Based on the UK Prospective Diabetes Study (UKPDS) and the Kumamoto study, most of the clinical practice guidelines for glycemic control target of type 2 diabetes mellitus (T2DM) since the early the 2000s have recommend the achievement of near-normoglycemia to prevent the onset or progression of diabetic microangiopathy [1,2]. The progressive nature of T2DM in the deterioration of pancreatic β-cell function was such that after 3 years, approximately 50% of patients could attain glycosylated hemoglobin (HbA1c) below 7% with monotherapy, and by 9 years, this declined to approximately 25% [1]. Early intensive insulin therapy in patients with newly diagnosed T2DM has favorable outcomes on the recovery and maintenance of β-cell function and protracted glycemic remission compared to treatment with oral hypoglycemic agents (OHAs) [3-6]; thus, supporting the rationale for the initiation of insulin therapy in subjects with T2DM whose glycemic targets cannot be achieved with lifestyle modifications and OHAs [7,8]. The Korean Diabetes Association recommends insulin therapy in two circumstances: timely initial treatment after the diagnosis of T2DM and in cases of OHA failure.

The initiation of insulin is recommended when severe hyperglycemia (more than 9% HbA1c) is detected at diagnosis, especially if hyperglycemic symptoms (polyuria or polydipsia) or any catabolic features (weight loss or ketosis) are present. Insulin should be also considered when adequate glycemic control is not obtained in patients with decompensated hepatic or renal insufficiency and when patients have suffered from myocardial infarction, stoke, or a major operation [9]. For patients with T2DM who fail to achieve the glycemic target with adequate treatment with OHAs, proceed to insulin injection therapy.

Insulin therapy has the powerful advantage of improving the glycemic control better than other OHAs, but it may be associated with risks of hypoglycemia and weight gain. Therefore, healthcare professionals need to provide comprehensive self-care education including insulin injection skills, self-monitoring of blood glucose, hypoglycemia management, and simple dosage adjustment before patients begin insulin therapy [8,10].

Basal insulin alone or in combination with OHAs is easy to administer and is the preferred choice. Basal insulin including both intermediate-acting and long-acting analogue alone is the most convenient initial insulin regimen. Because long-acting basal analogues (glargine, detemir, and degludec) are evident for reducing the risk of hypoglycemia compared to neutral protamine Hagedorn (NPH) in T2DM, they are preferable to NPH when there is a history of hypoglycemia [11]. Recently, several concentrated basal insulin preparations (U-300 glargine and U-200 degludec) were developed to allow a higher dose of basal insulin injection. Premixed insulin products contain both a basal and prandial component (NPH/regular 70/30, 70/30 aspart mix, 75/25, or 50/50 lispro mix), allowing coverage of both basal and prandial needs with a single injection. In the 4-T study [12,13], patients with T2DM (mean HbA1c 8.5% and 9-year duration of diabetes at baseline) were randomized to receive premixed insulin twice daily, prandial bolus insulin three times daily, or basal insulin once daily. In this study, median HbA1c were comparable after 3 years but weight gain and hypoglycemia were less with basal than with the other insulin regimens. In this study, however, 68% to 82% of the patients were taking a second type of insulin at the end of the research period [13]. From meta-analyses, a greater proportion of patients with T2DM achieved the HbA1c goal of < 7% with biphasic or prandial insulin compared to basal insulin, but these differences were not demonstrated after adjustment for the insulin dose [14,15]. The use of premixed or prandial insulin compared to basal insulin was associated with more hypoglycemia and weight gain. Even in previously insulin-treated patients, the frequency of hypoglycemia and weight gain increases as the number of insulin injections increases [15]. The contribution of fasting and postprandial glucose (PPG) on overall hyperglycemia or HbA1c is variable among patients. PPG excursions may play a more significant role in patients with early mild to moderate diabetes, whereas fasting hyperglycemia contributes more in patients as diabetes worsens, as represented by increasing concentrations of HbA1c, particularly at > 8.4% [16]. When physicians initiate insulin therapy, either basal insulin or premixed insulin [17,18], they should understand the pharmacodynamic profile of each formulation and also consider the baseline HbA1c in conjunction with submaximal doses of sulfonylurea (SU), blood glucose levels in a fasting or postprandial state, and symptoms of hyperglycemia (polyuria or polydipsia).

