Challenges in Choosing a Medication for Type 2 Diabetes

Introduction

The goals of diabetes care are to reduce patients' risk of acute and long-term complications, increase longevity, and improve health-related quality of life.

Relying on evidence from pivotal trials, tight glycemic control (hemoglobin A1c [HbA1c] ≤7.0%) has been the usual approach to reach these goals. Lately, the positive patient-important cardiovascular results of the EMPA-REG, LEADER, and SUSTAIN-6 trials, in addition to numerous new and upcoming medications, have produced even more enthusiasm. ¹ This, while exciting, has also conveyed some challenges in the patient-centered prescription for type 2 diabetes. Perhaps awareness of these situations may help clinicians decide what is best for each patient with diabetes.

Glycemic Control—“One-Size-Fits-All”?

There is a consensus that tight glycemic control reduces the risk of chronic, particularly microvascular, diabetes complications. Clinical practice guidelines, quality improvement interventions, quality-of-care measures, and patient-directed marketing have focused on this strategy. ^{2 –5} Under this assumption, the two-to-threefold risk of severe hypoglycemia and the increasing complexity of treatment regimens (i.e., burden of treatment) have been accepted by clinicians and endured by patients. ^2,3 However, evidence that tight glycemic control has a positive effect on important patient microand macrovascular outcomes remains uncertain and open to debete. ^{3,6 –8}

In the UKPDS trial there was an absolute risk reduction of 3.5% in “all-diabetes outcomes” in patients randomized to the intensively treated group; however, the composite outcome involved 14 individual outcomes with a different gradient of importance (e.g., from death to cataract extraction), and 2.7% of the absolute risk reduction was explained by a surrogate marker, that is, photocoagulation. ^9,10 Of note, in the ACCORD, ADVANCE, and VADT trials there was no difference between tight and conventional glycemic control in patient-important outcomes (i.e., outcomes that patient feel can care about) but instead in surrogate markers (e.g., microalbuminuria, photocoagulation, and nonclinical neuropathy, among others). ^{11 –13} More recently, a systematic review found that although the body of evidence reflects no impact of tight glycemic control on patient-important microvascular outcomes (end-stage renal disease, dialysis, renal death, blindness, and clinical neuropathy), there is an overwhelming consensus among experts and guideline panelists favoring this approach, perhaps driven by the positive effect of tight glycemic control on surrogate markers such as microalbuminuria or photocoagulation. ³ There is also no effect on all-cause mortality, cardiovascular mortality, stroke, or amputations. Yet, in meta-analysis and some primary studies, there is a 15% relative-risk reduction of nonfatal myocardial infarctions. ^3,6,14,15 Despite this, after a single study reported an increased cardiovascular risk with tight glycemic control (i.e., ACCORD), there is a more skeptical view with regard to its effect on macrovascular outcomes. ³

Which Medication to Choose and How?

To date, no diabetes drug per se has demonstrated to improve patient-important microvascular outcomes; however, three recent trials have reported a positive effect on cardiovascular outcomes. The EMPA-REG (empagliflozin), LEADER (liraglutide), and SUSTAIN-6 (semaglutide) trials have demonstrated a positive effect on cardiovascular outcomes. ¹ However, these were not head-to-head trials, but rather noninferiority transformed into superiority trials, which compared the active drug with placebo. Furthermore, patients in these studies were at high to very high cardiovascular risk; therefore, they do not represent the vast majority of patients with type 2 diabetes. Hence, the number-needed-to-treat in the low-to-moderate risk population would be considerably higher than that described in the original report. Finally, in these studies, with the exception of 0.1 mg of semaglutide (HbA1c 7.3%), the end of study HbA1c for patients randomized to the intensive group was around 8.0%, a finding that supports the lack of benefit of tight glycemic control in high-risk patients with diabetes and supports the notion of a different pathophysiological explanation rather than just glycemic control. ^{16 –18} Thus, today clinicians and patients with type 2 diabetes have a vast number of treatment options within multiple drug classes. The body of evidence shows a mix of potential benefits; however, agents differ in their safety profile, convenience, and out-of-pocket cost, a fact that will impact the decision-making process of patients and clinicians. For instance, today new technologies such as continuous-glucose monitoring and long-acting insulin have demonstrated to decrease the risk of hypoglycemia; this advantage must be weighted individually in terms of cost and convenience of its use. ^{19 –23}

The GRADE trial, a pragmatic study (i.e., comparative effectiveness trial) is currently evaluating, in addition to metformin, the long-term effects of four classes of medications for type 2 diabetes: insulin glargine, sitagliptin, liraglutide, and glimepiride. This trial, nonetheless, completely oversees patient-important outcomes (both the primary and secondary objectives of the study stem around the effect of these drugs to reduce HbA1c). ²⁴ In addition, the EASD, ADA, and AACE practice guidelines advocated a patient-centered approach in the care of individuals with diabetes. ²⁵ However, these guidelines fail to mention in what way or using which tools can patient-centered care in diabetes be practiced. Although life expectancy, hypoglycemia risk, disease duration, vascular complications, comorbidities, patient attitude, and social and economic resources have to be taken into account, this is only to establish an individualized HbA1c goal. Hence, the fact that patients with type 2 diabetes, by doing these, live longer independently, feel better, and live unhindered by complications remains uncertain.

Guidelines—Disease Specific and Context Blind?

Patients with type 2 diabetes seldomly arrive to the clinical encounter with a single illness. ²⁶ Therefore, patients with diabetes frequently suffer from multimorbidity, including obesity, hypertension, dyslipidemia, depression, chronic pain, cardiovascular disease, and renal impairment, among others. Although diabetes guidelines advocate for a certain HbA1c goal, it is important to acknowledge that, in addition to the already complex treatment regimen the patient has, reaching this goal will require usually new and more medications, monitorization, consultations, and an increased cost. Consequently, each treatment option not only involves a set of potential benefits and harms, but also brings an obligatory set of activities that require the patients' energy, attention, and time. Therefore, even with high-quality evidence to support a particular recommendation, straightforward decisions about adding a new treatment or intervention can become challenging in a patient who is affected by multimorbidity. In these situations, shared decision-making (SDM) (i.e., a patient-centered approach to medical decision-making in which patients and clinicians work together and engage in a deliberative dialogue about reasonable treatment options) can help clinicians uncover the patient's values, preferences, and context. ²⁷ This information can then be used to shape a treatment plan that is not only needed but also wanted and likely implemented. In SDM, both patients and clinicians are considered experts: the clinicians of the research evidence and the patients of how the illness is day-to-day lived and beard with. Tools (also called decision aids) that facilitate SDM are free and available at (http://shareddecisions.mayoclinic.org) and (https://decisionaid.ohri.ca/AZlist.html). The result of this interaction may also lead to a deintensification of therapy (including the option to forgo so-called indicated treatments) and an intensification of support interventions.

Conclusion

Today clinicians and patients with type 2 diabetes have a vast number of treatment options within multiple drug classes. To provide the best available treatment, clinicians must rely on the best available evidence. This is a difficult task not only because of uncertainties in the body of evidence regarding the impact of glycemic control and the best treatment regimen but also because the patients' comorbidities and context must be taken into account in the decision-making process. Still, clinicians and patients with type 2 diabetes have to make clinical decisions. A patient-centered approach such as SDM can facilitate this process by revealing the patients' values, preferences, and context to fit and accommodate a particular treatment plan into the patients' life. At the end, the hope is to have a kind and careful care for every patient with diabetes.