciraldi et
al (1986) showed that glucose transport was affected similarly for insulin and proinsulin in
rat adipocytes in vitro.
The in vitro results suggested that differential observations on glucose Ra and glucose Rd
from the in vivo studies could simply be due to the differences in distribution of proinsulin
in the liver compared to the periphery in comparison to insulin. The observation that
proinsulin exhibited apparent hepatoselectivity promoted the suggestion that the mechanism
of this differential effect could relate to the effect of the increased molecular size (proinsulin
9390 Da versus an insulin 5807 Da) on the rate of passage through the capillary walls and
hence to insulin receptors in intact tissues. Low receptor affinity per se does not convey
relative hepatoselectivity (Chap et al, 1987) In support of the hypothesis for the existence of
a capillary endothelial barrier is the observation that monomeric insulin analogues (eg B25
Asp, B9 Asp and B27Glu)(Brange et al, 1988, Kang et al, 1991) are absorbed more rapidly
from subcutaneous tissue in the reverse direction through the capillary endothelial barrier
than native insulin which at concentrations used for treatment subcutaneously are mostly
dimeric and hexameric. Hyperinsulinaemic euglycaemic clamp studies performed in pigs
demonstrated that monomeric insulin analogues (BIO Asp and B9Asp and B27 Glu) were
equipotent in stimulating glucose Rd and inhibiting glucose Ra in comparison to human
insulin (Ribel et al, 1990).
35
Albumin
pores
liver sinusoid
periphery
endothelium
no pores
Figure 3. Schematic representation of albumin binding of insulin detemir, and potential
mechanism for differences in effect on the liver and periphery.
Hepatic insulinisation
In normal physiology the pancreas releases its hormones into the hepatic vein these
hormones must pass the liver before reaching the systemic circulation (Figure 4.). 40-60%
of the insulin presented to the liver is removed in a single passage (Field et al 1973, Ishida et
al 1983, Chap et al 1987). Flepatic extraction of insulin is therefore potentially important for
control of the circulating concentration of the hormone. Using insulin pharmacologically in
the treatment of Type 1 diabetes, insulin is administered subcutaneously and absorbed
directly into the systemic circulation (Figure 5). In the absence of the portal-peripheral
gradient, glycaemic control is only achieved by relative portal-peripheral hyperinsulinaemia.
Many factors will influence hepatic exposure to subcutaneously delivered insulin including,
rate of absorption from the injection site into the venous system, the volume of distribution,
possible insulin extraction in the pulmonary and mesenteric capillary beds before reaching
36
Figure 4. Schematic Representation of Physiological Insulin Production.
food
Peripheral Tissuesi'oticrms
liisuiln
atllpos Glucose
Figure 5. Schematic Representaion of Pharmacological Replacement of Insulin.
food
Peripheral lissuKiPancreas
muscle
Insulin
adipose
Liver
Glucose
Insulin
37
the hepatic portal vein and the blood flow in the hepatic artery. However, concentration of
insulin after intraportal administration is affected only by portal vein flow (from the gut).
The effects of insulin on glucose and fat metabolism are not fully understood. Hormonal
control of glucose metabolism may be either direct by inhibition of glycogenolysis and or
gluconeogenesis in the liver (Hers and Ave, 1983; Chiasson et al 1976), or indirect by
restriction of flux of gluconeogenic substrates or rate of lipolysis from peripheral tissues,
(Giaccaetal, 1992).
Many studies have shown the importance of intraportal delivery of insulin. Intraportal
delivery of insulin to the liver has been shown to have an important role in the regulation of
hepatic glucose output in normal and diabetic dogs (Stevenson et al 1978, 1981 and 1983)).
Ishida et al (1984) demonstrated that in depancreatized diabetic dogs the liver is more
sensitive than peripheral tissue to small changes (a rise from less than 2|uU/ml to 9 and 16
pU/ml) in the plasma insulin concentration. Griffen et al (1987) performed a study
catheterising either the jugular vein or the portal vein (ie systemic versus portal insulin
delivery) in diabetic rats. They demonstrated that insulin exerts its metabolic and mitogenic
effects by direct action on the liver and established that the insulin delivery via the portal
vein is important for glucose homeostasis. Spangler, 1985 and Spangler et al, 1986)
demonstated that peripherally delivered vesicle encapsulated insulin preparations (VEI) and
insulin had an equivalent capacity to suppress hepatic glucose output but VEI was less
effective in stimulating glucose disposal.
Whilst these studies suggest that intraportal delivery of insulin is important in the control of
glucose Ra other data have suggested that systemic delivery of insulin may be more
effective in the suppression of glucose Ra. Glucose turnover studies with hyperinsulinaemic
euglycaemic clamping, in depancreatised dogs (Giacci et al 1992) or in normal dogs
combined with an infusion of somatostatin (Ader and Bergman 1990) demonstrated that
systemic delivery of insulin was more effective in the suppression of hepatic glucose output
than portal delivery.
38
In kidney pancreas transplanted patients, the surgical anastomoses are such that hormones
released from the transplanted pancreas drain into the systemic circulation. This surgical
method abolishes the normal portal-peripheral insulin gradient and also results in relative
hyperinsulinaemia in the periphery. Data from metabolic studies in patients with such
transplants compared to a control group using infusion of D[6 -^H] glucose and
bicarbonate and a meal enriched with D-[2-^H] glucose, suggest that systemic delivery of
insulin normalised the basal glucose concentrations to a similar extent as the control group
and there was a tendency for a more effective suppression of hepatic glucose Ra than in the
control group (Katz et al 1991).
Other studies find portal and systemic delivery of insulin to be equally effective in the
suppression of glucose Ra. A comparative study of glucose turnover performed in diabetic
rats which had received transplanted islets either under the kidney capsule (peripheral
delivery) or in the splenic pulp (portal delivery) (Kruszynska et al, 1985) indicated that
glucose turnover and recycling were not different between the groups. In another study the
direct and indirect action of insulin on carbohydrate metabolism during fasting and after an
oral glucose load has been investigated in a group of dogs in which the portal vein was
anastomosed to the inferior vena cava (peripheral delivery) (Kryshak et al 1990). This
resulted in insulin sectreted by the intact pancreas passing directly into the systemic
circulation, bypassing extraction by the liver and thereby producing hyperinsulinaemia in
the periphery. Glucose Ra and Rd, measured with D-[2-^H]glucose and D-[3-^H]glucose,
after a meal ingestion in these dogs, was not different from a control group suggesting that
glucose Ra may be controlled by a peripheral signal.
Cherrington in his review “control of glucose uptake and release by the liver in vivo”
(1997), discusses glucose homeostasis when fasting, as well as the effects of glucoregulatory
feedback loops in the regulation of hepatic glucose output. In particular the direct and
indirect role of insulin on hepatic glucose output is discussed. Evidence for indirect
mechanisms of insulin action on hepatic glucose output demonstrated by Bergman and
Giacca are presented as well as further studies by the author. Cherrington concludes that
39
“insulin thus appears to acutely inhibit liver glucose output through a direct effect on
glycogenolysis and an indirect effect on glycolysis/gluconeogenesis. When one considers
that the sinusoidal insulin level in a normal individual is 2-2.5 times the arterial insulin level
and that the liver is extremely sensitive to alterations in the portal insulin concentration, it
becomes evident that changes in insulin secretion must bring about the majority of their
effects via a direct action on the liver. It is also obvious, however, that when insulin is
administered peripherally, such as occurs in insulin-requiring diabetic individuals, the
indirect effects of the hormone on the liver will take on greater significance.”
One explanation for the conflicting conclusions of the experiments described in this section
could be the choice of mathematical model used for calculation of glucose Ra and Rd. The
results from most of these studies were calculated using Steele’s model (a non-physiological
mono-compartment model; Steele, 1959 and Steele et al, 1965) which has been shown to
underestimate glucose Ra and may introduce errors in the conclusions reached. More
experiments and better mathematical models of glucose kinetics are required to establish the
effects of insulin on glucose metabolism.
Long Acting insulin
Background
Initially insulin treatment consisted of subcutaneous injections of insulin shortly before
meals. In order to avoid multiple injections a number of attempts were made to prolong the
duration of action of insulin or to slow its absorption. Substances used were gum arabic
(1923), lecithin (1923) and oil suspensions (1925) (Rosskamp and Park 1999). These
attempts were mostly unsuccessful because of poor stability or side effects at the local
injection site.
40
The first successful insulin preparation with a prolonged action was protamine insulin
(Hagedom et al 1936). Here the principle of protracting the insulin action was to make the
insulin less soluble at the neutral pH of tissue fluid. When injected as a suspension, the
duration of action was roughly twice as long as that of regular unmodified insulin.
Protamine insulin is a complex of the fish protein protamine and insulin. When injected
subcutaneously proteolytic enzymes degrade the protamine and the insulin is absorbed. A
further way of protracting the action of insulin was the so-called lente series. These were
preformed amorphous or crystalline suspensions obtained by addition of small amounts of
zinc ions (Hallas-Moller 1956). Since their introduction in the 50s the technology of
prolonging insulin action has remained relatively unchanged, until now.
In the 1990s the link between metabolic control and development of micro vascular
complications became firmly established (DCCT). This has led to a renewed interest in
long-acting insulins on the basis of a new therapeutic concept. Now the objective is not
avoidance of multiple injections but the imitation of the physiological secretion of basal
insulin in healthy subjects, with additional mealtime insulin requirements taken care of by
short acting insulin bolus injections before meals. This basal-bolus concept is the
cornerstone of any kind of intensified insulin treatment, which leads to significantly better
metabolic control than one or two insulin injections per day can provide (Rosskamp and
Park 1999). However this therapy is not without its problems, currently available long
acting insulins were recognised not to be good at imitating physiological secretion of basal
insulin. They had peaks of action, and they were variable in absorption (Barnett, 2003).
Therefore renewed interest in prolonging the duration of action of insulin led to
developments along different strategic routes.
It was hoped that proinsulin would increase basal insulin concentrations, but concerns on
cardiovascular safety that arose from clinical studies meant this work was abandoned
(Owens et al, 2001). Proinsulin’s metabolite, des (64,65)-human proinsulin had an even
shorter period of action than NPH insulin and was discontinued (Owens et al, 2001).
Slowing absorption and clearance by strengthening hexameric interactions in the form of
cobalt (Ill)-insulin complex also proved unsuccessful (Burge and Schadel, 1997). However,
41
increase of the isoelectric point of insulin from 5.4 towards neutrality thereby causing
precipitation or crystallisation at the site of injection, delays absorption, and prolongs the
effect of insulin (Markussen et al, 1990). The first acid soluble extended action insulin
analogue made by this method was Novosol Basal (GlyA21, ArgB27, ThrBSO-NHz-insulin)
(Jorgenssen et al, 1989), but escalating dose requirements (reduced bioavailability) possibly
associated with local inflammatory reactions at the site of injection, meant that this insulin
was also abandoned.
On the basis of the same concept, a di-arginylinsulin analogue known as insulin glargine
was developed that had a much slower and more reproducible rate of subcutaneous
absorption than NPH insulin (Owens and Barnett, 2000), resulting in slower onset and
extended effect. An additional phenol molecule bound to the hexamer of crystalline insulin
glargine might further contribute to its protracted action. Plasma glucose and insulin
concentrations vary less between patients when insulin glargine is used compared to
ultralente insulin (Lepore et al, 2000, Heineman et al, 2000).
42
Insulin detemir
Physical and chemical properties
Insulin detemir is produced by a process including recombinant DNA technology using
Saccharomyces cerivisiae and chemical synthesis. The molecular formula is C267H402N 6 4S6
and the molecular weight is 5916.9 Da. In comparison to human insulin, the amino acid
residue at position B30 has been omitted, and a 14-C fatty acid chain has been attached to
position B29. The molecule is lipophilic due to the fatty acid chain. By adding zinc to the
formulation, the molecule is stabilized and hexamers are formed making it hydrophilic and
thereby soluble in water (Kurtzhals et al, 1995, Markussen et al, 1996).
Figure 6 .
ScheinniatDC Stractyire of InisylDim DetemSr
H2 0
H,N
Amino Acid R esidues A l - A21 COOH
Amino Acid R esidues B1 - B29 Lys® ® COOH
C =0I
(CH2)i2I
CH3
Pharmacology
Like native insulin, insulin detemir exists predominantly as hexamers in the prescence of
zinc and phenol. The fatty acid side-chains contribute to the aggregation of hexamers.
43
which results in delay in dissociation into monomers, and thereby may delay systemic
absorption (Havelund et al, 2004, Brunner et al, 2000), In monomeric state, the 14-C fatty
acid chain attached to position B29 binds to the fatty acid binding sites of albumin in the
subcutis, which further delays absorption into the bloodstream (Havelund et al, 2004,
Brunner et al, 2000).
Figure 7.
Structure of the insulin detemir dimer
Myristic acid
Myristic acidOlsen HB and Kaarsholm NC. Biochemistry 2000;39:11893-11900
44
Figure 8.
Di-hexamerisation of insulin detemir
Size exclusion chromatography of insulin detemir after exclusion of phenol (as expected in subcutis) reveals formation of dihexamers ->
hexameridihexamerequilibrium
Myristicacid
Olsen HB and Kaarsholm NC. Biochemistry 2000;39:11893-11900; Havelund S et al. Submitted 2003
Formulation
There have been three formulations of insulin detemir used during its clinical development
program. Formulation A was the first and contained 600 nmol/ml insulin detemir, and
glycerol as an isotonic agent. One dosing unit was defined as 6 nmol, the same as in current
human insulin products. Formulation B contained 1200 nmol/ml insulin detemir and had
mannitol as isotonic agent. One dosing unit was defined as 12 nmol. The final formulation
of insulin detemir formulation C contains 2400 nmol/ml, one dosing unit being defined as 24
nmol.
The constituents in formulations B and C are insulin detemir, phenol, m-Cresol, mannitol,
disodium phosphate, sodium chloride, zinc (as zinc acetate), hydrochloric acid and water for
injections. Both formulations are clear and colourless and are at pH 7.4. In this study we
have used formulation B. (Investigator’s Brochure, Soluble, basal insulin analogue Insulin
Detemir (NN304))
45
Efficacy pharmacology
The receptor binding properties of insulin detemir have been studied using soluble human
insulin receptors and the human hepatoma cell line HepG2. The affinity of insulin detemir
for the insulin receptor was about 18-46% relative to human insulin using soluble receptors,
(Markussen et al, 1996 and Kurtzalls et al, 2000) and 27% (when corrected for albumin
binding) using HepG2 cells. Correction for albumin binding is necessary to get a measure
for the receptor affinity, when assay buffers containing albumin are used, as only free (non
albumin bound) insulin detemir is available for binding to the receptor.
Insulin binding/dissociation kinetics were studied using Chinese Hamster ovary cells over
expressing the insulin receptor (CHO-hIR cells). Insulin detemir dissociates about 2 times
faster from the insulin receptor than human insulin (Kurtzhals et al, 2000).
Activation of the insulin receptor tyrosine kinase (IRTK), was studied using CHO-hIR cells.
Insulin detemir was approximately 9.5% as potent as human insulin in stimulating IRTK
activity (corrected for albumin binding) (Kurtzhals et al, 2000).
The IGF-1 receptor binding properties of insulin detemir were examined using soluble
human IGF-1 receptors, and using HepG2 cells. The affinity of insulin detemir for the IGF-
1 receptor was determined to 16% relative to human insulin using soluble receptors and to
41% (corrected for albumin binding) using HepG2 cells (Kurtzhals et al, 2000).
Pharmacokinetics and protein binding
The albumin binding properties of insulin detemir have been extensively studied using an
immobilised human serum albumin preparation (Kurtzalls et al, 1995 & 1997). The binding
constant of insulin detemir for HAS was determined to be 2.4 x 10-5 M-1 at 23 *C and 1.0 x
10-5 at 37®C in this assay system. Based on this binding constant, insulin detemir is
estimated to be 98.4 % albumin bound in plasma (37°C, 0.6mM albumin). This estimate is
46
consistent with in vitro studies using ^^^I-labelled insulin detemir in human plasma which
determined the plasma protein binding (98.8%) of insulin detemir.
The interaction between albumin binding of insulin detemir, fatty acids and selected drug
substances including warfarin, acetylsalicylic acid, diazepam, phenylbutazone, valproic acid,
ibuprofen, tolbutamide and glibenclamide was also examined in vitro using the immobilised
human serum albumin preparation (Kurtzalls et al, 1997). Insulin detemir binds to the fatty
acid binding sites of the albumin molecule. The binding of insulin detemir to the
immobilised human serum albumin preparation was independent of the binding of drugs in
the two major binding pockets located in domains HA and IIIA of the albumin molecule.
None of the tested drugs competed with the albumin binding of insulin detemir at clinically
relevant drug to albumin concentration ratios. In conclusion the in vitro studies do not
suggest any clinically relevant albumin binding interactions.
Figure 9.
Distinct binding sites on albumin
Nfvr2
Curry S et al. Nature Structural Biology 1998:5:827-835
47
Mitogenicity
The biological effects of insulin are both metabolic and mitogenic. The metabolic potency
was evaluated by measuring stimulation of lipogenesis in primary mouse adipocytes.
Insulin detemir stimulated lipogenesis with a potency of approximately 29% of human
insulin (corrected for albumin binding) (Kurtzalls et al, 2000). This is consistent with the
relative affinity to the insulin receptor.
The mitogenic potency of insulin detemir was evaluated using Chinese Hamster ovary cells,
K1 strain (CHO-Kl), human mammary cancer fibroblasts (MCF-7 cells) and human
osteosarcoma cells BIO (Saos/BlO cells), respectively. The mitogenic potency of insulin
detemir was determinined to be approximately 9% (CHO-Kl), 14.6% (MCF-7) and 11%
(Saos/BlO) relative to human insulin after correction for albumin binding (Kurtzhals et al,
2000).
Taken together the in vitro pharmacology studies indicate that insulin detemir is less potent
than human insulin in binding to insulin and IGF-1 receptors and in stimulating metabolic
and mitogenic responses.
Aims of this study
To determine whether insulin detemir by the nature of being a larger protein bound insulin
molecule has different effect on glucose and lipid metabolism when compared to NPH
insulin.
48
Chapter 2
Clinical methodology
Introduction to the glucose clamp technique
A particular problem encountered in studies investigating glucose metabolism is the
interrelationship between glucose concentration and insulin. Initial attempts were made to
make comparisons during a standard glucose tolerance test but both insulin and glucose
concentrations are in a state of flux, thus rendering it difficult to separate cause and effect. The
feedback loop between glucose and insulin is broken by controlling blood glucose
concentration by frequent measurement and altering an exogenous glucose infusion. Two
types of clamp can be performed:
i. Hyperglycaemic clamp in which blood glucose is held at hyperglycaemic levels and the
insulin response measured.
ii. Hyperinsulinaemic euglycaemic clamp in which an insulin infusion is given and blood
glucose is maintained with an exogenous infusion of glucose.
These two methods measure beta cell responsiveness and insulin action respectively. The latter
was used in the studies reported here to compare the actions of insulin detemir and NPH.
Isotope dilution
The tracer dilution technique was introduced by Stetton et al in 1951 for measuring glucose
turnover using a constant infusion of labelled glucose. Later, administration of a priming
injection of labelled glucose followed immediately by a constant infusion of the labelled
glucose was found to shorten the time required for mixing the labelled glucose in the body
49
glucose pool (Steel et al., 1956; Searle et al., 1954). Constant isotope infusions were also
used to measure the glucose appearance rate in non-steady state conditions during
hyperinsulinaemia (Steele et al., 1956). The tracer dilution technique in combination with a
constant insulin infusion and a euglycaemic clamp was used by Steele et al. (1965). This
method provided a means for the assessment of the differential effects of insulin on glucose
metabolism in the absence of any fluctuations in the concentration of glucose or counter-
regulatory hormones. Using these techniques in most circumstances the endogenous rate of
glucose appearance (Ra) equals the hepatic glucose production rate as apart from a small
contribution by the kidney, the liver is the only organ responsible for de novo glucose
production (Dmry et al\95V).
There are potential inaccuracies in these methods. For example during hyperinsulinaemic
euglycaemic clamps many workers have reported negative values for hepatic glucose
production, which is clearly impossible (Defronzo et al 1985, Butler et aï 1990, Ward et al
1990). Everyone has assumed when negative values are reached that hepatic glucose output is
zero but this calls into question the methodology used. The source of the errors in such studies
has stimulated much discussion and investigation. Potential sources of error include:
a) Inability of the simple model to cope with the complex in-vivo physiology of glucose
metabolism. The Steele model is a simple one compartment model with a fixed pool volume
which does not mirror the complex multicompartment physiology of glucose metabolism.
More sophisticated models have been proposed. However the original Steele model is still
used widely as it is easily understood and suited to individual clinical studies (Insel et al 1975,
Radzuik et al 1974, Steele et al 1974). The errors generated are proportional to the rate of
change of tracer enrichment and this can be offset by “spiking” the exogenous glucose infusion
with tracer to ensure uniform plasma enrichment during the study (Finegood et al 1987).
b) Tracer recycling. The isotope dilution technique assumes that once labelled tracer
has entered a metabolic pathway it does not return to the pool in its original molecular form. If
a tracer does return to the pool it is “recycling”. Different tracers lose their labels at different
positions in their metabolic pathways (Wolf 1984) and it is therefore possible for recycling to
50
occur leading to erroneous estimates for Ra and Rd (Bell et al 1986). For example glucose
enters the glycolytic pathway and is broken down to lactate, pyruvate or alanine which can be
converted back to glucose in the Cori cycle. This has been estimated to account for up to 20%
of the hepatic glucose production (Reichard et al 1963). Thus glucose labelled with deuterium
or tritium on carbon atoms 2, 3 or 6 are now the most commonly used tracers (See Figure 2 for
the steps at which labels are lost). [6 ,6 -^H2] glucose was used in the studies below. It loses one
of its labels in the pyruvate carboxylase step and one between malate and lumarate.
c)Isotope discrimination and tracer impurities. An assumption is made that the isotope
tracer is metabolised in an identical manner to unlabelled tracer. There is evidence in vitro to
suggest that some isotopes are metabolised in different ways (Argoud et al 1987). Finegood et
al (1988) have suggested that different arterio-venous specific activity of [ tf] glucose
suggested differential metabolism but others have shown that impurities in the tracer which are
metabolised several fold more slowly may have accounted for the reported a-v differences
(Butler et al 1990, Schwenk et al 1990). Thus highly purified tracers are essential for accurate
metabolic studies. In comparison to radioactively labelled glucose the stable isotope labelled
glucose is theoretically less likely to be subject to isotope discrimination as there is less of a
mass difference between it and the native glucose (Wolfe 1984).
Calculations
Rate of appearance (Ra), and Rate of disappearance (Rd) were calculated using the model
originally proposed by Steele (1959) and adapted by Finegood et al (1987) to account for the
addition of stable labelled tracer to the exogenous glucose infusate.
51
where Ey is the enrichment of the basal continuous [6 ,6 -^H2] glucose infusion in APE, Ep(t) and
Evar are the APE of the plasma and enriched glucose infusate respectively, ly is the continuous
tracer infusion rate (mg/kg/min, p is the pool fraction (ratio of effective glucose to actual space)
which was assumed to be 0.65, V is the glucose space assumed to be 220ml/kg of the body
weight, G(t) is the plasma glucose and Ivar is the exogenous variable glucose infusion rate.
These data were subjected to a smoothing procedure. The glucose concentration and the APE
of the enriched 2 0 % glucose infusâtes and the glucose concentrations of the tracer infusate
were all measured and these values used in the calculations.
Clamp protocol
Following screening subjects were assigned the lowest available randomisation number and
were randomised to a specific treatment sequence in a randomised block design which was
unknown to the investigator until the day before each study. The study was not blinded.
Subjects were studied on three separate occasions 7 to 28 days apart. Subjects received 1]
detemir 9nmol/kg(LD), 2] detemir 18nmol/kg (HD) or 3] human NPH insulin 0.3 lU/kg
(1.8nmol/kg) (NPH) (Novo-Nordisk) in random order. Subjects with diabetes received a
higher dose of NPH insulin 0.6IU/kg (3.6nmol/kg) instead of the 0.3 lU/kg (1.8nmol/kg)
dose. Subjects were asked to have a moderate carbohydrate rich meal the night before the
clamp. They were not allowed to perform strenuous exercise in the 48 hours preceding the
study, and were not allowed to consume alcohol for 24 hours before the study. They were
52
asked if they had taken any other medication that might interfere with the study drug, or
suffered any adverse events since the previous visit.
Subjects arrived at 8 in the morning following an overnight (10 hour) fast, they were
allowed to drink water freely throughout. Those subjects with Type 1 diabetes were asked
to omit their long/intermediate acting insulin for 24 hours prior to the study day. They were
given soluble human insulin (actrapid) to take the evening before the study but were asked
not to take any insulin after midnight the night before the study. Subjects with type 2
diabetes taking insulin were managed as those with type 1 diabetes. Those on oral
hypoglycaemic agents were asked to omit their oral hypoglycaemic medication for 48 hours
prior to the study.
