You are on page 1of 7

532 Diabetes Care Volume 39, April 2016

Anne L. Peters,1 Robert R. Henry,2,3

Diabetic Ketoacidosis With Payal Thakkar,4 Cindy Tong,4 and

Maria Alba4
Canagliflozin, a Sodium–Glucose
Cotransporter 2 Inhibitor, in
Patients With Type 1 Diabetes
Diabetes Care 2016;39:532–538 | DOI: 10.2337/dc15-1995

To assess the incidence of serious adverse events (AEs) of diabetic ketoacidosis
(DKA) with canagliflozin, a sodium–glucose cotransporter 2 inhibitor, as an add-on
to insulin in adults with type 1 diabetes.


In this 18-week, randomized, double-blind, phase 2 study, patients (N = 351; HbA1c
7.0–9.0% [53–75 mmol/mol]) on multiple daily insulin injections or continuous
subcutaneous insulin infusion received canagliflozin 100 or 300 mg or placebo
once daily. The incidence of ketone-related AEs, defined as any event from a
prespecified list of preferred terms (i.e., acidosis, blood ketone body increased,
blood ketone body present, DKA, diabetic ketoacidotic hyperglycemic coma,
ketoacidosis, ketonemia, ketonuria, ketosis, metabolic acidosis, urine ketone
body present), including serious AEs of DKA, was assessed based on AE reports.

At week 18, the incidence of any ketone-related AE with canagliflozin 100 and 300
mg was 5.1% (n = 6 of 117) and 9.4% (n = 11 of 117), respectively; no patients in the
placebo group experienced a ketone-related AE. The incidence of serious AEs of
DKA was 4.3% (n = 5 of 117) with canagliflozin 100 mg and 6.0% (n = 7 of 117) with 1
Keck School of Medicine of the University of
canagliflozin 300 mg; all serious events occurred in the presence of circumstances Southern California, Los Angeles, CA
that are known to potentially precipitate DKA (e.g., infection, insulin pump fail- 2
Center for Metabolic Research, VA San Diego
ure). Among the 12 patients with a serious AE of DKA, blood glucose levels ranged Healthcare System, San Diego, CA
from 9.4 to >44.4 mmol/L (170 to >800 mg/dL). Baseline characteristics were 4
University of California, San Diego, La Jolla, CA
Janssen Research & Development, LLC, Raritan,
generally similar in patients with and without a ketone-related AE. NJ
CONCLUSIONS Corresponding author: Anne L. Peters, momofmax@
Canagliflozin was associated with an increased incidence of serious AEs of DKA in
Received 11 September 2015 and accepted
patients with type 1 diabetes inadequately controlled with insulin. Mitigation 1 February 2016.
strategies are needed for use in future clinical trials to reduce the risk of DKA Clinical trial reg. no. NCT02139943, clinicaltrials
with canagliflozin treatment in patients with type 1 diabetes. .gov.
This article contains Supplementary Data online
Diabetic ketoacidosis (DKA) is a serious and potentially life-threatening complica-
tion of diabetes that is characterized by hyperglycemia, metabolic acidosis, and an
© 2016 by the American Diabetes Association.
increase in circulating ketones (1,2). DKA results from a combination of reduced Readers may use this article as long as the work is
insulin levels and increased levels of regulatory hormones, such as glucagon and properly cited, the use is educational and not for
cortisol (1,2). This decrease in insulin and increase in counterregulatory hormones profit, and the work is not altered. Peters and Associates 533

