Session 6, Health Care Technology Presenters: Kamakhya Das R. Dale - PDF document

Session 6, Health Care Technology Presenters: Kamakhya Das R. Dale Hall, FSA, CERA, MAAA SOA A Anti titr trust Disclaimer imer SO SOA A Presentatio ion D Discla laime

Recent technological and medical advances in diabetes Is it already time to challenge our pricing? DR KAMAKHYA DAS Chief Medical Underwriter, L&H, Asia Pacific, PartnerRe 17 June 2019

Global burden of diabetes 2

WHO: Global Burden of Diabetes Tren ends I In P Prev evalen ence e of D Diabet etes es, 1 1980-2014, 014, B By W WHO Region 3

WHO: Global Burden of Diabetes Tren ends I In P Prev evalen ence e of D Diabet etes es, 1 1980-2014, 014, B By C Country y Income Group 4

Obesity Epidemic 1994 1994 2015 2015 Patt ttern less ss/n /not t ye yet clea ear Patt ttern clea ear ! Sources: http://www.cdc.gov 24 November 2017 5

Recent medical advances impacting diabetes 6

Continuing evolution of precision health • Diagn gnosis • Moving from a single metabolite glucose to a heterogenous approach. • Com omplic lication ions • Shifting from traditional view of definite progression to complications to potential reversal of diabetes • Managem emen ent • Moving away from generalised treatment approaches to personalised treatments based on individual variability 7

Diagnosis of diabetes (Shift from a single metabolite glucose to a heterogenous approach.) 8

Diagnosis of diabetes • Classification of diabetes is undergoing a paradigm shift. (e.g. ANDIS study considers multiple factors like age at diagnosis, BMI, HbA1c, Insulin Resistance, Genotyping, etc to classify Diabetes into 5 different clusters Today’s classification ⇒ Potential Future Classification Ahlqvist E, Storm P, Karajamaki A, et al;Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of sixvariables. LancetDiabetes Endocrinol. 2018 Mar 1. pii: S2213-8587(18)30051-2. doi: 10.1016/S2213-8587(18)30051-2. 9

Output: From Spot Testing to Data Cloud HbA1c 1c Urine g e gluco cose Fasting g gluco cose Urine g e gluco cose HbA1c 1c Postprandia dial l gluco cose Urine g e gluco cose Fasting g gluco cose Patter ernof Fasting g gluco cose Hypo pogl glycem emia ias Postprandia dial l Patter ern o of Postprandia dial l gluco cose Urine g e gluco cose Hype pergly glycemia ias gluco cose 2017 2017 1967 1967 1977 1977 Ze Zeit 1767 1767 10

Complications of diabetes (Shift from definite progression to reversal) 11

Diabetes and Survival, according to Gender and Diabetes Status • Wha hat a are the di he diabetics dy s dying o g of ? • Biggest study ever: 97 prospective studies • n= 820,900 - no preexisting ⇒ Diabetes es ≈ 6 ye years r reduced life e ex expectancy (50 yr yr male) e) vascular disease ⇒ Smokin oking ≈ 10 ye years reduced ed l life ex expectancy ( y (50 yr yr mal male) • 123,205 deaths • Adjusted for age, sex, smoking status, BMI • Cause-specific deaths Seshasai SR, Kaptoge S, Thompson A, et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011;364:829-841[Erratum, N Engl J Med 2011;364:1281.] 12

International Trends of Diabetes Mortality and Complications (1995-2013)  Complications Diabetes (USA 1990-2010):  Heart attacks 68%  Strokes 53%  Amputations 52%  End stage renal disease 29%  Diabetic Blindness (1990-2000)  USA 25%  Israel 40%  Germany 50%  Mortality rates among Diabetics (1995-2013):  15% up to 40% (Every 10 Years) (USA, UK, Scotland, Canada, Taiwan, Israel) Causes? uses?  The improvement more in older adults (>65) 1. Bet etter er s screen eening 2. Bet etter er b beha ehavior Gregg, E. W., Sattar, N. & Ali, M. K. The changing face of diabetes complications. Lancet Diabetes Endocrinol. 4 , 537–547 (2016). 3. Bet etter er t trea eatmen ent 13

Diabetes Reversal- Key Observations • Counterpoint study: Low-calorie liquid formula diet • Mean weight change: 15.3 kg (in 8 weeks) • Plasma glucose normalized in a week Normalization of Liver Fat liver insulin Reawakening of the by 30% sensitivity β cell • Key ey Q Que uestions ns: • Could the return to normal glucose metabolism be maintained? • Would people with long-duration Type 2 diabetes benefit similarly? Taylor R 1 , Valabhji J 2 et al Prevention and reversal of Type 2 diabetes: highlights from a symposium at the 2019 Diabetes UK Annual Professional Conference. Diabet Med. 2019 Mar;36(3):359-365. doi: 10.1111/dme.13892. Epub 2019 Jan 25., 14

Management of diabetes (Shift from generalised to personalised treatment approach) 15

Personalised treatments based on individual variability Cluster Old New S ew Sub G b Group up Metabo Met bolic lic & & Vessel C el Complic licatio ions  High HbA1c 1 Type 1 Aut utoim immune ne Diabetes (SAID)  Insulin deficiency (impaired insulin production)  GADA-positive (glutamate acid decarboxylase antibodies)  Often insulin treatment in the short term  GADA-negative 2 Type 2 Insulin ulin-de defic icien ient diabetes (SIDD)  High HbA1c Similar to cluster 1  Low insulin secretion  Highes hest incid idenc nce o of early ly r ret etin inopathy  Often insulin treatment but time to reaching the treatment goal (HbA1c <52 mmol/mol) was longest  Insulin resistant 3 Type 2 Insulin ulin-resi sist stant Diabetes (SIRD)  Relatively low HbA1c  Persistent microalbuminuria  Highes hest incid idenc nce o of neph ephropathy ( (kidn idney)  Highest prevalence of non-alcoholic fatty liver disease  Almost no insulin treatment  Not insulin resistant 4 Type 2 Obes esit ity-related diabetes (MOD)  Almost no insulin treatment  “Hea ealt lthie ier” o obes besit ity?  Modest metabolic alterations 5 Type 2 Age ge -related diabetes  Almost no insulin treatment (MARD) Similar to cluster 4 Source: R&D PartnerRE Life&Health 16

Novel Medication Principles (from 2015) 1. 1. Glucagon L Like P Peptide 1 1 Ago gonist (GLP-1) ⇒ Incl. traditional therapy stimulate Insulin Release 2. 2. Sodium um-Gl Glucos ose C e Co-transporter Inhibitors T Typ ype 2 2 (SGLT2) ⇒ prevent Glucose Reabsorption in the Kidney 3. Di 3. Dipeptidyl yl P Peptidase 4 4 Inh nhibitors (DPP-4) ⇒ prevent Breakdown of GLP-1 ⇒ stimulate Insulin Release % ( ∅ f-up 3. SGLT2 I 2 Inhi hibitor V Vs Traditonal T Trea eatmen ent: Total mortality 32% ( 3.1 y 1 years) Zinman B, Wanner C, Lachin JM, et al. Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N Engl J Med 2015;373:2117-2128 June 7, 2019 17

AI in diabetes In God we trust, all others bring data William Denning 18

Life Cycle of New Technology AI Diabetes Control FGM (=Flash Glucose Monitoring) CGM (= Continuous Glucose Monitoring pectations GLP-1 (Glucagon Like Peptide 1 Agonist) New SGLT2 ( Sodium-Glucose subtypes Cotransporter Inhibitors Type 2) Expec Google Contact Lens  Apple Time me Smell Sensor Source: R&D PartnerRE Life&Health

Diabetes Management by AI May lead to breakthroughs (e.g. Artificial Pancreas) Database of similar glucose Data incorporated patterns of many diabetics Algorithm to decide the Sensors to into electronic best treatment option measure sleep, medical record mood, activity, food image analytics Optimization of Medical skin Diabetes interface which changes colours with change in blood sugar levels 20

Use of AI in Diabetes Management • Intelligent systems for glucose prediction and alarm generation • Clinical decision support tools to deal with the avalanche of data gathered by sensors. Data mining approaches for risk prediction and prevention of diabetes comorbidities • To build variety of solutions including closed loop systems • Deliver value based health care • Rrigorous understanding of the impact of a particular drug ug, dev evice or techno chnology to allow the evaluation of the potential impact of behaviours and treatments on cost Kerr, D, Axelrod, C, Hoppe, C, Klonoff, D. Diabetes and technology 2030: a utopian or dystopian future? Diabet Med. 2018:35(4):498-503. 21

Wearables Leapfrog T Technolo logy o on t n the he w way of e everyday life 2. 2. CGM M (= = Continuous Glucose M e Monitoring): 1. 1. FGM ( (=Flash G Glucose M e Monitoring): g): T T1+T2 T1 D Diabetes es Diabete etes  CGM: Real-time (always)  FGM: scan sensor (on demand) • Good accuracy • Backup • Alerts for Hypos • Suboptimal accuracy June 7, 2019 22

Diabetes Management by AI  Medtronic Gu Guardia ian C Connect C CGM GM System predic icts 98. 98.5% 5% accuracy  App pp o on A n Appl pple iOS de devices  “Personal D l Diabetes Assis istant”  Launch A April 2019 2019 June 7, 2019 23