On arrival their blood pressure and pulse were taken supine and standing, they were weighed
and body fat percentage estimated using electrical impedance (Tanita body fat scales, Tanita
UK Ltd, Yiewsley, Middlessex). They were then asked to lie or sit on a bed and an 18-
gauge cannula was inserted into each antecubital fossa if possible (other large veins in the
forearm or hand were used if there was no suitable vein in the antecubital fossa). One
cannula was used for blood sampling and one for the administration of the various infusâtes.
Baseline blood samples were taken, then a priming dose of 170mg [6 ,6 -^H2 ] glucose and
0.14 mg/kg D5 glycerol was given, followed by a continuous infusion 1.7 mg/min [6 ,6 -^H2 ]
glucose and 0.5 mg/kg/h D5 glycerol. Subjects with diabetes were given a tapering infusion
of human soluble insulin to bring their plasma glucose into the normal range (4-7mmol/l).
The infusion of human soluble insulin was tapered off by 25% from t = -10 to -5 min and by
a further 50% from t = -5 to 0 min and stopped at t = 0 min.
An 120 minute equilibration period was allowed to reach steady state tracer enrichment
before taking samples at -30, -20, -10 and time zero. For subjects with diabetes a 130
minute equilibration period was allowed and baseline blood samples were taken at -40, -30,
-20, -10 and time zero. Immediately after the blood sample was taken at time zero, subjects
were dosed with subcutaneous injection of trial insulin, either detemir 9 nmol/kg (LD),
detemir 18 nmol/kg HD or human NPH insulin 0.3 lU/kg (1.8 nmol/kg) (or 0.6IU/kg for
53
subjects with diabetes) (NPH) according to randomisation. Blood samples were taken at 15
minute intervals for measurement of plasma glucose, this was done at the bedside using a
blood glucose analyser (YSI 2300), the infusion rate of 20% dextrose (spiked with 400 mg
[6 ,6 -^H2 ] glucose per 500 ml 20% dextrose) was altered accordingly to clamp blood
glucose. The blood glucose was clamped between 4.0 mmoll'^ and 5.5 mmoll'^ for 16 hours
after subcutaneous insulin injection. In healthy subjects blood glucose was clamped to their
presenting blood glucose on their first study day.
Blood samples for glucose enrichment, insulin, c-peptide, glucagon, IGFBP-1, IGF-1,
NEFA and glycerol enrichment analyses were taken at regular intervals during the clamp
(see table 1). Following the 16 hour period the infusions were stopped, cannulae removed
and the subject was fed a carbohydrate rich meal and allowed to sleep until the following
morning, when they were allowed home. The clamp studies were then repeated on two
further dosing days separated by at least 7 days and not more than 28 days. On these dosing
days the subjects received the other doses according to the randomisation schedule.
34 (35 for subjects with diabetes) samples each requiring 2 ml of blood were taken at regular
intervals throughout the study for insulin, C-peptide and glucagon analyses. 34 (35 for
subjects with diabetes) samples each requiring 3 ml of blood were taken for glucose and
glycerol enrichment. 21 samples each requiring 2 ml blood were taken for IGFBP-1 and
IGF-1 analyses. A further 15 samples each requiring 1 ml of blood were taken for NEFA
analysis. Samples for measurement of blood glucose were midway between samples for
other assays at 15 minute intervals. In total 41 samples requiring 1 ml blood were taken for
glucose analysis. This is a total of 268 ml blood taken during each clamp study.
All blood samples were put immediately into bottles once the blood was drawn. The bottles
were then kept on ice. Every hour all the samples on ice were transferred to a refrigerated
centrifuge and centrifuged at 3060 revolutions per minute (2000g) for 10 minutes. The
samples were then immediately removed from the centrifuge, pippetted into the appropriate
aliquots and stored in a freezer at - 2 0 ®C.
54
Blood samples for insulin, insulin detemir, C-peptide and glucagon analysis were sent in
batches to Medi-lab in Copenhagen where the analyses were made. They were packed in
dry ice and sent by overnight courier service. Confirmation was received when the samples
arrived that they were still frozen and that the packing cases were intact.
Blood samples for glucose and glycerol enrichment as well as those for NEFA assay were
transferred in dry ice to St Thomas’ hospital by private transfer. All samples remained
frozen and were immediately placed in the freezer to await analysis.
Blood samples for analysis of IGFBP-1 and IGF-1 were kept at the Royal Surrey County
Hospital where the assays took place.
Table 1.
Blood Sampling TimesTime
Insulin,C-peptide SGIucagon
Glucose and Glycerol Concentration and enrichment NEFA
IGF-1&IGFBP-1
-160 X X X X-40 X X-30 X X-20 X X-10 X X X X0 X X X X30 X X X X60 X X X X90 X X X X120 X X X X150 X X X X180 X X X210 X X240 X X X X270 X X300 X X X X330 X X360 X X X X390 X X420 X X X450 X X480 X X X510 X X540 X X X X570 X X600 X X X630 X X660 X X X690 X X720 X X X X750 X X780 X X X X840 X X X X960 X X X X
55
Preparation of tracer solutions
Tracer solutions of [6 ,6 -^H2] glucose and [^Hs] glycerol (both 99% enrichment, sterile and
pyogen free from Phenome Siences Inc, Woburn, M.A., U.S.A. were used.
Laboratory assessments
Glucose concentration and enrichment
Blood samples were taken in fluoride oxalate tubes and whole blood initially and then
plasma glucose concentrations were measured on YSI 2300 STAT plus glucose analyser
(Yellow Springs instruments. Yellow Springs, Ohio). This method is based on glucose
oxidation. 25 pi of blood or plasma is aspirated by a sipper, which then transfers and injects
the sample into a buffer filled sample chamber. Some of the substrate diffuses through the
membrane where it contacts the immobilized oxidase enzyme, it is rapidly oxidised
producing hydrogen peroxide (reaction 1 )
Reaction 1
P-D-glucose + O2 glucose oxidase Glucono-Ô-lactone + H2O2
The hydrogen peroxide (H2O2) is in turn oxidised at the platinum anode producing electrons
(reaction 2 )
Reaction 2
H2O2 2 H^ + O2 + 2 e
A dynamic equilibrium is achieved when the rate of H2O2 production and the rate at which
H2O2 leaves the immobilised enzyme layer are constant and is indicated by a steady state
response. The electron flow is linearly proportional to the steady state concentration and
56
therefore the concentration of glucose. The platinum electrode is held at anionic potential
and is capable of oxidising many substances other than H2O2 . To prevent these reducing
agents from contributing to the sensor current the membrane contains an inner layer
consisting of a very thin film of cellulose acetate. The film readily passes H2O2 but excludes
chemical compounds with molecular weights above approximately 200. The analyser was
calibrated before use using glucose standard solution 25 mmol/1, and standard normal and
pathological serum with glucose concentrations of 4.5 mmol/1 and 13 mmol/1 respectively.
[6 ,6 -^H2] Glucose analysis
Glucose isotopic enrichment was determined by gas chromatography-mass spectrometry
(GC-MS) on a VG Trio-2 5890 (Hewlet-Packard, Woking, UK) using selected ion
monitoring of a glucose acetate boronate derivative (Bier et al, 1977). The samples were
first deproteinised by adding 50 pi of plasma sample to 0.5ml of ethyl alcohol and
centrifuging at 1000 rpm for 5 min. The supernatant was then removed and dried under a
continuous stream of N 2 at 50 °C. The glucose acetate boronate derivative was produced by
adding butylboronic acid in pyridine (10 mg/ml), heating this solution at 95°C for 30 min
and then cooling at room temperature. Acetic anhydride (100 pi) was then added and the
mixture left for 1 hour again at room temperature. This solution was dried under N2 and
dissolved in 500 pi of decane before being introduced into the GC-MS. The ions monitored
were of mass 297 and 299 representing [M-C4H9] and the corresponding fragment enriched
with two deuterium atoms respectively. The within assay coefficient of variation of
measurement of isotopic enrichment was < 2 %.
Glycerol concentration and enrichment
0.5 ml plasma was pipetted into plastic tubes and 1ml 3.5%SSA was added. Samples were
mixed and then allowed to stand for 15 to 30 minutes. They were then centrifuged at 2000
rpm for 20 minutes at 4®C. Meanwhile the ion exchange columns were prepared; 0.5 ml of
57
cation resin was added first to the column, then rinsed with 1 ml of distilled water, then 0.5
ml of anion resin was added and rinsed with 1 ml of distilled water. The supernatant from
the centrifuged tubes was then added to the columns and the elute collected in labelled vials.
Each column was rinsed with 3 ml distilled water. The vials were covered with parafilm and
frozen at -20®C. The samples were dried under a stream of dried compressed air at room
temperature until approximately 200pl of liquid remained (around 15 hours). At that point,
the sample temperature was increased to 50°C while continuing the air stream until the
sample completely dried (around 2 hours). lOOpl of pyridine and lOOpl of MTBSTFA were
added to the dried residue, and the mixture was vortexed. The solution was allowed to
remain at room temperature for 30 minutes allowing derivatisation to reach completion. The
derivatised sample was transferred into a plastic micro-injection autosampler vial and was
sealed.
After a 2- pi splitless injection of the derivatised sample onto the column, the oven was held
at the initial temperature of 100°C for 0.5 minutes, increased to 300°C at 25 °C /min and
maintained at 300 °C for 1.5 min. Thus the total run time was equal to 10 minutes with
glycerol and 5D glycerol eluting at 6 minutes. Using SIM, the major fragments monitored
for the tBDMS derivatives of glycerol and d5-glycerol were the [M-57] ion fragments,
377m/z and 382m/z respectively. With d5-glycerol used as the tracer, the enrichment
quantitated in plasma as the ratio of d5-glycerol:glycerol (ion abundance o f 377/382 m/z).
NEFA
NEFA was measured using the Wako NEFA C assay on the Cobas Fara II (Roche,
Welwyn Garden City, UK.) the method of which relies upon the acylation of coenzyme
A(CoA) by the fatty acids in the presence of added acyl-CoA synthetase (ACS). The acyl-
CoA produced is oxidized by added acyl-CoA oxidase (ACOD) with the generation of
hydrogen peroxide. Hydrogen peroxide, in the presence of peroxidase (POD) permits the
oxidative condensation of 3-methyl-N-ethyl-N-(b-hydroxyethyl)-aniline (MEHA) with 4-
aminoantipyrine to form a purple color which can be measured spectrophotometrically at '
550nm.
58
The within-run precision for this assay is 2.7%CV, 1.1 %CV and 1.1%CV for mean
concentrations of 0.33mmol/L, 0.62mmol/L and 0.99mmol/L, respectively. The linear
range of the assay is 0-2.0 mmol/L.
Insulin & insulin detemir, C-Peptide and Glucagon
The following assays were done at Medilab by the request of Novo Nordisk. Insulin detemir
was measured using the ELISA specific insulin detemir assay (Heineman et al 1999).
Human insulin was measured using the DAKO insulin ELISA assay (Inter-assay CV 4.2-
9.3%). C-peptide and glucagon were measured using DAKO ELISA assays (Inter-assay CV
3.3-5.7% for C-Peptide and 6 % for glucagon).
IGF-1
lGF-1 was measured using Nichols institute Diagnostics chemiluminescence immunoassay.
This assay is a two site chemiluminescence assay for the measurement of lGF-1 in human
serum. Acridium esters emit light upon treatment with hydrogen peroxide and an alkaline
solution. The Trigger 1 solution contains hydrogen peroxide in dilute acid and Trigger 2
solution contains dilute sodium hydroxide. The system automatically injects Trigger
solutions 1 and 2 into the wells which oxidse the acridium ester. The oxidised product is in
an excited state. The subsequent return to the ground state results in emission of light which
is quantified in 2 seconds and is expressed in relative light units (RLU) by the intergrated
system luminometer. The antibody to the C-terminal with the amino acid sequences of 62-
70 is biotinylated for capture and the antibody to the amino acid sequences of 1-23 and 42-
61 is labelled with acridium ester for detection.
The patient sample is acidified to separate lGF-1 from IGFBPs, then excess lGF-11 is added
in the assay to block IGFBP binding sites from recombining with the released lGF-1. The
acidified patient sample is incubated simultaneously with the biotinylated capture antibody,
excess lGF-11, and the acridium ester labelled tag antibody. During the first incubation.
59
IGF-1 in the sample forms a sandwich complex with the capture antibody and the acridium
labelled antibody. After the initial incubation period streptavidin coated magnetic particles
are added to the reaction mixture and a second incubation follows. The streptavidin coated
magnetic particles allow for a highly specific and efficient means of binding the sandwich
complex to the solid phase via the high affinity interaction between biotin and streptavidin.
Free labelled antibody is separated from the labelled antibody bound to the magnetic
particles by aspiration of the reaction mixture and subsequent washing. The wells
containing the washed magnetic particles are transported to the system luminometer, which
automatically injects Trigger 1 and Trigger 2, initiating the chemiluminescence reaction.
The light is quantitated by the luminometer and expressed as RLU. The amount of bound
labelled antibody is directly proportional to the concentration of IGF-1 in the sample.
The assay runs on the Nichols advantage speciality system. The sample is placed in the
designated sample racks and these racks placed in the sample loading area. The system
performs an assay specific on board dilution by first pipetting 13 pi of the sample and 400
pi of the acidification buffer into one well of the cuvette. 50 pi of the acidified sample
mixture is then pipetted and added to a mixture of 1 0 0 pi of of biotinylated goat polyclonal
antibody, 50 pi of the IGF-II solution and 50 pi of the acridium labelled goat polyclonal
antibody in the next well of the cuvette. The cuvette strip is transported to the incubation
chamber and incubated for 20 minutes at 37®C. After this initial incubation, 25 pi of
streptevidin coated magnetic particles are added to each well. The incubation is prolonged
for a further 10 minutes at 37®C. After the entire incubation the wells are washed and
measured in the measuring chamber. The coefficient of variation for this assay was 4.8%.
IGFBP-1
IGFBP-1 was measured using Oy Medix Biochemica Ab IGFBP-1 lEMA test. The
principle of this test is as follows. A monoclonal antibody specific to IGFBP-1 is
immobilized on microwell plates, and another monoclonal antibody, also specific to IGFBP-
1, is conjugated with horse-radish peroxidase (HRP). IGFBP-1 from the sample is bound to
the plates. After a washing step, HRP is added. After a second washing step, enzyme
60
substrate is added. The enzymatic reaction is proportional to the amount of IGFBP-1 in the
sample. The reaction is terminated by adding stopping solution. Absorbance is measured
on a plate reader.
80 pi of assay buffer was pipetted into each well. 2 0 pi of the serum sample was then
pipetted into the well. The plate was covered and incubated for 30 minutes at room
temperature on a plate shaker. The wells were then aspirated and washed 3 times with 300
pi of washing solution. 1 0 0 pi of enzyme conjugate was then pipetted into the wells and the
plate covered and incubated for a further 30 minutes at room temperature on the plate
shaker. The wells were then washed again as above and 100 pi of HRP solution was added
to each of the wells. The plate was covered again and incubated for 10 minutes at room
temperature on the plate shaker. The reaction was stopped by adding 50 pi of stopping
solution to each of the wells. The plate was shaken gently to mix the solutions. The
absorbance at 414 nm was then measured using a plate reader. The inter-assay CV was
6.4%.
Trial design
Type of trial
We performed a single centre randomised open label three period crossover trial with
subcutaneous injections of insulin detemir and NPH insulin using a euglycaemic clamp with
labelled glucose in healthy subjects and in subjects with Type 1 and Type 2 diabetes
mellitus. Insulin detemir was given in doses that were expected to cause the same blood
glucose lowering effect as the dose of NPH administered. An additional dose o f insulin
detemir was also studied. In healthy subjects this additional detemir dose was higher than
the comparative doses, but in the diabetic subjects the additional detemir dose was lower.
61
Rationale for trial design
The trial was open labelled. The insulins have different appearances and it was therefore not
possible to blind the trial. The crossover design was chosen to minimise the number of
subjects needed as each subject then becomes their own control and intra-subject variation is
smaller than inter-subject variation.
A clamp procedure with no pre-dose insulin infusion was used for healthy subjects. Type 1
and Type 2 diabetic subjects were expected to enter the trial sessions with high blood
glucose levels and thus the main issue was to lower their fasting blood glucose to predefined
clamp level. In order to obtain a clamp level as normal as possible, an insulin infusion was
started at the same time as the glucose and glycerol infusion. This insulin infusion was
tapered off before administration of the trial drug. Intravenous administered insulin is
rapidly eliminated and the effect on the endpoints of the trial was considered minimal. The
effect has been evaluated. Each subject was clamped at the same level at each visit.
Free fatty acid lipogenesis generally takes place immediately following an increase in
insulin concentration and stops when insulin levels are low. Measurements of NEFA were
made during the first hours after trial drug administration, at the maximum concentration
and during the last hours of the study. The concentrations of IGF-1 and IGFBP-1 were
measured hourly.
Rationale for dosage regimen
The effect of two dose levels of insulin detemir was investigated in order to allow for a dose
response relationship to be generated. One dose level of NPH insulin was administered for
comparison, and it was hoped that this dose would be equipotent to one of the detemir doses.
Pharmacodynamic trials with insulin detemir in healthy subjects indicated a difference in
blood glucose lowering potency of approximately a factor of five between insulin detemir
and NPH insulin. Thus the dose of insulin detemir expected to give the same blood glucose
lowering effect as NPH 0.3 lU/kg (1.8 nmol/kg) was 9 nmol/kg. The second dose of insulin
62
detemir was 18 nmol/kg and expected to give a similar blood glucose lowering effect as
NPH 0.6 lU/kg (3.6 nmol/kg).
Trial population
Number of subjects in each group
Sample size calculation was based on the peripheral glucose uptake (PGU) in a two-period
two-treatment sub-design of the three periods. Intrasubject coefficient of variation (CV) was
assumed to be maximum 25%. Considering a relative difference of 50%, and using a
significance level of 5%, the power would be 85% if eight subjects completed the trial and
70% if six subjects completed the trial.) Based on this sample size calculation eight subjects
were included in the trial. If more than two subjects were withdrawn, replacement subjects
were to be enrolled in order to obtain at least six completing subjects in each group (healthy
subjects and subjects with Type 1 and Type 2 diabetes). Therefore 18 subjects completing 3
studies.
Subjects were recruited based on the following inclusion and exclusion criteria:
Inclusion criteria
Healthy subjects
Signed informed consent obtained before any trial related activity.
Healthy male and female subjects between 19 and 60 years inclusive.
Considered generally healthy upon completion of medical history and physical examination.
Body mass index <30kg/m^
Fasting blood glucose < 6 mmol/1
Subjects with type 1 diabetes as above and/or
Age> 18 years
63
Type 1 diabetes according to ADA criteria (see introduction)
Duration of diabetes >12 months
HbAic < 12 %
C-peptide negative
Subjects with type 2 diabetes
Type 2 diabetes according to ADA criteria (see introduction)
Duration of diabetes >12 months
Age >35 years
BMI < 35 kg/m2
H bA ic<12%
Exclusion criteria
Healthy subjects
Participation in any other clinical trial involving other investigational products within the
last 3 months
Clinically significant abnormal haematology or biochemistry screening tests.
Any serious systemic infectious disease that occurred during the four weeks prior to the first
dose of the trial drug.
Any intercurrent illness that may effect blood glucose
Subject with a history of drug or alcohol dependence.
Use of prescription drugs except oral contraceptives within 3 weeks prior to the first dose of
the trial drug.
Use of non prescription drugs, except ordinary vitamin preparations within 3 weeks of the
first dose of the trial drug. Occasional use of analgesics was allowed up to 48 hours prior to
entrance in the clinical research centre.
Mental incapacity, unwillingness or language barriers precluding adequate understanding or
cooperation.
Blood donation of more than 500ml within the last 12 weeks.
64
Subjects with a first degree relative with diabetes mellitus.
History or presence of diabetes, cancer or any clinically significant cardiac, respiratory,
metabolic, renal, hepatic, gastrointestinal, endocrinological, dermatological, venereal
haematological, neurological or psychiatric disorder.
Previous anaphylactic reaction
History of significant drug allergy or hypersensitivity.
Females of child bearing potential not using an acceptable method of contraception which
include an intrauterine device (lUD) which has been in place for at least 3 months, or
sterilisation, or the oral contraceptive pill which should have been taken without difficulty
for at least 3 months or an approved hormonal implant or barrier methods.
Subjects with type 1 diabetes above and/or
Laser therapy for retinopathy within the last two months
Current treatment with systemic corticosteroids
Impaired hepatic fiinction measured as alanine aminotransferase (ALAT) > two times the
upper reference limit
Impaired renal function measure as creatinine >150 pM (1.7 mg/dL)
Cardiac problems defined as: 1) decompensated heart failure (New York Heart Association
class III and IV) and/or 2) angina pectoris and/or 3) myocardial infarction at any time
Uncontrolled treated/untreated hypertension as judged by the investigator
Subjects with type 2 diabetes as above and
Total insulin dosage >120 lU/day
Safety visit
Before screening took place the subjects were provided with written and oral information
about the study and the procedures involved. They were asked to sign an informed consent
form. All subjects were then asked a series of questions to determine their eligibility for the
trial with respect to the inclusion and exclusion criteria. The subject’s demographics (date
of birth, sex, ethnic origin and smoking habits) were recorded. A full medical history was
65
taken and the subject examined. The subject’s height and weight were recorded and their
body mass index calculated. Blood pressure and pulse were also measured and an
electrocardiograph (ECG) recorded. Blood was then taken for haematology and clinical
chemistry, and a urine sample sent for urinalysis and pregnancy test if relevant.
If the subject fulfilled all of the inclusion and exclusion criteria they were accepted into the
trial and an appointment was made for visit two within 5-21 days. Screen failures were
logged.
Subjects passing screening went on to perform the study days, which were held 7 to 21 days
apart from each other and were as described above.
Statistical analyses
Glucose Ra and Rd were calculated using a modified form of the Steele equation (Steele,
1959) adapted by Finegood et al, (1987) to account for the addition of stable labelled tracer
to the exogenous glucose infusate. The volume of distribution was 220 ml/kg and ratio of
effective glucose to actual space was 0.65. Glycerol Ra was calculated using the Steele
equation with a volume of distribution of 230ml/kg. Glucose Ra and Rd expressed as a
percentage of GIR at each time point was averaged over 960 minutes to determine the
proportion of the glucose lowering effect of each insulin due to a decrease in Ra and the
proportion due to an increase in Rd. All values are expressed as mean ± SEM. Statistical
analysis of the results was either by a 2 way ANOVA with subjects and treatment as factors
using baseline as a covariate or a paired students t-test.
66
Chapter 3
Results
Healthy Subjects
Subject demographics
10 healthy subjects 5 female and 5 male were screened. 4 female subjects and 4 male
subjects were randomised for the study. One male subject failed screening because he was
taking regular inhaled corticosteroids for asthma. One female subject passed screening, but
withdrew from the study before the first clamp. One female subject completed one clamp
study and then withdrew from further studies after finding the clamp procedure too arduous,
her results have not been included in the analysis as the data were not paired. 3 female
subjects and 4 male subjects completed all three clamps, they were aged 29.9 ± 3.4 years
with BMI 22.9 ± 0.72 kg/m^, they weighed 68.01 ± 2.88 kg and a body fat percentage of
24.71 ± 4.14 %. One of the subjects completing all three study days was not fasting for the
high dose study day. The results for this study have therefore not been included in our
analysis. The results presented therefore are for six subjects for high dose, and seven for
low dose and NPH.
Plasma Glucose
Euglycaemia was maintained throughout all clamp studies, there was no difference between
the plasma glucose during the three studies. Mean plasma glucose was clamped at
4.92+0.11, 4.95+0.08 and 5.01±0.14mmol/l in the LD, NPH and HD studies respectively.
The mean and standard differences of the individual coefficient of variation for glucose
67
during the clamp studies was 4.95+1.82 % for LD, 6.32±2.34 % for NPH and 7.82±3.76 %
for HD.
Figure 10. Graph of healthy subjects mean plasma glucose.
Healthy Subjects Plasma Glucose
I85
-180 -60 60 180 300 420 540 660 780 900 1020
Mean LD ■ Mean NPH Mean HD
Time mins
Glucose Infusion Rate
The total amount of glucose infused to maintain euglycaemia was similar for LD (2019 ±
374 mg/kg) and NPH (2102 ± 384 mg/kg)) (p=NS) demonstrating their equipotency (Fig
11). There was no difference between LD and NPH in the time to starting the glucose
infusion, peak glucose infusion rate or glucose infusion rate at 960 minutes. Glucose was
infused earlier following HD (40+7.5 min) than LD (61 ±9.6 min, p<0.03). The peak
glucose infusion rate following HD was higher (5.7±0.82 mg/kg/min) than LD (3.68+0.72
mg/kg/min, p<0.03). Total glucose infused was significantly higher following HD
(3267±388mg/kg) than LD (p<0.02). (Table 2)
68
Table 2. Healthy Subjects Plasma Glucose and Glucose Infusion Rate.
