5
Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:16381642 | DOI: 10.2337/dc15-1380 THE CASE AT HAND Recently, the U.S. Food and Drug Ad- ministration (FDA) issued a Drug Safety Communication that warns of an in- creased risk of diabetic ketoacidosis (DKA) with uncharacteristically mild to moderate glucose elevations (euglycemic DKA [euDKA]) associated with the use of all the approved sodiumglucose cotransporter 2 (SGLT2) inhibitors (1). This Communication was based on 20 clinical cases requiring hospitalization captured between March 2013 and June 2014 in the FDA Adverse Event Re- porting System database. The scarce clinical data provided suggested that most of the DKA cases were reported in patients with type 2 diabetes (T2D), for whom this class of agents is indicated; most likely, however, they were insulin- treated patients, some with type 1 dia- betes (T1D). The FDA also identi ed potential triggering factors such as inter- current illness, reduced food and uid intake, reduced insulin doses, and his- tory of alcohol intake. The following month, at the request of the European Commission, the European Medicines Agency (EMA) announced on 12 June 2015 that the Pharmacovigilance Risk As- sessment Committee has started a review of all of the three approved SGLT2 inhib- itors (canagliozin, dapagliozin, and em- pagliozin) to evaluate the risk of DKA in T2D (2). The EMA announcement claimed that as of May 2015 a total of 101 cases of DKA have been reported worldwide in EudraVigilance in T2D patients treated with SGLT2 inhibitors, with an estimated exposure over 0.5 million patient-years. No clinical details were provided except for the mention that all cases were seri- ous and some required hospitalisation. Al- though [DKA] is usually accompanied by high blood sugar levels, in a number of these reports blood sugar levels were only moderately increased(2). With this background, it is very timely that in this issue of Diabetes Care there are two articles on this subject. Erondu et al. (3) report cases of DKA in T2D from a large clinical development pro- gram and Peters et al. (4) discuss cases from clinical practice observations of T1D and T2D patients. It is not unusual that serious safety issues related to a new drug go unde- tected during the relatively short clinical development programs for regulatory drug approval. This is particularly true when the safety issue is unexpected, oc- curring as an off-target effect, or only emerges once the drug is used widely. If serious enough, the issue may require a label warning and a mitigation plan or even consideration of drug withdrawal. DKA is an overt serious clinical condition that may be missed only if presenting with mild to moderate hyperglycemia, as it may be the case with use of SGLT2 inhibitors, which could delay diagnosis and treatment and even accelerate the progressive metabolic deterioration. In- terestingly, the large clinical development programs of the three marketed SGLT2 inhibitors, comprising .40,000 T2D pa- tients, bore no clear signal of DKA. Erondu et al. (3), representing Janssen, the man- ufacturer of canagliozin, report a rela- tively low frequency of DKA (15 cases, 12 on canagliozin and 3 still blinded in the CANagliozin cardioVascular Assess- ment Study [CANVAS]) detected in a ret- rospective analysis of 17,596 participants in the development program up to May 2015. The estimated incidence ratesd0.5, 0.8, and 0.2 per 1,000 patient-years with canagliozin 100 mg, canagli ozin 300 mg, and comparator, respectivelydif under- whelming, are double with the SGLT2 inhibitor. Upon our inquiry, the other two manufacturers of approved SGLT2 inhibitors, AstraZeneca and Boehringer Ingelheim, provided preliminary (unpub- lished) gures that are even lower than the Janssen data. In more than 18,000 pa- tients exposed to dapagliozin in the ran- domized controlled T2D study program, including DECLARE (Dapagliozin Effect on Cardiovascular Events), the frequency of reported events suggestive of DKA (blinded and unblinded events) is less than 0.1%. Similarly in DECLARE, aim- ing for 17,150 patients randomized to dapagliozin or placebo, the total number 1 Dallas Diabetes and Endocrine Center at Medical City, Dallas, TX 2 Institute of Clinical Physiology, Consiglio Nazionale delle Ricerche, Pisa, Italy Corresponding author: Julio Rosenstock, [email protected]. © 2015 by the American Diabetes Association. Readers may use this article as long as the work is properly cited, the use is educational and not for prot, and the work is not altered. See accompanying articles, pp. 1680 and 1687. Julio Rosenstock 1 and Ele Ferrannini 2 1638 Diabetes Care Volume 38, September 2015 COMMENTARY

Euglycemic Diabetic Ketoacidosis: A Predictable ...Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:1638–1642

  • Upload
    others

  • View
    11

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Euglycemic Diabetic Ketoacidosis: A Predictable ...Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:1638–1642

Euglycemic Diabetic KetoacidosisA Predictable Detectable andPreventable Safety Concern WithSGLT2 InhibitorsDiabetes Care 2015381638ndash1642 | DOI 102337dc15-1380

THE CASE AT HAND

Recently the US Food and Drug Ad-ministration (FDA) issued a Drug SafetyCommunication that warns of an in-creased risk of diabetic ketoacidosis(DKA) with uncharacteristically mild tomoderate glucose elevations (euglycemicDKA [euDKA]) associated with the useof all the approved sodiumndashglucosecotransporter 2 (SGLT2) inhibitors (1)This Communication was based on 20clinical cases requiring hospitalizationcaptured between March 2013 andJune 2014 in the FDA Adverse Event Re-porting System database The scarceclinical data provided suggested thatmost of the DKA cases were reported inpatients with type 2 diabetes (T2D) forwhom this class of agents is indicatedmost likely however they were insulin-treated patients some with type 1 dia-betes (T1D) The FDA also identifiedpotential triggering factors such as inter-current illness reduced food and fluidintake reduced insulin doses and his-tory of alcohol intake The followingmonth at the request of the EuropeanCommission the European MedicinesAgency (EMA) announced on 12 June2015 that the Pharmacovigilance Risk As-sessment Committee has started a reviewof all of the three approved SGLT2 inhib-itors (canagliflozin dapagliflozin and em-pagliflozin) to evaluate the risk of DKA inT2D (2) The EMA announcement claimed

that as of May 2015 a total of 101 casesof DKA have been reported worldwide inEudraVigilance in T2D patients treatedwith SGLT2 inhibitors with an estimatedexposure over 05 million patient-yearsNo clinical details were provided exceptfor the mention that ldquoall cases were seri-ous and some required hospitalisation Al-though [DKA] is usually accompanied byhigh blood sugar levels in a number ofthese reports blood sugar levels wereonly moderately increasedrdquo (2)

With this background it is very timelythat in this issue of Diabetes Care thereare two articles on this subject Eronduet al (3) report cases of DKA in T2Dfrom a large clinical development pro-gram and Peters et al (4) discuss casesfrom clinical practice observations ofT1D and T2D patients

It is not unusual that serious safetyissues related to a new drug go unde-tected during the relatively short clinicaldevelopment programs for regulatorydrug approval This is particularly truewhen the safety issue is unexpected oc-curring as an off-target effect or onlyemerges once the drug is used widely Ifserious enough the issue may require alabel warning and a mitigation plan oreven consideration of drug withdrawalDKA is an overt serious clinical conditionthat may be missed only if presentingwith mild to moderate hyperglycemiaas it may be the case with use of SGLT2

