EMERGING THERAPIES: DRUGS AND REGIMENS

1672 Diabetes Care Volume 41, August 2018

Robert Ritzel,1 Stewart B. Harris,2

A Randomized Controlled Trial Helen Baron,3 Hermes Florez,4
Ronan Roussel,5 Melanie Espinasse,6
Comparing Efficacy and Safety of Isabel Muehlen-Bartmer,7
Nianxian Zhang,8 Monica Bertolini,6
Insulin Glargine 300 Units/mL Claire Brulle-Wohlhueter,6
Medha Munshi,9 and Geremia B. Bolli10
Versus 100 Units/mL in Older
People With Type 2 Diabetes:
Results From the SENIOR Study
Diabetes Care 2018;41:1672–1680 | https://doi.org/10.2337/dc18-0168

1
Klinikum Schwabing and Klinikum Bogenhausen,
OBJECTIVE Städtisches Klinikum München GmbH, Munich,
Germany
SENIOR compared the efficacy and safety of insulin glargine 300 units/mL (Gla-300) 2
The University of Western Ontario, London,
with glargine 100 units/mL (Gla-100) in older people (‡65 years old) with type 2 Ontario, Canada
3
diabetes. University of Southern California, Los Angeles,
CA
4
RESEARCH DESIGN AND METHODS Miller School of Medicine, University of Miami,
Coral Gables, FL
SENIOR was an open-label, two-arm, parallel-group, multicenter phase 3b trial 5
Bichat Hospital, Paris, France
designed to enroll ∼20% of participants aged ‡75 years. Participants were ran- 6
Sanofi, Paris, France
7
domized 1:1 to Gla-300 or Gla-100, titrated to a fasting self-monitored plasma Sanofi Deutschland GmbH, Frankfurt am Main,
glucose of 5.0–7.2 mmol/L (90–130 mg/dL). Germany
8
Sanofi R&D China, Beijing, China
9
RESULTS Joslin Diabetes Center, Harvard University,
Boston, MA
In total, 1,014 participants were randomized (mean age: 71 years). Comparable 10
Department of Medicine, Perugia University
reductions in HbA1c were observed from baseline to week 26 for Gla-300 (20.89%) Medical School, Perugia, Italy
and Gla-100 (20.91%) in the overall population (least squares mean difference: Corresponding author: Robert Ritzel, robert.ritzel@
0.02% [95% CI 20.092 to 0.129]) and for participants aged ‡75 years (20.11% klinikum-muenchen.de.
[20.330 to 0.106]). Incidence and rates of confirmed (£3.9 mmol/L [£70 mg/dL]) or Received 23 January 2018 and accepted 7 May
severe hypoglycemia events were low and similar between both treatment groups, 2018.
with lower rates of documented symptomatic hypoglycemia with Gla-300. The Clinical trial reg. no. NCT02320721, clinicaltrials
.gov.
lower risk of hypoglycemia with Gla-300 versus Gla-100 was more apparent in
This article contains Supplementary Data online
the subgroup aged ‡75 years versus the overall population. Significantly lower at http://care.diabetesjournals.org/lookup/
annualized rates of documented symptomatic (£3.9 mmol/L [£70 mg/dL]) hypo- suppl/doi:10.2337/dc18-0168/-/DC1.
glycemia were observed (Gla-300: 1.12; Gla-100: 2.71; rate ratio: 0.45 [95% CI This article is featured in a podcast available at
0.25–0.83]). http://www.diabetesjournals.org/content/diabetes-
core-update-podcasts.
CONCLUSIONS © 2018 by the American Diabetes Association.
Efficacy and safety of Gla-300 was demonstrated in older people (‡65 years of age) Readers may use this article as long as the work
is properly cited, the use is educational and not
with type 2 diabetes, with comparable reductions in HbA1c and similarly low or
for profit, and the work is not altered. More infor-
lower risk of documented symptomatic hypoglycemia versus Gla-100. A significant mation is available at http://www.diabetesjournals
benefit in hypoglycemia reduction was seen in participants aged ‡75 years. .org/content/license.
care.diabetesjournals.org Ritzel and Associates 1673