LD NPH HD
Plasma Glucose
mmol/14.92 ±0.11 4.95 1 0.08 5.0110.14
Total Amount of Glucose Infused,
mg/kg20191374* 21021384 32671388
Time to Start of Glucose Infusion,
minutes61 ±9.45* 6417.18 4017.35
Peak Glucose Infusion Rate,
mg/kg/min3.68 ± 0.72 * 4.0410.73 5.710.82
Glucose Infusion Rate At 960
minutes, mg/kg/min1.6410.23 * 1.4210.15 2.92 1 0.44
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 11. Graph of healthy subjects mean glucose infusion rates
Healthy Subjects Mean Glucose infusion Rates (GIR)
6
5
4
2
1
00 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Tim e m ins
■ Mean GIR LD■ Mean GIR NPH Mean GIR HD
69
Glucose Metabolism
The minimum glucose Ra was significantly lower (p<0.02) with LD (-0.10±0.15
mg/kg/min) than NPH (0.17±0.10 mg/kg/min) (Fig 12). Using ANOVA, glucose Ra with
LD was significantly lower than with NPH (mean difference 0.24mg.kg'\min'^)(CI 0.09,
0.39)(p<0.01). This difference was still significant (p<0.05) when corrected for baseline
(mean difference 0.25 mg.kg'\min"^ Cl 0.05, 0.44) using analysis of covariance.
Figure 12. Graph of healthy subjects mean glucose rate of appearance.
Healthy Subjects Mean Glucose Rate of Appearance (Ra)
2.5
MEAN LD Ra MEAN NPH Ra MEAN EŒ) Ra
S - h
S ^
> M 1 1 1 ^ ,3èO 420^4801540 600 660 720 780 840 900 960
Time mins
The mean of the maximum change from baseline of glucose Rd (for each individual) was
lower with LD (1.52±0.49 mg/kg/min) than NPH (1.97+0.52 mg/kg/min) but this did not
reach statistical significance.
70
Figure 13. Graph of healthy subjects mean glucose rate of disappearance.
Healthy Subjects Mean Glucose Rate of Disappearance (Rd)
,S
-60 0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
■MEANNPH Rd MEAN HD Rd MEANLDRd
Time mins
Figure 14. Graph of healthy subjects mean glucose Ra and Rd.
Healthy subjects Mean Glucose Ra & Rd
Glucose Ra and Rd mg/kg/m in
Time mins
-M E A N L D R d -M E A N N PH Rd
M EA N HD Rd- M EA N LD Ra- M EA N N PH Rii
M EA N HD Ra
Of the mean glucose lowering effect of insulin detemir over the 960 minutes (i.e. the
proportion of glucose infused needed to overcome each of Ra and Rd) 71.3±4% was due to
the decrease in Ra and 28.7±4% was due to the increase in glucose Rd, whereas for NPH
this was 63.3+3.9% from the decrease in Ra and 36.7+3.9% from the increase in Rd
(p<0 .0 1 ).
71
Figure 14. Bar charts showing the relative proportion of glucose lowering effect of each
insulin from effect on glucose Ra and glucose Rd.
Bar Chart to demonstrate the effects of 9nmol/kg detemir on glucose Ra&Rd
B Mean LD Rd
□ Mean LD Ra
3.5
c
0.5
-0.5
Time mins
Bar Chart to demonstrate the effects of 0.3IU/kg NPH on glucose Ra &Rd
4 - ___
r f l3.5 3
1.51 —
0.5 — 0 %
-0.5
_L
□ Mean NPH Rd□ Mean NPH Ra
_o o o c > > o o c > . o o o o o o o
Time mins
Iw
72
Figure 15. Pie chart showing the overall mean proportion of effect of each insulin on
glucose Ra and Rd.
Pie chart showing the mean proportion o f effect on glucose Ra and Rd following LD
29%
□ Ra ■ Rd
71%
Pie chart showing mean proportion o f effect on glucose Ra and Rd following NPH.
37%□ Ra ■ Rd
63%
The maximum Rd for HD was 5.21±0.71 mg/kg/min which was significantly greater than
with LD (p<0.03). HD appeared to suppress glycerol Ra more than LD but this difference
did not reach statistical significance. The maximum change from baseline was -1.31+0.20
pmol/kg/min with HD.
73
Table 3. Healthy Subjects Glucose Metabolism
LD NPH HD
Baseline Ra mg/kg/min 2.01 ±0.12 2.17 ±0.09 1.68 ±0.15
Minimum Ra mg/kg/min -0.10 ±0.15** 0.17±0.10 -0.69 ±0.16
Maximum change from
baseline R a mg/kg/min
-2.10 ±0.24 -2.00 ±0.18 -2.37 ± 0.26
AOC Ra mg/kg 1401.48 ±147.99 1328.69 ±135.80 1515.81 ±213.29
Baseline Rd mg/kg/min 2.06 ±0.12 2.18 ±0.08 1.75 ±0.18
Maximum Rd
mg/kg/min
3.58 ±0.54* 4.15 ±0.58 5.25 ±0.74
Maximum change from
baseline Rd mg/kg/min
1.52 ±0.49 1.97 ±0.52 3.51 ±0.75
AUC Rd mg/kg 564.42 ± 252.93 760.46 ± 247.27 1767.85 ±431.14
* Indicates a statistica ly significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
74
Figure 16. Graph of Healthy Subjects Glucose Tracer:Tracee ratio during the clamp studies
Healthy Subjects mean Plasma Glucose TraceriTracee ratio
5I
Mean LD ■ Mean NPH Mean HD
■180 -60 60 180 300 420 540 660 780 900 1020
Time mins
Glycerol metabolism
Mean glycerol Ra at baseline was similar with the three study doses at 1.75 ± 0.15
pmol/kg/min for LD, 1.91 ±0.33 pmol/kg/min for NPH and 1.94 ± 0.41 pmol/kg/min for
HD (p=NS). The point of maximum suppression following dosing was 0.93 ± 0.18
pmol/kg/min for LD, 0.81 ±0.18 pmol/kg/min for NPH and 0.63 ± 0.07 pmol/kg/min for
HD (p=NS). The maximum change from baseline for glycerol Ra was -0.82 ± 0.17
pmol/kg/min with LD, -1.10 ± 0.24 pmol/kg/min with NPH and 1.31 ± 0.20 pmol/kg/min
with HD. The area over the curve (AOC) effect on glycerol Ra was less for low dose than
NPH or high dose detemir, although not statistically significant. Following dosing with LD
AOC was -366.1 ± 145.52, following NPH it was -691.0 ± 215.44 and following HD it was
-945.6 ± 172.28.
75
Table 4. Healthy Subjects Glycerol Metabolism.
LD NPH HD
Baseline Ra pmol/kg/min 1.75 ±0.15 1.91 ±0.33 1.94 ±0.41
Minimum Ra pmol/kg/min 0.93 ±0.18 0.81 ±0.18 0.63 ±0.07
Maximum change from
baseline Ra pmol/kg/min
-0.82 ±0.17 -I.10±0.24 1.31 ±0.20
AOC Ra pmol/kg -366.1 ±145.52 -691.0 ±215.44 -945.6 ± 172.28
* Indicates a statistically significant difference between LD and I ID.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 17. Graph of healthy subjects mean rate of appearance of glycerol.
Healthy Subjects Mean Rate of apparence o f glycerol (Ra)
tI
0:&
-30 30 90 150 210 270 330 390 450 510 570 630 690 750 810 870 930
Mean HD ■ Mean LD - Mean NPH
76
NEFA
All three insulin doses suppressed NEFA concentrations. Plamsa NEFA concentrations at
baseline were 0.65 ± 0.08 mmol/1 for LD, 0.52 ±0.11 mmol/1 for NPH and 0.52 ± 0.12
mmol/1 for HD (p=NS). There was significantly more suppression of NEFA concentration
with NPH than with LD (mean difference O.lOmmol/1 Cl 0.03, 0.17)(ANOVA, p<0.02).
However there was no significant difference in the minimum NEFA concentrations
following LD (0.12 ± 0.04 pmol) or NPH (0.09 ± 0.03 pmol). Plasma NEFA concentration
following HD was significantly lower than with LD (ANOVA, p<0.01).
Table 5. Healthy Subjects NEFA results.
LD NPH HD
Baseline NEFA pmol/l 0.65 ± 0.08 0.52 ±0.11 0.52 ±0.12
Maximum change from
baseline NEFA pmol/1
0.54 ± 0.08 0.43 ±0.11 0.48 ±0.11
AOC NEFA pmol/1 -305.83 ± 59.8 -252.79 ± 89.9 -340.32 ± 90.73
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***lndicates a statistically significant difference between NPH and HD.
77
Figure 19. Graph of healthy subjects mean plasma NEFA concentrations.
Healthy Subjects Mean Plasma Non Esterified Fatty Acid Concentrations
i 8:e
-------------- ^
II
-180 -60 300 42060 180 540 660 780 900 1020
Mean LD • Mean NPH Mean HD
Time mins
IGF-1
IGF-1 levels did not alter during the studies. Mean IGF-1 levels at baseline were 24.15 ±
3.43 nmol/1 for LD, 24.97 ± 2.68 nmol/1 for NPH and 26.5 ± 2.98 nmol/1 for HD. Mean
IGF-1 level following injection with LD was 23.01 ± 2.46 nmol/1, 25.02 ± 2.53 nmol/1 for
HD and following NPH injection it was 24.13 ± 2.46 nmol/1, there was no difference
between these three groups.
Table 6. Healthy Subjects IGF-1 results.
LD NPH HD
Baseline IGF-1 concentration nmol/1 24.15+3.43 26J±:L98 24.97 ±2.68
Mean IGF-1 concentration nmol/1 23.01 ±2.46 24.13 ±2.46 25^ 2± 2^ 3
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 20. Graph of healthy subjects mean plasma IGF-1 concentrations.
78
Healthy Subjects Mean Plasma Insulin Like Growth Factor Concentrations (IGF-1)
35 1
i
-180 -60 60 180 300 420 540 660 780 900 1020
■ MEAN LD •MEAN NPH MEAN HD
Time mins
IGFBP-1
IGFBP-1 levels were suppressed by insulin. Mean IGFBP-1 levels at baseline were 7.25 ±
3.7 pg/l for LD, 3.08 ± 0.41 pg/1 for HD and 5.96 ± 3.48 pg/l for NPH study day. There
was no significant difference between these baseline values. They fell to a mean minimum
value of 1.76 ± 0.87 pg/1 following LD, 0.9 ± 0.65 pg/1 following HD and 1.57 ± 0.60 pg/1
following NPH insulin. The mean minimum was significantly lower following high dose
detemir than following low dose detemir (p<0.03), however it was not significantly different
to the mean minimum following NPH. There was no difference between the mean
minimums following NPH and low dose detemir. Using analysis of variance there was no
difference between the three insulins on effect on IGFBP-1 levels.
79
Table 7. Healthy Subjects IGFBP-1 results.
LD NPH HD
Baseline IGFBP-1 concentration pg/l 7.25 + 3.7 5.96 ±3.48 3.08 ±0.41
Minimum IGFBP-1 Plasma
concentration pg/1
1.76 ±0.87* 1.57 ±0.60 0.9 ±0.65
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 21. Graph of healthy subjects mean plasma IGFBP-1 concentrations.
Healthy subjects mean IGFBP-1 concentratrion
mo
-60-180 60 180 300 420 540 660 780 900 1020
■MEANLD ■MEAN NPH MEAN HD
Time mins
Insulin and insulin detemir
Mean AUC (0-960mins) was 1350.8±125.8, 2220.4+417.5 and 81.4±11.7 nmol.F^ x min for
LD, HD and NPH respectively. There was no individual difference in the time to reach
maximum concentration (Tmax) for the three insulins being 617+85 mins, 520±79 mins and
420±60 mins for LD, HD and NPH respectively. The maximum concentration (C max) was
greater for HD (4478±190 pmolf^) than LD (2345+222 pmolF*) (p<0.01). C max for NPH
was 122±14 pmolf \ A doubling in insulin detemir dose resulted in an approximately two
fold increase in AUC and Cmax
80
Figure 22. Graph of healthy subjects mean plasma insulin and insulin detemir
concentrations.
Healthy Subjects Plasma Insulin and Insulin Detemir Concentrations
4500
4000
3500
3000
I 2500cxI 2000I
1500
1000
500
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Mean LD Mean HD
■ Mean NPH
Time mins
C-Peptide
C-peptide at baseline was similar in each insulin study being 0.47±0.04 nmol/1, 0.55±0.08
nmol/1 and 0.31 ±0.13 nmol/1 for LD. NPH and HD respectively. C-peptide suppressed
following each insulin and there was no difference in this suppression between the insulins
with minimum c-peptide 0.25+0.05 nmol/1, 0.27+0.05 nmol/1 and 0.15+0.06 nmol/1
following LD, NPH and HD respectively.
81
Figure 23. Graph of healthy subjects mean plasma c-peptide concentrations.
Healthy Subjects Mean Plasma C-Peptide concentrations
0. ?
0.6
o 0.5-
-o 0.4
Iu 0.3
0.3
-40 360160 560 760 960
■ Mean LD • Mean NPH Mean HD
Time mins
Glucagon
The glucagon plasma concentrations did not change after dosing with insulin detemir and
NPH insulin, and there was no difference between insulin detemir and NPH insulin.
82
Type 1 Subjects
Subject demographics
5 male subjects and one female subject passed screening. The female subject suffered a
serious adverse event (cellulitis from infected cannulation site) following her second clamp
and did not complete her third clamp (the low dose study). Mean age of subjects 29.5 ± 2.39
years, they had had diabetes for 15.3 ± 3.86 years, they had reasonable diabetes control with
HbAlc of 8.45 ± 0.43 %. Their weight was 75.5 ± 1.83 kg and body fat percentage 21.5 ±
349%^
Plasma Glucose
Subjects arrived for the study day with high blood glucose concentrations having omitted
their long acting insulin for the previous 24 hours. However there was no difference in
plasma glucose levels on arrival for the study days. Plasma glucose levels were brought into
the clamp range using a tapering dose of soluble insulin. There was no difference in the
soluble insulin infusion rates on the separate dosing days. There was no difference in
plasma glucose concentrations following the subcutaneous administration of insulin (ie
during the clamp period) being 6.56 ± 0.21 mmol/1 following HD, 6.64 ± 0.2 mmol/1
following NPH and 6.55 ±0.19 mmol/1 following LD. The mean and standard differences of
the individual coefficient of variation for glucose during the clamp studies was 8.77±2.55 %
forLD, 13.35±5.31 % for NPH and 9.14±1.65 % for HD.
83
Figure 24, Graph of Type 1 Subjects Mean Plasma Glucose.
Type 1 subjects Plasma Glucose
60-180 -60 180 300 420 540 660 780 900 1020
Mean HD n=6 • Mean NPH n=6 ■ Mean LDn=5
Time mins
Glucose infusion rate
There was no difference in the time to starting glucose infusion between the 3 insulins with
time to start at 89 ± 26.5 mins for HD, 95.3 ± 24.0 for NPH and 136.4 ± 34.5 for LD. The
maximum glucose infusion rate was similar for the two equipotent doses at 2.9 ± 0.54
mg/kg/min for HD and 4.64 ± 0.79 mg/kg/min for NPH (p=NS), but lower following LD at
2.24 ± 0.47 mg/kg/min than NPH (p<0.04). There was no difference in the maximum GIR
following HD and LD. There was no difference in the time to reach maximum GIR for the
three insulins, nor was there a difference between the areas under the curve of GIR . AUC
for HD was 1537.04 ± 279.2 mg/kg, for NPH it was 2015.14 ± 407.7 mg/kg and for LD it
was 1019.4 ± 266.9 mg/kg. The GIR at the end of the studies was significantly lower
following LD administration at 0.11 ± 0.11 mg/kg/min than following HD at 1.45 ± 0.51
mg/kg/min (p<0.05), but not different to NPH at 0.91 ± 0.50 mg/kg/min.
84
Table 8. Type 1 Subjects Plasma Glucose and Glucose Infusion Rate.
LD NPH HD
Plasma Glucose mmoI/1 6.55 ±0.19 6.64 ± 0.2 6.56 ±0.21
Time to start of glucose
infusion mins
136.4 ±34.5 95.3 ±24.0 89 ± 26.5
Maximum glucose infusion
rate mg/kg/min
2.24 ±0.47 ** 4.64 ± 0.79 2.9 ±0.54
Area under the curve mg/kg 1019.4 ±266.9 2015.14 ±407.7 1537.04 ±279.2
Glucose infusion rate at 960
mins
0.11 ±0.11 * 0.91 ±0.50 1.45 ±0.51
* Indicates a statistically significant difference between LD and LID.
** Indicates a statistically significant difference between LD and NPH.
* Revindicates a statistically significant difference between NPH and HD.
Figure 25. Graph of Type 1 Subjects Mean Glucose Infusion Rates.
Type 1 subjects Glucose Infusion Rates (GIR)
aE 2.5
I0.5
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Mean HD n=6 • Mean NPH n=6 Mean LD n=5
Time mins
85
Glucose Metabolism
There was no differenee in hepatic glucose output (Ra) at baseline for the 3 insulins being
2.14±0.40 mg/kg/min for LD, 1.94±0.19 mg/kg/min for HD and 2.14± 0.18 mg/kg/min for
NPH. These insulins had similar effects suppressing glucose Ra with a mean change from
baseline o f -2.19+0.45 mg/kg/min for LD, -2.20 ± 0.30 mg/kg/min for HD and -2.53 ± 0.36
mg/kg/min for NPH (p=NS). Areas over the curve for suppression of Ra was also similar
between the 3 insulins being -1452.33± 401.61 mg/kg for LD, -1452.71 ± 268.83 mg/kg for
HD and-1461.18 ± 288.18 mg/kg for NPH (p=NS).
Figure 26. Graph of Type 1 Subjects Mean Glucose Rate of appearance.
Type 1 Subjects Glucose Rate of Appearance (Ra)
SIE
I-60 60 540 660 780 900 1020180 300 420
- Mean LD Ra (n=5) HD Ra (n=6)
- NPH Ra (n=6)
Time mins
There was no difference in peripheral glucose uptake (Rd) at baseline for the 3 insulins
being 1.93± 0.42 mg/kg/min for LD, 2.02 ± 0 .17 mg/kg/min for HD and 1.99 ± 0.20
mg/kg/min for NPH (p=NS).
86
Figure 27. Graph of Type 1 Subjects Mean Glucose Rate of Disappearance.
Ia:
aÜ
-60
Type 1 subjects Glucose Rate of Disappearance (Rd)
54t5 ■■
4T T T A T T T T
3.5
T Jr j T T T T T + M T
1.51
0.5rO-
- Mean LD Rd (n=5) HD Rd (n=6)
-NPH Rd(n=6)
60 180 300 420 540
Time mins
660 780 900 1020
There was significantly greater effect on change of Rd from baseline following NPH than
either HD (p<0.03) or LD (p<0.01), with the mean maximum change from baseline being
0.29 ± 0.46 mg/kg/min for LD, 0.83 ± 0.31 for HD and 2.22 ± 0.41 mg/kg/min for NPH.
There was no difference between HD and LD for change from baseline. There was also
significantly greater areas under the curve of Rd following NPH than HD (p<0.01) being
-234.32 ± 350 mg/kg for LD, 55.74 ± 197.5 mg/kg for HD and 678 ± 138.14 mg/kg for
NPH. There was no difference between the AUC for HD and LD nor LD and NPH (there
was great inter-subject variability in the AUC for LD).
87
Figure 28. Graph of Type 1 Subjects Mean Glucose Ra and Rd.
Type 1 subjects Mean Glucose Ra & Rd
5.5
60 180 300 420 540 660 780 900 1020
- Mean LD Ra (n=5) HD Ra (n=6)
- NPH Ra (n=6)■ Mean LD Rd (n=5) HDRd(n=6)
•NPHRd(n=6)
Time mins
Blood glucose was controlled before the injection of subcutaneous insulin by using
intravenous soluble insulin. The infusion was stopped just before the subcutaneous insulin
was injected. This technique led to poor maintenance of euglycaemia in the first 60 minutes
of the clamp. We have therefore calculated areas under/over the curve for Ra and Rd using
a baseline of 60 minutes post injection (ie calculating change from the 60 minute value).
The area over the curve (calculated in this way) of suppression of Ra was not significantly
different for insulin detemir (1057.8 ± 129.6 mg/kg) compared to NPH (1060.5 ± 188.8
mg/kg (p=NS). The area under the glucose Rd curve (from T=60) was also not significantly
different for the two insulins at 322.2 ± 197.1 mg/kg for detemir and 930.4 ± 290.7mg/kg
for NPH (p=0.06).
Figure 29. Graph of Type 1 Subjects Mean Glucose Ra and Rd Shown as change from 60
minutes
Type 1 subjects Mean Ra & Rd Shown as Change from 60 minutes
JLT
-0.5 6p-X nO 180 240 300 360 420 480 540 600 660 720 780-840 900 960
-2.5
Time mins
HDRa-NPHRaHDRd
■NPHRd
Figure 30. Graph of Type 1 Subjects Mean Glucose Tracer;Tracee Ratio.
Type 1 subjects M ean Plasm a G lucose Traceritracee R atio
■ M ean N P H
• M ean LD
M ean H D
-1 8 0 -120 -60 0 60 120 180 240 300 360 42 0 4 8 0 540 600 66 0 720 780 840 900 960
T im e m ins
89
Table 9. Type 1 Subjects Glucose Metabolism.
LD NPH HD
Baseline Ra
mg/kg/min
2.15±0.40 2.14±0.18 1.94±0.19
Maximum change
from baseline Ra
mg/kg/min
-2.19±0.45 2.53 ±0.36 -2.20 ±0.30
AOC Ra mg/kg 1452.33± 401.61 1461.18 ±288.18 1452.71 ±268.83
Baseline Rd
mg/kg/min
1.93± 0.42 1.99 ±0.20 2.02 ±0.17
Maximum change
from baseline Rd
mg/kg/min
0.29 ±0.46 ** 2.22 ± 0.41 0.83 ±0.31 ***
AUC Rd mg/kg -234.32 ± 350 678 ±138.14 55.74 ±197.5 ***
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Glycerol Metabolism
There was no difference in baseline glycerol Ra between the 3 insulins with 1.47 ± 0.26
mg/kg/min for LD, 2.61 ±0.72 mg/kg/min for HD and 3.27 ± 0.85 for NPH. The maximum
change in glycerol Ra from baseline was significantly greater with NPH than HD (p<0.05)
but not different from LD, nor was LD different from HD with maximum change in glycerol
Ra being -0.55 ± 0.22 mg/kg/min for LD, -1.09 ± 0.39 mg/kg/min for HD and -2.07 ± 0.67
mg/kg/min for NPH. There was no difference in the areas over the curve between the 3
insulins being -143.7 ±211.1 mg/kg/min for LD, -374.7 ± 277.6 mg/kg/min for HD and
-828.3 ± 439.5 mg/kg/min for NPH.
90
Table 10. Type 1 Subjects Glycerol Metabolism,
LD NPH HD
Baseline Ra pmol/kg/min 1.47 ±0.26 3.27 ±0.85 2.61 ±0.72
Maximum change from
baseline Ra pmol/kg/min
-0.55 ± 0.22 -2.07 ± 0.67 -1.09 ±0.39***
AOC Ra pmol/kg -143.7 ±211.1 -828.3 ±439.5 -374.7 ± 277.6
* Indicates a statistically significant difference between LD and FID.
** Indicates a statistically significant difference between LD and NPH.
***Indieates a statistically significant difference between NPH and HD.
Figure 31. Graph of Type 1 Subjects Mean Gycerol Rate of Appearance.
Type 1 Subjects Mean Glycerol Rate of Appearance (Ra)
c5'SbII6ai
O
-60 60 180 300 420 540 660 780 900 1020
Mean H ■ Mean N • Mean L
T im e m in s
NEFA Metabolism
There was no difference in baseline plasma NEFA concentrations between LD and NPH or
LD and HD, with baseline values for LD and NPH being 1.03 ± 0.25 pmol/1 and 1.03 ± 0.23
pmol/1 respectively. Baseline concentrations for HD were significantly lower than NPH at
0.57 ± 0.20 pmol/1 (p<0.04). The maximum change from baseline was greater with NPH
than HD (p<0.004) but not different between LD and NPH with maximum decrease in
91
plasma concentration from baseline being 0.62 ± 0.17 pmol/l, 0.90 ± 0.20 pmol/1 and 0.36 ±
0.14 pmol/l, for LD, NPH and HD respectively. There was significantly lower area over the
curve for HD than NPH (p<0.02), but no difference in area over the curve between LD and
NPH or LD and HD, with values being 262.3 ± 167.0 pmol/1, 537.4 ± 208.9 pmol/1 and
110.5 ± 126.2 pmol/l for LD, NPH and HD respectively.
Table 11. Type 1 Subjects NEFA results.
LD NPH HD
Baseline NEFA pmol/1 1.03 ±0.25 1.03 ±0.23 0.57 ±0.20 ***
Maximum change from
baseline NEFA pmol/I
0.62 ±0.17 0.90 ± 0.20 0.36 ±0.14 ***
AOC NEFA pmol/l 262.3 ± 167.0 537.4 ±208.9 110.5 ±126.2***
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 32. Graph of Type 1 Subjects Mean Plasma Non Esterified Fatty Acid
Concentrations.