inhibitors which could delay diagnosisand treatment and even accelerate theprogressive metabolic deterioration In-terestingly the large clinical developmentprograms of the three marketed SGLT2inhibitors comprising 40000 T2D pa-tients bore no clear signal of DKA Eronduet al (3) representing Janssen the man-ufacturer of canagliflozin report a rela-tively low frequency of DKA (15 cases12 on canagliflozin and 3 still blinded inthe CANagliflozin cardioVascular Assess-ment Study [CANVAS]) detected in a ret-rospective analysis of 17596 participantsin the development program up to May2015 The estimated incidence ratesd0508 and 02 per 1000 patient-years withcanagliflozin 100mg canagliflozin 300mgand comparator respectivelydif under-whelming are double with the SGLT2inhibitor Upon our inquiry the othertwo manufacturers of approved SGLT2inhibitors AstraZeneca and BoehringerIngelheim provided preliminary (unpub-lished)figures that are even lower than theJanssen data In more than 18000 pa-tients exposed to dapagliflozin in the ran-domized controlled T2D study programincluding DECLARE (Dapagliflozin Effecton Cardiovascular Events) the frequencyof reported events suggestive of DKA(blinded and unblinded events) is lessthan 01 Similarly in DECLARE aim-ing for 17150 patients randomized todapagliflozin orplacebo the total number

1Dallas Diabetes and Endocrine Center at Medical City Dallas TX2Institute of Clinical Physiology Consiglio Nazionale delle Ricerche Pisa Italy

Corresponding author Julio Rosenstock juliorosenstockdallasdiabetescom

copy 2015 by the American Diabetes Association Readers may use this article as long as the work is properly cited the use is educational and not for profitand the work is not altered

See accompanying articles pp 1680 and 1687

Julio Rosenstock1 and Ele Ferrannini2

1638 Diabetes Care Volume 38 September 2015

COMMEN

TARY

of reported blinded events of potentialDKAs is less than 01 (E JohnssonAstraZeneca personal communication)In a retrospective analysis of randomizedphase2and3empagliflozin trials (13000T2Dparticipants) therewere eight eventsconsistent with DKA with no imbalanceobserved between patients treated withempagliflozin 10 mg (two events) empa-gliflozin 25 mg (one event) and placebo(five events) In the cardiovascular out-come trial EMPA-REG Outcome withapproximately 7000 patients the fre-quency of reported blinded events of DKAis less than 01 (U Broedl BoehringerIngelheim personal communication)Of note the canagliflozin data reported

by Erondu et al (3) appeared to have agreater incidence of DKA but 6 out of the12 cases had evidence of latent autoim-mune diabetes in adults or T1D or testedpositive for GAD65 antibodies and per-haps some of the other cases may havebeen T2D misdiagnoses And even if thediagnosis was correct most of the patientswere on insulin treatment andwere part ofCANVAS suggesting a more advancedT2D stage with significant b-cell failureThe FDA did acknowledge that some

of the cases occurred in T1D where in-creasing off-label use of SGLT2 inhibi-tors has been observed most likely dueto the favorable insulin-independentglucose-lowering and weight-loss effectsIndeed preliminary proof-of-concept pi-lot studies in T1D have reported improve-ments in short-term glucose control withless glucose variability weight loss andlower insulin doses (5ndash7) Social mediahave disseminated initial favorable expe-riences and could have contributed tothe raised expectations leading tomany T1D patients discussing with theirphysicians the addition of an SGLT2 in-hibitor in an attempt to ameliorate theirdiabetes control Indeed despite new in-sulin analogs and technological improve-ments in insulin delivery devices andglucose monitoring systems T1D re-mains an intrusive and challenging dis-ease fraught with wide glucose swingsand hypoglycemic episodes that frustratepatients families and health care pro-viders There is no better example thanthe Diabetes Control and ComplicationsTrial (DCCT) Despite 6 years of monthlyvisits with outstanding diabetes treat-ment teams with limitless resources toachieve an HbA1c of 7 the T1D patientsin the intensive intervention group

