Cronfa - Swansea University Open Access Repository
_____________________________________________________________
This is an author produced version of a paper published in :
New England Journal of Medicine
Cronfa URL for this paper:
http://cronfa.swan.ac.uk/Record/cronfa30129
_____________________________________________________________
Paper:
Marso, S., Bain, S., Consoli, A., Eliaschewitz, F., Jódar, E., Leiter, L., Lingvay, I., Rosenstock, J., Seufert, J., Warren,
M., Woo, V., Hansen, O., Holst, A., Pettersson, J. & Vilsbøll, T. (2016). Semaglutide and Cardiovascular Outcomes in
Patients with Type 2 Diabetes. New England Journal of Medicine
http://dx.doi.org/10.1056/NEJMoa1607141
_____________________________________________________________
This article is brought to you by Swansea University. Any person downloading material is agreeing to abide by the
terms of the repository licence. Authors are personally responsible for adhering to publisher restrictions or conditions.
When uploading content they are required to comply with their publisher agreement and the SHERPA RoMEO
database to judge whether or not it is copyright safe to add this version of the paper to this repository.
http://www.swansea.ac.uk/iss/researchsupport/cronfa-support/
The
new england journal
of
medicine
n engl j med 375;19 nejm.org November 10, 2016
1834
From the Research Medical Center, Kansas
City, MO (S.P.M.); School of Medicine,
Swansea University, Swansea, United King-
dom (S.C.B.); Department of Medicine
and Aging Science and Center of Excel-
lence on Aging and Translational Medi-
cine, G. d’Annunzio University, Chieti-
Pescara, Italy (A.C.); CPClin Research
Center/Hospital Israelita Albert Einstein,
São Paulo (F.G.E.); Hospital Universitario
Quirón Salud Madrid, Facultad de Cien-
cias de la Salud, Universidad Europea de
Madrid, Madrid (E.J.); Li Ka Shing Knowl-
edge Institute and Keenan Research Cen-
tre for Biomedical Science, St. Michael’s
Hospital, University of Toronto, Toronto
(L.A.L.), and the University of Manitoba,
Winnipeg (V.W.) — both in Canada; Uni-
versity of Texas Southwestern Medical
Center (I.L.) and Dallas Diabetes Re-
search Center at Medical City (J.R.) —
both in Dallas; University of Freiburg
Medical Center, Faculty of Medicine, Uni-
versity of Freiburg, Freiburg, Germany
(J.S.); Physicians East, Greenville, NC
(M.L.W.); and Novo Nordisk, Søborg
(O.H., A.G.H., J.P.), and the Center for
Diabetes Research, Gentofte Hospital,
University of Copenhagen, Hellerup (T.V.)
— both in Denmark. Address reprint re-
quests to Dr. Marso at Cardiovascular
Services, HCA Midwest Health, Research
Medical Center, 2316 E. Meyer Blvd., Kansas
City, MO 64132, or at smarso@ gmail . com.
* A complete list of the investigators in the
Trial to Evaluate Cardiovascular and Other
Long-term Outcomes with Semaglutide
in Subjects with Type 2 Diabetes (SUS-
TAIN-6) is provided in the Supplemen-
tary Appendix, available at NEJM.org.
This article was published on September
16, 2016, at NEJM.org.
N Engl J Med 2016;375:1834-44.
DOI: 10.1056/NEJMoa1607141
Copyright © 2016 Massachusetts Medical Society.
BACKGROUND
Regulatory guidance specifies the need to establish cardiovascular safety of new dia-
betes therapies in patients with type 2 diabetes in order to rule out excess cardiovascu-
lar risk. The cardiovascular effects of semaglutide, a glucagon-like peptide 1 analogue
with an extended half-life of approximately 1 week, in type 2 diabetes are unknown.
METHODS
We randomly assigned 3297 patients with type 2 diabetes who were on a standard-
care regimen to receive once-weekly semaglutide (0.5 mg or 1.0 mg) or placebo for
104 weeks. The primary composite outcome was the first occurrence of cardiovascu-
lar death, nonfatal myocardial infarction, or nonfatal stroke. We hypothesized that
semaglutide would be noninferior to placebo for the primary outcome. The non-
inferiority margin was 1.8 for the upper boundary of the 95% confidence interval of
the hazard ratio.