Type 1 subjects M ean Plasma Non Esterified Fatty Acid (NEFA) concentrations
- Mean LD- M ean NPH M ean HD
-180 -60 60 180 300 420 540 660 780 900 1020
Time mins
92
IGF-1 Plasma Concentrations
There was a significant difference in the baseline between HD and NPH (p<0.02), but no
difference in baseline between LD and HD or NPH and LD, baseline values being 16.92 ±
2.32 nmol/1,19.24 ± 2.59 nmol/1 and 16.56 ± 2.51 nmol/1 for LD, NPH and HD respectively.
There was a significant rise from baseline after dosing with all three insulins (p<0.02 for
all), with a maximum IGF-1 plasma concentration of 22.36 ± 2.69 nmol/1, 22.72 ± 3.29
nmol/1 and 22.11 ±2.86 nmol/1 for LD, NPH and HD respectively. However there was no
difference in plasma IGF-1 concentrations between the three insulins.
Table 12. Type 1 Subjects IGF-1 Results.
LD NPH HD
Baseline IGF-1 concentration nmol/I 16.92 ±2.32 19.24 ±2.59 16.56 ±2.51
Mean IGF-1 concentration nmol/1 17.86 ±2.03 19.52 ±2.84 18.23 ±2.36
Maximum IGF-1 concentration nmol/1 22.36 ±2.69 22.72 ±3.29 22.11 ±2.86
* Indicates a statistically significant cifference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
93
Figure 33. Graph of Type 1 Subjects Mean Plasma Insulin Like Growth Factor-1
Concentrations.
Type 1 Subjects Mean Plasma Insulin Like Growth Factor-1 (IGF-1) concentrations
30 1
----------- 2 5 -
-T Î I f i l l [ Î I lJL 1 1
■ T
40
- 0-
■ Mean NPH - Mean LD Mean HD
-180 -60 60 180 300 420 540 660 780 900 1020Time mins
IGFBP-1 Plasma Concentrations
All three insulins suppressed IGFBP-1 concentrations. There was no significant difference
in the baseline or minimum concentrations, nor was there a difference in the change for
baseline or the areas over the curve. These results are shown in the table below.
94
Table 13. Type 1 Subjeets IGFBP-1 Results.
LD NPH HD
Baseline IGFBP-1
concentration pg/l
36.03 ± 8.07 28.99 ±9.00 31.91 ± 3.87
Minimum IGFBP-1
Plasma
concentration pg/1
4.14 ±0.64 4.38 ±0.80 3.26 ± 0.66
Maximum change from
baseline pg/1
31.89 ±7.65 24.61 ±8.52 28.64 ± 3.39
AOC 20528.4 ±4933.6 15842.3 ±7045.4 19244.0 ± 1912.1
* Indicates a statistica ly significant difference between LD and HD.
** Indieates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 34. Graph of Type 1 Subjects Mean Plasma Insulin Like Growth Factor Binding
Protein-1 Concentrations.
Type 1 Subjects Mean Plasm Insulin Like Growth Factor Binding Protein-1 (IGFBP-1)Concentrations
140 1
ffla
-60 420-180 60 180 300 540 780660 900 1020
- Mean LD • Mean NPH Mean HD
Time mins
95
Insulin and insulin detemir
Mean AUC (0-960mins) was 2301.8 ± 505 (nmol/l)*min, 3703.2 ± 230 (nmol/l)*min and
306.7 ± 94.26(nmol/l)*min for LD, HD and NPH respectively. There was no individual
difference in the time to reach maximum concentration (Tmax) for the three insulins being
390 ± 32.8 mins, 515 ± 67.2 mins and 374 ± 74 mins for LD, HD and NPH respectively.
The maximum concentration (C max) was greater for HD (6160 ± 231.4 pmolf^) than LD
(4182 ± 1013.4 pmoir^). C max for NPH was 517.7 ± 117.1 pm olf\ The increase in AUC
and Cmax was less than two-fold however, one subject (Subject 205) dosed with 9 nmol/kg
insulin detemir had a very high insulin detemir plasma concentration profile.
C-Peptide
All subjects with Type 1 diabetes were c-peptide negative.
Glucagon
The glucagon plasma concentrations did not change after dosing with insulin detemir and
NPH insulin, and there was no difference between insulin detemir and NPH insulin.
96
Type 2 Subjects
Subject demographics
5 male subjeets and 6 female subjects were screened. 3 male subjects and 3 female subjects
passed screening. Of the 5 subjects who failed screening, 4 failed because their body mass
index was too high, and 1 failed because of general ill health.
The mean age of subjects was 58.7 ± 2.85 years, they had diabetes for an average of 6.88 ±
0.41 years and had good control with a mean HbAlc of 6.57 ± 0.35 %. Their weight was
82.8 ± 8 kg and body fat percentage 33.4 ± 3.59 %.
Plasma Glucose
Subjects arrived for the study day with high blood glucose concentrations having omitted
their long acting insulin for the previous 24 hours or oral hypoglyeaemic agents for the
preceding 48 hours. However there was no difference in plasma glucose levels on arrival
for the study days (9.74 ± 1.79 mmol/1 for LD, 9.09 ± 1.09 mmol/1 for HD and 10.15 ± 2.33
for NPH). Plasma glucose levels were brought into the clamp range using a tapering dose of
soluble insulin. There was no difference in the soluble insulin infusion rates on the separate
dosing days nor was there a differenee in baseline glucose levels at 6.51 ± 0.33 mmol/1 for
LD, 6.26 ± 0.33 for HD and 6.28 ± 0.37 for NPH. There was no difference in plasma
glucose concentrations following the administration of insulin being 6.46 ± 0.25 mmol/1 for
LD, 6.27 ± 0.27 mmol/1 for HD and 6.23 ± 0.35 mmol/1 for NPH. The mean and standard
differences of the individual coefficient of variation for glucose during the clamp studies
was 5.19+1.32 % for LD, 6.8+3.08 % for NPH and 5.5+1.26 % for HD.
97
Figure 35. Graph of Type 2 Subjects Mean Plasma Glucose Concentrations.
Type 2 Subjects Mean Plasma Glucose Concetnrations
I%I
-240 -120 0 120 240 360 480 600 720 840 960
■ Mean LD■ Mean NPH Mean HD
Time mins
Glucose infusion rate
There was no difference in the time to start the glucose infusion, maximum glucose infusion
rate, area under the curve of the glucose infusion or glucose infusion rate after 960 minutes
between HD and NPH. Comparing HD and LD there was no significant difference in time
to start the glucose infusion, but the maximum glucose infusion rate was lower with LD
(p<0.05), the area under the curve was lower with LD (p<0.03) and the glucose infusion rate
at the end of the study (960 mins) was lower with LD (p<0.01). Comparing LD and NPH
the glucose infusion rate at 960 minutes was lower with LD (p<0.02), but the differences in
the other measurement parameters did not reach statistical significance. The time to the start
of the glucose infusion was 81.5 ± 10.82 mins for LD, 50.8 ± 7.54 mins for HD and 46.0 ±
10.82 mins for NPH. The maximum glucose infusion rate was 1.93 ± 0.76 mg/kg/min for
LD, 2.96 ± 0.51 mg/kg/min for HD and 3.29 ± 0.76 mg/kg/min for NPH. The area under the
curve (ie total glucose infused) was 1130.2 ± 211.2 mg for LD, 1868.5 ± 348.6 mg for HD
and 1674 ± 211.2 mg for NPH. The glucose infusion rates at the end of the study were 0.89
98
±0.13 mg/kg/min for LD, 2,18 ± 0.39 mg/kg/min for HD and 1.79 ± 0.13 mg/kg/min for
NPH.
Table 14. Type 2 Subjects Plasma Glucose and Glucose Infusion Rate.
LD NPH HD
Mean Plasma Glucose mmol/I 6.51 ±0.33 6 2 8 ± 0 3 7 6 2 6 ± 0 3 3
Time to start of glucose
infusion mins
81.5 ± 10.82 46.0 ± 10.82 50.8 ±7.54
Maximum glucose infusion rate
mg/kg/min1.93 ±0.76* 3.29 ±0.76 2.96 ±0.51
Area under the curve mg/kg 1130.2 ±211.2* 1674±211.2 1868.5 ±348.6
Glucose infusion rate at 960
mins
0.89 ±0.13 *, ** 1.79 ±0.13 2.18 ±0.39
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 36 Graph of Type 2 Subjects Mean Glucose Infusion Rates.
Type 2 Subjects Mean Glucose Infusion Rates (GIR)
3.5
0.5
-40 160 360 560 760 960
■ Mean LD Mean NPH Mean HD
Time mins
99
Glucose Metabolism
There were no statistically significant differences found between LD, HD and NPH when
analysed for baseline Ra, maximum change in Ra, AOC Ra, baseline Rd, maximum change
in Rd or area over the curve Rd, except the maximum change in glucose Ra for LD was 1.58
±0.13 mg/kg/min which was lower than that for NPH (which was 2.07 ±0.19 mg/kg/min)
(p<0.04), the maximum change in Ra from baseline for HD was 2.01 ± 0.28 mg/kg/min.
There were also no differences found using ANOVA to compare the data sets.
Figure 37. Graph of Type 2 Subjects Mean Glucose Rate of Appearance.
Type 2 subjects Mean Glucose Rate of Appearance (Ra)
i j
! -I ^a
300-60-0.5
180 420 540 660 780 900 1020
Time mins
■ Mean Ra LD■ Mean Ra NPH Mean Ra HD
100
Figure 38. Graph of Type 2 Subjects Mean Glucose Rate of Disappearance.
Type 2 Subjects Mean Glucose Rate of Disappearance (Rd)
-60 60 180 300 420 540 660 780 900 1020
• Mean Rd LD ■ Mean Rd NPH Mean Rd HD
Time mins
Figure 39. Graph of Type 2 Subjects Mean Glucose Ra and Rd.
Type 2 Subjects Mean Glucose Ra & Rd
3.5 1
-3 -
2:5-T T T t i l - Mean Ra NPH
■ Mean Rd NPH Mean Ra HD Mean Rd HD
■ Mean Ra LD • Mean Rd LD
420 5io^
Time mins
1020
Because of the technical difficulties with our clamp protocol in the changeover from
intravenous to subcutaneous insulin we also analysed the glucose metabolism data in these
type 2 subjects from time 60 minutes. When calculated from 60 min the area over the curve
of suppression of Ra was significantly greater for HD (984 ± 135 mg/kg) compared to NPH
(535 ± 79 mg/kg)(p<0.04). The area under the glucose Rd curve was 672 ± 255 mg/kg for
HD and 797 ± 149 mg/kg for NPH (p=NS).
101
Figure 40. Graph of Type 2 Subjects Mean Glucose Ra and Rd Expressed as a change from
T=60 minutes.
Type 2 Subjects Glucose Ra and Rd Expressed as Change from T=60mins
2.5
.5e 1.5
I 1I 0.5
§ 0
% -0.5
I:40 300 360 420 480 540 600 660 720 780 840 900 960
-1.5
Time mins
- NPH Ra change NPH Rd change HD Ra change HD Rd change
Figure 41. Graph of Type 2 Subjects Mean Plasma Glucose Tracer:Tracee Ratio.
T ype 2 Subjects M ean Plasm a G lucose T racerT racee R atio
0 .16 n
I I T t t I I t T t I I t I T T T I I I - I ---------------
1-------- 1--------ÔtT 4 -
■ M ean N P H
■ M ean L D
M ean H D
-180 -1 2 0 -60 0 60 120 180 240 300 360 420 48 0 540 60 0 660 720 78 0 840 90 0 960
Tim e m ins
102
Table 15. Type 2 Subjects Glucose Metabolism.
LD NPH HD
Baseline Ra mg/kg/min 1.64 ±1.99 1.66 ± 0.15 1.67 ± 0.26
Maximum change from
baseline Ra mg/kg/min
1.58 ±0.13 ** 2.07 ±0.19 2.01 ± 0.28
AOC Ra mg/kg 1015.57 ±102.91 1164.99 ±71.87 1242.29 ±200.76
Baseline Rd mg/kg/min 1.83 ±0.17 1.90 ±0.16 1.83 ±0.24
Maximum change from
baseline Rd mg/kg/min
0.33 ±0.17 1.32 ±0.49 1.12 ±0.53
AUC Rd mg/kg -70.9 ±153.94 270.3 ±219.65 463.63 ± 355.48
* Indicates a statistical y significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Glycerol Metabolism
There was no difference in the glycerol Ra at baseline for the three insulins, being 2.26 ±
0.33 pmol/kg/min for LD, 1.64 ± 0.23 pmol/kg/min for HD and 2.14 ± 0.55 pmol/kg/min
for NPH. There was also no difference in the maximum change from baseline between the
three insulins being 0.80 ± 0.26 pmol for LD, 0.72 ± 0.23 pmol for HD and 1.12 ± 0.41
pmol for NPH. There was however significantly lower area under the curve for HD glycerol
Ra than NPH glycerol Ra (p<0.04), but no difference in the areas under the curve between
LD and HD and LD and NPH. The areas under the curve were 321.53 ±271.7 pmol for LD,
329.34 ± 239.21 pmol for HD and 654.5 ± 345.3 pmol for NPH.
103
Table 16. Type 2 Subjects Glycerol Metabolism.
LD NPH HD
Baseline Ra pmol/kg/min 2.26 ± 0.33 2.14 ±0.55 1.64 ±0.23
Maximum change from
baseline Ra pmol/kg/min
0.80 ± 0.26 1.12 ±0.41 0.72 ± 0.23
AOC Ra pmol/kg 321.53 ±271.7 654.5 ±345.3 329.34 ± 239.21***
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
Figure 42. Graph of Type 2 Subjects Mean Glycerol Rate of Appearance.
Type 2 Subjects Mean Glycerol Rate of Appearance (Ra)
2t5-
285
-60 60 420180 300 540 660 780 900 1020
Mean H • Mean N ■ Mean L
Time mins
NEFA Metabolism
There was no difference between the plasma NEFA concentrations at baseline. Plasma
NEFA concentrations were more suppressed following HD than ED (p<0.05), but there was
104
no difference between HD and NPH or NPH and LD with minimum concentrations for the
three insulins being 0.23 ± 0.04 |Limol/l, 0.13 ± 0.03 |imol/l and 0.13 ± 0.05 imol/1 for LD,
HD and NPH respectively. There was no difference in the change from baseline between
the three insulins nor was there a difference in the areas over the curve.
Table 17. Type 2 Subjects NEFA Results.
LD NPH HD
Baseline NEFA pmol/I 0.55 ±0.12 0.48 ±0.12 0.43 ± 0.06
Maximum change from
baseline NEFA pmol/l
0.31 ±0.15 0.35 ±0.13 0.30 ±0.09
AOC NEFA pmol/l 75.48 ± 142.0 187.15 ±97.43 137.88 ±156.03
* Indicates a statistically significant difference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
Indicates a statistically significant difference between NPH and HD.
Figure 43. Graph of Type 2 Subjects Mean Plasma Non Esterified Fatty Acid
Concentrations.
Type 2 Subjects Mean Plasma Non Esterified Fatty acid (NEFA)Concentrations
180 -60 60 180 300 420 540 660 780 900
Mean LD Mean NPH Mean HD
Time mins
105
IGF-1 Plasma Concentrations
There was no difference in the baseline, maximum or mean concentrations of plasma lGF-1
between the three insulins. However there was a significant rise from baseline found in all
three insulins (p<0.03). The plasma concentrations are shown in the table below.
Table 18. Type 2 Subjects lGF-1 Results.
LD NPH HD
Baseline IGF-1 concentration nmol/1 15.03 ±0.94 13.83 ±0.69 14.44 ±0.97
Mean IGF-1 concentration nmol/1 15.36 ±0.74 14.98 ±0.86 15.59 ±0.92
Maximum IGF-1 concentration nmol/1 18.47 ±1.00 17.82± 1.11 19.92 ± 1.48
* Indicates a statistically significant cifference between LD and HD.
** Indicates a statistically significant difference between LD and NPH.
***lndicates a statistically significant difference between NPH and HD.
Figure 44. Graph of Type 2 Subjects Mean Plasma Insulin Like Growth Factor-1
Concentrations
Type 2 Subjcts Mean Plasma Insulin Like Growth Factor-1 (IGF-1) Concentrations
25
MeanNPFI ■ Mean LD Mean HD
'----------------------r - 0 -
-180 -60 60 180 300 420 540 660 780 900 1020
Time mins
106
IGFBP-1 Plasma Concentrations
All three insulins suppressed IGFBP-1 concentrations. There was no significant difference
in the baseline or minimum concentrations, nor was there a difference in the change for
baseline or the areas over the curve. These results are shown in the table below.
Table 19. Type 2 Subjects IGFBP-1 Results.
LD NPH HD
Baseline IGFBP-1
concentration pg/l
13.08 ±7.70 22.08 ±16.19 10.73 ±5.63
Minimum IGFBP-1 Plasma
concentration pg/1
3.8 ±1.75 5.72 ±2.88 2.92 ±1.51
Maximum change from
baseline pg/l
9.28 ± 6.00 16.36 ±13.33 7.82 ±4.14
AOC 4713 ±3106.1 9033.1 ±7870.3 3976.5 ±1981.8
* Indicates a statistically significant difference between LD and F[D.
** Indicates a statistically significant difference between LD and NPH.
***Indicates a statistically significant difference between NPH and HD.
107
Figure 45. Graph of Type 2 Subjects Mean Plasma Insulin Like Growth Factor Binding
Protein-1 Concentrations.
Type 2 Subjects Mean Plasma Insulin Like Growth Factor-1 (IGFBP-1) Concentrations
2
60-180 -60 180 300 420 540 660 780 900 1020
• Mean LD Mean HD
• Mean NPH
Time mins
Insulin and insulin detemir
Mean AUC (0-960mins) was 1577.6 ± 219.39 nmol, 3131.5 ± 261.92 nmol and 141.8 ±
31.76 nmol for LD, HD and NPH respectively. There was no individual difference in the
time to reach maximum concentration (Tmax) for the three insulins being 638.3 ± 86.98
mins, 625.3 ± 48.49 mins and 565.0 ± 62.57 mins for LD, HD and NPH respectively. The
maximum concentration (C max) was greater for HD (4996.7 ± 189.84 pmolF^) than LD
(2660.0 ± 311.02 pmoir^). C max for NPH was 208.8 ± 108.1 pmoH'\ A doubling in insulin
detemir dose resulted in an approximately two-fold increase in AUC and Cmax.
C-peptide
C-peptide at baseline was similar in each insulin study being 0.46±0.12 nmol/1, 0.49+0.17
nmol/1 and 0.52±0.14 nmol/1 for LD. NPH and HD respectively. C-peptide suppressed
following each insulin and there was no difference in this suppression between the insulins
with minimum c-peptide 0.32±0.08 nmol/1, 0.26+0.07 nmol/1 and 0.27+0.07 nmol/1
following LD, NPH and HD respectively.
108
The plasma concentrations of C-peptide decreased slightly after dosing with insulin detemir
and NPH insulin in subjects with type 2 diabetes.
Figure 46. Graph of type 2 Subjects Mean Plasma C-Peptide Concentrations.
Type 2 Subjects Mean Plasma C-Peptide Concentrations
0.8
0.7
0.6
0 0.51
^ 0.4
0.3
0.2
0.1
0
!
I . .T
r •: " ;; " ::' Î I T T
t i . . \^ N /-----
Mean NPH Mean HD Mean LD
0 120 240 360 480 600 720 840 960
Time Mins
Glucagon
The glucagon plasma concentrations did not change after dosing with insulin detemir and
NPH insulin, and there was no difference between insulin detemir and NPH insulin.
The mean glucagon plasma concentrations were highest in healthy subjects and lowest in
subjects with type 1 diabetes and subjects with type 2 diabetes were between the two other
subject groups.
109
Combined results
Subjects with diabetes are more insulin resistant than those without, and therefore the data
from the three groups of subjects (Healthy, Type 1 diabetes and Type 2 diabetes) were
analysed further by combining the results and comparing the effect of the equipotent doses
of detemir and NPH. The doses of insulin given for healthy subjects (9 nmol/kg insulin
detemir and 0.3 iu/kg NPH ) were half that of the doses given to the subjects with diabetes
(18 nmol/kg insulin detemir and 0.6 iu/kg NPH), in spite of this the doses required similar
amounts of glucose to be infused to maintain euglycaemia throughout all the groups. It may
therefore be justified to pool the data as shown below.
Figure 47. Graph of Mean Glucose Infusion Rates for Each Subject Group.
Mean Glucose Infusion Rates For Each Subject Group
S0Û60 2 s0 1 cÜ
0.5
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Healthy •Type 1 Type 2
Time mins
Plasma Glucose
The plasma glucose thoughout the clamp period (0-960mins) was lower in healthy subjects
than in those with type 1 diabetes (p<0.0001) or type 2 diabetes (p<0.0001) being 4.94 ±
0.06 mmol in the healthy group, 6.63 ±0.15 in subjects with Type 1 diabetes and 6.27 ±
0.22 in subjects with Type 2 diabetes. There was no difference between the plasma glucose
110
in the clamp period between the two groups of subjects with diabetes. With plasma glucose
pooled across the subject groups there was no difference in the plasma glucose for those
subjects treated with NPH (5.90 ± 0.22 mmol) compared with those treated with insulin
detemir (5.88 ± 0.21 mmol)(p=NS).
Glucose Infusion Rate
There were no differences in the glucose infusion rates either between the subject groups
(healthy, type 1, and type 2 subjects), or between the two treatments (detemir and NPH)
when analysed for total glucose infused, mean GIR, maximum GIR and GIR at 960 mins.
Table 20. Combined Glucose and Glucose Infusion Rate Data.
Healthy
subjects
Type 1
subjects
Type 2
Subjects
Detemir NPH
Area under
the curve
GIR, mg/kg
2055.99 ±
256.95
1776.1 ±
246.71
1771.5 ±
196.75
1819.0 ±
191.95
1936.3 ±
195.6
Maximum
GIR
mg/kg/min
3.79 ± 0.49 3.63 ±0.51 3.12 ±0.44 3.16 ±0.35 3.89 ± 0.42
Mean GIR
mg/kg/min
2.12 ±0.29 1.84 ±0.26 1.76 ±0.21 1.83 ±0.20 2.00 ± 0.22
GIR at 960
mins
mg/kg/min
1.64 ±0.15 1.40 ±0.39 1.98 ±0.20 1.83 ±0.23 1.51 ±0.42
I l l
Figure 48. Graph of Mean Glucose Infusion Rates for Each Insulin from the Combined
Data from each Subject Group.
Mean Glucose Infusion Rates for each insulin in all subject Groups
5
0.5
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Mean D • Mean NPH
Time mins
Glucose Metabolism
There were no significant differences found in the effect of the two insulins on glucose Ra
when compared at baseline, minimum Ra, maximum change from baseline, area over the
curve Ra or by ANOVA (p=0.06). Baseline Rd was similar for the two insulins but NPH
had a significantly higher maximum Rd compared with detemir (p<0.04) and also had a
greater change from baseline (p<0.02). There area under the curve for Rd appeared lower
for detemir, but the differences were not statistically significant. Using ANOVA the
differences in the effect on Rd did not quite reach statistical significance p=0.05. Of the
mean glucose lowering effect of insulin detemir over the 960 minutes 1.39 ± 0.08
mg/kg/min glucose was infused to overcome Ra (79%) and 0.37 ± 0.07 mg/kg/min glucose
was infused to overcome Rd (21%), whereas for NPH this was 1.29 ± 0.08 mg/kg/min for
Ra (65%) and 0.68 ± 0.10 mg/kg/min for Rd (35%)(p<0.001).
12
Table 2 1 . Combined Glucose Metabolism Data.
Detemir NPH
Baseline Ra mg/kg/min 1.87±0.11 1.95 ±0.08
Minimum Ra mg/kg/min -0.23 ± 0.09 -0.19 ±0.12
Maximum change from baseline Ra mg/kg/min 2.09 ±0.14 2.14±0.12
AOC Ra mg/kg 1375.3 ± 111.5 1282.2 ±84.1
Baseline Rd mg/kg/min 1.97±0.10 1.97 ±0.08
Maximum Rd mg/kg/min 3.18 ±0.25 3 ^ 8 ± 0 JT
Maximum change from baseline Rd mg/kg/min 1.21 ±0.25 1.90 ±0.29
AUC Rd mg/kg 367.38 ± 159.9 625.4 ± 140.4
Figure 49. Graph of Mean Glucose Ra and Rd for each insulin from the combined data from
each subject group.
Data from the three subject groups combined to show Mean Glucose Ra & Rd
Mean D Rd Mean D Ra Mean NPH Rd Mean NPH Ra
a
-60 0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
Time mins
13
Figure 50. Bar charts showing the relative proportion of glucose lowering effect from effect on glucose Ra and glucose Rd of each insulin using data combined from the three subject groups.
