Diabetes Care Volume 42, March 2019 349 COMMENTARY

Therapeutic Inertia and the Legacy Kamlesh Khunti and Samuel Seidu of Dysglycemia on the Microvascular and Macrovascular Complications of Diabetes Care 2019;42:349–351 | https://doi.org/10.2337/dci18-0030

Type 2 diabetes is characterized by a reduced costs and increased quality of significant cost avoidance of about continuous decline in b-cell function life (11). £340 million was apparent in the first with the resultant progressive loss of Besides patient-level characteristics 5 years, rising to about £5.5 billion after glycemic control over time (1). The such as multimorbidity and poor adher- 25 years of sustained improvement in hyperglycemic burden is associated ence to treatment recommendations, control. Reductions in microvascular with longer-term microvascular and clinical inertia has been suggested as complications were the main factors macrovascular complications. Early di- one key reason for not achieving driving the cost savings, with 74% of agnosis and intensive glycemic control targets. The term “clinical inertia” in cost avoidance from prevention of renal has conclusively demonstrated a re- most instances has been used in relation disease in people with type 1 diabetes duction in microvascular complications to failure to advance therapy when ap- and 57% of cost avoidance from reduc- (2,3) and possibly macrovascular com- propriate to do so (12). However, the tions in foot ulcers, amputations, and plications over a long period of obser- definition of therapeutic inertia is now neuropathy in people with type 2 di- vation (4), the so-called legacy effect also accepted to reflect the failure to abetes (14). (5). Even though there have been im- de-intensify treatment when appropri- In the study in this issue of Diabetes provements in survival from cardiovas- ate to do so (12). A number of studies Care using routine real-world data to cular disease in the general population, have shown that therapeutic inertia is explore the concept of legacy effects people with type 2 diabetes remain associated with worse microvascular of tight glycemic control on complica- at increased risk of cardiovascular mor- and macrovascular outcomes. In a cohort tions of diabetes, Laiteerapong et al. (15) tality compared with matched popula- study of 105,477 patients, mean HbA1c focused on newly diagnosed patients tion control subjects (6). A very high was 8.1% (65 mmol/mol) at diagnosis, who had poor glycemic control and proportion of patients fail to reach the 22% remained under poor glycemic con- 10 years of survival. They demonstrated recommended glycemic targets for a trol over 2 years, and 26% never received that HbA1c levels $6.5% ($48 mmol/mol) considerable period of time after the intensive treatment. A delay in intensive for the first year after diagnosis were diagnosis of diabetes (7–9), thus lead- treatment by 1 year in conjunction with associated with worse outcomes. They ing to the complications. Guidelines for poor glycemic control significantly in- showed that there was a relative in- the treatment of patients with type 2 creased the risk of myocardial infarction, crease in mortality of 29% in patients diabetes suggest that tight glycemic heart failure, stroke, and composite car- with a dysglycemic burden of HbA1c 7% control (defined as glycated hemoglo- diovascular events (hazard ratio 1.62 to ,8% in the first year of diagnosis bin [HbA1c] ,7.0% [53 mmol/mol]) [95% CI 1.46–1.80]) (7). In another study compared with HbA1c maintained at should be maintained from diagnosis using a computer simulation model de- ,6.5%. This increase rose to 32% if through active titration of combina- signed to translate surrogate end points the dysglycemic legacy was HbA1c tions of antihyperglycemic medica- into long-term health and economic out- $9% in the first year of diagnosis. The tions and lifestyle modification, as comes (the IMS CORE Diabetes Model achievement of glycemic control after appropriate (10). Modeling studies [13]) in a representative cohort of adults diabetes diagnosis is therefore sufficient suggest that when this is done through with type 1 or type 2 diabetes from the to establish remedial long-term risk of an individualized approach, it leads to U.K. primary care patient database, complications (15).

Diabetes Research Centre, University of Leicester, Leicester General Hospital, Leicester, U.K. Corresponding author: Kamlesh Khunti, [email protected] © 2019 by the American Diabetes Association. Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license. See accompanying article, p. 416. 350 Commentary Diabetes Care Volume 42, March 2019

Laiteerapong et al. (15) assessed the instance, the analysis did not differenti- the study population was ethnically di- associations between HbA1c ,6.5% ate between patients with or without verse, the analysis could have been (,48 mmol/mol), 6.5% to ,7.0% (48 to prior cardiovascular disease. As the strengthened even more by examination ,53 mmol/mol), 7.0% to ,8.0% (53 to UKPDS recruited newly diagnosed pa- of the outcomes by different ethnicities ,64 mmol/mol), 8.0% to ,9.0% (64 tients, most of whom had no prevalent and levels of deprivation. In an observa- to ,75 mmol/mol), or $9.0% ($75 cardiovascular disease, it would have tional study of this nature, the influence mmol/mol) for different periods of early been more informative to explore pa- of many more confounding factors, exposure (0–1, 0–2, 0–3, 0–4, 0–5, 0–6, tients with and without prevalent car- known and unknown, cannot be com- and 0–7 years) to demonstrate the re- diovascular disease in the current study. pletely excluded. medial long-term legacy benefits. HbA1c In a recent analysis of 15 years of There is currently a lot of interest in real- levels of $6.5% in the first year of the follow-up results of the Veterans Affairs world evidence. The findings in this study early exposure period carried a higher Diabetes Trial (VADT), patients with (15) support evidence from randomized risk for microvascular and macrovascular established cardiovascular disease re- controlled trials for the need to intensify complications than HbA1c ,6.5% in the cruited in that study did not demonstrate blood glucose–lowering therapy in a first year of glycemic exposure. For ex- the legacy benefits (21) demonstrated in timely manner to ensure that the benefits ample, for HbA1c ,7%, the hazard ratio the UKPDS, although such benefits were of reduced blood glucose are realized. It was 1.24 (95% CI 1.06–1.37) (15). This is a demonstrated after 10 years (22). An- is expected that sustained reduced HbA1c generous span of the whole spectrum other limitation is the sole emphasis on levels over a long period of time, achieved of poor glycemic control and allows for the early sustained intensive glycemic through the confines of the strict proto- adequate evaluation of the effects of control. In real clinical practice, there are cols of randomized control trials, will be good early control on glycemic legacy. marked variations in care standards and associated with reduced microvascular The UK Prospective Diabetes Study patients’ self-management, thus result- and macrovascular events (Fig. 1). In (UKPDS) demonstrated the legacy ing in variation in HbA1c levels, which has addition to therapeutic inertia in glycemic effects of HbA1c reduction below 7% also been associated with poor outcomes control, recent studies have also revealed (3). Additionally, the American Diabe- (23). Also, even though this study is associations of long-term glycemic vari- tes Association–European Association dubbed a real-world study, 34,737 of ability of HbA1c with microvascular and for the Study of Diabetes position state- the eligible 44,763 patients were in- macrovascular control (23). The mecha- ment also recommended HbA1c targets cluded in the analysis. The strength of nism for this is not entirely certain, of 6.5% in newly diagnosed patients (10). real-world studies is the inclusion of but HbA1c variability has been shown There are a number of strengths to this patients with varying characteristics, to increase urinary excretion of 8-iso- study. It is a population-based study reminiscent of those in real clinical prac- prostaglandin F2a, a