Insulin alone is not an ideal treatment for poorly controlled T2DM, as it is associated with both weight gain and inadequate glycemic control [19]. Various combinations of insulin with OHAs are now available and are widely used [20]. In patients with T2DM on maximum tolerated OHAs, metformin over placebo decreases weight gain, lowers insulin requirements, and improves glycemic control [19]. From meta-analyses including 2,171 adults with uncontrolled T2DM from 11 randomized controlled trials, adding insulin glargine to metformin monotherapy early in treatment may provide the greatest 24-week reductions in HbA1c and less symptomatic hypoglycemia benefits over regimens including SUs [21]. However, the Korean RCT comparing the commonly prescribed OHA combinations (1∶1∶1 basis with fixed doses of glimepiride, metformin, and glimepiride plus metformin) to use as an add-on therapy with insulin glargine showed a significant improvement in overall glycemic control with insignificant weight gain and the risk of hypoglycemia in the arm of combination therapy of metformin and glimepiride plus glargine insulin [22]. In this post hoc analysis of the Prospective Pioglitazone Clinical Trial in Macrovascular Events study (PROactive), pioglitazone use in combination with insulin resulted in sustained improved glycemic control with a rapid and sustained decrease in insulin doses compared to the placebo group [23]. More insulin-resistant patients (defined as poorly controlled T2DM despite high doses of insulin) in the pioglitazone plus insulin group showed the greatest glycosylated hemoglobin decline [23]. From meta-analyses including 3,092 patients from eight RCTs comparing pioglitazone in combination with any insulin-containing regimen compared to the same insulin regimen alone, pioglitazone confers a small advantage in terms of HbA1c in T2DM patients with previous inadequate glucose control but at the cost of increased hypoglycemia and weight gain [24]. Numerous studies of dipeptidyl peptidase 4 (DPP4) inhibitor add-on therapy compared to insulin showed significant improvement in glycemic control relative to the placebo without increasing hypoglycemia or body weight [15,25,26]. Sodium glucose cotransporter 2 (SGLT2) inhibitor is a novel insulin-independent OHA that reduces hyperglycemia by reducing proximal renal glucose reabsorption, causing urinary glucose excretion. The adjunctive use of a SGLT2 inhibitor improved glycemic control and reduced weight without increasing the risk of hypoglycemia and with lower insulin requirements, although potential side effects of urinary tract infection and euglycemic diabetes ketoacidosis should be considered [27-29]. In a covariate-adjusted indirect comparison using meta-regression analyses including five SGLT2 inhibitors and nine DPP4 inhibitors studies, SGLT 2 inhibitors achieved better glycemic control and greater weight reduction than DPP4 inhibitors without increasing the risk of hypoglycemia in patients with T2DM that is inadequately controlled with insulin [30]. When physicians initiate insulin therapy in patients with T2DM, metformin should be continued while other oral agents may be continued or discontinued on an individual basis, especially insulin regimens to avoid unnecessarily complex or costly OHA regimens.

In patients above the HbA1c target on basal insulin or premixed insulin once or twice daily, recommendations for further intensification, if needed, are outlined in Fig. 1 [31]. When physicians intensify an insulin regimen, they should consider the advantages and disadvantages such as flexibility, complexity, and frequency of hypoglycemia.

Intensified insulin might consist of dose titration and regimen modification. Once the initiation of an insulin regimen is stable, dose titration for adjusting insulin are made based on the fasting and PPG levels. If a patient is still above the HbA1c target with an acceptable fasting blood glucose level on titrated basal insulin, options for treatment intensification are either a single injection of rapid-acting insulin (lispro, aspart, or glulisine) at the largest meal, glucagon-like peptide-1 (GLP-1) receptor agonist (GLP-1RA), or switching to twice daily injections of premixed insulin. These recommendations were based on the non-inferior results of basal insulin + single injection of either rapid-acting insulin or GLP-1RA relative to twice daily premixed insulin [12,13,32-35]. Basal insulin plus GLP-1RA resulted in less hypoglycemia and weight loss compared to other insulin regimens [33,34]. If a patient is still above the HbA1c target on basal insulin + a single injection of rapid-acting insulin, naturally advancing to a basal-bolus regimen (≥ 2 times of rapid-acting insulin) should be considered [36]. If a patient is still above the HbA1c target on initial premixed insulin once or twice daily with dose titration, naturally advancing to premixed analog insulin 2 or 3 times daily has been found to be non-inferior to basal-bolus regimens with similar rates of hypoglycemia [17,37]. If a patient is still above the HbA1c target on an intensified insulin regimen, an option for treatment is switching one regimen with another (such as a basal-bolus regimen to premixed analog insulin three times daily or vice versa) [38,39].