Bar chart showing mean effect on Glucose Ra and Rd following injection with Detemir
B Mean D Rd □ Mean D Ra
Time mins
B a r c h a r t s h o w i n g m e a n e f f e c t o n G l u c o s e R a a n d R d f o l l o w i n g i n j e c t i o n w i t h N P H
| o ^
JLJJT r 4 4 l r i n T r r
■ M e a n N P H R d
□ M e a n N P H R a
UO 360 480 720 <%0
T i m e m i n s
114
Figure 51. Pie chart showing the overall mean proportion of effect of each insulin on
glucose Ra and Rd using data combined from the three subject groups.
Pie chart showing overall mean effect of Detemir on Glucose Ra and Rd
21%
79%
□ Ra ■ Rd
Pie chart showing overall mean effect of NPH on glucose Ra and Rd
35%
65%
□ Ra ■ Rd
15
Chapter 4
Discussion
In the 1990s the link between metabolic control and development of microvascular
complications was firmly established by the diabetes control and complications trial research
group (DCCT). The DCCT was a study in 1441 subjects with type 1 diabetes. They were
randomised to either intensive control (with basal bolus regime and strict targets) or
conventional treatment. Those in the intensive group had a mean HbAlc of 7.1% whereas
those in the conventional treatment group had an HbAlc of 9.0%. There was a lower rate of
complications in the intensive group with a reduction in: the risk of retinopathy of 45%,
development of microalbuminuria by 35% and clinical neuropathy by 60%. For every 10%
reduction in HbAlc (ie a reduction in HbAlc from 10% to 9%) there was about 35% risk
reduction of retinopathy, 25-44% risk reduction for nephropathy and 30% risk reduction for
neuropathy. However the risk of hypoglycaemia increased threefold in the intensive
treatment group. The United Kingdom Prospective diabetes study research group (UKPDS)
looked at intensive versus conventional treatment in patients with type 2 diabetes, and
showed risk reduction in microvascular endpoints, myocardial infarction, cataract extraction,
retinopathy and albuminuria.
These studies were large, well-controlled and highly regarded and demonstrated the
importance of good control for reducing the risk of complications. As a result of these
studies the objective now in the treatment of diabetes is not the avoidance of multiple
injections but the imitation of normal physiology.
116
Figure 52. Diagram of the Physiological Plasma Insulin Secretion Profile.
Physiological serum insulin secretion profileB reakfast Lunch Dinner
500-1
g 400 -
I300-
r
§
I* 2 0 0 -
s100
6 a.m. 10 a.m. 2 p.m. 6 p.m. 10 p.m. 2 a.m. 6 a.m.T im e
Polonsky KS et al. N E n g lJ M ed 1996;334:777-83
In people without diabetes insulin is released constantly from the pancreas at a “basal level”
with boluses of insulin released at meal times to cover additional prandial requirements.
Bolus (mealtime) insulin release from the pancreas limits hyperglycaemia after meals by an
immediate sharp rise and a peak at 1 hour (Bell et al., 1996). At first insulin regimes were
felt not to be satisfactory because the meal time bolus was with regular insulin. This insulin
when injected subcutaneously existed in hexameric form and therefore must dissociate into
dimers and monomers before it could be absorbed into the capillary circulation
(Cunningham et al, 1955). Conventional “short acting” or regular insulin had onset of
action about 1 hour after injection and peak action at 2-4 hours (Howey et al, 1994). Thus
patients were advised to inject insulin around 30 minutes before a meal (an instruction that
was rarely adhered to), and to take snacks between meals to compensate for prolonged
hyperinsulinaemia, which can lead to hypoglycaemia (Barnett, 2001). Rapid acting
analogues were designed to mimic the prandial release of insulin. They have fast absorption
times with onset of action after 20 to 30 minutes and peak action at about 1-2 hours, leaving
a shorter duration of action than regular insulin (Howey et al, 1994) and as such more
117
closely mimic physiological insulin release in response to meals. All the pivotal studies
involving several thousand patients with Type 1 and Type 2 diabetes showed significant
reductions in postprandial glucose excursions using the rapid acting insulin analogues
compared with human soluble insulin (Owens et al, 2000, Anderson et al 1997a, Anderson
et al, 1997b and Kang et al, 1991). This was entirely expected based on the more rapid
absorption and higher peak insulin levels demonstrated in earlier phase studies. These
registration studies, however did not show a significant improvement in glycated
haemoglobin, although there was a significant improvement in patient satisfaction (Barnet,
2001). The latter was primarily related to the fact that patients could take their rapid acting
insulin analogue immediately prior to eating a meal, rather than half an hour before.
Furthermore the incidence o f hypoglycaemia was reduced in those using rapid acting insulin
analogues compared to those using regular insulin as their bolus (Brunelle et al, 1998,
Holleman et al 1997).
Because these analogues did not normalise blood glucose control as hoped attention was
turned to basal insulins. It was recognised that little or no improvements in basal insulins
had occurred since the 1930’s when the concept of crystalline insulins as a method of
protraction was conceived. It is accepted that these longer acting insulins do not imitate the
basal insulin production from the pancreas, but rather have peaks of action after around 6-8
hours and had variable absorption, resulting in huge differences in daily basal insulin levels
(Home, 1999). Blood glucose variability correlates with mean glycaemic exposure and risk
of hypoglycaemia (Moberg et al, 1994), and also with mortality (Muggeo et al 2000).
Somogyi (1959) observed that the blood glucose and glycosuria of diabetic patients often
showed wide fluctuations even when both carbohydrate content of the diet and insulin dose
are unchanged.
With conventional crystalline suspensions of insulin some of this can be explained by poor
mixing techniques, with variations in the actual concentration of insulin injected varying in
one study from 5-214% (Bolli and Owens, 2000). Studies have shown that fluctuations in
absorption rate are the main cause for unpredictable insulin activity, accounting for as much
as 80% of the variability in the action of NPH insulin or insulin lente (Guthrie, 2001).
118
Laurizten et al (1979) injected subjects with radiolabelled NPH insulin over a 12 day period,
they found that absorption varied greatly from full to only 19% on some doses. Chen et al
(2003) also observed large variations in absorption of NPH insulin after subcutaneous
injection with 50% of the original dose of insulin found at the injection site for around 12
hours on one day and around 96 hours on another in the same patient. Luzio et al (2003)
also describes large variations in insulin absorption rates with 50% absorption between 7
and 18.2 hours for NPH and 10.2 and 42 hours for glargine. There is also considerable
variation in the absorption of NPH depending on the site of injection (Lauritzen et al, 1982),
and the time of day that that site is used (Chen et al, 2003).
The drive now was to design an insulin to mimic the secretion of basal insulin in healthy
subjects so to suppress glucose production between meals and overnight. Renewed interest
in prolonging the duration of action of insulin has led to developments along different
strategic routes. The expectation that proinsulin would increase basal insulin concentrations
was short lived because of concerns on cardiovascular safety that arose from clinical studies
(Owens et al, 2001). Proinsulin’s metabolite, des (64,65)-human proinsulin had an even
shorter period of action than NPH insulin and was discontinued (Owens et al, 2001).
Surprisingly, the strategy to slow absorption and clearance by strengthening hexameric
interactions in the form of cobalt (Ill)-insulin complex also proved unsuccessful (Burge et
al, 1997). However, increase of the isoelectric point of insulin from 5.4 towards neutrality
thereby causing precipitation or crystallisation at the site of injection, delays absorption, and
prolongs the effect of insulin (Markussenet al, 1987). The first acid soluble extended action
insulin analogue made by this method was Novosol Basal GlyA21, ArgB27, ThrB30-NH2-
insulin (Jorgensen et al, 1989), which was discontinued because of escalating dose
requirements (reduced bioavailability) possibly associated with local inflammatory reactions
at the site of injection. On the basis of the same concept, a di-arginylinsulin analogue
known as insulin glargine was developed that had a much slower and more reproducible rate
of subcutaneous absorption than NPH insulin (Owens et al, 2000), resulting in slower onset
and extended hypoglycaemic effect (Lepore et al, 2000, Heineman et al 2000). An
additional phenol molecule bound to the hexamer of crystalline insulin glargine might
further contribute to its protracted action.
119
Insulin glargine is the first of these long acting insulin analogues to make it to the market,
being launched in 2002. Plasma glucose and insulin concentrations vary less between
patients when insulin glargine is used compared to ultralente insulin (Lepore et al, 2000,
Scholtz et al 1999). Comparison of insulin glargine with NPH insulin in isoglycaemic
clamp studies shows that a bolus subcutaneous injection of NPH insulin (0.3iu/kg) produces
a definite peak effect within 3-6 hours, before returning to baseline at 16 hours. By contrast
insulin glargine plateaus at a concentration 2-3 times lower than NPH within 6-8 hours,
remains essentially unchanged for 24 hours (Heinemann et al, 2000) and is unaffected by the
site of injection (Owens and Barnett, 2000). In studies comparing insulin glargine with
NPH as part of a basal bolus regime, subjects using insulin glargine had a greater reduction
in fasting blood glucose (Pieber et al, 2000, Ratner et al, 2000, Rosenstock et al, 2000,
Raskin et al, 2000), less variability in blood glucose (Rosenstock et al, 2000, Raskin et al,
2000), a lower frequency of hypoglycaemia (predominantly nocturnal) (Pieber et al, 2000,
Ratner et al, 2000) and a fall in glycosylated haemoglobin (Pieber et al, 2000). However
injection site pain which is negligible in the NPH group is experienced in 6.1% with insulin
glargine, presumably because of the acidic pH (4.0) of insulin glargine (Raskin et al, 2000).
Insulin glargine has also been compared to NPH insulin in patients with type 2 diabetes.
Both insulins were given at bedtime in poorly controlled circumstances in which insulin
naïve patients were allowed to continue oral agents (Yki-Jarvinen and Westerbacka, 2000).
Overall glycated haemoglobin improved to the same extent in each group, however there
was a lower frequency of nocturnal hypoglycaemia in the glargine treated group. Similarly
in patients with type 2 diabetes changing from NPH once or twice daily to glargine at
bedtime, there was a lower risk of nocturnal hypoglycaemia and equivalent glycaemic
control (Rosenstock et al, 2001). In these studies no differences in doses were recorded, but
either no weight gain (Rosenstock et al, 2001), or less weight gain with insulin glargine
(Wang et al, 2003), than with NPH.
Insulin detemir is a long acting soluble insulin analogue, with a 14 C fatty acid chain
attached to position B29 and omission of the amino acid residue at position B30 of the
human insulin molecule (Brunner et al, 2000). The fatty acid residue associates with free
120
fatty acid binding sites on albumin in interstitial fluid and plasma (Kurtzhals et al, 1995)
developing an equilibrium between free and albumin bound analogue. Only the free fraction
of insulin detemir permeates the capillary wall entering the blood stream and binding to
albumin again. To reach peripheral target tissues (muscle and adipose) free insulin detemir
must again pass the capillary wall. The principle of protraction is based on slow release
from albumin and strong self-association at the injection site (Havelund et al, 2004, Brunner
et al, 2000). Insulin detemir uniquely among the long acting insulin analogues does not
form a precipitate at the site of injection and this property has been considered to be of key
importance in eliminating a potential source of variability (Lindholm, 2002). It is also likely
that plasma albumin binding may buffer the effect of short duration changes in the depot
absorption rate, thereby limiting the pharmacodynamic variability (Kurtzals, 2004). This
favourable action was found in a study reported by Heise et al (2004) and Bott et al (2003)
compared the pharmacokinetic, pharmacodynamic profiles and within subject variability of
NPH, detemir and glargine in patients with Type 1 diabetes. Subjects received 4 identical
doses of an insulin on four different dosing days, 18 subjects were randomised to each
insulin. In this study reproducibility of the glucose infusion rate curves were analysed and it
was found that the probability of reproducibility was worst for NPH insulin, followed by
glargine and that detemir had the most reproducible curve. The coefficient of variation for
pharmacodynamics as observed by glucose infusion rate and pharmacokinetics as assessed
by insulin concentration area under the curve was significantly lower with detemir than NPH
and glargine. This favourable action may result in lower frequencies of unpredicted hypo or
hyperglycaemia. A study reported by Pieber et al (2002) looking at duration of action,
pharmacodynamic profile and between subject variability of insulin detemir in subjects with
type 1 diabetes, found a linear dose response relationship for pharmacokinetic and
pharmacodynamic properties of detemir. It reported a longer duration of action at a dose of
0.4u/kg (20 hours) than NPH at a dose of 0.3iu/kg (13 hours). They also reported less
between subject variation.
Capillary endothelial cells in adipose and muscle limit the transfer of insulin detemir from
the circulation into the extravascular extracellular space (Havellund et al, 2004), whereas in
the liver the sinusoids are lined by highly fenestrated epithelial cells, between which there
121
are large gaps. The sinusoids also have no basal lamina. Because of the large gaps in the
endothelial cell layer and the absence of a continuous basal lamina, no significant barrier
exists between blood plasma in the sinusoid and the hepatocyte plasma membrane (Ross,
2003). Insulin detemir can therefore pass freely to this “space of Disse” (the perisinusoidal
space) Gray, 1973, Reichen, 1999). Hepatocytes will be freely exposed to insulin detemir
possibly resulting in a greater effect of this insulin analogue on the liver than on peripheral
tissue.
Insulin detemir’s binding properties may provide an alternative theory for insulin detemir’s
apparent hepatoselectivity. The albumin binding properties of insulin detemir have been
extensively studied using an immobilised human serum albumin (HSA) preparation
(Kurtzhals et al, 1995 and 1997). The binding constant of insulin detemir for HSA was
determined to be 2.4 x 10 M'^ at 23°C and 1.0 x 10 M'^ at 37°C in this assay system.
Based on this binding constant, insulin detemir is estimated to be 98.4% bound in plasma
(37°C, 0.6mM albumin), which is consistent with the above determined plasma protein
binding (98.8%) of insulin detemir (Kurtzhals et al, 1995). We can see therefore that
detemir has a higher binding affinity to albumin at colder temperatures. Core body
temperature is higher than that at the periphery (Kim et al, 1998). This may mean that a
higher proportion of detemir is free (that is not bound to albumin) in the core (ie at the liver)
whereas in the periphery (that is in adipose and muscle tissue) detemir will be more tightly
bound to albumin and less will pass through the capillary wall to act on target cells. This
may explain why detemir’s hepatoselective effects in people with type 2 diabetes is more
difficult to demonstrate, as these people tend to have a higher proportion of their adipose
tissue centrally (that is intra-abdominal fat deposits) where temperatures are more likely to
resemble core body temperatures, and thus detemir will be less bound to albumin and more
freely able to pass the capillary wall and act on target adipose cells.
122
Why is it important for an insulin to act more on the liver than in the periphery?
Glucose metabolism:
As discussed earlier the brain almost exclusively utilizes glucose as a source of energy
(Cryer 1997). The brain has high energy requirements (required to maintain the ionic
gradient across nerve membranes, critical for maintenance of brain activity) estimated at
approximately 1 mg/kg/min corresponding to about lOOg glucose per 24 hours, however it
has low energy reserves (enough for only a few minutes), and is unable to synthesize
glucose (Bolli, 1998). Intricate mechanisms to protect the supply of glucose to the brain
have therefore evolved. Only the liver and the kidney are able to release glucose. This
release results from the presence of glucose-6-phosphatase in these organs which liberates
glucose into the bloodstream (Bolli and Fanelli, 1999). It is generally believed that the liver
makes the major quantitative contribution during this state, whereas the kidney has a minor
role, except during prolonged starvation, when it is estimated that the kidney may contribute
almost half of glucose production (Bolli 1998, Cahill et al,1966, and Cahill, 1970).
Blood glucose is sensed in the brain (Frohman and Nagai, 1976, Ritter et al, 1981, Borg et
al, 1995), pancreas(Taborsky et al, 1998) and liver (Connolly et al, 1996), each of these
organs is involved in the counter-regulation response to low blood glucose in healthy
subjects. A fall in blood sugar concentration that may occur in the postabsorptive state leads
initially to a reduction in insulin secretion, by direct action of glucose on p-cells of the islets
of Langerhans (Cryer 1997, Bolli, 1998). This leads to a reduction in the plasma insulin
concentration in the portal circulation, and therefore a reduction in the suppression of
hepatic glucose output (or a rise in hepatic glucose output). If the blood glucose continues
to drop counter-regulatory hormones are stimulated and released.
As discussed earlier in the introduction section on glucose metabolism, all counter
regulatory responses to hypoglycaemia increase hepatic glucose output either directly alone
or in combination with an indirect effect by the stimulation of lipolysis (Bolli, 1998), some
also decrease peripheral glucose uptake. In normal physiology these defence mechanisms
123
are appropriate as insulin is released into the portal circulation and has its first effects at the
liver, and the liver is essential in the production of glucose to support brain metabolism
(Bolli, 1998). In insulin dependent diabetes no insulin is released by the pancreas
exogenous insulin is administered by subcutaneous delivery, it circulates peripherally first,
and is controlled by pharmacokinetics, with peak actions occurring 1 to 8 hours after
injection (Bolli, 1998). The insulin profile does not resemble physiological insulin profiles
of a normal 24 hour day, and is unable to respond to changing insuhn need (Amiel, 1998,
Cryer et al, 2003). In this situation the liver is relatively underinsulinised and peripheral
tissue over insulinised. In the event of hypoglycaemia, the first line of counter regulation
(that is a decrease in insulin release) is not possible, insulin levels remain inappropriately
high suppressing hepatic glucose output, increasing peripheral glucose uptake and inhibiting
lipolysis. Further counter regulation has evolved to exert itself primarily at the liver,
suppression of hepatic glucose output is thus easily overcome since in these circumstances
as the liver is relatively underinsulinsed however the peripheral over insulinisation presents
a more difficult obstacle, and therefore it could be argued that greater counter regulation has
to occur to prevent a further fall in plasma glucose and maintain the brain’s glucose supply.
This may cause later hyperglycaemia as a result of the sustained release of counter
regulatory hormones and their prolonged duration of action may be a factor in the overall
deterioration in glycaemic control seen in patients with frequent hypoglycaemia.
Hypoglycaemia is a common problem encountered in the treatment of diabetes. In one
study on 411 patients with type 1 diabetes (Pramming et a l, 1991) patients had 1.8 episodes
of mild hypoglycaemia per patient per week, and 0.027 episodes of severe hypoglycaemia
per patient per week. Others have suggested people with diabetes have an average of 2
episodes of symptomatic hypoglycaemia a week (Reichard, 1991, Macleod et al, 1993), and
an estimated 2-4% of deaths of people with diabetes are attributable to hypoglycaemia
(Laing et al, 1999). Socially this also causes problems to patients with diabetes with around
three times the percentage of people with type 1 diabetes being refused car insurance than
those without diabetes (Songer et a l, 1991). In the DCCT and UKPDS studies rates of
severe hypoglycaemia were higher for a given HbAlc in conventionally treated patients than
those on intensive treatment regimes, however those on intensive treatment regimes were
124
more likely to have more frequent hypoglycaemia. Hypoglycaemia seems to be the price to
be paid for improving rates of progression to complication.
Duration of diabetes is associated with decreases in the hormonal response to
hypoglycaemia (Bolli et al., 1983). This can lead to a vicious circle of hypoglycaemia
causing impaired physiological responses to hypoglycaemia, causing reduced awareness to
hypoglycaemia, causing increased vulnerability to further episodes and thus more
hypoglycaemia.
Lipid metabolism and weight gain:
The counterregulatory hormone response to insulin acts primarily by increasing glucose Ra.
This is less effective against exogenous insulin, which acts mainly to increase glucose Rd.
Thus an insulin analogue exerting greater effects on the liver than the periphery has the
potential to reduce hypoglycaemic episodes and may reduce weight gain through less
snacking to avoid hypoglycaemia. Peripheral hyperinsulinaemia increases peripheral
glucose uptake and lipogenesis and decreases lipolysis contributing to the weight gain
associated with insulin therapy (Andreani, 1999). Kruszynska et al (1985) investigated the
effects on carbohydrate metabolism of peripheral insulin delivery compared to portal insulin
delivery. They found in rats that hepatic glycogen concentrations are lower in rats receiving
insulin into the systemic circulation compared to those rats receiving insulin via the portal
circulation. Others have shown that sensors in the liver to increased intracellular ATP
concentrations affect eating behaviour via signals transmitted via vagus afferents from the
liver to the brain (Mayer, 1991, Freidman, 1995, Langhans, 1996). Although the control of
eating behaviour remains a highly complex and controversial subject with many factors
having been shown to influence it. Nonetheless recent large-scale clinical trials comparing
detemir and NPH provide support for the theory that by acting in a more physiological way
detemir has advantages over NPH. These trials have shown that nocturnal hypoglycaemic
episodes are reduced and weight gain is lower with insulin detemir when compared to NPH
insulin (Russell-Jones et a l , 2002, De Leeuw et al 2002, Vague et al 2003).
125
IGF-l/GH secretion:
In patients with type 1 diabetes growth hormone levels have been found to be raised
throughout the day with increased frequency and amplitude of pulsatility and exaggerated
responses to a wide variety of stimuli (Tamborlane et al, 1979). Although there is a marked
increase in growth hormone, IGF-1 levels are at the low end of normal or are suppressed
(Tan et al, 1986). IGF-1 levels are not normalised with intensive therapy or by maintainence
of good control (Amiel et al, 1984). Delivery of insulin via the portal system influences
hepatic IGF-1 production. In rats intraperitoneal insulin given to mimic portal insulin
increases IGF-1 mRNA expression and systemic IGF-1 levels (Russell-Jones DL, 1992),
Griffen et al (1987) also provided evidence that insulin delivered via the portal circulation is
more effective than systemic insulin at restoring IGF-I levels in diabetic rats.
Subcutaneously delivered insulin leads to unphysiological low portal insulin and this may
explain why, despite intensive s.c. insulin therapy, circulating IGF-I levels in diabetic
patients remain either low or at best low normal (Rudolf et al 1981). IGF-1 itself has a
negative feedback effect on GH secretion. It seems likely that in IDDM and NIDDM the
low level of IGF-1 is sensed at the hypothalamus/pituitary level, and GH is hypersecreted
(Simpson et al, 1998). This will lead to a distorted GH/IGF-1 axis with raised GH and
inappropriately low IGF-1. This elevated GH contributes to the insulin resistance and
causes the “dawn phenomenon” which is difficult to control (Fowelin, et al, 1991, Periello et
al, 1990). Changes in GH levels have been implicated in the development of microvascular
complications, particularly retinopathy but also nephropathy (Chiarelli et al, 2000). Low
levels of circulating IGF-1 have also been implicated to be associated with neuropathy
(Migdalis et al, 1995). Some authors have suggested that relative underinsulinisation of the
liver may play an important role in the development of microvascular complications
(Sonksen et al, 1993, Wurzberg et al 1995, Janssen et al, 2000). It is possible therefore to
speculate that an insulin that has greater effect at the liver may be associated with reduced
frequency of microvascular complications.
126
Is detemir hepatoselective?
From other studies; A study reported by Russell-Jones et al (2004) compared once daily
nocturnal injection of detemir and NPH in a 6 month multicentre randomised controlled
trial. They found that treatment for 6 months with insulin detemir resulted in slight lowering
of HbAlc, lowering of fasting plasma glucose, less fluctuation in blood glucose levels
especially at night, less nocturnal hypoglycaemia in the last 5 months of treatment and a
relative decrease in weight in comparison to the group treated with NPH. Vague et al (2003)
reported the efficacy and tolerability of detemir compared to NPH in a basal bolus regime
using asp art as the bolus insulin. They also reported that detemir had similar metabolic
control judged by HbAlc but found a lower within subject variation in fasting blood
glucose, a flatter and more stable blood glucose profile during the night, a lower risk of
hypoglycaemia, a relative decrease in weight gain and a good safety profile. De Leeuw et
al (2002) reported a study using NPH or detemir as a twice daily basal regime, and using
aspart as the bolus insulin. Again it was found that detemir had similar glycaemic control
over 12 months, significantly lower risk of hypoglycaemia and a significant and favourable
weight reduction in comparison to NPH. Pieber et al (2004) reported a study using a basal
bolus regime with aspart as the bolus insulin and twice daily basal injections of detemir or
NPH. The NPH was given at breakfast and bedtime, in one detemir arm detemir was given
as the NPH (ie breakfast and bedtime) and in the other it was given at breakfast and at
suppertime. Compared to NPH detemir provided significantly lower fasting plasma glucose
in both arms, significantly lower within subject variation in both arms and less weight gain.
Thus suggesting that detemir may be administered more flexibly than NPH. Home et al
(2004) reported a similar study with type 1 subjects randomised to receive their basal insulin
in a basal bolus regime as either NPH morning and bedtime, detemir morning and bedtime
or detemir 12 hours apart. Again they found detemir produced significantly lower fasting
plasma glucose, lower within subject variability, lower nocturnal hypoglycaemia, less
weight gain and flexible administration. Haak et al (2004) reported a study with type 2
individuals using a basal bolus regime with aspart as the bolus insulin, and twice daily NPH
or detemir. They found lower risk of hypoglycaemia, less weight increase and similar
glycaemic control. Hermansen et al (2004) compared an analogue basal bolus regime of
127
twice daily detemir (morning and bedtime) with aspart as the bolus insulin to a human
insulin regime of twice daily NPH (morning and bedtime) with actrapid as the bolus insulin.
