Comparison of Lys ,Pro -Human Insulin Analog and ... - Diabetes Care

S H O R T R E P O R T 

Comparison of Lys B28 ,Pro B29 -Human 

Insulin Analog and Regular Human 

Insulin in the Correction of Incidental 

Hyperglycemia 

FRITS HOLLEMAN, MD 

JOOPJ.G. VAN DEN BRAND, MSC 

ROLAND A.R.A. HOVEN, MSC 

Jos M. VAN DER LINDEN, MA 

INGEBORG VAN DER TWEEL 

JOOST B.L. HOEKSTRA, MD, PHD 

D. WLLLEM ERKELENS, MD, PHD 

OBJECTIVE — To obtain clinically applicable data on the effects of regular human insulin 

and the Lys B28 ,Pro B29 -human insulin analogue (lispro) on the correction of incidental hyperglycemia. 

RESEARCH DESIGN AND METHODS— The insulins were compared in a nonclamped 

randomized crossover study of 27 male IDDM patients. Hyperglycemia was induced 

by the withdrawal of the normal evening dose of insulin; the next morning patients fasted and 

received a single dose of study insulin according to a dosing nomogram. Blood glucose concentration 

and GR (a measure of glucose corrected for differences in administered insulin dose: 

GR = glucose concentration X BMI X insulin dose"" 1 ) were followed for 4 h. 

RESULTS — The time courses of blood glucose concentration and GR were significantly different 

after regular insulin in comparison with lispro (multiple analysis of variance, P < 

0.001). At t = 120 min, glucose concentrations had decreased 1.4 mmol/1 more with lispro than 

with regular insulin (95% confidence interval [CI] 0.6-2.3, P = 0.002). Similarly, GR had 

decreased 4.4 mol • kg • IU" 1 • m~ 5 more with lispro than with regular insulin (95% CI 

2.6-6.2, P < 0.001). The overall difference in glucose values was 0.87 mmol/1 (lispro < regular 

insulin, P = 0.036), and the overall difference in GR values was 1.96 mol • kg • 1U" 1 • m~ 5 

(lispro < regular insulin, P = NS). Unexpectedly, the intrinsic variability of GR was higher for 

lispro than for regular insulin. 

CONCLUSIONS — The more rapid action of lispro is an advantage in the correction of 

hyperglycemia, even though actual differences in glucose concentrations are smaller than 

suggested by previous clamped studies. 

The rapid and sufficient correction of 

incidental hyperglycemia is hampered 

by the long duration of action 

of regular human insulin and the inability 

to predict the effect of a certain amount of 

regular human insulin on blood glucose 

levels. The latter problem can largely be 

attributed to the high inter- and intra-individual 

variability in the absorption of reg- 

ular human insulin after subcutaneous 

injection (1-4). 

It would seem theoretically plausible 

that monomeric insulins, such as the 

Lys B28 ,Pro B29 -human insulin analog (lispro), 

show less variability in absorption than 

hexameric regular insulin (5). Such insulins 

could result in more clinically predictable 

behavior and possibly even a linear dose- 

From the Department of Internal Medicine (EH., J.J.G.B., R.A.R.A.H., J.B.L.H.), Diakonessenhuis, Utrecht; 

Eli Lilly and Company (J.M.L), Nieuwegein; Center for Biostatistics (IT.), Utrecht University, Utrecht; the 

Department of Internal Medicine (D.WE.), University Hospital, Utrecht, The Netherlands. 

Address correspondence and reprint requests to Frits Holleman, MD, Department of Internal Medicine, 

Diakonessenhuis, Bosboomstraat 1, 3582 KE Utrecht, The Netherlands. 

Received for publication 31 January 1996 and accepted in revised form 25 July 1996. 

(jR, relative measure of glucose; lispro, Lys B28 ,Pro B2c) -human insulin analog; MANOVA, multiple analysis 

of variance. 

effect relationship. The results of glucose 

clamp studies by Woodworth et al. (6) and 

Antsiferov et al. (7) seem to support this 

hypothesis. However, the practical applicability 

of these results is difficult to ascertain 

(5,8,9). 

This study was designed to compare 

the decrease in glucose concentrations 

after the subcutaneous administration of 

lispro and regular human insulin in a setting 

as close as possible to naturally occurring 

hyperglycemia. By using a well-controlled 

double-blind cross-over design, we 

tried to eliminate potential confounding 

factors. 

RESEARCH DESIGN AND 

METHODS — This study was approved 

by the Institutional Ethical Review Board. 

Informed consent was obtained from all 

participants. The trial was conducted 

according to the European Good Clinical 

Practice guidelines. 