RESULTS
At baseline, 2735 of the patients (83.0%) had established cardiovascular disease,
chronic kidney disease, or both. The primary outcome occurred in 108 of 1648 patients
(6.6%) in the semaglutide group and in 146 of 1649 patients (8.9%) in the placebo
group (hazard ratio, 0.74; 95% confidence interval [CI], 0.58 to 0.95; P<0.001 for non-
inferiority). Nonfatal myocardial infarction occurred in 2.9% of the patients receiving
semaglutide and in 3.9% of those receiving placebo (hazard ratio, 0.74; 95% CI, 0.51
to 1.08; P = 0.12); nonfatal stroke occurred in 1.6% and 2.7%, respectively (hazard ratio,
0.61; 95% CI, 0.38 to 0.99; P = 0.04). Rates of death from cardiovascular causes were
similar in the two groups. Rates of new or worsening nephropathy were lower in the
semaglutide group, but rates of retinopathy complications (vitreous hemorrhage, blind-
ness, or conditions requiring treatment with an intravitreal agent or photocoagulation)
were significantly higher (hazard ratio, 1.76; 95% CI, 1.11 to 2.78; P = 0.02). Fewer seri-
ous adverse events occurred in the semaglutide group, although more patients discon-
tinued treatment because of adverse events, mainly gastrointestinal.
CONCLUSIONS
In patients with type 2 diabetes who were at high cardiovascular risk, the rate of
cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke was signifi-
cantly lower among patients receiving semaglutide than among those receiving pla-
cebo, an outcome that confirmed the noninferiority of semaglutide. (Funded by Novo
Nordisk; SUSTAIN-6 ClinicalTrials.gov number, NCT01720446.)
ABSTR ACT
Semaglutide and Cardiovascular Outcomes
in Patients with Type 2 Diabetes
Steven P. Marso, M.D., Stephen C. Bain, M.D., Agostino Consoli, M.D.,
Freddy G. Eliaschewitz, M.D., Esteban Jódar, M.D., Lawrence A. Leiter, M.D.,
Ildiko Lingvay, M.D., M.P.H., M.S.C.S., Julio Rosenstock, M.D.,
Jochen Seufert, M.D., Ph.D., Mark L. Warren, M.D., Vincent Woo, M.D.,
Oluf Hansen, M.Sc., Anders G. Holst, M.D., Ph.D., Jonas Pettersson, M.D., Ph.D.,
and Tina Vilsbøll, M.D., D.M.Sc., for the SUSTAIN-6 Investigators*
Original A rticle
The New England Journal of Medicine
Downloaded from nejm.org on November 12, 2016. For personal use only. No other uses without permission.
Copyright © 2016 Massachusetts Medical Society. All rights reserved.
n engl j med 375;19 nejm.org November 10, 2016
1835
Semaglutide and Cardiovascular Outcomes in Diabetes
C
ardiovascular disease is the lead-
ing cause of death and complications in
patients with type 2 diabetes.
1
Recently,
trials evaluating a sodium–glucose cotransporter
2 inhibitor (empagliflozin) and a glucagon-like
peptide 1 (GLP-1) analogue (liraglutide) have
shown improved cardiovascular outcomes in pa-
tients with type 2 diabetes who were at high risk
for cardiovascular events.
2,3
Semaglutide, a GLP-1 analogue with an ex-
tended half-life of approximately 1 week (which
permits once-weekly subcutaneous administra-
tion),
4
is currently in development but not yet
approved for the treatment of type 2 diabetes.
Regulatory guidance specifies the need to estab-
lish the cardiovascular safety of new therapies
for type 2 diabetes in order to rule out excess
cardiovascular risk.
5
The preapproval Trial to
Evaluate Cardiovascular and Other Long-term
Outcomes with Semaglutide in Subjects with
Type 2 Diabetes (SUSTAIN-6) was designed to
assess the noninferiority of semaglutide as com-
pared with placebo in terms of cardiovascular
safety in patients with type 2 diabetes.
