SGLT2-Inhibitor Induced Euglycemic Ketoacidosis in Acute Surgical Patients
SGLT2-Inhibitor Induced Euglycemic Ketoacidosis in Acute Surgical Patients
SGLT2-Inhibitor Induced Euglycemic Ketoacidosis in Acute Surgical Patients
41
www.edoriumjournals.com/case-reports/jcrs
CASE REPORT
SERIES PEER REVIEWEDOPEN ACCESS
| OPEN ACCESS
blood pressure control and weight loss benefits have led to polyuria with urine output of up to 600 mL/h, associated
increasing the use of drug in the community. A significant with glycosuria. She needed significant potassium and
adverse effect that has been reported is euglycemic phosphate replacement intravenously of up to 20 mEq/
ketoacidosis. This is of particular significance to surgeons hr and 124 mg/hr respectively. Insulin/dextrose infusion
as the combination of long periods of fasting, surgical was continued for a further 48 hours with intravenous
stress and reduced insulin dosing in acute surgical electrolyte replacement before her urine output and
patients increases their risk of developing euglycaemic electrolytes improved. In total, she received a total of 23
ketoacidosis [1–6]. L of intravenous 5% dextrose in 72 hours, with associated
This report presents four acute surgical patients urine output of 27 L over that time. Urine dipstick showed
who developed euglycaemic ketoacidosis and explores significant glycosuria. Unfortunately, no formal urinary
the importance of early recognition of the condition by glucose, potassium or phosphate testing was performed.
surgical teams. Renal function was normal throughout the admission.
The patient was subsequently discharged on post-
operative day-7. Empagliflozin was permanently ceased.
CASE SERIES
Case 2
Case 1
A 70-year-old female presented with acute
A 43-year-old female presented for elective total cholecystitis. This was on a background of hypertension,
abdominal hysterectomy, bilateral salpingectomy and dyslipidaemia and type 2 diabetes mellitus. Her diabetes
incisional hernia repair. This was on a background of treatment consisted of metformin 1700 mg mane and
type 2 diabetes mellitus, hypercholesterolemia and 850 mg nocte, insulin aspart/aspart-protamine 30/70
hypertension. Her medications included sitagliptin 100 formulation 24 units mane and 20 units nocte and
mg daily, metformin 1000 mg daily, rosuvastatin 10 canagliflozin 300 mg daily. She had an acute kidney
mg daily, telmisartan 40 mg daily and empagliflozin 25 injury (AKI) (kidney disease: improving global outcomes
mg daily. She was fasted from midnight and underwent guideline stage 1) on admission with her serum creatinine
the procedure in the morning with no complications. of 1.11 mg/dL, resolving back to her baseline of 0.74 mg/
The patient was on a clear fluid diet to begin with and dL within 48 hours [7]. The patient spent 12 hours nil-
upgraded to a free fluid diet the following day at dinner. by-mouth on her admission day, followed by a clear fluid
She continued to receive all her medications including diet overnight. The patient was not charted insulin for
her oral hypoglycemic agents during the perioperative the first day of admission during her fasting state. On her
period. On the second postoperative day, she developed second day of admission it was noted that the patient had
significant anemia with a hemoglobin drop from 10.5 to a high anion gap metabolic acidosis with a pH of 7.27,
6.9 g/dL. Computed tomography (CT) imaging of her despite blood glucose levels ranging from 180–270 mg/
abdomen revealed a large rectus abdominis hematoma. dL. Further investigation revealed serum ketones of 20.92
The patient was again fasted, transfused one unit of mg/dL. An insulin infusion was commenced, successfully
packed red blood cells and underwent a laparotomy returning the ketones level to normal. Canagliflozin was
the following morning during which 1.5 L of blood was ceased and the patient advised to avoid SGLT2 inhibitors
drained from the hematoma. There was no evidence of in the future.
active bleeding.
