The Intravenous, Intraperitoneal, and
`Subcutaneous Routes of Insulin
`Delivery in Diabetic Man
`
`DAVID S. SCHADE, R. PHILIP EATON, NEAL FRIEDMAN, AND WILLIAM SPENCER
`
`SUMMARY
`Successful implantation of an artificial pancreas re-
`quires the infusion of insulin into an appropriate ana-
`tomic site. Three sites being actively investigated in-
`clude (1) intravenous (i.v.), (2) intraperitoneal (i.p.),
`and (3) subcutaneous (s.c.). This study compared the
`rate, magnitude, and duration of insulin absorption
`from these three absorption sites as assessed by the
`appearance of "free" insuljn into the plasma of 10 in-
`sulin-dependent diabetic subjects. The biologic effec-
`tiveness of insulin was assessed by the suppression
`of plasma glucose concentration following a 750-calo-
`rie meal.
`Our results suggest that i.v. delivered insulin pro-
`vides the most rapid increase in plasma free insulin
`concentration, followed by the i.p. and s.c. routes, re-
`spectively. In contrast, the elevation of plasma free in-
`sulin concentration was most prolonged with the s.c.
`route, followed by i.p. and i.v. routes, respectively.
`Compared with the i.v. and s.c. routes of insulin deliv-
`ery, only 50% of the i.p. delivered insulin appeared in
`the plasma. The onset of the biologic activity of the in-
`sulin delivered by the three different routes during the
`41/2-h observation period was most rapid for the i.v.
`and least rapid for the s.c. route. These results sug-
`gest that all three routes may be appropriate sites for
`delivery of insulin from an artificial pancreas.
`However, because of the difference in absorption ki-
`netics and the onset of biologic effectiveness of the
`delivered insulin, different quantities and timing of in-
`sulin delivery may be needed. DIABETES 28:1069-1072,
`December 1979.
`
`T he successful development and implantation of an
`
`artificial insulin delivery system into diabetic man
`requires an appropriate site for the infusion of insu-
`lin. The major anatomic sites being actively inves-
`
`From the University of New Mexico School of Medicine, Albuquerque, New
`Mexico, and Sandia Laboratories, Livermore, California.
`Address reprint requests to David S. Schade, M.D., University of New Mexico
`School of Medicine, Albuquerque, New Mexico 87131.
`Received for publication 19 July 1979.
`
`tigated include (1) intravenous (i.v.),1'2 (2) intraperitoneal
`(i.p.),3 and (3) subcutaneous (s:c.).4t5 Although there are ad-
`vantages and disadvantages to each administration site, to
`date no studies are available that compare the rate and de-
`gree of insulin absorption as measured by the appearance
`of insulin into the peripheral circulation.
`This study examined the increase in plasma free insulin
`in diabetic subjects receiving a prepro:
`concentration
`grammed infusion of exogenous insulin. Regular insulin (40
`U/ml) was infused into either the i.v., i.p., or s.c. site by
`means of a portable, rotary solenoid microliter-infusion
`pump. An assessment of the rapidity, magnitude, and dura-
`tion of insulin absorption from these three sites was made
`using a randomized paired-study protocol in 10 diabetic
`men lacking endogenous insulin secretion (no C-peptide
`secretion). In addition, the change in plasma glucose con-
`centration following ingestion of identical meals was also
`examined as a measure of the onset of biologic activity of
`the insulin delivered during the 41/2-h observation period.
`
`METHODS
`Subject population. Ten healthy, normal weight, male, in-
`sulin-dependent diabetic subjects participated
`in
`trie
`131/2-h studies. Ages ranged from 22 to 39 yr and all sub-
`jects were receiving either one or two injections per day of
`isophane insulin with or without the addition of regular insu-
`lin. On the evening before study, the subjects were admitted
`to the Clinical Research Center of the University of New
`Mexico School of Medicine. No subject received insulin
`between 7:30 p.m. and the start of the studies at 7:30 a.m.
`the next day. Before initiating these studies,
`informed
`written consent was obtained from each volunteer, and all
`studies were approved by the University of New Mexico
`Human Research Review Committee.
`Protocol. On the day of study, each subject received three
`identical meals, which corresponded in time to breakfast,
`lunch, and supper at 7:30 a.m.; noon, and 4:30 p.m. Each
`meal consisted of ham, fruit, bread, and milk comprising
`750 calories (40% carbohydrate, 40% fat, and 20% protein).