Diabetes is common in older adults, The basal insulin glargine is available and glycated hemoglobin (HbA1c) at
affecting an estimated 20% of people as insulin glargine 100 units/mL (Gla- screening of 7.5–11% (58–97 mmol/mol)
65–79 years of age (94 million people) 100; Sanofi, Paris, France) or insulin in insulin-naive participants or 7–10%
worldwide in 2015 (1). Prevalence of glargine 300 units/mL (Gla-300; Sanofi). (53–86 mmol/mol) in insulin-pretreated
type 2 diabetes increases with age owing Compared with Gla-100, Gla-300 provides participants. Exclusion criteria included
to attenuated b-cell function and insulin more stable and prolonged steady- the chronic use of short-acting insulin,
resistance (2). Coupled with the growing state pharmacokinetic and pharmaco- participants not on a stable basal insulin
proportion of older individuals in the global dynamic profiles, with blood insulin dose, and cognitive disorder and de-
population (3) is an increasing requirement levels that more closely resemble nor- mentia (Mini-Mental State Examination
for effective diabetes management for mal physiological conditions (18,19). score ,24). A list of key exclusion criteria
older people. The EDITION 1, 2, and 3 trials compared for the study is provided in Supplemen-
Therapies for lowering blood glucose the efficacy and safety of Gla-300 with tary Table 1.
include several classes of oral antihyper- Gla-100 in people with type 2 diabetes The study was approved by indepen-
glycemic drugs as well as injectables, and demonstrated comparable glycemic dent ethics review boards and conducted
such as rapid-acting and basal insulins control over 6 months, which was better according to Good Clinical Practice and
and glucagon-like peptide receptor ago- sustained with Gla-300 throughout the the Declaration of Helsinki. All participants
nists. When insulin is used, a fine balance 12-month study period, and a significantly provided written informed consent.
is required between achieving glycemic lower risk of confirmed (#3.9 mmol/L
control and avoiding hypoglycemia. Com- [#70 mg/dL]) or severe hypoglycemia Randomization and Treatment
pared with younger populations, older (definitions based on American Diabe- Participants were randomized 1:1 to re-
people with diabetes more often require tes Association [ADA] criteria [20]) at ceive Gla-300 or Gla-100. Randomization
insulin and are more prone to hypogly- night (0000–0559 h) and at any time was stratified by HbA1c at screening
cemia for multiple reasons such as erratic of day (24 h) with Gla-300 compared (,8.0/$8.0% [,64/$64 mmol/mol]),
food ingestion, insufficient adjustment with Gla-100. A recent post hoc analysis previous use of insulin (naive/pretreated),
of insulin dose, reduced responses to of participants $65 years of age with and use of sulfonylurea or meglitinide
counterregulatory hormones, lower blood type 2 diabetes from EDITION 1–3 con- at screening (yes/no).
glucose threshold for autonomic symp- firmed the comparable glycemic con- Basal insulins were self-administered
toms, and higher blood glucose threshold trol and reduction in hypoglycemia risk once daily. Evening administration was
for cognitive dysfunction. The latter leads for Gla-300 versus Gla-100 in this sub- recommended, although other times of
to impaired awareness of hypoglycemia group, with significantly lower rates of day were permissible provided adminis-
and an increased risk of severe hypo- nocturnal (0000–0559 h) hypoglycemia tration occurred at the same time each
glycemia (4–7). Cognitive impairment in (21). day 63 h. Starting doses were deter-
older adults may itself increase the risk The SENIOR study was the first pro- mined by previous dose history: insulin-
of hypoglycemia owing to difficulties in spectively designed clinical trial to ad- naive patients started on a dose of 0.2
diabetes management (8). Moreover, dress the efficacy and safety of basal units/kg, and participants pretreated
the frequent asymptomatic hypoglyce- insulin (insulin glargine) specifically in with basal insulin received a starting
mic episodes that may result from im- older people ($65 years of age) with dose equivalent to their median basal
paired awareness can further lower the type 2 diabetes. The study was also insulin dose of the 3 days before baseline,
threshold for autonomic symptoms and designed such that ;20% of the people unless receiving NPH insulin more than
lead to further hypoglycemic episodes enrolled would be $75 years of age to once daily, in which case a dose 20%
(9). explore Gla-300 versus Gla-100 treat- lower than the total daily dose was given.
Hypoglycemic events in older people ment in this population at elevated risk The insulin dose was adjusted every
are associated with an increased inci- of hypoglycemia (6) and its consequences 3–4 days to achieve a target fasting self-
dence of acute cardiovascular events, (4,6,10–15). monitored plasma glucose (SMPG) level
impaired cognitive function, dementia, of 5.0–7.2 mmol/L (90–130 mg/dL), the
hospitalizations, and mortality (4,6,10–15). RESEARCH DESIGN AND METHODS ADA-recommended target for healthy
Hypoglycemia and its consequences are Study Design older individuals (8). Dose adjustments
an even greater burden in people $75 SENIOR (NCT02320721) was a multina- were as follows: an increase of 3 units
years of age than in those 65–74 years tional, multicenter, phase 3b, active- if .7.2 mmol/L (.130 mg/dL) or a de-
of age, with hospitalization rates two- controlled, randomized, open-label, crease of 3 units (or an adjustment at
fold higher (16). Furthermore, the care two-arm parallel-group study in older the investigators’ discretion) if ,5.0
of older people is complicated by the people ($65 years of age) with type 2 mmol/L (,90 mg/dL) or in the event of
diverse range of functional ability, cog- diabetes comprising a 4-week screen- two or more symptomatic or one severe
nitive function, comorbid conditions, and ing period, followed by a 26-week treat- hypoglycemic episode in the preceding
frailty present in this group (8). Despite ment period, conducted in 162 centers week.
these considerations, there are few long- across 18 countries. Participants continued their previous
term studies in this population demon- Key inclusion criteria were age $65 antihyperglycemic drugs, if approved for
strating the possible benefits of better years, type 2 diabetes for $1 year treated use with insulin, with the exception of
glycemic, blood pressure, and lipid con- with a pharmacologic antihyperglycemic thiazolidinediones, which were stopped
trol (17). regimen for $8 weeks before screening, at randomization.
1674 Gla-300 in Older People With Type 2 Diabetes Diabetes Care Volume 41, August 2018

Figure 1—Mean (SE) HbA1c in the overall population (A) and in participants $75 years of age (B) and mean (SE) FPG in the overall population (C) and
in participants $75 years of age (D) by visit during the 26-week (W) treatment period (intent-to-treat population). BL, baseline; LS, least squares.

