Early insulin treatment in type 2 diabetes: what are the pros?

The prevalence of diabetes in the world is growing at an unprecedented rate and rapidly becoming a health concern and burden in both developed and developing countries (1). In addition, we are now witnessing an upsurge in the incidence of type 2 diabetes in children and adolescents, with the potential of translating into a future catastrophic disease burden as vascular complications of the disease begin affecting a younger population. Although there may be contention regarding the impact of lowering glycemia on macrovascular disease risk, there is strong consensus of the definite benefits of lowering blood glucose to reduce the risk of retinopathy and nephropathy in either type 1 or type 2 diabetes (2,3). Despite supporting data and multiple guidelines advanced by professional organizations, overall glycemic control falls far below expectations (4). Overall, <36% of individuals with diabetes are at recommended glycemic targets, with the most difficult-to-control cases represented by insulin-deficient individuals on insulin therapy to manage their diabetes (4). Furthermore, as β-cell dysfunction progresses over time, many patients with type 2 diabetes, treated with oral agents, fail to achieve or maintain adequate glycemic control. Unfortunately, in many of these cases, antiglycemic therapy is not adjusted or advanced, thereby exposing patients to prolonged hyperglycemia and the increased risk of diabetes-related complications. The term “clinical inertia,” which has come to define the lack of initiation, or intensification of therapy when clinically indicated (5), is most pronounced in the setting of insulin initiation. Subjects with type 2 diabetes, managed in a large integrated health care system, were initiated on additional blood glucose–lowering treatment only when the mean baseline A1C reached a value of 9.0% (6). Patients started on insulin had an even higher mean A1C of 9.6% and tended to have more severe baseline complications and comorbidities than those started on sulfonylurea, or metformin therapy. In addition, the higher the starting A1C when therapy was initiated or changed, the less likely the patient was of achieving adequate glycemic control (6). Although specialists are slightly more proficient than general practitioners in intensifying diabetes therapy when warranted (7), overall clinical inertia results in the majority of patients failing to achieve, or maintain, adequate metabolic goals from a period of months to several years (8,9). In summary, to improve these suboptimal metabolic outcomes, and reduce the risk of disease-related complications, more intensive management of glycemia is warranted, including the option of introducing insulin therapy earlier than the current widely practiced substandard of care.

INTRODUCTION OF INSULIN EARLIER IN THE TREATMENT PARADIGM

Typically, whereas introducing insulin therapy in a more timely fashion would significantly improve glycemic control among subjects with type 2 diabetes, the question of insulin initiation timing in relation to other antiglycemic therapies is the subject of considerable debate (10). While insulin administration has the potential of achieving the most effective reductions in glycemic control, the initiation of insulin therapy requires greater use of resources, time, and effort from provider and patient alike, compared with oral antidiabetic therapies (11). Patient resistance to the use of insulin therapy remains a challenge, especially in populations that may have misgivings and misconceptions regarding the role of insulin replacement in diabetes management.

Notwithstanding these issues, there are specific populations that would clearly benefit from early, aggressive, and targeted introduction of insulin therapy. For instance, patients presenting with significant hyperglycemia may benefit from timely initiation of insulin therapy that can effectively and rapidly correct their metabolic imbalance and reverse the deleterious effects of excessive glucose (glucotoxicity) and lipid (lipotoxicity) exposure on β-cell function and insulin action (12). In vitro studies have demonstrated that chronic hyperglycemia leads to increased production of reactive oxygen species, and subsequent oxidative stress, which appears to affect insulin promoter activity (PDX-1 and MafA binding) and results in diminished insulin gene expression in glucotoxic β-cells (13). Interestingly, in vitro experiments have shown that these glucotoxic effects occur in a continuum of glucose concentrations (no clear threshold effect), are reversible with reinstitution of euglycemic conditions, and result in the greatest recovery of β-cell function with shorter periods of exposure to hyperglycemia (14). Various studies have demonstrated improvement in insulin sensitivity and β-cell function after correction of hyperglycemia with intensive insulin therapy (15).

