Module 1 Richard M. Bergenstal, MD - PowerPoint Presentation

1. Module 1

2. Richard M. Bergenstal, MDA1CAGPAnalyze Personalize Act

3. Program in 3 ModulesModule 1: Analyze: Ambulatory Glucose Profile (AGP) Core CGM metrics Module 2: Personalize: Diabetes Management A1C to Glucose Management Indicator (GMI) to AGP Module 3: Act: Balance: rt CGM Retrospective patterns TIR TBR Pt. Experience Clinical Outcomes

4. Disclosures: Richard M. Bergenstal, MD I have participated in clinical research, been a member of a scientific advisory board, or served as a consultant for: Abbott Diabetes Care, Ascenia, CeQur, Dexcom, Eli Lilly, Hygieia, Johnson & Johnson, Medtronic, Novo Nordisk, Onduo, Roche, Sanofi, Senseonics and United HealthcareMy institution receives NIH funding: T1D (DCCT/EDIC) & T2D (GRADE) and Technology (SBIR with Hygieia) and automated insulin delivery systems (FLAIR)My employer, the nonprofit HealthPartners Institute, contracts for my services, and I receive no personal income from these activities.I am a volunteer for ADA, AACE, Endocrine Society and JDRF

5. How To Optimize Glucose Management ● Blood Glucose Monitoring (BGM) vs. ● Continuous Glucose Monitoring (CGM)288 daily data points14 days of 288 points/day in a CGM / AGP ReportA1C & BGM testing

6. Continuous Glucose Monitoring (CGM)Tiny filament sensor inserted under skin measuring interstitial glucose every 1-5 minutes and sent to receiver/phone to store dataLag time of 5-6 minutes between intravascular and interstitial compartmentsMovement towards factory calibration and non-adjunctive use (no BGM confirmation) and highest accuracy standard (interoperable CGM – iCGM) Three categoriesReal-time CGM: used continuously with alarms and alertsIntermittently scanned CGM: measure glucose continuously but displayed when patient swipes over the sensor with a reader or smart phone using cgm appl Professional CGM that is clinic owned and used for one blinded or unblinded sessionCarlson, et al. DTT 2017; 19:S4-S11; Basu, et. al Diabetes 2013; 62:4083–4087, 2013

7. Published 09/12/2019

8.

9. DexCom G5 Abbott FreeStyle Libre, Libre 2 and Libre Pro Eversense CGM (Eversense XL- Europe)Senseonics CGM Systems Medtronic Guardian Connect CGM(iPro2) G6 and G6 Pro

10. 2 Important Steps in 2019 & 2020 to achieve: 3 Key Advances in Standardizing CGM Data & Reports Battelino T, Danne D, Bergenstal R et alDiabetes Care online ahead print June 2019Endorsed by ADA, EASD, JDRF, AACE, AADE & others JUNE 2019 AGP, ambulatory glucose profile; CGM, continuous glucose monitoring; TIR, time in rangeJDRF, Juvenile Diabetes Research Foundation; EASD, European Association for the Study Diabetes; AACE, American Association of Clinical Endocrinologists; AADE, American Association Diabetes Educators ADA: Standards of Medical Care in Diabetes (January 2020) CGM revised standards – Glycemic Targets, chapter 610 core CGM metrics (including TIR’s) TIR targets “Use AGP to interpret CGM data & guide treatment decisions”January 2020!

11. Standardization of CGM metricsCV, coefficient of variation; GMI, glucose management indicator; TAR, time above range; TBR, time below rangeBattelino T, et al. Diabetes Care 2019;42:1593–6032019 core CGM metrics for clinical care1.Number of days CGM worn (minimum 10–14 days) 2.Percentage of time CGM is active (minimum 70% of data from 14 days) 3.Mean glucose4.GMI (Glucose Management Indicator)5.Glycemic variability (%CV) target ≤36%6.TAR: % of readings and time >250 mg/dl (>13.9 mmol/l)Level 2Very high7.TAR: % of readings and time >181–250 mg/dl (10.1–13.9 mmol/l) Level 1High8.TIR: % of readings and time 70–180 mg/dl (3.9–10.0 mmol/l) In range9.TBR: % of readings and time 54–69 mg/dl (3.0–3.8 mmol/l) Level 1Low10.TBR: % of readings and time <54 mg/dl (<3.0 mmol/l)Level 2Very low Use of AGP for CGM report