`
`DIABETES, VOL. 28, DECEMBER 1979
`
`1069
`
`MYLAN INST. EXHIBIT 1052 PAGE 1
`
`

`

`INSULIN DELIVERY ROUTES
`
`Between meals, free access to water was allowed but no
`food was ingested.
`The 131/2-h protocol was divided into three equal'time
`periods of 41/2 h each, during which each subject received
`one i.v., one i.p., and one s.c. insulin infusion study in ran-
`domized sequence. The protocol used in all studies em-
`ployed an initial 30-min high dose insulin infusion of 5 U of
`regular insulin beginning simultaneously with the start of the
`meal. Following the 30-min high dose insulin infusion, the
`infusion pump reverted to the low dose insulin infusion rate
`of 2 U/h for three additional hours. At the end of the 31/? h, the
`insulin delivery catheter was removed. The subsequent 60-
`min "no insulin" period permitted an assessment of contin-
`ued insulin absorption from the infusion site.
`All insulin protocols used a mini, peristaltic rotary sole-
`noid pump to deliver the insulin.6 This pump was portable
`(6.0 x 7.5 x 10 cm) and battery operated. The pump was
`driven in a pulsatile fashion by a rotary solenoid, each pulse
`rotating the pump head by 10 degrees and delivering a vol-
`ume of 2 /xl/pulse. Reproducibilityof the pump was ±5% for
`a delivery volume of 72 fx\. In all infusion studies, the insulin
`was stored in a 50-ml plastic reservoir connected by a 10-
`cm plastic tube to the rotary solenoid pump. Before starting
`the individual infusion studies, 1 ml of the U40 insulin in the
`reservoir was pumped through the plastic tubing to saturate
`insulin absorption sites.
`Infusion technique, insulin in all three studies was deliv-
`ered through a 22-gauge plastic Intracath (Deseret Pharma-
`ceutical Company, Sandy, Utah), 20 cm in length, attached
`to the insulin pump via a 40-cm connecting tube. In the i.v.
`delivery of insulin study, the 19-gauge trocar was inserted
`into an antecubital vein, permitting placement of the plastic
`catheter approximately 5 cm into the vein.
`In the i.p. insulin infusion studies, the 22-gauge Intracath
`was inserted 3 cm below the umbilicus into the i.p. space
`through a locally anesthetized area. The flexible plastic In-
`tracath catheter was "threaded" through a 19-gauge trocar
`into the peritoneal space to a depth of approximately 5 cm.
`This catheter was not rigid enough to pierce either a blood
`vessel or the wall of the intestine. Once the catheter was
`threaded through the trocar, the trocar was removed and
`taped to the abdominal skin. Sterile technique was used
`throughout the procedure. Following the 31/2-h peritoneal in-
`fusion, the catheter was removed and cultured. In the 10
`subjects undergoing this procedure, no complications or in-
`fections were encountered. Our experience indicates that
`this procedure is both safe and painless and agrees with
`published reports in the literature on the safety of this tech-
`nique.7
`In the s.c. insulin infusion study, an identical Intracath
`was placed subcutaneously approximately 4 cm above the
`umbilicus. The trocar was inserted approximately 5 cm into
`the s.c. tissue and then withdrawn, leaving the plastic cath-
`eter approximately 4 cm within the s.c. space.
`Assay methodology. Plasma glucose concentration was
`measured by glucose oxidase and plasma free insulin by
`double antibody radioimmunoassay following precipitation
`with 25% polyethylene glycol to remove endogenous insulin
`antibodies as previously described.3 C-peptide was as-
`sayed with a kit from Calbiochem (Van Nuys, California).8
`Statistical assessment. The changes in plasma free.insu-
`lin concentration in these diabetic subjects were assessed
`
`125
`
`off
`
`— intravenous delivery
`— intraperitoneal
`••
`--- subcutaneous
`••
`
`Figure 1. The changes in plasma free insulin concentration (top) and
`glucose concentration (bottom) during programmed delivery of 11 U of
`regular insulin by three different routes in diabetic man. The infusion
`of insulin (U40-regular) was begun at a high dose (10 U/h) at time 0
`and continued for 30 min, when the infusion pump automatically
`reverted to the low dose (2 U/h) for an additional 3 h. Simultaneously
`with the start of the high dose insulin infusion, each of the 10 diabetic
`subjects began to eat a 750-calorie meal (arrow). At the end of the
`infusion period (31/2 h) the insulin infusion was stopped for 60 min
`(off). Each of the 10 diabetic subjects participated in an intravenous,
`intraperitoneal, and subcutaneous insulin infusion study in a
`randomized sequence. The mean ± SEM of these 10 subjects is
`shown in the figure.