escalated back to an HbA1c of 8 in theposttrial years (8) In a recent report fromthe T1D Exchange clinic registry (whichprovides the best cross-sectional USdata) the average HbA1c was 8 andonly 30 achieved a goal HbA1c of7severe hypoglycemia occurred in 9ndash20of patients per year depending on ageand diabetes duration overweightobesity was present in 68 of patientsand interestingly DKA still occurred at10 per year in patients aged 13ndash26years and at 4ndash5 per year in the olderpatients (9) Therefore it is not surpris-ing that given the burden of T1D and itschallenging unmet needs the pharma-cological properties of SGLT2 inhibitorsprompted clinical development pro-grams seeking regulatory approval andattracted off-label use in T1D The datapresented in this issue of Diabetes Carein T2D (3) and in particular the casesassociated with T1D as presented byPeters et al (4) do provide a good op-portunity to discuss how these agentsmodulate the pathophysiology leadingto DKA Thus it is in this context thatwe need to analyze the potential prob-lem of euDKA associated with SGLT2 in-hibitors to provide a more realistic andpractical perspective

THE PATHOPHYSIOLOGY

Ketosis results from restriction of carbo-hydrate usage with increased relianceon fat oxidation for energy productionThe pathogenesis of DKA is well estab-lished (10) Briefly absolute insulin de-ficiency leads to reduced glucoseutilization and enhanced lipolysis in-creased delivery of free fatty acids(FFAs) to the liver coupled with raisedglucagon levels promotes FFA oxidationand production of ketone bodies Inboth T1D and T2D DKA presents withmarked hyperglycemia (250 mgdLtypically 350ndash800mgdL) profuse glycos-uria (2ndash4 mg z min21 z kg21) and hyper-ketonemia (plasma b-hydroxybutyrate42ndash110 mmolL) (1112) The hypergly-cemia of DKA is associated with extremeinsulin resistance manifesting itself asmarkedly (70) reduced tissue glucosedisposal and increased endogenous glu-cose production (EGP) (12)

euDKA was originally defined as DKAwith plasma glucose levels300 mgdLoccurring in young T1D patients two-thirds of whom were female (13) Theprimary cause was reduced availability

of carbohydrate possibly in conjunctionwith reduced insulin dose The euDKAreported in T2D patients with SGLT2 in-hibitor treatment has a different originFull-dose SGLT2 inhibition induces arapid increase in urinary glucose excre-tion ranging 50ndash100 gday equally inmen and women and lasting slightly lon-ger than 24 h (14) In a typical 60-year-oldoverweight T2D patient (BMI 28 kgm2)consuming 50 of daily calories ascarbohydrate (15) this glucose lossamounts to 17ndash34 of estimated carbo-hydrate intake in men and 22ndash44 inwomen Of note in a comparative studyof Japanese and European T2D patientstreated with an SGLT2 inhibitor urinaryglucose excretion was if anything largerin the former (averaging 110 gday)than in the latter (60 gday) groups (16)thus in the Japanese group (BMI 25 kgm2)the glucose loss through the urine repre-sented 47 of estimated daily carbohy-drate intake in men and 57 in womenIn general depending on body size glo-merular filtration rate and degree of hy-perglycemia SGLT2-induced glucose losscan make up a substantial fraction ofdaily carbohydrate availability

Abstracting from a study in well-controlled drug-na ıve or metformin-treated T2D patients on chronic therapywith an SGLT2 inhibitor (17) plasma glu-cose levels decreased by 20ndash25 mgdLboth in the overnight fasted state andfollowing a mixed meal As glucose is thechief stimulus for insulin release under allcircumstances plasma insulin levels also fell(by 10 pmolL fasting and 60 pmolLpostmeal) In contrast plasma glucagonconcentrations increased significantlypartly because of a diminished paracrineinhibition by insulin (18) and possibly alsobecause of decreased SGLT2-mediatedglucose transport intoa-cells (19) As a con-sequence thecalculatedprehepatic insulin-to-glucagon molar concentration ratiodropped from9 to 7molmol in the fastingstate and from 29 to 24 molmol duringthe meal This hormonal shift which re-leases inhibition of gluconeogenesis inthe liver (20) augmented EGP both inthe fasting state and during the meal(17) Insulin sensitivity however wasimproved as also shown with the useof the euglycemic insulin clamp as a re-sult of attenuated glucotoxicity (21)