Subjects had type 1 diabetes and were randomised in an open label trial for 18 weeks. They
found the analogue regime lowered HbAlc more than the human insulin regime and caused
less nocturnal hypoglycaemia.
From this study: The equivalence in glucose infusion rate profiles is of key importance
when interpreting the comparative effects of two insulin preparations on glucose Ra and
glucose Rd. Insulin exhibits a different proportion of action at each of these sites at varying
concentrations. In diabetic patients (Brown et al 1978) and in diabetic dogs (Issekutz et al
1974) it has been demonstrated that peripheral infusion of insulin at low rates of
0.5mU/kg/min without euglycaemia reduces hepatic glucose output without increase in
glucose Rd. Thus at low concentrations of insulin the majority of the glucose lowering
effect occurs by suppression of glucose Ra, with only a small effect on glucose Rd. At
higher concentrations the majority of action occurs by glucose Rd once glucose Ra has
effectively been completely suppressed (Sonksen and Sonksen, 2000). Therefore results
from studies that compare the effects of two insulin preparations on glucose Ra and Rd
where the glucose infusion profile is not equivalent must be treated with caution as the
difference in effect may be purely related to dose.
In this study we have demonstrated in healthy subjects that 9nmol/kg insulin detemir is
equipotent to 0.3 iu/kg NPH, with a very similar time action profile, we have also
demonstrated in subjects with Type 1 diabetes that 18nmol/kg insulin detemir is equipotent
to 0.6 iu/kg NPH, with a similar time action profile There were small but significant
differences in the effect of this dose of insulin detemir on glucose Ra and a lesser non
significant effect on glucose Rd compared to the equipotent dose of NPH suggesting a
greater effect on liver than on peripheral tissue. In the group with type 1 diabetes there were
also small but significant differences in the effect of this dose of insulin detemir on glucose
Rd but no measurable differences in the effect on glucose Ra (which at this dose of insulin
is almost completely suppressed) suggesting NPH has a greater effect on the periphery than
an equipotent dose of insulin detemir.
128
The reduced effect of insulin detemir on peripheral tissues was confirmed by the
measurements of NEFA concentration. The suppression of NEFA concentrations by insulin
detemir was less than with an equipotent dose of NPH, both in healthy subjects and to a
lesser extent in those with type 1 diabetes (the effect in subjects with type 1 diabetes was
further clouded by apparent although non statistical differences at baseline). It has been
suggested that insulin’s suppressive effect on peripheral lipolysis with a decrease in NEFA
concentrations is an important signal mediating an indirect effect of insulin on glucose Ra
(Bergman, 2000). The demonstration in this study, that detemir has greater effect on
glucose Ra but less effect on NEFA concentration than an equipotent dose of NPH insulin,
suggests that the action of detemir on glucose Ra is independent of NEFA delivery to the
liver.
In type 2 subjects the results are more difficult to determine. There was great inter
individual variation in the response to the insulins. There were also technical difficulties
with interpreting the clamp results as both the Ra and Rd for each individual fell after the
baseline bloods had been taken and the injection given at time 0 reaching a nadir at around
60 minutes, suggesting that the patients were not at steady state at the start of the clamp.
This is probably a result of the technical difficulties associated with the study design of
controlling the blood glucose using intravenous insulin and converting to subcutaneous
insulin at time 0 even though the delayed absorption of these insulins would mean that
subjects would effectively be without insulin for a short period at the beginning of the study.
It is not clear why the effect of this is more pronounced in the subjects with Type 2 diabetes.
However when analysed using time 60 minutes as baseline there was greater effect on
glucose Ra with insulin detemir than the equipotent dose of NPH, and a small (although not
significant) reduction in the effect on glucose Rd for detemir compared to NPH.
Looking at the effect on lipid metabolism in subjects with type 2 diabetes findings are
similar to those in subjects with type 1 diabetes and healthy subjects, in that there was less
suppression of glycerol Ra with insulin detemir compared with NPH (although this
difference again was clouded by apparent although non statistical differences at baseline)
129
and less effect on NEFA suppression, although the latter did not reach statistical
significance. There were no measurable differences in these insulins effects on IGF-1 or
IGFBP-1 in subjects with type 1 or type 2 diabetes.
When analysed together as a combined set the data from these 19 subjects show the glucose
lowering effect of these insulins are not different. Insulin detemir suppresses hepatic
glucose output equivalently but has less effect on peripheral glucose uptake. This further
supports our hypothesis.
The clinical implications of an insulin that may act in a more hepatoselective way are
exciting. As described above by acting more physiologically detemir may give protection
from profound hypoglycaemia and may reduce the frequency of hypoglycaemic episodes.
Studies so far indeed indicate this reduced hypoglycaemia risk. This is of great importance
to patients in whom the fear of hypoglycaemia is often a barrier to improved control and
thus lowered risk of micro and macrovascular complications. There have been no studies
with long enough follow up to determine the long term benefits of detemir with respect to its
potential effects on reducing microvascular complications as a result of its differential effect
on IGF-1. These advantages may only become apparent with decades of use of this insulin.
Limitations of this work:
This study had many limitations, lack of resources and man-power meant that it was not
feasible to attempt to achieve more than 6 analysable sets of data in each group of patients.
This in itself meant that we performed nearly 1400 hours of euglycaemic clamping. With
low subject numbers it was difficult at times to demonstrate statistically significant
differences in our results although similar trends ran throughout our findings.
Constraints of metabolic study:
Metabolic studies often provide such results with day to day intra-individual variation
making interpretation more difficult. Also the calculations used in these studies rely on
130
assumptions to determine glucose Ra and glucose Rd, and data entered in these calculations
undergoes a smoothing program which may result in important differences being minimised.
As described above it is of utmost importance that the glucose infusion rate profiles be as
equivalent as possible for comparison between two insulins (ie their equipotency must be as
near as possible) for interpretation of differences to be truly due to differences in insulin
action and not merely differences due to dose. We were exceptionally fortunate that in the
healthy subjects the glucose infusion rates were inseparable and thus the small differences
that we found could be truly attributed to effect rather than dose.
These differences appear small but if each dose results in 10% greater action at the liver than
the periphery as demonstrated in the healthy subjects in this study, then the cumulative
effect over the years of administration may be clinically relevant.
The results from the two groups of patients with diabetes has been more difficult to
interpret. The studies were technically much more difficult, in order to avoid “hangover
effect” from any other long acting insulin, patients were asked to omit their long acting
insulin from 24 hours prior to the study. However we know from data presented above that
NPH has highly variable absorption, and therefore it is possible that some of the NPH
previously injected was still acting thus accounting for the greater intra-individual variation
in results in this group of patients. Also lowering blood glucose after presentation on the
study day was not without difficulties, as rapid lowering meant difficulties arose during the
baseline period and looking at the data from both diabetic groups it is clear that subjects
were not at steady state at baseline or for the 30-60 minutes following subcutaneous
injection as the calculations clearly show falls in both Ra and Rd during this time period.
The type 2 patient subject group also shows considerable variation. The very nature of this
group of patients is variable given that they represent a spectrum of diabetes from insulin
resistance to decreased insulin production. Therefore as a group their results are more
variable. This means that the small differences we demonstrated in other groups in this
study were more difficult to demonstrate in this group of patients. Also this group consisted
of some patients with higher abdominal girth (although this was not actually measured) and
131
as previously discussed this may lead to a greater effect of detemir on central deposits of
adipose tissue and thus less relative effect on the liver. Given these considerations it may
have been prudent to increase the numbers of subjects studied in the type 2 group, or
controlled for other variables such as abdominal girth.
In conclusion this study shows that when compared to an equipotent dose of NPH, detemir
has a greater effect on glucose metabolism in the liver than in peripheral tissues. This
reduced effect of insulin detemir on peripheral tissues was confirmed by less of an effect on
NEFA concentrations than NPH. This suggests that this insulin analogue has the potential to
imitate the normal insulin gradient, thus mimicking normal insulin physiology more closely
than any other insulin currently available. This may mean that this insulin has therapeutic
advantages over other currently available insulin preparations.
132
Chapter 5
References
Adams MJ, Blundell TL, Dodson EJ, Dodson GG, Vijayan M, Baker EN, Harding MM,
Hodgkin DC, Rimmer B, Sheat S. Structure of 2 zinc crystals. Nature 224:491-495.1969.
Ader M and Bergman RN. Peripheral effects of insulin dominate suppression of fasting
hepatic glucose production. Am J Physiol 258: El 020-El 032.1990.
Agius L, Chowdery MH, Davis SN, Alberti KGMM. Regulation of ketogenesis,
gluconeogenesis, and glycogen synthesis by insulin and proinsulin in rat hepatocyte
monolayer cultures. Diabetes. 35:1286-1293. 1986.
Allen FM. Studies concerning glycosuria and diabetes. Bradley RF, Krall LP. Eds
Cambridge, MA, Harvard Univ. Press. 1913.
American Diabetes Association. Report of the expert committee on the diagnosis and
classification of diabetes mellitus. Diabetes Care 22 (Supp 1). 1999.
Amiel S, Archibald H, Chusney G et al: Ketone infusion lowers hormonal responses to
hypoglycaemia: Evidence for acute cerebral utilization of a non glucose fuel. Clin Sci
81:189.1991.
Amiel SA, Sherwin RS, Hintze RE, Gertner JM, Press M, Tamborlane WV. Effect of
diabetes and its control on insulin-like growth factors in the young subject with type 1
diabetes. Diabetes. 33:1175-1179. 1984.
Amiel SA. Hypoglycaemia avoidance-technology and knowledge. Lancet. Aug
15;352(9127):502-3.1998.
133
Anderson JH Jr, Brunelle RL, Koivister VA et al. Improved mealtime treatment of diabetes
mellitus using an insulin analogue. Clin Theraneut. 19:62-72.1997b.
Anderson JH Jr, Brunelle RL, Koivister VA et al. Multicentre insulin lispro study group.
Reduction of postprandial hyperglycaemia and frequency of hypoglycaemia in IDDM
patients on insulin analogue treatment. Diabetes 46:265-270. 1997a.
Andreani D. Lights and shadows of insulin treatment seen by a senior diabetologist. Exp.
Clin. Endocrinol. Diabetes 107 (Supp 2) S1-S5. 1999.
Argoud GM, Schade DS, and Eaton RP. Underestimation of hepatic glucose production by
radioactive and stable stracers. Am J Phvsiol. 252: E606-E615. 1987.
Ashcroft FM and Ashcroft SJH. Properties and functions of ATP-sensitive K-channels.
Cell Sis. 2:197-214.1990.
Ashcroft FM, Harrison DE and Ashcroft SJH. Glucose induces closure of single potassium
channels in isolated rat pancreatic p-cells. Nature 312:446-448. 1984.
Baltensperger K., Kozma LM, Cherniak AD, Klarlund JK, Chawla A, Banerjee U and Czech
MP. Binding of Ras activator son of the sevenless to insulin receptor substrate-1 specific
signalling complexes. Science 260:1950-1952. 1993.
Barnett AH. Rapid acting insulin analogues. Insulin Made Easv (Medical Education
Partnership) Ed AH Barnett. 2001.
Bell GI, Pictet RL, Rutter WJ, Cordell B, Tischer E, Goodman HM. Sequence of the human
insulin gene. Nature 284:26-32. 1980.
Bell GI, Pilkis SJ, Weber IT, Polonsky KS. Glucokinase mutations, insulin secretion, and
diabetes mellitus. Annu Rev Phvsiol.:58:171 -86. 1996.
134
Bell JI, Hockaday TDR. Diabetes Mellitus. Oxford Textbook of Medicine (third edition).
1448-1504.1996.
Bell PM , Firth RG and Rizza RA. Assesment of insulin action in insulin dependent
diabetes mellitus using [6- '^C] glucose, [3-^H] glucose and [2-^H] glucose: differences in the
apparent pattern of insulin resistance depending on the isotope used. J Clin Invest 78: 1479-
1486.1986.
Bier DM, Arnold KJ, Sherman WR, Holland WH, Holmes WF, Kipnis DM. In vivo
measurement of glucose and alanine metabolism with stable isotope tracers. Diabetes
26(11):1005-1015. 1977.
Bergman RN: Non-esterified fatty acids and the liver: why is insulin secreted into the portal
vein? Diabetologia 43:946-952,2000.
Berhanu P, Olefsky JM, Tsai P, Saunders DT and Brandenburg D. Internalization and
molecular processing of insulin receptors in isolated rat adipocytes. Proc. Natl. Acad. Sci.
USA. 79: 4069-4073.1982.
Bliss M. The discovery of insulin. Lily. 1982.
Blundell TL, Bedarkar S, Rinderknecht E and Humbel RE. Insulin like growth factor: a
model for tertiary structure accounting for immunoreactivity and receptor binding. Proc.
Natl. Acad. Sci. USA. 75:180-184. 1978.
Blundell TL, Cutfield JF, Cutfield SM, Dodson EJ, Dodson GG, Hodgkin DC, Mercola DA
and Vijayan M. Atomic postions in rhombohedral 2-zinc insulin crystals. Nature 231:506-
511. 1971.
135
Blundell TL, Dodson GG, Hodgkin DC and Mercola DA. Insulin: The structure in the
crystal and its reflection in chemistry and biology. Adv. Protein Chem. 26:279-402. 1972.
Bolli G, de Feo P, Compagnucci P, Cartechini MG, Angeletti G, Santeusanio F, Brunetti P,
Gerich JE. Abnormal glucose counterregulation in insulin-dependent diabetes mellitus.
Interaction of anti-insulin antibodies and impaired glucagon and epinephrine secretion.
Diabetes. Feb:32r2k 134-41. 1983.
Bolli G, De Feo P, Perriello G, De Cosmo S, Ventura M, Campbell P and Gerich J. Role of
hepatic autoregulation in defense against hypoglycaemia in humans. J Clin. Invest. 75:1623-
1631.1985.
Bolli GB. Counter regulatory mechanisms in insulin-induced hypoglycaemia in
humans:relevance to the problem of intensive treatment of IDDM. J Paed Endoc & Metab
11:103-115. 1998.
Bolli GB, Fanelli CG. Physiology of glucose counter regulation to hypoglycaemia.
Endocrinol & Metab clinics of North America. 28 (3):467-492.1999.
Bolli GB, Owens, DR. Insulin glargine. Lancet Vol 356 (9228) 443-445. 2000.
Bolli GB: The pharmacokinetic basis of insulin therapy in diabetes mellitus. Diab Res and
Clin Prac 6: S3-S16, 1989.
Boni-Schnetzler M, Schmid, Meier PJ, Froesch ER Insulin regulates insulin-like growth
factor I mRNA in rat hepatocytes. Am J Phvsiol: 260:E846-E851. 1991.
Borg W, Sherman R, During M et al: Local ventromedial hypothalamus glycopaenia triggers
counterregulatory hormone release. Diabetes 44:180.1995.
136
Bomfeldt KE, Amqvist HJ, Enberg B, Mathews LS, Norstedt G. Regulation of insulin-like
growth factor I and growth hormone receptor gene expression by diabetes and nutritional
state in rat tissues. J Endocrinol: 122, 651-656. 1989.
Bortz WM, Paul P, Haff AC, Holmes WL. Glycerol turnover and oxidation in man. J Clin
Invest.. 51:1537-1546. 1972.
Bott S, Tusek C, Jacobsen L, Kristensen A, Heise T. Insulin Detemir Reaches Steady-State
after the First Day of Treatment and Shows a Peakless Time-Action Profile with Twice
Daily-Applications. Diabetes. 52 SUPPLEMENT 1:A112, June 2003.
Bowes SB, Benn JJ, Scobie IN, Umpleby AM, Lowy C and Sonksen PH. Glucose
metabolism in patients with Cushing’s syndrome. Clin Endocrinol. 34:311-316. 1991.
Bradbury JH and Brown LR, Nuclear-magnetic-resonance-spectroscopic studies of the
amino groups of insulin. Eur. J. Biochem. 76: 573-582.1977.
Brange J, Ribel U, Hansen JF, Dodson G, Hansen MT, Havelund S, Melberg SG, Norris F,
Norris K, Snel L, Soresen AR, Voigt HO. Monomeric insulins obtained by protein
engineering and their medical implications. Nature. 333:679-682.1988.
Britton HG. Permeability of the human red cell to labelled glucose. J. Phvsiol. 170:1-20.
1964.
Brown PM, Tompkins CV, Juul SM, and Sonksen PH. Mechanism of action of insulin in
diabetic patients: a dose related effect on glucose production and utilization. Brit. Med J . 1 :
1239-1242.1978.
Brownsey RW, Denton RM. Evidence that insulin activates fat cell acetyl-CoA carboxylase
by increased phosphorylation at a sepcific site. Biochem J. 202:77-86. 1982.
137
Brunelle RL, Llewlyn J, Anderson JH et al. Meta-analysis of the effect of insulin lispro on
severe hypoglycaemia in patients with type 1 diabetes. Diab Care. 21: 1726-1731. 1998.
Brunner G A, Sendhofer G, Wutte A, Ellmerer M, Sogaard B, Siebenhofer A, Hirschberger
S, Krejs GJ, Pieber TR: Pharmacokinetic and pharmacodynamic properties of long-acting
insulin analogue NN304 in comparison to NPH in humans. Exp Clin Endocrinol Diabetes
108: 100-5,2000.
Bucchini D, Ripoche M-A, Stinnakre M-G, Desbois P, Lores P, Monthioux E, Absil J,
Lepesant J-A, Pictet R and Jami J. Pancreatic expression of human insulin gene in
transgenic mice. Proc. Natl Acad Sci USA. 83: 2511-2515. 1986.
Burge MR, Schade DS. Insulins. Endoc. And Metab. Clin. Of N. Am. 26(3):575-598. 1997.
Bums and Palade GE. Studies on blood capillaries II. Transport of ferritin molecules across
the wall of muscle capillaries. J Cell Biol. 37:277-299. (1968b).
Bums RR and Palade GE. Studies on blood capillaries I. J Cell Biol 37:244-276. 1968a.
Butler PC, Kmshak EJ, Schwenk WF, Haymond MW and Rizza RA. Hepatic and
extrahepatic responses to insulin in NIDDM and non-diabetic humans. Assessment in
absence of artefact introduced by tritiated non-glucose contaminants. Diabetes 39:217-225.
1990.
Cahill GF Jr, Herrera MG, Morgan AP et al. Hormone fuel inter-relationship during fasting.
J Clin. Invest 45:1751.1966.
Cahill GF Jr. Starvation in man. N Engl. J Med 282:668. 1970.
138
Campbell PJ, Mandarino LJ and Gerich JE. Quantitation of the relative impairment in
actions of insulin on hepatic glucose production and peripheral glucose uptake in non
insulin dependent diabetes mellitus. Metabolism 37:15-21. 1988.
Carpentier J-L, Gazzano H, Van Obberghen E, Fehlmann M, Freychet P and Orci L.
Intracellular pathways followed by the insulin receptor covalently coupled to ^ I-
photoreactive insulin during internalisation and recycling. J. Cell. Biol. 102:989-996. 1986.
Carpentier, J-L., Gorden, P., Barazzone, P., Freychet, P., Le Cam, A. and Orci, L.
Intracellular localization of 1251-labelled insulin in hepatocytes from intact rat liver. Proc.
Natl. Acad. Sci USA. 76:2803-2807.1979b.
Carpentier, J-L., Gorden, P., Freychet, P., Le Cam, A. and Orci, L. Lysosomal association of
internalised enzymes is concentrated in cis Golgi cistemae. J Clin. Invest. 63;1249-1261.
1979 a.
Caspar DLD, Clarage J, Salunke DM and Clarage M. Liquid like movements in crystalline
insulin. Nature 332: 659-662.1988.
Chan BL, Lisanti MP, Rodriguez-Boulan E, and Saltiel AR. Insulin-Stimulated release of
lipoprotein lipase by metabolism of its phosphoinositol anchor. Science 241:1670-1672.
1998.
Chan SJ, Keim P and Steiner DF. Cell free synthesis of rat preproinsulins; charachterisation
and partial amino acid sequence determination. Proc. Natl. Acad. Sci. USA. 73: 1964-1968.
1976.
Chance RE, Elios RM and Bromer WW. Porcine proinsulin: Charachterization and amino
acid sequence. Science 161:165-167. 1968.
139
Chap Z, Ishida T, Chou J, Hartly CJ, Entman ML, Brandenberg D, Jones RH and Field JB..
First pass hepatic extraction and metabolic effect of insulin and insulin analogues. Am J
Phvsiol 252: E209-E217.1987.
Chen JW, Christiansen JS, Lauritzen T. Limitations to subcutaneous insulin administration
in type 1 diabetes. Diabetes Obes Metab. Jul;5(4):223-33. 2003.
Cherrington AD. Banting Lecture 1997. Control of Glucose Uptake and Release by the Liver In vivo. Diabetes 48:1198-1214.
Chiarelli F, Santilli F and Mohn A. Role of growth factors in the development of diabetic
complications. HpnnRes (53):53-67.2000.
Chiasson JL, Liljenquist JE, Finger FE, Lacy WW and Nashville MD. Differential
sensitivity of glycogenolysis and gluconeogenesis to insulin in dogs. Diabetes 25:283-291.
1976.
Ciraldi TP, Brady D and Olefsky JM. Kinetics of biosynthetic human proinsulin action in
isolated rat adipocytes. Diabetes 32: 627-632. 1986.
dem enti F and Palade GE. Intestinal capillaries. I. Permeability to peroxidase and ferritin. J
Cell. Biol. 41: 33-58.1969.
Cobelli C, Mari A and Ferrannini E. Non-steady state: error analysis of Steele’s model and
developments for glucose kinetics. Am J Phvsiol. 252: E679-689. 1987.
Connolly CC, Myers SR, Neal DW et al:In the absence of counterregulatory hormones the
increase in hepatic glucose production during insulin induced hypoglycaemia in dogs is
initiated by the liver rather than the brain. Diabetes 45:1805,1996.
Consoli A, Kennedy F, Miles J, Gerich J. Determination of the Krebs cycle metabolic
carbon exchange in vivo and its use to estimate the individual contributions of
140
gluconeogenesis and glycogenolysis to overall glucose output in man. J Clin. Invest.
80:1303-1310.1987.
Cowan JS and Hetanyi Jr G. Glucoregulatory responses in normal and diabetic dogs
recorded by a new tracer method. Metabolism 20:360-372. 1971.
Cryer PE, Davis, SN, Shamoon H. Hypoglycaemia in diabetes. Diabetes Care 26(6): 1902-
1912. 2003.
Cryer PE. Regulation of glucose metabolism in man. J Int. Med. 229 Suppl 2 31-39. 1991.
Cryer PE: Hvnoglvcaemia. Pathophvsioloev. Diagnosis and Treatment. New York, Oxford
Univ Press, 1997.
Cunningham LW, Fischer RL, Vestling CS. A study of the binding of zinc and cobalt by
insulin. J Am Chem Soc.77:5703-5705. 1955.
Cutfield FJ, Cutfield SM, Dodson EJ, DodsonGG and Sabesan MN. Low resolution crstal
structure and purification by affinity chromatography. J. Mol. Biol. 87:23-30. 1974.
De Bolo RC, Steele R, Altszuler N, Dunn A and Bishop JS. On the hormonal regulation of
carbohydrtae metabolism; studies with C*" glucose. Rec Prog Horm Res 19: 445-485. 1963.
de Leeuw I, Vague P, Selam J-L, Skeie S, Elte JWF, Lang H and Draeger E: Lower risk of
nocturnal hypoglycaemia and favourable weight development in type 1 subjects after 12
months of treatment with insulin detemir vs NPH insulin. Diabetologia 45 (Suppl 2):799A,
2002.
DeFronzo RA, Andres R, Bledsoe TA, Boden G, Faloona GA and Tobin JD. A test of the
hypothesis that the rate of fall in glucose concentration triggers counter regulatory hormonal
responses in man. Diabetes 26: 445-52. 1977.
141
DeFronzo RA, Gunnarsson R, Bjorkman O, Olsson M, Wahren J. Effects of insulin on
peripheral and splanchnic glucose metabolism in noninsulin-dependent (type II) diabetes
mellitus. J Clin Invest : 76: 149-155.1985.
Denton RM, Randle PJ, Bridges BJ, Cooper RH, Kerbey AL, Pask HT, Severson DL,
Stansbie D and Whitehouse S. Regulation of mammalian pyruvate dehydrogenase. Mol.
Cell. Biochem. 9: 27-53. 1975.
Drury DR, Wick AN, Bancroft RW, McKay EM. Glucose metabolism in the
hepatectomized rabbit. Am J Phvsiol: 164:207. 1951.
Farquhar, M.G. Multiple pathways of exocytosis, endocytosis and membrane recycling:
validation of a Golgi route. Fed. Proc. 42: 2407-2413. 1983.
Fehlmann, M., Carpentier, J-L., Le Cam, A, Thamm, P., Saunders, D., Brandenberg, D,
Orci, L and Freychet, P. Biochemical and morphological evidence that the insulin receptor
is internalised with insulin in hepatocytes. J. Cell. Biol. 93: 82-87. 1982a.
