Analogos de Insulina en DM 1 2017

REVIEWS
Insulin analogues in type1 diabetes

mellitus: getting better all the time
Chantal Mathieu, Pieter Gillard and Katrien Benhalima
Abstract | The treatment of type1 diabetes mellitus consists of external replacement of the
functions of cells in an attempt to achieve blood levels of glucose as close to the normal
range as possible. This approach means that glucose sensing needs to be replaced and levels
of insulin need to mimic physiological insulin-action profiles, including basal coverage and
changes around meals. Training and educating patients are crucial for the achievement of
good glycaemic control, but having insulin preparations with action profiles that provide
stable basal insulin coverage and appropriate mealtime insulin peaks helps people with type1
diabetes mellitus to live active lives without sacrificing tight glycaemic control. Insulin
analogues enable patients to achieve this goal, as some have fast action profiles, and some
have very slow action profiles, which gives people with type1 diabetes mellitus the tools to
achieve dynamic insulin-action profiles that enable tight glycaemic control with a risk of
hypoglycaemia that is lower than that with human short-acting and long-acting insulins.
This Review discusses the established and novel insulin analogues that are used to treat
patients with type1 diabetes mellitus and provides insights into the future development of
insulin analogues.
Type1 diabetes mellitus (T1DM) is an autoimmune dis- portal system (primarily targeting the liver), and that
order in which insulin-producing cells are destroyed no feedback or suppression of insulin release is possi-
by the immune system. Secretion of insulin (which ble when levels of glucose fall, which increases the risk
controls the metabolism of carbohydrates, proteins of hypoglycaemia1,2.
and lipids) is tightly regulated by feedback systems that Insulin preparations have come a long way since
enable stable control of metabolism, thus preventing the discovery of insulin, from purified animal insulins
hypoglycaemia, hyperglycaemia, protein catabolism, to human insulins produced by genetically modified
lipolysis and the formation of ketone bodies1,2. organisms to insulin analogues that enable an improved
Basal insulin secretion maintains metabolism in fit between insulin-action profiles and glucose excur-
an anabolic state. Upon food intake, cells, driven by sions (that is, fluctuations in levels of glucose). In this
direct sensing of glucose through glucose transporter Review, we detail the established and novel insulin ana-
type2 (GLUT2) receptors on their surface (as well as logues that are used to treat T1DM, and provide insights
by neural signals and incretin signalling), release insulin into future developments of insulin analogues.
into the blood to promote the uptake of carbohydrates,
proteins, peptides and lipids into other cells. The effect From animal to human insulin
of insulin on peripheral glucose uptake in muscle and The human insulins come in different types: rapid-
the rapid inhibition of gluconeogenesis and glyco- acting (regular) insulin, slow-acting neutral protamine
genolysis in the liver result in a decrease in blood levels Hagedorn (NPH) insulin or zinc-based insulin. Most
of glucose, which causes cells to stop synthesizing and patients in the 1980s were treated with mixtures of regu-
Clinical and Experimental
secreting insulin1,2. lar and NPH or zinc-based insulin (mixed by the patients
Endocrinology, University
of Leuven, Herestraat49, A major challenge of insulin replacement in patients or in a premix form) that were administered twice daily
Leuven3000, Belgium. with T1DM is mimicking the insulin-action profiles of (before breakfast and dinner)2.
Correspondence to C.M. cells maintaining basal levels and achieving peak The Diabetes Control and Complications Trial
[email protected] levels at mealtimes. The two major weaknesses of exter- (DCCT) demonstrated that intensive insulin therapy
doi:10.1038/nrendo.2017.39 nal insulin replacement are that insulin is adminis- resulting in tight glycaemic control prevented the micro-
Published online 21 Apr 2017 tered peripherally, whereas cells secrete insulin in the vascular complications that are associated with T1DM3,4.
NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 1
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
Key points the concentration of regular insulin also influences the

onset and duration of action of the preparation. As such,
Established rapid-acting and long-acting insulin analogues have enabled more hyperconcentrated U500 regular insulin has an action
patients with type1 diabetes mellitus to reach better glucose targets, with lower profile that is right-shifted compared with that of regular
hypoglycaemia rates and a better quality of life than was possible with short-acting insulin, with a delayed onset of action and a duration of
and long-acting human insulin
action of 610 h (REF.6).
In patients who are prone to severe hypoglycaemia, using a full analogue regimen is
rapidly cost saving and should therefore be the standard of care in all patients with
Zinc and NPH insulins
type1 diabetes mellitus
Zinc and NPH insulins are formed by the addition
The new long-acting insulin analogues insulin glargine U300 and insulin degludec
of zinc or protamine, respectively, to regular insulin,
have shown increased stability, which translates to a reduced risk of nocturnal
hypoglycaemia and increased flexibility in timing of administration
which results in lumps where the insulin molecules are
linked to these substances (resulting in an inhomogene-
Faster and shorter acting insulin analogues are needed for use in insulin pumps and
future artificial pancreas systems; fast-acting insulin aspart, a new formulation of
ous suspension in vials of the substances), which causes
aspart, is well advanced in clinical development their action profile to be prolonged. The major drawback
of these insulins is the variability in their action pro-
file, which is partially attributable to the need for resus-
These findings established intensive insulin therapy, using pension of the insulin in the vial before it is injected;
regular insulin administered before each meal and basal however, variability is still present under fully controlled
insulin administered at bedtime, as the gold standard laboratory conditions of resuspension7. In subcutaneous
in the treatment of people with T1DM3,4. tissue, regular insulin hexamers are released from the
The DCCT also highlighted the limitations of insu- zinc or NPH depots in a stochastic way over several
lin replacement therapy in T1DM, in particular when hours, which causes a highly variable insulin-release
using human insulins: patients who underwent inten- action profile, with durations of action ranging from a
sive insulin therapy had a threefold increase in the risk few hours to more than 24 h (REF.8) (FIGS1,3). Variation
of hypoglycaemia, particularly of severe hypoglycaemia. in duration of action is seen in these insulins, as well as
Moreover, intensively treated patients had a 30% higher variation in the strength of the insulin action and their
risk of becoming overweight than non-intensively treated action profiles, with peak levels of release sometimes
patients. These adverse effects are probably due to the happening soon after injection (causing early nocturnal
absence of negative feedback on insulin release from sub- hypoglycaemia and necessitating snacks before bed-
cutaneous depots once insulin is injected but could also time). Another major limitation of these insulins is that
be attributable to the mismatch between insulin-action they do not cover the basal insulin needs for the full 24 h
profiles of the human preparations and mealtime or basal in many patients8,9 (FIG.3).
levels, which does not occur with the action profile of
functioning cells in a healthy person3 (BOX1). First-generation insulin analogues
In the past 20years, rapid-acting and long-acting insulin
Regular insulin analogues have been designed to mimic the action pro-
Regular insulin has the same structure as insulin produced files in insulin secretion of pancreatic cells more closely
by cells: six monomers of insulin, each of which con- than previous preparations1,2 (BOX2).
sists of an Achain and a Bchain linked by two disulfide
bridges (with an additional disulfide bridge between Rapid-acting insulin analogues
two amino acids in the Achain) that are positioned The first rapid-acting insulin analogues were designed
around a zinc ion and form a hexamer. When injected to create less-stable insulin hexamers, creating insulins
into the bloodstream, these hexamers immediately dis- that would more readily become monomeric or would
sociate into monomers and are able to interact with the even be monomeric in solution, thus moving into the
insulin receptor on target tissues, which means that bloodstream more rapidly after subcutaneous injection
the glucose-lowering effect of intravenous regular insu- than human regular insulin. This action profile enables
lin is almost immediate. However, when regular insulin a shortening of the time between injection and start of
is injected into subcutaneous tissue, the hexamers must the meal, and thus provides a better match than human
dissociate into monomers before resorption into the regular insulin between the insulin-action profile and
bloodstream can happen (FIG.1). Thus, a delay in onset the glucose excursion that is caused by the meal10. Three
of action of the glucose-lowering effect occurs with rapid-acting insulin analogues are available for clinical
subcutaneous injection of regular insulin (depending use in Europe and the USA: insulin lispro, insulin aspart
on several factors, such as the site of injection, blood and insulin glulisine.
flow and temperature), which causes variability in action
and a mismatch between the insulin-action profile that Insulin lispro. The molecular structure of insulin lispro
results from injecting regular insulin immediately before differs from that of regular human insulin by a switch
a meal and the glucose excursion caused by the meal in the order of proline and lysine at residues 28 and 29
(FIG.2). On the basis of this mismatch, patients who use of the Bchain5,11 (FIG.4a). This change destabilizes hexa-
regular insulin as their mealtime insulin are advised to merization, and dissociation into dimers and mono-
allow 1530 min between the injection and the start of mers occurs swiftly, which enables an uptake through
the meal, which is inconvenient in daily life2,5. Of note, the blood vessels that is more rapid than that of human
2 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
regular insulin5 (FIG.1). Pharmacokinetic studies show to that of insulin lispro, with postprandial levels of glu-
that the peak plasma concentration of insulin lispro in cose that are notably lower than those achieved with
the first hour after injection is twofold higher than that regular human insulin1923.
of human regular insulin and that the time to maximum
concentration of insulin lispro is less than half that of Insulin glulisine. The molecular structure of insulin glu-
human regular insulin5,11 (FIG.2). The concentration lisine differs from that of human regular insulin by the
of insulin lispro decreases to levels <20% of peak con- replacement of asparagine with lysine and of lysine with
centrations 4 h after injection, whereas absorption of glutamic acid at residue 3 and residue 29 of the Bchain,
human regular insulin is still ongoing at this point 5,11. respectively 5,24 (FIG.4a). In contrast to insulin lispro and
Taken together, these pharmacokinetic data show that insulin aspart, the formulation of insulin glulisine con-
regular insulin and insulin lispro have a similar area tains polysorbate20 instead of zinc24. Whereas insulin
under the curve, but the curve is shifted to the left for lispro and insulin aspart are stable in subcutaneous pump
insulin lispro. Pharmacodynamic data show that, com- catheters, issues of clotting and catheter obstructions
pared with the administration of human regular insulin are more frequent with insulin glulisine, which makes
at the same time, insulin lispro leads to a lower postpran- insulin lispro and insulin aspart the preferred insulin
dial glycaemic peak, shorter time to peak and lower total analogues for use in pumps25. In most pharmacokinetic
glucose excursion for the 04 h period. These character- pharmacodynamic studies that compared insulin glulis-
istics enable insulin lispro to be injected within 15 min ine with insulin lispro or insulin aspart, insulin glulisine
of starting a meal5,11. In contrast to human regular insu- had a slightly faster onset of action than the other ana-
lin, the site of injection of insulin lispro is less impor- logues24,26,27. The faster onset of action of insulin glulisine
tant with regard to speed of onset of action. However, was particularly noticeable in patients with obesity 28,29.
for rapid-acting insulin analogues, abdominal wall The zinc-free formulation of insulin glulisine might be
injections are advised, as the absorption from deltoid the reason behind the faster onset of action24,26, as zinc
and femoral administrations is slower than that from might delay the absorption and action of insulin lispro
abdominal administration and results in an increased and insulin aspart by slowing down the dissociation into
duration of action for both regular insulin and insulin monomers after injection5 (FIG.1).
lispro according to pharmacokinetic and pharmaco-
dynamic data5,12. The faster absorption and faster onset Rapid-acting insulin analogues in clinical trials and
of the glucose-lowering effect of rapid-acting insulin real life. Overall, clinical studies in patients with T1DM
analogues lead to postprandial levels of glucose that are that have compared first-generation rapid-acting insu-
considerably lower than those triggered by mealtime lin analogues with human regular insulin at mealtimes
administration of human regular insulin5,1214. show minor improvements in levels of HbA1c, with a
reduced risk of hypoglycaemia, particularly severe noc-
Insulin aspart. The molecular structure of insulin turnal hypoglycaemia5,13,14,30,31. However, a major issue
aspart differs from that of human regular insulin by with these trials is the fact that they were carried out in
the replacement of proline with aspartic acid at resi- the absence of good basal insulin optimization, mostly
due 28 of the Bchain5,15,16 (FIG.4a). The pharmaceuti- because basal insulin analogues were not available when
cal formulation of insulin aspart similar to that of the trials were carried out. This weakness points to the
insulin lispro contains glycerine, metacresol, zinc importance of having both good mealtime and basal
and phenol, and has disodium hydrogen phosphate as coverage in T1DM. Later studies of novel insulin ana-
the buffer 5. The pharmacokineticpharmacodynamic logues have taken more care in optimizing basal insulin
action profile of insulin aspart is similar to that of insu- therapy before introducing the new mealtime analogue32.
lin lispro15, and most studies show that insulin lispro Using the Diabetes Treatment Satisfaction
and insulin aspart have similar effects on lowering levels Questionnaire, patients with T1DM considered treat-
of glucose without a difference in the time to maximum ment with a rapid-acting insulin analogue at mealtime
concentration of insulin5,17,18; thus, insulin aspart can to enable more flexibility and noted the reduced risk of
also be injected 15 min before the start of a meal5,15. hypoglycaemia as an asset13 (BOX3).
The clinical action profile of insulin aspart is also similar A Cochrane Database systematic review published in
2016 considered only a few of the clinical studies that
were carried out using these analogues as valuable and
Box 1 | The need for tight glucose control in type1 diabetes mellitus showed that rapid-acting insulin analogues induced only
Intensive insulin therapy that results in tight glucose control can prevent the
a minor reduction of 0.15% in levels of HbA1c compared
microvascular complications of diabetes in people with type1 diabetes mellitus with regular human insulin in patients with T1DM31. No
(T1DM). Studies such as the Diabetes Control and Complications Trial have established differences in the frequency of hypoglycaemia were seen
intensive insulin therapy as the gold standard in the treatment of patients with T1DM with insulin lispro or insulin glulisine compared with
but also demonstrated the most important shortcomings of using exogenous insulin in human regular insulin31. The Cochrane review also found
striving for strict glucose control: that is, the risk of hypoglycaemia (particularly severe no clear evidence for a substantial effect of these rapid-
and nocturnal hypoglycaemia) and weight gain. The advent of insulin analogues has acting insulin analogues on health-related quality of life
had a major effect on patient care, mainly on reducing the risk of hypoglycaemia. or weight gain31. In the few head-to-head clinical trials
However, achieving normoglycaemia in a safe way remains a challenge in patients with that compared the different rapid-acting insulin ana-
T1DM, even with the newest insulin analogues.
logues, no clinically significant differences were seen in
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
Degludec Insulin glargine Insulin glargine Insulin Human Human Insulin lispro, Fast-acting
U300 U100 detemir NPH insulin regular aspart and insulin aspart
insulin glulisine
In the vial
Dihexamers Hexamers Hexamers Dihexamers Hexamer Hexamers Hexamers and Hexamers

