`New Directions in Drug
`Develop111ent: Mixtures,
`Development: Mixtures,
`Analogues, and Modeling
`Analogues, and Modeling
`
`JOHN A. GALLOWAY, MD
`JOHN A. GALLOWAY, MD
`
`Success m modem medical research is acheived wlieu bask and clinical infonnation
`Success in modern medical research is acheived when basic and clinical information
`about a given disorder converges, either intentionally or fonuitously, with the
`about a given disorder converges, either intentionally or fortuitously, with the
`availability of technology or other means to design and apply interventions for the
`availability of technology or other means to design and apply interventions for the
`disorder in question. A prime example is the discovery of insulin and its replacement
`disorder in question. A prime example is the discovery of insulin and its replacement
`in patients with IDDM in 1923. Seven decades later, the focus of diabetes manage(cid:173)
`in patients with IDDM in 1923. Seven decades later, the focus of diabetes manage-
`ment is on improvement in metabolic control to forestall the chronic complications
`ment is on improvement in metabolic control to forestall the chronic complications
`of the disease and improve the quality of life of patients with the disease. Metabolic
`of the disease and improve the quality of life of patients with the disease. Metabolic
`control is being addressed through the development of insulin analogues using
`control is being addressed through the development of insulin analogues using
`sophisticated techniques to understand "the chemistry of insulin and to modify it
`sophisticated techniques to understand the chemistry of insulin and to modify it
`using rDNA technology. The objective of these efforts is to simulate nonnal insulin
`using rDNA technology. The objective of these efforts is to simulate normal insulin
`secretion with subcutaneously injected agonists. Quality-of-life needs are being
`secretion with subcutaneously injected agonists. Quality- of-life needs are being
`addressed with delivery devices, insulin mixtures, and insulin analogues. Although
`addressed with delivery devices, insulin mixtures, and insulin analogues. Although
`none of these improvements parallel the discovery of insulin, they do provide an
`none of these improvements parallel the discovery of insulin, they do provide an
`optimistic outlook for patients with diabetes mellitus.
`optimistic outlook for patients with diabetes mellitus.
`
`W hen the progress made in the
`W hen the progress made in the
`
`tated the use of conventional intensive
`tated the use of conventional intensive
`treatment of diabetes mellitus
`insulin treatment programs and CSU or
`treatment of diabetes mellitus
`insulin treatment programs and CSII or
`during the past two decades is
`pumps. Acceptance of CSll has been
`during the past two decades is
`pumps. Acceptance of CSII has been
`considered, the needs and challenges for
`based largely on the belief, against which
`considered, the needs and challenges for
`based largely on the belief, against which
`no effective counter can be found, that
`the final decade of this century become
`the final decade of this century become
`no effective counter can be found, that
`evident. For instance, the introduction of
`optimizing blood glucose control is most
`evident. For instance, the introduction of
`optimizing blood glucose control is most
`efficiently achieved by simulating normal
`patient self-monitoring of blood glucose
`patient self-monitoring of blood glucose
`efficiently achieved by simulating normal
`insulin secretion.
`and laboratory measurement of glycosy(cid:173)
`insulin secretion.
`and laboratory measurement of glycosy-
`The use of intensified regimens
`lated hemoglobin called attention to the
`The use of intensified regimens
`lated hemoglobin called attention to the
`failure of commonly used regimens to
`has been limited by three significant fac(cid:173)
`failure of commonly used regimens to
`has been limited by three significant fac-
`tors. First, until very recently there has
`nonnalize blood glucose. The shoncom(cid:173)
`tors. First, until very recently there has
`normalize blood glucose. The shortcom-
`ings of these regimens led to and facili-
`been a paucity of incomrovenible evi-
`ings of these regimens led to and facili-
`been a paucity of incontrovertible evi-
`•••••••••••••••••••••••••••••••••••••••••••••••••••••••••
`From the Lilly Research Laboratories, Eli lilly, Indianapolis; and the Depanment of Medicine,
`From the Lilly Research Laboratories, Eli Lilly, Indianapolis; and the Department of Medicine,
`Indiana University School of Medicine, Indianapolis, Indiana.
`Indiana University School of Medicine, Indianapolis, Indiana.
