Are We There Yet?
Eda Cengiz, MD, MHSAssistant Professor of Pediatrics
Yale University School of Medicine
Update on Artificial Pancreas Project
Artificial Pancreas / Bionic Pancreas ?
utilizing electronic devices and mechanical parts to assist humans in performing difficult,
dangerous, or intricate tasks, as by supplementing or duplicating parts of the body
Artificial intelligence runs the algorithm
•Sensor signals transmitted to a laptop computer that displays the sensor glucose and calculates rate of insulin delivery
•Rate of insulin delivery is transmitted to the insulin pump
Determine insulin requirement real time, deliver proper insulin to achieve euglycemia.
The Promise of Artificial Pancreas
•Glucose Sensor•Algorithm•Insulin Pump & Insulin
Medtronic Closed-loop System c. 2011
Insulin Pump
CGM
Laptop Controller
What is taking so long ?
During a Typical Clinic Visit at Yale Diabetes Center…
How is it going with the Artificial Pancreas Dr. C?
When is it going to be ready?
Challenges & Solutions: Artificial Pancreas (Closed-Loop System)
•Progress to date
•Pitfalls
•Solutions
•Future closed-loop studies(hot off the press!)
JDRF Road Map to Artificial Pancreas
What is the plan?
Artificial Pancreas
Do we need to wait until we have the Fully Automated Artificial Pancreas?
Artificial Pancreas
Progress to date
SENSOR-AUGMENTED PUMP THERAPY
CGM improved A1c, but not in everyone
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
8 - 14 yrs 15 - 24 yrs > 25 yrs
RT-CGM Control
p=0.29 p=0.52 p<0.001
JDRF CGM Study Group. N Engl J Med 2008; 359:1464-76.
Average Days of CGM Use by Age Group
CGM Use
0%
20%
40%
60%
80%
Age ≥25 Age 15-24 Age 8-14
Per
cent
age
of s
ubje
cts
<4.0 days/week
4.0-<6.0 days/week
≥6.0 days/week
Change in glycated hemoglobin-0.8
-0.6
-0.4
-0.2
0.0
0.2
Cha
nge
in g
lyca
ted
hem
oglo
bin
JDRF CGM Study Group. N Engl J Med 2008; 359:1464-76.
Bergenstal RM, Tamborlane WV, Ahmann A, et al. N Engl J Med. Doi:10.1056/NEJMoa1002853.
8.3%
7.3%
7.5% 7.5% 7.5%
8.0% 8.0%8.1% 8.1%
7.0%
7.5%
8.0%
8.5%
0 3 6 9 12
Months
A1C
Medtronic STAR 3 Sensor-Augmented Pump Trial
Values are means ± SE. Comparisons between SAP group and MDI group are significant for each time period (P<0.001).
• The SAP group achieved a greater A1C reduction vs. MDI at 3 months and sustained it over 12 months
A1C Reduction for SAP and MDI Groups
= MDI= SAPn = 244 n = 241
- 0.6P<0.001
∆ -0.2
∆ -0.8
- 0.6P<0.001
Bergenstal RM, Tamborlane WV, Ahmann A, et al. N Engl J Med. Doi:10.1056/NEJMoa1002853.
A1C Reduction Correlates to Increased Sensor Use
Values are the difference between the means ± SE. p=0.003 for association between sensor wear and A1C reduction at 1 year. Only 7 participants had sensor use of 20% or less, with a change in A1C of -0.43 at 1 year vs. baseline.
• The majority of patients used sensors ≥61% of the time • Patients who used sensors ≥81% of the time reduced their mean A1C by
1.2% at 1 year vs. baseline
-0.19
-0.64-0.79
-1.21-1.5
-1
-0.5
021-40% 41-60% 61-80% 81-100%
Frequency of Sensor Use (% of Time)
Cha
nge
in A
1C a
t 1 Y
ear v
s B
asel
ine
n =27 n =46 n =108 n =56
DirecNet / TrialNet Metabolic Control Study
• Does the rapid normalization of BG levels at the time of diagnosis of diabetes, followed by super-intensive control of BG levels, help to preserve residual beta-cell function?
• Use of an artificial pancreas in subjects AT DIAGNOSIS for 3-4 days to rapidly normalize BG levels, followed by sensor-augmented pump therapy x 2 years
Sensor and BG Levels During CL TherapySensor and BG Levels During CL Therapy
Mean Sensor Glucose Levels Prior to, Mean Sensor Glucose Levels Prior to, During, and Following CL TherapyDuring, and Following CL Therapy
3 mth f/u: A1c = 5.9 %
6 wk f/u
6 wk f/u6 mth f/u: A1c = 6.0 %
“typical” tracing12 mth f/u: A1c = 6.3 %
Artificial Pancreas
Progress to Date
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL
HYPOGLYCEMIA
Prevention of Hypoglycemia with AP
• Insulin suspension prevented low BG (<60mg/dl) in 78% of the suspensions.