Postoperatively, the patient required noradrenaline
to maintain her blood pressure and was transferred to
Case 3
the ICU. At this point she was noted to have a metabolic A 45-year-old female presented with cholelithiasis
acidosis with a pH of 7.02, HCO3 5 mEq/L, and base and underwent an uncomplicated cholecystectomy. This
excess of 24.6 mEq/L. Blood glucose was only 181.98 was on a background of hypertension, dyslipidemia,
mg/dL and ketones 33.12 mg/dL. Serum Ppotassium bell’s palsy and type 2 diabetes mellitus with normal
was 3.1 mEq/L and phosphate was as low as <0.31 renal function. Treatment for diabetes mellitus consisted
mg/dL. A diagnosis of SGLT-2i induced euglycemic of metformin and dapagliflozin 10 mg daily. Day-1
ketoacidosis was suspected. Empagliflozin was ceased, postsurgery, the patient developed sinus tachycardia up
insulin and dextrose infusion commenced as well as to 150 bpm, and in the emergency call that followed it was
aggressive intravenous electrolyte replacement and discovered she had a high anion gap metabolic acidosis
total parenteral nutrition at 30 mL/h. She then became with a pH 7.19, bicarb 8 mEq/L and ketones 34.86 mg/dL,
febrile with signs of systemic inflammatory response despite a blood glucose level 144 mg/dL. The patient was
syndrome (SIRS), requiring ongoing vasopressor, and commenced on an insulin infusion and the ketoacidosis
was commenced on intravenous piperacillin-tazobactam subsequently resolved. The patient was subsequently
for presumed sepsis. Over the next 24 hours, the patient’s commenced on regular insulin and dapagliflozin was
ketosis and acidosis resolved, though she developed ceased permanently.
Table 1: Blood pathology results: comparison of results at time of presentation to results at discharge
Case BSL On Presentation On Discharge HbA1c Date
(mg/dL) Bicarb Ketones pH Lactate Bicarb Ketones pH Lactate %
(mEq/L) ( mg/dL) ( mg/dL) (mEq/L) ( mg/dL) ( mg/dL)
1 181.8 5 33.12 7.02 12.61 22 0.0 7.48 6.31 - April 2017
2 273.6 10 20.92 7.27 17.12 25 5.81 7.36 7.21 - Sept 2016
3 135 10 32.54 7.19 12.61 27 0.58 7.40 7.21 8.0 May 2016
4 205.2 14 25.56 7.30 7.21 26 1.74 7.39 11.71 10.4 July 2016
inhibitors likely also predispose to the development of an example of this. The surgical patient was kept nil by
ketoacidosis indirectly through suppression of renal mouth for an extended period, with her insulin dose
ketone oxidation [2]. initially withheld and then restarted at a reduced dose.
This accelerates the process of ketogenesis, as discussed
above. A second, related, risk factor for development of
Our Cases euglycemic ketoacidosis is restriction of carbohydrate
The four cases presented in this report have some availability [3]. This can occur in situations of inter-
significant differences in presentation, severity and current illness or, significantly in the cases discussed,
management issues. Case 1 was clearly the most severe, prolonged fasting and surgical stress [2, 3, 16].
requiring a four day intensive care admission, though the A classic acute surgical patient will often unavoidably
case was also complicated by sepsis and an abdominal experience extended periods of fasting and reduced
hematoma. In each case, the ketosis and acidosis resolved insulin dosing combined with the stress of both the
within 24 hours of ceasing the SGLT-2 inhibitor and presenting condition and potentially surgery. For this
commencing insulin/dextrose infusion. Case 1, however, reason, the perioperative period could be considered a
differed from the others in the extended period of polyuria perfect storm for precipitating euglycemic ketoacidosis.
and electrolyte disturbances in the form of hypokalemia Early recognition of the condition by surgeons will
and hypophosphatemia. become increasingly important as the use of this new
We hypothesize that the significant hypokalemia and drug-class becomes more prevalent.