Outcomes experiencing hypoglycemic events and definitions (20). Data are presented for
The primary efficacy end point was annualized rates of hypoglycemia at documented symptomatic hypoglycemia,
change in HbA1c from baseline to week either threshold (#3.9 mmol/L [#70 severe hypoglycemia, and as a combina-
26. The main secondary efficacy end points mg/dL] and ,3.0 mmol/L [,54 mg/dL]) tion of confirmed or severe hypoglyce-
were the percentage of participants at any time of day [24 h] and at night mic events.
with one or more confirmed (#3.9 [0000–0559 h] over 26 weeks of treat- Participant-reported outcomes (PROs)
mmol/L [#70 mg/dL]) or severe hypo- ment), and the percentage of partici- were assessed for mobility scores taken
glycemic events occurring at any time of pants achieving HbA1c ,7.5 and ,7.0% from the EQ-5D, the EuroQol Research
day (24 h) or at night (0000–0559 h or (58 and 53 mmol/mol) at the end of Foundation’s questionnaire that assesses
2200–0859 h) over 26 weeks of treat- 26 weeks of treatment. A composite five dimensions of healthdmobility, self-
ment. Other secondary end points were end point of HbA1c target achievement care, usual activities, pain/discomfort, and
assessed in the overall population and (,7.5 or ,7.0% [58 and 53 mmol/ anxiety/depressiondas a single utility in-
in a subgroup of participants $75 years mol]) without confirmed (#3.9 mmol/L dex and on a visual analog scale. Safety
of age. These included change in fasting [#70 mg/dL]) or severe hypoglycemia outcomes included treatment-emergent
plasma glucose (FPG) from baseline to was also assessed. Hypoglycemia was adverse events (TEAEs) and severe ad-
week 26, the percentage of participants defined based on the ADA workgroup verse events (SAEs).
care.diabetesjournals.org Ritzel and Associates 1675

Data Analysis and Statistics diabetes duration (,10/$10 years), his- Change in HbA1c in Participants ‡75 Years
A sample size of 460 randomized partic- tory of diabetic neuropathy (yes/no), of Age
ipants for each treatment group provided and estimated glomerular filtration rate Within the subpopulation of participants
98% power to show noninferiority of Gla- categories ($60, 30 to ,60, or ,30 mL/ $75 years old, similar reductions in HbA1c
300 versus Gla-100 in HbA1c change from min/1.73 m2). Mean HbA1c and hypogly- from baseline to week 26 were observed,
baseline to week 26 on the basis of a true cemia (incidence and rates) by mobility from 8.17% (SD 0.89) to 7.29% (0.84)
difference between the two groups of PRO score subgroup (“no problems”/“slight (65.9 [9.9] to 56.2 [9.2] mmol/mol) for
zero and a noninferiority margin of 0.3%. to extreme problems”) were also per- Gla-300 and 8.18% (0.97) to 7.39% (0.87)
This calculation assumes a common SD formed as exploratory analyses. HbA1c (65.9 [10.64] to 57.3 [9.5] mmol/mol) for
of 1.1%, with a one-sided t test at the by baseline mobility PRO score subgroup Gla-100 (Fig. 1B).
2.5% significance level. Change in HbA1c was assessed using a similar approach to
Plasma Glucose
(primary end point) and FPG from baseline the primary analysis. Analyses of hypo-
Comparable reductions in mean FPG
to week 26 were assessed using a multiple glycemia by baseline category subgroups
were observed for both treatment groups
imputation approach for missing data were descriptive. Safety outcomes in-
from baseline to week 26 in the overall
and ANCOVA using fixed categorical ef- cluded TEAEs and SAEs.
population (Fig. 1C), with reductions of
fects of treatment group, randomization
1.68 (SD 0.12) mmol/L and 1.77 (0.14)
strata, and continuous fixed covariates RESULTS
mmol/L (30.4 [2.2] and 31.8 [2.4] mg/dL,
of baseline value. Least squared mean Study Participants respectively). Similar results were ob-
and least squared mean differences were A total of 1,014 participants were ran- served in participants $75 years of age
combined using the Rubin formula. domized to receive Gla-300 (n = 508) or (Fig. 1D).
The main secondary end point and Gla-100 (n = 506) (Supplementary Fig. 1); Reduction in average 5-point SMPG
HbA1c target achievement, with or with- of these, 135 (26.6%) and 106 (20.9%) and bedtime SMPG was comparable be-
out hypoglycemia, was analyzed using participants were $75 years of age tween treatment groups in the overall
the Cochran–Mantel–Haenszel method for Gla-300 and Gla-100, respectively. population and in participants $75 years
with treatment group as a factor and Overall, baseline characteristics were of age (data not shown).
stratified on randomization strata. Non- similar for the Gla-300 and Gla-100
inferiority and, subsequently, superior- groups (Supplementary Table 3), al- Hypoglycemia
ity of Gla-300 compared with Gla-100 though participants $75 years of age Confirmed (£3.9 mmol/L [£70 mg/dL]) or
was assessed for the primary and main had a lower mean estimated glomer- Severe Hypoglycemia
secondary end points using a hierarchical ular filtration rate, longer duration of Superiority of Gla-300 versus Gla-100 for
step-down testing procedure (Supple- diabetes, a higher overall rate of diabetes- the main secondary efficacy end points,
mentary Table 2), with two-sided tests related complications, and a smaller pro- the proportion of participants in the
for superiority performed at the level portion were previously treated with intent-to-treat population experiencing
of a = 0.05. Analyses of safety outcomes insulin compared with the overall pop- one or more confirmed (#3.9 mmol/L
were descriptive. Analyses were per- ulation (Supplementary Table 3). [#70 mg/dL]) or severe hypoglycemic
formed for the overall population and
events occurring during either 2200–
for the subgroup of participants $75 Glycemic Control
years of age. Change in HbA1c
0859 h or 0000–0559 h, was not detected
Exploratory analyses of the percent- The primary objective of noninferiority (Table 1).
age of patients experiencing one or of Gla-300 versus Gla-100 for change in For the safety population, the pro-
more confirmed (#3.9 mmol/L [#70 HbA1c from baseline to week 26 was portion of participants experiencing
mg/dL]) or severe hypoglycemic events met (Fig. 1A). Mean (SD) HbA1c decreased one or more confirmed or severe hypogly-
in the overall population were analyzed similarly in both treatment groups, from cemic events was similar for both treat-
by prespecified baseline categories of 8.20% (0.91) for Gla-300 and 8.22% (0.92) ment groups at both the #3.9 mmol/L
age (65–70, 70–75, or $75 years), BMI for Gla-100 (66.1 [9.9] vs. 66.3 [10.1] (#70 mg/dL) and ,3.0 mmol/L (,54
(,30/$30 kg/m2), HbA1c (,8.0/$8.0% mmol/mol) at baseline to 7.31% (0.91) mg/dL) thresholds at any time of day
[,64/$64 mmol/mol]), sulfonylurea or and 7.28% (0.84) (56.4 [9.5] vs. 56.0 [9.2] (24 h) (Fig. 2A). Similar results were ob-
meglitinide use at screening (yes/no), mmol/mol), respectively. served at night (0000–0559 h) (Fig. 2A).

