A terápiás hatékonyság értékelésének szempontjai az újabb antidiabetikus készítmények tükrében Kocsis Győző Péterfy Sándor utcai Kórház-Rendelőintézet és Baleseti Központ Diabetes Ambulancia Budapest

Miről lesz szó? Tények és számok 2-es típusú cukorbetegség korunk egészségügyének egyik legnagyobb kihívása A hatékony kezeléssel szembeni elvárások Jelenleg alkalmazott kezelés eredményei és költségei Következtetések, összefoglalás

2017.04.08. IDF Diabetes Atlas 5th Edition 2012 Update 3 3

2017.04.08. The figure presents the top 10 countries for numbers of people with diabetes in millions. All but two of these countries are middle-income countries and rapidly developing. Combined, these countries make up 75% of the total prevalence of diabetes in the world. Urbanisation and the accompanying changes in lifestyle are the main drivers of the epidemic in addition to changes in population structure where more people are living longer. The health systems of most of these countries are not equipped to deal with the rapidly rising burden of diabetes. IDF Diabetes Atlas 5th Edition 2012 Update 4 4

2017.04.08. A staggering 50% of people with diabetes do not know they have the disease. Diabetes can go many years without showing symptoms, or symptoms may be misdiagnosed as other conditions, meanwhile high blood glucose is causing damage to major organs in the body. Complications such as cardiovascular disease, neuropathy, retinopathy, and kidney disease are irreversible once they develop and can mean serious disability for the person who experiences them. Regions where the overall prevalence of diabetes is relatively low, such as Africa, have the some of the highest percentages of people who are undiagnosed. This is often because of a complete lack of awareness of the disease both in the public and the health community. IDF Diabetes Atlas 5th Edition 2012 Update 5 5

2017.04.08. Estimating deaths from diabetes is difficult because it is rarely reported as the cause of death in medical records. However, studies using the risk of death due to diabetes in different populations allow us to generate more realistic estimates of the true mortality burden. The estimates are almost certainly an underestimate but show us that a substantial proportion of deaths occur in people under the age of 60 and many even under the age of 50. This is well before the life expectancy in many countries and is a result of the serious complications that can develop in untreated diabetes. IDF Diabetes Atlas 5th Edition 2012 Update 6 6

2017.04.08. Looking at diabetes deaths against spending for diabetes care shows us the impact of a lack of investment very starkly. In countries where very little is spent on diabetes, the rate of death is almost double that in high-income countries where the vast majority of money for diabetes is spent. IDF Diabetes Atlas 5th Edition 2012 Update 7 7

A cukorbetegség előfordulása Magyarországon A cukorbetegség előfordulása a 20-69 éves magyar ffi: 10%, nő: 6% 55 év felett 16-20 % Jermendy Gy et al. A cukorbetegség és az emelkedett éhomi vércukor prevalenciája a hazai felnőttkorú lakosság körében: reprezentatív keresztmetszeti szűrővizsgálat eredményei. Magyar Belorvosi Arch 2008,61:203-7

Az egészségügyre fordított kiadások a GDP százalékában Magyarország GDP-je 2012-es KSH adatok alapján 28 252 milliárd Ft volt 150 milliárd USD dollár (Word Bank) A GDP 7,6%-át (2 146 milliárd Ft) fordítottuk egészségügyi kiadásokra 2012-ben. Ugyanez az arány 2005-ben 8,2% volt, EU-s átlag ’95-től 2009-ig folyamatosan nő Mo. és EU-s átlag közötti leg- kisebb különbség 2003-ban volt (0,2%pont) A legnagyobb különbség pedig 2009-ben volt (2,17%pont)

Összes egészségügyi kiadások a GDP arányában, 2010 2017.04.08. Összes egészségügyi kiadások a GDP arányában, 2010 Public expenditure= közkiadás: központi költségvetésből (általában adó) és/vagy társadalom biztosításból (általában járulék) finanszírozott egészségügyi kiadások Private expenditure= magánkiadás: magánbiztosítások (ezek lehetnek önkéntes, vagy kötelező magánbiztosítások) Public expenditure= közkiadás: központi költségvetésből (általában adó) és/vagy társadalom biztosításból (általában járulék) finanszírozott egészségügyi kiadások Private expenditure= magánkiadás: magánbiztosítások (ezek lehetnek önkéntes, vagy kötelező magánbiztosítások) Forrás: Health at a Glance Europe 2012 10 10

Egészségügyi kiadás vs. egészségügyi mutatók Korai halálozás : A korai halálozásból eredő várható élettartam-csökkenés szintén Magyarországon a legmagasabb (WHO-HFA) Korai halálozás: A 65. életév betöltése előtt bekövetkezett halálozás. (Ádány et al 2003)

Egészségügyi kiadás vs. egészségügyi mutatók 2017.04.08. Egészségügyi kiadás vs. egészségügyi mutatók Betegség specifikus halálozás: a cukorbetegség okozta halálozás 2005 óta folyamatosan csökken, aránya továbbra is meghaladja a Visegrádi országok eredményeit Forrás: OECD Statistics 12 12

A diabétesz „járvány” Cukorbetegek száma 13 A diabétesz „járvány” Cukorbetegek száma A 2-es típusú cukorbetegség rizikó faktorai Region China: 98 million Japan & Korea: 10.5 million International Operations: 201.5 million Europe: 45 million North America: 27 million Obesity Poor diet Physical inactivity Advancing age Family history of diabetes Ethnicity High blood glucose during pregnancy affecting the unborn child A cukorbetegség becsült költsége Global diabetes healthcare costs in 2013: USD 548 billion, projected to exceed USD 627 billion by 2035 Global diabetes healthcare costs accounted for 10.8% of total healthcare costs worldwide in 2013 Deaths due to diabetes in 2013: 5.1 million 2. The diabetes pandemic 1. International Diabetes Federation. IDF Diabetes Atlas, first edition, 2000:8. 2. International Diabetes Federation. IDF Diabetes Atlas, sixth edition, 2013:11.

Novo Nordisk NovoMix <presenter name> <reference> A 2-es típusú cukorbetegség és szövődményeinek kezelése jelentős gazdasági teher 14 €1,373 1.3x increase €1,723 €3,355 2.4x €3,436 2.5x increase €5,642 4.1x All insured patients Diabetes, no complications Diabetes, microvascular complications Diabetes, macrovascular complications Diabetes, micro- and macrovascular complications Annual costs per patient1 In 2013, ~10.8% (~$548b USD) of the total healthcare expenditure worldwide was spent on treating diabetes2 Based on an exchange rate of 1 Euro = 1.4156 US dollars. Exchange rate as of 30 Oct 2011 1Liebl et al. Dtsch. Med. Wschr 2001;126:585–89 (CODE-2 Study); 2International Diabetes Federation. IDF Diabetes Atlas, 6th edn. Brussels, Belgium, 2013: http://www.idf.org/diabetesatlas Diabetes is an increasing and widespread chronic disease causing considerable costs for the health care system. In the CODE-2® Study (Costs of Diabetes in Europe –Type 2) the total expenses for type 2 diabetics in Germany were evaluated and analyzed for the first time. Patients and Methods: The CODE-2®-study has been performed in eight European countries. In the German arm of the study, medical, demographic, and economic data of 809 patients were obtained retrospectively for a one year period, using face-to-face interviews with 135 physicians. These results were projected for the overall population of type 2 diabetes patients in Germany. Results: The annual costs caused by type 2 diabetes patients in Germany in 1998 amount to 31.4 billion DM. The majority of these costs (61%) were covered by statuatory and private health insurance. The annual expenses of the statuatory Health Insurance (SHI) for these patients amounted to 18,5 billion DM. These costs divided in 50% spent for inpatient treatment, 13% for ambulatory care, and 27% for medication. Diabetes medication (Insulin, oral antidiabetic drugs) accounted for only 7% of total SHI costs. Only 26% of all diabetic patients were adjusted to HbA1c values <6,5% according to the therapeutic targets of the European Diabetes policy group. 50% of the type 2 diabetic patients exhibited severe macro- and/or microvascular complications. The costs per patient – compared to the average expenses for SHI insured patients – increased with complication state from the 1.3- fold (no complications) up to the 4.1-fold (macro- and microvascular complications). Conclusions: The overall costs for patients with type 2 diabetes are higher than expected from previous estimates. Diabetes related complications and concomitant diseases are the predominant reasons for these high costs. Control of blood glucose is inadequate for the majority of diabetic patients. To prevent longterm complications, an optimized treatment of type 2 diabetes is imperative not only from a medical but also from a health economics point of view.

A megfelelő anyagcsere kontroll elérése világszerte kihívást jelent: átlagos HbA1c 2-es típusú cukorbetegségben China 9.5%11 India 8.7-9.6%9,11 Japan 7.05-9.5%11 Korea 7.9-8.7%4 Russia 9.5%11 Canada 7.36-8.6%11 Latin 7.6%1 America US 7.2%7 Germany 8.42-9.2%8 Greece 8.911-9.7%3,8 Italy 8.4%11 Poland 8.9%11 Portugal 9.7%3 Romania 9.9%3 Spain 9.2%8 Sweden 8.7%3 Turkey 10.6%3 UK 8.610-9.8%2 NB. This data is taken from (mostly) epidemiological and observational studies. The inclusion criteria from some of the studies means that the data cannot be interpreted as the mean HbA1C for all people with Type 2 diabetes in each particular study. However, the slide provides a useful overview of the extent of the problem of glycaemic control for a population of people with diabetes in each country and globally. The following notes provide information about the studies and their inclusion / exclusion criteria that may skew the data. 1Lopez Stewart et al, 2007 A multi-centre, cross sectional, epidemiological survey across 9 countries in Latin America. General Practitioners were asked to provide care and control data for patients with type 2 diabetes, ages 18-75 years. Information provided included demographics, medical and medication history, laboratory exams and information on the challenges of patient management. 2Kostev & Rathmann, 2013 Longitudinal data from general practices in Germany and UK (Disease Analyser) from 1995 to 2010 were analysed. Patients with type 2 diabetes who started their insulin treatment from 2005 to 2010 were analysed regarding the time from first diabetes diagnosis (index date) and the first insulin prescription, including 6368 patients (age 68[SD:12] years) in Germany and 1998 patients (age 64[12] years) in UK. 3Oguz et al, 2013 Study name ‘TREAT’. This was a prospective 24 month, observational study in patients with type 2 diabetes initiating insulin in clinical practice. Patient characteristics were collected at baseline and metabolic outcomes at 3, 6, 12, 18 and 24 months after initiation. 985 patients were enrolled, 886 at baseline and 734 at 24 months. 4Ko et al, 2007 People with type 2 diabetes (n=547) who were hospitalised from Dec 1999 to Dec 2000 were randomly assigned to two groups. 219 undertook an inpatient structured intensive diabetes education programme (SIDEP) and the remaining patients received conventional glycemic control without intensive education. After discharge patients were monitored regularly. Laboratory data were obtained, and adherence to self-care behavior was determined. 5Arai et al, 2009 This was a cross-sectional survey. 8112 clinics and hospitals (randomly) across Japan were asked to participate in this study, 721 agreed. A total of 15,652 patients aged 15-97 with type 1 and type 2 diabetes were enrolled. 6Harris et al, 2005 This was a national, cross-sectional study to provide insight into the care and treatment of type 2 diabetes in the Canadian primary care setting. Participating primary care physicians completed chart audits for the first 10 patients attending their clinic. 7Hoerger et al, 2007 Three consecutive waves of the National Health and Nutrition Examination Survey (NHANES) were used to examine trends in HbA1C. 8Liebl et al, 2012 Data were collected over 24 months from patients requiring insulin initiation as part of their usual care, in this prospective, observational study. Patients were recruited at primary and secondary care centres that routinely treat large numbers of people with type 2 diabetes. Table 1. 9Shah et al, 2009 Also called the IMPROVE study – this is an open-label, non-randomised, observational study to determine the safety and efficacy of Biphasic insulin aspart (BIAsp 30). This was carried out in 11 countries – this report includes the findings from India only. All patients with type 2 diabetes requiring insulin and considered suitable for BIAsp 30 therapy based on their physician’s clinical judgement were eligible to enter the study. Data was recorded at baseline (including demographics, detailed medical history, physicians-cited reasons for staring BIAsp 30 treatment, and the chosen treatment regimens. 10 Blak et al, 2012 This was a retrospective analysis of patients with type 2 diabetes within the Health Improvement Network UK primary care database. Patients with type 2 diabetes receiving basal insulin between Jan and June 2006 were followed until June 2009. 11Valensi et al, 2009 Also called the IMPROVE study – this is an open-label, non-randomised, observational study to determine the safety and efficacy of Biphasic insulin aspart (BIAsp 30). This was carried out in 11 countries. All patients with type 2 diabetes requiring insulin and considered suitable for BIAsp 30 therapy based on their physician’s clinical judgement were eligible to enter the study. Data was recorded at baseline (including demographics, detailed medical history, physicians-cited reasons for staring BIAsp 30 treatment, and the chosen treatment regimens. 1Lopez Stewart et al, 2007; 2Kostev & Rathmann, 2013; 3Oguz et al, 2013; 4Ko et al, 2007; 5Arai et al, 2009 – Type 1 & 2 Diabetes; 6Harris et al, 2005; 7Hoerger et al, 2008 – Type 1 & 2 Diabetes; 8Liebl et al, 2012; 9Shah et al, 2009; 10 Blak et al. 2012; 11Valensi et al, 2008

A „felező szabály” (The ‘Rule of Halves’) According to the rule of halves*, only around 6% of people with diabetes live a life free from diabetes-related complications. *Hart J.T., Rule of Halves: implications of increasing diagnosis and reducing dropout for future workload and prescribing costs in primary care, Br J Gen Pract 1992, March; 42(356):116–119, and W.C.S. Smith, A.J. Lee, I.K. Coombie, H. Tunstall-Pedoe, Control of blood pressure in Scotland: The rule of halves, Br. Med. J, 300 (1990): 981–983. **Actual rates of diagnosis, treatment, targets and outcomes vary in different countries.

Klinikai vizsgálatok szigorúan kontrollált körülményei között, még inzulinnal sem éri el sok beteg a megfelelő vércukorkontrollt, a treat to target megközelítés ellenére sem Detemir NPH Glargine a 1b 2b 3 4 20 weeks 24 weeks 52 weeks aHbA1c responders <7% are reported for combined treatment target groups of FPG = 3.9–5.0 mmol/L and FPG = 4.4–6.1 mmol/L bHbA1c ≤7.0% without hypoglycaemia 1Blonde et al. Diabetes Obes Metab 2009;11:623-631; 2Eliaschewitz et al. Arch Med Res 2006;37:495-501; 3Riddle et al. Diabetes Care 2003;26:3080–3086; 4Rosenstock et al. Diabetologia 2008: 51:408-416; Blonde (2009) - TITRATE™ study looked into 2 treatment arms using 3.9–5.0 or 4.4–6.1 mmol/L FPG targets using a patient-directed algorithm. A majority of subjects in both titration groups achieved ADA/EASD recommended guidelines of <7% HbA1c. Riddle (2003) – Overall target achievers of <7% Rosenstock (2008) - 52-week multinational, randomised, open-label, parallel-group, non-inferiority trial compared clinical outcomes with basal insulin analogues detemir and glargine in patients with type 2 diabetes

Jó anyagcsere kontroll számít: kedvező hatás a 20-30 éves követés utáni eredményekben is UKPDS eroriginal results: Intensive vs. conventional treatment 2007 (30 years) 10-year post-trial follow-up (non-interventional) 1977–1991 Randomisation 1997 (20 years) 9%* 12%* 15%* 16% Any diabetes-related endpoint Myocardial infarction Microvascular disease A glykaemiás kontroll javítása 2-es típusú cukorbetegségben csökkenti a szövődmények kockázatát 24%* 25%* Animated slide. The long-term data for the UKPDS trial. Long-term follow-up of the population studied in the UKPDS study demonstrated that the patients reaching strict glycaemic control soonest had the lowest incidence of cardiovascular diseases, including myocardial infarction. Notably, patients in the intensive therapy group, those being treated with sulphonylurea and insulin, had better glucose levels in the early stages of their therapy, and benefits seem to persist a decade later. *p<0.05; intensive vs. conventional treatment Adapted from Holman et al. N Engl J Med 2008;359:1577–89; UKPDS Study Group. Lancet 1998;352:837–53

HbA1c range reduction (%) A jelen terápiás lehetőségei: Jelentős különbség van a HbA1c csökkentő hatásukban HbA1c range reduction (%) Compound Insulin1 GLP-1 RA†2 0.0 Metformin3 Glinides4 α-GIs5 DPP-IV I6 SUs7 TZDs8 -4.0 -1.0 -2.0 -3.0 -0.5 -1.5 -2.5 -3.5 1.5 – 3.5% 0.97 - 1.51% 0.14 - 1.40% 0.17 – 0.60% 0.44 – 0.74% 0.60 – 0.70% 0.30 – 1.70% 0.80 – 1.20% Insulin and GLP-1 receptor agonists are effective glucose lowering agents †In combination with metformin. α –GI, Glucosidase inhibitor; DPP-IV I, DPP-IV inhibitors; TZD, thiazolidinediones; SU, sulfonylurea; GLP-1, glucagon-like peptide-1 receptor. 1Nathan et al. Diabetes Care 2009;32:193-203; 2Victoza summary of product characteristics, 2012; 3Glucophage package insert, 2009;4Prandin package insert, 2011;5Precose package insert, 2011;6Januvia package insert, 2013; 7Glucotrol prescribing information, 2013; 8Actos prescribing information, 2011. Insulin: No dose limit, rapidly effective (Nathan et al. 2009) GLP-1 receptor agonist: Victoza Summary of product characteristics, 2012, pg. 9,11. Metformin: Table 2, pg 1, Table 3 & Table 7, pg 2. Glucophage PI, Bristol-Myers Squibb, 2009. Based on 24-29 weeks of monotherapy, with up to 2250 mg/day Glinides: Page 2, Prandin PI, Novo Nordisk 2011. Reduction based on 24 weeks of 8mg/day monotherapy. Reduction based on mean change from baseline. α-Gis: Table 1, pg 2, Precose PI, Bayer, 2011. Reduction based on monotherapy with up to the maximum recommended daily dose of 100 mg. Reduction represent placebo-subtracted differences. DPP-IV inhibitors: page 12, Table 4, pg 13, Januvia PI, Merck, August 2013. Reduction based on 24 weeks of 100mg/day monotherapy. Reductions represent mean change from baseline. SUs: pages 13-15, Tables 3 and 4. Amaryl prescribing information, January 2013. Sanofi-Aventis US. Reductions based on monotherapy treatment, 14-22 weeks, up to the maximum recommended dose of 8mg/day. TZDs: pages 29-32, Tables 17-19. Actos prescribing information, 2011. Takeda Pharmaceuticals America, Inc. Reductions based on monotherapy treatment, 16-26 weeks, up to 45mg/day.

A kombinációs kezelés (metforminhoz adott második gyógyszer) hatékonyság szerinti értékelése network meta-analízis alapján 20 1. csoport: inzulinok (BOT, bifázisos inzulin) -0,88% és -1,07% GLP-1 agonista -1,02% 2. csoport: SU: -0,82% TZD: -0,82% glinid: -0,71% DPP-4 gátló: -0,69% HbA1c változás a kiindulási értékekhez képest a placebóhoz viszonyítva értendő A hatékonysági eredményeket mutatja az első látásra bonyolultnak tűnő network metaanalízis ábra is. Ezek alapján a SU és a DPP4-gátlók között nem volt számottevő különbség a hatékonyság tekintetében. Összefoglalva: a metformin mellé adott DPP-4 gátló kombinációs terápia hatékonysága közel azonos a metformin mellé adott SU, TZD, glinid kombinációs kezelés hatékonyságával (átlagos HbA1c csökkenés: -0,69%). Hypoglykaemia kockázat és testsúlyra gyakorolt hatás alapján azonban e kombinációs kezelés kedvezőbbnek bizonyult, mint a hasonló hatékonysággal rendelkező a MET + SU, TZD, glinid kombinációs kezelések. 39 were included in this meta-analysis Liu és mtsai, Diabetes, Obesity and Metabolism 2012; 14: 810–820 20

A kombinációs kezelés értékelése a biztonságosság szempontjából (hipoglikémia rizikó) A placebo-csoporthoz képest szignifikánsan nagyobb hipoglikémia kockázat: Gyógyszer Esélyhányados SU 8,86 glinid 10,51 bázis-inzulin 4,77 bifázisos inzulin 17,78 A placebo-csoporthoz képest a hipoglikémia kockázat nem számottevő: Gyógyszer Esélyhányados TZD 0,5 alfa-glukozidáz gátló 0,4 DPP-4 gátló 1,13 GLP-1 agonista 0,92 Liu és mtsai, Diabetes, Obesity and Metabolism 2012; 14: 810–820

Gyógyszer Ts. növekedés Ts. csökkenés Ts. semleges SU +2,17 kg glinid A kombinációs kezelés (metforminhoz adott második gyógyszer) értékelése testsúlyváltozás szempontjából (network meta-analízis) Gyógyszer Ts. növekedés Ts. csökkenés Ts. semleges SU +2,17 kg glinid +1,4 kg TZD +2,46 kg bázis inzulin +1,38 kg bifázisos inzulin + 3,41 kg GLP-1 agonista -1,66 kg α-glukozidáz-gátló -1,01 kg DPP4-inhibitor +0,23 kg Liu és mtsai, Diabetes, Obesity and Metabolism 2012; 14: 810–820

A kezelést intenzifikálását hátráltató tényezők Megnövekedett orvosi terhek3 Beteg-együttműködés hiánya2 Bonyolult rezsim1 Testsúly növekedés1 Hypoglykaemia1 Új kezelési stratégiákra van szükség a jobb vércukor-kontroll eléréséhez, a mellékhatások minimalizálása mellett 1Kunt and Snoek. Int J Clin Pract 2009; 63(Suppl. 164):6–10;2Vijan et al. J Gen Intern Med 2005; 20:479-482; 3Cuddihy et al. Diabetes Educ 2011; 37(1):111-23

A hypoglykaemiától való félelem gyakran vezet az inzulin dózis csökkentéséhez Retrospectív vizsgálat 1-es és 2-es típusú cukorbetegekkel Many patients are willing to accept suboptimal glycaemic control and decrease their insulin dose following a hypoglycaemic event n=202 T1DM and 133 insulin-treated T2DM patients Conducted in 4 Canadian Centres (Ontario and Quebec) in 2003 Self-administered questionnaire Total patient sample, n=335 (type 1 diabetes, n=202; type 2 diabetes, n=133) Leiter et al. Can J Diabetes 2005;29:186–92 Many patients who experience hypoglycaemia alter their behaviour to avoid further episodes. Compensating behaviours such as reducing their insulin dose or missing doses to avoid hypoglycaemia are likely to worsen glycaemic control. Objective: To assess the impact of mild, moderate and severe hypoglycemia and fear of future hypoglycemic episodes on patients with type 1 or insulin-treated type 2 diabetes. A self-administered questionnaire was used to collect retrospective data on frequency of hypoglycemia and its impact on glycemic management, treatment of hypoglycemia and post hypoglycemia lifestyle infringements. Two-hundred-two type 1 and 133 insulin-treated type 2 diabetes patients enrolled. Following a mild or moderate hypoglycemic episode, more type 1 diabetes patients reported increased fear of future hypoglycemia (37.8%) than insulin treated type 2 diabetes patients (29.9%). However, subsequent to a severe hypoglycemic episode, 84.2% of type 2 vs. 63.6% of type 1 diabetes patients reported greater fear of future hypoglycemia. The most common management strategy for hypoglycemia of any severity was self-treatment. Patients with type 2 diabetes reported “sometimes” or “always” modifying their insulin dose 57.5% of the time following severe and 43% of the time following mild or moderate hypoglycemia episodes.

Kulcsfontosságú vizsgálatokban a kezelés intenzifikálás növelte a testsúlyt Cross-sectional, mean values 0.0 2.5 5.0 7.5 3 6 9 12 15 Years from randomisation Weight (kg) *Intensive therapy UKPDS1 20-year study duration ACCORD2 7-year study duration 28% of patients who gained weight, gained >10 kg VADT3 8-year study duration Mean of 8.2 kg increase *Aim for intensive group (treated with either SU or insulin): FPG<6 mmol/L. Aim for conventional group: best achievable FPG with diet alone, drugs added only if there were hyperglycaemic symptoms or FPG>15 mmol/L. 1 Adapted from UKPDS 33. Lancet 1998;13:837–53; 2 ACCORD Study Group. NEJM 2008;358(24): 2545-2559; 3 Duckworth et al. NEJM 2009;360:129-39. It is pertinent to discuss some of the barriers associated with intensive glycaemic control therapies. The UKPDS demonstrated that even though intensive therapy with either SU or insulin did yield greater improvements in HbA1c, it was also associated with increase in body weight compared to conventional therapy with diet. Background Subjects: 3,867 newly diagnosed patients with type 2 diabetes Asymptomatic after 3 months' diet (FPG) 6.0-15.0 mmol/L randomised to either intensive treatment group (insulin or sulphonylurea) or conventional diet treatment group Aggregate endpoints measured: Any diabetes-related endpoint (sudden death, death from hyperglycaemia or hypoglycaemia, fatal or non-fatal myocardial infarction, angina, heart failure, stroke, renal failure, amputation [of at least one digit], vitreous haemorrhage, retinopathy requiring photocoagulation, blindness in one eye, or cataract extraction) Diabetes-related death (death from myocardial infarction, stroke, peripheral vascular disease, renal disease, hyperglycaemia or hypoglycaemia, and sudden death) All-cause mortality

A 2-es típusú cukorbetegség ideális kezelése 2017.04.08. A 2-es típusú cukorbetegség ideális kezelése Hatékony  tartósan biztosítja a célértékeket Biztonságos  nem okoz hypoglykaemiát (vércukorszint-függő hatás) A kardiovaszkuláris kockázati tényezőket nem változtatja vagy javítja  hypertonia, testsúly, lipidek… Somogyi A és mtsai. Orvosképzés 2010;85:145–154 A 2-es típusú cukorbetegség ideális kezeléséhez nem csak az anyagcsere kontrollra van szükség, hanem a CV rizikó tényezők csökkentésére is (pl. testsúly, hipoglikémia stb.) Referencia: Somogyi A és mtsai. Orvosképzés 2010;85:145–154. 26 26

A racionális gyógyszer kombináció Kombinációs kiszerelés kivitelezhető a különböző, stabil asszociációs formák miatt Liraglutid IDegLira Inzulin degludek Az egyedi, liraglutid és inzulin degludek kombináció várható előnyei: Glykaemiás kontroll elérése a teljes napra, mind az éhomi időszakban, mind az étkezések után Titrálás, kedvező biztonságossági profil Napi 1x-i, adagolás egy penben

IDegLira: Combination in a single daily injection Max dose Subcutaneous injection 3 mL pre-filled pen Fixed ratio of IDeg (100 U/mL) and liraglutide (3.6 mg/mL) 1.8 mg Liraglutide Titrate Insulin titration to achieve glycaemic control Insulin degludec 50 U 50 U IDeg + 1.8 mg liraglutide 50 dose steps 10 U IDeg + 0.36 mg liraglutide 10 dose steps 1 U IDeg + 0.036 mg liraglutide 1 dose step 20 U IDeg + 0.72 mg liraglutide 20 dose steps

DUAL™ I Study design Patients with type 2 diabetes (n=1663) Mean FPGa Main phase – 26 weeks Extension phase – 26 weeks IDegLira + met ± pio (n=834) IDegLira + met ± pio (n=665) Titrate to target Starting dose: 10 dose steps/units Patients with type 2 diabetes (n=1663) IDeg + met ± pio (n=414) IDeg + met ± pio (n=333) Liraglutide 1.8 mg + met ± pio (n=415) Liraglutide 1.8 mg + met ± pio (n=313) Inclusion criteria Type 2 diabetes Insulin-naïve treated with metformin ± pioglitazone HbA1c 7.0–10.0% Stratification: HbA1c ≤8.3; HbA1c >8.3 BMI ≤40 kg/m2 Age ≥18 years* 26 weeks 52 weeks Randomised 2:1:1 Open label Mean FPGa Dose change mmol/Lb mg/dLb dose steps/U <4.0 <72 −2 4.0−5.0 72−90 >5.0 >90 +2 *Singapore, age ≥21 years; FPG, fasting plasma glucose; IDeg, insulin degludec; IDegLira, insulin degludec/liraglutide combination; met, metformin; pio, pioglitazone; NN9068-3697; IDegLira vs IDeg vs liraglutide in type 2 diabetes extension

DUAL™ I extension: Mean daily doses IDeg component 62 U 39 U 53 U 38 U Daily dose (U) IDegLira (n=825) IDeg (n=412) Week 26 Week 52 Time (weeks) Liraglutide component 1.8 mg 1.4 mg 1.8 mg 1.4 mg Daily dose (mg) Week 26 Week 52 IDegLira (n=825) Liraglutide (n=412) Time (weeks) Mean values with error bars (standard error mean) based on safety population and last observation carried forward (LOCF) imputed data Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I: Jelentős HbA1c javulás IDegLira (n=833) IDeg (n=413) Liraglutid (n=414) 26 hét 52 hét −1.28% ∆HbA1c EOT 7.0% −1.44% 6.9% *p<0.0001 vs. IDeg és vs. liraglutide −1.91%* 6.4% 7.1% 6.9% 6.4% ∆HbA1c EOT −1.21% −1.40% −1.84%* *p<0.0001 vs. IDeg és vs. liraglutide HbA1c (%) 0.0 Idő (hét) Mean values with error bars (standard error mean) based on FAS and LOCF imputed data p values are from an ANCOVA model; EOT, end of treatment Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I: HbA1c célértéket elérő betegek aránya IDegLira (n=833) IDeg (n=413) Liraglutid (n=414) 26 hét 52 hét 100 80 60 40 20 p<0.0001 Betegek aránya (%) HbA1c <7.0% HbA1c ≤6.5% 100 80 60 40 20 p<0.0001 Betegek aránya (%) HbA1c <7.0% HbA1c ≤6.5% ADA/EASD HbA1c target <7.0%; AACE HbA1c target ≤6.5% Values based on FAS and LOCF-imputed data; p-values are from a logistic regression model Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I extension: FPG over time IDegLira (n=833) IDeg (n=413) Liraglutide (n=414) Week 26 Week 52 −1.75 mmol/L ∆FPG EOT 7.3 mmol/L −3.61 mmol/L 5.8 mmol/L −3.62 mmol/L* 5.6 mmol/L −1.67 mmol/L ∆FPG EOT 7.3 mmol/L −3.40 mmol/L 6.0 mmol/L −3.45 mmol/L* 5.7 mmol/L FPG (mmol/L) *p<0.0001 vs. liraglutide Time (weeks) Mean values with error bars (standard error mean) based on FAS and LOCF imputed data p values are from an ANCOVA model Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I: 9 pontos vércukor profil 26 hét Liraglutid IDeg PG (mmol/L) IDegLira 52 hét Liraglutid IDeg IDegLira Breakfast +90 min Lunch Evening meal Bedtime 04:00 Breakfast next day p=0.0042 p=0.0104 p=0.0036 p=ns p<0.0001 Liraglutid 52 hét IDeg PG kiugráso (mmol/L) IDegLira Breakfast† Lunch† Evening meal† PG increment = change in pre-meal level to 90 min post-meal. Mean values based on FAS and LOCF-imputed data †p-values from separate ANCOVAs; 9P-SMBG, 9-point self-measured blood glucose; PG, plasma glucose Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I: Méréssel megerősített hypoglykaemia IDegLira (n=825) IDeg (n=412) Liraglutid (n=412) 6.9% 6.4% HbA1c 7.1% Rate ratio: 8.52 p<0.0001 Rate ratio: 0.63 Rate ratio: 7.61 p<0.0001 Rate ratio: 0.68 p=0.002 6.9% 6.4% HbA1c 7.0% Esemény/beteg Idő (hét) Cumulative hypoglycaemia data are from safety analysis set (SAS). Rate ratios and p values are from a negative binomial model using FAS. HbA1c data are mean from FAS with LOCF Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

Testsúly változás (kg) DUAL™ I: Testsúly IDegLira (n=833) IDeg (n=413) Liraglutid (n=414) 26 hét 52 hét −2.80 kg p<0.0001 2.66 kg p<0.0001 −2.22 kg p<0.0001 2.44 kg p<0.0001 Testsúly változás (kg) Idő (hét) Mean values with error bars (standard error mean) based on FAS and LOCF imputed data Estimated treatment differences and p values are from an ANCOVA model Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I extension: Percentage of subjects with nausea IDegLira (n=825) IDeg (n=412) Liraglutide (n=412) Subjects (%) Time (weeks) Data are from SAS Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I extension: Summary of adverse events (AEs) IDegLira IDeg Liraglutide Week 26 Week 52 Number of exposed subjects 825 412 Percentage of subjects with AEs† 63.2% 71.2% 60.2% 70.6% 72.6% 77.2% AE rate per patient-year of exposure 4.8 4.1 4.3 3.8 6.4 5.1 Percentage of subjects with serious AEs† 2.3% 4.6% 1.9% 5.3% 3.4% 5.8% Serious AE rate per patient-year of exposure 0.05 0.07 0.09 Four major adverse cardiovascular events (MACE) were reported in the IDegLira group and one in each of the IDeg and liraglutide groups* No medullary thyroid carcinomas were reported, and there were no confirmed thyroid neoplasms* Two events of acute pancreatitis (one also a pancreas cancer) in the liraglutide group* †Percentage of subjects with ≥1 event; *as assessed by external adjudication committee Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

Summary and conclusions of 52-week data Compared to insulin degludec, IDegLira was associated with: Significantly greater reduction in HbA1c Significantly lower risk of hypoglycaemia No weight gain Compared to liraglutide, IDegLira was associated with: Significantly greater reduction in FPG Fewer GI adverse events 52-week data provide further support for the sustainability of the glucose-lowering effect and longer-term safety of IDegLira Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

IDegLira: The combination of insulin degludec and liraglutide in a once-daily single injection independent of meals HbA1c (%) FPG (mmol/L) Weight (kg) Confirmed hypoglycaemia +2.66* +752%* + vs. Lira vs. Lira vs. IDeg DUALTM I extension -1.21 vs. IDeg -1.67 -1.40 -2.80* - -1.84* -3.45† -3.40 -37%* IDegLira IDeg Liraglutide *p<0.0001 vs. IDeg and liraglutide; †p<0.0001 vs. liraglutide Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

IDegLira: The combination of insulin degludec and liraglutide in a once-daily single injection independent of meals + HbA1c (%) FPG (mmol/L) Weight (kg) - -1.84* -1.40 -1.21 -3.45† -3.40 -1.67 Confirmed hypoglycaemia -37%* DUALTM I extension +2.66* +661%† vs. Lira vs. Lira vs. IDeg vs. IDeg -2.80* + - -1.90§ -0.89 -3.46‡ -2.58 -2.7 n.s DUALTM II 0.0 IDegLira IDeg Liraglutide *p<0.0001 vs. IDeg and liraglutide; †p<0.0001 vs. liraglutide; ‡p=0.002; §p<0.0001 vs. IDeg; n.s, non-significant Buse et al. ADA 2013: 65-OR; Gough et al. EASD 2013: 219-OR; Buse J et al. IDF 2013: OP-0082 Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

Összefoglalás I. DUAL™ I vizsgálat Insulin degludekhez képest, az IDegLira mellett: Szignifikánsan jobb HbA1c csökkenés Szignifikánsan alacsonyabb hypoglykaemia kockázat Nincs testsúly növekedés Liraglutidhoz képest, az IDegLira mellett: Szignifikánsan jobb éhomi vércukor csökkenés Ritkább GI mellékhatások IDegLira kombinálva az inzulin degludek és a liraglutid hatását egy injekcióban, összességében jelentősebb glykaemiás kontroll javulást eredményez alacsony hypoglykaemia kockázat és testsúlynövekedés nélkül Buse et al. ADA 2013: 65-OR; Gough et al. EASD 2013: 219-OR

2017.04.08. Összefoglalás II: 43 Mikor és hogyan illeszük az új terápiás eljárásokat a hazai cukorbeteg ellátásba? Korai felismerés, betegoktatás, orvosoktatás, korai terápiás intenzifikálás megfelelő készítményekkel, egyénre szabott kezeléssel, holisztikus megközelítéssel. Folyamatos minőségi kontroll elengedhetlen Rendszeres frissítése a finanszírozási prokotokolloknak Összhangban legyen a hazai szakmai ajánlásokkal Nem a hyperglikémiát „gyógyítjuk”, hanem a cukorbeteget és annak minden betegségét 43

Conclusions DUAL™ II Superior glycaemic control of IDegLira vs. IDeg at an equivalent insulin dose Significant contribution of the liraglutide component to the overall effect of IDegLira Adverse event and tolerability profile of IDeg were consistent with previous findings No apparent or unexpected adverse event or tolerability issues for IDegLira were reported Buse J et al. IDF 2013: OP-0082

Presentation title Date Backups

Many patients are not at glycaemic goal despite insulin therapy THIN database Mean HbA1c level: 8.3% at median follow-up of 117 days Retrospective cohort study with UK primary care patients with type 2 diabetes n=4045; mean age 62.6 years Mean baseline HbA1c 9.6% 52.4% initiated basal, 41.5% premixed, 4.0% basal-bolus and 2.1% prandial insulin A retrospective cohort study of patients with type 2 diabetes selected from the THIN database, which collects data from general practices in the UK, evaluated outcomes following insulin initiation. Methods A retrospective cohort study was performed using quality-checked patient data from The Health Improvement Network database. Eligible patients who initiated insulin for the first time between 2004 and 2006 were grouped into four cohorts according to the type of insulin regimen initiated. Data on patient characteristics, metabolic and clinical outcomes and health-care resource use were collected at baseline and during 6 months of follow-up. Results In total, 4045 eligible adults [2269 male, 1776 female; mean age 62.6 13.3 years; mean baseline HbA1c (9.6%±2.0%)] initiated insulin. Approximately half (52.4%) initiated insulin as basal insulin only, 41.6% as premixed only, 4.0% as basal-bolus and 2.1% as prandial insulin only. Among patients with ≥180 days follow-up (n = 3815), the initial insulin regimen was not changed during follow-up in 75.1% of patients, while 13.7% discontinued, 7.0% switched and 4.7% intensified insulin therapy. The mean change in HbA1c was -1.3%, n = 2881, with 17.3% of patients achieving an HbA1c of 7% (n = 3024). The mean weight change was +0.9 kg (n = 2345). At a median of 117 days, the mean HbA1c was 8.3%±1.6%. The majority of patients did not achieve their HbA1c target: only 17.3% of patients achieved an HbA1c level of <7.0% and 30.0% achieved an HbA1c level of <7.5%. Conclusions Basal and premixed insulin were the most common types of insulin initiated and in most patients no changes were made to the initial regimen over 6 months. However, few patients achieved glycemic control targets. Adapted from Blak et al. Diabetic Medicine 2012;29(8):e191-198

Insulin receptor activation Complementary actions of GLP-1 and insulin target underlying pathophysiology of type 2 diabetes Brain Energy intake Satiety Neuroprotection GLP-1 analogue Basal insulin Heart Cardiac function Skeletal muscle Glucose disposal Pancreas Glucose-dependent insulin and glucagon secretion Insulin synthesis Liver Hepatic glucose production Liver Hepatic glucose output Adipose tissue Insulin receptor activation GI tract Gastric emptying GLP-1, glucagon-like peptide-1 Baggio, Drucker. Gastroenterol 2007;132:2131–57

DUAL™ I extension: Baseline characteristics IDegLira IDeg Liraglutide Full analysis set (FAS), n 833 413 414 Male/female, % 52.2/47.8 48.4/51.6 50.2/49.8 Age, years 55.1 (9.9) 54.9 (9.7) 55.0 (10.2) Weight, kg 87.2 (19.0) 87.4 (19.2) 87.4 (18.0) BMI, kg/m2 31.2 (5.2) 31.2 (5.3) 31.3 (4.8) Duration of diabetes, years 6.63 (5.13) 6.99 (5.31) 7.15 (6.09) HbA1c , % 8.3 (0.9) 8.3 (1.0) FPG, mmol/L FPG, mg/dL 9.2 (2.4) 165.8 (43.2) 9.4 (2.7) 169.4 (48.6) 9.0 (2.6) 162.2 (46.8) OAD at screening (% of patients): Met Met + pio 83.0 17.0 83.1 16.9 81.6 18.1 Values are mean unless otherwise stated; values in brackets indicate standard deviation; aCalculated, not measured; OAD, oral antidiabetic drug Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ I extension: Subject disposition 834 IDegLira randomised main phase 414 IDeg randomised main phase 415 liraglutide randomised main phase 665 included in extension 333 included in extension 313 included in extension 621 (74.5%) Completers 305 (73.7%) Withdrawals: 44 (5.3%) AE: 5 (0.6%) Ineffective therapy: 0 (0%) Non-compliance: 2 (0.2%) Withdrawal criteria: 19 (2.3%) Other: 18 (2.2%) Withdrawals: 28 (6.8%) AE: 1 (0.2%) Non-compliance: 0 (0%) Withdrawal criteria: 14 (3.4%) Other: 13 (3.1%) Withdrawals: 28 (6.7%) AE: 2 (0.5%) Non-compliance: 1 (0.2%) Withdrawal criteria: 16 (3.9%) Other: 9 (2.2%) 2:1:1 randomisation 285 (68.7%) 342 liraglutide main phase completers 734 IDegLira main phase completers 366 IDeg main phase completers AE, adverse event; Withdrawals based on subjects included in extension Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

Combining therapies offers opportunities to enhance efficacy and diminish side effects GLP-1 RA monotherapy + Basal insulin GLP-1 RA/insulin combined - HbA1c FPG PPG Weight Hypoglycaemia Nausea For illustrative purposes only, not meant to quantify or imply magnitude of change in either direction

Healthy Eating, Weight Control, Increased Physical Activity ADA/EASD Position Statement Combination of insulin and GLP-1 RA therapy is supported in current diabetes treatment algorithms Healthy Eating, Weight Control, Increased Physical Activity Initial drug monotherapy Metformin Two drug combinations Sulfonylurea Thiazolidine-dione DPP-4 inhibitor GLP-1 receptor agonist Insulin (usually basal) Sulfonylurea + TZD DPP-4-i GLP-1-RA Insulin or Thiazolidine-dione + SU DPP-4-i GLP-1-RA Insulin or DPP-4 Inhibitor + SU TZD Insulin or GLP-1 receptor agonist + SU TZD Insulin or Insulin (usually basal) + TZD DPP-4-i GLP-1-RA or Three drug combinations More complex insulin strategies Insulin (multiple daily doses) Key points from ADA/EASD position statement on management of T2D: Glycemic targets & BG-lowering therapies must be individualized Diet, exercise, & education: foundation of any T2DM therapy program Unless contraindicated, metformin = optimal 1st-line drug After metformin, data are limited; combination therapy with 1-2 other oral / injectable agents; minimize side effects Ultimately, many patients will require insulin therapy alone or in combination with other agents to maintain BG control All treatment decisions should be made with the patient (focus on preferences, needs & values) Comprehensive CV risk reduction - a major focus of therapy Adapted from: Inzucchi et al. Diabetes Care 2012;35:1364-79

Liraglutide, a once daily human GLP-1 analogue Shares a 97% amino acid identity with human GLP-11,2 Less susceptible to degradation by DPP-4 due to modifications of the native-GLP-1 molecule T1/2 = 13 hrs Phase 3 clinical development programme: LEAD Reduction in HbA1c Rapid and sustained reduction in FPG Decrease in postprandial glucose Low rates of hypoglycaemia Significant reduction in body weight Liraglutide is a once-daily human GLP-1 analogue In creating liraglutide, two modifications to the amino acid sequence of GLP-1 are made: a fatty acid is acylated to lysine at position 26 and the lysine at position 34 is replaced with arginine. These modifications result in increased self-association (which slows absorption from the subcutaneous depot), albumin binding and reduced susceptibility to DPP-IV, which combine to prolong its plasma half-life, protracting its action. Thus, the problem of the short half-life, which is the major clinical drawback of native GLP-1, is overcome. For comparison, the plasma half-life of exenatide is 4 to 6 hours (reviewed in Nauck and Drucker, 2006) The LEAD programme consists of LEAD trials 1 to 5 as part of the Phase 3a and LEAD-6 as part of the Phase 3b clinical development programme. A total of 4,456 subjects were included, recruited at more than 600 sites in 40 countries, of which 2,739 patients were treated with liraglutide in these trials. Reference Knudsen et al. J Med Chem 2000;43:1664–69 T1/2, half life 1. Knudsen et al. J Med Chem 2000;43:1664–1669; 2. Degn et al. Diabetes 2004;53:1187–1194

Patients with type 2 diabetes (n=987) Adding detemir to liraglutide Insulin-naïve patients with type 2 diabetes Study design HbA1c <7.0% Liraglutide 1.8 mg + met* (n=498) Observational group Liraglutide 1.8 mg Liraglutide 1.2 mg 1 week IDet + Liraglutide 1.8 mg + met* (n=162) Randomised treatment group Liraglutide 0.6 mg 1 week HbA1c ≥7.0% Patients with type 2 diabetes (n=987) Liraglutide 1.8 mg + met* (n=161) Randomised control group 12 week run-in 26 weeks Inclusion criteria Type 2 diabetes Treated with metformin (≥1500 mg) or MET (≥1500 mg) + SU (≤50% max. dose) for ≥3 months HbA1c 7.5–10.0% (MET only) or 7.0-8.5% (MET + SU) Age 18-80 years Observational group: ∆HbA1c from BL = -1.12% DeVries evaluated the novel treatment intensification sequence: addition of liraglutide to metformin in type 2 diabetes followed by intensification with basal insulin (detemir) if HbA1c ≥ 7%. Participants were insulin-naïve adults with type 2 diabetes treated for at least 3 months or more, with ≥ 1500 mg/day metformin and HbA1c values of 7.0-10.0% or with metformin and sulfonylurea (less than or equal to half of the maximum approved dose) and HbA1c values of 7.0–8.5%. The run-in period consist of 988 participants from North America and Europe uncontrolled on metformin 6 sulfonylurea, sulfonylurea was discontinued and liraglutide 1.8 mg/day added for 12 weeks. Subsequently, those with HbA1c ≥ than 7% were randomized 1:1 to 26 weeks’ open-label addition of insulin detemir to metformin + liraglutide (n = 162) or continuation without insulin detemir (n = 161). Patients achieving HbA1c < 7% continued unchanged treatment (observational arm). The primary end point was HbA1c change between randomized groups. Met, metformin ≥1500 mg/day DeVries et al. Diabetes Care 2012; 35:1446–1454

Adding detemir to liraglutide: HbA1c Run-in (Weeks −12 to 0) Randomised period (Weeks 0 to 26) Change in HbA1c (%) Baseline: RC: 8.29 RT: 8.22 (final value 7.5) RC: 0.76 (final value 7.1) RT: 1.13 Randomised control group (RC; MET + liraglutide 1.8 mg) Randomised treatment group (RT; MET + IDet + liraglutide 1.8 mg) HbA1c and FPG data are from FAS; weight change and confirmed hypoglycaemia are from SAS Confirmed hypoglycaemia - Data are from the safety analysis set and exclude one outlier (who experienced 25 minor hypoglycaemic episodes). DeVries et al. Diabetes Care 2012; 35:1446–1454

Adding detemir to liraglutide: Confirmed hypoglycaemia Run-in period (Weeks 12 to 0) Randomised period (Weeks 0 to 26) *p=0.004 Randomised control group (RC; MET + liraglutide 1.8 mg) Randomised treatment group (RT; MET + IDet + liraglutide 1.8 mg) HbA1c and FPG data are from FAS; weight change and confirmed hypoglycaemia are from SAS Confirmed hypoglycaemia - Data are from the safety analysis set and exclude one outlier (who experienced 25 minor hypoglycaemic episodes). DeVries et al. Diabetes Care 2012; 35:1446–1454

Obese, type 2 diabetes (n=30) Adding liraglutide to high dose intensive insulin therapy Small, single site sample of type 2 diabetes patients requiring >100 U insulin/day Liraglutide + insulin Obese, type 2 diabetes (n=30) Standard insulin up-titration 24 weeks Randomised 1:1 Inclusion criteria HbA1c > 6.5% Intensive insulin therapy of >100 units of insulin/day, either by basal-bolus or CSII ± Met Liraglutide initiated at 0.6 mg daily and increased to 1.2 or 1.8 mg Rappaport et al. EASD 2013; Abstract #900

Total daily insulin dose Adding liraglutide to high dose intensive insulin therapy Small, single site sample of type 2 diabetes patients requiring >100 U insulin/day Weight Total daily insulin dose Liraglutide + insulin Insulin only Baseline 6 months Liraglutide + insulin Insulin only * p<.0001 † p=.435 * p<.0001 † p=.320 Liraglutide + insulin resulted in similar improvements in HbA1c compared to insulin up-titration alone In obese subjects with type 2 diabetes on intensive insulin therapy with >100 units of insulin per day, the addition of Liraglutide to insulin resulted in the benefits of  weight loss, reduction in TDID, increased percent of time in the euglycemic range (data not shown) and reduced glycemic variability by CGM (data not shown). The addition of Liraglutide to insulin resulted in similar improvements in HbA1c compared to insulin up-titration alone. Rappaport et al. EASD 2013; Abstract #900 HbA1c (%) Baseline 6 months Liraglutide + insulin 7.76% 7.14% Insulin only 7.88% 7.38%

Insulin degludec: from injection to subcutaneous depot Insulin degludec injected Phenol from the vehicle diffuses quickly, and hexamers link up via single side-chain contacts Long multi-hexamers assemble Phenol Zn2+ Jonassen et al. Pharm Res 2012;29:2104–14

Monomers are absorbed from the depot into the circulation Insulin degludec: slow release from subcutaneous depot Zn2+ Subcutaneous depot Insulin degludec multi-hexamers Zinc diffuses slowly causing individual hexamers to disassemble, releasing monomers Monomers are absorbed from the depot into the circulation Jonassen et al. Pharm Res 2012;29:2104–14

Half-life of insulin degludec is twice as long as that of insulin glargine IDeg 0.8 U/kg IGlar 0.8 U/kg * Insulin degludec Insulin glargine 0.4 U/kg 0.6 U/kg 0.8 U/kg Half-life (hours) 25.9 27.0 23.6 11.5 12.9 11.9 Mean half-life 25.4 12.5 *Insulin glargine was undectable after 48 hours Results from 66 patients with type 1 diabetes (T1D) IDeg, insulin degludec; IGlar, insulin glargine Heise et al. Diabetologia 2011;54(Suppl. 1):S425 Insulin degludec was detectable in the serum for at least 120 hours following the final dose, whereas, for most subjects, insulin glargine fell below the lower limit of quantification after 36–48 hours post-dosing. Lower limit of quantification (LLoQ) was 20 pmol/L for both insulin assays, which was ~0.3% of maximum degludec concentration and ~8% of maximum for glargine. N=66 patients with type 1 diabetes randomised in this double-blind, two-period, crossover study in which they received either 0.4, 0.6 or 0.8 U/kg of degludec and glargine once daily for 8 days. Subjects underwent a 42 hour euglycaemic clamp (5.5 mmol/l) after the last dose at the end of each treatment period (day 8). There was a 7-21 day washout period between the two periods. Blood samples were drawn for assessment of steady-state pharmacokinetics for 120 hours after the last dose . 60

Flat time-action profile in type 2 diabetes at steady state 0.8 U/kg 0.6 U/kg 0.4 U/kg Type 2 diabetes, 49 patients, randomised, 2-period, 12-day trial Variability was assessed at steady state by clamps on days 6 and 12; GIR, glucose infusion rate Heise et al. Diabetes Obes Metab 2012; 2012;14:944-50 The pharmacodynamics (PD) variability of IDeg was determined under steady-state conditions by Heise and colleagues.

IDegLira: Clinical development programme Phase 3a trials DUAL™ I Combination compared to the mono-components added on to OAD DUAL™ II Combination compared to IDeg in patients previously treated with basal insulin Phase 3b trials DUAL™ III Switch from (daily) GLP-1 receptor agonist therapy vs placebo DUAL™ IV IDegLira add-on to SU vs placebo DUAL™ V Basal insulin optimisation vs IGlar DUAL I - Combination product compared to individual components as required by EU guidelines DUAL II - add-on to OAD FDA requirement to demonstrate combination product HbA1c superiority to insulin degludec

Hypoglycaemia classification Suspected hypoglycaemia or routine PG measurement Patient able to treat self? Yes No PG <3.1 mmol/La (56 mg/dL) No Yes Not classified as confirmed hypoglycaemia in this trial Confirmed hypoglycaemia Minor hypoglycaemia Severe hypoglycaemia a:With or without symptoms Nocturnal confirmed = Between 00:01 and 05:59am (both inclusive) Gough et al. ADA 2014: 65-OR; Gough et al. EASD 2014: 78-OR

DUAL™ II: Double-blind trial in type 2 diabetes patients uncontrolled on basal insulin Titrate to target FPG 4–5 mmol/L Starting dose: 16 dose steps/units Maximum dose 50 dose steps / 50 units IDegLira + met (n=199) Patients with type 2 diabetes (N=398) IDeg + met (n=199) Randomised 1:1 Double-blind 26 weeks Inclusion criteria Type 2 diabetes HbA1c 7.5–10.0% BMI ≥27 kg/m2 Age ≥18 years Basal insulin (20-40U) + metformin +/- SU or glinides Titration algorithm: IDegLira and IDeg Mean fasting PG Dose change mmol/L dose steps or U <4.0 -2 4.0–5.0 >5.0 +2 Male and female subjects aged 18 years or older with type 2 diabetes, treated with basal insulin in combination with 1–2 OADs (metformin, or metformin + SU, or metformin + glinides) were randomly allocated in a 1:1 manner to one of the two treatment groups (insulin degludec/liraglutide or insulin degludec). The subjects must have been diagnosed with type 2 diabetes and been on stable treatment for at least 90 days prior to screening. N, number of randomised subjects, excluding subjects from Site 105 (8 subjects for IDegLira and 7 subjects for IDeg); Fasting PG, self-measured using a glucometer which was calibrated to convert blood glucose measurements to plasma glucose values Buse J et al. IDF 2013: OP-0082

DUAL™ II: Baseline characteristics IDegLira IDeg Full analysis set (FAS), n 199 Male, Female, % 56.3/43.7 53.3/46.7 Age, years 56.8 (8.9) 57.5 (10.5) Weight, kg 95.4 (19.4) 93.5 (20.0) BMI, kg/m2 33.6 (5.7) 33.8 (5.6) Duration of diabetes, years 10.30 (6.01) 10.91 (7.04) HbA1c , % 8.7 (0.7) 8.8 (0.7) FPG, mmol/L 9.7 (2.9) 9.6 (3.1) Basal insulin at screening (% of patients): Basal insulin + met Basal insulin + met + SU or glinides 47.7 52.3 49.2 50.8 The treatment groups were overall well matched with respect to baseline demographics and characteristics. Approximately 75% of the subjects were in the 40–65 years age group and 21.8% were older than 65 years of age. Mean HbA1c was 8.7% in the IDegLira group and 8.8% in the IDeg group. Duration of diabetes was approximately 10 years in the IDegLira group and approximately 11 years in the IDeg group. No difference in family history of diabetes was observed between treatment groups. Values are mean unless otherwise stated. Values in brackets indicate standard deviation Buse J et al. IDF 2013: OP-0082

DUAL™ II: Mean daily doses IDeg (n=199) IDegLira (n=199) IDeg: 45U IDegLira: 45U ~65% at max dose of IDegLira, of which ~60% were at target (<7.0%) Daily dose (U) ~67% at max dose of IDeg, of which ~18% were at target (<7.0%) Time (weeks) After 26 weeks of treatment: the actual median dose was 50 units for IDegLira (dose range of 12−50 units) and IDeg (dose range of 12−50 units) the actual mean dose of insulin was 45 units for both IDegLira and IDeg The majority of subjects in both treatment groups (65.3% with IDegLira and 67.3% with IDeg) reached the maximum allowed dose of 50 dose steps/50 units during the trial. There were no clinically relevant differences between the prescribed and actual or recommended insulin doses with IDegLira or IDeg. Total ≥50U: 134 persons Out of 199 total: 134/199 = 67.3% on max dose, 24/134= 17.9% responders (<7%) Mean values with error bars (standard error mean) based on SAS and LOCF imputed data Estimated treatment differences are from an ANCOVA analysis IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: HbA1c over time IDeg (n=199) IDegLira (n=199) Time (weeks) HbA1c (%) 0.0 ∆HbA1C EOT 8.0% -0.89% -1.90% 6.9% p<0.0001 The largest decrease in HbA1c was observed with IDegLira. The mean HbA1c at baseline was 8.7% in the IDegLira treatment group and 8.8% in the IDeg treatment group. After 26 weeks of treatment, the mean HbA1c decreased by 1.9%-point to 6.9% on IDegLira and by 0.89%-point to 8.0% on IDeg. Mean values with error bars (standard error mean) based on FAS and LOCF imputed data; p-values are from an ANCOVA EOT, end of trial; IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: Percentage of patients to target IDegLira (n=199) IDeg (n=199) p<0.0001 p<0.0001 % of patients reaching glycaemic targets HbA1c <7.0% HbA1c ≤6.5% Asterisks denote statistically significant differences. Values based on FAS and LOCF-imputed data; p-values are from a logistic regression model IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: FPG over time IDeg (n=199) IDegLira (n=199) ∆FPG EOT 7.0 mmol/L ∆FPG EOT -2.58 FPG (mmol/L) 6.2 mmol/L p=0.0019 -3.46 0.0 Time (weeks) Baseline FPG was similar between treatments. For both treatment groups, the greatest change in FPG was observed during the first 8 weeks of treatment, with the most pronounced decrease observed with IDegLira. FPG continued to decrease for subjects on IDegLira or IDeg until a plateau was reached after 16 weeks and 12 weeks of treatment, respectively. From baseline to Week 26, FPG: decreased by 3.46 mmol/L to 6.2 mmol/L with IDegLira decreased by 2.58 mmol/L to 7.0 mmol/L with IDeg The estimated treatment difference for IDegLira vs. IDeg was –0.73 mmol/L, p=0.0019, which demonstrated a statistically significantly greater reduction in FPG for IDegLira compared to IDeg after 26 weeks of treatment. Mean values with error bars (standard error mean) based on FAS and LOCF imputed data; p-values are from an ANCOVA EOT, end of trial; IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: Self-measured blood glucose Week 0 IDeg Breakfast +90 min Evening meal Breakfast next day Lunch Bedtime 04:00 Glucose (mmol/L) 0.0 IDegLira Week 26 IDeg IDegLira * At baseline, the SMBG 9-point profiles appeared similar for both treatments groups. After 26 weeks of treatment, concentrations had decreased for both treatments. The profile for IDegLira showed both lower pre-prandial (before meals) and post-prandial (90 min after meal consumption) glucose concentrations than with IDeg. Dotted line represents the glycaemic range for IDegLira Mean values based on FAS and LOCF-imputed data; p-values from an ANCOVA IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: Confirmed hypoglycaemia IDeg (n=199) IDegLira (n=199) Rate ratio: 0.66 p=0.13 Cumulative episodes per subject The definition for hypoglycaemia was the same in DUAL™ I and DUAL™ II Time (weeks) The observed rate of confirmed hypoglycaemic episodes over time was similar between IDegLira and IDeg during the first part of the trial (from Week 1 to Week 6), followed by a lower rate of hypoglycaemic episodes with IDegLira than with IDeg during the latter part of the trial (from Week 6 to Week 26). The estimated treatment ratio for IDegLira vs. IDeg was 0.66, p=0.13. Mean values based on SAS Estimated rate ratios and p-values are from a negative binomial model IDeg maximum dose was 50 units per day; Buse J et al. IDF 2013: OP-0082

DUAL™ II: Confirmed hypoglycaemia IDegLira (n=199) % patients (# patients) Rate episodes/ PYE Confirmed 24.1% (48/199) 1.53 IDeg (n=199) % patients (# patients) Rate episodes/ PYE 24.6% (49/199) 2.63 IDegLira vs. IDeg Rate ratio ∆Risk 0.66 34% SAS; % patients, proportion of patients with events; # patients, number of patients with events; PYE, patient-years of exposure The percentage of subjects with hypoglycaemia was similar between the treatment groups. The observed rate of confirmed hypoglycaemic episodes appeared lower for subjects treated with IDegLira compared to subjects treated with IDeg. The difference in the estimated rates was driven by few subjects in the IDeg group, who had a large number of episodes and the difference was not statistically significant. The estimated treatment ratio for IDegLira vs. IDeg was 0.66, p=0.13. This represented a 34% reduction in the risk of hypoglycaemia with IDegLira vs IDeg. Buse J et al. IDF 2013: OP-0082

DUAL™ II: Nocturnal confirmed hypoglycaemia IDeg (n=199) IDegLira (n=199) ns Cumulative episodes per subject Time (weeks) IDeg IDegLira HbA1c (%) 8.0 6.9 HbA1c responders <7.0% (%) 23.1 60.3 Nocturnal confirmed hypoglycaemic episodes over time were similar for IDegLira (21.8 episodes per 100 PYE) and IDeg (32.2 events per 100 PYE). There was no statistically significantly difference between the 2 treatment groups. Mean values based on SAS Estimated rate ratios and p-values are from a negative binomial model; Nocturnal defined at 00:01 to 05:59 (both inclusive); IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: Change in body weight over time IDeg (n=199) IDegLira (n=199) Absolute change in mean body weight 0.0 kg −2.51 kg p<0.0001 Change in body weight (kg) -2.7 kg Time (weeks) The estimated mean treatment difference between IDegLira and IDeg was –2.51 kg, p<0.0001, which demonstrates a statistically significant greater reduction in body weight on IDegLira compared to IDeg. Mean values with error bars (standard error mean) based on FAS and LOCF imputed data Estimated treatment differences an p-values are from an ANCOVA analysis IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082

DUAL™ II: Summary of adverse events (AEs) IDegLira IDeg Number of exposed subjects 199 Percentage of subjects with AEs† 57.8 61.3 AE rate per patient-year of exposure 4.0 3.6 Percentage of subjects with serious AEs† 3.5 5.5 Serious AE rate per patient-year of exposure 0.12 0.14 †Percentage of subjects with ≥1 event One major adverse cardiovascular event (MACE) was reported in the IDegLira group and 2 MACE events were reported in the IDeg group No medullary thyroid carcinomas were reported and there were no confirmed thyroid neoplasms (as assessed by external adjudication committee) One event of pancreatic carcinoma metastatic in the IDeg group (as assessed by external adjudication committee) No confirmed adjudicated pancreatitis adverse events by EAC classification IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082 3 of the confirmed cardiovascular events were major cardiovascular events (MACEs). 2 events were ‘myocardial infarction’ and 1 event was ‘stroke’. 2 events in 2 subjects in the IDeg group were confirmed to be neoplasms by EAC classification EAC classification – pancreatic carcinoma metastatic Subject no. 752012 A 62-year-old male subject treated with IDeg was hospitalised with pancreatic carcinoma metastatic (Day 163) Medical history: Uncomplicated T2DM, hypercholesterolaemia and hypertension  The subject was hospitalised due to low back pain that radiated to right lower quadrant and peri-umbilical region The subject had lost approximately 15 pounds over the last 6 months CT scan and biopsies revealed pancreatic adenocarcinoma with metastasis to the lung and liver. The patient was not a candidate for surgery Outcome: Not recovered, considered to be a chronic condition

% of patients experiencing nausea DUAL™ II: Percentage of patients experiencing nausea in this double-blinded trial IDeg (n=199) IDegLira (n=199) % of patients experiencing nausea Time (weeks) IDeg maximum dose was 50 units per day Buse J et al. IDF 2013: OP-0082 For subjects treated with IDegLira, nausea was more frequent during the first 12 weeks of the treatment period compared to last 12 weeks of the treatment period. For subjects treated with IDeg, events of nausea were evenly distributed over the trial period. Nausea was transient in nature for most subjects in both treatment groups. The median duration of nausea was 6.0 days for IDegLira and 8.5 days for IDeg. Results based on SAS