Methods
Trial Design and Oversight
We performed a randomized, double-blind,
placebo-controlled, parallel-group trial at 230 sites
in 20 countries. The trial protocol, available with
the full text of this article at NEJM.org, was ap-
proved by the institutional review board and
ethics committee at each participating center.
All patients provided written informed consent.
Patients were randomized in a 1:1:1:1 ratio to
receive either 0.5 mg or 1.0 mg of once-weekly
subcutaneous semaglutide or volume-matched
placebo, which maintained blinding within dose.
The trial consisted of a planned observation
period of 109 weeks for all patients (a 104-week
treatment period with a 5-week follow-up period)
in which patients who had prematurely discon-
tinued a study treatment were also included.
The sponsor, Novo Nordisk, designed the
study. Data were gathered by the site investiga-
tors, and the sponsor performed site monitoring,
data collection, and data analysis. An indepen-
dent data and safety monitoring committee per-
formed ongoing surveillance and had access to
all the data in an unblinded fashion.
All the authors had confidential access to the
final trial results and actively contributed to
manuscript preparation. A working group that
included the first and last authors wrote the first
draft of the manuscript, which was revised and
approved by all the authors, who made the deci-
sion to submit the manuscript for publication.
The authors assume responsibility for the accu-
racy and completeness of the data and vouch for
the fidelity of the trial to the protocol. Editorial
support was funded by the sponsor and provided
by independent medical writers under the guid-
ance of the authors.
Patients
Patients with type 2 diabetes and a glycated hemo-
globin level of 7% or more were eligible if they
had not been treated with an antihyperglycemic
drug or had been treated with no more than two
oral antihyperglycemic agents, with or without
basal or premixed insulin. Key inclusion criteria
were an age of 50 years or more with established
cardiovascular disease (previous cardiovascular,
cerebrovascular, or peripheral vascular disease),
chronic heart failure (New York Heart Associa-
tion class II or III), or chronic kidney disease of
stage 3 or higher or an age of 60 years or more
with at least one cardiovascular risk factor (as
defined in Table S1 in the Supplementary Appen-
dix, available at NEJM.org).
Key exclusion criteria included treatment with
a dipeptidyl-peptidase 4 inhibitor within 30 days
before screening or with a GLP-1–receptor ago-
nist or insulin other than basal or premixed
within 90 days before screening; a history of an
acute coronary or cerebrovascular event within
90 days before randomization; planned revascu-
larization of a coronary, carotid, or peripheral
artery; or long-term dialysis. (A complete list of
exclusion criteria is provided in Table S2 in the
Supplementary Appendix.)
Procedures
The randomization of patients was stratified ac-
cording to cardiovascular disease status (estab-
lished cardiovascular or chronic kidney disease
or cardiovascular risk factors only), insulin treat-
ment (none, basal insulin only, or premixed in-
sulin), and estimated glomerular filtration rate
(≤30 ml or >30 ml per minute per 1.73 m
2
of
body-surface area) at screening. A fixed dose-
escalation procedure was used, with a starting
dose of 0.25 mg for 4 weeks that escalated to
0.5 mg for 4 weeks until the maintenance dose
(0.5 mg or 1.0 mg) was reached. No change in
A Quick Take is
available at
NEJM.org
The New England Journal of Medicine
Downloaded from nejm.org on November 12, 2016. For personal use only. No other uses without permission.
Copyright © 2016 Massachusetts Medical Society. All rights reserved.
n engl j med 375;19 nejm.org November 10, 2016
1836
The
new england journal
of
medicine
the maintenance dose of either semaglutide or
placebo was permitted during the treatment
period.
Patients were scheduled for quarterly site visits
during the trial. All investigators were encour-
aged to treat all the patients according to local
guidelines to achieve the most effective glycemic
control (Table S3 in the Supplementary Appen-
dix), and additional noninvestigational antihyper-
glycemic medication (nonincretin-based therapy)
could be added or adjusted.