Table 1—Percentage of participants experiencing one or more confirmed or severe hypoglycemic events over 26 weeks of
treatment (intent-to-treat population)
% participants RR Gla-300 vs. Gla-100a
Confirmed or severe hypoglycemia Gla-300 (n = 508) Gla-100 (n = 506) RR 95% CI P value (CMH)
2200–0859 h 48.3 47.7 1.01 0.890–1.153 0.8415
0000–0559 h 20.2 22.5 0.90 0.706–1.140 d
Any time of the day (24 h) 59.4 62.7 0.95 0.859–1.046 d
Confirmed hypoglycemia defined as #3.9 mmol/L (#70 mg/dL). Relative risk (RR) stratified by randomization strata of screening HbA1c (,8.0/$8.0%),
a

randomization strata of previous use of insulin (naive/pretreated), randomization strata of use of sulfonylurea or meglitinides at screening (yes/no),
using a Cochran–Mantel–Haenszel (CMH) methodology.
1676 Gla-300 in Older People With Type 2 Diabetes Diabetes Care Volume 41, August 2018

Annualized event rates for confirmed A trend toward lower annualized event Within the subgroup of participants
(#3.9 mmol/L [#70 mg/dL]) or severe rates of hypoglycemia with Gla-300 versus $75 years of age, the incidence of con-
hypoglycemia were similar between Gla-100 was observed at both glyce- firmed or severe hypoglycemic events for
treatment groups, at night (0000–0559 h) mic thresholds. These differences were both glycemic thresholds and docu-
or any time of day, but a trend toward statistically significant for anytime (24 h) mented symptomatic hypoglycemia for
lower rates with Gla-300 compared with for the #3.9 mmol/L (#70 mg/dL) and the higher threshold were comparable
Gla-100 was observed at the lower thresh- for the ,3.0 mmol/L (,54 mg/dL) thresh- between treatment groups at any time
old of ,3.0 mmol/L (,54 mg/dL) (Fig. 3A). old (Fig. 3A). of day (24 h). The incidence of any time
Severe Hypoglycemia (24 h) documented symptomatic (,3.0
Documented Symptomatic Hypoglycemia
The incidence and annualized rate of mmol/L [,54 mg/dL]) hypoglycemia
The results reported for documented
severe hypoglycemia were low, with was significantly lower with Gla-300 com-
symptomatic hypoglycemia were in keep-
four participants (0.8%; 0.02 events per pared with Gla-100 (1.5% vs. 10.4%;
ing with those seen for confirmed or
participant-year) in the Gla-300 group and relative risk 0.33; 95% CI 0.12–0.88)
severe hypoglycemia. A comparable pro-
three participants (0.6%; 0.01 events per (Fig. 2B).
portion of participants experienced one
participant-year) in the Gla-100 group re- Statistically significant reductions in
or more events in both treatment groups
porting one event each. annualized rates of hypoglycemia at
at any time of day (24 h) and at night
(0000–0559 h); however, fewer partici- Hypoglycemia in Participants ‡75 Years any time of day (24 h) were observed for
pants experienced a hypoglycemic event of Age Gla-300 compared with Gla-100 for con-
at the lower glycemic threshold of ,3.0 In general, the proportion of participants firmed (,3.0 mmol/L [,54 mg/dL]) or
mmol/L (,54 mg/dL) with Gla-300 com- $75 years of age experiencing one or severe hypoglycemia and documented
pared with Gla-100, although this was more hypoglycemic events was similar to symptomatic hypoglycemia at either
not significant (Fig. 2A). the overall SENIOR population. glycemic threshold (Fig. 3B). Nocturnal

