Assessment of an Integrated Continuous Glucose Monitor and Insulin Injection Port in Type 2 Diabetes

Study Goal: Carry out 7-day study in 18 subjects with T2D who use MDI. Compare use of the PDT CGM injection port with standard site-rotated MDI injections; test potential insulin incompatibility of long-and short-acting insulin.

Experimental Protocol: Approval will be obtained from OHSU IRB after the IDE is cleared by FDA. Subjects with T2D (n=18 in total) using MDI will be recruited from Dr. Castle's OHSU Diabetes Clinic.

Informed consent and insulin regimen: Informed consent and signing will precede a screening visit which will include history, physical exam, complete blood count, electrolytes, BUN, creatinine, liver panel, calcium and phosphorus. Those who pass will be pseudo-randomized to injecting insulin through the sensing port during days 2-4 and crossing over to standard subcutaneous injections during days 5-7 or vice-versa. Prior to starting insulin pen usage, a visit will be scheduled with a diabetes research staff member or a certified diabetes educator to demonstrate the proper pen usage and port mating. Before the study begins, Dr. Castle will modify the MDI regimen to include aspart or rapid aspart insulin (Fiasp) before meals and long-acting (degludec) at bedtime. Dr. Henrik Bengsston from Novo has clarified that degludec insulin (neutral pH) is compatible with Novo fast-acting preparations.

Study Goals: (1) Demonstrate that CGM accuracy over 7d during injections via the CGM-MDI port is not inferior to accuracy during remote SC injections; (2) Demonstrate that algorithmic compensation for the dilution artifact will maintain sensor accuracy in the hour after boluses; (3) Validate the safety of the novel CGM-MDI port; (4) assess port longevity, and (5) assess subject satisfaction with port usage.

In addition, there is a possibility that degludec insulin remaining in the SC depot will alter the pharmacodynamics of the aspart given the next day. The study design will allow investigators to address this possibility by evaluating post-prandial glycemic control on Day 7 after the first meal: in 50% of cases, the prior degludec will have been given via the port into the same SC depot; in the other 50%, the degludec will have been given at a distant SC site by standard injection.

Skin Issues: Allergic Skin reactions, Infection, Dislodgement: It is now recognized that allergic reactions to sensor adhesives are common (in as many as 35% of sensor users). For this reason, the acrylate adhesive will not contain IBOA, ethyl cyanoacrylate, or colophonium. As a further precaution against contact dermatitis, a breathable barrier film will be placed under the device (A layer of IV Prep (Smith and Nephew) will first be placed on clean, dry skin). At study's end, the subjects will be asked about surrounding redness, itching, or pain, and the site will be observed and photo-graphed (without identifying information) after device removal. To reduce the risk of bacterial inoculation, each subject will prepare the septum roof with an alcohol wipe, letting it dry prior to injection. Since many people struggle with keeping devices attached to their skin, each subject will have a Grifgrips overlay with a hole in center, placed over their ports.

CLINICAL STUDY, Day 1: Each subject will arrive in the fasted state unless BG at home prior to travel is less than 90 mg/dl, in which case subject will take 15 g of carbohydrate (and recheck BG) before traveling to the clinic. After arrival at OHSU, each subject will have two CGM's placed widely apart, a Dexcom G6 and a PDT CGM-MDI port. A catheter will be placed in a forearm vein for blood withdrawal and kept warm (arterialized) either with a hand warming box set to 55° C or with warmed towels. Time 0 is set at the first bite of the first meal (approximately 1000 h). Both meals will be high carbohydrate mixed meals consisting of 9 kcal/kg and half-ml arterialized venous blood samples will be obtained at time 0 and every 10 minutes through the remainder of the study, and analyzed for whole blood glucose using a YSI 2300 glucose analyzer. The first insulin injection will be administered by a staff member via the sensing port (teaching demonstration); the second will be given by the subject, with staff supervision. Each injection will be given just prior to the first bite of the meal. The second meal (same composition) will be given at minute 210. At minute 420, the IV catheter will be removed, the in-clinic experiment terminated, and the subject discharged (BG will be checked and a snack given if needed prior to travel home). The third meal of the day will be taken at home.

Days 2-6: To give insulin, subjects will use either use the sensing port or remote SC sites according to the pseudo-randomization schedule ('pseudo-randomization' defined as 50% of subjects start at home with standard injections and 50% start with port injections). Each subject, regardless of randomization, will give 11 injections through the port and 12 via the standard subcutaneous route. Subjects will use the Dexcom and BG data to manage their diabetes (experimental CGM data will not be visible to subjects). Subjects will measure BG with a Contour Next meter and strips (supplied) before each meal, 1.5 hours after each meal, at bedtime, and whenever symptoms dictate. Subjects will keep a record of symptoms, carbohydrate intake, and insulin doses, and will be encouraged to call research staff with questions or concerns. A provision will be made for possible port occlusion detection: In the case of persistent hyperglycemia (over 250 mg/dl) during port usage, unexplained by dietary choices, subjects will be asked to given injections by the standard route instead and phone the research staff for additional guidance.

Day 7: The in-clinic protocol will be the same as Day 1, except that the sensing port and the Dexcom CGM will already be in place. After the study, all devices will be removed. The research staff will assess the skin site by (1) photodocumentation (avoiding identifying marks such as tattoos) and (2) use of the Draize scale to quantify inflammation.

User Experience: At study's end, each subject will be asked to answer the Insulin Delivery Satisfaction Survey, a validated 18-question survey developed by Polonsky and Fisher (Polonsky, Fisher et al. 2015).