Outcomes
The primary composite outcome was the first
occurrence of death from cardiovascular causes,
nonfatal myocardial infarction (including silent),
or nonfatal stroke. Prespecified secondary out-
comes included the first occurrence of an ex-
panded composite cardiovascular outcome (death
from cardiovascular causes, nonfatal myocar-
dial infarction, nonfatal stroke, revascularization
[coronary or peripheral], and hospitalization for
unstable angina or heart failure), an additional
composite outcome (death from all causes, non-
fatal myocardial infarction, or nonfatal stroke),
the individual components of the composite out-
comes, retinopathy complications, and new or
worsening nephropathy. Each outcome, except
for peripheral revascularization, was adjudicated
in a blinded fashion by an external, independent
event-adjudication committee.
Continuous efficacy and safety outcomes were
assessed as the change from baseline to week
104. From baseline to week 109, we assessed
serious and nonserious adverse events and hypo-
glycemic episodes, which were defined as severe
(according to American Diabetes Association
criteria
6
) or as confirmed on analysis of plasma
glucose (with symptomatic hypoglycemia defined
as <56 mg per deciliter [3.1 mmol per liter]).
Neoplasm and pancreatitis events were adjudi-
cated. (All definitions of adjudicated events are
provided in Table S4 in the Supplementary Ap-
pendix.)
Statistical Analysis
The prespecified statistical analysis plan is avail-
able with the protocol at NEJM.org. We based
the sample size for the trial on an assumed an-
nual primary-event rate of 1.98% in each group,
a dropout rate of less than 10.0%, a mean in-trial
observation time of 2.1 years, and a true hazard
ratio of 1.00. We determined that the enrollment
of 3260 patients would be required to determine
the primary outcome in at least 122 patients and
provide a power of 90% to reject a hazard ratio
of at least 1.80 at the 0.05 level of significance.
The prespecified analysis for the primary out-
come was a Cox proportional-hazards model, with
pooled treatment (semaglutide vs. placebo) as a
fixed factor, and categorized according to all
possible combinations of stratification factors
used for randomization. The primary hypothesis
was for noninferiority for the primary outcome.
Such noninferiority was confirmed if the upper
boundary of the two-sided 95% confidence inter-
val of the hazard ratio was below the noninferior-
ity margin of 1.80.
5
Testing for superiority for the
primary outcome was not prespecified or adjusted
for multiplicity. We conducted prespecified sen-
sitivity analyses of the primary outcome, using
alternative patient selection and data-censoring
strategies for exposure to treatment, and per-
protocol sensitivity analyses were performed post
hoc. (Details regarding analysis sets are provided
in the Supplementary Appendix.)
The primary outcome was evaluated in sub-
groups according to demographic and disease
measures at baseline. We evaluated the effect of
dose on the primary outcome by repeating the
primary analysis with the four treatment groups
(semaglutide doses of 0.5 mg and 1.0 mg and
corresponding placebo doses) using volume-
matched treatment comparisons. Efficacy and
safety outcome analyses were prespecified to
include the four treatment groups.
All P values are two-sided, with a level of 0.05
considered to indicate statistical significance.
P values other than that for the primary hypoth-
esis have not been adjusted for multiplicity and
have been calculated to test for null hypotheses
of no difference. All results were analyzed on an
intention-to-treat basis that included the full
analysis set (i.e., all patients who underwent
randomization according to the planned treat-
ment), with the exception of adverse events lead-
ing to premature discontinuation, which were
included in the as-treated safety analysis.
Results
Patients
From February 2013 through December 2013, a
total of 4346 patients were screened, and 3297
The New England Journal of Medicine
Downloaded from nejm.org on November 12, 2016. For personal use only. No other uses without permission.
Copyright © 2016 Massachusetts Medical Society. All rights reserved.
n engl j med 375;19 nejm.org November 10, 2016
1837
Semaglutide and Cardiovascular Outcomes in Diabetes
underwent randomization; of these patients, 3232
(98.0%) attended the last follow-up visit at an
investigator site, were contacted by telephone, or
died during the trial. The date of the last patient
visit was March 15, 2016. Vital status was known
for 99.6% of the patients by the end of the trial
(Figs. S1A and S1B in the Supplementary Ap-
pendix).
The median observation time was 2.1 years.