A total of 27 healthy male IDDM 

patients aged 20-65 years with HbAlc 

Table 1—Dosing nomogram 

Glucose 

range 

(mmol/1) 

05-10 

10-15 

15-18 

18-21 

21-24 

24-27 

27-30 

Insulin dosage 

per kg/m 2 

0.1 

0.2 

0.3 

0.4 

0.5 

0.6 

0.7 

(NovoFine, Novo Nordisk, Bagsvaerd, Denmark). 

The pens were prefilled, blinded, 

and sealed by the hospital pharmacist. 

Venous blood samples were taken at t 

- 20, 40, 60, 90, 120, 150, 180, 210, and 

240 min postinjection. Whole blood glucose 

was measured using an APEC glucose 

analyzer (Danvers, MA). HbAlc was measured 

using high-performance liquid chromatography 

(Biograd, Anaheim, CA). 

Since the insulin dosages were determined 

by the nomogram, we decided to 

obtain a relative measure which expresses 

the fall in glucose in such a way that it corrects 

for the differences in the BMI and the 

administered dose. This relative measure 

of glucose, GR (mol • kg • IU" 1 • m~ 5 ), was 

calculated by dividing the glucose values 

(in millimoles per liter) at all time points 

by the insulin dose actually administered 

(in international units) and multiplying 

this figure with the measured BMI (in kilograms 

per meters squared). 

Data were statistically analyzed using 

the SPSS computer program. 

Differences between the time course of 

lispro and regular human insulin were evaluated 

using an analysis of variance with 

repeated measures. The within-patient variability 

was defined as the square root of the 

error variance after the correction for the 

decrease of the GR values in time. 

RESULTS— The participants had a 

median age of 34 years (range, 20-60 

years), a median duration of diabetes of 8 

years (range, 0-30 years), an average 

HbAlc of 7.7 ± 1.1%, and a BMI of 24.4 ± 

1.7 kg/m 2 . Thirteen patients were randomized 

to the lispro-regular insulin 

sequence, and fourteen were randomized 

to the regular-lispro insulin sequence. 

There were no significant differences in 

baseline characteristics. The average dose 

of insulin administered was 7.9 ± 3.5 IU 

2 

Absolute dose (IU) of insulin administered by BMI (kg/m ) 

20 21 22 23 24 25 26 27 

2 2 2 2 2 2 3 3 

4 4 4 5 5 5 5 5 

6 6 7 7 7 7 8 8 

8 8 9 9 10 10 10 11 

10 10 11 11 12 12 13 14 

12 13 14 14 14 15 16 16 

14 15 15 16 17 17 18 19 

of regular human insulin and 7.9 ± 3.7 IU 

of lispro. 

The initial glucose values for regular 

human insulin ([Glue], = 0, reg) an d lispro 

([Gluc]( = 0 us) were 16.7 ± 4.7 mmol/1 and 

16.4 ± 5.4 mmol/1, respectively. Four 

hours after injection, [Glue], = 240, a-g was 

8.7 ± 2.3 mmol/1, and [Gluc], = 24o, us was 

8.5 ± 2.5 mmol/1. The average values for 

glucose concentration are plotted against 

time in Fig. 1. 

The differences in glucose values 

between regular human insulin and lispro 

were dependent on time (multiple analysis 

of variance [MANOVA], P < 0.001). 

Overall, the average difference between 

[Gluc]reg and [GlucJiis from t = 0 min to t = 

240 min using MANOVA was 0.87 

mmol/1 (lispro < regular human insulin, 

95% CI 0.07-1.68, P = 0.036). The average 

fall in glucose concentration from t = 0 

min to t = 120 min was 4.8 mmol/1 for reg- 

24,0 

20,0 

16,0 

12,0 

I 4,0 

> 

0,0 

0 20 40 60 80 

Figure 1—Glucose concentrations (means 

according to the nomogram at t = 0 min. 

< 

4 

< 

Holleman and Associates 

ular human insulin and 6.2 mmol/1 for 

lispro (95% CI 0.6-2.3, P - 0.002). 

At each time point after t = 60 min, a 

good correlation coefficient was found for 

the individual fall in glucose concentration 

plotted against the dose of insulin administered 

per kilograms per meters squared 

(all r > 0.89 and P < 0.001) for both regular 

human insulin and lispro. 

The relative measure of glucose, GR 

(mol • kg • IU" 1 • m ~0, at t -•• 0 min was 

55.2 ± 12.5 mol • kg • IU ! • m ' for regular 

human insulin and 54.7 ± 10.2 mol • 

kg • IU" 1 • m~ 5 for lispro. At t - 240 min, 

GR,K8 was 30.3 ±11.9 mol • kg • IU ' • 

m~ 5 and GRiiLS was 31.6 ± 15.1 mol • kg • 

IU" 1 • m" 5 . The average values of GR are 

represented in Fig. 2. 

Again, differences between regular 

human insulin and lispro were timedependent 

(P < 0.001). Overall, the average 

difference between GRrc,, and GRj,s 

from t = 0 min to t = 240 min using 

MANOVA was 1.96 mol • kg • IU ' • m * 

(lispro < regular human insulin, 95% CI 

-3.25-7.18, P = 0.45). The average fall in 

GR from t = 0 min to t = 120 min was 14.9 

mol • kg • IU" 1 • m" i for regular human 

insulin and 19.3 mol • kg • IU ' • m ' for 

lispro (95% CI 2.6-6.2, P < 0.001). 

The within-patient variability of all 

values of GR from t = 0 min to t = 240 min 

was 2.68 mol • kg • IU" 1 • m ' for regular 

human insulin and 3.58 mol • kg • IU ' • 

m~ 5 for lispro. The within-patient variability 

of the values of GR from t - 0 min to t - 

120 min was 2.25 mol • kg • IU ' • m ' 

A 

4 

*- . 

100 120 140 

time (min.) 

• 

" . 

i 

< 

— • 

Li .pro 

Hi lmuli iR 

— — 

> 

^—+ , 

> 

- A - 

160 180 200 220 240 

SD) after the administration of the study insulins 

DIABKTES CARE, VOLUME 19, NUMBER 12, DECEMBER 1996 1427 

•

Comparison of regular human insulin and lispro 

70,0 

60,0 

^50,0 

40,0 

30,0 

20,0 

10,0 

0,0 

< 

i 

< 

A 

i I 

• 

• 

i 

< i 

- — 

A 

•—-. 

— • -^ ^ • . 

' * 

< > 

—• 

Li spro 

—* 

Hi imuli lR 

0 20 40 60 80 100 120 140 160 180 200 220 240 

time (min.) 

Figure 2—Values o/GR (means ± SD) after the administration of the study insulins according to the 

nomogram att = O min. 

for regular human insulin and 2.68 mol • 

kg • IU" 1 • m" 5 for lispro. 

The fall in glucose concentration at a 

given time point that would be the result 

of the administration of 1 IU insulin per 

kg/m 2 at t = 0 min can be expressed as 

AGR. The differential of the fall in GR 

against time, AGR/A,, is a representation of 

the in vivo time-action profile of the 

administered insulins (Fig. 3). 

CONCLUSIONS— In this study, we 

used a relative measure of glucose, GR, that 

expresses the decrease in glucose concentration 

per 1 IU/BMI. This mathematical 

conversion might not be justified if, as 

13 

0,25 

§0,05 

0,00 

L 

JA 

\ 

r-A V\ 

1 

1 

published observations indicate (6,10), 

increased doses of regular insulin result in 

relatively decreasing effects on glucose disposal. 

However, we found linear correlations 

between the administered dose per 

kg/m 2 and the fall in glucose for regular 

human insulin and lispro, supporting the 

feasibility of our approach. 

The total decrease in glucose concentration 

and GR over 240 min was similar 

for both insulins, though there seems to be 

a residual effect of regular human insulin 

that lasts beyond 240 min (Figs. 2 and 3). 

The time courses of regular human insulin 

and lispro were significantly different. As 

illustrated in Figs. 2 and 3, the effect per 

\ 

\ 

Ns 

-* 

Li spro 

—* 

H jmuli riR 

0 20 40 60 80 100 120 140 160 180 200 220 240 

time (min.) 

Figure 3—The differential curve of AGR against time, AGR/A,, representing the time-action profiles 

of the studied insulins. 

• 

• 

IU/BMI in the first 120-min period was 

significantly higher for lispro than for regular 

human insulin. Also, the peak effect of 

lispro, as reflected in AGR/A(, was about 

150% of the peak effect of regular human 

insulin. These results are compatible with 

previously published data on lispro 

(6,7,11-14). 

From a clinical point of view, the differences 

in the decrease of glucose 

between regular human insulin and lispro 

are modest: maximally ~1.4 mmol/1 at t = 

120 min. The overall difference during the 

whole time period was less (~0.9 

mmol/1), while the difference in GR (~2 

mol • kg • IU" 1 • m~ 5 ) was not even 

significant. Thus, it must be emphasized 

that large differences in insulin concentrations 

or glucose infusion rates found in 

previous clamp studies do not imply similarly 

large differences in glucose concentration. 