`Address correspondence and reprint requests to John A. Galloway, MD, Lilly Research
`Address correspondence and reprint requests to John A. Galloway, MD, Lilly Research
`Laboratories, Mail drop, 2128, Lilly Corporate Center, Indianapolis, IN 46285 .
`Laboratories, Mail drop, 2128, Lilly Corporate Center, Indianapolis, IN 46285.
`IDDM, insulin-dependent diabetes mellitus; NIDDM, non-insulin-dependent diabetes mel(cid:173)
`IDDM, insulin-dependent diabetes mellitus; N1DDM, non-insulin-dependent diabetes mel-
`litus; CSU, continuous subcutaneous insulin infusion; DCCT, Diabetes Control and Compli(cid:173)
`litus; CSII, continuous subcutaneous insulin infusion; DCCT, Diabetes Control and Compli-
`cations Trial; HPI, human proinsulin: IGF-1, insulin-like growth factor-I; BHCP, biosynthetic
`cations Trial; HPI, human proinsulin; IGF-1, insulin-like growth factor-1; BHCP, biosynthetic
`human (-peptide; HGP, hepatic glucose production; 0/, coefficient of variation.
`human C-peptide; HGP, hepatic glucose production; CV, coefficient of variation.
`
`dence, based on controlled, long-term
`dente, based on controlled, long-term
`clinical trials, that lowering blood glu(cid:173)
`clinical trials, that lowering blood glu-
`cose will forestall development of the
`cose will forestall development of the
`chronic complications of diabetes. For(cid:173)
`chronic complications of diabetes. For-
`tunately, the results of the DCCT (1)
`tunately, the results of the DCCT (1)
`point unmistakably to the importance of
`point unmistakably to the importance of
`glycemic control in preventing the pro(cid:173)
`glycemic control in preventing the pro-
`gression of retinopathy, nephropathy,
`gression of retinopathy, nephropathy,
`and neuropathy in patients with IDDM.
`and neuropathy in patients with IDDM.
`Although the DCCT results are also rel(cid:173)
`Although the DCCT results are also rel-
`evant to NID D M, the findings of the U. K
`evant to NIDDM, the findings of the U.K
`Prospective Diabetes Study (2), which
`Prospective Diabetes Study (2), which
`will be available in 1994, should be par(cid:173)
`will be available in 1994, should be par-
`ticularly instructive regarding the bene(cid:173)
`ticularly instructive regarding the bene-
`fits and risks of glycemic control in
`fits and risks of glycemic control in
`NIDDM patients. It should be pointed
`NIDDM patients. It should be pointed
`out, however, that based on random, un(cid:173)
`out, however, that based on random, un-
`controlled trials and anecdotal experi(cid:173)
`controlled trials and anecdotal experi-
`ence, the impressions are extant that gly(cid:173)
`ence, the impressions are extant that gly-
`cemic control is very important in IDDM
`cemic control is very important in IDDM
`and useful in NIDDM. These notions will
`and useful in NIDDM. These notions will
`be supported by the results of the afore(cid:173)
`be supported by the results of the afore-
`mentioned controlled clinical trials. Sec(cid:173)
`mentioned controlled clinical trials. Sec-
`ond, glycemic control, particularly in in(cid:173)
`ond, glycemic control, particularly in in-
`dividuals with IDDM, imposes enonnous
`dividuals with IDDM, imposes enormous
`burdens on the patient's life-style and
`burdens on the patient's life-style and
`quality of life. Thus, those who choose to
`quality of life. Thus, those who choose to
`self-monitor their blood glucose are in a
`self-monitor their blood glucose are in a
`constant state of vigilance with respect to
`constant state of vigilance with respect to
`their well- being in general, and the pos(cid:173)
`their well-being in general, and the pos-
`sibility of hypoglycemia in particular.
`sibility of hypoglycemia in particular.