Figure 2—Incidence (number and percentage of participants experiencing one or more hypoglycemic events) for the overall population ($65 years
of age) (A) and for participants $75 years of age (safety population) (B). RR, relative risk.
care.diabetesjournals.org Ritzel and Associates 1677

Figure 3—Annualized rates (events per participant-year) of hypoglycemia for the overall population ($65 years of age) (A) and for participants
$75 years of age (safety population) (B).

(0000–0559 h) incidence and rates were HbA1c Target Achievement and Efficacy and Safety by Mobility PRO
comparable between treatment groups Composite End Point Score
for both confirmed and severe and for do- The proportion of participants achiev- The majority of participants (;60%) in
cumented symptomatic hypoglycemia at ing HbA1c ,7.5% (,58 mmol/mol) and both treatment groups reported no mo-
the #3.9 mmol/L (#70 mg/dL) threshold ,7.0% (,53 mmol/mol) was compara- bility problems, and few participants
(Fig. 3B). ble between treatment groups for the (,1.0%) reported extreme problems
No severe hypoglycemic events were overall population and for participants at baseline. No difference in mean
observed in participants $75 years of $75 years of age (Supplementary Table HbA1c change from baseline to week
age. 4). Similar results were observed for 26 was observed between participants
the composite end points of HbA1c tar- who reported no mobility problems at
Hypoglycemia Incidence by Baseline get achievement without confirmed (#3.9 baseline and those who reported slight
Characteristics mmol/L [#70 mg/dL]) or severe hypo- to extreme problems (Supplementary Fig.
The results observed in the overall pop- glycemia, although significantly more 4). No differences were observed in
ulation were consistent across the base- participants in the $75 years age group the incidence or rates of hypoglycemia
line subgroups, with a similar percentage achieved HbA1c ,7.5% (,58 mmol/mol) between treatment groups at any time
of participants experiencing confirmed or without confirmed (#3.9 mmol/L [#70 of day or at night for participants who
severe hypoglycemia and documented mg/dL]) or severe hypoglycemia with reported mobility problems at baseline
symptomatic hypoglycemia (#3.9 mmol/L Gla-300 versus Gla-100. and for those who did not (Supplementary
[#70 mg/dL]) with Gla-300 compared with The incidence of confirmed or severe Fig. 5).
Gla-100 irrespective of age, BMI, HbA1c, hypoglycemic events and documented
previous insulin use, sulfonylurea or me- symptomatic events (#3.9 mmol/L [#70 Adverse Events
glitinide use, diabetes duration, nutri- mg/dL]) was similar between treatment TEAEs were reported in 299 partici-
tional status, diabetic neuropathy, or groups, irrespective of HbA1c level achieved pants (58.9%) in the Gla-300 group
kidney function (Supplementary Fig. 2). at week 26 (Supplementary Fig. 3). and in 304 (60.2%) in the Gla-100 group
1678 Gla-300 in Older People With Type 2 Diabetes Diabetes Care Volume 41, August 2018