Fehlmann, M., Carpentier, J-L., Van Obberghen, E., Freychet, P., Thamm, P., Saunders, D.,
Brandenberg, D. and Orci L. Internalised insulin receptors are recycled to the cell surface in
rat hepatocytes. Proc. Natl. Acad Sci. USA. 79:5921-5925. 1982b.
FeligP. The glucose-alanine cycle. Metabolism 22: 179-207. 1973.
Field JB. Extraction of insulin by the liver. Annu Rev Med 20: 309-314. 1973.
Finegood DT, Bergman RN and Vranic M Estimation of endogenous glucose production
during hyperinsulinaemic euglcyaemic glucose clamps. Comparison of unlabelled and
labelled exogenous glucose infiisates. Diabetes 36: 914-924. 1987.
142
Finegood DT, Bergman RN, and Vranic M. Modelling error and apparent isotope
discrimination confound estimation of endogenous glucose production during euglycaemic
clamps. Diabetes: 37: 1025-1030. 1988.
Firth RG, Bell PM, Marsh HM, Hansen I and Rizza RA. Postprandial hyperglycaemia in
patients with non-insulin dependent diabetes mellitus: the relationship between hepatic and
extrahepatic insulin resistance and obesity. Metabolism 36: 1091-1095. 1986.
Flakoll PJ, Zheng M, Vaughan S, Borel MJ: Determination of stable isotopic enrichment
and concentration of glycerol in plasma via gas chromatgraphy-mass spectrometry for the
estimation of lipolysis in vivo. J Chromatographv B Biomed Sci Appl 744: 47-54. 2000.
Flock EV, Tyce G, Owen CA. Regulatory effects of insulin on liver and brain glucose
metabolism. Endocrinol 84:392-406.1969.
Fowelin J, Attvall S, von Schenk H, Smith U, Lager I. Charachterisation of the insulin-
antagonistic effect of GH in man. Diabetologia;34:500-506. 1991.
Friedman MI. Control of energy intake by energy metabolism. Am J Clin Nutr. 62 (5
Suppl):1096S-1100S, 1995.
Frohman L, Nagai K: Cnetral nervous system-mediated stimulation of glucagons secretion
in the dog following 2-deoxy-glucose. Metabolism 25:1449.1976.
Galloway JA, Hooper SA, Spadlin CT, Howey DC, Frank BH, Bowsher RR and Anderson
JH. Biosynthetic human proinsulin. Diabetes Care 15 666-692. 1992.
Gammeltoft S, Van Obberghen E. Protein kinase activity of the insulin receptor. Biochem
L 2 3 5 (l):l- ll . 1986.
143
Garber AJ, Cryer PE, Santiago JV, Haymond JW, Pagliara AS and Kipnis DM. The role of
adrenergic mechanisms in the substrate and hormonal response to insulin induced
hypoglycaemia in man. J Clin. Invest. 58:7-15.1976.
Garlick DG and Renkin EM Transport of large molecules form plasma to interstitial fluid
and lymph in dogs. Am J Phvsiol. 219:1595-1605. 1970.
Garratt CJ, Brown D and Evans RM. Direct evidence for an effect of insulin on intracellular
metabolism. Int J. Biochem. 10:943-945. 1979.
Gerich JE. Insulin glargine: Long acting basal insulin analog for improved metabolic
control. Curr Med Res Qpin Jan flL31-37. 2004.
Giacca A, Fisher SJ, Qing Shi Z, Gupta R, Lavina H, Lickley A and Vranic M. Importance
of peripheral insulin levels for insulin induced suppression of glucose production in
depancreatized dogs. J Clin Invest 90: 1769-1777. 1990.
Giacca A, Gupta R, Efendic S, Hall K, Skotter A, Linckley L and Vranic M (1990).
Differential effects of IGF-1 and insulin glucoregulation and fat metabolism in
depancreatised dogs. Diabetes 39: 340-347. 1992.
Glaubler HS, Revers R, Henry R, Schmeiser L, Wallace P, Kolterman OG, Cohen RM,
Rubenstein AH, Galloway JA, Frank BH and Olefsky JM. In vivo inactivation of proinsulin
action on glucose disposal and hepatic glucose production in normal man. Diabetes 35: 311-
317. 1986.
Goldstein S, Harp JB, Phillips LS. Nutrition and somatomedin XXII: molecular regulation
of insulin-like growth factor I during fasting and refeeding in rats. J Mol Endocrinol: 6, 33-
43. 1991.
Gray H: Gravs Anatomv 35th ed. 1973, p i308-1309
144
Griffen SC, Russel SM, Katz LS, Nicoll CS. Insulin exerts metabolic and growth promoting
effects by a direct action on the liver in vivo: Clarification of the fiinctional significance of
the portal vascular link between the beta cells of the pancreatic islets and the liver. Proc
Natl. Acad Sci 84: 7300-7304. 1987.
Guthrie R. Is there a need for a better basal insulin? Clin Diab 19:66-70. 2001.
Haak T, Tiengo A, Draeger E, Suntum M, Waldhausl W. Lower within subject variability of
fasting blood glucose and reduced weight gain with insulin detemir compared to NPH
insulin in patients with type 2 diabetes. Diab Obes Metab In press. 2004.
Hagedom HC, Jensen BN, Krarup NB, Wodstrug I. Protamine insulinate. J Am Med Assoc
106;177-180.1936.
Hallas-Moller K. The lente insulins. Diabetes 5:7-14.1956.
Hamilton-Wessler M, Ader M, Dea M, Moore D, Jorgensen PN, Markussen J, Bergman RN.
Mechanism of protracted metabolic effects of fatty acid acylated insulin, NN304, in dogs:
retention of NN304 by albumin. Diabetologia. Oct;42(10): 1254-63. 1999.
Havellund S, Ribel U, Plum A, Jonassen IB, Volund A, Markussen J and Kurtzals P. The
mechanism of protraction of insulin detemir, a long acting, acylated analogue of human
insulim Diabetes abstract supplement for 64th scientific sessions, P462,2004.
Heindreich K, Bemhau P, Brandenburg D, Olefsky JM. Degradation of insulin receptors in
rat adipocytes. Diabetes 32:1001-1009. 1983.
Heinemann L, Linkeschova R, Rave K, Hompesch B, Sedlak M, Heise T. Time action
profile of the long acting insulin analogue insulin glargine (HOE901) in comparison with
those of NPH and placebo. Diabetes Care. 23(5):644-9. 2000.
145
Heinemann L., Sinha K., Weyer C., Loftager M., Hirsberger S., Heise T. Time-action profile
of the soluble fatty acid acylated long-acting insulin analogue NN304. Diabetic Medicine
16(4)332-338, 1999.
Heise T, Nosek L, Ronn BB, Endahl L, Heinemann L, Kapitza C and Draeger E. Lower
within-Subject Variability of Insulin Detemir in Comparison to NPH Insulin and Insulin
Glargine in Subjects with Type 1 Diabetes. Diabetes. 53:1614-1620. 2004.
Hermansen K, Fontaine P, Kukolja KK, Peterkova V, Leth G, Gall MA. Insulin analogues
(insulin detemir and insulin aspart) versus traditional human insulins (NPH insulin and
regular human insulin) in a basal bolus therapy for patients with type 1 diabetes.
Diabetologia. 47(3):622-629.2004.
Hermansen K, Madsbad S, Perrild H, Kristensen A and Axelsen M. Comparison of the
Soluble Basal Insulin Analog Insulin Detemir With NPH Insulin A randomized open
crossover trial in type 1 diabetic subjects on basal-bolus therapy. Diabetes Care 24:296-301,
2001 .
Hers AG and Ave L. Gluconeogenesis and related aspects of glycolysis. Annu Rev
Biochem 52: 617-553. 1983.
Hildebrandt P: Subcutaneous absorption of insulin in insulin dependent diabetic patients.
Influence of species, physico-chemical properties of insulin and physiological factors.
Danish Medical Bulletin 38: 337-346,1991.
Hill AV. The diffusion of oxygen and lactic acid through tissues. Proc. Rov. Soc. IBioll
104B:39-96. 1928.
Hodgkin DC. The Banting Memorial Lecture 1972. The structure of insulin. Diabetes
21:1131-1150. 1972.
146
Holland R, Hardie DG, Glegg RA and Zammit VA. Evidence that glucagon mediated
inhibition of acetyl CoA carboxylase in isolated adipocytes involvesincreased
phosphorylation of the enzyme by cyclic-AMP-dependent protein kinase. Biochem J.
226:139-149. 1985.
Holleman F, Schmitt H, Rottiers R et al. Reduced frequency of severe hypoglycaemia and
coma in well-controlled IDDM patients treated with insulin lispro. Diab Care: 20: 1827-
1832. 1997.
Home P, Bartley P, Rusell-Jones, Hanaire-Boutin H, Heeg JE, Abrams P, Landin-Olsson M,
Hylleberg B, Lang H, Draeger E. Insulin Detemir Offers Improved Glycémie Control
compared with NPH insulin in people with Type 1 Diabetes. Diabetes Care. 27(5): 1081-
1087. 2004.
Home P. Insulin Glargine: the first clinically useful extended acting insulin in half a
century? Exp. Qpin. Invest. Drugs. 0999). 8131307-314. 1999.
Hordern V, Wright J, Umpleby M, Jacobsen L, Russell-Jones D. Stable Isotope Studies
Show Effect of Insulin Detemir and NPH on Hepatic Glucose Output and Peripheral
Glucose Uptake after Subcutaneous Administration in Healthy Subjects. Diabetes. 50
SUPPLEMENT 2:A504-A505, June 2001.
Hordern VM, Wright JE, Umpleby AM, Jacobsen LV, Draeger E, Russell-Jones DL.
Stable isotope studies show differences in effect of insulin detemir and NPH on hepatic
glucose output and peripheral glucose uptake after subcutaneous administration in subjects
with Type 1 diabetes. Diabetes 51 Suppl A292. 2002.
Hordern VM, Wright JE, Umpleby M, Jacobsen LV, Russell-Jones. Stable Isotope Studies
Show Differences in Effect of Insulin Detemir and NPH on Hepatic Glucose Output after
147
Subcutaneous Administration in Subjects with Type 2 Diabetes Mellitus. Diabetes 52
Supplement 2370 PO. 2003.
Hoskins RG. Endocrinology. The glands and their functions. Kegan Paul, Trench and
Trubner & Co. Ltd. 1941.
Hovorka R, Powrie JK, Smith GD, Sonksen PH, Carson ER and Jones RH. Five
compartmental model of insulin kinetics and its use to investigate action of chloroquine in
NIDDM. Am J Phvsiol. 265: E162-E175. 1993.
Howell SL and Tyhurst M. The cytoskeleton and insulin secretion. Diab/metab Rev.
1&2:107-123. 1986.
Howell SL. Role of ATP in the intracellular translocation of proinsulin and insulin. Nature
New Biol. 235:85-87. 1972.
Howell SL, Jones PM and Persaud SJ. Regulation of insulin secretion: the role of second
messengers. Diabetologia 37 (Supp 2): S30-S35. 1994.
Howey DC, Bowsher RR, Brunelle RL. Lys(B28),Pro(B29)-human insulin: A rapidly
absorbed analogue of human insulin. Diabetes: 43: 396-402. 1994.
Insel PA, Liljenquist JE, Tobin JD, Sherwin RS, Watkins P, Andres R and Berman M.
Insulin control of glucose metabolism in man: a new kinetic analysis. J Clin Invest: 55:
1057-1066. 1975.
Ishida T, Chap Z, Chou J, Lewis RM, Hartly Cj, Entman ML and Field JB. Differential
effects of oral, peripheral intravenous and intraportal glucose on hepatic glucose uptake and
insulin and glucagon extraction in conscious dogs. J Clin Invest 72: 590-601. 1983.
148
Ishida T, Chap Z, Chou J, Lewis RM, Hartly CJ, Entman ML and Field JB. Effect of portal
and peripheral venous insulin infusion on glucose production and utilization in
depancreatized conscious dogs. Diabetes 33: 984-990. 1984.
Issekutz B Jr, Issekutz TB, Elahi D and Borkow I. Effect of insulin infusion on glucose
kinetics in alloxan - streptozotocin diabetic dogs. Diabetologia 10: 323-328. 1974.
Janssen JA, Lamberts SW. Circulating IGF-I and its protective role in the
pathogenesis of diabetic angiopathy. Clin Endocrinol (OxQ. 52:1-9,2000.
Jorgensen S, Vaag A, Langkjaer L, Hougard P, Markussen J. Novosol
Basahpharmacokinetics of a novel soluble insulin analogue. BMJ.. 12;299(6696):415-9.
1989.
Juul, S.M., Jones RH, Evans JL, Neffe J, Sonksen PH and Brandenberg D. Evidence for an
early degradative event to the insulin molecule following binding to hepatocyte receptors.
Biochem. Biophvs. Acta 856:310-319.1986.
Kang S, Brenge J, Burch A, Voland A, Owens DR. Subcutaneous insulin absorption
explained by insulin’s physicochemical properties. Evidence from absorption studies of
soluble human insulin and insulin analogues in humans. Diabetes Care 14: 942-948. 1991.
Kang S, Creagh FM, Peters JR et al. Comparison of subcutaneous soluble human insulin
and insulin analogues (AspB9, GluB27, AspBlO, AspB28) on meal-related plasma glucose
excursions in type 1 diabetic subjects. Diab. Care: 14: 571-577. 1991.
Kamovsky MJ. The ultrastructural basis of capillary permeability studied with peroxidase
as a tracer. J Cell Biol. 35:213-236. 1967.
Kasuga M, Karlsson FA and Kahn CR. Insulin stimulates the phosphorylation of the 95000
dalton subunit of its own receptor. Science 215:185-187. 1982a.
149
Kasuga M, Zick Y, Blithe DL, Crettaz M and Kahn CR. Insulin stimulates tyrosine
phosphorylation of the insulin receptor in a cell free system. Nature (Lond) 298: 667-669.
1982b.
Katz H, Homan M Velosa J, Robertson P and Rizza R. Effects of pancreas transplantation
on postprandial glucose metabolism. N Eng J Med. 325: 1275-1283. 1991.
Keefer LM, Piron M-A and De Meyts P. Human insulin prepared by recombinant DNA
techniques and native human insulin interact identically with insulin receptors. Proc. Natl.
Acad Sci USA. 78:1391-1395. 1981.
Kim H, Richardson C, Roberts J, Gren L and Lyon JL. Cold hands, warm heart. Lancet
351:1492. 1998.
Kitabchi A, Jones G, Duckworth W. Effect of hydrocortisone and corticotrophin on glucose
induced insulin and proinsulin secretion in man. J Clin Endocrinol. Metab. 37:79-84. 1973.
Koch CA, Anderson D, Moran MF, Ellis C and Pawson T. SH2 and SH3 domains: elements
that control interactions of cytoplasmic signalling proteins. Science 252:668-674. 1991.
Krebs HA and Johnson WA. The role of citric acid in intermediate metabolism oin animal
tissue. Enzvmologia. 48: 148-156. 1937.
Kruszynska YT, Home PD and Alberti KGMM. Comparison of portal and peripheral
insulin delivery on carbohydrate metabolism in streptozotocin diabetic rats. Diabetologia
28: 167-171. 1985.
Kryshak EJ, Butler PC, Marsh C, Miller A, Barr D, Polonsky K, Perkins JD and Rizza R.
Pattern of postprandial carbohydrate metabolism and effects of portal and peripheral insulinI
delivery. Diabetes 39: 142-148. 1990.
150
Kuerzel GU, Shukle U, Scholtz HE, Pretorius SG, Wessels DH, Venter C, Potgieter MA,
Lang AM, Koose T, Bernhard E. Biotransformation of insulin glargine after subcutaneous
injection in healthy subjects. Curr Med Res Qpin 19GL 34-40. 2003.
Kuhne MR, Pawson T, Lienhard GE, and Feng GS. The insulin receptor substrate 1
associated with the SH2-containing phototyrosine phosphatase SYP. J Biol. Chem.
268:11479-11481. 1993.
Kurtzhals P. Engineering predictability and protraction in a basal insulin analogue: the
pharmacology of insulin detemir. Int J Obesity 28(s2): S23-S28.
Kurtzhals P, Havelund S, Jonassen I, Kiehr B, Larsen UD, Ribel U, Markussen J: Albumin
binding of insulins with fatty acids: characterization of the ligand-protein interaction and
correlation between binding affinity and timing of the insulin effect in vivo. Biochem J
312: 725-731,1995.
Kurtzhals P, Havelund S, Jonassen I, Kiehr B, Ribel U, Markussen J. Albumin binding and
time action of acylated insulins in various species. J Pharm Sci. Mar;85(3):304-8. 1996.
Kurtzhals P, Havelund S, Jonassen I, Markussen J. Effect of fatty acids and selected drugs
on the albumin binding of a long-acting, acylated insulin analogue. J Pharm Sci.
Dec;86(12):1365-8. 1997.
Kurtzhals P, L Schaffer, A Sorensen, C Kristensen, I Jonassen, C Schmid and T Trub.
Correlations of receptor binding and metabolic and mitogenic potencies of insulin analogs
designed for clinical use. Diabetes. Vol 49, Issue 6 999-1005.
Kurtzhals P, Ribel U. Action profile of cobalt(III)-insulin. A novel principle of protraction
of potential use for basal insulin delivery. Diabetes. Dec;44(12):1381-5. 1995.
151
Laing SP, Swerdlow AJ, Slater SD, Botha JL, Burden Ac, Waugh NR, Smith AW, Hill RD,
Bingley PJ, Patterson CC, Qiao Z, Keen H. The British Diabetic Association Cohort Study,
II: cause specific mortality in patients with insulin trated diabetes mellitus. Diab Med.
16(6):446-7. 1999.
Lang, H. 1. Standi, E. 2. Roberts, A. 3. Insulin detemir shows favourable weight
development and reduced nocturnal hypoglycaemia compared to NPH over 12 months in
subjects with Type 1 diabetes. Diabetic Medicine Supplement. 20 Supplement 2:105, April
2003.
Langhans W. Role of the liver in the metabolic control of eating: What we know and what
we do not know. Neuroscience and Biobehavioral reviews. 20(1): 145-53.1996.
Lamer J, Cheng K, Scwartz C, Dubler R, Creasy S. Chemical mechanism of insulin action
via proteolytic formation of mediator peptides. Mol Cell Biochem. 40:155-161. 1981.
Lauritzen T, Faber OK, Binder C: Variation in 1251-insulin absorption and blood glucose
concentration. Diabetologia 17: 291-5, 1979.
Lauritzen T, Pramming S, Gale EA, Deckert T, Binder C. Absorption of isophane (NPH)
insulin and its clinical implications. Br Med J (Clin Res EdL Jul 17;285(6336): 159-62. 1982.
Lindholm A. New insulins in the treatment of diabetes mellitus. Best Pract Res Clin
Gastroenterol. Jun;16(3):475-92. 2002.
Lenzen S and Bailey C. Thyroid hormones, gonadal and adrenocortical steroids and the
function of the islets of Langerhans. Enocr. Rev 5:411-434. 1984.
Lepore M, Pampanelli S, Fanelli C, Porcellati F, Bartocci L, Di Vincenzo A, Cordoni C,
Costa E, Bmneti P, Bolli GB. Pharmacokinetics and pharmacodynamics of subcutaneous
injection of long acting human insulin analogue glargine, NPH insulin, and ultralente human
152
insulin and continuous subcutaneous infusion of insulin lispro. Diabetes 49(12):2142-8.
2000.
Levy JC, Brown G, Mathews DR and Turner RC. Hepatic glucose output in humans
measured with labelled glucose to reduce negative errors. Am J Phvsiol. 257: E531-E540.
1989.
Liljenquist JL, Mueller G, Cherrington A, Perry J and Rabinowitz D. Hyperglycaemia per
se (insulin and glucagon withdrawn) can inhibit hepatic glucose production in man. J Clin
Endocrinol. Metab. 48:171-175. 1979.
Luzio SD, Beck P, Owens DR. Comparison of the subcutaneous absorption of insulin
glargine (Lantus) and NPH insulin in patients with Type 2 diabetes. Horm Metab Res.
Jul;35(7):434-8. 2003.
MacCracken J. From ants to analogues, puzzles and promises in diabetes management.
Diabetes. 101 (4) 138-150. 1997.
Macleod KM, Hepburn DA, Frier BM,; Frequency and morbidity of severe hypoglycaemia
in insulin treated diabetic patients. Diab Med 10:238-245,1993.
Maran A, Cranston I, Lomas J et al: Protection by lactate of cerebral function during
hypoglycaemia. Lancet 343:16. 1994.
Mari A. Estimation of rate of appearance in the non-steady state with a two compartmental
model. Am J Phvsiol. 263: E400-E415. 1992.
Marker JC, Hirsch IB, Smith LJ, Parvin CA, Holloszy JO and Cryer PE. Catecholamines in
prevention of hypoglycaemia during exercise in humans. Am J Phvsiol. 260:E705-E712.
1991.
153
Markussen J, Havelund S, Kurtzhals P, Andersen AS, Halstrom J, Hasselager E, Larsen UD,
Ribel U, Schaffer L, Vad K, Jonassen L Soluble, fatty acid acylated insulins bind to albumin
and show protracted action in pigs. Diabetologia. Mar;39(3):281-8. 1996.
MarkussenJ, Hougaard P, Ribel U, Sorensen AR, Sorensen E. Soulble, prolonged acting
insulin derivatives. I. Degree of protraction and crystallizability of insulins substituted in
the termini of the B-chain. ProteinEng. 1(3)205-13. 1987.
Marshall s. Green A, Olefsky JM. Evidence for recycling of insulin receptors in isolated rat
adipocytes. J Biol. Chem. 256:11464-11470. 1981.
Mathews LS, Norstedt G, Palmiter RD Regulation of insulin-like growth factor I gene
expression by growth hormone. Proc Natl Acad Sci USA: 83:9343-9347. 1986.
Matschinsky FM and Bedoya FJ. Metabolism of pancreatic islets and regulation of insulin
and glucagon secretion. In Endocrinol. Vol 2 (Ed De Groot, LJ) Second Edition. WB
Saunders Company, Philadelphia ppl290-1303. 1989.
Mayer J. Bulletin of the New England Medical Center, Volume XIV, April-June 1952: The
glucostatic theory of regulation of food intake and the problem of obesity (a review). Nutr
Rev 49f2L46-48. 1991.
Mayerson HS, Wolfram CG, Shirely HH Jr and Wasserman K. Regional differences in
capillary permeability. Am J Phvsiol. 198:155-160. 1960.
McCarthy DA, Maegawa H, Lee J, Dull TJ, Ullrich A, and Olefsky JM. A mutant insulin
receptor with defective tyrosine kinase displays no biologic activity and does not undergo
endocytosis. J Biol Chem. 262:14663-14671. 1987.
McGregor-Robertson J. The Household Phvsician. A family guide to the preservation of
health and to the domestic treatment of ailments and disease, with chapters on food and
drugs and first aid in accidents and injuries. 2° Edition. 303-304. 1898.
154
McMahon MM, Schwenk WF, Haymon MW, and Rizza RA. Underestimation of glucose
turnover measured with [6-^H]- and [6,6-^H^]- but not [6- " C] glucose during
hyperinsulinaemia in humans. Diabetes 38; 97-107. 1989.
Migdalis IN, Kalogeropoulou K, Kalantzis L, Nounopoulos C, Bouloukos A, Samartzis M.
Insulin like growth factor-1 and IGF-1 receptors in diabetic patients and neuropathy. Diab
Med. 12:823-827. 1995.
Moberg EA, Lins PE, Adamson UK. Variability of blood glucose levels in patients with type
1 diabetes mellitus on intensified insulin regimens. Diabete Metab. Nov-Dec;20(6):546-52.
1994.
Mokdad Ah, Ford ES, Bowman BA. Diabetes trends in the US: 1990-1998. Diabetes Care
23(9): 1278-1283. 2000.
Muggeo M, Bolli G, Bompiani G, Brunetti P, Capani F, Cavallo-Perin P, Comaschi M,
Cordera R, Crepaldi G, Cucinotta D, Di Mario U, Fedele D, Ferrannini E, Giorgino R,
Giugliano D, Gnudi A, Mancini M, Menzinger G, Pagano G, Pontiroli AE, Pozza G,
Santeusanio F, Tiengo BA, Trovati M, Vigneri R. Glycémie control and cardiovascular
diseases in Type 2 diabetes mellitus. Beyond fasting glycemia and glycosylated hemoglobin.
Diabetes Nutr Metab. Aug: 13('4L 182-5. 2000.
Naito M and Wisse E. Filtration effect of endothelial fenestrations on chylomicron transport
in neonatal rat liver sinusoids. 1978.
Neely RDG, Rooney DP, Atkinson AB, Sheridan B, Ennis CN, Trimble ER and Bell PM.
Underestimation of glucose turnover determined using [6-3H] glucose tracer in non-steady
state: the role of a tritiated tracer impurity. Diabetologia 33:681-687. 1990.