`For this reason, interventions or treat(cid:173)
`For this reason, interventions or treat-
`ment improvements that the person
`ment improvements that the person
`without diabetes may perceive as trivial
`without diabetes may perceive as trivial
`· or marginal may be highly useful co the
`or marginal may be highly useful to the
`patient with diabetes. Third, factors re(cid:173)
`patient with diabetes. Third, factors re-
`lated co compliance and the inability or
`lated to compliance and the inability or
`unwillingness of patients to make the
`unwillingness of patients to make the
`necessary quality- of-life sacrifices un(cid:173)
`necessary quality- of-life sacrifices un-
`doubtedly constitute the major impedi(cid:173)
`doubtedly constitute the major impedi-
`ment to normalization of blood glucose
`ment to normalization of blood glucose
`in most patients (3). Nonetheless, it is
`in most patients (3). Nonetheless, it is
`clear that certain properties of commer(cid:173)
`clear that certain properties of commer-
`cially available insulin preparations pre(cid:173)
`cially available insulin preparations pre-
`clude the simulation of normal insulin
`clude the simulation of normal insulin
`secretion and its metabolic benefits. For
`secretion and its metabolic benefits. For
`this reason, the principal manufacturers
`this reason, the principal manufacturers
`of insulin are engaged in vigorous com(cid:173)
`of insulin are engaged in vigorous com-
`petition to produce insulin agonists that
`petition to produce insulin agonists that
`can more closely replicate endogenous
`can more closely replicate endogenous
`insulin secretion or its effects.
`insulin secretion or its effects.
`
`16
`16
`
`DIABETES URE, VOLUME 16, SUPPLEMENT 3, DECEMBER 1993
`DIABETES CARE, VOLUME 16, SUPPLEMENT 3, DECEMBER 1993
`
`MYLAN EXHIBIT - 1029
`Mylan Pharmaceuticals, Inc. v. Bausch Health Ireland, Ltd. - IPR2022-00722
`
`
`
`Fed
`
`Fasted-
`
`100 —
`80 .V.
`60 .
`40 I
`20 I
`
`120
`1110
`100
`
`90 1 80
`
`170
`60
`614-21)14,7,14
`1-4,1
`8 2 5
`9
`I
`0
`81F
`
`Figure 1—Serum insulin and blood glucose re-
`sponses of normal subjects to a standard diabetic
`diet (30 kcal/kg) given as 2/7 with each main
`meal and 1/7 at bedtime. From Galloway and
`Chance (4). © by Elsevier.
`
`This paper reviews the rationale
`for and methods used to develop specific
`insulin analogues, and describes how
`certain insulin mixtures and analogues
`may have a positive effect on the quality
`of life of patients with NIDDM as well as
`those with IDDM.
`
`NORMAL INSULIN SECRETION
`AND ITS SIMULATION SY
`INJECTED "INSULINS"— As sug-
`gested above, the most efficient method
`for achieving metabolic control in IDDM
`patients is to simulate normal endoge-
`nous insulin secretion. It is therefore ap-
`propriate to identify the essential com-
`ponents of endogenous insulin secretion
`in the nondiabetic individual, review the
`deficiencies of presently available insulin
`preparations, and indicate what might or
`should be done to simulate it or its effects
`with insulin agonists manufactured by
`rDNA technology.
`As indicated in Fig. 1, normal
`insulin secretion consists of basal and
`meal-stimulated components. The func-
`
`tion of basal insulin secretion, which is
`usually in the range of 5-10 U/ml
`(30-60 pM) (Fig. 1), is to restrain HGP
`in the postabsorptive state. Because fast-
`ing blood glucose has an extremely high
`correlation with HGP and is the base on
`which prandial glycemia is added during
`the next 24 h, simulation of this phase of
`insulin secretion is highly desirable, par-
`ticularly in individuals with NIDDM. In
`Lilly clinical trials of patients with IDDM
`as well as NIDDM, the single blood glu-
`cose value that correlated best with gly-
`cohemoglobin was the value taken after
`an overnight fast and collected at the
`study site (4). The function of the other
`component of endogenous insulin secre-
`tion, the meal-stimulated phase, is to
`promote disposal of ingested nutrients,
`principally glucose, into the periphery.
`As indicated in Fig. 1, this phase, which
`is characterized by a rise in serum hor-
`mone to concentrations of 80-120 U/ml
`(480 to 720 pM), is exquisitely sensitive
`to, and therefore synchronized with, the
`rise in blood glucose in response to
`meals. In addition, once the meal glyce-
`mia has subsided, there is a prompt re-
`turn to the basal or postabsorptive level.
`The physiological importance of proper
`timing of the premeal insulin dose has
`been clearly demonstrated in both
`NIDDM (5) and IDDM (6).
`As suggested above, presently
`available insulins or insulin regimens are
`generally incapable of simulating the
`basal or meal-stimulated components of
`normal insulin secretion. For instance, as
`indicated in Fig. 2, even ultralente pork
`insulin, which presumably would behave
`like ultralente human insulin, is not truly
`a basal insulin, as it has peaks that occur
`15-20 h after injection. With respect to
`the meal-stimulated component, al-
`though the serum hormone profile and
`glucodynamic effects of neutral regular
`human insulin, which is the fastest-
`acting commercially available insulin,
`would be expected to simulate that of
`normal insulin secretion, Fig. 3 shows
`that it does not. Here it is clear that the
`peak effects of neutral regular human in-
`
`Galloway
`
`23
`22
`1 21
`
`I 18
`• 17
`16
`it 15
`114
`e 13
`i 12
`11
`
`O
`
`8 18 24 303640 48 52 56 60
`
`64 68 72
`
`Insulin
`Inliketsd
`
`Time )hour.)
`
`23
`22
`21
`'
`183 17
`s.
`I
`tu 11
`8
`
`14
`13
`
`O 2 4 6 8 10 12
`20
`16
`Time [hours)
`
`24
`
`28
`
`32
`
`36
`
`Insulin
`Injected
`
`Figure 2—Mean serum free insulin responses
`of 6 IDDM patients to ultralente pork insulin
`(0.8 U/hg subcutaneously). Patients were main-
`tained eugbIcemic with a glucose clamp tech-
`nique. A Responses to dosing at 0, 24, and 48 h.
`B: The response over 36 h to a single dose at time
`0. Courtesy of J.E. Gerich.
`
`sulin, 0.2 U/kg subcutaneously, do not
`occur until 3 to 4 h after injection and
`are present as long as 8 h. The adminis-
`tration of regular insulin by CSII would
`seem to be an ideal method for mimick-
`ing normal insulin secretion. However,
`Fig. 4 shows that patients with IDDM on
`CSII treated with basal-bolus programs
`have higher than normal serum insulin
`concentrations throughout the day and
`night, probably as a result of the long-
`acting nature of the premeal bolus infu-
`sions of neutral regular insulin.
`The discussion above has focused
`on the acute benefits of simulating nor-
`mal insulin secretion on glucose metab-
`olism. However, because conventional
`insulin treatment usually results in aver-
`age serum insulin concentrations 2. •4
`times normal (7), and hyperinsulinemia
`
`DIABETES CARE, VOLUME 16, SUPPLEMENT 3, DECEMBER 1993
`
`17
`
`
`
`New directions in drug development
`
`Iwo
`
`6
`800
`S
`C . 600
`00
`200
`
`Hunk.lin A A,
`
`Hyrnuiln R SC
`
`3
`
`4
`Hour&
`
`5
`
`Figure 3—Pharrnacodynamic responses of 6
`normal volunteer subjects to neutral regular hu-
`man insulin rVNA (Humulin 170 0.2 U/hg by
`bolus intravenous infusion and by subcutaneous
`injection. Subjects were kept eugb,cemic with a
`Biostator. Note that the peak effect of subcuta-
`neous Humulin occurs after 3 h and that glucose
`is still being infused at 8 h after the subcutaneous
`injection. From Galloway (8). by the American
`Diabetes Association.
`
`may be a significant risk factor for coro-
`nary heart disease (8,9), an important
`long-term advantage to simulating nor-
`mal insulin secretion may be reduction
`in chronic hyperinsulinemia.
`
`METHODS TO DEVELOP INSULIN
`ANALOGUES— With the deficien-
`cies of conventional insulins described,
`the discussion now turns to methods
`used in the development of improved
`
`too
`
`_
`77
`2860
`
`80
`
`CSII
`COnsenlional Insulin Irealmenl
`
`40
`
`20
`
`2
`
`7 9 11 13 15 17 19 21 23 1 3 5
`Clock Time (hours)
`
`Figure 4—Mean serum free insulin concen-
`trations in 5 patients with 1DDM who received
`pork insulin either by CS11 or with conventional
`twice-daily regimens of NPH and neutral insulin
`at 0730 and 0430. With CS11, meal boluses were
`given at 0700 and 1100 and at 1630 and 2130.
`Basal insulin was 50-60% of the total daily dose.
`From Galloway and Chance (4). t) by Elsevier.