The difference in the pathophysiologyof DKA versus SGLT2 inhibitorndashinducedeuDKA is schematized in Fig 1 In euDKA

carediabetesjournalsorg Rosenstock and Ferrannini 1639

insulin deficiency and insulin resistanceare milder (and insulin resistance mayactually be improved) therefore glu-cose overproduction and underutiliza-tion are quantitatively lesser than inDKA More importantly renal glucoseclearance (ie the ratio of glycosuriato prevailing glycemia) is twice as largewith euDKA than with DKA In factfrom previous studies of patients ad-mitted with DKA it can be calculatedthat renal glucose clearance averaged03 mL z min21 z kg21 (12) whereas inT2D patients it rose from a near-negligible value in the baseline studyto 06 mL z min21 z kg21 with SGLT2treatment (17) Thus it is the entityof glycosuria viz the height of hyper-glycemia that marks the difference be-tween the two metabolic states Thisdifference can actually be amplified asDKA frequently occurs in patients withimpaired renal function (and hence lessglycosuria) (11) while SGLT2 inhibitorsmay be used in patients with glomeru-lar hyperfiltration (and more abundantglycosuria) (5)Ketoacidosis follows with the same

sequence of events in euDKA as inDKA Thus in SGLT2-treated T2D patients(17) the lower insulin-to-glucagon ratiostimulated lipolysis (circulating FFAswere 40 higher during the meal) andenhanced lipid oxidation (by 20 on av-erage) at the expense of carbohydrateoxidation (which fell by 60) In theface of lower substrate (glucose)

concentrations nonoxidative glucosedisposal (ie glycogen synthesis andlactate release) also fell by 15 Theaugmented FFA delivery to the liver re-sulted in mild stimulation of ketogene-sis whereby both fasting and meanpostmeal b-hydroxybutyrate levelsrosetwofold higher than in the baselinestudy (though not exceeding 1 mmolL)conversely plasma lactate levels de-creased 20 a readout of reducedcarbohydrate utilization (E Ferranniniet al unpublished data) Had insulindeficiency been more profounddascan happen in T1D patientsdor hadcarbohydrate availability been drasti-cally restricted this mild ketosis wouldhave evolved toward ketoacidosis withdecreased blood pH and bicarbonateand increased anion gap (12) It mustbe noted that in the only study thatlooked at insulin administration (22)ddating back to 1951dit was found thatinsulin decreased the transport maxi-mum for glucose in patients with diabe-tes implying that starting or escalatingexogenous insulin treatment would in-duce glycosuria in its own right Whilethese studies have not been repeatedand the mechanism of this insulin effecthas not been investigated it is never-theless intriguing that insulin may in-tensify SGLT2-induced glycosuria

All in all euDKA is pathophysiologi-cally similar to DKA except for thecircumstancedSGLT2-induced glycosuriadthat ldquoartificiallyrdquo lowers plasma glucose

levels and predisposes to increasedketogenesis

THE CLINICAL LESSON

The evidence reviewed above suggeststhat the risk of bona fide euDKA (and notsimple ketosis) in T2D related to the useof SGLT2 inhibitors will probably turnout to be very low with an ldquoacceptablerdquofrequency Still physicians and patientsneed to be made aware that such riskmay be increased in long-standing T2Dpatients with marked b-cell insuffi-ciency or in latent autoimmune diabetesin adults with rapid evolution towardT1D and during prolonged starvationafter surgery or during intercurrentillness In T1D however the euDKArisk appears to be more concrete forreasons entirely within the pathophysi-ology 1) in T1D patients hyperglycemiatypically is higher than in T2D patients2) in early T1D glomerular filtration ratemay be increased 3) insulin may en-hance the effect of SGLT2 inhibition onglycosuria and 4) changes in insulindose are not infrequent and may be in-appropriate for the amount and kind ofcarbohydrate intake