INTENSIVE INSULIN TREATMENT AND β-CELL FUNCTION

A number of trials have evaluated the strategy of implementing short-term aggressive insulin replacement as first-line therapy in the management of hyperglycemia in newly diagnosed type 2 diabetes (Table 1), with the goal of improving and preserving β-cell function, reducing insulin resistance, and maintaining optimal glycemic control through disease “remission” (16–18). In these studies, intensive insulin therapy was delivered via multiple daily insulin injections, or insulin pump therapy (continuous subcutaneous insulin infusion), over a period of 2–3 weeks, with achievement of euglycemia in ∼90% of subjects on completion of insulin treatment. After insulin withdrawal, patients were maintained on diet therapy only, with 42–69% maintaining euglycemia 12 or more months after treatment. Patients who achieved and maintained long-term euglycemia tended to have a better response to insulin therapy, as well as associated improvements in β-cell function, including first-phase insulin release, as measured by homeostasis model assessment of β-cell function (HOMA-B) and intravenous glucose tolerance tests.

Table 1

Baseline characteristics and outcomes of patients with type 2 diabetes receiving temporary insulin therapy at disease diagnosis

	n	Age	BMI	Baseline A1C (%)	Insulin dose (units · kg−1 · day−1)	Days to glycemic control	Duration insulin therapy (weeks)	% Early responders	% Sustained responders	Weight change
Ilkova et al. (17)	13	50	27	11.2	0.61	1.9	2	92	69 (26 months)	0.4 kg
Li et al. (16)	126	50	25	10.0	0.7	6.3	2	90	42 (24 months)	−0.04 kg/m2
Ryan et al. (18)	16	52	31	11.8	0.37–0.73	<14	2–3	88	44 (12 months)	−0.5 kg/m2

Early responders are subjects who achieved euglycemia with insulin treatment, and late responders are subjects who maintained long-term euglycemia without pharmacotherapy after the initial insulin treatment.

Improvements in β-cell function and insulin action have also been reported when euglycemia is achieved with noninsulin therapies (19). Unfortunately, as illustrated by the U.K. Prospective Diabetes Study, long-term glycemic control in type 2 diabetes is difficult to maintain, regardless of the therapeutic intervention due, in part, to progressive loss of β-cell function over time. The recently published A Diabetes Outcome Progression Trial (ADOPT) demonstrated longer maintenance of glycemic control in patients using a thiazolidinedione (rosiglitazone) compared with glyburide or metformin monotherapy, although β-cell function, as measured by HOMA-B was no different at the end of the trial between the rosiglitazone and sulfonylurea groups (20); the benefits in durability of control seemed to have been a result of improved insulin sensitivity.

A recent study comparing intensive insulin therapy (multiple daily insulin injections or continuous subcutaneous insulin infusion) with oral hypoglycemic agents (glicazide and/or metformin) in newly diagnosed patients with type 2 diabetes provided some provocative results (21). In this trial, 92% of 382 subjects with poorly controlled diabetes achieved glycemic targets (fasting and 2-h postprandial capillary glucose levels of <110 mg/dl and <144 mg/dl, respectively) within an average of 8 days from start of therapy (Table 2). Treatment was withdrawn after 2 weeks of normoglycemia, followed by diet and exercise management. A greater proportion of patients randomized to intensive insulin therapy achieved glycemic targets and did so in a shorter period compared with oral agent therapy (Table 2). Shortly after discontinuing antiglycemic treatment, measures of first-phase insulin release, HOMA-B and HOMA-IR were similar among all treatment groups. By the end of 1 year, remission rates were significantly higher in the groups that had received initial insulin therapy (51 and 45% in the continuous subcutaneous insulin infusion and multiple daily insulin injections groups, respectively), compared with 27% in the oral therapy group. Whereas in the oral agent group, acute insulin response at 1 year declined significantly compared with immediate post-treatment, it was maintained in the insulin treatment groups. Of note, responders typically had higher BMI, less baseline hyperglycemia, and greater responsiveness to therapy than nonresponders.

Table 2

Baseline characteristics and clinical outcomes comparing subjects treated with insulin or oral agent therapies lasting for 2 weeks after achievement of normoglycemia

	Continuous subcutaneous insulin infusion	Multiple daily injections	Oral agents
n	133	118	101
Age (yrs)	50	51	52
BMI (kg/m2)	25	24	25
Baseline A1C (%)	9.8	9.7	9.5
% Achieving euglycemia	97	95	83
Time to euglycemia (days)	4	5.6	9.3
Daily drug doses	0.68 units/kg (mean)	0.74 units/kg (mean)	Glicazide 160 mg + metformin 1,500 mg (max median)
Δ in AIR* (pmol · l−1 · min−1)	951	800	831
AIR (median) in remission groups at 1 year	809	729	335†

From Weng et al. (21).

*Change in median AIR (acute insulin response) between baseline and treatment end.

†P < 0.05 compared with continuous subcutaneous insulin infusion.