(Supplementary Table 5), with infections years of age compared with the overall used in the EDITION trials (22–25) or the
(25.8% and 29.7% for Gla-300 and Gla- SENIOR population (predominantly age- less frequent SMPG monitoring required
100, respectively) the most commonly related changes that may be expected), for SENIOR, which, coupled with the
reported TEAEs. The incidence of TEAEs whether these differences contributed to lower perception of hypoglycemia
in participants $75 years of age was the more pronounced hypoglycemia symptoms in older individuals (4,5),
similar to the overall $65 years popula- benefit with Gla-300 versus Gla-100 in may have allowed for undetected hypo-
tion (59.3% for Gla-300 and 54.7% for Gla- the $75 years population is unknown glycemic events to have occurred. Im-
100) (Supplementary Table 5). because this is outside the remit of the paired awareness of hypoglycemia in
present study. older people ($65 years of age) may
CONCLUSIONS When hypoglycemia incidence was also explain the lack of nocturnal hypo-
SENIOR was the first prospective study assessed according to baseline charac- glycemia risk reduction with Gla-300
designed to evaluate the efficacy and teristics (age, BMI, HbA1c, previous in- versus Gla-100 in the present study, de-
safety of basal insulin in older people sulin use, sulfonylurea or meglitinide use, spite this being a consistent finding re-
with type 2 diabetes, with ;20% of diabetes duration, nutritional status, di- ported in all EDITION studies (22–25). For
participants $75 years of age (4). Re- abetic neuropathy, or kidney function), example, a study using continuous glucose
sults of SENIOR demonstrate that the no differences were observed in hypo- monitoring to examine hypoglycemia in a
efficacy and safety profile of Gla-300 glycemia risk between treatment groups population aged $69 years old reported
and Gla-100 observed in the EDITION for any subgroup. In addition, hypo- that 93% of events were unrecognized
studies (22–25) is also apparent in older glycemia risk profiles did not differ (26). Given that the risk of hypoglycemia
adults ($65 years of age) with type 2 between participants who reported mo- in older people is likely to be underesti-
diabetes. The incidence of TEAEs was bility problems and those who did not. mated, the trend toward a greater re-
similar between treatment groups for the Hypoglycemia is associated with mor- duction in hypoglycemic risk observed
overall population and for participants bidities that may lead to physical dys- with Gla-300 versus Gla-100, which was
$75 years of age (54.7–60.2%) and function, and this can create a cyclical consistent with those observed in the
was consistent with that observed for relationship between hypoglycemia and overall EDITION population (27,28) and
the older population of the EDITION increased frailty (4). As such, it is reas- in the older population of the EDITION
studies (21). suring to note that the hypoglycemia risk studies (21), may be beneficial in older
Reductions in HbA1c were comparable profiles for both insulins did not differ people and will be the focus of future real-
between treatment groups, with .50% between participants who experienced world investigations.
of participants in either age group mobility problems and those who did In keeping with ADA recommendations
achieving HbA 1c of ,7.5%. Reduc- not. Furthermore, good glycemic con- for older people (8), the SENIOR study
tions in FPG were also comparable be- trol was demonstrated in both groups, used a titration algorithm with a higher
tween treatment groups and between suggesting that the observed comparable glycemic target of 5.0–7.2 mmol/L
age groups, indicating that insulin was hypoglycemia profiles were not due to (90–130 mg/dL) compared with 4.4–
adjusted similarly in the overall pop- hypoglycemia avoidance through poor 7.2 mmol/L (80–130 mg/dL) recommen-
ulation and in participants $75 years glycemic control. ded for the overall population, owing
of age. The small differences in hypoglyce- to the diversity of cognitive function,
A consistent trend for reduced hy- mia risk reductions observed between frailty, comorbidities, polypharmacy,
poglycemia for Gla-300 versus Gla-100 Gla-300 and Gla-100 in the overall popula- and pathophysiology observed in older
was observed in the overall SENIOR tion in SENIOR may be due, in part, to adults. Despite this, similar reductions in
population, with statistically significant the low number of hypoglycemic events HbA1c and FPG were observed in SENIOR
reductions seen for rates of documented observed. Participants in EDITION 2, for compared with the overall EDITION popu-
symptomatic hypoglycemia at any time example, experienced a greater number lation with type 2 diabetes (27,28) and
of day (24 h). Interestingly, reductions in of hypoglycemic events compared with the combined older ($65 years) popu-
hypoglycemia for Gla-300 versus Gla-100 SENIOR participants (Supplementary lations of EDITION 1–3 studies (21), where
were more pronounced in individuals Table 6), and significant differences in a lower glycemic target (4.4–7.2 mmol/L
$75 years of age, who are known to be hypoglycemia rates and incidence were [80–130 mg/dL]) was used (22–25).
at a greater risk of hypoglycemia (4), observed between Gla-300 and Gla-100 Furthermore, .50% of participants in
with a significantly lower risk of docu- (25). This hypothesis is supported by SENIOR achieved a target HbA1c of ,7.5%
mented symptomatic hypoglycemia at the comparable low incidence of hy- (,58 mmol/mol), similar to HbA1c tar-
any time of day (24 h) observed with Gla- poglycemia in SENIOR and EDITION 3 get achievement observed in partici-
300 compared with Gla-100 in this sub- (Supplementary Table 6), with similarly pants $65 years of age from EDITION
group. This is an important finding in an few significant differences in hypogly- 1–3 (21). Hence, the results from SENIOR
understudied and particularly vulnera- cemia risk observed between treatment show that Gla-300 can be used in older
ble population and might be considered groups for both studies (22,23). Factors people with type 2 diabetes, in accor-
when deciding among basal insulin op- underpinning the low number of hypo- dance with the ADA safety recommen-
tions in older adults, particularly in those glycemic events reported in SENIOR may dations to use higher glycemic targets,
older than 75 years. Although some include the higher glycemic treatment without compromising efficacy.
differences were observed in baseline target set in SENIOR (5.0–7.2 mmol/L Limitations of SENIOR include the
characteristics in the participants $75 [90–130 mg/dL]) compared with that open-label design (owing to the different
care.diabetesjournals.org Ritzel and Associates 1679