155
Nilsson LH and Hultman E. Liver glycogen in man - the effect of total starvation or a
carbohydrate poor diet followed by carbohydrate refeeding. Scand. J. Clin. Lab. Invest
32:325-330. 1973.
Nolan C, Margoliash E, Peterson JL and Steiner DF. The structure of bovine proinsulin. J
Biol Chem 246:2780-2795. 1971.
Norwich KH, Radziuk J, Lau D and Vranic M. Experimental validation of non-steady state
rate measurements using a tracer infusion method and insulin ans tracer and tracee. Can J
Phvsiol Pharmacol. 52 508-521. 1974.
Nuqahan N, Campbell P, Kennedy F, Miles J, Gerich J. Insulin dose response
characteristics for suppression of glycerol release and conversion to glucose in humans.
Diabetes 35: 1326-1331. 1986.
Olsen HB, Kaarsholm NC. Structural effects of protein lipidation as revealed by LysB29-
myristoyl, des(B30) insulin. Biochemistrv. Oct 3;39(39):11893-900. 2000.
Orci L, Halban P, Amherdt M, Ravazzola M, Vassali JD and Perrelet A. Nonconverted,
amino acid analog-modified proinsulin stays in a Golgi-derived clatherin-coated membrane
compartment. J Cell. Biol. 99: 2187-2192. 1984.
Orci L, Like AA, Amherdt M, Blundell B, Kanazawa Y, Marliss EB, Lambert AE,
Wollheim CB and Renold AE. Monolayer cell culture of neonatal rat pancreas: An
ultrastructural and biochemical study of functioning endocrine cells. J. Ultrastr. Res.
43:279-297. 1973.
Orci L, Ravazzola M, Amherdt M. Conversion of proinsulin to insulin occurs co-ordinately
with acidification of maturing secretory granules. J. Cell Biol. 103:2273-2281. 1986.
156
Orci L, Ruffener C, Malaisse-Lagae F, Blundell B, Amherdt M, Bataille D, Freychet P and
Perrelet A. A morphological approach to surface receptors on islet and liver cells. Israel J
Med Sci. 11:635-655. 1975.
Owens B, Barnett AH. Designer insulins; Have they revolutionised insulin therapy? In:
Betteridge DJ (Ed) Diabetes: Current Perspectives. London: Martin Dunitz; 223-266. 2000.
Owens DR. Humuan insulin. Clinical pharmacological studies in normal man. 1986.
Owens DR, Zinman B, Bolli GB. Insulins today and beyond. Lancet. 358(9283):739-46.
2001.
Owerbach D, Bell GI, Rutter WJ and Shows TB. The insulin gene is located on the short
arm of chromosome 11 in humans. Diabetes 30:267-270. 1981.
Oyer PE, Cho S, Peterson JD and Steiner DF. Studies on human proinsulin. Isolation and
amino acid sequence of the human pancreatic C-peptide. J Biol Chem 246:1375-1386.
1971.
Palade GE. Blood capillaries of the heart and other organs. Circulation 24:368-384. 1961.
Parker JC, Perry MA and Taylor AE. Permeability of the microvascular barrier. In: Staub
NC and Taylor AE (Ed). Edema. Raven Press, New York. 1984.
Patzelt C, Labrecque AD, Duguid JR, Carrol RJ, Keim PS, Heinrikson RL and Steiner DF.
Detection and kinetic behaviour of preproinsulin in pancreatic islets. Proc. Natl. Acad Sci.
USA. 75: 1260-1264. 1978.
Periello G, De Feo P, Torlone E, Fanelli C, Santeusanio F, Brunetti P, Bolli GB. Nocturnal
spikes in growth hormone secretion cause the dawn phenomenon in type 1 (insulin
157
dependent) diabetes mellitus by decreasing hepatic and extrahepatic sensitivity to insulin in
the absence of insulin wasting. Diabetologia. 33:52-59. 1990.
Peterson JD, Coulter CL, Steiner DF, Emdin SO, Falkmer S. Structural and crystallographic
observations of the hag fish insulin. Nature TLond.) 251:239-240. 1974.
Pieber T, Grill Valdemar, Kristensen A, Draeger E. Treatment with Insulin Detemir Allows
Flexible Timing of Administration in Subjects with Type 1 Diabetes. Diabetes. 52
SUPPLEMENT 1 : A130, June 2003.
Pieber T, Grill V, Kristensen A, Draeger E. Timing o f dose-administration of insulin
detemir is flexible and results in lower and less variable plasma glucose levels and less
weight gain than NPH insulin. Diab Med. In Press. 2004.
Pieber TR, Eugene-Jolchine I, Derobert E. Efficacy and safety of HOE901 versus NPH
insulin in patients with Type 1 diabetes. The European Study Group of HOE 901 in Type 1
Diabetes. Diabetes Care 23(2): 157-62. 2000.
Pieber TR, Plank J, Goerzer E, Sommer R, Wutte A, Sinner F, Bodenlenz M, Endahl L,
Draeger E, Zdravkovic M. Duration of Action, Pharmacodynamic Profile and between-
Subject Variability of Insulin Detemir in Subjects with Type 1 Diabetes. Diabetes. 51
SUPPLEMENT 2:A53, June 2002.
Pieber, T. 1. Plank, J. 1. Sommer, R. 1. Goerzer, E. 1. Wutte, A. 1. Sinner, F. 1. Bodenlenz,
M. 1. Endahl, L. 2. Draeger, E. 2. Zdravkovic, M. 2. Protracted pharmacodynamic profile
and high between-subject predictability with insulin detemir in subjects with Type 1
diabetes. Diabetic Medicine Supplement. 20 Supplement 2:105-106, April 2003.
Powrie JK, Smith GD, Shojaee-Moradie F, Sonksen PH and Jones RH. Mode of action of
chloroquine in patients with non insulin dependent diabetes mellitus. Am J Phvsiol. 260: E
897-904. 1991.
158
Polonsky KS, Stuns J, Bell GI. Seminars in medicine of the Beth Israel Hospital, Boston.
Non-insulin dependent diabetes mellitus - a genetically programmed failure of the beta cell
to compensate for insulin resistance. N Engl J Med. 334G2k777-783. 1996.
Pramming S, Thorsteinsson B, Bendston I, Binder C. Symptomatic hypoglycaemia in 411
type 1 diabetic patients. DiabetM ed Apr;8(3):217-22. 1991.
Radzuik J, Norwich KH, and Vranic M. Fed Proc Fed Am Soc Exp Biol 33: 1855-1864.
1974.
Randle PJ. Glucokinase and candidate genes for type 2 (non-insulin dependent) diabetes
mellitus. Diabetologia 36: 269-275. 1993.
Randle PJ, Hales CN, Garland PB and Newsholme EA. The glucose fatty acid cycle. Its
role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Diabetologia
36:269-275. 1963.
Rangasami JJ, Greenwood DC, McSporran B, Small PJ, Patterson CC, Waugh NR. Rising
incidence of Type 1 Diabetes in Scottish children. Arch Dis Child 77:210-213. 1997.
Raskin P, Klaff L, Bergenstal R, Halle JP, Donley D, Mecca T. A 16 week comparison of
the novel insulin analogue insulin glargine (HOE 901) and NPH insulin used with insulin
lispro in patients with type 1 diabetes. Diab Care. 23(11).T666-71. 2000.
Ratner Re, Hirsch IB, Neifing JL, Garg SK, Mecca TE, Wilson CA. Less hypoglycaemia
with glargine in intensive insulin therapy for Type 1 diabetes. US Study group of Insulin
Glargine in Tupe 1 Diabetes. Diab Care. 23(5):639-43. 2000.
Reichard GA, Moury NF, Hochella NJ, Patterson AL, and Weinhouse S. Quantitative
estimation of the Cori cycle in the human. J Biol Chem: 238: 495-501. 1963.
159
Reichard P, Berglund B, Britz A, Cars I, Nilsson BY, Rosenqvist U; Intensified
conventional insulin treatment retards the microvascular complications of insulin-dependent
diabete mellitus (IDDM): the Stockholm Diabetes Intervention Study (SDIS) after 5 years. J
Intern Med 230: 101-108,1991.
Reichen J: The role of the sinusoidal endothelium in liver function. News Phvsiol Sci 14:
117-121,1999.
Renkin EM. Transport of large molecules across capillary walls. Phvsiologist 7:13-28.
1964.
Renkin EM. Lymph as a measure of composition of interstitial fluid. In: Fishman AP and
Renkin EM (ed). Pulmonarv edema. Williams and Wilkins, Baltimore 145-159. 1979.
Revers RR, Henry R, Schmeiser RHL, Kolterman OG, Cohen R, Bergenstal R, Polonsky K,
Japan J, Rubenstein AH, Frank B, Gallway JA and Olefsky JM. The effects of biosyntheitic
human proinsulin on carbohydrate metabolism. Diabetes 33: 762-770. 1984.
Ribel U, Hougaard P, Drejer K and Sorensen AR. Equivalent in vivo activity of insulin
analogues and human insulin despite different in vitro potencies. Diabetes 39: 1033-1039.
1990.
Ritter R, Slusser P and Stone S: Glucoreceptors controlling feeding and blood
glucose:location in the hindbrain. Science 213:451.1981.
Rizza RA, Mandarino LJ and Gerich JE. Cortisol induced insulin resistance in man:
Impaired suppression of glucose production and stimulation of glucose utilization due to a
postreceptor defect of insulin action. J Clin Endocrinol. And Metab. 54:131-138. 1982.
160
Rizza RA, Mandarino LJ, and Gerich JE. Dose-response charachteristics for effects of
insulin on production and utilization of glucose in man. 1981.
Rose DW, Saltiel AR, Majumdar M, Decker SJ and Olefsky JM. Insulin receptor substrate-
1 is required for insulin mediated mitogenic signal transduction. Proc Natl Acad Sci 91:
797-801. 1994.
Rosenberg PL, Lane WS, Backer JM, White MF, Kahn CR. Purification and partial
sequence analysis of ppl85, the major cellular substrate of the insulin receptor tyrosine
kinase. JBiol. Chem 266:8302-8311. 1991.
Rosenstock J, Park G, Zimmerman J; US insulin Glargine (HOE 901) Type 1 Diabetes
Investigator Group. Basal insulin glargine (HOE 901) versus NPH insulin in patients with
type 1 diabetes on multiple daily insulin regimens. US Insulin Glargine (HOE 901). Type 1
Investigator Group. Diabetes Care. 23 (8): 1137-42. 2000.
Rosenstock J, Schwaltz SL, Clark CM Jr, Park GD, Donley DW, Edwards MB. Basal
insulin therapy in type 2 diabetes:28 week comparison of glargine (HOE901) and NPH
insulin. Diabetes Care. 24(4):631-6. 2001.
Ross, M, Kaye GI, Pawlina W. Histology a text and atlas. 4th e d , p540-541. 2003.
Rosskamp RH and Park G. Long acting insulin analogues. Diabetes Care 22 (S2) B109-
B113. 1999.
Rothman DL, Magnusson I, Katz LD, Shulman RG and Shulman GI. Quantitation of
hepatic glycogenolysis and gluconeogenesis in fasting humans with 13C NMR. Science
254:573-576. 1991.
161
Rudolf MCJ, Sherwin RS, Markowitz R, Bates SE, Genel M, Hochstadt J, Tamborlane WV.
Effect of intensive insulin treatment on linear growth in the young diabetic patient. J Pediatr:
101,333-339. 1981.
Russell-Jones D, Draeger E, Simpson R, Bolinder J, Hylleburg B. Effects of once daily
insulin detemir or NPH insulin on blood glucose control in people using a basal bolus
regimen. Clin Ther. 26:724-736. 2004.
Russell-Jones D, Bolinder J, Simpson R, Stades A, Stender A, Hylleberg B, Draeger E.
Once Daily Dosing with Insulin Detemir Offers Advantages Compared to NPH Insulin in
Subjects with Type 1 Diabetes. Diabetes. 52 SUPPLEMENT LA132, June 2003.
Russell-Jones D, Bolinder J, Simpson: Lower and more predictable fasting blood glucose
and reduced risk of nocturnal hypoglycaemia with once daily insulin detemir versus NPH in
subjects with Type 1 diabetes. Diabetologia 45 (Suppl 2): 147A, 2002.
Russell-Jones D, Bolinder J, Simpson: Lower and more predictable fasting blood glucose
and reduced risk of nocturnal hypoglycaemia with once daily insulin detemir versus NPH in
subjects with Type 1 diabetes. Diabetologia 45 (Suppl 2): 147A, 2002.
Russell-Jones DL, Rattray M, Wilson VJ et al . Intraperitoneal insulin is more potent than
subcutaneous insulin at restoring hepatic IGF-1 mRNA levels in the diabetic rat: a functional
role for the portal vascular link. J Mol Endocrinol. 9:257-263. 1992.
Russell-Jones DL, Rattray M, Wilson VJ, Jones RH, Sonksen PH Thomas CR.
Intraperitoneal insulin is more potent than subcutaneous insulin at restoring hepatic insulin
like growth factor-I mRNA levels in the diabetic rat: a functional role for the portal vascular
link. JM ol Endocrinol. Dec:9f3):257-63. 1992.
Ryle AP, Sanger F, Smith LF, and Kitai R. The disulphide bonds of insulin. Biochem. J.
60:541-556. 1955.
162
Schlichtkrull J. Insulin crystals. Munksgaard, Copenhagen. 1956.
Scholtz HE, van Niekirk M, Meyer BH, Rosenkranz B. An assessment of the variability in
the pharmacodynamics (glucose lowering effect) of HOE901 compared to NPH and
ultralente human insulins using the euglyaemic clamp technique. Diabetologia (supple 1):
A 235.1999.
Schwenk WF, Burtler PC, Haymond MW and Rizza RA. Underestimation of glucose
turnover corrected with high-performance liquid chromatography purification of [6-
3H]glucose. Am J Phvsiol: 258: E228-E233. 1990.
Searle GL, Strisower EH and Chaikoff IE. Glucose pool and glucose space in the normal
and diabetic dog. Am J Phvsiol. 176: 190-194. 1954.
Shojaee-Moradie F, Jackson NC, Jones RH, Mallet AI, Hovorka R, Umpleby AM:
Quantitative Measurement of 3-0 Methyl D glucose by gas chromatography mass
spectrometry as a measure of glucose transport in vivo. J Mass Snectrometrv 31: 961-966,
1996.
Simpson HE, Umpleby AM, Russell-Jones DL. Insulin like growth factor-1 and diabetes.
A review. Growth hormone and IGF research. f8L83-95. 1998.
Skeie, S. 1. De Leeuw, I. 2. Vague, P. 3. Selam, J. L. 4. Elte, J. W. F. 5. Lang, H. 6. Draeger,
E. 6. Insulin detemir poses lower risk of nocturnal hypoglycaemia and less weight gain
compared with NPH after 12 months' treatment in subjects with Type 1 diabetes. Diabetic
Medicine Supplement. 20 Supplement 2:27, April 2003.
Somogyi M. Exacerbation of diabetes by excess insulin action. J Med. Feb;26(2): 169-91.
1959.
163
Songer TJ, La Porte RE, Dorman JS, Orchard TJ, Brecker DJ, Drash AL. Health, life, and
automobile insurance charachteristics in adults with IDDM. Diabetes Care. 14(4):318-24.
1991.
Sonksen P and Sonksen J: Insulin: Understanding its action in health and disease. Br J of
Anaesthesia 85:69-79, 2000.
Sonksen PH, Russell-Jones D, Jones RH. Growth hormone and diabetes mellitus. A review
of sixty-three years of medical research and a glimpse into the future? Horm Res. 40(1-
3):68-79,1993.
Spangler RS, Triebwasser KC and Wilson ML. Effect of change on hepatic specific nature
of vesicle encapsulated insulin in conscious dogs. Am J Phvsiol 251 : E591-E596. 1986.
Spangler RS. Selctive insulinization of liver in conscious dogs. Am J Phvsiol. 249: E152-
E159. 1985.
Standi E, Roberts H, Lang H. Long term efficacy and safetyof insulin detmir in subjects
with Type 1 diabetes. Favourable weight development and risk reduction of nocturnal
hypoglycaemia. Diabetes 51 Suppl A l l 5. 2002.
Steele R. Influence of glucose loading and of injected insulin on hepatic glucose output.
Ann N Y Acad Sci 82: 420-430. 1959.
Steele R, Bishop JS, Dunn A, Altszuler N, Rathgeb I and De Bodo C. Inhibition of hepatic
glucose production in the normal dog. Am J Phvsiol. 208(2): 301-306. 1965.
Steele R, Rostami H and Altszuler N. A two-compartmental calculator for the dog glucose
pool in the non-steady state. Fed Proc: 33: 1869-1876. 1974.
164
Steele R, Wall JS, DeBodo RC, Altszuler N. Measurement of size and turnover rate of body
glucose pool by isotope dilution method. Am J Phvsiol: 187:15-24. 1956.
Steiner DF. Cocrystallization of proinsulin and insulin. Nature 243: 528-530. 1973.
Steiner DF and Oyer PE. The biosynthesis of insulin and probable precursoe of insulin by a
human islet adenoma. Proc. Nat. Acad. Sci. USA. 57:473-480. 1967.
Steiner DF, Kemmler W, Tager HS and Peterson JD. Proteolytic processing in the
biosynthesis of insulin and other proteins. Fed Proc 33: 2105-2115. 1974.
Stetten D, Welt ID, Ingle DJ and Morley EH. Rates of glucose production and oxidation in
normal and diabetic rats. JBiol Chem: 192: 817-830. 1951.
Stevenson RW, Parson JA and Alberti KGM. Insulin infusion into the portal and peripheral
circulation of unanaesthetised dogs. Clin Endocrinol 8: 335-347. 1978.
Stevenson RW, Parson JA and Alberti KGM. Comparison of the metabolic responses of
portal and peripheral infusions of insulin in diabetic dogs. Metabolism 30: 745-752. 1981.
Stevenson RW, Parson JA and Alberti KGM. Effect of intraportal and peripheral insulin on
glucose turnover and recycling in diabetic dogs. Am J Phvsiol. 244: E190-E195. 1983.
Sugden MC, Holness MJ and Palmer TN. Fuel selection and carbon flux during the starved
to fed transition. Biochem. J. 263: 313-323. 1989.
Sun XJ, Rothenberg PL, Kahn CR, Backer JM, Araki E, Wilden PA, Cahil DA, Goldstein
BJ and White MF. The structure of the insulin receptor substrate IRS-1 defines a unique
signal transduction protein. Nature 352:73-77. 1991.
165
Sussman KE, LeitnerJW, Draznin B. Cytosolic free calcium concentrations in normal
pancreatic islet cells. Diabetes 36:571-571. 1987.
Taborsky JG Jr, Ahren B, Havel PJ: Autonomic mediation of glucagons secretion during
hypoglycaemia: Implications for impaired alpha cell responses in type 1 diabetes. Diabetes
47:995,1998.
Tamborlane W, Sherwin R, Kovisto V, Hendler R, Genel M, Felig P, Normalisation of
growth hormone and catecholamine response to exercise in juvenile onset diabetic subjects
treated with a portable insulin pump. Diabetes ;24:317-327. 1975.
Tan K, Baxter RC. Serum IGF-1 levels in diabetic patients: the effect o f age. J Clin
Endocrinol. Metab. 63:651-655. 1986.
Tanti, J-F, Gremeaux T, Van Obberghen E and Le Marchand-Brustel, Y. Serine/threonine
phosphorylation of the insulin receptor substrate 1 modulates insulin receptor signalling. J
Biol Chem. 6051-6057. 1994.
The Diabetes Control and Complications Trial Research Group: The effect of intensive
treatment of diabetes on the development and progression of long-term complications in
insulin-dependent diabetes mellitus. N Engl J Med 329:977-986.1993.
The United Kingdon Prospective Diabetes Study Research Group: Intensive blood-glucose
control with sulphonylureas or insulin compared with conventional treatment and risk of
complications in patients with type 2 diabetes. Lancet 352:837-853,1998.
Thiebaud D, Jacot E, De Fronzo RA, Maeder E, Jequier E and Felber J-P. The effect of
graded doses of insulin on total glucose uptake, glucose oxidation and glucose storage in
man. Diabetes 31: 957-963. 1982.
166
Tomkins CV, Brandenburg D, Jones RH and Sonksen PH. Mechanism of action of insulin
and insulin analogues. A comparison of the hepatic and peripheral effects on glucose
turnover and of insulin, proinsulin and three insulin analogues modified at positions A l and
B29. Diabetologia 20: 94-101. 1981.
Tsemg KY and Kalhan SC. Estimation of glucose carbon recycling and glucose turnover
with [U-13C]glucose. Am J Phvsiol. 245: E476-E482. 1983.
Tyko, B and Maxfield, F.R. Rapid acidification of endocytic vesicles containing p2-
macroglobulin. Cell 28:643-651. 1982.
Tyler Fizzell R, Hendrick GK, Biggers DW, Lacy DB, Donahue DP, Green DR, Carr RK,
Williams PE, Stevensen RW, Charrington AD. Role of gluconeogenesis in sustaining
glucose production during hypoglycaemia caused by continuous insulin infusion in
conscious dogs. Diabetes. 37:749-759. 1988.
Ullrich SA, Shine J, Chirgwin J, Pictet R, Tischer E, Rutter WJ, Goodman HM. Rat insulin
genes : construction of plasmids containing the coding sequences. Science 196(4296): 1313-
9. 1977.
Umpleby AM, Sonksen PH. Measurement of the turnover of substrates of carbohydrate and
protein metabolism using radioactive isotopes. Baillieres Clin Endocrinol Metab.
Nov;l(4):773-96.1987.
Vague P, SelamJ-L, Skeie S, De Leeuw I, Elte JWF., Haahr H, Kristensen A and Draeger E.
Insulin Detemir Is Associated With More Predictable Glycémie Control and Reduced Risk
of Hypoglycemia Than NPH Insulin in Patients With Type 1 Diabetes on a Basal-Bolus
Regimen With Premeal Insulin Aspart. Diabetes Care 26:590-596, 2003.
VandeHaar MJ, Moats-Staats BM, Davenport ML, Walker JL, Ketelslegers JM, Sharma BK,
Underwood LE. Reduced serum concentrations of insulin-like growth factor-I (IGF-I) in
167
protein-restricted growing rats are accompanied by reduced IGF-I mRNA levels in liver and
skeletal muscle. J Endocrinol: 130, 305-312. 1991.
Veneman T, Mitrakou A, Mokan M et al: Effect of hyperketonaemia and
hyperlacticacidaemia on symptoms, cognitive dysfunction and counter-regulatory hormone
responses during hypoglycaemia in normal humans. Diabetes 43:1311,1994.
Wang F, Carabino JM, Vergara CM. Insulin Glargine: A systematic review of a long acting
insulin analogue. Clin Ther. 25f6kl541-77. 2003.
Ward GM, Walters JM, Aitken PM, Best JD and Alford FP. Effects of prolonged pulsatile
hyperinsulinaemia in humans. Enhancement of insulin sensitivity. Diabetes: 39: 501-507.
1990.
Wareham NJ. Epidemiology of diabetes.Medicine 30flh l 1-13. 2002.
White MF, Moran R, and Kahn CR. Insulin rapidly stimulates tyrosine phosphorylation of a
M r- 185,000 protein in intact cells. Nature 318:183-186. 1985.
Wisse E. An electron microscope study of the fenestrated endothelial lining of rat liver
sinusoids. J Ultra Str Res 31:125-150. 1970.
Wissig and Williams MC. The permeability of muscle capillaries to microperoxidase. J Cell
Biol. 76:341-359. 1978.
Wissig SL, and Charonis AS. Capillary ultrastructure. In. Staub NC and Taylor AE (ed)
Edema. Raven Press, New York. 1984.
Wissig SL. Identification of the small pore in muscle capillaries. Acta Phvsiol. Scand
[Suppl 1], 463:33-44. 1979.
168
Wolfe RR. Tracers in Metabolic Research, Laboratory and Research Methods in Biology
and Medicine. Volume 9. Alan R Liss Inc, New York, 1984.
Wood Sp, Blundell TL, Woolmer A, Lazarus NR and Neville RW. The relation of
conformation and association of insulin receptor binding; X-ray and circular dichroism
studies on bovine and histricomorph insulins. Eur. J. Biochem. 55:531-542. 1975.
Wurzburger MI, Prelevic GM, Sonksen PH, Wheeler M, Balint-Peric L. Effect of
recombinant human growth hormone treatment on insulin-like growth factor
(IGF-I) levels in insulin-dependent diabetic patients. Acta Diabetol. 32:131-4,1995.
Yip CC, Yeung CWT, Moule ML. Photoaffinity label of insulin receptor proteins of liver
plasma membrane preparations. Biochemistry. 19:70-76. 1988.
Yipe CC, Yeung CWT, Moule ML. Photoaffinity labelling of insulin receptor of rat
adipocyte plasma membrane. JBiol Chem. 253: 1743-1745. 1978.
Yki-Jarvinen H, Westerbacka J. Vascular actions of insulin in obesity. J Obes Relat Metab
Disord. Jun;24 Suppl 2:S25-8. 2000.
169