Rates of premature treatment discontinuation
were similar across groups (20% overall) (Table
S5 in the Supplementary Appendix). The mean
percentage of time during which patients re-
ceived semaglutide was 86.5% (87.7% among
those receiving 0.5 mg and 85.3% among those
receiving 1.0 mg), and the mean percentage dur-
ing which patients received placebo was 89.5%
(89.4% among those receiving 0.5 mg and 89.6%
among those receiving 1.0 mg).
Demographic and clinical characteristics of
the patients at baseline were similar across treat-
ment groups (Table 1, and Table S6 in the Supple-
mentary Appendix). Of the 3297 patients, 2735
(83.0%) had established cardiovascular disease
(including chronic kidney disease of stage 3 or
higher), 1940 patients (58.8%) had established
cardiovascular disease without chronic kidney dis-
ease, 353 (10.7%) had chronic kidney disease only,
and 442 (13.4%) had both cardiovascular disease
and kidney disease; 17% of the patients had car-
diovascular risk factors and were 60 years of age
or older. The overall mean duration of type 2
diabetes was 13.9 years, and the mean glycated
hemoglobin level was 8.7%. The use of anti-
hyperglycemic and cardiovascular medications
was well balanced between the groups (Tables
S7A and S8A in the Supplementary Appendix).
Most patients (93.5%) were taking antihyperten-
Characteristic
Semaglutide
(N = 1648)
Placebo
(N = 1649)
Total
(N = 3297)
0.5 mg
(N = 826)
1.0 mg
(N = 822)
0.5 mg
(N = 824)
1.0 mg
(N = 825)
Age — yr 64.6±7.3 64.7±7.1 64.8±7.6 64.4±7.5 64.6±7.4
Male sex — no. (%) 495 (59.9) 518 (63.0) 482 (58.5) 507 (61.5) 2002 (60.7)
Body weight — kg 91.8±20.3 92.9±21.1 91.8±20.3 91.9±20.8 92.1±20.6
Type 2 diabetes
Duration — yr 14.3±8.2 14.1±8.2 14.0±8.5 13.2±7.4 13.9±8.1
Glycated hemoglobin — % 8.7±1.4 8.7±1.5 8.7±1.5 8.7±1.5 8.7±1.5
Cardiovascular risk factors
Systolic blood pressure — mm Hg 136.1±18.0 135.8±17.0 135.8±16.2 134.8±17.5 135.6±17.2
Diastolic blood pressure — mm Hg 77.1±9.8 76.9±10.2 77.5±9.9 76.7±10.2 77.0±10.0
Low-density lipoprotein cholesterol — mg/dl† 81.6±47.1 83.3±41.2 80.9±48.1 83.6±45.9 82.3±45.6
Never smoked — no. (%) 390 (47.2) 364 (44.3) 391 (47.5) 348 (42.2) 1493 (45.3)
History of cardiovascular disease — no. (%)
Ischemic heart disease 493 (59.7) 495 (60.2) 510 (61.9) 496 (60.1) 1994 (60.5)
Myocardial infarction 266 (32.2) 264 (32.1) 267 (32.4) 275 (33.3) 1072 (32.5)
Heart failure 201 (24.3) 180 (21.9) 190 (23.1) 206 (25.0) 777 (23.6)
Ischemic stroke 89 (10.8) 89 (10.8) 96 (11.7) 109 (13.2) 383 (11.6)
Hemorrhagic stroke 28 (3.4) 24 (2.9) 27 (3.3) 29 (3.5) 108 (3.3)
Hypertension 772 (93.5) 771 (93.8) 756 (91.7) 760 (92.1) 3059 (92.8)
* Plus–minus values are means ±SD unless otherwise indicated. Differences in baseline characteristics were assessed with the use of analysis
of covariance for continuous characteristics and logistic regression for categorical characteristics. There were no significant differences be-
tween the groups except for the duration of type 2 diabetes (P = 0.048). To convert the values for cholesterol to millimoles per liter, multiply
by 0.02586.
† Values are geometric means and coefficients of variation.
Table 1. Characteristics of the Patients at Baseline.*
The New England Journal of Medicine
Downloaded from nejm.org on November 12, 2016. For personal use only. No other uses without permission.
Copyright © 2016 Massachusetts Medical Society. All rights reserved.