Performance of an Artificial Pancreas System for Young Children with Type 1 Diabetes

Abstract

Background:

Young children 5–8 years old with type 1 diabetes (T1D) exhibit clear needs for improved glycemic control but may be limited in their ability to safely interact with an artificial pancreas system. Our goal was to evaluate the safety and performance of an artificial pancreas (AP) system among young children with T1D.

Research Design and Methods:

In a randomized, crossover trial, children with T1D age 5–8 years were enrolled to receive on separate study periods (in random order) either the UVa AP using the DiAs Control Platform software with child-resistant lock-out screens (followed as an out-patient admission) or their usual insulin pump+continuous glucose monitor (CGM) care at home. Hypoglycemic events and CGM tracings were compared between the two 68-h study periods. All analyses were adjusted for level of physical activity as tracked using Fitbit devices.

Results:

Twelve participants (median age 7 years, n = 6 males) completed the trial. Compared to home care, the AP admission resulted in increased time with blood glucose (BG) 70–180 mg/dL (73% vs. 47%) and lower mean BG (152 mg/dL vs. 190 mg/dL), both P < 0.001 after adjustment for activity. Occurrence of hypoglycemia was similar between sessions without differences in time <70 mg/dL (AP 1.1% ± 1.1%; home 1.6% ± 1.2%). There were no adverse events during the AP or home study periods.

Conclusions:

Use of an AP in young children was safe and resulted in improved mean BG without increased hypoglycemia. This suggests that AP use in young children is safe and improves overall diabetes control.

ClinicalTrials.gov registration number: NCT02750267.

Introduction

Young children age 5–8 years with type 1 Diabetes (T1D) have clear needs but unique challenges regarding the application of artificial pancreas (AP) technologies. Young children may benefit from an AP system, given their current deficits in glycemic control that include both significant hypoglycemia and suboptimal HbA1c levels.^1,2 A recent AP trial in this age range reduced hypoglycemia without improvement in glycemic control,³ as demonstrated in older children^4

–10 and adults.^10

–13

One limitation pertinent to use of an AP in young children is their undeveloped abilities to control and interact with the AP system, which poses potential safety issues. During the hours that these children are away from their parents, the child—and potentially his or her friends—could have access to the system's control settings, including the ability to change inputs related to insulin sensitivity, insulin-to-carbohydrate ratios, and ingested carbohydrate amounts. Accidental or purposeful tampering with such a system could result in delivery of inappropriate amounts of insulin that could result in potentially dangerous hypoglycemia or hyperglycemia with ketosis. Many available insulin pumps have parental controls such as lock-out screens for this purpose. However, to our knowledge, no AP system has addressed this potential danger.

Our goal was to assess the safety, feasibility, and efficacy of our AP system in 5- to 8-year-old children and their families. We hypothesized that such a system would lower mean blood glucose (BG) and increase the time in the target range. Because of the potential for a young child's mishandling of the system's settings, we utilized an adapted system with parental lock-out controls. If effective at improving diabetes control, such a young child AP could potentially revolutionize diabetes care for this group with significant clinical needs.

Materials and Methods

Study participants were recruited from the University of Virginia (UVa) Center for Diabetes Technology database, the University of Virginia Pediatric Diabetes Clinic, and clinicaltrials.gov (NCT02750267) direct contact. The study was approved by the FDA (IDE No. G160047) and the University of Virginia Institutional Review Board. All participants and their parents provided informed consent/assent, respectively. Inclusion criteria were as follows: age 5–8 years old, inclusive; clinical diagnosis of T1D with use of daily insulin for ≥1 year; use of insulin pump ≥6 months; and HbA1c 5.0%–10.5%.

Exclusion criteria included severe hypoglycemia within the last 3 months (severe hypoglycemia defined as hypoglycemia with loss of consciousness or seizures); diabetic ketoacidosis within last 6 months; anemia; conditions that may increase risk of hypoglycemia (e.g., known history of arrhythmia, stroke, seizure, syncope, adrenal insufficiency, or neurological disease); or use of a medication that lowers heart rate.