12. AGP reportMetrics and targetsAGP profile (14 days)Daily views http://www.agpreport.org/agp/agpreports

13. www.AGPreport.org 12345678910 1Beck RW, Bergenstal RM, Cheng P, et al. J Diabetes Sci Technol 2019;13:614-626TIR – A1C correlation1TIR 70% ≈ A1C 7%TIR 50% ≈ A1C 8%10% ΔTIR ≈ 0.5% ΔA1C

14. MGLRMORE GREEN LESS RED AGP Report: Do I have room for improvement?More than70%Less than1%Less than4%

15. AGP 14 day profileAGP Report: Where are the out of range values?Target RangeHighVery HighLowVery LowAddress Hypoglycemia FirstPursue Hyperglycemia Next

16. AGP Report Daily Glucose Profileshttp://www.agpreport.org/agp/agpreports

17. Not flat, not narrow, not in range What are we striving for in a CGM/AGP profile? Flat, not narrow, not in range Flat, narrow, not in range Flat, narrow and in range!Each 5% in TIR is clinically significant FNIR

18. AGP reportMetrics and targetsMetrics and targetsAGP profile (14 days)AGP profile (14 days)Daily views Daily views http://www.agpreport.org/agp/agpreportsMORE GREEN LESS RED FNIRPatternsFlatNarrowIn Range

19. Module 2

20. Program in 3 ModulesModule 1: Analyze: Ambulatory Glucose Profile (AGP) Core CGM metrics Module 2: Personalize: Diabetes Management 3 Ways to Personalize Diabetes Management ABC (D3 – quality measure D3,D4, D5) CKG (management algorithms)GMI (bridge A1C to TIR)

21. Worry about the bigger risk picture: optimize the D3,D4,D5ABC no smoking ASA if +CVDD3 D4 D5Clinic leadership says: Pay for Performance / Pay for Quality Trends in Diabetes Management Among US Adults:1999–2016Fang M. J Gen Intern Med 35(5):1427–34, Jan. 2020 A1C BP Chol (LDL) ABC’c

22. D5-Individual Components N >60,000; Minneapolis, MN What innovation can get us off this plateau?

23. Program in 3 Modules 3 Ways to Personalize Diabetes Management ABC (D3 – quality measure D3,D4, D5) CKG (management algorithms)GMI (bridge A1C to TIR)

24. 64 yo T2D A1C 8.4 % On metformin Atorvastatin, ACE IeGFR 44Alb:Cr ratio: 310 CKG Algorithm C K G

25. 64 yo T2D A1C 8.4 % On metformin Atorvastatin, ACE IeGFR 44Alb:Cr ratio: 310 CKG CVD CKD Glucose

26. Program in 3 Modules 3 Ways to Personalize Diabetes Management ABC (D3 – quality measure D3,D4, D5) CKG (management algorithms)GMI (bridge A1C to TIR)

27.

28. TIRA1CA1C Management Era CGMManagement Era Why not just use A1C?

29. DCCT: Relationship of HbA1c toRisk of Microvascular ComplicationsRelative RiskRetinopathyNephropathyNeuropathyHbA1c (%)151311975316789101112

30. 51%End DCCT “HbA1c Era”53%68%28%Population Health or Accountability Measure

31. HbA1c – The gold standard?Aijan RA. Diabetes Technol Ther 2017;19:S27–S36Beck RW, Bergenstal RM, Fallacy of Average: How Using A1C Alone to Assess Glycemic Control Can Be Misleading D Care Aug 2017 40:994-999. Easy to measure and standardizeRelatively cheapPredictive of vascular complicationsOften used to guide management decisions AdvantagesLimitationsHbA1cThe gold standard There is a need for metrics beyond HbA1c to address these limitationsHbA1c is a poor indicator of hypoglycemia & glucose variabilityPotential confounding by some conditions (e.g. renal failure, liver disease, iron deficiency hemoglobinopathies, varying RBC lifespans, genetics, rates of hemoglobin glycation)Only provides an average measure of glycemia