Another study comparing early and continued insulin treatment versus oral agent therapy (glibenclamide) over a period of 2 years in recently diagnosed patients with type 2 diabetes showed better long-term glycemic control and β-cell function in the insulin-treated group (22). There was no difference in weight gain between insulin and oral agent therapy and no reported cases of severe hypoglycemia, reflecting easier-to-manage glycemia, probably as a result of better endogenous insulin production.

POTENTIAL PHYSIOLOGICAL EFFECTS OF INSULIN REPLACEMENT THERAPY

What could account for some of the differences in β-cell function seen in studies with early aggressive insulin therapy? A study evaluating the anti-inflammatory effects of an insulin infusion on obese subjects without diabetes demonstrated suppression of nuclear factor κB. Nuclear factor κB is the key transcription factor responsible for the transcription of proinflammatory cytokines, adhesion molecules and enzymes responsible for producing reactive oxygen species (23). As a consequence, insulin infusion significantly suppressed generation of reactive oxygen species and decreased concentrations of plasma soluble intercellular adhesion molecule-1 (sICAM-1), monocyte chemo-attractant protein-1 (MCP-1), and plasminogen activator inhibitor-1 (PAI-1), among other observed anti-inflammatory actions (24).

Could the timing of the intervention affect the metabolic response to insulin therapy? For example, loss of first-phase insulin response, possibly as a consequence of glucotoxicity, is evident with fasting plasma glucose concentrations >115 mg/dl (25). Often, when diabetes is diagnosed, fasting plasma glucose levels are usually significantly higher, and may have been so for quite some time (26), exposing β-cells to chronic hyperglycemia and consequent β-cell decompensation (13). It could be hypothesized that early aggressive physiologic insulin replacement with both prandial and basal coverage results in rapid improvement in glucolipotoxicity, reduction of the inflammatory milieu, and consequent greater preservation of β-cell function. Some of these improvements in β-cell function were also evident after rigorous management with glyburide and metformin.

INSULIN REPLACEMENT OPTIONS AND STRATEGIES

Whereas the use of insulin therapy in newly diagnosed subjects with type 2 diabetes appears to be associated with a low risk of hypoglycemia and weight gain, the use of algorithm-driven insulin replacement in more advanced disease is often associated with a greater incidence of weight gain and hypoglycemia. Individualizing the insulin prescription may minimize some of these adverse outcomes. Using the A1C status of a patient, the fasting blood glucose, and if available, the postprandial glucose could assist the provider in individualizing insulin replacement. Published trials in suboptimally controlled insulin-naive type 2 diabetes seem to indicate that basal insulin replacement yields similar effectiveness, but with less weight gain, and hypoglycemia risk than basal/prandial or mixed insulin strategies, when baseline A1C is ≤8.5% (27). Thus, in a patient whose predominant glycemic burden occurs overnight and whose A1C level is within 1–2% points of target, starting with a low dose of basal insulin (0.2 units · kg−1 · day−1) and adjusting the dose to achieve fasting blood glucose levels <110–130 mg/dl often proves an effective strategy. With higher A1C levels, replacing prandial insulin, with or without basal insulin coverage, results in greater A1C reduction than basal-only replacement, albeit at the expense of more weight gain and hypoglycemia (28). For example, patients inadequately controlled on basal insulin can be started on one or more doses of rapid-acting insulin (0.05 units/kg/meal) before one or more meals (usually the largest meals), and the insulin dose titrated to achieve postprandial blood glucose levels <180 mg/dl. A basal/bolus insulin replacement, giving patients flexible prandial dosing instructions, as opposed to fixed doses of premeal insulin, has been shown to be associated with equivalent glycemic control, but with less weight gain (29). Furthermore, the use of basal insulin analogs (glargine or detemir) is associated with less hypoglycemia (especially nocturnal hypoglycemia) and, in the case of insulin detemir, less weight gain than human NPH insulin (27,30).

CONCLUSIONS

In summary, aggressive and often temporary use of insulin therapy at disease onset in type 2 diabetes is associated with effective glycemic control with minimal weight gain and hypoglycemia. Early restitution of physiologic insulin secretion and glycemic control could be, in theory, followed by therapies to prolong maintenance of euglycemia, such as thiazolidinediones- (20) or glucagons-like peptide 1–based interventions (to date not clinically tested). A more timely and selective introduction of insulin replacement therapy, as β-cell function progresses, could facilitate the achievement and maintenance of euglycemia and thus reduce disease-associated complications.