stimulates the release of free fatty acids of free fatty acids that are converted to exclusion criteria were a history of
from adipose tissue, which are then fur- ketones (19,20). SGLT2 inhibition has type 2 diabetes; DKA or a severe hypo-
ther oxidized into ketone bodies in the also been shown to increase glucagon glycemic event (defined as an event that
liver. The accumulation of ketone bod- secretion, potentially through reduc- required assistance from another per-
ies, primarily b-hydroxybutyrate and tions in insulin dose and a possible son or resulted in seizure or loss of con-
acetoacetate, leads to ketonemia and direct effect of SGLT2 inhibitors on pan- sciousness) within 6 months before
metabolic acidosis (1,3). creatic a-cells; increases in glucagon fur- randomization; myocardial infarction,
DKA occurs predominantly in patients ther alter the insulin-to-glucagon ratio unstable angina, revascularization pro-
with type 1 diabetes but can also be a to stimulate gluconeogenesis and ke- cedure, or cerebrovascular accident
rare complication in those with type 2 tone production (19–22). Reductions in #12 weeks before screening; history
diabetes (1). DKA may also affect pa- the insulin-to-glucagon ratio in patients of New York Heart Association class
tients with latent autoimmune diabetes treated with SGLT2 inhibitors may III–IV cardiac disease; uncontrolled hyper-
of adulthood and those with ketosis- increase susceptibility to DKA in the tension; estimated glomerular filtration
prone diabetes (i.e., patients without presence of risk factors (e.g., pump rate ,70 mL/min/1.73 m2; or treatment
the typical presentation of autoimmune malfunctions, carbohydrate restriction, with an antihyperglycemic agent other
type 1 diabetes) (4,5). Most cases of increased alcohol consumption) that than insulin within 12 weeks before
DKA are associated with precipitating may not otherwise result in DKA (20). screening. Patients were randomly as-
factors, such as infection, intercurrent On the basis of an analysis of the U.S. signed in a 1:1:1 ratio to canagliflozin
illness, surgical stress, and poor compli- T1D Exchange clinic registry, the fre- 100 or 300 mg or placebo once daily
ance with insulin therapy (1,6). Infection quency of DKA is ;5% in patients $26 taken before the first meal of the day.
is the most common precipitating fac- years of age with type 1 diabetes and is Further details on randomization and
tor, occurring in 30–50% of cases of as high as 10% in patients aged 13–26 blinding have been previously reported
DKA (6). years (23); therefore, understanding (15).
Both the U.S. Food and Drug Ad- the risk of DKA associated with SGLT2 The study was conducted in accor-
ministration and European Medicines inhibition is important if these medica- dance with ethical principles that comply
Agency have issued statements warning tions are eventually approved for use in with the Declaration of Helsinki and are
that treatment with sodium–glucose co- patients with type 1 diabetes in clinical consistent with good clinical practices
transporter 2 (SGLT2) inhibitors may be practice. and applicable regulatory requirements.
associated with an increased risk of DKA The efficacy and safety of the SGLT2 The study protocol and amendments
(7,8). SGLT2 inhibition lowers blood glu- inhibitor canagliflozin was assessed in a were approved by institutional review
cose through an insulin-independent phase 2 study as an add-on to insulin in boards and independent ethics com-
mechanism of action by lowering the re- patients with type 1 diabetes; details of mittees at participating institutions.
nal threshold for glucose and increasing the study have been previously reported All patients provided written informed
urinary glucose excretion (UGE), which (15). Briefly, canagliflozin provided consent before participation.
results in mild osmotic diuresis and net reductions in HbA1c, body weight, and
caloric loss (9–11). SGLT2 inhibitors are insulin dose, with no increase in hypogly- Insulin Therapy
approved to treat type 2 diabetes, yet cemia over 18 weeks. This article de- Before randomization, patients with
initial studies indicate that they may im- scribes the incidence of serious adverse HbA 1 c #8.0% (#64 mmol/mol) at
prove glycemic control in patients with events (AEs) of DKA with canagliflozin in screening were recommended, at the
type 1 diabetes as an adjunct to insulin patients with type 1 diabetes based on discretion of the investigator, to reduce
therapy (12–17). Rare, but serious, cases data from that study. their basal insulin dose by 20%, and pa-
of DKA have been reported in patients tients with HbA1c .8.0% (.64 mmol/mol)
with either type 1 or type 2 diabetes RESEARCH DESIGN AND METHODS at screening were recommended to re-
treated with SGLT2 inhibitors in clinical Study Design and Patients duce their basal insulin dose by 10%. Al-
practice (18). In some of these cases, the This study was a randomized, multicen- gorithms for titrating basal and bolus
presentation of DKA was atypical be- ter, placebo-controlled, double-blind, insulin doses to achieve prespecified
cause blood glucose levels were only phase 2 trial that consisted of a 2-week prebreakfast, prelunch, predinner, and
moderately elevated. Patients treated prerandomization period followed by an bedtime glucose levels were provided
with SGLT2 inhibitors may be at a 18-week double-blind treatment phase as a general guideline for investigators
greater risk for DKA because increased and a 2-week follow-up period for all (15). Compliance with insulin therapy
UGE (ranging from 50 to 100 g/day) can patients. Eligible patients were men was assessed based on patient records of
result in lower hyperglycemia than and women with type 1 diabetes for basal and bolus insulin doses at specified
might be expected in the setting of $1 year who had an HbA 1c of 7.0– time points.
DKA (19). Increased UGE lowers plasma 9.0% (53–75 mmol/mol) at screening
glucose levels, which may cause patients and were on a stable insulin regimen End Points and Assessments
with type 1 diabetes on background in- with multiple daily insulin injections or The primary efficacy end point of the
sulin to lower their insulin doses to continuous subcutaneous insulin infu- study was the proportion of patients at
avoid hypoglycemia. The reduction in in- sion for $8 weeks before screening. Pa- week 18 with an HbA1c reduction $0.4%
sulin suppresses glucose oxidation and tients were 25–65 years of age with a ($4.4 mmol/mol) and no increase in
stimulates lipolysis and the mobilization BMI of 21–35 kg/m2 at screening. Key body weight relative to baseline (15).
534 DKA With Canagliflozin in Type 1 Diabetes Diabetes Care Volume 39, April 2016