Study participants underwent two periods of BG monitoring in random order: home care using usual insulin pump settings (at home for 3 days) or AP use (at hotel for 3 days), a similar study design to a recent high-profile AP trial.¹⁰ The AP system as tested consisted of the following wirelessly connected devices: (1) a Tandem t-slim insulin pump (Tandem Diabetes Care, San Diego, CA), (2) DexCom (San Diego, CA) G4 Platinum with Share continuous glucose monitor (CGM) sensor, and (3) a smart phone running the DiAs Control Platform software (Supplementary Fig. S1; Supplementary Data are available online at www.liebertpub.com/dia).¹¹ Participants also wore a Fitbit Charge HR (Fitbit, Inc., San Francisco, CA) to monitor the degree of activity during both periods.

Those participants who had their home-care period before the AP period had their CGM placed between 3 and 5 h of starting the home assessment; those who had the AP phase first received the CGM between 3 and 5 h of starting the AP. Both groups randomized to AP first or home first continued with the same sensor and proceeded to the alternate phase after a washout period of 6–8 h. This design ensured that half of participants started the AP period during the first 24 h of placement of the CGM sensor (which has lower BG accuracy compared to the subsequent 6 days) and the other half started home care during this first 24-h period after sensor placement.

During the AP period, participants were connected to a remote monitoring system, with BG tracings and insulin administration followed by study personnel online. Ketones were assessed for BG values >250 mg/dL. CGM devices were calibrated before breakfast and dinner. Parents were counseled before the home period to match the mix of lower and higher intensity activities and food intake planned for the AP period.

The AP phase was held at a local resort; all 12 participants had at least one parent staying with the child in his/her room and at each meal. Parents were otherwise not required to be present for activities but generally were. The AP portion involved a mixture of activities such as crafts, viewing movies, miniature golf, going for a short walk, riding an electronic bull and climbing a wall, as well as games in a field. Meals were at ∼8 am, 12 pm, and 6 pm with a mid-afternoon snack. Further snacks were allowed on an individual basis if the parents reported that this was part of their usual home routine. During the AP portion, parents helped the child in selection of food and performed all carbohydrate counting and entering in of carbohydrate information into DiAs.

BG measurements were performed before meals, before bed, and for system alarms (for imminent hypoglycemia and sustained hyperglycemia). The AP system issued a low-BG alarm for a predicted BG <70 within a 15-min period. High-glucose alarm was not based on a specific BG threshold but was instead based on a difference in observed versus predicted insulin resistance. During the home-care period, the DexCom devices were required to have audible alarms for hypo- and hyperglycemia, with settings that were permitted based on parental preference to no lower than 60 mg/dL for low-BG alert and no higher than 300 mg/dL for high-BG alert.

During the AP period, a hypoglycemia alert prompted a finger stick self-monitored BG (SMBG) assessment (SMBG). For SMBG that was <70 mg/dL during the day and <80 mg/dL at night, participants immediately consumed 4–16 g of carbohydrates. BG was rechecked 15 min later and additional treatments administered until BG was >80 mg/dL. During the home-care period, parents were instructed to treat hypoglycemia according to their usual practice.

This study used a version of DiAs that had been altered for safer use in young children via placing password-protected lock-out screens for participant and pump settings and carbohydrate ingestion. Passwords were changed daily by combining the child's study number and current date. Parents were told the pattern of this password but were instructed not to tell this to the child. Parents were further instructed to observe whether they noted the child gaining access to the system through discovering the password. Parents were permitted to allow their child to enter carbohydrate data under parental supervision. DiAs screens displaying the BG tracings and entering data regarding BG (e.g., during a low BG) did not require the password for access.

The AP system was otherwise similar to control algorithms in prior DiAs trials, including running an AP algorithm with a glycemic target of 160 mg/dL during the daytime and 120 at night, with decay between these between 11 pm and 2 am.^4,14

Statistics

The primary outcome of the study was percent time in target range 70–180 mg/dL. Secondary outcomes were percent of CGM BG values below 70 mg/dL, number of hypoglycemic episodes (<70 mg/dL), and mean BG number of episodes requiring treatment and Low Blood Glucose Index. Comparison between AP and home care was performed using paired t-tests and nonparametric Wilcoxon signed rank test when necessary (e.g., hypoglycemia count); all results are provided as mean ± SE of the mean, AP versus home care, unless specified otherwise. Following comparison of Fitbit-measured activity levels between study periods, the analysis for differences in glycemic control between study periods was performed again using ANOVA, with activity level (total steps and heart rate in separate analyses) as covariates. This comparison allowed generation of P-values for difference between study periods, following adjustment for potential interperiod differences in activity level.