`
`by frequent venous blood sampling from the arm contra-
`lateral to that which received the i.v. insulin. Four-milliliter
`blood samples were withdrawn via a #19 scalp vein needle
`every 15 or 30 min throughout the study, as shown in Figure
`1. Patency of this scalp vein needle was maintained by con-
`tinuous infusion of isotonic saline (1 ml/min).
`The mean ±SE for insulin and glucose for all studies are
`plotted in Figure 1. To quantitate the amount of insulin ap-
`pearing
`in the peripheral circulation during the 4!/2-h
`period, the increase in insulin concentration above the ini-
`tial starting concentration at time 0 was assessed by inte-
`gration. To estimate the rapidity of onset of biologic effec-
`tiveness of the administered insulin, the increase of plasma
`glucose concentration over baseline levels was compared
`with Student's t test for paired data.
`
`RESULTS
`Mean plasma free insulin concentration (Figure 1). In the
`i.v. insulin infusion study, mean plasma free insulin concen-
`tration rose to a maximum of 119 ± 13 /xU/ml by 15 min
`post initiation of infusion and then decreased rapidly fol-
`lowing termination of the high dose infusion to an approxi-
`mate background level of 30 ^U/ml. When the insulin infusion
`
`1070
`
`DIABETES, VOL. 28, DECEMBER 1979
`
`MYLAN INST. EXHIBIT 1052 PAGE 2
`
`

`

`DAVID S. SCHADE AND ASSOCIATES
`was stopped at 31/2 h, the plasma free insulin concentration
`decreased further to approximately 18 jLtU/ml, 5 /u,U/ml
`below the initial starting basal free insulin concentration of
`22 ± 3/LiU/ml (P < 0.01).
`In the i.p. insulin infusion study, mean plasma free insulin
`concentration rose to a maximal concentration of 37 ± 6
`/All/ml by 45 min postinitiation of the high dose insulin infu-
`sion followed by a slow decline over a period of approxi-
`mately 1 h, attaining a level of 26 /All/ml during the low dose
`insulin infusion. Following termination of the infusion, a fur-
`ther decline in plasma free insulin concentration was ob-
`served to a level of 18 ± 1 /ull/ml, not different from the ini-
`tial starting basal free insulin concentration of 19 ± 3 ju,U/ml
`(P > 0.05).
`In the s.c. insulin infusion study, mean plasma free insulin
`concentration rose to 40 ± 4 /uAJ/ml by 2 h postinitiation of
`high dose infusion followed by a slow decline in plasma free
`insulin concentration. After the infusion of insulin had
`ceased for 60 min, the plasma free insulin concentration re-
`mained elevated at 29 ± 2 /ill/ml (P < 0.001).
`In each study, plasma free insulin concentration demon-
`strated a rise and subsequent fall. However, the timing and
`magnitude of this change depended on the route of insulin
`delivery. At 30 min following the initiation of insulin infusion,
`plasma free insulin concentration in the i.p. study (34 ± 4
`jLtU/ml) was significantly greater than in the s.c. one (21 ± 2
`/ill/ml) and significantly less than in the i.p. study (117 ± 10
`AiU/ml) (P < 0.05) (Figure 1). By 2 h, however, the plasma
`free insulin concentration in the s.c. study (41 ± 3 /u,U/ml)
`had exceeded the concentration in both the i.p. (26 ± 3
`juAJ/ml) and i.v. studies (29 ± 2 /uU/ml) (P < 0.05). This late
`elevation in plasma free insulin concentration persisted
`throughout the remainder of the study such that at 41/2 h, the
`insulin concentration in the s.c. study (29 ± 2 ^ill/ml) was
`significantly greater than in the i.p. (18 ± 1 /u,U/ml) and the
`i.v. ones (16 ± 2 /xU/ml) (P < 0.01).
`Integrated insulin concentration (Table 1). Although the
`same quantity of exogenous insulin was infused into all sub-
`jects in each study (11 U), the quantity of insulin appearing
`in the plasma significantly differed, depending on the route
`of administration. During the 41/2-h i.v.