This may partly explain why the 

use of lispro, despite its more physiological 

insulin profile, has not yet led to major 

improvements in HbAlc. 

We unexpectedly found a higher 

within-patient variability for lispro than 

for regular human insulin, even when we 

restricted our analysis to the first 120-min 

period. 

Given the time frame and the randomization 

procedure used, these differences 

can hardly be attributed to the 

residual effects of the patients' own 

insulin. Maybe the uniformly monomeric 

state makes the diffusion of lispro more 

susceptible to subtle changes in local 

blood flow than the diffusion of regular 

insulin, which is buffered by the equilibrium 

mixture of monomeric and polymeric 

forms. However, given the conflicting 

results of Antsiferov (7), this point 

requires further investigation. 

In conclusion, as in most other studies, 

the lispro insulin analog showed a 

more rapid action profile than regular 

human insulin. Lispro resulted in lower 

overall glucose values, but had a higher 

within-patient variability than regular 

human insulin. The absolute differences in 

effect on glucose for regular human 

insulin and lispro were not very large. Still, 

the early termination of insulin action is an 

advantage in the rapid and safe correction 

of incidental hyperglycemia. 

Acknowledgments— This study was financially 

supported by the Dutch branch of Eli Lilly 

& Company, Nieuwegein, The Netherlands. 

1428 DIABETES CARE, VOLUME 19, NUMBER 12, DECEMBER 1996

References 

1. Binder C, Lauritzen T, Faber O, Pramming 

S: Insulin pharmacokinetics. Diabetes Care 

7:188-199,1984 

2. Lauritzen T: Pharmacokinetic and clinical 

aspects of intensified subcutaneous insulin 

therapy. PhD thesis. Copenhagen, Denmark, 

University of Copenhagen, 1985 

3. Skyler JS: Insulin pharmacology. Med Clin 

North Am 72:1337-1354,1988 

4. De Meijer PHEM, Lutterman JA, van't Laar 

A: The absorption of subcutaneously 

injected insulin. Neth] Med 34:210-227, 

1989 

5. Furler SM, Kraegen EW: Quantitative 

aspects of subcutaneous insulin absorption. 

Diabetic Med 6:657-665, 1989 

6. Woodworth J, Howey D, Lutz S, Santa P, 

Brady P, Bowsher R: [LYS(B28), 

PRO(B29)] human insulin: dose ranging 

versus Humulin R. Diabetologia 36 (Suppl. 

1):A155,1993 

7. Antsiferov M, Woodworth JR, Mayorov A, 

Ristic S, Dedov I: Lower within patient 

variability in post-prandial glucose-excursion 

with lispro insulin analog compared 

with regular insulin. Diabetes 44 (Suppl. 

1):214A, 1995 

8. Boyle PJ: Glucose clamp investigations: the 

ups and downs. Diabetes Care 17:239- 

241, 1994 

9. DeFronzo RA, Tobin JD, Andres R: Glucose 

clamp technique: a method for quantifying 

insulin secretion and resistance. Am 

J Physiol 237(3):E214-E223, 1979 

10. Heinemann L, Richter B: Clinical pharmacology 

of human insulin. Diabetes Care 16 

(Suppl. 3):90-100, 1993 

11. Ter Braak EWMT, Bianchi R, Erkelens 

DW: More suitable action profile for premeal 

treatment of Lys(B28)Pro(B29) 

Holleman and Associates 

insulin analogue than regular insulin. Diabetologia 

36 (Suppl. l):A30,1993 

12. Radziuk J, Davies J, Pye S, DiMarchi R, 

Shields J, Chance R: Subcutaneous 

absorption of fast-acting insulin analogs: 

kinetics and bio-effectiveness. Diabetologia 

35 (Suppl. 1):A3,1992 

13. Howey DC, Bowsher RR, Brunelle RL, 

Woodworth JR: [Lys(B28), Pro(B29)]human 

insulin: a rapidly absorbed analogue 

of human insulin. Diabetes 

43:396-402,1994 

14. Torlone E, Fanelli C, Rambotti AM, Kassi 

G, Modarelli F, Di Vincenzo A, Epifano L, 

Ciofetta M, Pampanelli S, Brunetti P, Bolli 

GB: Pharmacokinetics, pharmacodynamics 

and glucose-counterregulation following 

subcutaneous injection of the 

monomeric insulin analogue [Lys(B28), 

Pro(B29)] in IDDM. Diabetologia 37:713- 

720,1994 

DIABETES CARE, VOLUME 19, NUMBER 12, DECEMBER 1996 1429

Comparison of Lys ,Pro -Human Insulin Analog and ... - Diabetes Care

Create successful ePaper yourself

Delete template?

Save as template?