Change from baseline in insulin dosage, inhibition (i.e., urinary tract infections, given the wide range and small sample
after the initial 10–20% downtitration, AEs related to osmotic diuresis and vol- size, it is not possible to make conclu-
was assessed at week 18. Overall safety ume depletion) (15), consistent with the sions regarding the time to onset of a
and tolerability were assessed based on safety profile reported in studies of serious AE of DKA. One patient with an
AE reports, safety laboratory tests, vital canagliflozin in patients with type 2 di- initial serious AE of DKA and a subse-
sign measurements, and physical exam- abetes. The incidence of AEs leading to quent nonserious ketone-related AE
inations. A ketone-related AE was de- discontinuation was low across groups (i.e., increased urine ketones) discontin-
fined as any event from a prespecified (one patient in the canagliflozin 100 ued the study. All patients experiencing
list of preferred terms (i.e., acidosis, mg group, two in the canagliflozin a serious AE of DKA also had a coexisting
blood ketone body increased, blood 300 mg group, and none in the placebo condition at the time of the event that
ketone body present, DKA, diabetic ke- group). The incidence of serious AEs may have contributed to the develop-
toacidotic hyperglycemic coma, ketoaci- was 7.7% and 6.8% with canagliflozin ment of DKA, including infection (influ-
dosis, ketonemia, ketonuria, ketosis, 100 and 300 mg, respectively; no patients enza, pneumonia, infusion-site infection,
metabolic acidosis, urine ketone body treated with placebo experienced a serious food poisoning), reduction in insulin dose
present). Patients were counseled on AE. This difference was mainly driven by (insulin pump failure or malfunction, non-
how to recognize and monitor signs serious AEs of DKA. compliance with insulin dosing), and re-
and symptoms of DKA and received in- duction in food intake.
structions on when to contact the study Incidence of Ketone-Related AEs and
site or seek medical attention per local Serious AEs of DKA Characteristics of Patients With and
guidelines. Guidelines for the diagnosis, The incidence of any ketone-related AE Without a Ketone-Related AE
treatment, and prevention of DKA pro- at week 18 with canagliflozin 100 and Baseline demographic and disease char-
vided to investigators and patients are 300 mg was 5.1% (n = 6 of 117) and acteristics were generally similar in pa-
included in the Supplementary Data. Pa- 9.4% (n = 11 of 117), respectively; no tients with and without a ketone-related
tients were instructed to test their urine patients in the placebo group experi- AE (Table 3). The mean baseline HbA1c
for ketones if blood glucose was .13.9 enced a ketone-related AE (Table 1). was 8.0% (64 mmol/mol) and 7.9%
mmol/L (.250 mg/dL) or if they felt ill, The majority of ketone-related AEs oc- (63 mmol/mol) in patients with and
even if blood glucose levels were normal. curred after 1 month of treatment with without a ketone-related AE, respec-
Furthermore, patients were advised on canagliflozin 100 and 300 mg (Supple- tively; the mean duration of type 1
how to appropriately manage insulin mentary Fig. 1). The incidence of serious diabetes was 22 years in both groups.
and diet during illness and how to moni- AEs of DKA that required hospitalization History of DKA was not a predictor
tor for glucose and ketones. Supplies for was 4.3% (n = 5 of 117) with canagliflozin of whether a patient experienced a
serum ketone measurements were not 100 mg and 6.0% (n = 7 of 117) with ketone-related AE in this study. Of the
provided. Interruption of the study drug canagliflozin 300 mg. In these 12 pa- 17 patients with a ketone-related AE,
during an illness was based on the dis- tients, blood glucose levels at the time 12 were female. At week 18, reductions
cretion of the investigator or treating of hospitalization ranged from 9.4 to from baseline in body weight were sim-
physician. .44.4 mmol/L (170 to .800 mg/dL); 5 ilar with canagliflozin 100 and 300 mg in
of these patients had blood glucose lev- patients with (–2.7% and –5.5%, respec-
Statistical Analyses els ,13.9 mmol/L (250 mg/dL), the typ- tively) and without (–3.2% and –5.1%,
Safety analyses were conducted by us- ical glucose threshold used to define respectively) a ketone-related AE. In
ing the modified intent-to-treat analysis DKA. Three of the 12 patients had pH the 12 patients with a serious AE of
set (i.e., all patients who were random- values ,7.0 at the time of hospitaliza- DKA, changes from baseline in body
ized and received one or more doses of tion. Specific details of the 12 patients weight at the closest study assessment
double-blind study drug) and included with serious AEs of DKA, including labo- before the event (Table 2) were similar
all reported AEs with onset during the ratory results, are presented in Table 2. to the observed changes in patients with
treatment phase (i.e., treatment-emergent The median time to a serious AE of DKA or without ketone-related AEs at week
AEs). Statistical testing of comparisons was 116 days (range 27–130 days) with 18. Most of these patients (n = 8) had a
of canagliflozin versus placebo was not canagliflozin 100 mg and 32 days (range weight loss of ,5.0%, and the largest
performed (not prespecified) for safety 5–110 days) with canagliflozin 300 mg; body weight reduction observed was
analyses; therefore, P values are not
Table 1—Summary of the incidence of ketone-related AEs and serious AEs of DKA
RESULTS Placebo CANA 100 mg CANA 300 mg
Overall Safety (n = 117) (n = 117) (n = 117)
At week 18, the overall incidence of AEs Any ketone-related AE* 0 6 (5.1) 11 (9.4)†
was higher in the canagliflozin 100 and Serious AEs of DKA requiring hospitalization 0 5 (4.3) 7 (6.0)
300 mg groups (55.6% and 67.5%, re- Nonserious ketone-related AEs‡ 0 1 (0.9) 5 (4.3)
spectively) than in the placebo group Data are n of patients (%). CANA, canagliflozin. *Including DKA, ketoacidosis, and urine ketone
(54.7%) and were mainly associated body present. †One patient had an initial serious DKA event and a subsequent nonserious AE of
with an increased incidence of AEs increased urine ketones. ‡Including ketoacidosis in two patients and the presence of urine
ketones in four patients.
related to the mechanism of SGLT2
Peters and Associates