32. Factors Impacting Laboratory Measured A1CFactors impacting A1C measurement on an individual basis:Race and ethnicity (e.g. A1C in African Americans is about 0.2-0.4 % points higher)Factors impacting RBC life spanHemoglobinopathyBlood lossHemolysisPregnancyVariability in glycation of hemoglobin Hgb variants, metabolic factors or substances affecting assayVariability in average glucose levels over the short-termOther undefined variables impacting the measure (relatively common)Radin MS. J Gen Intern Med. 2014;29(2):388‐394.Bergenstal et al Racial Differences in Relationship of Glucose and A1C Annals Int. Med 2167: 365-374, 2017

33. RBC lives in circulation 100 days and average age of a RBC is 49 daysBut RBC mean age varies between people by 15-20% If glucose is 200 mg/dL all day and expect A1c 8.6% RBC mean age 38 days get Lab A1c =7.5%RBC mean age 60 days get Lab A1c =9.9%

34. HbA1C% Time HypoGlu. VariabilityTIR(%)< 70 mg/dL(CV %)70 – 180 mg/dL6.76.76.7169264253836951CGM Profile or A1C?

35. Estimated A1C (eA1C)Estimates what the laboratory A1C would be based on BGM or CGM derived mean glucoseEquation: eA1C = mean glucose in mg/dL + 46.7/28.7 eA1C= 200 mg/dL + 46.7/ 28.7 = 8.6%Nathan et al. Diab Care. 2008; 31:1473-1478. Mean glucose eA1C126 mg/dL6%154 mg/dL7%183 mg/dL8%212 mg/dL9%240 mg/dL10%269 mg/dL11%

36. Your patient has a mean blood glucose of 185 mg/dL, what is their approximate A1C?7.5%8.0%8.5%9.0%9.5%10.0%

37. Beck R, Bergenstal R et al. Fallacy of Average… D. Care Aug. 2017 Perspectives in Care In both data bases an individual with a mean glucose of 185 mg/dL we could predict to have an A1c of: 7%or 8%or 9%

38. Mean CGM Glucose by HbA1c - Overall (N=208)

39. www.AGPreport.org 4 1Beck RW, Bergenstal RM, Cheng P, et al. J Diabetes Sci Technol 2019;13:614-626

40. Meet Bill 7.8Lab A1C7.8%65 y.o. with type 2 DM 15 yrsCurrent therapy:Metformin 2000 mg/dayLispro 8-12 units/mealGlargine 30 units at bedtimeImpaired awareness of hypoglycemiaBill and his physician are frustrated because which one do you believe? Why the big difference between the two values? What should his A1C and eA1C targets be?

41. Food and Drug Administration (FDA) Concerns and RecommendationseA1C maybe be helpful, but clinicians and patients confused over discordance between lab A1C and eA1CeA1C terminology implies more direct relationship between lab A1C and eA1C than there really isFDA decided nomenclature of eA1C had to change to avoid confusion and mandated that eA1C be removed from CGM reports in the U.S.FDA asked a group of diabetes technology experts to come up with new terminology with the following caveatsNo use of “estimated” or “A1C”No use of word “index” to avoid confusion with glycemic index Bergenstal et al. Diab Care. 2018; 41:2275-2280.

42. There is a much bigger difference in glucose derivied to A1c mismatch

43. Why GMI? A New Way to Gauge Glycemic Management Replaces estimated A1C (eA1C)Avoids confusion when eA1C and laboratory A1C do not matchGMI is less “judgmental” and conveys it was derived from glucose valuesGMI and laboratory A1C may not match due to differences in lifespan of RBC, timing of laboratory A1C vs. CGM data (remember A1C is variable)For the math majors!Bergenstal et al. Diab Care. 2018; 41:2275-2280.

44. Difference Between GMI and Laboratory A1C 19% of the time values match51% of the time absolute difference will be ≥0.3%28% of the time absolute difference will be ≥0.5%n=528Bergenstal et al. Diab Care. 2018; 41:2275-2280.

45.

46.

47. “FDA believes the authors’ proposal to reshape the narrative around this metric through the proposed new terminology is an important step for the proper use of the metric and that the inclusion of GMI in certain diabetes management devices (e.g., software) could potentially offer additional insight into glycemic control.”