Results

Participant characteristics

Trial participants consisted of 12 children 5–8 years old (n = 6 males) with median (intraquartile range) characteristics as follows: age 7 (6.75, 7.25) years, duration of diabetes 3.7 (2.9, 5.7) years, HbA1c 7.7% (7.1, 7.8), and total daily insulin 20.3 (15.7, 22.2) U (Table 1).

Table 1.

Participant Characteristics

Characteristic	Value ^a
Number	12
Female, number (%)	6 (50)
Age (years)	7 (6.75, 7.25)
BMI (percentile)	67 (52.5, 84.8)
Diabetes duration (years)	3.7 (2.9, 5.7)
HbA1c (%)	7.7 (7.1, 7.8)
TDI (U/day)	20.3 (15.7, 22.2)
TDI (U/kg/day)	0.75 (0.58, 0.82)

Values reported as median (interquartile range).

TDI, total daily insulin.

Activity level, AP versus home care

Despite our attempts to produce similar activity levels between study periods, analysis of Fitbit data revealed that participants during the AP period had significantly more activity, including total steps (AP 41,002 ± 12,661 vs. home 21,046 ± 14,354), steps/min (AP 10.4 ± 3.2 vs. home 5.3 ± 3.6), and average heart rate (AP 96.9 ± 6.2 beats/min vs. home 92.9 ± 7.3 beats/min). We therefore performed analyses of BG-related comparisons using ANOVA in models that included activity measure (total steps and heart rate in separate analyses) covariates.

BG control, AP versus home care

There were no episodes of severe hypoglycemia or other adverse events during the trial. During the AP period, there were few interruptions to the system, with participants remaining in closed-loop mode 96.7% ± 0.8% of the time. Overall control was improved during the AP period compared to the home-care period (Fig. 1).

FIG. 1.

BG characteristics between study phases. Frequency of BG measurements as recorded by continuous glucose monitor in individual BG ranges during the AP and home-care study periods. AP, artificial pancreas; BG, blood glucose.

There were increases in percent time 70–180 mg/dL (AP 73.1% ± 2.7%; home 46.9% ± 2.7%, P < 0.001, P = 0.002 after adjustment for level of activity) and 80–140 mg/dL (AP 45.9% ± 1.0%; home 24.8% ± 2.0%, P < 0.001, P < 0.001 after adjustment for activity) and decreases in hyperglycemia >180 mg/dL (AP 25.8% ± 2.8%; home 51.5% ± 0.8%, P < 0.001, P < 0.001 after adjustment for activity) and >250 mg/dL (AP 6.3% ± 5.4%; home 20.8% ± 9.9%, P < 0.001, P < 0.001 after adjustment for activity). Mean BG was lower during the AP period (AP 152 ± 4.2 mg/dL; home 190 ± 7.4 mg/dL, P < 0.001, P < 0.001 after adjustment for activity). The difference in BG control between the AP and home periods was more striking at night (mean BG for AP 134 ± 4.2 mg/dL; home 200 ± 9.3 mg/dL, P < 0.001; percent time 70–180 mg/dL for AP 87.0% ± 2.6%; home 42.4% ± 7.3%, P < 0.001) than during the day (mean BG for AP 165 ± 5.3 mg/dL; home 183 ± 8.4 mg/dL, P < 0.05; percent time 70–180 mg/dL for AP 63.5% ± 3.1%; home 50.5% ± 4.6%, P < 0.05). The number of hypoglycemic events was overall low in both study periods without differences in time <70 mg/dL (AP 1.1% ± 1.1%; home 1.6% ± 1.2%). Total daily insulin was similar between periods (AP 0.681 ± 0.036 U/kg/day; home 0.733 ± 0.050 U/kg/day, P > 0.05) (Table 2).

Table 2.