`
`"insulins," or insulin analogues. (It
`should be noted that the first analogue of
`human insulin was beef insulin, which
`has been in use for 70 years.) Basically,
`two approaches can be used in develop-
`ing insulin analogues. The first is to iden-
`tify an "insulin" that has been found to
`have attractive pharmacological proper-
`ties and adapt it for human use. An ex-
`ample of this process is the selection and
`development of HPI (10). Studies in an-
`imals and humans demonstrated that
`pork proinsulin was a soluble, interme-
`diate- acting insulin agonist that ap-
`peared to be relatively hepatospecific.
`Thus, in proinsulin there seemed to be
`the possibility of developing an interme-
`diate-acting insulin agonist that was free
`of protamine or excess zinc, and with
`effects on the liver that might be
`uniquely efficacious in NIDDM, a disor-
`der characterized by excessive HGP
`among other defects (8,11). Accordingly,
`HPI was produced by rDNA technology
`and clinical studies were undertaken.
`Pharmacology studies in dogs,
`normal volunteer subjects, and patients
`with NIDDM demonstrated that, com-
`pared with insulin for equivalent sup-
`pression of HGP, the effect of HPI on
`peripheral glucose disposal was less than
`that of insulin; that is, consistent with
`the finding with pork proinsulin, HPI
`was relatively hepatospecific (10). Al-
`though the differences between HPI and
`insulin were clear and distinct in all
`studies, they were quantitatively un-
`impressive. However, an unexpected
`pharmacological finding was that the in-
`trapatient CV in serum hormone concen-
`trations and glycemic response was
`—50% of that observed with insulin. De-
`spite these apparent pharmacological
`benefits, controlled clinical trials failed to
`demonstrate better glycemic control in
`the HPI-treated group (10).
`Of note is that as the multicenter
`clinical trials were nearing completion, a
`report occurred indicating that HPI was
`more active when injected into the ab-
`domen than into the arm or thigh, sug-
`gesting the possibility of partial uptake
`
`by the portal vein with direct access to
`the liver. The potential importance of
`this finding is evidenced by another pre-
`liminary report that showed that HPI was
`equipotent with human insulin when ad-
`ministered intraperitoneally (12). Al-
`though injection sites were random in
`the multicenter trials, at one center
`where the investigator used just the ab-
`dominal site, HPI was found to be supe-
`rior to insulin (13). However, before an
`appropriate study to evaluate the impor-
`tance of injection sire on the bioactivity
`of HPI could be undertaken, a potentially
`significant toxicity issue arose. Specifi-
`cally, in one study of insulin-naive pa-
`tients, in which —70 were randomized to
`insulin and 70 to HPI, there were 6 myo-
`cardial infarctions, including 2 deaths, in
`the HPI-treated group after 1 year of
`treatment, but none in the insulin-
`treated group (10). This finding, com-
`bined with the failure to demonstrate
`improved metabolic control with HPI
`versus insulin across the multicenter
`clinical trials, led to the suspension of
`development of HPI. Although a causal
`relationship between the use of HPI and
`the cardiovascular events was not estab-
`lished, it should be noted that because
`the hypoglycemic potency of HPI was
`only 4,0 to 5.7 U/mg, serum hormone
`concentrations in HP1-treated patients
`were exceedingly high—up to 1000
`times normal. Nonetheless, experience
`with HPI combined with an expanding
`literature linking the possible mitogenic
`effects of hyperinsulinemia to atheroscle-
`rosis (9,10,14) has clearly added the is-
`sues of mitogenicity, including both
`atherogenicity and, by implication, car-
`cinogenicity, to the developmental agenda
`for insulin analogues. Indeed, in vitro
`studies have suggested that certain insulin
`analogues have greater effects (binding to
`vascular smooth muscle and/or cell prolif-
`eration) than human insulin or HP1 (15).
