a b s t r a c t

Introduction: Euglycemic diabetic ketoacidosis (EuDKA) associated with Sodium-Glucose Cotransporter-2 inhib- itor (SGLT2i) use has been described but remains poorly understood. Data on Emergency Department (ED) pre- sentation, resource utilization, and Safety outcomes for these patients are lacking. We report a case series of patients diagnosed with EuDKA in the ED.

Methods: An electronic medical record search identified adult patients presenting to a large Tertiary ED with EuDKA. They were screened for concurrent use of SGLT2i. Clinical presentation, resource utilization, safety, and disposition data were collected and described.

Results: Five patients were included for analysis. Median age [range] was 57 [43-73] years. Presenting symptoms included nausea, vomiting, fatigue, and altered mental status. Initial results included: Serum glucose 191 mg/dL [176-215], venous pH 7.01 [6.95-7.30], serum HCO3 8 mEq/L [6-13], anion gap 27 [26-31], serum beta- hydroxybutyrate 9.9 mmol/L [9.2-12.3], and urine ketones 150 [150-150]. Patients remained on an insulin infu- sion for 18.77 h [11.25-56.48]. There were zero episodes of hypoglycemia and one episode of hypokalemia while on Insulin infusion. Time to resolution of metabolic acidosis was 23.82 h [15.45-24.77].

Discussion: We report a case series of ED patients with EuDKA associated with SGLT2i use, and describe presen- tation characteristics, resource utilization, and safety outcomes. Emergency physicians should be aware of the as- sociation between SGLT2i use and EuDKA. An appropriate work-up should be pursued for patients taking an SGLT2i who present with symptoms suggestive of DKA, including nausea, vomiting, malaise, and altered mental status, or are noted to have an unexplained elevated Anion gap metabolic acidosis.

(C) 2021

1. Introduction

Diabetic ketoacidosis results from a metabolic imbalance sec- ondary to low serum effective circulating insulin levels and elevated levels of counter-regulatory hormones such as glucagon, epinephrine, and cortisol. The outcome of this hormonal mismatch is ketone produc- tion with subsequent development of anion gap metabolic acidosis [1]. Euglycemic diabetic ketoacidosis (EuDKA) has been observed most fre- quently in patients with Type 1 diabetes mellitus (T1DM), with some cases in patients with type 2 diabetes mellitus (T2DM) also reported [1]. Laboratory diagnostic parameters for both EuDKA and DKA include pH < 7.30, HCO₃ < 18 mEq/L, and anion gap >10. However, significant hyperglycemia is absent in EuDKA, and the condition is diagnosed if serum glucose <250 mg/dL [1,2].

Known triggers for EuDKA include factors that lead to a state of car- bohydrate deficit, which causes an imbalance between low insulin con- centration and high concentration of counterregulatory hormones,

* Corresponding author.

E-mail address: [email protected] (H.A. Puls).

resulting in a shift of metabolism into ketosis while maintaining normoglycemia. Examples include gastroparesis, fasting, ketogenic diet, pregnancy, pancreatitis, surgery, infections, insulin pump use, alco- holism, glycogen storage disorders, and cirrhosis [2]. Recently, use of sodium-glucose cotransporter 2 inhibitors (SGLT2i) was identified as a trigger for EuDKA, especially when associated with triggers listed above [3].

SGLT2i were introduced in 2014 and are approved for glycemic con- trol in T2DM [4]. Use of SGLT2i increased from 0.8% of adults with DM in 2014 (the first SGLT2i approval year) to 4.4% in 2015, [5] and recent lit- erature shows potential for use in other conditions like T1DM, heart fail- ure with reduced ejection fraction, and chronic kidney disease [6-8]. SGLT2i used in the United States include canagliflozin, dapagliflozin, and empagliflozin. With SGLT2i, euglycemia is maintained through glucosuria and SGLT2i-triggered hypoinsulinemia [9]. EuDKA associated with SGLT2i use has been infrequently reported, with the largest case se- ries to date including 13 patients [3]. This series estimated a rate of 1.8 cases of SGLT2i-associated EuDKA per 1000 patient-years, though ac- knowledged that the uncertainty of this diagnosis makes the exact inci- dence unknown. A separate case series identified nine patients with 13