Table 2—Summary of patients with serious AEs of DKA

1 2 3 4 5 6 7 8 9 10 11 12
Treatment group CANA 100 mg CANA 100 mg CANA 100 mg CANA 100 mg CANA 100 mg CANA 300 mg CANA 300 mg CANA 300 mg CANA 300 mg CANA 300 mg CANA 300 mg CANA 300 mg
Sex M F F F F M M M F F F F
Age (years) 34 28 27 41 41 56 41 44 54 60 28 42
Diabetes duration 16.5 18.7 10.9 28.5 23.0 45.3 32.4 16.2 14.3 20.0 17.7 34.0
(at randomization)
History of DKA Yes Yes No No Yes No No No No No No No
Baseline HbA1c 7.9 (63) 8.2 (66) 8.5 (69) 6.9 (52) 8.6 (70) 7.5 (58) 7.9 (63) 8.3 (67) 7.9 (63) 7.9 (63) 8.5 (69) 8.3 (67)
(% [mmol/mol])
Baseline body 88.5 74.1 84.7 92.4 89.0 78.0 95.3 64.9 84.8 65.9 60.7 91.0
weight (kg)
Percentage change in 23.6 21.8 0.4 24.8 27.4 23.5 0.8 21.4 22.6 20.5 NA‡ 22.6
body weight at
before DKA event
Onset day relative 116 77 27 116 130 110 27 28 103 47 5 32
to first dose
pH 6.9 7.2 7.3 7.3 6.9 7.1 7.0 NA NA 6.9 7.2 7.2
Ketones: 3BOH 3BOH: Blood ketones Blood ketones ur $80 (high) NA 3BOH: .5 mmol/L AC: 320 mg/dL; AC: large; ur +++ AC: positive Blood ketones ur ++++ NA
(blood), AC 6.3 mmol/L NOS: 150 mg/dL NOS: 0.71 (high) ur +++ (moderate); NOS: .20
(blood), and/or ur ++ (high)
urine ketone*
Anion gap (mmol/L) .35 22 14.6 19.6 .37 27 NA 31 .28 NA NA NA
Bicarbonate NA 11.5 15 12 ,5 6.1 7.2 NA NA 6 5.5 10.1
Blood glucose 33.2 (598) 10.8 (194) 31.3 (563) 11.0 (199) 44.8 (807) 12.1 (218) 22.3 (401) 19.4 (.350) 27.0 (486) .36.1 (.650) 13.5 (243) 9.4 (170)
Precipitating factors c Pneumonia c Failure of the c Noncompliance c Insulin pump c Pneumonia c Influenza with c Low-carbohydrate c Interruption of c Sepsis c Gastroenteritis/ c Death in family c Viral
insulin infusion with insulin malfunction c Possible vomiting and diet with insulin infusion secondary to abdominal c Uncontrolled infection
pump set the therapy, with insufficient weight loss of reduction in by insulin pump infection at sepsis diet with c Tooth
day before last dose of insulin related 4.5 kg before insulin dose for 6 h the day the insulin noncompliance extraction
hospitalization, basal insulin to insulin hospitalization before pump with insulin and root
which resulted in administered pump hospitalization injection site therapy and canal
hyperglycemia on day 25 malfunction c Increased improper
before the pump c Excessive alcohol monitoring of

was fixed consumption of consumption blood glucose

alcohol the day
3BOH, 3-b-hydroxybutyrate; AC, (aceto-) acetone; CANA, canagliflozin; CSII, continuous subcutaneous insulin infusion; MDI, multiple daily insulin injection; NA, not assessed; NOS, not otherwise specified; ur, urine ketone; +, positive result. *Ketone data are
reported using units provided in hospital records. †At the time of hospital admission. ‡The only body weight measurement taken before the DKA event was at baseline.
536 DKA With Canagliflozin in Type 1 Diabetes Diabetes Care Volume 39, April 2016