Blood Glucose Parameters by Study Period

	AP	Home	P ^a	P ^b
Percent time with BG in the following range
70–180 mg/dL	73.1% ± 2.7%	46.9% ± 2.7%	<0.001	0.002
80–140 mg/dL	45.9% ± 1.0%	24.8% ± 2.0%	<0.001	<0.001
>180 mg/dL	25.8% ± 2.8%	51.5% ± 0.8%	<0.001	<0.001
>250 mg/dL	6.3% ± 5.4%	20.8% ± 9.9%	<0.001	<0.001
<70 mg/dL	1.1% ± 1.1%	1.6% ± 1.2%	NS	NS
Number of events <70 mg/dL	3.3 ± 1.0	4.0 ± 1.0	NS	NS
Total daily insulin dose (U/day)	18.6 ± 1.6	19.9 ± 1.6	NS	NS
Total daily insulin dose (U/kg/day)	0.68 ± 0.04	0.73 ± 0.04	NS	NS
Mean BG (mg/dL)	152 ± 4.2	190% ± 7.4%	<0.001	<0.001
Fitbit data
Total steps	41,002 ± 12,661	21,046 ± 14,354	<0.001	NA
Steps/min	10.4 ± 3.2	5.3 ± 3.6	<0.001	NA
Heart rate	96.9 ± 6.2	92.9 ± 7.3	<0.001	NA

Comparison of AP to home period using paired t-test.

Comparison of AP to home period using ANOVA taking Fitbit activity into account.

AP, artificial pancreas; BG, blood glucose; NA, not applicable; NS, not significant (P > 0.05).

When the first 24 h of AP use was compared to the remaining time period, there was a tendency toward improvement in the percent of time in target range during the later stage (76% vs. 67%, P = 0.052). This did not appear to be due to CGM performance, in that mean absolute relative difference (MARD) was not different between periods (14.5% ± 1.2% vs. 15.1% ± 1.2%).

Parents reported 0 instances of the children discovering the password or gaining access to AP control screens.

Discussion

We demonstrated improved glycemic control among children 5–8 years old using an AP system that resulted in an increase in time in the range of 70–180 mg/dL without increased hypoglycemia, addressing two main obstacles to improved glycemic control in this age group. Whereas a prior study demonstrated reduced hypoglycemia for young children on an AP without reducing mean BG,³ this is the first study to demonstrate improved time in target range on a young-child-specific AP. These data thus indicate the potential to use AP systems to improve overall control in young children despite their potential challenges of high insulin sensitivity and rapid fluctuations of BG with physical activity. Given suboptimal control overall among children this age, with mean HbA1c values of 8.4,¹ use of an AP may over time contribute to lower risks of long-term complications of T1D.¹⁵

The higher degree of activity during the AP session was likely due to a limitation of the experimental design, with the AP period having a different setting from the home period, with additional opportunities for physical activity. While this design was also used in a recent high-profile report of AP efficacy,¹⁰ this approach lacks the rigor of a comparison of identical settings for both periods. Nevertheless, the improved time in range during AP use did not appear to be due to the increased activity alone, as demonstrated by (1) the increase in time in target range persisted in models adjusted for physical activity, suggesting that these levels were due to AP use and not merely differences in physical activity; (2) there was no increase in overall hypoglycemia during the AP period despite more physical activity; (3) during the home-care period, parents administered corrective doses of insulin (as per usual routine) to address any hyperglycemia.

The passcode-protected lock-out screens appeared to be effective in preventing undesired access of young children to AP device settings and insulin delivery screens, in that parents reported that they did not observe their child using these screens unsupervised nor suspect that their child had discovered the password. Even though many parents trust their child with diabetes management activities in this age range,¹⁶ the potential exists for error and elevated risk of hypoglycemia.¹⁷

The degree of improvement in BG control between AP and home care among young children in terms of increase in time in target range was overall similar to what has been noted for adolescents⁹ and adults¹¹ on this system. In addition, this trial in young children using this control-to-range algorithm (increasing time in target range) differed from a model predictive control algorithm tested in young children in a prior trial (which resulted in a decrease in hypoglycemia without changes in time in target range³) in a similar way to differences noted between these algorithms in adults.^11,18 Overall, these comparisons suggest that algorithms originally designed for use in adults function similarly in young children. While additional adaptations for young child physiology (including risk of rapid decrease in BG during exercise) may improve BG control further on an AP, it appears that current systems already provide reasonable control.