protamine multimers
A er administration
under the skin
Multihexamer Concentrated Hexamer Dihexamers Hexamer Hexamers, Hexamers, Hexamers,

(>5,000 kDa) hexamer aggregates albumin protamine dimers and dimers and dimers and
aggregates monomers monomers monomers
In the capillaries
absorption
Rate of
Very slow Very fast

Duration
of action
Long Short
Arginine and Insulin

Albumin Polysorbate 20 Protamine Zn
nicotinamide monomer
Figure 1 | Different determinants of absorption and duration of action absorption rate and prolongs the duration of action. Neutral protamine
of human and analogue insulins. Degludec forms weak hexamers in Hagedorn (NPH) insulin co-crystalizes with protamine, both in the
solution in the vial and stable multihexamers after administration at the pharmaceutical preparation and at the injection site, slowing absorption
injection depot, thereby slowing its absorption. Reversible binding to and action. The classic rapid-acting insulin analogues (lispro, aspart and
albumin in the circulation further prolongs its action. Insulin glargine U300 glulisine) dissociate into dimers and monomers more rapidly than does
precipitates at physiological pH, forming compact aggregates at the human regular insulin, causing a more rapid absorption and shorter
injection depot, leading to a reduced surface area from which absorption duration of action. For glulisine, polysorbate20 is used as a stabilizing agent,
can occur, causing slow absorption and prolonged duration of action. and formation of hexamers is prevented by absence of zinc (Zn). More rapid
Insulin glargine U100 also precipitates at physiological pH but is less absorption and earlier action of fast-acting insulin aspart is caused by
compact than insulin glargine U300. Insulin detemir forms weak dihexamers addition of arginine and nicotinamide to the formulation, thereby
in the vial and strong dihexamers at the injection depot. Reversible binding increasing the rate of formation of monomers at the injection depot
to albumin, both at the injection depot and in circulation, further slows the and increasing the rate of absorption.
glycaemic control or in the frequency of hypoglycaemia caused by taking human regular insulin at the evening
between these analogues1921. In particular, despite the meal). Translation into lowering levels of HbA1c was not
somewhat faster onset of activity of insulin glulisine, no spectacular, but quality of life of patients with T1DM
differences in clinical efficacy have been seen between improved with the use of rapid-acting insulins that
insulin glulisine and the other insulin analogues24,2628. enabled injections closer to meals and were less disrup-
Despite the sometimes disappointing data coming tive in daily life13,22,33. Studies have shown that patients
from individual trials and meta-analyses, the clinical have a personal preference for rapid-acting insulin
impact of these rapid-acting insulin analogues in real life analogues over human regular insulin22,33. Responses
has been dramatic, mainly because they prevent post- to the Diabetes Treatment Satisfaction Questionnaire
prandial hyperglycaemia and late hypoglycaemia (espe- showed that patients with T1DM perceived therapy with
cially hypoglycaemia in the early night period, which is insulin analogues to be more flexible and reduced the
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
40 involved patients carrying pumps administering insulin

35
lispro, insulin aspart and insulin glulisine, occlusions
Regular insulin
were rare, and the incidence was similar for the three
Serum levels of insulin (IU/l)
30 Rapid-acting insulin aspart rapid-acting insulin analogues in the first 72 h; however,

Insulin aspart, insulin lispro after this time, the incidence of occlusions increased
25 and insulin glulisine
substantially, particularly with insulin glulisine40. This
20 result might be due to the fact that insulin glulisine
has lower physicochemical stability than the other two
15
rapid-acting analogues. A systematic review published in
10 2013 that aimed to determine the stability and perfor-
mance of rapid-acting insulin analogues in CSII in outpa-
5
tients also concluded that the risk of occlusion is higher
0 with insulin glulisine than with the two other rapid-
0 120 240 360 480 acting insulin analogues when the infusion duration
Time (min)
extends beyond approximately 3days25, which means
Figure 2 | Pharmacokinetic action profiles of rapid-acting Nature Reviews
insulins. The| Endocrinology
schematic that insulin aspart and insulin lispro are currently the
shows the pharmacokinetic action profiles of rapid-acting insulins7,102. Compared with preferred rapid-acting insulin analogues for use in
human regular insulin, insulin aspart, insulin lispro and insulin glulisine have a faster people with T1DM usingCSII.
onset of action, a higher peak level and a shorter duration of action. Compared with the
other rapid-acting insulin analogues, the curve for rapid-acting insulin aspart is shifted to
Long-acting insulin analogues
the left, with a similar area under the curve, but a faster onset and earlier peak. Care must
be taken when interpreting the curves, as experimental settings in which the data were
The first long-acting insulin analogues, insulin glargine
gathered differed between studies. IU, international units. and insulin detemir, were designed to provide more
stable basal insulin-action profiles and longer, as well
as better, 24 h coverage of the insulin needs of patients
perceived risk of hypoglycaemia compared with ther- compared with human long-acting insulins. Although
apy with human regular insulin13. In addition, the use of insulin glargine and insulin detemir are very different
rapid-acting insulin analogues in patients with T1DM basal insulins, they show dramatic improvements in var-
is cost-effective34,35. Thus, these rapid-acting insulin iability in their insulin-action profiles and duration of
analogues have become the standard of care in people action compared with NPH insulin7,41,42. Clinically, this
with T1DM36. In the clinic, choices of one analogue over improvement in variability translates to an important
another are mostly driven by the characteristics of the reduction in the risk of nocturnal hypoglycaemia; how-
tools they come with (for example, ease of use of pen ever, no improvements in levels of HbA1c have been con-
or the availability of concentrated forms), their compat- sistently observed4347. The introduction of these basal
ibility with continuous subcutaneous insulin infusion insulins has had an effect on the quality of life of patients
(CSII) or approval in special populations. with T1DM, particularly because of the reduction in the
risk of nocturnal hypoglycaemia1 (BOX3).
Rapid-acting insulin analogues in special popula-
tions. Insulin lispro and insulin aspart have similar Insulin glargine. Insulin glargine, currently the most fre-
effects on metabolic control and pregnancy outcomes quently prescribed long-acting insulin analogue, was the
as human regular insulin and are approved for use in first basal insulin analogue approved for clinical use and
pregnancy 37,38. A large randomized controlled trial that its mechanism of protracted action is precipitation in the
included 322 patients has shown that insulin aspart is at subcutaneous tissue, which forms aggregates that lead
least as safe and effective as human regular insulin when to long term release48. This precipitation only happens
used in basal-bolus therapy with NPH insulin in preg- at neutral pH, whereas, at acid pH (in the vial), insu-
nant women with T1DM and provides benefits in terms lin glargine is soluble49,50. This feature was achieved by
of postprandial glucose control and prevention of severe shifting the isoelectric point to pH6.7 through the addi-
hypoglycaemia38. The safety and efficacy of insulin glu- tion of two arginine molecules to the amino terminus
lisine in pregnancy have not been investigated in large of the Bchain. In addition, a substitution of asparagine
clinical trials; therefore, insulin glulisine is not approved with glycine at residue 21 of the Achain was introduced
for use in pregnancy. (FIG.4b). Clinically, insulin glargine has a time-action
All rapid-acting insulin analogues are approved profile that is not only longer but also flatter than that
for use in children, but the minimum age limits vary of NPH insulin43,51 (FIG.3). The mean duration of insu-
depending on available data from studies and regula- lin glargine action is 2224 h under single-dose condi-
tory approval. Insulin lispro has no minimum age limit, tions7,49 and 2425.6 h under steady-state conditions50,52.
whereas the minimum age limit is 2years for insulin Consequently, once-daily dosing is effective in most, but
aspart and 6years for insulin glulisine39. not all, patients8,44,53.
All three rapid-acting insulin analogues are approved Insulin glargine has an action profile that is less
for use in CSII. However, studies using insulin glu- variable than that of NPH insulin7, which translates to
lisine indicate that this analogue is less stable and has a decrease in the risk of hypoglycaemia in clinical tri-
a higher occlusion rate of catheters than the other two als, in particular nocturnal hypoglycaemia45, whereas
rapid-acting analogues25,40. In a laboratory setting that head-to-head studies fail to demonstrate superiority in
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
of insulin glargine (particularly under steady-state con-