marker of oxidative with a large number (34,737) of partic- tice, as opposed to the refined selection stress (25). Wide variations of HbA1c could ipants included in the analysis. The an- of patients in randomized controlled reflect poor adherence to therapeutic and alytical methodology was very thorough, trials (24). In this instance, it is unclear lifestyle interventions, a more compli- capturing not just the raised levels of whether the remaining 22.4% of the cated clinical course, or a subpar organi- HbA1c but also the periods of dysglycemic patients excluded would have behaved zational process of diabetes care. Figure 1 exposures. Another strength of this study differently. Therefore, the exclusion of demonstrates the theoretical adverse ef- was that it included an ethnically diverse such a large proportion of the study fects of both a legacy of dysglycemia as a population. Most studies on this subject population limits the generalizability of result of therapeutic inertia (potentially are usually conducted in predominantly the results of this study. Finally, although avoided through the confines of strict Caucasian populations, thus making the results less generalizable to other pop- ulations. The study also highlights in its analysis all diabetes complications, in- cluding microvascular complications, macrovascular complications, and mor- tality. Previous publications on this sub- ject have usually just focused on macrovascular (7) or microvascular com- plications (16) or surrogate cardiovascu- lar risk factors. The authors also focused mainly on the effect of glycemic control rather than the effects of specific anti- diabetes medications. Hopefully, this ap- proach ruled out potential confounding by the pleiotropic effects of newer gly- cemic agents with extraglycemic cardio- fi – Figure 1—A schematic representation of the effects of early intensive glycemic control in vascular bene ts (17 20). preventing initial microvascular complications and then macrovascular complications several Despite these strengths, the study is years later. Failure to initially control and maintain glycemia at diagnosis or sustained glycemic not without limitations. In the first variability leads to the dysglycemic legacy of diabetes complications. care.diabetesjournals.org Khunti and Seidu 351

randomized controlled trial protocols) 5. Khunti K, Kosiborod M, Ray KK. Legacy bene- Diabetes & Aging Study). Diabetes Care 2019; fits of blood glucose, blood pressure and lipid 42:416–426 and variability of HbA1c control. Future studies should be designed to determine control in individuals with diabetes and cardio- 16. Laiteerapong N, Karter AJ, Moffet HH, et al. vasculardisease:timetoovercome multifactorial Ten-year hemoglobin A1c trajectories and out- outcomes in people intensified to indi- therapeutic inertia? Diabetes Obes Metab 2018; comes in type 2 diabetes mellitus: The Diabetes & vidualized targets with reduced glycemic 20:1337–1341 Aging Study. J Diabetes Complications 2017;31: variability. 6. Rawshani A, Rawshani A, Franzen´ S, et al. 94–100 Mortality and cardiovascular disease in type 1 17. Neal B, Perkovic V, Mahaffey KW, et al.; and type 2 diabetes. N Engl J Med 2017;376: CANVAS Program Collaborative Group. Canagli- 1407–1418 flozin and cardiovascular and renal events Funding. The authors acknowledge support 7. Paul SK, Klein K, Thorsted BL, Wolden ML, in type 2 diabetes. N Engl J Med 2017;377: from National Institute for Health Research Khunti K. Delay in treatment intensification in- 644–657 (NIHR) Collaboration for Leadership in Applied creases the risks of cardiovascular events in 18. Marso SP, Daniels GH, Brown-Frandsen K, – Health Research and Care East Midlands and patients with type 2 diabetes. Cardiovasc Dia- et al.; LEADER Steering Committee; LEADER Trial the NIHR Leicester Biomedical Research Centre. betol 2015;14:100 Investigators. Liraglutide and cardiovascular out- Duality of Interest. K.K. has acted as a consul- 8. Khunti K, Wolden ML, Thorsted BL, Andersen comes in type 2 diabetes. N Engl J Med 2016; tant and speaker for Novartis, Novo Nordisk, M, Davies MJ. Clinical inertia in people with 375:311–322 fi Sano , Eli Lilly, and Merck Sharp & Dohme. He has type 2 diabetes: a retrospective cohort study of 19. Marso SP, Bain SC, Consoli A, et al.; SUSTAIN- received grants in support of investigator and more than 80,000 people. Diabetes Care 2013; 6 Investigators. Semaglutide and cardiovascular investigator-initiated trials from Novartis, Novo 36:3411–3417 outcomes in patients with type 2 diabetes. N fi fi Nordisk, Sano , Eli Lilly, P zer, Boehringer In- 9. Yam FK, Adams AG, Divine H, Steinke D, Jones Engl J Med 2016;375:1834–1844 gelheim, and Merck Sharp & Dohme. S.S. has MD. Clinical inertia in type 2 diabetes: a retro- 20. Zinman B, Wanner C, Lachin JM, et al.; EMPA- acted as a consultant, advisory board member, spective analysis of pharmacist-managed diabe- REG OUTCOME Investigators. Empagliflozin, car- fi and speaker for Novo Nordisk, Amgen, Sano , Eli tes care vs. usual medical care. Pharm Pract diovascular outcomes, and mortality in type 2 Lilly, Merck Sharp & Dohme, Boehringer Ingel- (Granada) 2013;11:203–210 diabetes. N Engl J Med 2015;373:2117–2128 heim, AstraZeneca, Janssen, Napp, and Novartis. 10. American Diabetes Association. Standards 21. Wiitala WL, Reaven PD, Emanuele NV, He has received research grants from Janssen. of Medical Care in Diabetesd2019. Diabetes Gerstein HC. The Veterans Affairs Diabetes Trial Care 2019;42(Suppl. 1):S1–S183 (VADT) at 15 years. Symposium presented at the References 11. Laiteerapong N, Cooper JM, Skandari MR, 78th Scientific Sessions of the American Diabetes 1. U.K. Prospective Diabetes Study Group. U.K. et al. Individualized glycemic control for US Association, 24 June 2018, Orlando, FL Prospective Diabetes Study 16. Overview of adults with type 2 diabetes. Ann Intern Med 22. Hayward RA, Reaven PD, Wiitala WL, et al.; 6 years’ therapy of type II diabetes: a progressive 2018;168:170–178 VADT Investigators. Follow-up of glycemic con- disease. Diabetes 1995;44:1249–1258 12. Khunti K, Davies MJ. Clinical inertiadtime to trol and cardiovascular outcomes in type 2 di- 2. Patel A, MacMahon S, Chalmers J, et al.; reappraise the terminology? Prim Care Diabetes abetes. N Engl J Med 2015;372:2197–2206 ADVANCE Collaborative Group. Intensive blood 2017;11:105–106 23. Gorst C, Kwok CS, Aslam S, et al. Long-term glucose control and vascular outcomes in pa- 13. Fritzen K, Heinemann L, Schnell O. Modeling glycemic variability and risk of adverse outcomes: tients with type 2 diabetes. N Engl J Med 2008; of diabetes and its clinical impact. J Diabetes Sci a systematic review and meta-analysis. Diabetes 358:2560–2572 Technol 2018;12:976–984 Care 2015;38:2354–2369 3. UK Prospective Diabetes Study (UKPDS) 14. Baxter M, Hudson R, Mahon J, et al. Esti- 24. Mc Cord KA, Al-Shahi Salman R, Treweek S, Group. Effect of intensive blood-glucose control mating the impact of better management of et al. Routinely collected data for randomized with on complications in overweight glycaemic control in adults with type 1 and type 2 trials: promises, barriers, and implications. Trials patients with type 2 diabetes (UKPDS 34). Lancet diabetes on the number of clinical complications 2018;19:29 1998;352:854–865 and the associated financial benefit. Diabet Med 25. Monnier L, Mas E, Ginet C, et al. Activation of 4. Holman RR, Paul SK, Bethel MA, Matthews DR, 2016;33:1575–1581 oxidative stress by acute glucose fluctuations Neil HAW. 10-year follow-up of intensive glu- 15. Laiteerapong N, Ham SA, Gao Y, et al. The compared with sustained chronic cose control in type 2 diabetes. N Engl J Med legacy effect in type 2 diabetes: Impact of early in patients with type 2 diabetes. JAMA 2006;295: 2008;359:1577–1589 glycemic control on future complications (The 1681–1687