Table 3—Baseline demographic and disease characteristics of patients with and without a ketone-related AE
Patients with a ketone-related AE* Patients without a ketone-related AE
CANA 100 mg CANA 300 mg Placebo CANA 100 mg CANA 300 mg
(n = 6) (n = 11) (n = 117) (n = 111) (n = 106)
Male 1 (16.7) 4 (36.4) 63 (53.8) 68 (61.3) 61 (57.5)
Female 5 (83.3) 7 (63.6) 54 (46.2) 43 (38.7) 45 (42.5)
Age (years) 32.8 6 6.9 45.8 6 11.2 42.0 6 11.9 42.5 6 11.6 42.5 6 10.9
% 7.9 6 0.6 8.0 6 0.3 7.9 6 0.6 7.9 6 0.5 8.0 6 0.5
mmol/mol 63 6 6.6 64 6 3.3 63 6 6.6 63 6 5.5 64 6 5.5
Duration of type 1 diabetes (years) 18.8 6 6.2 23.3 6 10.4 23.3 6 11.0 22.2 6 11.7 21.7 6 10.7
Prior DKA 3 (50.0) 1 (9.1) 14 (12.0) 10 (9.0) 14 (13.2)
Data are n (%) or mean 6 SD. CANA, canagliflozin. *No patients in the placebo group experienced a ketone-related AE.

–7.4% in a patient treated with canagli- compared with a decrease in those of insulin therapy). There were no appar-
flozin 100 mg; two patients had small without a ketone-related AE. Results ent differences in baseline characteristics
increases (,1.0%) in body weight, and were similar when change in insulin in patients with a ketone-related AE com-
one did not have a body weight mea- dose was determined relative to insulin pared with those without a ketone-
surement postbaseline before the DKA levels at screening (i.e., before the initial related AE that would predict greater or
event. 10–20% downtitration). Whether these lesser risk. Most patients did not have a
In the overall study population, cana- changes in insulin dose are meaningful history of DKA, including those who
gliflozin was associated with reductions given the small number of patients with experienced a serious AE of DKA in this
in daily insulin dose from baseline at week a ketone-related AE and the large SDs study. Although more men than women
18 (15). For the 12 patients with a serious observed is unclear. were enrolled in this study, more
AE of DKA, data on changes in insulin dose women than men experienced a ketone-
at the time of the event were not avail- CONCLUSIONS related AE; however, the incidence of
able. Relative to baseline insulin levels Treatment with canagliflozin 100 and DKA generally is reported to be similar in
(i.e., after the protocol-specified 10–20% 300 mg for 18 weeks was associated men and women (1). Insulin dose reduc-
downtitration), mean reductions in total with an increased incidence of serious tions appeared to be greater over 18
insulin dose at week 18 were greater AEs of DKA in patients with type 1 di- weeks in patients with a ketone-related
with canagliflozin 300 mg in patients abetes on background insulin therapy. AE than in those without, but this anal-
with a ketone-related AE than in those All patients with a serious AE of DKA ysis is of limited value because insulin
without a ketone-related AE (Table 4); had precipitating factors that likely con- doses were not recorded at the time
however, an increase in total insulin tributed to the development of DKA of the event. Although DKA typically is
dose was seen with canagliflozin 100 mg (e.g., infection, decrease in carbohy- characterized by hyperglycemia, blood
in patients with a ketone-related AE drate intake, interruption/reduction glucose monitoring was not sufficient to

Table 4—Change in insulin dose in patients with and without a ketone-related AE at week 18
Patients with a ketone-related AE* Patients without a ketone-related AE
CANA 100 mg CANA 300 mg Placebo CANA 100 mg CANA 300 mg
(n = 6) (n = 11) (n = 117) (n = 111) (n = 106)
Insulin dose at baseline† (IU/day)
Total insulin 52.0 6 19.4 52.4 6 21.6 57.8 6 29.4 51.8 6 17.9 52.9 6 21.5
Change at week 18 (%) 15.0 6 72.5 214.9 6 17.7 5.1 6 32.6 25.0 6 24.3 26.1 6 48.3
Basal insulin 27.1 6 12.6 26.6 6 13.3 27.6 6 14.5 26.7 6 10.5 27.1 6 13.9
Change at week 18 (%) 27.3 6 30.5 29.7 6 12.4 15.9 6 33.2 23.2 6 18.0 26.9 6 24.4
Bolus insulin 25.3 6 12.1 26.9 6 13.5 30.4 6 20.3 25.5 6 11.1 27.0 6 13.0
Change at week 18 (%) 22.8 6 70.7 222.5 6 36.9 25.5 6 35.6 0.3 6 57.3 214.3 6 33.9
Insulin dose at screening‡ (IU/day)
Total insulin 56.8 6 20.5 56.4 6 22.4 62.6 6 31.0 56.8 6 19.3 57.0 6 22.2
Change at week 18 (%) 4.3 6 55.0 220.0 6 15.9 23.7 6 28.0 213.8 6 21.4 213.7 6 42.9
Basal insulin 32.0 6 13.5 30.6 6 14.3 32.5 6 16.5 31.8 6 12.0 31.2 6 15.0
Change at week 18 (%) 218.8 6 20.3 220.0 6 11.0 22.1 6 28.9 219.0 6 16.1 219.5 6 22.5
Bolus insulin 25.3 6 12.1 26.9 6 13.5 30.4 6 20.2 25.5 6 11.1 27.0 6 13.0
Change at week 18 (%) 22.3 6 63.2 222.5 6 36.9 25.5 6 35.6 0.3 6 57.3 214.1 6 33.7
Data are mean 6 SD. CANA, canagliflozin. *No patients in the placebo group experienced a ketone-related AE. †After the initial protocol-specified
10–20% downtitration. ‡Before the initial protocol-specified 10–20% downtitration. Peters and Associates 537