48. Clinical ScenariosWhen GMI and Laboratory A1C Will Not Likely AgreeGMI higher than laboratory A1CShort periods of acute hyperglycemiaExamples: illness, steroid therapy, DKAGMI lower than laboratory A1CShort periods of lower than normal glucose readingsExamples: starting low carbohydrate diet, extreme physical activity, starting new diabetes medication

49. Talking Your Patients about GMI and Laboratory A1CBergenstal et al. Diab Care. 2018; 41:2275-2280.

50. Mismatch: Lab A1c and blood glucose Lab A1c vs. Predicted A1c from FPG (HGI) (hemoglobin glycation index) Lab A1c vs. Predicated A1c from fructosamine Lab A1c vs. Predicated A1c from CGM mean glucose (GMI) or old estimated A1cLab A1c vs. Predicated A1c from BGM mean glucose (estimated A1c) Lab A1c vs. A1c adjusted / corrected for calculated RBC lifespan RBC lifespan and Hb glycation rate

51. Impact of HbA1c-Mean Glucose Discordance The ACCORD StudyRCT evaluating an intensive management approach to lower HbA1c to <6% in type 2 diabetesN=10,251, mean age 62 yrs, median HbA1c 8.1%Trial stopped early because of increased mortality in intensively-treated groupPost hoc analysis by Hempe et al (Diabetes Care 2015)Participants with HbA1c higher than estimated HbA1c (from fasting glucose correlation with A1c for the population) more likely to experience severe hypoglycemia than those with HbA1c lower than HbA1c estimated from mean glucoseHigher mortality rate present only in intensive group participants with HbA1c level higher than HbA1c estimated from mean glucose

52. Hemoglobin glycation index may predict CKD risk in diabetes A high hemoglobin glycation index is associated with an increased risk for chronic kidney disease during 10 years of follow-up among treatment-naive adults with prediabetes or type 2 diabetes.2,187 adults with prediabetes or type 2 diabetes, using data from the Korean Genome and Epidemiologic Study. HGI as the measured HbA1c minus predicted HbA1c, calculated from the linear relationship between HbA1c and FPG. Third HGI tertile group showed the highest adjusted HR for incident CKD (HR = 1.57; 95% CI, 1.06–2.34), w ith results persisting after adjustment for HbA1c. .“In our study, a higher HGI was significantly associated with incident CKD even after adjusting for traditional risk factors for CKD and HbA1c levels,” Kim said. “These findings suggest that the HGI may have an additional impact on the decline in renal function beyond the influence of HbA1c level.”EASD September 2020 Oral PresentationWonjin Kim, MD, endocrinology at CHA University School of Medicine in Seoul, South Korea

53. A1CA1C Management Era CGMManagement Era GMIAGPTIR

54. Module 3

55. Program in 3 ModulesModule 1: Analyze: Ambulatory Glucose Profile (AGP) Core CGM metrics Module 2: Personalize: Diabetes Management A1C to Glucose Management Indicator (GMI) to AGP/TIR Module 3: Act: Balance: rt CGM Retrospective patterns TIR TBR Pt. Experience Clinical Outcomes

56. What Patients with Type 2 Diabetes Should be Considered For CGM? American Diabetes A. 7. Diabetes Technology: Standards of Medical Care in Diabetes‐2020. Diabetes Care. 2020;43(Suppl 1):S77‐S88.

57. IDC CGM in Type 2 Diabetes RecommendationsHighest to Lowest Benefit Highest benefit:Patient taking multiple daily injections (MDI).Patients with severe hypoglycemia (needing assistance to treat) or with frequent mild to significant hypoglycemia (BG <70 mg/dL, <54 mg/dL).Patients with mismatch A1c and GMI High benefit:Patients treated with insulin and/or sulfonylurea regardless of A1C.Clinician/educator desire CGM for behavior, lifestyle, regimen modifications.Patients with high BG fluctuations/variability.Patients with diabetes complications (e.g. gastroparesis, renal impairment, diminished visual acuity).Insufficient glucose data to make decision on regimen change.

58. IDC CGM in Type 2 Diabetes RecommendationsHighest to Lowest Benefit Moderate benefit:Patients frequently treated with glucocorticoids (3 times or more/year).Situations where barriers to SMBG exist including SMBG avoidance, visual impairment, illness, cognitive issues, care giver assisting with care. Dramatic change in existing activity (e.g. knee replacement, stroke, sleep apnea, bariatric surgery).Less benefit:The patient desires CGM to improve health and does not meet other criteria.Finger sticks are an issue due to occupation (e.g. musicians). Newly diagnosed patients initiating medical nutrition and activity therapy with metformin and other therapies that don’t cause hypoglycemia.