5 NPH insulin (1216 h) Insulin glargine U300 (~32 h) ditions); however, the duration of action of both insulins
Insulin glargine U100 (~24 h) Insulin degludec approaches 24 h in most, but not all, patients with T1DM59.
GIR (mg/kg per min)
4
Insulin detemir (~2024 h) U100 and U200 (~42 h) In clinical studies in patients with T1DM57, and in real-
life settings, this shorter duration of action translates to
3 a higher proportion of people using twice-daily insulin
detemir compared with insulin glargine to maintain
2 full basal coverage46.
A large-scale, repeated-clamp study compared within-
1 patient variability in the glucose-lowering response
from injection to injection in adults with T1DM using
NPH insulin, insulin glargine or insulin detemir as basal
0
0 2 4 6 8 10 12 14 16 18 20 22 24 insulin7. The lowest variability was reported for insulin
Time (h) detemir, which was fourfold and twofold more stable
Nature Reviews | Endocrinology than NPH insulin and insulin glargine, respectively 7.
Figure 3 | Pharmacodynamic action profiles of long-acting insulins. The schematic
Similar data were reported in children60.
shows the pharmacodynamic action profiles of long-acting insulins in steady state9,75,78.
The pharmacokinetic action profiles of these insulins cannot be compared because the Clinically, studies that investigate insulin detemir in
acylated insulins (insulin detemir and insulin degludec) are mostly bound to albumin and patients with T1DM report a notable decrease in overall
total concentrations of insulin do not yield helpful information on action profiles. and nocturnal hypoglycaemia but no minimal reduc-
Therefore, this figure shows the pharmacodynamic action profiles from studies that were tion in levels of HbA1c compared with NPH insulin57,61.
carried out in steady state, as these yield the most useful information for the clinician Meta-analyses show an overall small (about 0.20.4%)
using these insulins on a daily basis in patients with type1 diabetes mellitus. The but notable reduction in levels of HbA1c in patients
difference in action profile between a single injection and steady state is particularly treated with detemir compared with patients treated
important for those insulins with half-lives above 12 h (insulin glargine U300 and insulin with NPH62,63. A 52-week study that compared insu-
degludec). Care must be taken when interpreting the curves, as experimental settings in lin detemir and insulin glargine showed similar rates
which the data were gathered differed between studies. GIR, glucose infusion rate; NPH,
of hypoglycaemia57, whereas a 26-week study showed
neutral protamine Hagedorn.
less nocturnal and severe hypoglycaemia in patients
using twice-daily insulin detemir than in patients using
lowering levels of HbA1c (REF.43,45). Insulin glargine and once-daily insulin glargine56. Therefore, in patients
NPH insulin have similar effects on weight, generally with T1DM who have a high risk of nocturnal hypo-
resulting in weight gain43,45. glycaemia, insulin detemir might be the preferred
long-acting insulin analogue.
Insulin detemir. Insulin detemir is a pH-neutral, solu- Intriguingly, in all studies that compared insulin
ble basal insulin analogue in which the threonine has detemir with NPH insulin, less weight gain and even
been removed from residue 30 of the Bchain of human a small weight loss were seen in patients treated with
regular insulin and a 14-carbon myristoyl fatty acid has insulin detemir 62,63. This difference was not observed in
been added to lysine at residue 29 of the Bchain, which the head-to-head studies of insulin detemir and insulin
facilitates self-association of insulin detemir molecules glargine56,57. Of interest, the weight advantage was great-
into dihexamers at the injection site and reversible bind- est in those patients who received insulin detemir once
ing to albumin in tissue and the bloodstream (FIG.4b). daily 64. The reasons for this relative reduction in weight
The formation of dihexamer and the binding to albu- gain are not understood but might be related to a slight
min are some of the mechanisms that are involved in the hepato-preferential effect65 or satiety effects on the central
increased length of action of insulin detemir; however, nervous system66.
we do not completely understand how insulin detemir
has a long duration of action54 (FIG.1). The concentra- Long-acting insulin analogues in special populations
tion of insulin in insulin detemir is four times higher and real-life settings. Most studies of insulin glargine
than that of human regular insulin because its molar in pregnancy are small and retrospective, and include
potency is lower than that of human regular insulin and women with T1DM, T2DM and gestational diabetes;
other insulin analogues41,55,56. This feature has few clin- however, a systematic review and meta-analysis of these
ical implications, as the required doses are only slightly studies found no safety issues67. Therefore, regulatory
higher than those of NPH insulin or insulin glargine56,57. bodies allow continuation of insulin glargine during
The mean duration of action of insulin detemir is pregnancy if required to achieve desired glycaemic
21.5 h in patients with T1DM (which is slightly shorter control36,68. A head-to-head study of insulin detemir
than that of insulin glargine) on the basis of data obtained versus NPH insulin has reported on the safety of insulin
using a dose of 0.4units perkg in a single-dose clamp detemir in pregnant women with T1DM69. The study
study 8,58 (FIG.3). The studies evaluating the duration of was too small to enable conclusions on fetal outcomes,
effect of basal insulins are heavily debated, and subtle dif- but fasting plasma levels of glucose improved with insu-
ferences in technique of clamping, as well as differences in lin detemir without an increased incidence of hypo-
single-dose versus steady-state studies, add to the confu- glycaemia, which supports the use of insulin detemir
sion42. Even so, looking at the overall data, insulin detemir as the long-acting insulin analogue of choice in preg-
seems to have a duration of action that is shorter than that nancy 69. Considering the importance of tight glycaemic
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
control in women with T1DM from the first days of As rapid-acting insulins have a start of action that
pregnancy to the moment of delivery, the insulin ana- is too slow, particularly for patients who use pumps,
logue with the action profile that best fits the insulin improvements are still needed. This shortcoming has
needs of the patient has to be chosen by the clinician become obvious with the introduction of algorithms for
in charge. Insulin glargine is approved by the FDA for dose adaptation by smart insulin pumps and with the
use in children from 6years of age and by the European development of closed-loop or semi-closed-loop systems.
Medicines Agency (EMA) from 2years of age, and insu- In contrast to early suggestions from pharmacokinetic
lin detemir is approved for use in children from 2years pharmacodynamic studies, the current rapid-acting
of age (these age limits are the result of clinical study data insulins still take at least 10 min to start their glucose-
and regulatory approval)70,71. lowering action, thus still forcing patients to wait at
The choice of which long-acting insulin analogue least a few minutes between injections and starting the
to use in patients with T1DM in special populations is meal for full matching of the insulin-action profile to
driven by the action profile and the regulatory approval meal-induced glucose excursions75,76. The use of sensor
in that particular group of patients but also by the tools technology has been particularly helpful in uncovering
it comes with (such as the pen it comes with or the avail- this shortcoming 75,76, which has created a demand for
ability of more concentrated forms). As most patients faster-acting and shorter-acting insulin analogues, espe-
with T1DM will require twice-daily insulin detemir cially for use in artificial pancreas systems. In addition,
when it is used as the basal insulin, most individuals will the first-generation long-acting insulin analogues (insu-
prefer once-daily insulin glargine. However, some guide- lin glargine and insulin detemir) do not achieve full 24 h
lines, such as the UK NICE guideline, advocate the use of basal insulin coverage in all patients with T1DM, which
insulin detemir in patients with T1DM on the basis of its has resulted in the need for twice-daily administration
superior action profile with respect to the risk of hypo- in some patients42. Another, more important, problem
glycaemia, particularly during the night, compared with is the variability in insulin-action profile that is still
using insulin glargine72. an issue for the current long-acting insulin analogues,
especially insulin glargine, which can contribute to
The need for more insulin analogues hypoglycaemia (particularly nocturnal hypoglycaemia).
The introduction of the rapid-acting and long-acting In 2014 and 2015, a biosimilar insulin glargine
insulin analogues described so far has changed the lives was approved by the EMA and the FDA, respectively,
of patients with T1DM; more patients are reaching bet- for use in people with T1DM77,78. Pharmacokinetic-
ter glucose targets, with hypoglycaemia rates decreasing pharmacodynamic studies confirmed full similarity of
and quality of life improving compared with previous the action profile to the original insulin glargine79,80.
regimens, particularly when full analogue regimens A limited clinical trial programme has confirmed iden-
are used5,47,73. In a head-to-head study, a mean of 22.1 tical clinical efficacy and safety of the biosimilar and the
fewer episodes of hypoglycaemia per patient-year original insulin8183. As this new long-acting insulin
were recorded, and the frequency of nocturnal hypo- is identical in clinical effect to the original insulin, the
glycaemia was also statistically significantly reduced biosimilar will not be discussed in more detailhere.
in patients treated with the full analogue compared with
patients using human insulins47. New basal insulins
In our opinion, the use of full analogue regimens At present, two new basal insulins are available for clin-
should be standard of care in all patients with T1DM. ical use in Europe and the USA: insulin glargine U300
In a small but elegant study (HypoAna trial) that was and insulin degludec (U100 and U200).
carried out in Denmark in people with T1DM who
experienced recurrent severe hypoglycaemia, it was Insulin glargine U300
demonstrated that people treated with the full analogue When human regular insulin is concentrated, its phar-
regimen (insulin detemir and insulin aspart) had a 29% macokineticpharmacodynamic action profile is altered.
reduction in the rate of severe hypoglycaemia episodes Compared with the non-concentrated form, concen-
(absolute reduction of 0.5 episodes per year) compared trated insulin has a similar time to onset of action,
with people receiving human insulins (NPH and regular but the glucose-lowering effect is longer than that of
insulin)74. Of note, using a full analogue regimen in this the non-concentrated form, as a result of protracted
vulnerable population of people who are prone to severe release from the injection site. At high doses, the blood
hypoglycaemia is cost saving 35. glucose-lowering effect of human regular insulin U500
is extended to up to 21 h, whereas it is 18 h for human
regular insulin U100 (REF.6). A probable explanation
Box 2 | The need to achieve a physiological insulin-action profile for this phenomenon is that, compared with the non-
Insulin analogues enable people with type1 diabetes mellitus to live more flexible lives concentrated form, more-compact conglomerates of
than do previous regimens, as using rapid-acting insulin analogues reduces the time insulin are formed under the skin, which decreases the
needed between the injection of a mealtime insulin and a meal, and obviates the need surface area from which dissociation of insulin molecules
for inter-meal snacking. The long-acting insulin analogues have enabled improved can occur and increases the distance to capillaries84.
stability of replacement of basal insulin needs, reducing the risk of nocturnal A similar effect has been achieved by concentrat-
hypoglycaemia compared with human insulin preparations. The major remaining ing insulin glargine from the usual U100 to a U300
challenge is the absence of physiological feedback on insulin supply.
formulation (FIG.1). Compared with insulin glargine
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
a S S
Human 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
regular A chain Gly Ile Val Glu Gln Cys Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn
insulin
S S
S S
B chain Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 30
S S