hypophosphatemia in Case 1 was caused by a combination
of intracellular shift secondary to the extended period of
insulin infusion as well as urinary loss of potassium and Recognizing and managing the problem
phosphate. Unfortunately, urinary electrolytes were not The clinical presentation of euglycemic ketoacidosis
measured to confirm this theory. is often fairly non-specific. Patients may complain of
The cause for the extended period of polyuria is less abdominal pain, nausea, vomiting, lethargy and malaise
clear. Empagliflozin was administered until day two [17]. All of these are common symptoms for other surgical
post-procedure, with the final dose given the day prior conditions, making diagnosis more difficult. More severe
to her ICU admission. Empagliflozin has been shown presentations may involve significant clinical dehydration
to have an extended effect on urinary glucose excretion and mental obtundation [17]. An arterial or venous blood
(UGE) when compared to dapagliflozin and canagliflozin gas, looking especially at pH and lactate level, should be
[14]. Though the half-life of empagliflozin in the serum taken. Blood ketones should be measured as well, even
is only 10–19 hours, UGE is maintained long after with normal or only slightly raised blood glucose.
plasma concentrations diminish [15]. While plasma Once the diagnosis is established, the SGLT2 inhibitor
empagliflozin concentrations peak 2 hours post dose, should be ceased and the patient commenced on an
UGE peaks 7 hours post dose and only reduces to a third insulin infusion with adequate fluid resuscitation [18].
by 60 hours. However, by 60 hours post dose the plasma Electrolyte disturbances are common and should be
concentrations of empagliflozin are less than a 50th of corrected accordingly [18].
maximum [14]. This shows that empagliflozin’s tissue
effect of inhibition of SGLT2 action continues well beyond
its virtual elimination from the blood. It is likely that this CONCLUSION
prolonged effect of empagliflozin on UGE, couple with
the large volumes of intravenous dextrose, resulted in an Given the increasing use of sodium-glucose co-
osmotic diuresis. transporter 2 inhibitors and the unique combination of
While euglycemic ketoacidosis has been reported in risk factors that surgical patients commonly possess, it is
patients not taking SGLT2 inhibitors in the past, it is an important that surgeons are familiar with the condition of
exceedingly rare condition in type 2 diabetes mellitus and euglycaemic ketoacidosis caused by these drugs and are
SGLT2 inhibitors are by far the most commonly reported able to diagnose and treat it early. This report presents
cause. The presentations in these cases are consistent with four cases of SGLT2 inhibitor induced euglycaemic
previously reported cases of SGLT2 inhibitors induced ketoacidosis in surgical patients and explores the
euglycaemic ketoacidosis, however a different previously significance of the condition in surgical patients.
unreported cause cannot be completely excluded.
*********
Contributing factors Author Contributions
Risk factors or scenarios that may precipitate the above Aaron Matthew Hawkins – Substantial contributions
adverse reactions are important to recognize, allowing to conception and design, Acquisition of data, Analysis
early diagnosis and treatment. Insulin deficiency is a and interpretation of data, Drafting the article, Revising
major precipitating factor, and the most common scenario it critically for important intellectual content, Final
for this is a reduction in exogenous insulin dose by either approval of the version to be published
the patient or health practitioner [3]. Case 2 illustrated
Richard V Jackson – Analysis and interpretation of data, 5. Zinman B, Wanner C, Lachin JM, et al. Empagliflozin,
Drafting the article, Revising it critically for important cardiovascular outcomes, and mortality in type 2
intellectual content, Final approval of the version to be diabetes. N Engl J Med 2015 Nov 26;373(22):2117–
published 28.
6. Heerspink HJ, Desai M, Jardine M, Balis D, Meininger
Hayden White – Analysis and interpretation of data,
G, Perkovic V. Canagliflozin slows progression of renal
Drafting the article, Revising it critically for important function decline independently of glycemic effects. J
intellectual content, Final approval of the version to be Am Soc Nephrol 2017 Jan;28(1):368–75.
published 7. Summary of recommendation statements. Kidney Int
Deepak L Vardesh – Substantial contributions to Suppl (2011) 2012 Mar;2(1):8–12.
conception and design, Acquisition of data, Analysis 8. Hediger MA, Rhoads DB. Molecular physiology of
and interpretation of data, Drafting the article, Revising sodium-glucose cotransporters. Physiol Rev 1994
it critically for important intellectual content, Final Oct;74(4):993–1026.
approval of the version to be published 9. Australian Statistics on Medicine 2015. Edited
by Scheme APB. Australia: Drug Utilisation Sub-
Committee (DUSC) of the Pharmaceutical Benefits
Guarantor Advisory Committee; 2015.
