Real-world study shows benefits of semaglutide in GLP-1RA-naïve adults with T2D

Once-weekly semaglutide significantly reduced HbA_1c and weight in GLP-1RA*-naïve adults with type 2 diabetes (T2D), the retrospective SPARE** study reveals.

“[We] investigated real-world clinical outcomes during the first year of semaglutide availability in Canada, in patients with T2D attending a referral-based specialist practice,” said the researchers. “[Our findings showed] statistically and clinically significant 3- to 6-month reduction in HbA_1c and weight, regardless of semaglutide dose or order of semaglutide therapy. Blood pressure (BP) and lipids also improved, and there was no change in reported incidence of hypoglycaemia.”

At a mean follow-up of 4.9 months, there were significant reductions in HbA_1c (mean, −1.03 percent) and weight (mean, −3.9 kg; p<0.001 for both). [Diabetes Obes Metab 2020;doi:10.1111/dom.14117]

Marked HbA_1c reductions were also seen with semaglutide 1.0 mg vs 0.25–0.5 mg (least squares [LS] mean change, −0.24 percent; p<0.001), among noninsulin vs insulin users (LS mean change, −0.21 percent; p=0.003), and those who had a 6-month vs a 3-month follow-up (LS mean change, −0.18 percent; p<0.01).

The HbA_1c reductions were significant regardless whether semaglutide was added to one, two, or ≥3 OHAs*** (LS mean change, −1.35, −1.18, and −0.76 percent, respectively), or insulin (LS mean change, −0.87 percent; p<0.001 for all). A similar trend was seen for weight reduction (LS mean change, −4.4, −4.2, and −3.4 percent [one, two, and ≥3 OHAs, respectively] and –3.6 percent [insulin]; p<0.001 for all).

Body mass index (−1.3 kg/m²), BP (−4.0 mm Hg [systolic] and −1.5 mm Hg [diastolic]), and lipid profile (−0.4 mmol/L [triglycerides], −0.2 mmol/L [LDL cholesterol], and −0.4 mmol/L [non-HDL cholesterol]; p<0.001 for all) also dropped significantly following initiation of semaglutide.

“Therapies for diabetes that improve not only glycaemia but also weight, BP, and lipids may have additional benefit in the risk reduction of cardiovascular and microvascular complications associated with diabetes,” said the researchers.

Baseline and follow-up rates of hypoglycaemia (at least one incidence weekly) were similar in the overall cohort (4.3 percent vs 4.7 percent; p=0.70), as well as among those on sulphonylurea (2.4 percent vs 3.4 percent; p=0.41) or insulin (7.5 percent vs 8.4 percent; p=0.65).

Nearly 20 percent discontinued semaglutide, which was lower than other real-world assessments of exenatide, dulaglutide, and liraglutide. [Diabetes Obes Metab 2019;21:1-10; Diabetes Obes Metab 2017;19:953-961; Curr Med Res Opin 2018;34:995-1003] “[Our] participants … were attending multidisciplinary specialist clinics operating in a well-resourced public health system, which may have improved persistence,” said the researchers.

Randomized vs real-world

SPARE comprised 937 participants (mean age 57.1 years) who initiated and maintained semaglutide during follow-up. Despite the significant findings, when compared against SUSTAIN 1 – an RCT comparing semaglutide against placebo – the drops in HbA_1c (particularly with the 0.5- and 1.5-mg doses) in SPARE were smaller. [Lancet Diab Endocrinol 2017;5:251-260]

However, SUSTAIN 1 had younger and treatment-naïve participants with shorter disease duration and longer on-treatment time. “[Conversely, our participants had multiple OHAs] and comorbidities. Furthermore, RCTs provide additional healthcare provider support, as well as scheduled dispensing of a therapy [that is] not routinely available to the larger populations being assessed in real-world evaluations,” the researchers pointed out.

Nonetheless, real-world studies are now recognized by the US FDA as vital support for regulatory decision making. [N Engl J Med 2016;375:2293-2297; www.fda.gov/regulatory-information/search-fda-guidance-documents/use-real-world-evidence-support-regulatory-decision-making-medical-devices, accessed September 1, 2020] “Real-world studies have the advantage of including people with a broad age range, different socio-economic backgrounds, multiple comorbidities, and using polytherapy, which may make the results more generalizable to real-world clinical practice,” they said.

However, the lack of randomization and having a select patient cluster from a specialist endocrinology practice may have limited generalizability of the findings, they pointed out. “We only reported outcomes in GLP-1RA-naïve individuals, and it is not known how the results would differ in those who switched from another GLP-1RA to semaglutide.”

Future trials with longer follow-ups are thus warranted to explore the impact of semaglutide in other patient subgroups and compare semaglutide against other diabetes drug classes, they said.