Insulin 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
lispro
Gly Ile Val Glu Gln Cys Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn
S S
S S
Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Lys Pro Thr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 30
S S
Insulin 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
aspart
S S
S S
Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Asp Lys Thr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 30
S S
Insulin 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
glulisine
S S
S S
Phe Val Lys Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Glu Thr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 30
b NPH
S S
1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
insulin
A chain Gly Ile Val Glu Gln Cys Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn
S S
S S
B chain Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 30
S S
Insulin 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
glargine
Gly Ile Val Glu Gln Cys Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Gly
S S
S S
Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr Arg Arg
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
S S
Insulin 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
detemir
S S
S S
Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29 NH
Insulin S S
degludec 1 2 3 4 5 7 8 9 10 12 13 14 15 6 17 18 19 20 21
S S
S S
Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6 17 18 19 20 21 22 23 24 25 26 27 28 29
O
H
N
HO O
O
O OH
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
Figure 4 | Amino acid structure of short-acting and long-acting insulins. All insulin also increased with insulin glargine U300 compared
analogues are created from the basic structure of human regular insulin. a|Rapid-acting with insulin glargine U100, but the study only included a
insulins are designed to decrease the formation of hexamers and are typically window of 3 h around the usual injectiontime.
constructed through amino acid exchanges. b|Long-acting insulins are created through There is no clinical experience with the use of insulin
exchange of amino acids to shift the isoelectric point (insulin glargine) or through
glargine U300 in pregnant women or children. However,
addition of free fatty acid moieties that enhance formation of dihexamers and
multihexamers, as well as binding to albumin (insulin detemir and insulin degludec).
considering the fact that glargine U300 is the same mol-
NPH, neutral protamine Hagedorn. ecule as insulin glargine U100, no specific safety issues
for this formulation are to be expected, and clinicians
should use the insulin analogue with the action profile
U100, insulin glargine U300 has an extended glucose- most suitable for the individual patient.
lowering action profile (FIG.3). After 1week at steady Overall, insulin glargine U300 is an important
state (0.4units perkg perday), the half-life of insulin step forward in basal insulin coverage in patients with
glargine U300 was 19.0 h compared with 13.5 h for insu- T1DM. Now, full 24 h coverage can be reached in all
lin glargine U100 (REF.85). In these pharmacokinetic patients, with increased stability translating to a reduced
pharmacodynamic studies in patients with T1DM, the risk of nocturnal hypoglycaemia and increased flexibility
duration of action even extended to over 32 h (from ~24 h in timing of administration. In addition, insulin glargine
with the U100 formulation) with a pharmacokinetic is administered with a disposable pen, which is a major
pharmacodynamic action profile for insulin glargine asset for ease of use and should avoid all confusion
U300 that is flatter than that for insulin glargine U100 around dosing of this concentrated insulin. Patients and
(REF.86). However, this study also showed a decrease in health-care professionals should be aware and cautioned
the biopotency of insulin glargine U300 at steady state against using regular insulin syringes (labelled for dosing
(27% less than insulin glargine U100)86, which sug- U100 insulins) for this concentrated insulin.
gests that the dose needs to be adjusted when switching
patients from insulin glargine U100 to insulin glargine Insulin degludec
U300. The observed variability for insulin glargine U300 Insulin degludec is long acting owing to a novel method,
is lower than that for insulin glargine U100 (REF.87). A as it relies on the formation of multihexamer chains fol-
lower increase in levels of glucose (based on continuous lowing injection in the subcutaneous tissue (FIG.1). This
glucose monitoring) in the last 4 h of the 24 h injection formation is achieved through the manipulation of the
interval, smoother average 24 h glucose profiles irre- human regular insulin molecule, with the removal of
spective of injection time and reduced nocturnal hypo- threonine at residue 30 of the Bchain and the addition of a
glycaemia were observed in the group who received 16-carbon fatty acid at residue 29 of the Bchain through
insulin glargine U300 compared with the changes seen a glutamic acid spacer (FIG.4b) This change also enables
in patients who received insulin glargine U100 (REF.87). albumin to bind to the insulin molecule, which contributes
Clinical studies that use insulin glargine U300 in to its ultra-long and stable action profile93 (FIG.3).
patients with T1DM are limited in number, with only In a head-to-head 42 h glucose-clamp study in
two studies available, one of which was exclusively in patients with T1DM, the mean half-life of insulin deglu-
Japanese patients88,89. This limited development pro- dec action was 25.4 h (versus 12.1 h for insulin glargine)
gramme of insulin glargine U300 is due to the fact that with a duration of insulin degludec action at steady state
regulatory bodies allowed a shorter programme, as using once-daily administration of 0.4units perkg of at
the insulin molecule was the same as the one that had least 42 h (REF.93) (FIG.3). The coefficient of variation of
already been extensively studied with insulin glargine the glucose-lowering effect of insulin degludec was four
U100. In EDITION 4, patients with T1DM treated times lower than that of insulin glargine at the same dose
with basal-bolus therapy were randomly assigned to (20% versus 82%), with a more even distribution of the
receive insulin glargine U100 or insulin glargine U300 glucose-lowering effect over 24 h for insulin degludec94,95.
as basal insulin89. No differences in levels of HbA1c were As demonstrated through modelling and later con-
observed, and the study only showed a benefit of the firmed in clinical studies, a once-daily administration
U300 formulation in decreasing the incidence of noc- of such a long-acting insulin does not result in the accu-
turnal hypoglycaemia during the first 8weeks of the mulation of active insulin but leads to the build-up of
study (the titration phase) and not during the mainte- a stable reserve of insulin, resulting in a stable basal
nance phase; this finding is in contrast to reports on insulin-action profile96. Clinically, however, this feature
insulin degludec9092. No firm conclusions on severe means that patience is required when adapting doses,
hypoglycaemia could be drawn as too few events were as 3days are required for insulin degludec to reach the
recorded. The increase in body weight was 0.6 kg lower stable plateau after the first injection96.
in patients who were given insulin glargine U300 than in Clinically, the long and stable action profile of insu-
those given insulin glargine U100 (P < 0.4). Basal doses lin degludec translates to fewer episodes of nocturnal
of insulin at the end of the study were higher in patients hypoglycaemia with insulin degludec than with insu-
who were given insulin glargine U300 than in those on lin glargine in patients with T1DM90,91,97. Use of either
insulin glargine U100 (0.47units perkg perday versus drug can result in similar levels of HbA1c, but with a 25%
0.40units perkg perday). No difference in efficacy or lower rate of confirmed nocturnal hypoglycaemic events
safety was present between morning and evening injec- in patients receiving insulin degludec than in those
tions of insulin glargine U300 (REF.89). Flexibility was on insulin glargine. However, the trend for the rate of
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
Box 3 | Pros and cons of insulin preparations in type1 diabetes mellitus Insulin degludec is available commercially in two for-
mulations: U100 and U200. In contrast to other insu-
Rapid-acting insulins lins (human regular insulin U500 and insulin glargine
Rapid onset of action favours insulin analogues U300), altering the concentration of insulin degludec
Postprandial glucose control favours insulin analogues has not altered its pharmacokineticpharmacodynamic
Short action profile favours insulin analogues action profile100, which means that patients can switch
Risk of diurnal hypoglycaemia favours insulin analogues between the formulations without dose adaptations. The
Need for snacks favours insulin analogues reason why the concentration of insulin degludec has no
effect on the action profile is unclear. However, protrac-
Risk of nocturnal hypoglycaemia favours insulin analogues
tion of the action profile relies on hexamer formation
Flexibility (injection with meal) favours insulin analogues
rather than on aggregation, which could be an explana-
Cost favours human insulin tion, as the release of zinc from multihexamer chains is
Experience favours human insulin the rate-limiting step for absorption of insulin degludec
Availability favours human insulin and this process is not dependent on concentration101,102.
Availability of concentrated forms favours both forms At present, insulin degludec is approved globally for
Use of continuous subcutaneous insulin infusion favours insulin analogues use in children older than 1year of age. A study in young
Long-acting insulins children and adolescents (117years of age) indicates
that the metabolic advantages in children are similar
Duration of action favours insulin analogues
to those in adults (that is, a reduced risk of nocturnal
Flat profile favours insulin analogues
hypoglycaemia), with an intriguing observation of a
Variability favours insulin analogues decreased number of ketotic episodes in children treated
Risk of hypoglycaemia favours insulin analogues with insulin degludec103. At present, no studies on the
Need for evening snacks favours insulin analogues use of insulin degludec in pregnancy are available.
Risk of nocturnal hypoglycaemia favours insulin analogues Overall, the very long-acting and stable action profile
Flexibility favours insulin analogues of insulin degludec, the extensive data from clinical trials
Cost favours human insulin and the ease of use of the disposable pen in which it comes
Experience favours human insulin make this insulin, in our opinion, the current preferred
Availability favours human insulin basal insulin analogue in patients with T1DM. However,
a major challenge in many health-care systems is the
Availability of concentrated forms favours insulin analogues
price of insulin degludec, which prevents many people
with T1DM from using this insulin. However, studies that
include factors other than just the cost of the insulin have
overall confirmed hypoglycaemia was higher in patients shown that insulin degludec can be a cost-effective alter-
receiving insulin degludec but did not reach statistical native to other basal insulin analogues in patients with
significance. Importantly, a decrease (35%) in severe T1DM. The lower costs are mainly driven by the lower
hypoglycaemia in the phase in which the dose of insulin daily dose of insulin degludec and the reduced risk of
was fully titrated was observed in those patients receiving severe hypoglycaemic events with insulin degludec104,105.
insulin degludec92. These findings were confirmed in the
SWITCH1 study, in which patients with T1DM who were Pegylated lispro
using insulin degludec or insulin glargine as their basal Despite promising metabolic results in the development
long-acting insulin were compared in a double-blind programme of a new concept to prolong the duration of
manner, with a reduced overall risk of hypoglycaemia in action of insulin analogues pegylation no pegylated
those patients receiving insulin degludec98. Of interest, basal insulin is available for clinical use at the moment,
the very long action profile of insulin degludec enables and it is doubtful whether this path is a viable method to
extreme flexibility in administration of the basal insu- develop novel insulin analogues. The concept of pegyl-
lin. Indeed, whereas insulin glargine and insulin detemir ation is built on the search for a liver-specific insulin,
need to be injected at the same time every day to maintain which would hold the promise of enhanced glucose-
appropriate basal insulin coverage in people with T1DM, lowering potential and reduced risk of hypoglycaemia