detect DKA in this study because some control with concomitant reduction in no longer available, and/or changes
serious events occurred with minimal insulin requirement, and the potential were not recorded by the treating phy-
increases in blood glucose. As a result, increase in glucagon secretion associ- sician in hospital records.
some patients received a misdiagnosis ated with SGLT2 inhibition, which leads Because of the potentially life-threat-
from their health care providers, thus to a decrease in the insulin-to-glucagon ening nature of DKA in patients with
delaying proper treatment. ratio and promotes ketogenesis. type 1 diabetes, further development
DKA has also been reported in pilot A limitation of this study was that very of SGLT2 inhibitor therapy as a treat-
studies of the SGLT2 inhibitors sotagli- little was known about the risk of DKA in ment for type 1 diabetes should proceed
flozin and empagliflozin in patients with patients with type 1 diabetes treated with with caution. On the basis of the findings
type 1 diabetes (13,14). All cases (n = 2 SGLT2 inhibitors when the study was initi- from this phase 2 study, potential miti-
of 16 patients treated with sotagliflozin ated. Although it was acknowledged that gation strategies for future clinical trials
and n = 2 of 40 patients treated with patients may be at an increased risk for in patients with type 1 diabetes may in-
empagliflozin) were associated with developing DKA, no specific monitoring clude routine monitoring of blood and
precipitating causes (i.e., gastroenteri- guidelines or mitigation strategies were urine ketones throughout the study;
tis, insulin pump failure). Similar find- implemented other than providing instruc- use of lower canagliflozin doses; inter-
ings have also been reported in clinical tions on the importance of ketone moni- ruption of treatment during any illness
practice. A retrospective study in 27 pa- toring and sick-day management. These or if undergoing major surgical proce-
tients with type 1 diabetes treated with instructions were not specific to the study dures; discontinuation of treatment if
canagliflozin 100 mg reported two cases but followed standard American Diabetes ketone levels remain elevated despite
of DKA that developed due to insulin Association recommendations and were increases in insulin dose; and provi-
pump failure and cessation of insulin implemented at the beginning of the sion of more detailed algorithms for
therapy (16). A recent case series de- study. Additionally, patients were pro- ketone and blood glucose assess-
scribed 13 episodes of DKA in patients vided urine ketone test strips at the begin- ments, insulin dose adjustments, car-
with type 1 (n = 7) and type 2 (n = 2) di- ning of the study, which were replaced if bohydrate intake, and when to contact a
abetes treated with SGLT2 inhibitors (18). needed. Study investigators were re- physician. In addition, regulatory scrutiny
In all cases, patients had near-normal minded of the importance of ketone of such trials will be important to ensure
blood glucose levels and contributing monitoring and sick-day management patient safety before SGLT2 inhibitor
factors (e.g., reduced insulin dose in throughout the study and used their judg- therapy is approved for use in type 1
most patients with type 1 diabetes, sur- ment regarding participant reinforcement. diabetes.
gery in both patients with type 2 dia- Investigators were also required to review Additionally, patients, treating physi-
betes). DKA in patients with type 2 reported serious AEs throughout the cians, and emergency care providers
diabetes is rare but may occur in times study, including cases of DKA. must be made aware of the potential
of acute illness, such as trauma, surgery, Patients were instructed to monitor risks of DKA with SGLT2 inhibitor ther-
or infection (1). A separate analysis of se- urine ketones as part of standard sick- apy. Patients taking SGLT2 inhibitors
rious AEs of DKA in the canagliflozin day management guidelines (i.e., during and their physicians should closely mon-
type 2 diabetes clinical trial program an illness or if blood glucose levels itor serum ketones when patients are ill,
(n = 17,596 patients) demonstrated that were $13.9 mmol/L [$250 mg/dL]). Al- have reduced insulin and/or food in-
DKA occurred at a low frequency though DKA typically is characterized by take, or experience any circumstance
(,0.1%) in canagliflozin-treated pa- blood glucose levels .13.9 mmol/L that is known to be associated with an
tients with an overall incidence of 0.5, (.250 mg/dL), bicarbonate levels ,18 increased risk of DKA (e.g., increased al-
0.8, and 0.2 per 1,000 patient-years in mmol/L, and the presence of ketones cohol consumption, unusual strenuous
the canagliflozin 100 and 300 mg and com- (2), physicians did not use strict criteria or prolonged physical exercise, intense
parator groups, respectively (24). In that for diagnosing DKA. However, as de- physical or psychological stress). In
analysis, the majority of canagliflozin- scribed previously, DKA in patients these circumstances, serum ketones
treated patients with a serious AE of treated with SGLT2 inhibitors can occur should be monitored regardless of con-
DKA were on insulin and had precipi- with minimal symptoms (18), and pa- comitant blood glucose levels. Should
tating factors similar to those reported tients may not have the characteristic ketone elevation occur, patients should
in the current study (e.g., infection, marked increase in blood glucose, thus increase insulin and fluid intake and
noncompliance with insulin therapy); 6 delaying diagnosis and treatment. consider interruption of canagliflozin
of the 10 canagliflozin-treated patients Another limitation of this study is that treatment during times of illness and
showed evidence of autoimmune diabe- the protocol did not instruct patients to before undergoing surgical proce-
tes (i.e., latent autoimmune diabetes of proactively record details of their treat- dures. Furthermore, patients should
adulthood, type 1 diabetes, GAD65 an- ment regimen that might pertain to seek medical attention when having
tibody positive). The underlying mecha- DKA, such as changes in insulin doses. symptoms of DKA or elevated ke-
nisms of increased DKA incidence in As a result, it was not possible to obtain tone levels that cannot be safely self-
patients treated with SGLT2 inhibitors reliable information regarding the exact managed by increasing insulin and/or
are not fully understood but may be reduction in insulin dose preceding a food intake.
related to the insulin-independent re- DKA event in most cases because pa- In summary, treatment with canagliflo-
duction of blood glucose through in- tients may have been unable to remem- zin for 18 weeks was associated with an
creased UGE, which allows for glycemic ber these changes, the pump data were increased incidence of ketone-related
538 DKA With Canagliflozin in Type 1 Diabetes Diabetes Care Volume 39, April 2016