This reasonable control was achieved using traditional U-100 insulin. A previous group testing AP use in young children used insulin diluted 1:4, showing a reduction in hypoglycemia but similar average glucose levels compared to usual care.³ Another group assessed overnight effectiveness of an AP system in young children age 3–6 years, using either U-100 or diluted insulin (to 20 IU/mL), finding overall similar BG control between groups, with mean BG of 122 mg/dL in both groups but with a nonsignificant tendency toward fewer episodes of mild hypoglycemia in the group using diluted insulin (1 episode vs. 6 episodes, P = 0.09).¹⁹ Overall, these findings suggest that standard insulin concentration is likely adequate in this age range.

We noted a tendency toward improved glycemic control on the AP after the first 24 h of AP use. This is of unclear cause. While the system tested does have an adaptive property, these adaptations have a time constant of 2 days to 6 days and would not have altered control parameters over the time frame of this study. The more difficult glycemic control during AP initiation may have instead been due to a higher degree of glucose variability at the start of the study compared to afterward, although more investigation is needed.

This study had several limitations. The most significant of these was that the study periods were in different settings. While parents were instructed to provide select food choices between study phases, we were unable to track the quantity and quality of food intake, potentially contributing to higher intake of carbohydrate during one study period versus the other. In addition, there was a lack of blinding regarding AP use—which remains a necessary design feature to allow parents to intervene with additional insulin doses when they note high blood sugars. However, this lack of blinding likely resulted in doses of correction factor of insulin during the home phase (but not during the AP phase), which would have been expected to lower BG levels during the home phase, biasing the data toward improved BG during this phase.

That we noted lower BG during the AP phase supports the notion that the AP does a more consistently better job of this than seen in home care. The alarms were of a different nature between the AP and the CGM during the home-care period, with predictive low-BG alarms in the AP period, compared to threshold CGM alarms that were permitted to vary in the home-care period. In theory, this could have resulted in earlier intervention for hypoglycemia in the AP period. By protocol, during the AP period, we did not intervene with treatment for low BG until the BG was ≤70 mg/dL; nevertheless, earlier awareness of an impending hypoglycemia may have resulted in earlier treatment for hypoglycemia during the AP period.

Overall, though, this would have contributed to higher and not lower mean glucose levels; the time spent with BG <70 mg/dL was very low in both treatment periods and altering time <70 would not have significantly changed the time in range 70–180 mg/dL. In addition, the activity monitor we used, the Fitbit Charge HR, does not have separate software modes for young children, potentially contributing to its limitations in exercise assessments related to heart rate or step counts.²⁰ Nevertheless, we used these activity monitors primarily as a means of detecting differences in activity between study periods, and any limitations in detecting HR and steps would be presumed to apply similarly between study periods.

We assessed the child's potential access to locked-out screens only by querying parents if they had witnessed their child gaining access or revealing the password; it remains possible that children gained access to the locked-out screens during a time when they were not being observed by parents. Finally, this study lacked systematic use of a questionnaire to query parents regarding their impressions of the system, including its function and specific features.

In conclusion, we noted that young children with T1D exhibited increased time in target range on an AP system without increased hypoglycemia, compared to usual home care. Given a suboptimal number of children in the target range for BG, this supports the notion that the AP could be useful in improving overall control for children in this age range. Further long-term studies of safety and efficacy are still needed.

Footnotes

Acknowledgments

The authors thank Tandem, Inc. (San Diego, CA) and DexCom (San Diego, CA) for providing access to their technology and Mary Oliveri (University of Virginia), who was the primary contact for the regulatory submission to the FDA and UVa IRB. Dr. Daniel R. Cherñavvsky is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. This study was supported by UVa LaunchPad.

Author Disclosure Statement

M.D.B. and B.P.K. receive research support from Tandem, DexCom, Roche, Sanofi, Ascencia (former Bayer Diabetes), and BD. M.D.B. consults for Roche, Sanofi, Ascencia, and DexCom. B.P.K. consults for Sanofi. Both M.D.B. and B.P.K. are founders of TypeZero Technologies, Inc. D.R.C. is part-time Chief Medical Officer for TypeZero. M.D.D., C.W., E.G.E., and L.L.K. have no conflicts of interest to declare.