insulin degludec can be injected with time intervals and weight gain compared with currently available
between two injections as small as 8 h to as long as 40 h, insulin analogues. This search is based on the fact that,
as long as it is administered once daily 91. Importantly, this when a large polyethylene glycol (PEG) polymer chain is
flexibility does not influence the effectiveness or safety attached to an insulin molecule, the hydrodynamic size
of the basal insulin, as the glucose-lowering potency was of the molecule is substantially increased, thus leading to
similar and the nocturnal-hypoglycaemia advantage easy penetrance through the fenestrated capillaries of the
remained present91. This increased flexibility is important liver but much less penetrance through the capillaries of
to patients with T1DM, as it enables them to live active other tissues, such as fat or muscle. This molecule would
lives91. A small observational study demonstrated that, have a reduced rate of clearance, which would prolong its
in people with T1DM switching from twice-daily insulin glucose-loweringaction.
glargine or insulin detemir to once-daily insulin deglu- The pegylated insulin created by Eli Lilly, PEGLispro,
dec, levels of HbA1c, risk of hypoglycaemia and treatment fulfilled the promise of increased hepato-preferential
satisfaction improved, and doses of insulin decreased99. effects106 and was the first insulin of its kind to enter a
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
clinical development programme. Pharmacokinetic with T1DM experienced fewer nocturnal hypoglycae-
pharmacodynamic studies showed a long half-life of mic attacks (relative risk:0.62) and used less insulin than
23days107,108, with PEGLispro having less within-patient those using insulin detemir once or twice daily as basal
variability in the glucose-lowering effect than insulin insulin and insulin aspart at meals117.
glargine U100 (REF.109).
In the open-label IMAGINE1 trial, patients with Future developments
T1DM who received PEGLispro had better lowering of Several projects on novel insulin analogues or at least
levels of HbA1c but a higher rate of severe hypoglycaemic novel concepts in insulin administration are under
events compared with those who received insulin investigation and are close to being available to clinicians
glargine U100 (REF.110). The larger, blinded IMAGINE3 and patients118.
trial confirmed the superior glucose-lowering effect,
with less weight gain and a nonsignificant trend to lower Rapid-acting insulin analogues
rates of severe hypoglycaemic events, of PEGLispro com- A novel formulation of insulin lispro has been intro-
pared with insulin glargine111,112. However, a major issue duced (U200), which is primarily aimed at enabling
that arose in patients with T1DM, as well as in patients injections of reduced volume in patients who need
with T2DM, was abnormalities on liver function tests, high doses of insulin (mainly patients with T2DM).
which led to uncertainty about the safety of using pegyl- Bioequivalence and comparative pharmacodynamics
ation as a protraction mechanism and termination of have been demonstrated, and the results are in contrast
the programme113,114. To determine whether the safety to the expectations that were based on experiences with
issues were a direct toxic effect or a consequence of the more concentrated human regular insulin119.
liver-preferred insulin action (accumulation of fat in The pharmaceutical industry has taken several routes
the liver that could lead to nonalcoholic steatohepatitis in the pursuit of the perfect match between insulin-
in the long term), long and extensive studies would have action profiles and meal-induced glucose excursions,
been needed, which was considered unachievable. and in particular to find insulins that have a faster on/off
The story of PEGLispro opened the discussion on action profile. Early on, Halozyme attempted to alter the
desirability of a liver-specific insulin. Clearly, an essen- injection site in such a manner that entry of the injected
tial part of insulin action is suppression of peripheral insulin into the draining blood vessels would be enhanced
lipolysis. When suppression of glucose output by the (for example, enzymatically by hyaluronidase). No clini-
liver is not accompanied by some suppression of lipol- cal data are available, and the programme is not active.
ysis in fat tissue, the liver will be overwhelmed by free Several companies have altered the excipients in which
fatty acids coming in via that route, and thus liver ste- existing rapid-acting insulin analogues are presented.
atosis, and even steatohepatitis, will inevitably ensue114. For example, Biodel developed an ultra-fast-acting insu-
Achieving a liver-preferred action, favouring the liver, lin on the basis of human regular insulin, now suspended
but still affecting fat to some extent is the challenge for in EDTA, citrate and magnesium sulfate, to increase the
future analogue development. rapidity of onset of action, but this programme has also
been halted120123.
Co-formulations A novel rapid-acting insulin analogue with the
The structure of insulin degludec and the way it is most advanced clinical development is fast-acting
formulated have made it possible to develop a combi- insulin aspart, in which the excipients have been
natorial insulin in which insulin degludec and insu- altered to enable a faster onset of action. Fast-acting
lin aspart are co-formulated at a ratio of 70:30 (insulin insulin aspart includes the original molecule of insu-
degludecaspart). Insulin degludec and insulin aspart lin aspart set in a new formulation that contains two
remain separate entities in solution, and size-exclusion well-known excipients, nicotinamide and arginine.
chromatography studies have found no evidence of phys- The addition of these excipients results in a stable
ical or chemical interaction between the two insulins in formulation with an initial absorption after subcuta-
co-formulation115. The glucose-lowering effect of once- neous injection that is faster than that of standard insu-
daily insulin degludecaspart is characterized by a peak lin aspart 118. In a pharmacokinetic study, fast-acting
action from the insulin aspart present in the solution and insulin aspart had a faster onset of action than insulin
a separate basal action that lasts more than 30 h at steady aspart (4.9 min versus 11.2 min) and reached the 50%
state from the insulin degludec116. maximum concentration more quickly (20.7 min ver-
In our opinion, the need for this combinatorial insu- sus 31.6 min)118,124,125. The greatest difference occurred
lin is limited in T1DM, as the basis of insulin therapy in during the first 15 min, when the area under the curve
T1DM is flexibility in dose adjustment of mealtime and was 4.5-fold greater with fast-acting insulin aspart than
basal components of the therapy dependent on varia- with insulin aspart. Both insulins had similar time to
ble factors such as meals, glycaemic levels and exercise. maximum concentration, total exposure and maximum
This flexibility is partially lost with a fixed-combination concentration. Pharmacodynamic analysis showed that
insulin such as insulin degludecaspart. Even so, clini- fast-acting insulin aspart had a greater glucose-lowering
cal trials in patients with T1DM have been carried out, effect within 90 min after dosing than insulin aspart.
and a 52-week study demonstrated that, when insulin Both insulins had similar total and maximum glucose-
degludecaspart was administered at the main meal, lowering effects, which indicates that they have similar
combined with insulin aspart at the other meals, people overall potency but that a shift of action to earlier time
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
points has occurred with fast-acting insulin aspart. these insulin analogues is also an important step in the
These ultra-fast pharmacodynamic properties of development of workable algorithms for concepts to
fast-acting insulin aspart are similar in elderly patients develop an artificial pancreas.
(65years of age) and children (617years of age) with
T1DM compared to the properties in young adults Other concepts
(1835years of age)125,126. Intensive research is ongoing in the field of creating even
A group of patients in whom fast-acting insulin longer-lasting insulin preparations131, exploiting novel
aspart might be of most use is users of CSII. In a ran- technologies such as antibody-linked insulins to achieve
domized, double-blind, crossover trial in users of weekly administrable insulin preparations. These insulins
CSII, the pharmacokineticpharmacodynamic action have a protracted action profile as a result of decreased
profiles for fast-acting insulin aspart were left shifted clearance. Data are preliminary, and research is still in the
compared with those for insulin aspart, with onset of animal-model phase132,133.
action 11.1 min earlier for fast-acting insulin aspart; Another research avenue is the development of
however, the end of insulin action was also left shifted smart insulins, which are insulins that would be released
by 24 min127. In a double-blind, randomized, crossover from depots under the skin when levels of glucose rise.
trial in 43 patients with T1DM, CSII delivery of fast- Currently, developments are happening based on resin-
acting insulin aspart had a greater glucose-lowering embedded insulins, as well as on lectin-bound insulins
effect than insulin aspart after a meal test. In addition, that would be released from lectin-binding areas when
continuous glucose-monitoring results showed that levels of glucose rise134,135.
patients who received fast-acting insulin aspart spent Finally, the goal of administering insulin through
less time with low levels of glucose (<70 mg/dl) than methods other than parenterally is still being pursued.
those who received insulin aspart 128. Inhaled insulin is available but is used infrequently 136,137.
The first clinical study results in patients with T1DM The first reports on the development of oral insulin
showed that 6months of treatment with fast-acting insu- preparations are beginning to appear 137,138.
lin aspart enabled patients to achieve a level of HbA1c
that was significantly lower than the level in patients Conclusions
who used insulin aspart (a difference in HbA1c levels of Achieving normoglycaemia without hypoglycaemia and
0.15%), with lower postprandial glucose excursions32. No excessive weight gain in patients with T1DM remains an
differences in the risk of hypoglycaemia were observed. elusive goal, but the advent of insulin analogues has had
In this study, postprandial administration of fast-acting a large effect, enabling intensive insulin therapy without
insulin aspart showed similar glycaemic control to being too disruptive to daily life. Rapid-acting insulin
preprandial administration of standard insulin aspart, analogues, in particular the ultra-rapid-acting insulins,
which suggests that postprandial administration of can be administered shortly before meals, giving bet-
this rapid-acting insulin would become possible with- ter coverage of mealtime-induced glucose excursions
out compromising efficacy or safety 32. This insulin was than human regular insulin. In addition, basal insulin
approved by the EMA in January 2017 (REF.129). analogues are becoming more stable and provide better
Some researchers are underwhelmed by the gain of coverage of basal insulin needs for people with T1DM
just a couple of minutes in speed of onset with this novel than human insulin. Clinical studies on individual agents
ultra-rapid-acting insulin. However, in our opinion, in show small advantages, mainly in prevention of hypo-
real-life settings, these few minutes make a considerable glycaemia, when treating patients to achieve similar gly-
difference to the lives of people with T1DM. From the caemic targets. In particular, studies comparing human
introduction of the original rapid-acting insulin ana- regular insulin with full insulin analogue regimens
logues, we have learned that a gain of a few minutes clearly demonstrate the superiority of these agents in
in onset of insulin action leads to improved prandial the treatment of patients with T1DM. Improvements
coverage of the meal-induced glucose excursions and in rapidity in onset of rapid-acting insulin analogues, sta-
enables patients to administer their insulin injection at bility of long-acting insulin analogues, mode of admin-
mealtimes at a more convenient time. However, contin- istration and glucose sensitivity for action would further
uous glucose monitoring has shown that many patients contribute to improved glycaemic control in patients
now inject just at the start of the meal, or even dur- withT1DM.
ing or after meals. Gaining another 45 min in speed The novel insulin analogues are an important step
of onset of insulin action will contribute to improved in the path to achieving tight glycaemic control without