The corresponding author is the guarantor of submission. 10. Australian Statistics on Medicine 2014. Edited
by Scheme APB. Australia: Drug Utilisation Sub-
Conflict of Interest Committee (DUSC) of the Pharmaceutical Benefits
Authors declare no conflict of interest. Advisory Committee; 2014.
11. Nyirjesy P, Zhao Y, Ways K, Usiskin K. Evaluation
Copyright of vulvovaginal symptoms and Candida colonization
in women with type 2 diabetes mellitus treated with
© 2017 Aaron Matthew Hawkins et al. This article
canagliflozin, a sodium glucose co-transporter 2
is distributed under the terms of Creative Commons inhibitor. Curr Med Res Opin 2012 Jul;28(7):1173–8.
Attribution License which permits unrestricted use, 12. Weir MR, Januszewicz A, Gilbert RE, et al. Effect of
distribution and reproduction in any medium provided canagliflozin on blood pressure and adverse events
the original author(s) and original publisher are properly related to osmotic diuresis and reduced intravascular
credited. Please see the copyright policy on the journal volume in patients with type 2 diabetes mellitus. J
website for more information. Clin Hypertens (Greenwich) 2014 Dec;16(12):875–
82.
13. Drug Safety Communication: FDA warns that
SGLT2 inhibitors for diabetes may result in a serious
REFERENCES condition of too much acid in the blood. Edited by
Administration UFaD: FDA; 2015.
1. Peters AL, Buschur EO, Buse JB, Cohan P, Diner
14. Liu JJ, Lee T, DeFronzo RA. Why Do SGLT2 inhibitors
JC, Hirsch IB. Euglycemic diabetic ketoacidosis: A
inhibit only 30-50% of renal glucose reabsorption in
potential complication of treatment with sodium-
humans? Diabetes 2012 Sep;61(9):2199–204.
glucose cotransporter 2 inhibition. Diabetes Care
15. Gangadharan Komala M, Mather A. Empagliflozin
2015 Sep;38(9):1687–93.
for the treatment of Type 2 diabetes. Expert Rev Clin
2. Palmer BF, Clegg DJ, Taylor SI, Weir MR. Diabetic
Pharmacol 2014 May;7(3):271–9.
ketoacidosis, sodium-glucose transporter-2 inhibitors
16. Venkatesh B, Moore G, Gill D, Kelly W. Diabetic
and the kidney. J Diabetes Complications 2016
ketoacidosis precipitated by therapy with antidiabetic
Aug;30(6):1162–6.
agents SGLT2 inhibitors: Two cases. Crit Care Resusc
3. Burke KR, Schumacher CA, Harpe SE. SGLT
2015 Dec;17(4):280–2.
inhibitors: A systematic review of diabetic ketoacidosis
17. Davis SN, Umpierrez GE. Diabetic ketoacidosis in
and related risk factors in the primary literature.
type 2 diabetes mellitus–pathophysiology and clinical
Pharmacotherapy 2017 Feb;37(2):187–94.
presentation. Nat Clin Pract Endocrinol Metab 2007
4. Liu XY, Zhang N, Chen R, Zhao JG, Yu P. Efficacy and
Nov;3(11):730–1.
safety of sodium-glucose cotransporter 2 inhibitors
18. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN.
in type 2 diabetes: A meta-analysis of randomized
Hyperglycemic crises in adult patients with diabetes.
controlled trials for 1 to 2 years. J Diabetes
Diabetes Care 2009 Jul;32(7):1335–43.
Complications 2015 Nov–Dec;29(8):1295–303.