https://doi.org/10.1016/j.ajem.2021.01.033

0735-6757/(C) 2021

episodes of EuDKA associated with SGLT2i use from practices across the United States, further highlighting the infrequency of this condition [10]. While cases of EuDKA associated with SGLT2i use have been re- ported, EuDKA remains a poorly understood and infrequently reported pathology, and data regarding treatment and outcomes are sparse. In this case series, our objective was to further this body of literature via describing a series of adult Emergency Department (ED) patients with EuDKA associated with SGLT2i use, and detail presentation characteris-

tics, resource utilization, and safety outcomes.

1. Methods

This is a case series of adult (>18 years) ED patients with EuDKA as- sociated with SGLT2i use managed in an ED-based Intensive Care Unit

Table 1

Summary of results.

Variable Values

Presenting data

Patients (count) 5

Age [median (values)] 57 years

(43, 53, 57, 63, 73)

Male [count (%)] male 3 (60%)

Hemoglobin A1c on presentation [median (values)] 8.1%

(6.4, 7.7, 8.1, 8.2, 11.7)

Duration of SGLT2i therapy before episode [median 20 months (values)] (17, 20, 21)

Most common presenting symptoms (%) Nausea (60%), vomiting

(60%), fatigue (40%)

Heart rate on presentation [median (values)] 116 bpm

(104, 115, 115, 116, 121)

(ED-ICU) [11]. The institutional review board at the University of Mich- igan reviewed and approved this study (IRB HUM00183169). The study

setting is a large tertiary academic center with approximately 75,000

Mean arterial pressure on presentation [median (values)]

Median lab values on presentation

93 mmHg

(87, 88, 93, 102, 104)

adult ED visits per year.

Charts from patients presenting to the ED between February 2015 and January 2020 were retrospectively screened and cases were identi- fied via an ED encounter diagnosis of DKA, use of a protocolized DKA treatment order set, and initial laboratory values compatible with EuDKA (serum glucose <250 mg/dL, pH < 7.30, HCO₃ < 18, anion gap

>10) [1,2]. Charts of patients identified through primary screening were then manually reviewed, and cases were included in this case se- ries if the primary team or endocrinology consulting service notes de- scribed concern for EuDKA associated with SGLT2i while the patient was in the ED or ED-ICU.

Data extracted from charts included clinical presentation, response

Glucose [median (values)] 191 mg/dL

(176, 190, 191, 192, 215)

pH [median (values)] 7.01

(6.95, 7.01, 7.01, 7.10, 7.30)

HCO3^- [median (values)] 8 mEq/L

(6, 7, 8, 9, 13)

Anion gap [median (values)] 27

(26, 27, 27, 28, 31)

Serum beta-hydroxybutyrate [median (values)] 9.9 mmol/L

(9.2, 9.4, 9.9, 11.4, 12.3)

Urine ketones [median (values)] 150 mg/dL

(150, 150, 150, 150, 150)

Time on insulin infusion [median (values)] 18.77 h

(11.25, 11.45, 18.77, 21.43,

56.48)

to treatment, and resource utilization outcomes. Data regarding clinical presentation included age, gender, symptoms, long-term oral hypogly- cemic medication use, hemoglobin A1c, vital signs, serum glucose, serum pH, serum HCO₃, anion gap, beta-hydroxybutyrate, and urine ke-

Time interval to resolution of metabolic acidosis (pH > 7.3 and HCO3 > 18 mEq/L) [median (values)]

Disposition from ED-ICU

23.82 h

(15.45, 23.82, 24.77)

tones. Data evaluating response to treatment and resource utilization included time on insulin infusion, frequency of hypoglycemia or hypokalemia while on insulin infusion, time to resolution of metabolic acidosis, ED-ICU length of stay, hospital length of stay, and ED-ICU

Admission to general medicine ward 3 (60%)

Admission to ICU 1 (20%)

Discharge to home 1 (20%)

ED-ICU length of stay [median (values)] 22.7 h

(16.9, 18.6, 22.7, 28.3, 28.5)

disposition.

Baseline data, vital signs, laboratory values, time on insulin infusion, and disposition information were obtained via automatic electronic health record data retrieval. Patients’ symptoms and the number

Hospital length of stay for admitted patients [median (values)]

1.8 days

(0.1, 0.8, 2.8, 13)

of oral hypoglycemic medications at the time of presentation were ex- tracted via manual chart review. Numerical data are reported as median (range). Categorical data are reported as count (percentage [%]).

1. Results

Table 1 summarizes the findings of this study. We identified five pa- tients with EuDKA and concurrent SGLT2i use, and included them for analysis. Median age [range] was 57 [43-73] years, and three (60%) were male. Three patients had confirmed type 2 diabetes mellitus, one patient had confirmed type 1 diabetes mellitus, and one patient was thought to have type 2 diabetes mellitus but was undergoing evaluation for possible type 1 diabetes mellitus. All patients were prescribed three or more oral hypoglycemic medications. Three of the five patients had subcutaneous insulin as part of their home regimen. Median hemoglo- bin A1c on presentation was 8.1% [6.4-11.7]. Median duration of outpa- tient SGLT2i therapy before the episode of EuDKA was 20 [17-21] months. The most common presenting symptoms included nausea (60%), vomiting (60%) and fatigue (40%), while one patient (20%) had altered mental status. All patients were tachycardic on presentation (median heart rate: 116 bpm [104-121]) and median presenting mean arterial pressure was 93 mmHg [87-104].

Median [range] values for laboratory values on presentation in- cluded: glucose 191 mg/dL [176-215], pH 7.01 [6.95-7.30], HCO₃

8 mEq/L [6-13], anion gap 27 [26-31], serum beta-hydroxybutyrate

9.9 mmol/L [9.2-12.3], and urine ketones 150 mg/dL [150-150].

Patients remained on an insulin infusion for a median of 18.77 h [11.25-56.48]. There were no episodes of hypoglycemia (serum glucose

<70 mg/dL) and one episode of hypokalemia (serum K+ < 3.3 mEq/L) while on insulin infusion. Median time to resolution of metabolic acidosis, defined as pH > 7.3 and HCO₃ > 18 mEq/L, was 23.82h [15.45-24.77].

Dispositions from the ED-ICU included one (20%) admission to ICU because of concomitant sepsis, one (20%) Discharge home from the ED-ICU, and three (60%) admissions to a general medicine ward. The median ED-ICU length of stay was 22.7 h [16.9-28.5] and median hospi- tal length of stay for admitted patients was 1.8 days [0.1-13.0].

1. Discussion

EuDKA is a rare but important complication of diabetes associated with SGLT2i use. With increasing indications and frequency of use of SGLT2i, it is likely that SGLT2i-associated EuDKA will continue to in- crease in prevalence. We present a case series describing presentation characteristics, resource utilization, and safety outcomes of five cases of SGLT2i-associated EuDKA.

At our institution, it is standard to transfer nearly all ED patients with DKA to our ED-ICU [12], with 380 cases managed in a 3.5-year period [13]. Of these, we identified only 5 patients with SGLT2i-associated

EuDKA. This highlights the relative infrequency of this diagnosis. The rarity of EuDKA, in combination with its unintuitive relative normo- glycemia, may lead to initial misdiagnosis by treating clinicians. This may lead to morbidity and mortality if unaddressed.

Thus, Emergency Physicians (EPs) should be aware of the associa- tion between SGLT2i use and EuDKA. An appropriate work up should be pursued for patients taking an SGLT2i who present with symptoms suggestive of DKA, including nausea, vomiting, malaise, and altered mental status, or are noted to have an unexplained elevated anion gap metabolic acidosis. Recognition of this condition should prompt consul- tation with endocrinology and initiation of insulin infusion with con- comitant glucose administration to prevent hypoglycemia. Similar to standard hyperglycemic DKA management, it seems reasonable to delay insulin infusion in EuDKA until serum potassium is >3.3 mEq/L to avoid iatrogenic hypokalemia and arrythmias [1], although data re- garding nuanced management of EuDKA is sparse. All five patients in this case series had SGLT2i discontinued at discharge as recommended by the American Association of Clinical Endocrinologists and American College of Endocrinology [14]. EPs should be aware of these recommen- dations and should discontinue SGLT2i in patients with EuDKA in con- sultation with an endocrinologist.

Resource utilization and patient safety outcomes of a cohort of pa- tients with classic hyperglycemic DKA managed in the same ED-ICU set- ting have been reported previously [12,13]. The small sample size of the present case series of patients with EuDKA prevents reasonable statisti- cal comparisons between those with hyperglycemic DKA and EuDKA, though may prove hypothesis generating.

Our study was limited by multiple factors. The retrospective design limits interpretations to association rather than causation. The small sample size (n = 5) and single center setting limit generalizability of ob- served results. As we identified patients via an EHR search, it is possible that other patients with EuDKA from SGLT2i use were managed at our institution over the study period and not identified via this method.

These observations suggest patients with EuDKA may have a similar course and experience similar outcomes as patients with DKA. Given the small sample size, retrospective nature, and single center setting, these relative similarities can be viewed as hypothesis-generating rather than confirmatory, and should be interpreted accordingly. Future research is needed to further quantify outcomes of patients treated for EuDKA as compared to hyperglycemic DKA.

1. Conclusion

We report a case series of ED patients with EuDKA associated with SGLT2i use. Presentation characteristics, resource utilization, and safety outcomes appear largely similar to data previously published for a co- hort of classic DKA patients treated in the same setting. As EuDKA is an infrequently reported clinical entity, our report contributes to the un- derstanding of ED management and clinical course for these patients. Further data from larger observational studies are needed to clarify re- source utilization and safety outcomes in EuDKA. Emergency physicians should be aware of the association between SGLT2i use and EuDKA, and an appropriate work-up should be pursued for patients taking an SGLT2i who present with symptoms suggestive of DKA or are noted to have an unexplained elevated anion gap metabolic acidosis.

Declarations of Competing Interest

None.

Meeting Information

This work has not previously been presented or published elsewhere.

Funding Information

The study did not receive funding.

CRediT authorship contribution statement Henrique A. Puls: Conceptualization, Methodology, Formal analysis,

Investigation, Resources, Writing - original draft, Writing - review & editing, Visualization, Project administration. Nathan L. Haas: Concep- tualization, Methodology, Writing - original draft, Writing - review & editing, Supervision, Project administration. Brian Franklin: Resources, Writing - original draft, Visualization. Nik Theyyunni: Conceptualiza- tion, Resources, Writing - original draft. Carrie E. Harvey: Conceptuali- zation, Methodology, Resources, Writing - original draft, Supervision.

Acknowledgements

The authors would like to acknowledge the Joyce and Don Massey Family Foundation for their generous support in creation of the Joyce and Don Massey Family Foundation Emergency Critical Care Center. The authors would also like to acknowledge Stephanie Laurinec and Amanda Melvin for their administrative support and role in this project.

References

1. Kitabchi AE, Umpierrez GE, Miles JM, et al. Hyperglycemic crises in adult patients with diabetes. Diabetes Care. 2009;32(7):1335-43. https://doi.org/10.2337/dc09- 9032.
2. Plewa MC, Bryant M, King-Thiele R. Euglycemic Diabetic Ketoacidosis. StatPearls [In- ternet]. Treasure Island (FL): StatPearls Publishing; 2020 Apr 30 2020 Jan-. PMID: 32119457.
3. Meyer E, Gabb G, Jesudason D. SGLT2 inhibitor-associated euglycemic diabetic ketoacidosis: a south Australian clinical case series and Australian spontaneous ad- verse event notifications. Diabetes Care. 2018;41:e47-9.
4. Tahrani AA, Bailey CJ, Del Prato S, et al. Management of type 2 diabetes: new and fu- ture developments in treatment. Lancet. 2011;378(9786):182-97. https://doi.org/ 10.1016/S0140-6736(11)60207-9.
5. Raval AD, Vyas A. National Trends in diabetes medication use in the United States: 2008 to 2015. J Pharm Pract. 2020 Aug;33(4):433-42. https://doi.org/10.1177/

0897190018815048 Epub 2018 Dec 20 30572757.

1. Herring R, Russell-Jones DL. SGLT2 inhibitors in Type 1 diabetes: is this the future? Diabet Med. 2018;35(12):1642-3. https://doi.org/10.1111/dme.13791.
2. Packer M, Anker SD, Butler J. Et al; EMPEROR-reduced trial investigators. Cardiovas- cular and renal outcomes with Empagliflozin in heart failure. N Engl J Med. 2020 Oct 8;383(15):1413-24. https://doi.org/10.1056/NEJMoa2022190 Epub 2020 Aug 28 32

865377.

1. Heerspink HJL, Stefansson BV, Correa-Rotter R, et al; DAPA-CKD Trial Committees and Investigators. Dapagliflozin in patients with chronic kidney disease. N Engl J Med. 2020 Oct 8;383(15):1436-46. https://doi.org/10.1056/NEJMoa2024816 Epub

2020 Sep 24 32970396.

1. Yu X, Zhang S, Zhang L. Newer perspectives of mechanisms for Euglycemic diabetic ketoacidosis. Int J Endocrinol. 2018;2018:7074868.
2. Peters AL, Buschur EO, Buse JB, Cohan P, Diner JC, Hirsch IB. Euglycemic Diabetic Ketoacidosis: A potential complication of Treatment With Sodium-Glucose Cotransporter 2 Inhibition. Diabetes Care. 2015 Sep;38(9):1687-93. https://doi. org/10.2337/dc15-0843 Epub 2015 Jun 15. PMID: 26078479; PMCID: PMC4542270.
3. Gunnerson KJ, Bassin BS, Havey RA, et al. Association of an Emergency Department- Based Intensive Care Unit With Survival and Inpatient Intensive Care Unit Admis- sions. JAMA Netw Open. 2019 Jul 3;2(7):e197584. https://doi.org/10.1001/ jamanetworkopen.2019.7584 PMID: 31339545; PMCID: PMC6659143.
4. Haas NL, Gianchandani RY, Gunnerson KJ, et al. The two-bag method for treatment of diabetic ketoacidosis in adults. J Emerg Med. 2018 May;54(5):593-9. https:// doi.org/10.1016/j.jemermed.2018.01.004 Epub 2018 Apr 5 29628184.
5. Haas NL, Whitmore SP, Cranford JA, et al. An emergency department-based intensive care unit is associated with decreased hospital and intensive care unit utilization for diabetic ketoacidosis. J Emerg Med. 2020 Apr;58(4):620-6. https://doi.org/10.1016/ j.jemermed.2019.10.005 Epub 2019 Dec 13 31843318.
6. Handelsman Y, Henry RR, Bloomgarden ZT, et al. American Association of Clinical Endocrinologists and American College of endocrinology position statement on the association of SGLT-2 inhibitors and diabetic ketoacidosis. Endocr Pract. 2016 Jun; 22(6):753-62. https://doi.org/10.4158/EP161292.PS Epub 2016 Jun 1 27082665.