59. What is the best way to Act on data from CGM Report?Retrospective patterns rt CGM Think FastCorrective / preventive action Use Arrows for insulin dosing Note Trends due to food or exercise Think Slow AGP Report Treat to MGLR Treat for Hypo first, then HyperWork toward FNIR Nobel Prize in Economics

60. What is the best way to Act on data from CGM Report?Retrospective patterns rt CGM Think FastCorrective / preventive action Use Arrows for insulin dosing Note Trends due to food or exercise Think Slow AGP Report Treat to MGLR Treat for Hypo first, then HyperWork toward FNIR

61. American Diabetes Association Standards of Medical Care in Diabetes 2020. Diabetes Care. January 2020 Volume 43, Supplement 1

62. AGP Report 1 T2D; non-insulin3 months ago: A1C 8.2%; TIR 55%MetforminSGLT2 inhibitor due to GFR 58Reluctant to start 3rd medPatient asked for 3 months to make lifestyle changes

63. 2Observe3Learn

64.

65.

66. AGP Report 2Visit 2 – (4 mo. later)A1c 7.0 %; TIR 78%MetforminSGLT2 inhibitorChanges made with CGM Saw what impacted his glucose. Focused on:Taking meds as prescribedGave up sweet tea at lunch Followed plate methodIncreased activity, 4x/wkClinicial Outcomes Patient Experience

67. What is the best way to Act on data from CGM Report?Retrospective patterns rt CGM Think FastCorrective / preventive action Use Arrows for insulin dosing Note Trends due to food or exercise Think Slow AGP Report Treat to MGLR Treat for Hypo first, then HyperWork toward FNIR

68. Johnson ML, Martens T, Carlson A., Criego A, Simonson G. Bergenstal RM Diabetes Technol Ther. 2019 Jun;21(S2):S217-S225

69. Met/SUMet72-year-old femaleT2D for 12 yearsBMI 32 A1C 7.9eGFR >60 + History of ASCVD W B L D BT AGP Report 14 days Jean’s first CGM report www.AGPreport.org7.2%7.9%Step 8: Action plan Step 7: Compare to past AGP Step 9: AGP into EHR & to patient GLP-1 RA or SGLT-2.

70. Jean’s Second Professional CGMMetMet72-year-old femaleT2D for 12 yearsBMI 32 A1C 7.1eGFR >60 + History of ASCVDAGP Report 14 days W B L D BT Weekly Semaglutide7.1%6.5%

71. GMI TIR & Standardizationof CGM Reports in Clinical PracticeAnders Carlson, Amy Criego, Thomas Martens, Richard BergenstalEndocrinol Metab Clin N Am 49 (2020) 95–107GMI 7.2%60 yo man with T2D No Hx: CVD or CKDA1c – 7.5% Wt 95 kgMetformin 2000/d Glimepiride 4 mg/d Glargine 70 units q HS Classic – “stair step” picture of: “over basalinization” Start meal-time insulin Cover 1 or 3 meals/day?Reduce glargineStop SU

72. American Diabetes Association Standards of Medical Care in Diabetes 2020. Diabetes Care. January 2020 Volume 43, Supplement 1What if you think really slow – carefully –what did that new ADA Algorithm say?

73. American Diabetes Association Standards of Medical Care in Diabetes 2020. Diabetes Care. January 2020 Volume 43, Supplement 1

74. GMI TIR & Standardizationof CGM Reports in Clinical PracticeAnders Carlson, Amy Criego, Thomas Martens, Richard BergenstalEndocrinol Metab Clin N Am 49 (2020) 95–107GMI 7.2%60 yo man with T2D No Hx: CVD or CKDA1c – 7.5% Wt 95 kgMeds: GLP 1- not tolerated Metformin 2000/d Glimepiride 4 mg/d Glargine 70 units q HS Classic – “stair step” picture of: “over basalinization” Start GLP1 agonist Stop SUReduce basal insulin dose

75. A1CAGPSummary Consider crossing the bridge from A1c management era to the CGM/AGP management era 2. Learn to Analyze the 10 core CGM metrics in the AGP ● MGLR ● Hypo First ● FNIR 3. Personalize treatment plan by considering ● ABC’s ● CKG Algo ● GMI 4. Act on AGP by balancing ● rt & retrospective ● TIR & TBR ● Pt Exp & Clinical Outcomes