`Because of the attractiveness of an
`intermediate-insulin agonist that would
`be soluble and contain no excess zinc,
`researchers proceeded with the develop-
`ment of a normal metabolite of HPI, des
`
`18
`
`DIABETES CARE, VOLUME 16, SUPPLEMENT 3, DECEMBER 1993
`
`
`
`62
`
`61
`SER
`LEU
`GLN
`
`-COOH
`
`63
`
`61
`
`•44
`
`SIN
`
`CIA
`
`nn.—
`
`Mena.*
`
`me•ma
`
`NH,- Al
`
`A2
`
`A3
`
`64
`
`65
`
`NHi Al
`
`A2
`
`A3
`
`62
`
`De
`
`63e
`
`-COOH
`
`Galloway
`
`35
`
`34
`
`NH2-
`
`33
`
`-COOH
`
`THR
`
`LYS
`PRO
`29
`28
`
`GLU
`
`35
`
`NH
`
`ALA
`
`34
`
`GLU
`
`33
`-COOH
`) 32
`
`A
`A
`
`31
`
`30
`
`THR
`LYS
`29
`
`PRO
`28
`
`Figure 5—Metabolites or intermediates formed by the conversion of HPI to insulin. The des 64,65 HP1 (dPRO) form (shown upper left) is the result
`of a split between 65 and 66 (or Al and C65), followed by removal of the Arg and Lys residues at positions 64 and 65. From Galloway et al. (10).
`(`) by the American Diabetes Association.
`
`64,65 HPI (des 64,65 HPI, or "dPRO")
`(Fig. 5). Data generated in humans with
`this analogue are very preliminary,
`whereas studies in dogs (16) suggest that
`des 64,65 HPI is intermediate acting and
`has a hypoglycemic potency comparable
`to that of insulin. Therefore, serum hor-
`mone concentrations after administra-
`tion of des 64,65 HPI should be signifi-
`cantly lower than those seen after HPI.
`The fact that animal studies comparing
`des 64,65 HPI with insulin and HPI (17)
`have disclosed that the hepatic extraction
`coefficient and hepatic half-life of the des
`64,65 analogue is closer to that of insulin
`than HPI reduces the likelihood that the
`des 64,65 analogue will be hepatospe-
`cific. On the other hand, the consistent
`absorption kinetics seen with HPI would
`be expected with des 64,65 HPI.
`The second method for develop-
`ing improved insulins is that of com-
`puter modeling, which is usually per-
`
`formed by a computational chemist. This
`is a multicomponent technique that uses
`a computer programmed with physico-
`chemical and biological information de-
`veloped on insulin over several decades.
`Computer modeling allows the chemist
`to generate virtually infinite iterations of
`the possibilities of intramolecular rela-
`tionships to explain or extrapolate the
`behavior of insulin or its analogues un-
`der various conditions without actually
`undertaking arduous and expensive pre-
`clinical or clinical tests (18-20). This
`use of a computer can generate vital
`information concerning stability, self-
`association, and pharmacological activ-
`ity, including receptor binding, mitoge-
`nicity, and immunogenicity.
`Modeling (computational chem-
`istry) is usually used in conjunction with
`experimental chemistry. An example is
`the development of the Lys,Pro analogue
`of human insulin, a fast-acting insulin
`
`analogue that has been synthesized in the
`Lilly Research Laboratories (5,21). For
`instance, the long-acting nature of neu-
`tral regular insulin (described above) has
`been understood to be attributed to its
`tendency to self-associate into dimers,
`tetramers, hexamers, and polymers, with
`absorption of subcutaneously injected
`insulin occurring only after it has disso-
`ciated into a less aggregated form (6). It
`was noted that 1GF-I has many structural
`similarities to insulin in the COOH-
`terminus of the B-chain but does not
`self- associate. Therefore, attention was
`focused on an important difference be-
`tween the two polypeptides B28 and B29
`on the B-chain, where the sequence is
`Pro,Lys for insulin and Lys,Pro for IGF-I
`(22). Using molecular modeling, with
`help from a Cray 2 supercomputer, it
`was possible to visualize and compre-
`hend the dynamics of the intramolecular
`behavior that results in self-association.
`
`DIABETES CARE, VOLUME 16, SUPPLEMENT 3, DECEMBER 1993
`
`19
`
`
`
`New directions in drug development
`
`Lys (B28), Pro (B29) — Human Insulin
`
`A-CHAIN
`
`3
`
`4
`
`B-CHAIN
`
`5
`
`2
`
`3
`
`6
`
`7
`
`S
`
`10
`
`11
`
`21
`
`20
`
`18
`
`17
`
`11
`
`12 13
`
`14
`
`30
`
`PRO
`29
`LYS
`28
`
`27
`
`25 26
`
`3
`
`4
`
`15
`
`17
`
`18 19 20
`
`Figure 6 —Lys(B28), Pro(B29)-Human insulin. This analogue is identical to human insulin except
`at positions B28 and B29 where the sequence of the two residues has been reversed and are in the same
`order as in IGF-1, a polypeptide which does not self associate. Courtesy of R.E. Chance.