`infusion study,
`
`TABLE 1
`Integrated area of free insulin above basal concentration
`h/100)
`
`Subject
`
`Intravenous
`
`Intraperitoneal
`
`Subcutaneous
`
`1. N.V.
`2. E.G.
`3. J.A.
`4. M.P.
`5. J.G.
`6. J.D.
`7. D.B.
`8. J.S.
`9. I P.
`10. D.B.
`Mean
`Sem
`
`4,825
`12,837
`5,437
`13,975
`4,437
`8,175
`9,262
`8,850
`11,637
`8,000
`+ 8,743
`~ 1,046
`
`2,275
`5,262
`1,040
`5,625
`3,537
`2,112
`3,050
`4,862
`4,775
`3,850
`+ 3,638
`-
`480
`
`8,187
`8,300
`9,362
`9,162
`3,262
`3,525
`5,650
`8,637
`9,150
`8,600
`+ 7,383
`-
`743
`
`<
`
`P <
`
`0.001
`
`>*
`
`P <
`
`0.001
`
`*
`
`p vnnc
`i
`^ U.Uu
`
`in the
`8743 ± 1046 /-ill/ml • h/100 of insulin appeared
`plasma. This quantity was twice that of insulin appearing in
`the plasma when delivered via the i.p. route (3636 ± 480
`/nil/ml • h/100) (P < 0.001). When the 11 U of insulin was
`delivered via the s.c. route, 7383 ± 743 /xU/ml • h/100 ap-
`peared in the plasma, significantly greater than that during
`the i.p. route (P < 0.001) but not different from that quantity
`appearing during the i.v. infusion of insulin (P > 0.05).
`Plasma glucose concentration (Figure 1). Plasma glucose
`concentration rose in all studies following ingestion of the
`750-calorie meal. However, the rise was most rapid during
`the s.c. administration of insulin and least rapid during i.v.
`administration. Statistically, this difference became signifi-
`cant at 45 min when plasma glucose concentration had
`risen by 4.1 ±1.2 mmol/L in the s.c. study, 1.6 ± 0.7 mmol/L
`in the i.p. study, and declined by -0.9 ± 0.8 mmol/L in the
`i.v. study (P < 0.05 for all comparisons). This statistical dif-
`ference remained throughout the initial 75-min observation
`period. This rate of rise corresponded inversely to the rate of
`rise in plasma free insulin concentration in the three differ-
`ent delivery route studies (Figure 1, top).
`Plasma C-peptide concentration. Plasma C-peptide con-
`centration was assayed in all plasma samples. In no subject
`was C-peptide detectable in the plasma at any observation
`point during these studies.
`
`DISCUSSION
`This study compared the rapidity, duration, and magnitude
`of insulin absorption during insulin infusion via the i.v., i.p.,
`and s.c. delivery routes. Our results demonstrate that i.v. in-
`sulin attains maximal plasma concentration by 15 min, fol-
`lowed by the i.p. (45 min) and the s.c. routes (120 min), re-
`spectively. With
`respect
`to
`the duration of
`insulin
`absorption, absorption from the s.c. site was most pro-
`longed in contrast to the very rapid decline in plasma insu-
`lin concentration following termination of i.v. insulin deliv-
`ery. The magnitude of the integrated rise in plasma insulin
`was comparable in both the i.v. and s.c. infusion routes, but
`was reduced by 50% with peritoneal delivery of insulin. In
`spite of this "loss," i.p. delivery of insulin was as effective, if
`not more effective, than s.c. delivery in suppressing the
`meal-related hyperglycemia.
`.
`This is the first study demonstrating that i.p. insulin deliv-
`ery may control postprandial hyperglycemia in healthy dia-
`betic men. Intraperitoneal insulin delivery has several char-
`acteristics that make it an attractive alternative to i.v. and
`s.c. insulin delivery. First, the absorption of insulin is rela-
`tively rapid such that rapid control of hyperglycemia may be
`achieved. Second, the peritoneal space is extravascular
`such that catheter thrombosis is minimized.9 Third, the peri-
`toneal space is a potentially large anatomic space into
`which an artificial pancreas could be implanted without dis-
`figurement of the recipient.