pen injectors) and the lack of functional Nordisk, and Sanofi and is on the speakers bu- of dementia in older patients with type 2 dia-
mobility assessments. Although mobility reau for Novo Nordisk and Sanofi. H.F. is a con- betes mellitus. JAMA 2009;301:1565–1572
sultant for Sanofi. R.Ro. is on the advisory panel 12. Geller AI, Shehab N, Lovegrove MC, et al.
was assessed using PROs, the standardized of Amgen, AstraZeneca, Sanofi, Merck Sharp & National estimates of insulin-related hypogly-
EQ-5D questionnaire used may not have Dohme, Eli Lilly, Janssen, Novo Nordisk, Physi- cemia and errors leading to emergency depart-
had sufficient sensitivity to detect any ogenex, and Danone Research and is a consultant ment visits and hospitalizations. JAMA Intern
differences in mobility score during the for Sanofi. M.E., I.M.-B., N.Z., M.B., and C.B.-W. Med 2014;174:678–686
are employees of Sanofi and disclose stocks and 13. Kagansky N, Levy S, Rimon E, et al. Hypoglyce-
course of the study. Owing to the rela-
shares of Sanofi. M.M. is a consultant for Sanofi. mia as a predictor of mortality in hospitalized el-
tively small number of hypoglycemic G.B.B. is on the Sanofi advisory panel, is a derly patients. Arch Intern Med 2003;163:1825–1829
events reported, the power to detect consultant for Novartis, and is on the Eli Lilly 14. Majumdar SR, Hemmelgarn BR, Lin M,
significant differences in hypoglycemia speakers bureau. No other potential conflicts of McBrien K, Manns BJ, Tonelli M. Hypoglycemia
risk was low. In addition, further analyses interest relevant to this article were reported. associated with hospitalization and adverse
Author Contributions. R.Ri., S.B.H., H.B., H.F., events in older people: population-based cohort
are required to explain the lower hypo- R.Ro., M.E., I.M.-B., N.Z., M.B., C.B.-W., M.M., and study. Diabetes Care 2013;36:3585–3590
glycemia risk with Gla-300 compared G.B.B. contributed to interpreting the findings and 15. Stepka M, Rogala H, Czyzyk A. Hypoglycemia:
with Gla-100 observed in participants to writing, reviewing, and editing the manuscript. a major problem in the management of diabetes
aged $75 years. R.Ri., M.E., I.M.-B., and N.Z. developed the in the elderly. Aging (Milano) 1993;5:117–121
In summary, Gla-300 demonstrated initial concept for this analysis. H.B., H.F., and 16. Lipska KJ, Ross JS, Wang Y, et al. National
M.M. were responsible for data acquisition. R.Ri. trends in US hospital admissions for hypergly-
good efficacy and safety in older people is the guarantor of this work and, as such, had cemia and hypoglycemia among Medicare ben-
with type 2 diabetes, particularly in full access to all the data in the study and takes eficiaries, 1999 to 2011. JAMA Intern Med 2014;
those of advanced age ($75 years). In responsibility for the integrity of the data and 174:1116–1124
the overall population, comparable re- the accuracy of the data analysis. 17. Lakey WC, Barnard K, Batch BC, Chiswell K,
ductions in HbA 1c and lower rates of Tasneem A, Green JB. Are current clinical trials
References in diabetes addressing important issues in dia-
documented symptomatic hypoglyce- betes care? Diabetologia 2013;56:1226–1235
mia were observed with Gla-300 versus 1. The International Diabetes Federation. IDF Di-
abetes Atlas. 7th ed. [Internet], 2015. Available 18. Becker RH, Dahmen R, Bergmann K,
Gla-100. Although the risk of hypoglyce- from https://www.idf.org/e-library/epidemiology- Lehmann A, Jax T, Heise T. New insulin glargine
mia was similar in all prespecified base- research/diabetes-atlas/13-diabetes-atlas-seventh- 300 unitszmL21 provides a more even activity
line characteristic subgroups, a greater edition.html. Accessed 8 November 2016 profile and prolonged glycemic control at
2. Helman A, Avrahami D, Klochendler A, et al. steady state compared with insulin glargine
reduction in the rate of documented 100 unitszmL21. Diabetes Care 2015;38:637–643
Effects of ageing and senescence on pancreatic
symptomatic hypoglycemia with Gla- b-cell function. Diabetes Obes Metab 2016; 19. Porcellati F, Lucidi P, Andreoli AM, et al.
300 versus Gla-100 was observed in 18(Suppl. 1):58–62 More physiological circulating insulin and mod-
the subgroup of people $75 years of 3. National Institute on Aging and World Health ulation of hepatic glucose production with in-
Organization. Global health and aging [Internet], sulin glargine U300 vs. U100 in type 1 diabetes.
age. Results of SENIOR confirm those
2011. Available from http://www.who.int/ageing/ Diabetes 2017;66:A269
previously observed in both the overall 20. American Diabetes Association Workgroup
publications/global_health.pdf?ua51. Accessed
and the older population of the EDITION 20 July 2017 on Hypoglycemia. Defining and reporting hy-
1–3 studies (21–23,25), with similar re- 4. Abdelhafiz AH, Rodrı́guez-Ma~ nas L, Morley JE, poglycemia in diabetes: a report from the