`
`It was ascertained that the reversal of the
`Pro,Lys sequence in the B28-29 region
`of insulin (Fig. 6) increased the free-
`energy barrier to self-association. Mod-
`eling also predicted that the Lys,Lys an-
`alogue would be close to native insulin in
`self-association and showed that the po-
`sition of the Pro in the sequence was
`critical to controlling self-association. As
`indicated first in studies in dogs (23)
`(Fig. 7), in normal humans (24) the ab-
`sorption and glucodynamic effects of the
`Lys,Pro analogue of human insulin,
`Lys,Pro human insulin, are significantly
`more rapid in onset that those of human
`insulin rDNA (Fig. 6). Of interest is the
`fact that the slight change in the strut
`cure of human insulin had an insignifi-
`cant effect on affinity for IGF-1 receptors
`or insulin receptors on placental mem-
`branes (21,25). Thus, it would appear
`that the Lys-Pro analogue and other fast-
`acting analogues (6,26) will have the
`pharmacological properties necessary for
`producing a rapid, timely, and adequate
`increase in serum hormone concentra-
`tion to promote disposal of ingested nu-
`trients efficiently. Table 1 shows the pro-
`
`file of an ideal fast-acting insulin
`analogue.
`An issue not addressed above is
`that of immunogenicity. The importance
`of lack of immunogenicity is based on
`studies in patients with and without an-
`tibodies to animal insulins, which have
`demonstrated that antibodies may delay
`the effect of injected insulin (27). This
`being the case, if a fast-acting analogue is
`immunogenic, antibodies could be
`formed that would delay and therefore
`
`Table 1—Features of an ideal fast-acting
`insulin analogue
`Primary metabolic function
`Promotion of peripheral glucose disposal
`Specifications
`Time-action profile:
`Onset <1 h after subcutaneous
`injection
`Duration of <4 h
`Metabolic > > > mitogenic effects
`Nonimmunogenic
`Chemically stable
`Mixable with long-acting insulin and
`insulin analogues
`
`neutralize the pharmacodynamic effect it
`was designed to deliver. One can only
`speculate on the potential importance of
`the proimmunogenic effects of episodic
`exposure, a sine qua non of fast-acting
`analogue use, and the anti-immunogenic
`effects of reducing the size of the peptide
`that might stimulate antibody formation.
`As indicated above, based on the
`pharmacokinetic shortcomings of ul-
`tralente insulin and the suspension of
`trials with HPI, a significant clinical need
`exists for a basal insulin agonist. Table 2
`lists the features identified as being de-
`sirable in the ideal long-acting insulin
`analogue. Novo-Nordisk has reported an
`analogue (NovoSol basal) (28) with ab-
`sorption and duration of action appar-
`ently substantially slower and more con-
`sistent than that of ultralente beef insulin
`and without peaks. However, the ab-
`sence of additional information on this
`analogue precludes further discussion.
`Because of the extraordinary challenge of
`
`80
`70
`E 60
`5 4, 50
`40
`6 30
`2
`
`10
`Ji
`0
`-60 -30
`0
`
`110-
`.100-
`2
`90-
`.6 80-
`4•5i 70-
`g BO-
`?
`
`‘.5 50:
`
`• LYSZIPP0204.ul.n. Noe
`• Nouirol moults, HI IONS
`
`-
`30 60 00 120 150 190 210 240 270
`MInuIes
`
`101'
`
`40/
`0
`-60 -30 0
`
`• elS284140194, 44.4...8
`
`A Houtral repukt/ NI ,DNA
`
`30 60 90 120 150 180 210 240 270
`Minutes
`
`Figure 7—Serum insulin and blood glucose re-
`sponses of normal fasted anesthetized male bea-
`gle dogs after subcutaneous administration of 0.1
`U/kg of neutral regular human insulin rDNA or
`the Lys,Pro analogue of human insulin. From
`Galloway et al. (23). © by Elsevier.
`
`20
`
`DIABE-rFs CARE, VOLUME 16, SUPPLEMENT 3, DECEMBER 1993
`
`
`
`Gallaway
`
`Table 2—Features of an ideal long-acting
`insulin analogue
`Primary metabolic function
`Suppression of HGP
`Specifications
`Time-action profile:
`Onset >4 h after subcutaneous
`injection
`Duration of >24 h
`Once-daily injection
`Low intrasubject CV of response
`Metabolic >>> mitogenic effects
`Hypoglycemic potency = or > human
`insulin
`Nonimmunogenic
`Chemically stable
`Mixable with short-acting insulin and
`insulin analogues
`
`developing an analogue to meet the spec-
`ifications listed in Table 2, it seems un-
`likely that such a compound will be
`available in the immediate future. In ad-
`dition, the possibility of developing a
`long-acting analogue that could simulate
`the pulsatility of normal-basal insulin se-
`cretion (29) seems exceedingly remote.
`
`HUMAN C-PEPTIDE— A substance
`prepared by rDNA (30) that has attrac-
`tive physiological properties, as opposed
`to HP1, which has desirable features but
`only in pharmacological doses, is BHCP.
`Although C- peptide is cosecreted from
`normal a in quantities that are equimolar
`with insulin and is present in the circu-
`lation, until recently, no physiological
`function had been assigned to this pep-
`tide. For this reason, the C-peptide de-
`ficiency that accompanies insulin defi-
`ciency in 1DDM has been regarded as
`having no consequence. Recently, inves-
`tigators in Stockholm reported that acute
`infusions of BHCP to mimic normal
`plasma concentration of C-peptide in
`C- peptide-negative patients with IDDM
`reduced glomerular hyperfiltration with-
`out affecting renal plasma flow (31). In a
`preliminary report, this group indicated
`that the addition of BHCP, administered
`for 30 days by CSII in amounts equimo-
`
`lar with insulin in patients with IDDM,
`reduced glomerular hyperfiltration and
`capillary leakage (measured by vitreous
`fluorophotometry) and possibly im-
`proved metabolic control (32). In vitro
`studies have shown that BHCP promoted
`the transport of 3-O-methylglucose in a
`dose-dependent manner in muscle spec-
`imens taken from healthy volunteer sub-
`jects (33) but not in muscle from
`NIDDM patients (34). These findings, if
`confirmed, could have important impli-
`cations for treatment of patients with
`IDDM. However, because C- peptide
`concentrations usually are normal or el-
`evated in patients with NIDDM (35),
`C-peptide replacement therapy is un-
`likely to be of benefit. Nonetheless, if the
`findings in IDDM are confirmed, serious
`consideration will be given to adding
`BHCP to insulin for patients with IDDM.
`
`QUALITY-OF-LIFE ISSUES —
`The above discussion has focused on
`how improvements in the pharmacoki-
`netics and/or pharmacodynamics of in-
`sulin preparations might result in im-
`proved metabolic control, which, in
`turn, may reduce the complications of
`diabetes. No less important to the patient
`with diabetes are interventions and tech-
`niques that will improve his or her qual-
`ity of life. Undoubtedly, the most impor-
`tant tool in recent years for improving
`the quality of life of patients with diabe-
`tes mellitus has been self-monitoring of
`blood glucose. Two other developments,
`however, also deserve mention: pen in-
`jectors and premixtures of NPH and reg-
`ular insulin. The availability of these
`penlike devices, which accurately deliver
`insulin (regular alone or in combination
`with NPH) from a prefilled cartridge,
`give patients a simple, convenient
`method for delivering insulin without
`having to carry insulin vials (36). These
`devices undoubtedly have improved
`quality of life for patients by facilitating
`multiple-dose insulin therapy (37,38).
`Although occasional reports indicate that
`use of pen injectors improves metabolic
`control (39), in general, metabolic con-
`
`trol with this device has not been shown
`to be superior to that achieved with in-
`tensive conventional treatment. None-
`theless, the convenience afforded to pa-
`tients by pen injectors is undeniable.
`The availability of insulin mix-
`tures, especially 70% NPH and 30% reg-
`ular insulin, has provided patients with a
`convenient method for taking two insu-
`lins and has obviated errors inherent in
`the multiple-step procedure of self -
`mixing. The use of premixtures of insulin
`may or may not improve metabolic con-
`trol. Bell et al.