postprandial control and, as important, to more flexi- hypoglycaemia in patients with T1DM. However, even
bility in peoples lives without compromising efficacy or with the improved action profiles of the insulin analogues,
safety of the insulin regimens. In our opinion, particu- insulin therapy in patients with T1DM remains a chal-
larly in patients with T1DM, the introduction of ultra- lenge, with the risk of hypoglycaemia and weight gain still
rapid-acting insulin analogues will make a difference. present. In the treatment of patients with T1DM, educa-
The group in which this difference of a couple of min- tion, improvements in glucose monitoring and devices
utes is even more crucial is users of CSII, in particular that assist patients in insulin delivery and decision mak-
those who use sensor-augmented pumps. With this ing have a crucial role for embedding T1DM in their lives
technology, everything depends on the rapidity of onset until the ideal insulin preparation, able to fully mimic the
and offset of the insulin in the pump130. Availability of physiological insulin secretion of cells, is discovered.
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
1. Atkinson,M.A., Eisenbarth,G.S. & Michels,A.W. 22. Home,P.D., Lindholm,A., Riis,A. & European 41. Brunner,G.A. etal. Pharmacokinetic and
Type1 diabetes. Lancet 383, 6982 (2014). Insulin Aspart Study Group. Insulin aspart versus pharmacodynamic properties of long-acting insulin
2. Donner,T. Insulin pharmacology, therapeutic human insulin in the management of long-term analogue NN304 in comparison to NPH insulin in
regimens and principles of intensive insulin therapy. blood glucose control in type1 diabetes mellitus: humans. Exp. Clin. Endocrinol. Diabetes 108, 100105
Endotext https://www.ncbi.nlm.nih.gov/books/ a randomized controlled trial. Diabet. Med. 17, (2000).
NBK278938/ (2000). 762770 (2000). 42. Porcellati,F., Bolli,G.B. & Fanelli,C.G.
3. The Diabetes Control and Complications Trial 23. Raskin,P., Guthrie,R.A., Leiter,L., Riis,A. & Pharmacokinetics and pharmacodynamics of basal
Research Group. The effect of intensive treatment of Jovanovic,L. Use of insulin aspart, a fast-acting insulin insulins. Diabetes Technol. Ther. 13 (Suppl. 1), S15S24
diabetes on the development and progression of long- analog, as the mealtime insulin in the management of (2011).
term complications in insulin-dependent diabetes patients with type1 diabetes. Diabetes Care 23, 43. Rosenstock,J., Park,G., Zimmerman,J. & U.S. Insulin
mellitus. N.Engl. J.Med. 329, 977986 (1993). 583588 (2000). Glargine (HOE 901) Type1 Diabetes Investigator
4. Nathan,D.M. & DCCT/EDIC Research Group. 24. Becker,R.H., Frick,A.D., Burger,F., Potgieter,J.H. & Group. Basal insulin glargine (HOE 901) versus NPH
The diabetes control and complications trial/ Scholtz,H. Insulin glulisine, a new rapid-acting insulin insulin in patients with type1 diabetes on multiple
epidemiology of diabetes interventions and analogue, displays a rapid time-action profile in obese daily insulin regimens. Diabetes Care 23, 11371142
complications study at 30years: overview. non-diabetic subjects. Exp. Clin. Endocrinol. Diabetes (2000).
Diabetes Care 37, 916 (2014). 113, 435443 (2005). 44. Albright,E.S., Desmond,R. & Bell,D.S. Efficacy of
5. Home,P.D. The pharmacokinetics and 25. Kerr,D., Wizemann,E., Senstius,J., Zacho,M. & conversion from bedtime NPH insulin injection to
pharmacodynamics of rapid-acting insulin analogues Ampudia-Blasco,F.J. Stability and performance of once- or twice-daily injections of insulin glargine in
and their clinical consequences. Diabetes Obes. Metab. rapid-acting insulin analogs used for continuous type1 diabetic patients using basal/bolus therapy.
14, 780788 (2012). subcutaneous insulin infusion: a systematic review. Diabetes Care 27, 632633 (2004).
6. de la Pena,A. etal. Pharmacokinetics and J.Diabetes Sci. Technol. 7, 15951606 (2013). 45. Ratner,R.E. etal. Less hypoglycemia with insulin
pharmacodynamics of high-dose human regular U-500 26. Heise,T. etal. Insulin glulisine: a faster onset of action glargine in intensive insulin therapy for type1
insulin versus human regular U-100 insulin in healthy compared with insulin lispro. Diabetes Obes. Metab. diabetes. U.S. study group of insulin glargine in type1
obese subjects. Diabetes Care 34, 24962501 9, 746753 (2007). diabetes. Diabetes Care 23, 639643 (2000).
(2011). 27. Arnolds,S. etal. Insulin glulisine has a faster onset of 46. Dornhorst,A. etal. Safety and efficacy of insulin
7. Heise,T. etal. Lower within-subject variability of action compared with insulin aspart in healthy detemir in clinical practice: 14-week follow-up data
insulin detemir in comparison to NPH insulin and volunteers. Exp. Clin. Endocrinol. Diabetes 118, from type1 and type2 diabetes patients in the
insulin glargine in people with type1 diabetes. 662664 (2010). PREDICTIVE European cohort. Int. J.Clin. Pract. 61,
Diabetes 53, 16141620 (2004). 28. Luzio,S., Peter,R., Dunseath,G.J., Mustafa,L. & 523528 (2007).
8. Heise,T. & Pieber,T.R. Towards peakless, Owens,D.R. A comparison of preprandial insulin 47. Hermansen,K., Dornhorst,A. & Sreenan,S.
reproducible and long-acting insulins. An glulisine versus insulin lispro in people with type2 Observational, open-label study of type1 and type2
assessment of the basal analogues based on diabetes over a 12- h period. Diabetes Res. Clin. diabetes patients switching from human insulin to
isoglycaemic clamp studies. Diabetes Obes. Metab. Pract. 79, 269275 (2008). insulin analogue basal-bolus regimens: insights from
9, 648659 (2007). 29. Bolli,G.B. etal. Comparative pharmacodynamic and the PREDICTIVE study. Curr. Med. Res. Opin. 25,
9. Lucidi,P. etal. Pharmacokinetics and pharmacokinetic characteristics of subcutaneous 26012608 (2009).
pharmacodynamics of therapeutic doses of basal insulin glulisine and insulin aspart prior to a standard 48. [No authors listed.] Top 50 pharmaceutical products
insulins NPH, glargine, and detemir after 1week of meal in obese subjects with type2 diabetes. by global sales. PMLiVE http://www.pmlive.com/top_
daily administration at bedtime in type2 diabetic Diabetes Obes. Metab. 13, 251257 (2011). pharma_list/Top_50_pharmaceutical_products_by_
subjects: a randomized cross-over study. 30. Garg,S.K., Rosenstock,J. & Ways,K. Optimized global_sales (2017).
Diabetes Care 34, 13121314 (2011). basal-bolus insulin regimens in type1 diabetes: 49. Lepore,M. etal. Pharmacokinetics and
10. Brange,J., Owens,D.R., Kang,S. & Volund,A. insulin glulisine versus regular human insulin in pharmacodynamics of subcutaneous injection of long-
Monomeric insulins and their experimental and combination with basal insulin glargine. Endocr. Pract. acting human insulin analog glargine, NPH insulin,
clinical implications. Diabetes Care 13, 923954 11, 1117 (2005). and ultralente human insulin and continuous
(1990). 31. Fullerton,B. etal. Short-acting insulin analogues subcutaneous infusion of insulin lispro. Diabetes 49,
11. Heinemann,L. etal. Prandial glycaemia after a versus regular human insulin for adults with type1 21422148 (2000).
carbohydrate-rich meal in typeI diabetic patients: diabetes mellitus. Cochrane Database Syst. Rev. 6, 50. Porcellati,F. etal. Pharmacokinetics and
using the rapid acting insulin analogue [Lys(B28), CD012161 (2016). pharmacodynamics of the long-acting insulin analog
Pro(B29)] human insulin. Diabet. Med. 13, 625629 32. Russell-Jones,D. etal. Fast-acting insulin aspart glargine after 1week of use compared with its first
(1996). improves glycemic control in basal-bolus treatment for administration in subjects with type1 diabetes.
12. ter Braak,E.W. etal. Injection site effects on the type 1 diabetes: Results of a 26-week multicenter, Diabetes Care 30, 12611263 (2007).
pharmacokinetics and glucodynamics of insulin lispro active-controlled, treat-to-target, randomized, 51. Hilgenfeld,R. etal. Controlling insulin bioavailability
and regular insulin. Diabetes Care 19, 14371440 parallel-group trial (Onset 1). Diabetes Care by crystal contact engineering. Diabetologia 35, A193
(1996). http://dx.doi.org/10.2337/dc16-1771 (2017). (1992).
13. Tamas,G. etal. Glycaemic control in type1 diabetic 33. DeVries,J.H. etal. A randomized trial of insulin 52. Klein,O. etal. Albumin-bound basal insulin analogues
patients using optimised insulin aspart or human aspart with intensified basal NPH insulin (insulin detemir and NN344): comparable time-action
insulin in a randomised multinational study. supplementation in people with type1 diabetes. profiles but less variability than insulin glargine in
Diabetes Res. Clin. Pract. 54, 105114 (2001). Diabet. Med. 20, 312318 (2003). type2 diabetes. Diabetes Obes. Metab. 9, 290299
14. Valle,D., Santoro,D., Bates,P., Scarpa,L. & Italian 34. Shafie,A.A., Ng,C.H., Tan,Y.P. & (2007).
Multicentre Lispro Study Group. Italian multicentre Chaiyakunapruk,N. Systematic review of the cost 53. Ashwell,S.G., Gebbie,J. & Home,P.D. Optimal
study of intensive therapy with insulin lispro in 1184 effectiveness of insulin analogues in type1 and type2 timing of injection of once-daily insulin glargine in
patients with type1 diabetes. Diabetes Nutr. Metab. diabetes mellitus. Pharmacoeconomics 35, 141162 people with type1 diabetes using insulin lispro at
14, 126132 (2001). (2017). meal-times. Diabet. Med. 23, 4652 (2006).
15. Lindholm,A., McEwen,J. & Riis,A.P. Improved 35. Pedersen-Bjergaard,U. etal. Short-term cost- 54. Havelund,S. etal. The mechanism of protraction of
postprandial glycemic control with insulin aspart. effectiveness of insulin detemir and insulin aspart in insulin detemir, a long-acting, acylated analog of
A randomized double-blind cross-over trial in type1 people with type1 diabetes who are prone to human insulin. Pharm. Res. 21, 14981504
diabetes. Diabetes Care 22, 801805 (1999). recurrent severe hypoglycemia. Curr. Med. Res. Opin. (2004).
16. Home,P.D., Barriocanal,L. & Lindholm,A. 32, 17191725 (2016). 55. Pieber,T.R. etal. Duration of action,
Comparative pharmacokinetics and 36. American Diabetes Association. 8. Pharmacologic pharmacodynamic profile and between-subject
pharmacodynamics of the novel rapid-acting insulin approaches to glycemic treatment. Diabetes Care 40, variability of insulin detemir in subjects with type1
analogue, insulin aspart, in healthy volunteers. S64S74 (2017). diabetes. Diabetes 51, A53 (2002).
Eur. J.Clin. Pharmacol. 55, 199203 (1999). 37. Gonzalez Blanco,C., Chico Ballesteros,A., 56. Pieber,T.R. etal. Comparison of insulin detemir and
17. Plank,J. etal. A direct comparison of insulin aspart Gich Saladich,I. & Corcoy Pla,R. Glycemic control insulin glargine in subjects with type1 diabetes using
and insulin lispro in patients with type1 diabetes. and pregnancy outcomes in women with type1 intensive insulin therapy. Diabet. Med. 24, 635642
Diabetes Care 25, 20532057 (2002). diabetes mellitus using lispro versus regular insulin: (2007).
18. Homko,C., Deluzio,A., Jimenez,C., Kolaczynski,J.W. a systematic review and meta-analysis. Diabetes 57. Heller,S., Koenen,C. & Bode,B. Comparison of insulin
& Boden,G. Comparison of insulin aspart and lispro: Technol. Ther. 13, 907911 (2011). detemir and insulin glargine in a basal-bolus regimen,
pharmacokinetic and metabolic effects. Diabetes Care 38. Mathiesen,E.R. etal. Maternal glycemic control with insulin aspart as the mealtime insulin, in patients
26, 20272031 (2003). and hypoglycemia in type1 diabetic pregnancy: with type1 diabetes: a 52-week, multinational,
19. Bartolo,P.D. etal. Better postprandial glucose a randomized trial of insulin aspart versus human randomized, open-label, parallel-group, treat-to-target
stability during continuous subcutaneous infusion with insulin in 322 pregnant women. Diabetes Care 30, noninferiority trial. Clin. Ther. 31, 20862097
insulin aspart compared with insulin lispro in patients 771776 (2007). (2009).
with type1 diabetes. Diabetes Technol. Ther. 10, 39. Pozzilli,P. etal. Continuous subcutaneous insulin 58. Plank,J. etal. A double-blind, randomized, dose-
495498 (2008). infusion in diabetes: patient populations, safety, response study investigating the pharmacodynamic
20. Bode,B. etal. Comparison of insulin aspart with efficacy, and pharmacoeconomics. Diabetes Metab. and pharmacokinetic properties of the long-acting
buffered regular insulin and insulin lispro in Res. Rev. 32, 2139 (2016). insulin analog detemir. Diabetes Care 28, 11071112
continuous subcutaneous insulin infusion: a 40. Kerr,D., Morton,J., Whately-Smith,C., Everett,J. (2005).
randomized study in type1 diabetes. Diabetes Care & Begley,J.P. Laboratory-based non-clinical 59. Koehler,G. etal. Pharmacodynamics of the long-
25, 439444 (2002). comparison of occlusion rates using three rapid- acting insulin analogues detemir and glargine
21. Dreyer,M. etal. Efficacy and safety of insulin glulisine acting insulin analogs in continuous subcutaneous following single-doses and under steady-state
