New England Journal of Medicine SEPTEMBER 22, 2022

Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE)

The GRADE Study Research Group

Bottom Line

In a randomized, pragmatic trial of adults with early type 2 diabetes on metformin, the addition of insulin glargine or liraglutide was more effective than glimepiride or sitagliptin at achieving and maintaining glycemic control (HbA1c <7.0%).

Key Findings

1. Over a mean follow-up of 5.0 years, insulin glargine and liraglutide demonstrated superior glycemic durability compared to glimepiride and sitagliptin, with the primary outcome (HbA1c ≥7.0%) occurring less frequently in the glargine and liraglutide arms.

2. The cumulative incidence of reaching the primary HbA1c threshold was 26.5 per 100 participant-years for glargine and 26.1 per 100 participant-years for liraglutide, compared to 30.4 for glimepiride and 38.1 for sitagliptin (P<0.001 for overall difference).

3. Severe hypoglycemia was rare across all groups but occurred more frequently with glimepiride (2.2%) compared to sitagliptin (0.7%), glargine (1.3%), and liraglutide (1.0%).

4. No material differences were found among the treatment groups for microvascular outcomes or cardiovascular disease events over the study duration.

Study Design

Design

RCT

Open-Label

Sample

5,047

Patients

Duration

5.0 yr

Median

Setting

Multicenter, US

Population Adults with type 2 diabetes (duration <10 years), aged ≥30 years, treated with metformin (1,000–2,000 mg/day) with HbA1c between 6.8% and 8.5%.

Intervention Randomized to one of four glucose-lowering medications added to metformin: insulin glargine, glimepiride, liraglutide, or sitagliptin.

Comparator Other arms within the four-group randomized trial (active comparative effectiveness).

Outcome Time to failure to maintain an HbA1c level <7.0%, confirmed by a subsequent measurement.

Study Limitations

• The study was unmasked, which could introduce bias in the reporting of subjective outcomes or in the management of glycemic goals by participants and providers.

• SGLT-2 inhibitors, now a standard second-line treatment, were not included in this trial, limiting the applicability of these findings to current clinical practice guidelines.

• The pragmatic design allowed for variability in clinical practice at 36 different centers, which, while increasing real-world applicability, may introduce heterogeneity.

• Most participants were unable to maintain glycemic targets over the long term, regardless of the treatment group, reflecting the progressive nature of type 2 diabetes.

Clinical Significance

The results provide important comparative data for choosing second-line glucose-lowering agents for patients inadequately controlled on metformin, suggesting that insulin glargine and liraglutide may offer superior long-term glycemic durability compared to sitagliptin or glimepiride.

Historical Context

Prior to GRADE, most diabetes medication trials were placebo-controlled and evaluated medications as monotherapy or in limited head-to-head scenarios, leaving significant uncertainty about the optimal second-line treatment strategy after metformin. GRADE was designed to fill this evidence gap by directly comparing four major classes of medications in a real-world, long-term clinical setting.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

Compare the mechanisms of action of the four study drugs (glargine, liraglutide, glimepiride, and sitagliptin) and explain why insulin and GLP-1 receptor agonists might inherently provide more durable glycemic control than DPP-4 inhibitors or sulfonylureas.

Key Response

Glargine (basal insulin) provides exogenous replacement regardless of beta-cell function; Liraglutide (GLP-1 RA) enhances glucose-dependent insulin secretion and slows gastric emptying; Glimepiride (sulfonylurea) stimulates beta cells but is associated with secondary failure due to 'beta-cell burnout'; Sitagliptin (DPP-4 inhibitor) has a relatively weak glucose-lowering effect. The study demonstrated that 'physiologic' or more potent agents (insulin/GLP-1) maintain targets longer as the underlying disease progresses.

Resident

In a patient with T2DM currently on metformin with an HbA1c of 7.5%, the GRADE study suggests adding either glargine or liraglutide over sitagliptin. What are the key clinical trade-offs regarding weight and hypoglycemia risk that should guide this specific selection?

Key Response

The GRADE study found that while glargine and liraglutide were most effective for glycemic durability, glargine was associated with the most weight gain and a higher risk of severe hypoglycemia compared to the others. Liraglutide patients experienced more weight loss but higher rates of gastrointestinal side effects. These findings allow for individualized therapy based on patient comorbidities and tolerance.

