Implantable batteryless device for on-demand and pulsatile insulin administration

Many implantable systems have been designed for long-term, pulsatile delivery of insulin, but the lifetime of these devices is limited by the need for battery replacement and consequent replacement surgery. Here we propose a batteryless, fully implantable insulin pump that can be actuated by a magnetic field. The pump is prepared by simple-assembly of magnets and constituent units and comprises a drug reservoir and actuator equipped with a plunger and barrel, each assembled with a magnet. The plunger moves to noninvasively infuse insulin only when a magnetic field is applied on the exterior surface of the body. Here we show that the dose is easily controlled by varying the number of magnet applications. Also, pump implantation in diabetic rats results in profiles of insulin concentration and decreased blood glucose levels similar to those observed in rats treated with conventional subcutaneous insulin injections.


Supplementary Figure 6 Reproducibility assessment of the MDP after a refilling procedure
For this evaluation, we first simulated consumption of insulin by intentionally withdrawing 0.4 ml of the insulin solution in the drug reservoir and then injected the same volume of fresh insulin solution through 3 the septum port of the MDP using a 30G needle. (a) Schematic description of the refilling procedure. (i) The refill port is located with a localization magnet (ML) with a donut shape and polarity opposite that of MR; (ii) a syringe needle is inserted through the refill port, and 0.4 ml of insulin solution (109 U/ml) is 6 injected; and (iii) the drug reservoir is filled with the insulin solution. (b) In vitro insulin release profiles before and after a refilling procedure. The MDP was fully immersed in phosphate-buffered saline (PBS; pH 7.4) at 37 °C. Initially, three actuations were conducted at 10-min intervals. After each actuation, the 9 amount of released insulin was measured. Then, a refilling procedure was performed, and the experiments were repeated. Three MDPs were tested for this experiment. Error bars are s.d. (c, d) In vivo profiles of plasma insulin concentration and glucose level in blood before and after the refilling procedure (n = 3). The 12 refilling procedure was performed while the MDP was implanted in streptozotocin (STZ)-induced diabetic rats. The septum port was found from outside of skin using a localization magnet (ML) with a polarity opposite that of MR. To measure the plasma insulin concentration, blood was collected at 60 min after 15 actuation. The blood glucose levels were measured at -1 and 120 min after actuation. Error bars are s.d. In addition to a short acting insulin mainly used in this study (i.e., MDP_1A (short acting insulin)), we additionally prepared the MDP filled with a rapid acting or long acting insulin formulation (NovoRapid 1 or Lantus 2 , respectively) to give the animal groups of MDP_1A (rapid acting insulin) or MDP_1A (long acting 6 insulin), respectively. At 1 day after the MDP was implanted in STZ-induced diabetic rats, the blood glucose level was obtained at -1 to 720 min after insulin administration by a single actuation of the MDP.

Supplementary
A similar dose of insulin was administered for all experimental groups (rapid acting: 0.74 U; short acting: 9 0.80 U; and long acting 0.74 U) . For each animal group, 4 rats were employed for statistics. Error bars are s.d. For all formulations, the decrease in blood glucose level was apparent after actuation and as expected, a specific profile of blood glucose level was observed for each type of the insulin formulations filled in the 12 MDP. For rapid acting insulin, the blood glucose level decreased and increased back more rapidly. For long acting insulin, the blood glucose level dropped relatively slowly and this lowered level was maintained for a longer period.

Supplementary Figure 9 Profiles of blood glucose level with multiple daily actuations of the MDP
To give an insight of insulin delivery after each of the meal times, the MDP was actuated twice with an 3 interval of 12 h after implantation. After the first actuation, a glucose level dropped and increased back to a high level as expected with diabetic rats, which was observed to be repeated in a similar pattern after the second actuation (n=4). Error bars are s.d.

Supplementary Table 1 Actuation ability of the MDP according to the gap between the external device and MDP
The MDP could be actuated at gaps of up to 3 mm and in this range, the MDP could infuse the same amount 3 of insulin, as observed in our in vitro performance test. To be actuated with the skin gap of 1 mm, the MDP herein needs an external magnet with 3000 G. Therefore, considering the range of a magnetic field available in a regular life style (< 2 G) 3 , a chance for an accidental activation of the MDP is expected to be not high.

Supplementary Table 2 Insulin amount left in fibrotic capsules
To assess the insulin amount possibly left in the fibrotic capsule, we biopsied the whole capsule including the MDP at two different days after implantation, i.e., at 16 days and 60 days, respectively (n = 5; 16 days, 3 n = 4; 60 days). At each day, the MDP was actuated twice consecutively (1.6 U insulin delivery) and after 360 min, the biopsy was performed. From each of the biopsied capsule, we extracted the MDP and the surrounding capsule tissue was fully immersed in 5 ml of pH 7.4 PBS at 37 o C for 6 h. Then, the supernatant 6 was analyzed with HPLC, as described in the Methods, to measure the insulin amount. Our results revealed that more than 97.5% of the total amount of delivered insulin was diffused out from the fibrotic capsule during the first 360 min after actuations. Values ± s.d.

Supplementary Table 3 Inflammatory markers in plasma
Inflammatory markers in plasma, such as IL-1b, IL-6 and TNF-α, were measured at 30 and 60 days after MDP implantation (n = 3; 30 days, n = 3; 60 days). For all animals, elevation of inflammatory markers was 3 not observed and their levels in plasma were not different from the ones with the control animal group (i.e., the animals without MDP implantation). 6

Supplementary Table 4 Amount of released insulin with varied catheter lengths
The MDP without a catheter (i.e., a catheter, 0 cm in length) and the ones connected with a catheter, 10 and 15 cm in length, respectively, were each actuated under the in vitro experimental condition, as depicted in

Theoretical assessment of magnetic forces between the plunger and barrel magnets
The attraction force between the plunger and barrel magnets (MP and MB, respectively) was measured to be 1.25 N (Advanced force measurement 9830, Interface, USA), which could fix their position to not release 3 insulin during the period of no actuation. With the mass of the plunger used in this study (9.8 x 10 -4 kg), therefore, the acceleration needed to overcome this attraction force and move the plunger was calculated to be 1.28 x 10 3 m s -2 , which was more than 100 times larger than gravitational acceleration (9.8 m s -2 ). In this