Annals of Internal Medicine | 2019
In patients 50 years with type 2 diabetes and CV disease or risk factors, oral semaglutide did not increase CV events
Abstract
Question In patients 50 years of age with type 2 diabetes and cardiovascular (CV) disease or risk factors, does oral semaglutide increase risk for CV events? Methods Design Randomized placebo-controlled trial (Peptide Innovation for Early Diabetes Treatment [PIONEER] 6 trial). ClinicalTrials.gov NCT02692716. Allocation Concealed.* Blinding Blinded* (patients, clinicians, and outcome adjudicators). Follow-up period Median 16 months. Setting 214 clinical centers in 21 countries. Patients 3183 adults who had type 2 diabetes (mean age 66 y, 68% men) and were 50 years of age with established CV disease or chronic kidney disease or were 60 years of age with CV risk factors only. Exclusion criteria included New York Heart Association class 4 heart failure; long-term or intermittent hemodialysis or peritoneal dialysis, or severe renal impairment; planned coronary-, carotid-, or peripheral artery revascularization; proliferative retinopathy or maculopathy resulting in active treatment; treatment with a glucagon-like peptide-1 receptor agonist (GLP-1RA), dipeptidyl peptidase 4 inhibitor, or pramlintide in the past 90 days; or myocardial infarction (MI), stroke, or hospitalization for unstable angina or transient ischemic attack in the past 60 days. Intervention Oral semaglutide, target dose 14 mg/d, achieved using a dose-escalation schedule (n =1591), or placebo (n =1592). Outcomes Major adverse cardiac events (composite of death from CV or undetermined causes, nonfatal MI, or nonfatal stroke). Secondary outcomes included components of the primary composite outcome, hospitalization for unstable angina, hospitalization for heart failure, and death from any cause. The prespecified noninferiority limit for the primary outcome was 1.8 (upper boundary of the 2-sided 95% CI around the hazard ratio). Patient follow-up >99% (intention-to-treat analysis). Main results The results are in the Table. Conclusion In patients 50 years of age with type 2 diabetes and cardiovascular (CV) disease or risk factors, oral semaglutide did not increase risk for CV events compared with placebo. Oral semaglutide vs placebo in adults 50 y of age with type 2 diabetes and cardiovascular (CV) disease or risk factors Outcomes Events/100 person-y At a median 16 mo Oral semaglutide Placebo RRR (95% CI) NNT (CI) Major adverse cardiac events 2.9 3.7 21% (11 to 42) Not significant Death from CV or undetermined causes 0.7 1.4 51% (8 to 73) 104 (73 to 664) Nonfatal stroke 0.6 0.8 26% (57 to 65) Not significant Death from any cause 1.1 2.2 49% (16 to 69) 74 (52 to 226) Hospitalization for heart failure 1.0 1.2 14% (54 to 52) Not significant RRI (CI) NNH Nonfatal myocardial infarction 1.8 1.5 18% (27 to 88) Not significant Hospitalization for unstable angina 0.5 0.3 56% (40 to 299) Not significant Abbreviations defined in Glossary. RRR, RRI, NNT, and CI calculated from placebo event rates and hazard ratios in article. Death from CV or undetermined causes, nonfatal myocardial infarction, or nonfatal stroke. Criterion for noninferiority was met because the upper limit of the 2-sided 95% CI of the hazard ratio (0.79, 95% CI 0.57 to 1.11) was <1.8. Commentary The REWIND, PIONEER 6, and PIONEER 4 trials provide evidence for benefits of GLP-1RAs for treating type 2 diabetes. GLP-1RAs have several benefits beyond glucose lowering, such as weight loss and negligible hypoglycemia risk, which make them an attractive treatment option for patients with type 2 diabetes. With results from REWIND, we now have evidence that 3 GLP-1RAs on the market are also associated with CV benefit in patients with, or at high risk for, CV disease (1, 2). Liraglutide was the first GLP-1RA shown to reduce risk for CV events in a population of whom the majority had CV disease and high CV risk (4-y placebo event rate 15%) (1). A subsequent trial showed that subcutaneous semaglutide reduced risk for CV events in a lower-risk population (4-y placebo event rate 8.9%) (2), and its weekly administration offered greater convenience than daily liraglutide for some patients. The REWIND trial showed that another weekly GLP-1RA, dulaglutide, reduced risk for major CV events in patients at substantially lower CV risk than in other trialsonly 32% of patients had CV disease, and the 5-year placebo event rate was 2.7%. These findings suggest that the CV benefits of GLP-1RAs probably extend to patients with type 2 diabetes and CV risk factors, rather than only patients with CV disease. As with the CV benefits of some SGLT2 inhibitors (3, 4), the CV benefits of GLP-1RAs are probably not a result of glucose lowering alone (5). Both GLP-1RAs and SGLT2 inhibitors have beneficial effects on weight and blood pressure, and GLP-1RAs may also improve endothelial function, endothelial cell responses to ischemia, and platelet function (6). All approved GLP-1RAs are administered by subcutaneous injection, which, along with cost and gastrointestinal side effects, remains a barrier to their use. However, evidence is accumulating for the safety and efficacy of oral semaglutide, the first orally administered GLP-1RA. It has been shown to be superior to the commonly used oral medication, sitagliptin, for HbA1c and weight reduction (7). Results of PIONEER 6 provide evidence of the CV safety of oral semaglutide. Patients who received oral semaglutide had nonsignificantly lower rates of CV events and death than placebo-treated patients, with relative risk reductions comparable to those of REWIND and other GLP-1RA trials (1, 2). Although noninferiority was confirmed, statistically significant CV superiority was not shown. Because the trial was designed to test for noninferiority, its short duration (16 mo) and low number of events probably limited the power to determine a CV benefit of oral semaglutide. PIONEER 4 was the first trial to directly compare oral semaglutide with another GLP-1RA. Compared with liraglutide, the most commonly used GLP-1RA, oral semaglutide showed a similar reduction in HbA1c and, surprisingly, greater weight loss. Gastrointestinal side effects and discontinuation rates were similar between the 2 treatments, making oral semaglutideonce it is approveda comparably effective and tolerable GLP-1RA option for patients wanting to avoid injections. That it must be taken daily on an empty stomach with half a glass of water 30 minutes before food intake may be cumbersome for some patients, who may favor the convenience of weekly subcutaneous semaglutide or dulaglutide. Patients at higher CV risk may also prefer the weekly options due to their proven CV benefit. The results of these trials add to the growing evidence regarding GLP-1RA treatment to help guide treatment decisions. These results further endorse the CV benefits of GLP-1RAs in lower-risk populations and provide support for oral semaglutide as a promising alternative to current subcutaneously administered agents. Whether oral semaglutide has comparable CV-lowering benefits to subcutaneous GLP-1RAs remains to be confirmed.