• Non resulted in hyperglycemia
• Reduction in CGM hypoglycemia alarms
A Semi-Closed-Loop System: The Paradigm® Veo™*
• Integrated sensor• Improved Calibration Routines• Glucose trend• Alarms
– Outside target zone– Predictive– Missing insulin bolus
• Minimum basal rate of 0.025 U/h• Hypoglycemia suspend
– Suspend for 2 hours– Re-suspend after 4 hours if
needed
* Investigational device. Limited by U.S. law to investigational use.
Keenan et al., J. Diabetes Sci. Tech., 2010; 4(1):111-118Buckingham et al., Diabetes Technol. Ther., 2009; 11:93-97Attia et al., Diabetes Care, 1998; 21:817-821.Guerci et al., J Clin Endocrinol Metab, 1999; 84:2673-2678.Zisser, Diabetes Care, 2008; 31:238-239.
S L I D E 22
Automatic Low-Glucose Suspend
LGS off LGS onMean Glucose (mg/dL) 145 ± 23 148 ± 19
Time < 70 per day (min) 101 ± 68 58 ± 33 *
Time > 140 per day (min) 651 ± 240 639 ± 182
Number of excursions < 70 and < 40 mg/dL reduced with LGS
* p=0.002
Danne T, Diabetes Technol Ther 2011; 13: 1129-1134
Artificial Pancreas
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL HYPOGLYCEMIA
PUMP SUSPENSION FORPREDICTED
HYPOGLYCEMIA
Progress to Date
CL2-MW 9/3/08
Time in Minutes Beginning at 9:30 PM
0 100 200 300 400 500 600
Ser
um G
luco
se (m
g/dl
)
20
40
60
80
100
120
140
160
Y A
xis
2
0.0
0.5
1.0
1.5
2.0
Controller GlucoseYSI Basal Insulin
ROC = -.36 mg/dl-min3 Alarm, Threshold 80 mg/dl, Horizon 35 minAutomatic pump suspension for predicted hypoglycemia
Exercise AP Study objective
• To evaluate whether use of a AP system reduces the risk of delayed (nocturnal) hypoglycemia following antecedent daytime exercise
Subject recruitment, consent, enrollment
(n = 12)
Closed-Loop
Open-LoopOpen-Loop
Closed-Loop
Exercise Study Protocol Schematic
48-hour evaluation period2 overnights
48-hour evaluation period2 overnights
Plasma BG q30 min, insulin q30 min x 3h @ meals
Nocturnal Hypoglycemia
Closed LoopOpen Loop
0
5
10
15
20
25
3
22
All Nocturnal Hypo
Num
ber
of T
reat
men
ts G
iven
p=0.05
1
14
Night Following Exercise
p=0.06
Glucose Frequency DistributionNight after sedentary condition
1 % 99 % 0% 3 % 90 % 7 %
P<0.0001
3 % 93 % 4 %8 % 76 % 16 %
P<0.0001
Glucose Frequency DistributionNight following afternoon exercise
Artificial Pancreas
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL HYPOGLYCEMIA
PUMP SUSPENSION FORPREDICTED HYPOGLYCEMIA
BASAL/BOLUS ACTIVATION
FOR HYPERGLYCEMIA
Progress to Date
Conceptual Scheme for “Treat-to-Range”
Artificial Pancreas
Progress to Date
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL HYPOGLYCEMIA
PUMP SUSPENSION FORPREDICTED HYPOGLYCEMIA
BASAL/BOLUS ACTIVATIONFOR HYPERGLYCEMIA
CLOSED-LOOP BASAL
MANUAL MEALS
Studies of Overnight CL
• Increased time in target BG
• Reduction of hypoglycemia
• Incorporation of daytime challenges– Exercise– Alcohol– Pregnancy
Hybrid control improves performance
6A Noon 6P MidN 6A Noon 6P0
100
200
300Closed Loop (N=8)
meals
setpoint
Hybrid CL (N=9)
Glu
cose
(mg/
dl)
Mean Daytime Peak PPFull CL 147 58 154 60 219 54Hybrid 138 49 143 50 196 52
Weinzimer SA. Diabetes Care 2008; 31:934-939.
Late post-prandial hypoglycemia in CL
6 12 18 24 30 36 420
3
6
9
12
15
18
0
50
100
150
Plasma InsulinEVP
Insu
lin (U
/h)
Insulin ( U/ml)
6 12 18 24 30 36 420
100
200
300
0
25
50
SG CHO
Glu
cose
(mg/
dl)
Artificial Pancreas
Progress to date
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL HYPOGLYCEMIA
PUMP SUSPENSION FORPREDICTED HYPOGLYCEMIA
BASAL/BOLUS ACTIVATIONFOR HYPERGLYCEMIA
CLOSED-LOOP MULTIHORMONAL
FULL CLOSED-LOOP CONTROL
CLOSED-LOOP BASALMANUAL MEALS
AP Multi-hormonal Approach
• Can the addition of pramlintide improve the performance of a CL system by reducing the peak post-prandial glucose excursions?
Pramlintide
• Analog of human amylin
• Co-secreted with insulin from -cell
• Used as adjunct to insulin in T1D to reduce post-prandial glycemic excursions
– Delay gastric emptying– Suppress endogenous glucagon
Subject recruitment, consent, enrollment
(n = 8)
Closed-Loop
Closed-Loop + Pramlintide
(30 mcg per meal)
Closed-Loop + Pramlintide
(30 mcg per meal)
Closed-Loop
Pramlintide Study Protocol Schematic
24-hour evaluation period3 meals (BF, L, Din)
24-hour evaluation period3 meals (BF, L, Din)
Plasma BG q30 min, insulin q30 min x 3h @ meals
Glucose excursions with/without pramlintide
6 9 12 15 18 21 24 27 30 330
100
200
300
Control DaySymlin Day
Time (hrs)
Blo
od G
luco
se (m
g/dL
)
Summary and conclusions
• Pramlintide had modest effect on prandial glucose
• Would require manual injection or at best, manual bolus
• Faster insulin absorption / action clearly needed
Artificial Pancreas
Progress to date
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL HYPOGLYCEMIA
PUMP SUSPENSION FORPREDICTED HYPOGLYCEMIA
BASAL/BOLUS ACTIVATIONFOR HYPERGLYCEMIA
CLOSED-LOOP MULTIHORMONAL
FULL CLOSED-LOOP CONTROL
CLOSED-LOOP BASALMANUAL MEALS
?
What do we need to achieve AP?
• Better accuracy, user interface, reliability.
• Better algorithm• One site for CGM &
Insulin
Single Port, Multiple Sensor
• One site, two catheters for insulin infusion and glucose sensing
• Single port for CGM and insulin
• CGM with an optic sensor back up
What do we need to achieve AP?
• Better accuracy, user interface, reliability.
• Better algorithm• One site for CGM &
Insulin
• Faster acting insulin
Dangerous Delays in Insulin Action after SC Injection
1. Delays due to the chemical properties of insulin.
1. Tissue delays
Insulin Time-Action Curves
0 30 60 90 120 150 180 210 240 270 3000
10
20
30
40
50
60
70
80
90
100
0
1
2
3
4
5
6
7
InsulinGIR
Time (min)
Plas
ma
Insu
lin (u
U/m
L) GIR
(mg/kg/m
in)
•The InsuPatchTM device applies controlled heat around the insulin infusion site.
InsuPatch
Temp. Sensor
with InsuPatch activation:
Aspart insulin bolus maximum effect was 35 min earlier compared
to the same dose bolus without InsuPatch activation .
Peak aspart insulin action curve shifted to the left .
The Effect of InsuPatch on The Effect of InsuPatch on Insulin Action Insulin Action
51
Hyaluronidase Mechanism of Action
The Accu-Chek DiaPort is a port system for continuous intraperitoneal insulin infusion.
DiaPort
The catheter tip is placed in the peritoneal cavity where the insulin is directly infused.
Pathway to Closed-Loop (CL)
Pump
Platform and Connectivity for Ambulatory Studies
Artificial Pancreas
Progress to date
SENSOR-AUGMENTED PUMP THERAPY
PUMP SUSPENSION FORACTUAL HYPOGLYCEMIA
PUMP SUSPENSION FORPREDICTED HYPOGLYCEMIA
BASAL/BOLUS ACTIVATIONFOR HYPERGLYCEMIA
CLOSED-LOOP MULTIHORMONAL
FULL CLOSED-LOOP CONTROL
CLOSED-LOOP BASALMANUAL MEALS
?Outpatient studies
The “Dream” Study
• http://www.youtube.com/watch?v=9HMx8yy2nVw
Thank you!
• Yale Closed Loop Team– Stu Weinzimer– Jennifer Sherr– William Tamborlane– Grace Kim– Miladys Palau– Camille Michaud– Lori Carria– Amy Steffen– Kate Weyman– Melinda Zgorski– Eileen Tichy