in patients with type1 diabetes. Horm. Metab. Res. insulin infusion catheters using low flow rates. conditions in patients with type1 diabetes.
37, 702707 (2005). J.Diabetes Sci. Technol. 2, 450455 (2008). Diabetes Obes. Metab. 16, 5762 (2014).
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
60. Danne,T. etal. Insulin detemir is characterized by a 80. Linnebjerg,H. etal. Pharmacokinetics of the long- 99. Galasso,S. etal. Switching from twice-daily glargine
more reproducible pharmacokinetic profile than acting basal insulin LY2605541 in subjects with or detemir to once-daily degludec improves glucose
insulin glargine in children and adolescents with varying degrees of renal function. Clin. Pharmacol. control in type1 diabetes. An observational study.
type1 diabetes: results from a randomized, double- Drug Dev. 5, 216224 (2016). Nutr. Metab. Cardiovasc. Dis. 26, 11121119 (2016).
blind, controlled trial. Pediatr. Diabetes 9, 554560 81. Ilag,L.L. etal. Evaluation of immunogenicity of 100. Korsatko,S. etal. A comparison of the steady-state
(2008). LY2963016 insulin glargine compared with Lantus pharmacokinetic and pharmacodynamic profiles of
61. De Leeuw,I. etal. Insulin detemir used in basal-bolus insulin glargine in patients with type1 or type2 diabetes 100 and 200 U/mL formulations of ultra-long-acting
therapy in people with type1 diabetes is associated mellitus. Diabetes Obes. Metab. 18, 159168 (2016). insulin degludec. Clin. Drug Investig. 33, 515521
with a lower risk of nocturnal hypoglycaemia and less 82. Hadjiyianni,I. etal. Efficacy and safety of LY2963016 (2013).
weight gain over 12months in comparison to NPH insulin glargine in patients with type1 and type2 101. Jonassen,I. etal. Design of the novel protraction
insulin. Diabetes Obes. Metab. 7, 7382 (2005). diabetes previously treated with insulin glargine. mechanism of insulin degludec, an ultra-long-acting
62. Tricco,A.C. etal. Safety, effectiveness, and cost Diabetes Obes. Metab. 18, 425429 (2016). basal insulin. Pharm. Res. 29, 21042114 (2012).
effectiveness of long acting versus intermediate acting 83. Linnebjerg,H. etal. Duration of action of two insulin 102. Haahr,H. & Heise,T. A review of the pharmacological
insulin for patients with type1 diabetes: systematic glargine products, LY2963016 insulin glargine and properties of insulin degludec and their clinical
review and network meta-analysis. BMJ 349, g5459 Lantus insulin glargine, in subjects with type1 relevance. Clin. Pharmacokinet. 53, 787800
(2014). diabetes mellitus. Diabetes Obes. Metab. 19, 3339 (2014).
63. Frier,B.M., Russell-Jones,D. & Heise,T. A comparison (2016). 103. Thalange,N. etal. Insulin degludec in combination
of insulin detemir and neutral protamine Hagedorn 84. Sindelka,G., Heinemann,L., Berger,M., Frenck,W. with bolus insulin aspart is safe and effective in
(isophane) insulin in the treatment of diabetes: & Chantelau,E. Effect of insulin concentration, children and adolescents with type1 diabetes.
a systematic review. Diabetes Obes. Metab. 15, subcutaneous fat thickness and skin temperature on Pediatr. Diabetes 16, 164176 (2015).
978986 (2013). subcutaneous insulin absorption in healthy subjects. 104. Evans,M., Chubb,B. & Gundgaard,J. Cost-
64. Rosenstock,J. etal. A randomised, 52-week, Diabetologia 37, 377380 (1994). effectiveness of insulin degludec versus insulin
treat-to-target trial comparing insulin detemir with 85. Becker,R.H., Nowotny,I., Teichert,L., Bergmann,K. glargine in adults with type1 and type2 diabetes
insulin glargine when administered as add-on to & Kapitza,C. Low within- and between-day variability mellitus. Diabetes Ther. http:dx.doi.org/10.1007/
glucose-lowering drugs in insulin-naive people with in exposure to new insulin glargine 300 U/ml. s13300-017-0236-9 (2017).
type2 diabetes. Diabetologia 51, 408416 (2008). Diabetes Obes. Metab. 17, 261267 (2015). 105. Landstedt-Hallin,L., Gundgaard,J., Ericsson,A. &
65. Hordern,S.V. & Russell-Jones,D.L. Insulin detemir, 86. Becker,R.H. etal. New insulin glargine 300 Ellfors-Zetterlund,S. Cost-effectiveness of switching to
does a new century bring a better basal insulin? UnitsmL-1 provides a more even activity profile and insulin degludec from other basal insulins: evidence
Int. J.Clin. Pract. 59, 730739 (2005). prolonged glycemic control at steady state compared from Swedish real-world data. Curr. Med. Res. Opin.
66. Herring,R. etal. Effect of subcutaneous insulin with insulin glargine 100 UnitsmL-1. Diabetes Care http://dx.doi.org/10.1080/03007995.2016.1277194
detemir on glucose flux, lipolysis and 38, 637643 (2015). (2017).
electroencephalography in type1 diabetes. 87. Bergenstal,R.M. etal. Comparison of insulin glargine 106. Henry,R.R. etal. Basal insulin peglispro
Diabetes Obes. Metab. 17, 11001103 (2015). 300 U/mL and 100 U/mL in adults with type1 demonstrates preferential hepatic versus peripheral
67. Pollex,E., Moretti,M.E., Koren,G. & Feig,D.S. diabetes: continuous glucose monitoring profiles and action relative to insulin glargine in healthy subjects.
Safety of insulin glargine use in pregnancy: a systematic variability using morning or evening injections. Diabetes Care 37, 26092615 (2014).
review and meta-analysis. Ann. Pharmacother. 45, Diabetes Care 40, 554560 (2017). 107. Sinha,V.P. etal. Single-dose pharmacokinetics and
916 (2011). 88. Matsuhisa,M. etal. Sustained glycaemic control and glucodynamics of the novel, long-acting basal insulin
68. Blumer,I. etal. Diabetes and pregnancy: an endocrine less nocturnal hypoglycaemia with insulin glargine LY2605541 in healthy subjects. J.Clin. Pharmacol.
society clinical practice guideline. J.Clin. Endocrinol. 300U/mL compared with glargine 100U/mL in 54, 792799 (2014).
Metab. 98, 42274249 (2013). Japanese adults with type1 diabetes (EDITION JP 1 108. Caparrotta,T.M. & Evans,M. PEGylated insulin
69. Mathiesen,E.R. etal. Maternal efficacy and safety randomised 12-month trial including 6-month lispro, (LY2605541) a new basal insulin analogue.
outcomes in a randomized, controlled trial comparing extension). Diabetes Res. Clin. Pract. 122, 133140 Diabetes Obes. Metab. 16, 388395 (2014).
insulin detemir with NPH insulin in 310 pregnant (2016). 109. Buse,J.B. etal. Randomized clinical trial comparing
women with type1 diabetes. Diabetes Care 35, 89. Home,P.D. etal. New insulin glargine 300 Units/mL basal insulin peglispro and insulin glargine in patients
20122017 (2012). versus glargine 100 Units/mL in people with type1 with type2 diabetes previously treated with basal
70. Tan,C.Y., Wilson,D.M. & Buckingham,B. Initiation of diabetes: a randomized, phase 3a, open-label clinical insulin: IMAGINE 5. Diabetes Care 39, 92100
insulin glargine in children and adolescents with trial (EDITION 4). Diabetes Care 38, 22172225 (2016).
type1 diabetes. Pediatr. Diabetes 5, 8086 (2004). (2015). 110. Garg,S. etal. A randomized clinical trial comparing
71. Thalange,N., Bereket,A., Larsen,J., Hiort,L.C. & 90. Heller,S., Mathieu,C., Kapur,R., Wolden,M.L. & basal insulin peglispro and insulin glargine, in
Peterkova,V. Treatment with insulin detemir or NPH Zinman,B. A meta-analysis of rate ratios for nocturnal combination with prandial insulin lispro, in patients
insulin in children aged 25 yr with type1 diabetes confirmed hypoglycaemia with insulin degludec versus with type1 diabetes: IMAGINE 1. Diabetes Obes.
mellitus. Pediatr. Diabetes 12, 632641 (2011). insulin glargine using different definitions for Metab. 18 (Suppl. 2), 2533 (2016).
72. National Institute for Health and Care Excellence. hypoglycaemia. Diabet. Med. 33, 478487 (2016). 111. Garg,S. etal. Greater HbA1c reduction with basal
Type1 diabetes in adults: diagnosis and management. 91. Mathieu,C. etal. Efficacy and safety of insulin insulin peglispro (BIL) versus insulin glargine (GL) in an
NICE https://www.nice.org.uk/guidance/ng17 (2015). degludec in a flexible dosing regimen versus insulin open-label, randomised study in type1 diabetic
73. Ashwell,S.G. etal. Improved glycaemic control with glargine in patients with type1 diabetes (BEGIN: Flex patients: IMAGINE 1 [abstract 3]. Diabetologia 58
insulin glargine plus insulin lispro: a multicentre, T1): a 26-week randomized, treat-to-target trial with a (Suppl. 1), S2 (2015).
randomized, cross-over trial in people with type1 26-week extension. J.Clin. Endocrinol. Metab. 98, 112. Bergenstal,R.M. etal. Randomized, double-blind
diabetes. Diabet. Med. 23, 285292 (2006). 11541162 (2013). clinical trial comparing basal insulin peglispro and
74. Pedersen-Bjergaard,U. etal. Effect of insulin 92. Ratner,R.E. etal. Hypoglycaemia risk with insulin insulin glargine, in combination with prandial insulin
analogues on risk of severe hypoglycaemia in patients degludec compared with insulin glargine in type2 and lispro, in patients with type1 diabetes: IMAGINE 3.
with type1 diabetes prone to recurrent severe type1 diabetes: a pre-planned meta-analysis of phase Diabetes Obes. Metab. 18, 10811088 (2016).
hypoglycaemia (HypoAna trial): a prospective, 3 trials. Diabetes Obes. Metab. 15, 175184 (2013). 113. Cusi,K. etal. Different effects of basal insulin
randomised, open-label, blinded-endpoint crossover 93. Heise,T. etal. Comparison of the pharmacokinetic and peglispro and insulin glargine on liver enzymes and
trial. Lancet Diabetes Endocrinol. 2, 553561 pharmacodynamic profiles of insulin degludec and liver fat content in patients with type1 and type2
(2014). insulin glargine. Expert Opin. Drug Metab. Toxicol. 11, diabetes. Diabetes Obes. Metab. 18 (Suppl. 2),
75. Taki,K. etal. Analysis of 24-hour glycemic excursions 11931201 (2015). 5058 (2016).
in patients with type1 diabetes by using continuous 94. Heise,T. etal. A new-generation ultra-long-acting 114. Munoz-Garach,A., Molina-Vega,M. & Tinahones,F.J.
glucose monitoring. Diabetes Technol. Ther. 12, basal insulin with a bolus boost compared with insulin How can a good idea fail? Basal insulin peglispro
523528 (2010). glargine in insulin-naive people with type2 diabetes: [LY2605541] for the treatment of type2 diabetes.
76. Maia,F.F. & Araujo,L.R. Efficacy of continuous a randomized, controlled trial. Diabetes Care 34, Diabetes Ther. 8, 922 (2017).
glucose monitoring system (CGMS) to detect 669674 (2011). 115. Atkin,S., Javed,Z. & Fulcher,G. Insulin degludec and
postprandial hyperglycemia and unrecognized 95. Heise,T., Nosek,L., Bottcher,S.G., Hastrup,H. & insulin aspart: novel insulins for the management of
hypoglycemia in type1 diabetic patients. Haahr,H. Ultra-long-acting insulin degludec has a flat diabetes mellitus. Ther. Adv. Chronic Dis. 6, 375388
Diabetes Res. Clin. Pract. 75, 3034 (2007). and stable glucose-lowering effect in type2 diabetes. (2015).
77. European Medicines Agency. Abasaglar (previously Diabetes Obes. Metab. 14, 944950 (2012). 116. Heise,T. etal. Distinct prandial and basal glucose-
Abasria). EMA http://www.ema.europa.eu/ema/ 96. Heise,T. & Meneghini,L.F. Insulin stacking versus lowering effects of insulin degludec/insulin aspart
index.jsp?curl=pages/medicines/human/medicines/ therapeutic accumulation: understanding the (IDegAsp) at steady state in subjects with type1
002835/human_med_001790.jsp&mid= differences. Endocr. Pract. 20, 7583 (2014). diabetes mellitus. Diabetes Ther. 5, 255265 (2014).
WC0b01ac058001d124 (2014). 97. Heller,S. etal. Insulin degludec, an ultra-longacting 117. Hirsch,I.B., Franek,E., Mersebach,H., Bardtrum,L.
78. US Food and Drug Administration. FDA approves basal insulin, versus insulin glargine in basal-bolus & Hermansen,K. Safety and efficacy of insulin
Basaglar, the first follow-on insulin glargine product treatment with mealtime insulin aspart in type1 degludec/insulin aspart with bolus mealtime insulin
to treat diabetes. FDA https://www.fda.gov/ diabetes (BEGIN Basal-Bolus Type1): a phase 3, aspart compared with standard basal-bolus treatment
NewsEvents/Newsroom/PressAnnouncements/ randomised, open-label, treat-to-target non-inferiority in people with type1 diabetes: 1-year results from a
ucm477734.htm (2015). trial. Lancet 379, 14891497 (2012). randomized clinical trial (BOOST(R) T1). Diabet. Med.
79. Linnebjerg,H. etal. Comparison of the 98. Lane,W.S. etal. Switch1: reduced hypoglycaemia 34, 167173 (2017).
pharmacokinetics and pharmacodynamics of with insulin degludec (IDeg) vs. insulin glargine (IGlar), 118. Heise,T. etal. Faster-acting insulin aspart: earlier
LY2963016 insulin glargine and EU- and US-approved both U100, in patients with T1D at high risk of onset of appearance and greater early
versions of lantus insulin glargine in healthy subjects: hypoglycaemia: a randomized, double-blind, crossover pharmacokinetic and pharmacodynamic effects than
three randomized euglycemic clamp studies. trial [abstract LB-87]. American Diabetes Association insulin aspart. Diabetes Obes. Metab. 17, 682688
Diabetes Care 38, 22262233 (2015). (2016). (2015).
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