AEs, including serious AEs of DKA, in pa- responsibility for the integrity of the data and the 12. Henry RR, Rosenstock J, Edelman S, et al.
tients with type 1 diabetes. The increased accuracy of the data analysis. Exploring the potential of the SGLT2 inhibitor
dapagliflozin in type 1 diabetes: a randomized,
risk of DKA may be related to reductions
References double-blind, placebo-controlled pilot study.
in insulin dose and/or increases in insulin Diabetes Care 2015;38:412–419
1. Kitabchi AE, Umpierrez GE, Miles JM, Fisher
resistance during times of illness or stress. JN. Hyperglycemic crises in adult patients with 13. Perkins BA, Cherney DZ, Partridge H, et al.
Mitigation strategies are expected to re- diabetes. Diabetes Care 2009;32:1335–1343 Sodium-glucose cotransporter 2 inhibition and
duce the risk of DKA in patients with 2. Trachtenbarg DE. Diabetic ketoacidosis. Am glycemic control in type 1 diabetes: results of an
type 1 diabetes in future clinical trials, Fam Physician 2005;71:1705–1714 8-week open-label proof-of-concept trial. Dia-
3. Perilli G, Saraceni C, Daniels MN, Ahmad A. betes Care 2014;37:1480–1483
and mechanistic preclinical studies are
Diabetic ketoacidosis: a review and update. Curr 14. Sands AT, Zambrowicz BP, Rosenstock J,
currently ongoing to understand the re- et al. Sotagliflozin, a dual SGLT1 and SGLT2 in-
Emerg Hosp Med Rep 2013;1:10–17
lationship between canagliflozin and de- 4. Leslie RD, Kolb H, Schloot NC, et al. Diabetes hibitor, as adjunct therapy to insulin in type 1
velopment of DKA. classification: grey zones, sound and smoke: Ac- diabetes. Diabetes Care 2015;38:1181–1188
tion LADA 1. Diabetes Metab Res Rev 2008;24: 15. Henry RR, Thakkar P, Tong C, Polidori D,
511–519 Alba M. Efficacy and safety of canagliflozin, a
5. Laugesen E, Østergaard JA, Leslie RD; Danish sodium glucose co-transporter 2 inhibitor, as
Acknowledgments. The authors thank all in-
vestigators, study teams, and patients for par- Diabetes Academy Workshop and Workshop add-on to insulin in patients with type 1 diabe-
ticipating in this study. Speakers. Latent autoimmune diabetes of the tes mellitus. Diabetes Care 2015;38:2258–2265
Funding. R.R.H. has received support from a adult: current knowledge and uncertainty. Dia- 16. Argento NB, Nakamura K. Glycemic effects
Department of Veterans Affairs Merit Review bet Med 2015;32:843–852 of SGLT-2 inhibitor canagliflozin in type 1 diabe-
grant and the VA San Diego Healthcare System. 6. Umpierrez GE, Murphy MB, Kitabchi AE. Di- tes using Dexcom G4 Platinum CGM. Endocr
Duality of Interest. This study was supported abetic ketoacidosis and hyperglycemic hyperos- Pract. 2 November 2015 [Epub ahead of print].
by Janssen Research & Development, LLC. Med- molar syndrome. Diabetes Spectr 2002;15: DOI: 10.4158/EP151016.OR
ical writing support was provided by Kimberly 28–36 17. Pieber TR, Famulla S, Eilbracht J, et al. Em-
Fuller, PhD, of MedErgy HealthGroup and was 7. Food and Drug Administration. FDA drug pagliflozin as adjunct to insulin in patients with
funded by Janssen Global Services, LLC. Canagli- safety communication: FDA warns that SGLT2 type 1 diabetes: a 4-week, randomized, placebo-
flozin was developed by Janssen Research & inhibitors for diabetes may result in a serious controlled trial (EASE-1). Diabetes Obes Metab
Development, LLC, in collaboration with condition of too much acid in the blood [article 2015;17:928–935
Mitsubishi Tanabe Pharma Corporation. A.L.P. online], 2015. Available from http://www.fda 18. Peters AL, Buschur EO, Buse JB, Cohan P,