`In these studies, only short-term insulin absorption kinet-
`ics were examined, which may not necessarily reflect kinet-
`ics of long-term delivery. Furthermore, because our protocol
`was not designed to optimize glucose control by the three
`routes of insulin delivery, additional studies will be required
`to determine the most effective route in controlling plasma
`glucose concentration. However, since a relationship exists
`between the timing of the insulin increase in the plasma and
`the rise of plasma glucose concentration,10 our results sug-
`
`DIABETES, VOL. 28, DECEMBER 1979
`
`1071
`
`MYLAN INST. EXHIBIT 1052 PAGE 3
`
`

`

`INSULIN DELIVERY ROUTES
`
`gest that the infusion of s.c. insulin be begun before the start
`of the meal, as has been employed.4-5 Furthermore, since
`both our results and the results of others11 demonstrate a
`prolonged absorption of insulin at the s.c. site following ter-
`mination of the insulin infusion, a reduced rate of delivery
`between meals should be evaluated when using this route.
`Our results also suggest that improved glucose control
`might be achieved during the i.p. route if the insulin is ad-
`ministered at a greater concentration for a shorter period of
`time following the initiation of the meal. Whatever the route
`of insulin delivery used by an artificial pancreas, knowledge
`of the kinetics of insulin absorption and the onset of biologic
`activity should result in improved control of meal-induced
`hyperglycemia.
`
`ACKNOWLEDGMENTS
`This investigation was supported by grants from the Clinical
`Research Center Program, DRR, NIH grant RR-00997-02,
`the Kroc Foundation, NIAMD-19998-01, NICHD grant IP50-
`HD11327-01, and the Department of Energy. David S.
`Schade is the recipient of Research Career Development
`Award 1KZ4AM00260-02.
`The technical assistance of Enid Pinero and Jerry Towle
`is appreciated. The mechanical and electrical engineering
`expertise of Raymond Bair, Gary Carlson, and Jerry Love in
`the design, fabrication, and testing of the insulin mini infu-
`
`sion pump used in these studies is gratefully acknowl-
`edged.
`
`REFERENCES
`1 Albisser, A. M., Botz, C. K., and Leibel, B. S.: Blood glucose regula-
`tion using an open loop insulin delivery system in pancreatectomized dogs
`given glucose infusions. Diabetologia 76:129-33, 1979.
`2 Irsigler, K., and Kritz, H.: Long-term continuous intravenous insulin
`therapy with a portable insulin dosage-regulating apparatus. Diabetes
`28:196-203, 1979.
`3 Schade, D. S., Eaton, R. P., Spencer, W., et al.: The peritoneal ab-
`sorption of insulin in diabetic man: a potential site for a mechanical insulin
`delivery system. Metabolism 28:195-97, 1979.
`4 Pickup, J. C, Keen, H., Parsons, J. A., et al.: Continuous subcutane-
`ous insulin infusion: an approach to achieving normoglycaemia. Br. Med. J.
`7:204-07, 1978.
`5 Tamborlane, W. V., Sherwin, R. S., Genel, M., et al.: Reduction to nor-
`mal of plasma glucose in juvenile diabetes by subcutaneous administration
`of insulin with a portable infusion pump. N. Engl. J. Med. 300:573-78, 1979.
`6 Carlson, G. A., Shafer, B. D., Urenda, R. S., et al.: A new low-power
`high-reliability infusion pump. Presented at the 7th Annual New England
`(Northeast) Bioengineering Conference, March 22-23, 1979.
`7 Berger, W. J., Jr.: Evaluation of "intracath" method of abdominal
`paracentesis. Am. Surg. 35:23-26, 1969.
`8 Kuzuya, H, Blix, P. M., Horwitz, D. L, et al.: Determination of free and
`total insulin and C-peptide in insulin-treated diabetics. Diabetes 26:22-29,
`1977.
`
`9 Yoffey, J. M., and Courtice, F. C: Lymphatics, Lymph and the Lym-
`phomyeloid Complex. New York, Academic Press, 1970, p. 295.
`10 Eaton, R. P., Spencer, W., Schade, D. S., et al.: Diabetic glucose
`control: matching plasma insulin concentration to dietary and stress hyper-
`glycemia. Diabetes Care 7:40-44, 1978.
`11 Slama, G., Buu, K. N. P., Tchobroutsky, G., et al.: Plasma insulin and
`C-peptide levels during continuous subcutaneous insulin infusion. Diabetes
`Care 2:251-55, 1979.
`
`1072
`
`DIABETES, VOL. 28, DECEMBER 1979
`
`MYLAN INST. EXHIBIT 1052 PAGE 4
`
`