REVIEWS
119. de la Pena,A. etal. Bioequivalence and comparative diabetes using continuous subcutaneous insulin 138. Fonte,P., Araujo,F., Reis,S. & Sarmento,B.
pharmacodynamics of insulin lispro 200 U/mL infusion: a randomised, double-blind, crossover trial. Oral insulin delivery: how far are we? J.Diabetes
relative to insulin lispro (Humalog) 100 U/mL. Diabetes Obes. Metab. 19, 208215 (2017). Sci. Technol. 7, 520531 (2013).
Clin. Pharmacol. Drug Dev. 5, 6975 (2016). 128. Bode,B.W., Johnson,J.A., Hyveled,L., Tamer,S.C. &
120. Muchmore,D.B. & Vaughn,D.E. Review of the Demissie,M. Improved postprandial glycemic control Author contributions
mechanism of action and clinical efficacy of with faster-acting insulin aspart in patients with type1 C.M. and K.B. researched data for the article, contributed to
recombinant human hyaluronidase coadministration diabetes using continuous subcutaneous insulin discussion of the content, wrote the article and reviewed
with current prandial insulin formulations. infusion. Diabetes Technol. Ther. 19, 2533 (2017). and/or edited the manuscript before submission. P.G. contrib-
J.Diabetes Sci. Technol. 4, 419428 (2010). 129. European Medicines Agency. Fiasp. EMA http://www. uted to discussion of the content and reviewed and/or edited
121. Muchmore,D.B. & Vaughn,D.E. Accelerating and ema.europa.eu/ema/index.jsp?curl=pages/medicines/ the manuscript before submission.
improving the consistency of rapid-acting analog human/medicines/004046/human_med_002063.
insulin absorption and action for both subcutaneous jsp&mid=WC0b01ac058001d124 (2017). Competing interests statement
injection and continuous subcutaneous infusion using 130. Cengiz,E., Bode,B., Van Name,M. & C.M. serves or has served on the advisory panel for
recombinant human hyaluronidase. J.Diabetes Sci. Tamborlane,W.V. Moving toward the ideal insulin for AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb,
Technol. 6, 764772 (2012). insulin pumps. Expert Rev. Med. Devices 13, 5769 Eli Lilly and Company, Intrexon, Janssen Pharmaceuticals,
122. Krasner,A. etal. A review of a family of ultra-rapid- (2016). Hanmi Pharmaceuticals, Mannkind, Medtronic, Merck Sharp
acting insulins: formulation development. 131. Zaykov,A.N., Mayer,J.P. & DiMarchi,R.D. Pursuit of and Dohme Ltd., Novartis, Novo Nordisk, Pfizer, Sanofi,
J.Diabetes Sci. Technol. 6, 786796 (2012). a perfect insulin. Nat. Rev. Drug Discov. 15, 425439 Roche Diagnostics and UCB. KU Leuven has received
123. Pandyarajan,V. & Weiss,M.A. Design of non- (2016). research support for C.M. from Abbott, Eli Lilly and Company,
standard insulin analogs for the treatment of diabetes 132. Wang,Y., Shao,J., Zaro,J.L. & Shen,W.C. Proinsulin- Intrexon, Merck Sharp and Dohme Ltd., Novartis, Novo
mellitus. Curr. Diab. Rep. 12, 697704 (2012). transferrin fusion protein as a novel long-acting insulin Nordisk, Roche Diagnostics and Sanofi. C.M. serves or has
124. Heise,T., Pieber,R.R., Danne,T., Erlichsen,L. & analog for the inhibition of hepatic glucose served on the speakers bureau for AstraZeneca, Boehringer
Haahr,H. Faster onset and greater early exposure and production. Diabetes 63, 17791788 (2014). Ingelheim, Eli Lilly and Company, Merck Sharp and Dohme,
glucose-lowering effect with faster-acting insulin 133. Phillips,N.B., Whittaker,J., Ismail-Beigi,F. & Novartis, Novo Nordisk and Sanofi. P.G. has served on the
aspart versus insulin aspart: a pooled analysis in Weiss,M.A. Insulin fibrillation and protein design: advisory panel for AstraZeneca, Lilly, Merck Sharp and
subjects with type1 diabetes [abstract 929-P]. topological resistance of single-chain analogs to Dohme Ltd., Novo Nordisk and Sanofi. P.G. has served on the
American Diabetes Association (2016). thermal degradation with application to a pump speakers bureau for AstraZeneca, Bristol-Meyers Squibb,
125. Heise,T. etal. A comparison of pharmacokinetic and reservoir. J.Diabetes Sci. Technol. 6, 277288 (2012). Boehringer Ingelheim, Janssen Pharmaceuticals, Lilly,
pharmacodynamic properties between faster-acting 134. Chou,D.H. etal. Glucose-responsive insulin activity Novartis, Novo Nordisk and Sanofi. K.B. has served on the
insulin aspart and insulin aspart in elderly subjects by covalent modification with aliphatic phenylboronic advisory panel for AstraZeneca, Merck Sharp and Dohme
with type1 diabetes mellitus. Drugs Aging 34, 2938 acid conjugates. Proc. Natl Acad. Sci. USA 112, Ltd. and Novo Nordisk. K.B. has served on the speakers
(2017). 24012406 (2015). bureau for AstraZeneca, Bristol-Meyers Squibb, Boehringer
126. Fath,M. etal. Faster-acting insulin aspart provides 135. Baeshen,N.A. etal. Cell factories for insulin Ingelheim, Janssen Pharmaceuticals, Lilly, Novartis and Novo
faster onset and greater early exposure versus insulin production. Microb. Cell Fact. 13, 141 (2014). Nordisk. KU Leuven has received research grants for K.B.
aspart in children and adolescents with type1 136. Santos Cavaiola,T. & Edelman,S. Inhaled insulin: from AstraZeneca, Janssen Pharmaceuticals, Merck Sharp
diabetes mellitus. Pediatr. Diabetes http://dx.doi. a breath of fresh air? A review of inhaled insulin. and Dohme Ltd., Novartis, Novo Nordisk and Sanofi.
org/10.1111/pedi.12506 (2017). Clin. Ther. 36, 12751289 (2014).
127. Heise,T., Zijlstra,E., Nosek,L., Rikte,T. & Haahr,H. 137. Heinemann,L. Insulin pens and new ways of insulin Publishers note
Pharmacological properties of faster-acting insulin delivery. Diabetes Technol. Ther. 16 (Suppl. 1), Springer Nature remains neutral with regard to jurisdictional
aspart versus insulin aspart in patients with type1 S44S55 (2014). claims in published maps and institutional affiliations.
!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

Analogos de Insulina en DM 1 2017

Uploaded by

Copyright:

Available Formats

Analogos de Insulina en DM 1 2017

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analogos de Insulina en DM 1 2017

Uploaded by

Copyright:

Available Formats

REVIEWS

Insulin analogues in type1 diabetes

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 1

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

Key points the concentration of regular insulin also influences the

2 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 3

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

Dihexamers Hexamers Hexamers Dihexamers Hexamer Hexamers Hexamers and Hexamers

Multihexamer Concentrated Hexamer Dihexamers Hexamer Hexamers, Hexamers, Hexamers,

Very slow Very fast

Arginine and Insulin

4 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

40 involved patients carrying pumps administering insulin

30 Rapid-acting insulin aspart rapid-acting insulin analogues in the first 72 h; however,

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 5

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

of insulin glargine (particularly under steady-state con-

6 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 7

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

8 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 9

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

10 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 11

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

12 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 13

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

14 | ADVANCE ONLINE PUBLICATION www.nature.com/nrendo

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

NATURE REVIEWS | ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION | 15

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

You might also like

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"

!,(++- 4 +(2'#12 (,(3#" /13 .$ /1(-%#1 341# ++ 1(%'32 1#2#15#"