References

Beck

, Tamborlane

, Bergenstal

, et al.: The T1D exchange clinic registry. J Clin Endocrinol Metab, 2012; 97:4383–4389.

Miller

, Foster

, Beck

, et al.: Current state of type 1 diabetes treatment in the U.S.: updated data from the T1D exchange clinic registry. Diabetes Care, 2015; 38:971–978.

Del Favero

, Boscari

, Messori

, et al.: Randomized summer camp crossover trial in 5- to 9-year-old children: outpatient wearable artificial pancreas is feasible and safe. Diabetes Care, 2016; 39:1180–1185.

Cherñavvsky

, DeBoer

, Keith-Hynes

, et al.: Use of an artificial pancreas among adolescents for a missed snack bolus and an underestimated meal bolus. Pediatr Diabetes, 2016; 17:28–35.

Tauschmann

, Allen

, Wilinska

, et al.: Home use of day-and-night hybrid closed-loop insulin delivery in suboptimally controlled adolescents with type 1 diabetes: a 3-week, free-living, randomized crossover trial. Diabetes Care, 2016; 39:2019–2025.

Wadams

, Cherñavvsky

, Lteif

, et al.: Closed-loop control for pediatric type 1 diabetes mellitus. Diabetes Manage, 2015; 5:23–35.

Forlenza

, Buckingham

, Maahs

: Progress in diabetes technology: developments in insulin pumps, continuous glucose monitors, and progress towards the artificial pancreas. J Pediatr, 2016; 169:13–20.

Phillip

, Battelino

, Atlas

, et al.: Nocturnal glucose control with an artificial pancreas at a diabetes camp. N Engl J Med, 2013; 368:824–833.

, Breton

, Keith-Hynes

, et al.: Overnight glucose control with an automated, unified safety system in children and adolescents with type 1 diabetes at diabetes camp. Diabetes Care, 2014; 37:2310–2316.

10.

Russell

, El-Khatib

, Sinha

, et al.: Outpatient glycemic control with a bionic pancreas in type 1 diabetes. N Engl J Med, 2014; 371:313–325.

11.

Anderson

, Raghinaru

, Pinsker

, et al.: Multinational home use of closed-loop control is safe and effective. Diabetes Care, 2016; 39:1143–1150.

12.

Leelarathna

, Dellweg

, Mader

, et al.: Day and night home closed-loop insulin delivery in adults with type 1 diabetes: three-center randomized crossover study. Diabetes Care, 2014; 37:1931–1937.

13.

Brown

, Kovatchev

, Breton

, et al.: Multinight “bedside” closed-loop control for patients with type 1 diabetes. Diabetes Technol Ther, 2015; 17:203–209.

14.

, Buckingham

, DeSalvo

, Shanmugham

, et al.: Day-and-night closed-loop control using the unified safety system in adolescents with type 1 diabetes at camp. Diabetes Care, 2016; 39:e106–e107.

15.

The Diabetes Control and Complications Trial Research Group: Lifetime benefits and costs of intensive therapy as practiced in the diabetes control and complications trial. JAMA, 1996; 276:1409–1415.

16.

Siminerio

, Albanese-O'Neill

, Chiang

, et al.: Care of young children with diabetes in the child care setting: a position statement of the American Diabetes Association. Diabetes Care, 2014; 37:2834–2842.

17.

Nabors

, Lehmkuhl

, Christos

, Andreone

: Children with diabetes: perceptions of supports for self-management at school. J Sch Health, 2003; 73:216–221.

18.

Kovatchev

, Renard

, Cobelli

, et al.: Safety of outpatient closed-loop control: first randomized crossover trials of a wearable artificial pancreas. Diabetes Care, 2014; 37:1789–1796.

19.

Elleri

, Allen

, Tauschmann

, et al.: Feasibility of overnight closed-loop therapy in young children with type 1 diabetes aged 3–6 years: comparison between diluted and standard insulin strength. BMJ Open Diabetes Res Care, 2014; 2:e000040.

20.

Wallen

, Gomersall

, Keating

, et al.: Accuracy of heart rate watches: implications for weight management. PLoS One, 2016; 11:e0154420.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB