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 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 glucose 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.