Fellow

The GRADE trial used a primary metabolic endpoint of 'cumulative incidence of an HbA1c level of 7.0% or higher.' How does this 'time-to-failure' model refine our understanding of the natural history of T2DM treatment compared to traditional cross-sectional HbA1c reductions seen in shorter phase III trials?

Key Response

Most T2DM trials look at 26-52 week HbA1c changes. GRADE's 'failure' design emphasizes the progressive nature of T2DM. It highlights that even with high-potency agents, many patients eventually cross the 7.0% threshold, suggesting that the focus should shift from 'achieving' a target to 'sustaining' it through early combination therapy or sequential intensification.

Attending

The GRADE study showed a clear glycemic benefit for glargine and liraglutide over glimepiride and sitagliptin, yet the cardiovascular outcome differences were minimal. How should this affect your counseling of a 'low-risk' patient regarding the choice of second-line agent?

Key Response

While GRADE was not powered for cardiovascular outcomes, the lack of significant differences in the primary MACE (Major Adverse Cardiovascular Events) composite across the four groups suggests that for patients without established CVD or high risk, the choice of agent can be driven primarily by glycemic durability, cost, and side-effect profile rather than an immediate expectation of macrovascular protection.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

Critique the use of a 'pragmatic trial' design in GRADE, specifically regarding the standardized titration protocols for glargine versus the fixed-dose or clinical-judgment-based management of oral agents. How might this introduce bias in comparative effectiveness?

Key Response

Pragmatic trials aim for real-world applicability, but the glargine arm in GRADE used a rigorous titration protocol (to reach a fasting glucose target). This may favor the glargine arm's performance relative to agents like sitagliptin, which have no titration potential, or glimepiride, where titration is often limited by clinician fear of hypoglycemia, potentially overestimating the 'relative' effectiveness of insulin compared to actual community practice.

Journal Editor

As a reviewer, how would you address the significant limitation of the GRADE study's omission of SGLT2 inhibitors, and does this absence render the results less impactful for modern clinical decision-making?

Key Response

A tough reviewer would flag that by the time GRADE results were published, SGLT2 inhibitors had become standard of care for many T2DM patients. However, the editorial significance remains high because GRADE provides the most robust evidence to date for the relative durability of the other four major classes, establishing a 'benchmark' against which future SGLT2i head-to-head durability data must be compared.

Guideline Committee

Based on the GRADE findings, should the 'Step 2' algorithm in T2DM guidelines be modified to explicitly prefer GLP-1 RAs and Basal Insulin over DPP-4 inhibitors for glycemic durability, regardless of cardiovascular risk status?

Key Response

Current ADA Standards of Care prioritize GLP-1 RAs and SGLT2is for those with ASCVD or CKD. GRADE provides high-level evidence (Level A) to further support the preference for GLP-1 RAs and insulin over DPP-4 inhibitors and SUs for the general population to delay 'treatment failure.' The guidelines should emphasize that while DPP-4 inhibitors are well-tolerated, they are significantly less effective at maintaining long-term glycemic targets.

Clinical Landscape

Noteworthy Related Trials

1998

UK Prospective Diabetes Study (UKPDS)

n = 5102 · Lancet

Tested

Intensive blood glucose control (sulfonylurea or insulin)

Population

Patients with newly diagnosed type 2 diabetes

Comparator

Conventional dietary therapy

Endpoint

Diabetes-related endpoints (death or complications)

Key result: Intensive therapy significantly reduced the risk of microvascular complications by 25%.

2008

ACCORD Trial

n = 10251 · NEJM

Tested

Intensive glycemic control (target HbA1c <6.0%)

Population

T2DM patients with high cardiovascular risk

Comparator

Standard glycemic control (target HbA1c 7.0-7.9%)

Endpoint

Composite of nonfatal myocardial infarction, nonfatal stroke, or death from cardiovascular causes

Key result: Intensive therapy increased the rate of mortality and did not significantly reduce major cardiovascular events.

2008

ADVANCE Trial

n = 11140 · NEJM

Tested

Intensive glucose control (target HbA1c <=6.5%)

Population

Patients with type 2 diabetes at high risk of vascular disease

Comparator

Standard glucose control

Endpoint

Composite of macrovascular and microvascular events

Key result: Intensive glucose control significantly reduced the incidence of major microvascular events, primarily due to a reduction in nephropathy.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis