Use of Sodium Glucose Cotransporter 2 Inhibitors in the Hands of Cardiologists
With Great Power Comes Great Responsibility
This article has a correction. Please see:
- cardiovascular diseases
- diabetes mellitus
- heart failure
- prevention and control
- sodium-glucose transporter 2
The EMPA-REG OUTCOME trial (BI 10773 [Empagliflozin] Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients) showed beneficial effects of a sodium glucose cotransport-2 inhibitor (SGLT2i) on overall and cardiovascular mortality and heart failure (HF) hospitalizations in patients with type 2 diabetes mellitus.1 It is important to note that the benefits of empagliflozin were observed in patients across chronic kidney disease stages despite an expected attenuation of its hemoglobin A1c–lowering effects in patients with an estimated glomerular filtration rate <60 mL·min−1·1.73 m−2. While the mechanisms responsible for these benefits with an SGLT2i are being elucidated, the results of this pivotal study will inevitably affect clinical cardiology practice. Moreover, given the apparent glucose-independent effects of empagliflozin on cardiovascular outcomes,1 cardiologists may wish to recommend or prescribe an SGLT2i for patients with type 2 diabetes mellitus who have prevalent atherosclerotic cardiovascular disease. Recommendations for SGLT2is in cardiology guidelines will further empower frontline cardiologists. However, SGLT2is are relatively new medications, until now prescribed primarily by primary care clinicians and endocrinologists. Hence, many cardiologists are not yet familiar with the benefits and risks of these agents.
Accordingly, the aim of this Perspective is to provide an overview of the mechanism of action and side-effect profile of SGLT2is and specifically to provide guidance for SGLT2i use when combined with diuretics. We are concerned about the potential for untoward ramifications of broad-based, untargeted use of a new cardioprotective therapy with potential for additive side effects from concomitant cardiorenal therapies such as diuretics.2 The risk for such events is usually higher than reported in clinical trials of select patients and investigators, which may lead to a negative profile and lost opportunity.
The SGLT2is decrease hemoglobin A1c by ≈0.7% to 1.0% and induce weight loss through glycosuria.3 Concomitant natriuresis and osmotic diuresis by these agents contract plasma volume by ≈7%, contributing to the ≈5/2–mm Hg antihypertensive effects, 30% to 40% albuminuria reduction through intrarenal hemodynamic effects, decreased HF risk, and loop-diuretic sparing effects.1
The most common side effect of glycosuria is genitourinary infections, mostly genital mycotic infections but in some studies bacterial urinary tract infections, as well. Thus, when SGLT2is are prescribed, we advise preemptive counseling of common sense urinary hygiene, especially in those with prior mycotic genital infections and those at risk, including women and uncircumcised men.
There is also potential for volume depletion and altered renal function, especially in elderly patients with chronic kidney disease taking renin-angiotensin-aldosterone system inhibitors or loop diuretics. Natriuresis-related risks, however, were not reported in the first outcome trial1 but instead in postmarketing analyses, both subject to limitations for quantifying safety events. Nevertheless, the risk for effective circulating volume depletion is an issue to be appreciated and avoided.
An additional safety concern with SGLT2is is the development of diabetic ketoacidosis, which can present clinically with lower-than-expected blood glucose levels (10–15 mmol/L), thereby delaying the recognition of this complication. It is encountered mainly when the drugs are used off label in patients with type 1 diabetes mellitus but has also been reported in type 2 diabetes mellitus. To mitigate diabetic ketoacidosis risk, patients prescribed SGLT2is should be educated about common precipitants (periods of illness, reduced oral intake, reduced insulin doses, alcohol use), about regular home glucose monitoring, and about adjunctive point-of-care home ketone monitoring in appropriate clinical settings to detect and treat this complication as early as possible.
As a result of these potential adverse events, diabetes clinical practice guidelines have added SGLT2is to the list of “sick-day medications” that require special attention during intercurrent illnesses, including temporarily withholding these agents when appropriate,4 to avoid volume contraction and diabetic ketoacidosis. Physicians are encouraged to routinely provide safety information to their patients on the risk of dehydration, when to potentially hold the SGLT2i, when to increase glucose (or ketone) monitoring, and when to seek medical attention.4 We caution against using SGLT2is in patients with type 1 diabetes mellitus until further data are available, but if an SGLT2i is used in this setting (or in patients with type 2 diabetes mellitus who are felt to be ketosis prone), patients should be counseled to perform ketone monitoring when suggestive symptoms develop and especially on sick days.
Beyond this routine advice is clinician decision making on diuretic management when SGLT2is are initiated (Figure). Background diuretic management has not heretofore been mandated in SGLT2i clinical trial protocols, so it is unclear whether such adjustments might have affected the cardioprotective effects observed in EMPA-REG OUTCOME. However, rational consideration of diuretic use seems prudent because SGLT2i with empagliflozin exerts a loop-diuretic sparing effect.1 Diuretic management strategies for patients with overt hypovolemia or hypervolemia are suggested at the top of the Figure. In euvolemic patients, it is important to assess blood pressure. If a patient is euvolemic and hypotensive, we advise caution in initiating SGLT2i, despite no reported blood pressure thresholds for holding therapy in outcome trials. We suggest that euvolemic normotensive patients remain on thiazide therapy. This suggestion is based on data that concomitant thiazide therapy did not further reduce blood pressure beyond the reduction achieved with SGLT2i.3 In contrast, in patients taking loop diuretics, reducing the dose by 50% can be considered. Last, in patients with hypertensive euvolemia, diuretics can be continued and monitored.
We emphasize that SGLT2i should not be started in patients with acute decompensated HF or dynamic volume status outside of research settings until further data are available.5 Similarly, there are limited efficacy and safety data for SGLT2i among patients with severe systolic dysfunction or treatment with sacubitril/valsartan, an HF drug that results in natriuresis. In both scenarios, we advise close monitoring of blood pressure after SGLT2i initiation. With the caveat that mortality benefits in EMPA-REG OUTCOME were observed down to an estimated glomerular filtration rate of 30 mL·min−1·1.73 m−2, current labeling states that empagliflozin/canagliflozin should be discontinued when the estimated glomerular filtration rate is <45 mL·min−1·1.73 m−2 and dapagliflozin stopped when it is <60 mL·min−1·1.73 m−2. Further caution is required in patients with a history of recurrent or severe genital tract infections, urosepsis, or chronic hyponatremia because of the risk of compounding the impact of diuretics on antidiuretic hormone release.
In conclusion, in selected patients, substantial cardiovascular risk reduction with SGLT2i may be achieved, but clinicians should proceed with caution in high-risk patients, especially those taking diuretics. Future mechanistic studies and trials should endeavor to study the risk of volume depletion in patients with or at risk for HF on diuretic agents.
The authors thank Muhammad Mamdani and Silvio Inzucchi for their helpful review of earlier versions of this manuscript. No compensation was received from any individual for their assistance.
Sources of Funding
Dr Cherney is supported by funding from the Canadian Institutes for Health Research. Dr Udell is supported by a Heart and Stroke Foundation National New Investigator/Ontario Clinician Scientist Award and by funding from the National Heart, Lung, and Blood Institute (U01HL130163-01); Heart and Stroke Foundation of Canada (G-15-0009034; ERLI024); Ontario Ministry of Research and Innovation (ER15-11–037); Women’s College Research Institute and Department of Medicine, Women’s College Hospital; Peter Munk Cardiac Center, University Health Network; and Department of Medicine and Heart and Stroke Richard Lewar Center of Excellence in Cardiovascular Research, University of Toronto.
All authors have completed and submitted the International Committee of Medical Journal Editors Form for Disclosure of Potential Conflicts of Interest. Dr Cherney has received speaker/consultant fees from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Sanofi, and Merck and has received operational funding from AstraZeneca, Boehringer Ingelheim, Janssen, and Merck. Dr Udell has received consulting fees from Amgen, Boehringer-Ingelheim, Janssen, Merck, and Sanofi Pasteur, as well as honoraria from AstraZeneca, Janssen, Novartis (steering committee), and Servier.
The opinions in this article are not necessarily those of the editors or of the American Heart Association.
Circulation is available at http://circ.ahajournals.org.
- © 2016 American Heart Association, Inc.
- Zinman B,
- Wanner C,
- Lachin JM,
- Fitchett D,
- Bluhmki E,
- Hantel S,
- Mattheus M,
- Devins T,
- Johansen OE,
- Woerle HJ,
- Broedl UC,
- Inzucchi SE
- Heerspink HJ,
- Perkins BA,
- Fitchett DH,
- Husain M,
- Cherney DZ
- Harper W,
- Clement M,
- Goldenberg R,
- Hanna A,
- Main A,
- Retnakaran R,
- Sherifali D,
- Woo V,
- Yale JF,
- Cheng AY
- Sairaku A,
- Nakano Y,
- Kihara Y