has been an investigator, speaker, and/or con- .gov/Drugs/DrugSafety/ucm446845.htm. Accessed Diner JC, Hirsch IB. Euglycemic diabetic ketoaci-
sultant for Abbott Diabetes Care, Amgen, 19 May 2015 dosis: a potential complication of treatment
AstraZeneca, Becton Dickinson, Biodel, Bristol- 8. European Medicines Agency. Review of diabe- with sodium-glucose cotransporter 2 inhibition.
Myers Squibb, Boehringer Ingelheim, CVS/ tes medicines called SGLT2 inhibitors started. Risk Diabetes Care 2015;38:1687–1693
Caremark, Janssen, Lexicon, Eli Lilly, Medtronic, of diabetic ketoacidosis to be examined [article on- 19. Rosenstock J, Ferrannini E. Euglycemic di-
Merck, Novo Nordisk, OptumRx, Sanofi, Takeda, line], 2015. Available from http://www.ema abetic ketoacidosis: a predictable, detectable,
and Thermalin Diabetes. R.R.H. has received and preventable safety concern with SGLT2 in-
grant support from Hitachi, Janssen, Eli Lilly, human/referrals/SGLT2_inhibitors/human_referral_ hibitors. Diabetes Care 2015;38:1638–1642
Sanofi-Aventis, and ViaCyte and is a consul- prac_000052.jsp&mid=WC0b01ac05805c516f. 20. Taylor SI, Blau JE, Rother KI. SGLT2 inhibi-
tant/advisory board member for Alere, Amgen, tors may predispose to ketoacidosis. J Clin En-
Accessed 25 June 2015
AstraZeneca, Boehringer Ingelheim, ClinMet, docrinol Metab 2015;100:2849–2852
9. Devineni D, Morrow L, Hompesch M, et al.
Eisai, Elcelyx, Gilead, Intarcia Therapeutics, Ionis 21. Ferrannini E, Muscelli E, Frascerra S, et al.
Canagliflozin improves glycaemic control over
Pharmaceuticals, Janssen, Merck, Novo Nordisk,
28 days in subjects with type 2 diabetes not Metabolic response to sodium-glucose cotrans-
and Sanofi-Aventis. P.T., C.T., and M.A. are full-
optimally controlled on insulin. Diabetes Obes porter 2 inhibition in type 2 diabetic patients.
time employees of Janssen Research & Develop-
Metab 2012;14:539–545 J Clin Invest 2014;124:499–508
ment, LLC. No other potential conflicts of
10. Sha S, Devineni D, Ghosh A, et al. 22. Bonner C, Kerr-Conte J, Gmyr V, et al. In-
interest relevant to this article were reported.
Author Contributions. A.L.P., R.R.H., and P.T. Canagliflozin, a novel inhibitor of sodium glu- hibition of the glucose transporter SGLT2 with
contributed to the study conduct; data acquisi- cose co-transporter 2, dose dependently re- dapagliflozin in pancreatic alpha cells triggers
tion, analysis, and interpretation; and drafting, duces calculated renal threshold for glucose glucagon secretion. Nat Med 2015;21:512–
review, and approval of the manuscript. C.T. excretion and increases urinary glucose excre- 517
contributed to the data analysis and interpreta- tion in healthy subjects. Diabetes Obes Metab 23. Beck RW, Tamborlane WV, Bergenstal RM,
tion and drafting, review, and approval of the 2011;13:669–672 Miller KM, DuBose SN, Hall CA; T1D Exchange
manuscript. M.A. contributed to the study de- 11. Polidori D, Sha S, Ghosh A, Plum-Mörschel Clinic Network. The T1D Exchange clinic regis-
sign and conduct; data acquisition, analysis, L, Heise T, Rothenberg P. Validation of a novel try. J Clin Endocrinol Metab 2012;97:4383–4389
and interpretation; and drafting, review, and method for determining the renal threshold for 24. Erondu N, Desai M, Ways K, Meininger G.
approval of the manuscript. A.L.P. and M.A. are glucose excretion in untreated and canagliflozin- Diabetic ketoacidosis and related events in the
the guarantors of this work and, as such, had full treated subjects with type 2 diabetes mellitus. canagliflozin type 2 diabetes clinical program.
access to all the data in the study and take J Clin Endocrinol Metab 2013;98:E867–E871 Diabetes Care 2015;38:1680–1686