ductions in HbA1c observed despite the Sinclair AJ. Hypoglycemia in older people - a less American Diabetes Association Workgroup on
well recognized risk factor for frailty. Aging Dis Hypoglycemia. Diabetes Care 2005;28:1245–1249
less stringent glycemic target used in 21. Yale JF, Aroda VR, Charbonnel B, et al. Older
2015;6:156–167
SENIOR, indicating that Gla-300 is suit- 5. Matyka K, Evans M, Lomas J, Cranston I, people with type 2 diabetes: glycemic control and
able for use in this vulnerable population. Macdonald I, Amiel SA. Altered hierarchy of hypoglycaemic risk with new insulin glargine
protective responses against severe hypoglyce- 300 U/mL. Diabetes 2015;64(Suppl. 1):A261
mia in normal aging in healthy men. Diabetes 22. Bolli GB, Riddle MC, Bergenstal RM,
Care 1997;20:135–141 Wardecki M, Goyeau H, Home PD; EDITION
Acknowledgments. The authors thank the 6. Kirkman MS, Briscoe VJ, Clark N, et al. Di- 3 Study Investigators. Glycaemic control and
study participants, trial staff, and investigators abetes in older adults. Diabetes Care 2012;35: hypoglycaemia with insulin glargine 300U/mL
for their participation. Editorial and writing 2650–2664 versus insulin glargine 100U/mL in insulin-naı̈ve
assistance was provided by Hannah Brown of 7. Dhaliwal R, Weinstock RS. Management people with type 2 diabetes: 12-month results
Fishawack Communications. of type 1 diabetes in older adults. Diabetes from the EDITION 3 trial. Diabetes Metab 2017;
Duality of Interest. This study was sponsored by Spectr 2014;27:9–20 43:351–358
Sanofi. Sanofi was responsible for designing and 8. American Diabetes Association. Standards of 23. Bolli GB, Riddle MC, Bergenstal RM, et al.;
coordinating the trial, monitoring clinical sites, Medical Care in Diabetesd2017. Diabetes Care EDITION 3 Study Investigators. New insulin glar-
collecting and managing data, and performing all 2017;40(Suppl. 1):S1–S135 gine 300 U/ml compared with glargine 100 U/ml
statistical analyses. The clinical trial considered in 9. Cryer PE. Mechanisms of hypoglycemia- in insulin-naı̈ve people with type 2 diabetes on
this analysis was sponsored by Sanofi, Paris, France. associated autonomic failure and its compo- oral glucose-lowering drugs: a randomized con-
Sanofi also funded the editorial and writing assis- nent syndromes in diabetes. Diabetes 2005;54: trolled trial (EDITION 3). Diabetes Obes Metab
tance of Fishawack Communications. R.Ri. is a con- 3592–3601 2015;17:386–394
sultant for AstraZeneca, Merck, Novo Nordisk, 10. Sinclair A, Morley JE, Rodriguez-Ma~ nas L, 24. Riddle MC, Bolli GB, Ziemen M, Muehlen-
Sanofi, and Servier and is on the speakers bureau et al. Diabetes mellitus in older people: position Bartmer I, Bizet F, Home PD; EDITION 1 Study
for AstraZeneca, Boehringer Ingelheim, Eli Lilly, statement on behalf of the International Asso- Investigators. New insulin glargine 300 units/mL
Merck, Novartis, Novo Nordisk, and Sanofi. ciation of Gerontology and Geriatrics (IAGG), versus glargine 100 units/mL in people with
S.B.H. is a consultant for Abbott, AstraZeneca, the European Diabetes Working Party for Older type 2 diabetes using basal and mealtime insulin:
Boehringer Ingelheim, Eli Lilly, Janssen, Merck, People (EDWPOP), and the International Task glucose control and hypoglycemia in a 6-month
Novo Nordisk, and Sanofi and receives research Force of Experts in Diabetes. J Am Med Dir Assoc randomized controlled trial (EDITION 1). Diabe-
support from Abbott, AstraZeneca, Boehringer 2012;13:497–502 tes Care 2014;37:2755–2762
Ingelheim, Eli Lilly, Janssen, Novo Nordisk, and 11. Whitmer RA, Karter AJ, Yaffe K, Quesenberry 25. Yki-Järvinen H, Bergenstal R, Ziemen M,
Sanofi. H.B. is a consultant for Intarcia, Novo CP Jr., Selby JV. Hypoglycemic episodes and risk et al.; EDITION 2 Study Investigators. New insulin
1680 Gla-300 in Older People With Type 2 Diabetes Diabetes Care Volume 41, August 2018

glargine 300 units/mL versus glargine 100 units/mL glycemic control. Arch Intern Med 2011;171: 28. Ritzel R, Roussel R, Giaccari A, Vora J, Brulle-
in people with type 2 diabetes using oral agents 362–364 Wohlhueter C, Yki-Jarvinen H. Better glycaemic
and basal insulin: glucose control and hypo- 27. Ritzel R, Roussel R, Bolli GB, et al. Patient- control and less hypoglycaemia with insulin
glycemia in a 6-month randomized controlled level meta-analysis of the EDITION 1, 2 and glargine 300 U/mL vs glargine 100 U/mL:
trial (EDITION 2). Diabetes Care 2014;37:3235– 3 studies: glycaemic control and hypoglycaemia 1-year patient-level meta-analysis of the
3243 with new insulin glargine 300 U/ml versus glar- EDITION clinical studies in people with type 2
26. Munshi MN, Segal AR, Suhl E, et al. Frequent gine 100 U/ml in people with type 2 diabetes. diabetes. Diabetes Obes Metab 2018;20:541–
hypoglycemia among elderly patients with poor Diabetes Obes Metab 2015;17:859–867 548