We submit that this potential compli-cation related to SGLT2 inhibition is pre-dictable detectable and preventable (ormitigable) so that the balance of benefitsand risks favors the use of SGLT2 inhibi-tors in the T1D population which is indesperate need of adjunct therapies Itis predictable because the persistent gly-cosuria induced by SGLT2 inhibition setsoff a sequence of metabolic changes thatare obligatory quantitative consequencesof a large glucose subtraction from thebody glucose pool In particular enhancedketogenesis is seen already even in non-diabetic subjects receiving SGLT2 inhibi-tors and is inscribed on already modestlyraised plasma b-hydroxybutyrate levels inpatients with diabetes (E Ferrannini et alunpublished observations) This back-ground ketonemia is asymptomatic andclinically irrelevant in most T2D patientsbut is certainly more of a concern in T1Dpatients who are already prone to de-velop ketosis under circumstances ofreduced insulin doses stress and inter-current illnesses and following a hypo-glycemic episode and during prolongedfasting or starvation

Thus one can envision a sequence ofclinical events that could evolve into afull-blown episode of DKA (Fig 2) First

Figure 1mdashEssential pathophysiology of DKA and euDKA consequent of the use of SGLT2 inhib-itors TGD tissue glucose disposal UGCr urinary glucose clearance rate

1640 Commentary Diabetes Care Volume 38 September 2015

inappropriate reductions of insulin dosesor any factor that may increase insulindemand such as stress a sick day oreven alcohol intake may induce hyper-ketonemia Under these circumstancesinitially patients may just not feel wellor experience some malaise and perhapsmild nausea with no vomiting Their firstimpulse is to check their blood glucosebecause of the persistent glycosuriaglycemia will be only mildly elevated sothat they would tend to reduce or with-hold insulin and avoid eating These ma-neuverswill accelerate ketone productionand metabolic decompensation towardDKA Themetabolic picture will be furthercompounded by the volume depletioncaused by the persistent glycosuria andvomitingThe cases reported by Peters et al (4)

exemplify some of the factors that trig-gered DKA most commonly they wereinsulin reductions low caloric and fluidintake intercurrent illness and alcoholuse Timewas wasted because of delayeddiagnosis due todeceptively ldquoacceptablerdquoblood glucose levels Peters et al (4) doprovide a further understanding of theclinical clues that can contribute to theearly detection and help raise awarenessof the potential for SGLT2 inhibition tocause euDKA in T1D We believe thateuDKA is in fact easily detectable becausereliable tools are currently available tomonitor ketonuria and ketonemia andshould be recommended to be used at

any time an SGLT2 inhibitorndashtreated pa-tient feels unwell regardless of the ambi-ent glucose levels This should be part ofany educational element for those treatedwith an SGLT2 inhibitor If detectablethen euDKA is preventable because de-tection of significant ketonuria andorketonemia any time symptoms such asnausea andor vomitingdor even justmalaisedappear especially after alcoholintake or a recent cut in insulin dose canprompt advice to maintain vigorous fluidintake and to consume carbohydrates toallow at least full-dose insulin therapy un-til the ketosis resolves Patients shouldtemporarily stop the SGLT2 inhibitor con-tact their medical provider and take sup-plemental boluses of rapid insulin alongwith liquids and carbohydrates Even ifpatients are unable to adjust the insulindose euDKA can bemitigable by drinkingand eating as tolerated without fear ofhyperglycemia and seeking prompt med-ical attention for parenteral fluid replace-ment and insulin therapy

In any event T1D patients who chooseto take this medication off-label shouldsign an ad hoc informed consent thatmakes them fully aware of the potentialfor euDKA the precipitating factors thewarning symptoms and signs and thepreventative measures to adopt

CONCLUSIONS

The ongoing long-term randomizedplacebo-controlled studies will provide

the necessary information on the safetyand efficacy of SGLT2 inhibitors in T1D(as well as insulin-treated T2D) Regula-tory scrutiny of these compounds willbalance their beneficial impact on over-all glycemic control glycemic variabilityand weight management against therisk of hypoglycemia and overall safetyincluding risk of euDKA A reduction ininsulin dose should not be regarded as apositive outcome in itself and should beachieved by slow gentle decrements si-multaneously to avoid hypoglycemiaand sliding toward euDKA Hopefullythese clinical development programswill quickly expand so as to offer to pa-tients and physicians a potential adjunctagainst the day-to-day managementchallenges of such a demanding disease

Duality of Interest JR has participated inadvisory boards and received honoraria or con-sulting fees fromMerck Sanofi Novo Nordisk EliLilly MannKind Corporation GlaxoSmithKlineTakeda Daiichi Sankyo Boehringer IngelheimJanssen Lexicon and Intarcia and has receivedresearch grants from Merck Pfizer Sanofi NovoNordisk Eli Lilly GlaxoSmithKline Takeda NovartisAstraZeneca Janssen Daiichi Sankyo MannKindCorporation Bristol-Myers Squibb BoehringerIngelheim Lexiconand Intarcia EFhasbeenamem-ber of the scientific advisory board or a speakerfor Boehringer Ingelheim Merck AstraZenecaEli Lilly Johnson amp Johnson Takeda Novo NordiskGlaxoSmithKline and Sanofi and has received re-search grant support from Eli Lilly and BoehringerIngelheim No other potential conflicts of inter-est relevant to this article were reported

References1 US Food and Drug Administration DrugSafety Communication FDA warns that SGLT2 in-hibitors for diabetes may result in a serious con-dition of too much acid in the blood [Internet] 15May 2015 Available from httpwwwfdagovdownloadsDrugsDrugSafetyUCM446954pdfAccessed 22 June 20152 European Medicines Agency Review of dia-betes medicines called SGLT2 inhibitors startedrisk of diabetic ketoacidosis to be examined [In-ternet] 12 June 2015 Available from httpwwwemaeuropaeudocsen_GBdocument_libraryReferrals_documentSGLT2_inhibitors__20Procedure_startedWC500187926pdf Accessed 22June 20153 Erondu N Desai M Ways K Meininger GDiabetic ketoacidosis and related events in thecanagliflozin type 2 diabetes clinical programDiabetes Care 2015381680ndash16864 Peters AL Buschur EO Buse JB Cohan PDiner JC Hirsch IB Euglycemic diabetic ketoaci-dosis a potential complication of treatmentwith sodiumndashglucose cotransporter 2 inhibi-tion Diabetes Care 2015381687ndash16935 Perkins BA Cherney DZ Partridge H et alSodium-glucose cotransporter 2 inhibition andglycemic control in type 1 diabetes results of an

Figure 2mdashDemonstration of the cascade of clinical events and metabolic changes that contrib-ute sequentially to progressive clinical deterioration and development of full-blown episodes ofeuDKA BG blood glucose CHO carbohydrate TGs triglycerides

carediabetesjournalsorg Rosenstock and Ferrannini 1641

8-week open-label proof-of-concept trial Dia-betes Care 2014371480ndash14836 Henry RR Rosenstock J Edelman S et al Ex-ploring the potential of the SGLT2 inhibitordapagliflozin in type 1 diabetes a randomizeddouble-blind placebo-controlled pilot studyDiabetes Care 201538412ndash4197 Sands AT Zambrowicz BP Rosenstock Jet al Sotagliflozin a dual SGLT1 and SGLT2inhibitor as adjunct therapy to insulin in type1 diabetes Diabetes Care 2015381181ndash11888 Nathan DM DCCTEDIC ResearchGroup TheDiabetes Control and Complications TrialEpidemiology of Diabetes Interventions and Com-plications study at 30 years overview DiabetesCare 2014379ndash169 Miller KM Foster NC Beck RW et al T1DExchange Clinic Network current state of type 1diabetes treatment in the US updated datafrom the T1D Exchange clinic registry DiabetesCare 201538971ndash97810 Kitabchi AE Umpierrez GE Murphy MB Di-abetic ketoacidosis and hyperosmolar state InInternational Textbook of Diabetes Mellitus 4thed DeFronzo RA Ferrannini E Zimmet P Alberti

KGMM Eds New York John Wiley amp Sons Ltd2015 p 799ndash81411 Barrett EJ DeFronzo RA Bevilacqua SFerrannini E Insulin resistance in diabetic keto-acidosis Diabetes 198231923ndash92812 Luzi L Barrett EJ Groop LC Ferrannini EDeFronzo RA Metabolic effects of low-dose in-sulin therapy on glucose metabolism in diabeticketoacidosis Diabetes 1988371470ndash147713 Munro JF Campbell IW McCuish ACDuncan LJP Euglycaemic diabetic ketoacidosisBMJ 19732578ndash58014 Sha S Devineni D Ghosh A et alCanagliflozin a novel inhibitor of sodium glucoseco-transporter 2 dose dependently reduces cal-culated renal threshold for glucose excretion andincreases urinary glucose excretion in healthysubjects Diabetes Obes Metab 201113669ndash67215 Hall KD Chow CC Estimating changes infree-living energy intake and its confidence in-terval Am J Clin Nutr 20119466ndash7416 Ferrannini E Veltkamp SA Smulders RAKadokura T Renal glucose handling impact ofchronic kidney disease and sodium-glucose co-transporter 2 inhibition in patients with type 2diabetes Diabetes Care 2013361260ndash1265

17 Ferrannini E Muscelli E Frascerra S et alMetabolic response to sodium-glucose cotrans-porter 2 inhibition in type 2 diabetic patients JClin Invest 2014124499ndash50818 Maruyama H Hisatomi A Orci L GrodskyGM Unger RH Insulin within islets is a physio-logic glucagon release inhibitor J Clin Invest1984742296ndash229919 Bonner C Kerr-Conte J Gmyr V et al In-hibition of the glucose transporter SGLT2 withdapagliflozin in pancreatic alpha cells triggersglucagon secretion Nat Med 201521512ndash51720 Burgess SC Regulation of glucose metabo-lism in liver In International Textbook of Diabe-tes Mellitus 4th ed DeFronzo RA Ferrannini EZimmet P Alberti KGMM Eds New York JohnWiley amp Sons Ltd 2015 p 193ndash21021 Merovci A Solis-Herrera C Daniele G et alDapagliflozin improves muscle insulin sensitiv-ity but enhances endogenous glucose produc-tion J Clin Invest 2014124509ndash51422 Farber SJ Berger EY Earle DP Effect of di-abetes and insulin of the maximum capacity ofthe renal tubules to reabsorb glucose J Clin In-vest 195130125ndash129

1642 Commentary Diabetes Care Volume 38 September 2015

Page 2: Euglycemic Diabetic Ketoacidosis: A Predictable ...Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:1638–1642

of reported blinded events of potentialDKAs is less than 01 (E JohnssonAstraZeneca personal communication)In a retrospective analysis of randomizedphase2and3empagliflozin trials (13000T2Dparticipants) therewere eight eventsconsistent with DKA with no imbalanceobserved between patients treated withempagliflozin 10 mg (two events) empa-gliflozin 25 mg (one event) and placebo(five events) In the cardiovascular out-come trial EMPA-REG Outcome withapproximately 7000 patients the fre-quency of reported blinded events of DKAis less than 01 (U Broedl BoehringerIngelheim personal communication)Of note the canagliflozin data reported

by Erondu et al (3) appeared to have agreater incidence of DKA but 6 out of the12 cases had evidence of latent autoim-mune diabetes in adults or T1D or testedpositive for GAD65 antibodies and per-haps some of the other cases may havebeen T2D misdiagnoses And even if thediagnosis was correct most of the patientswere on insulin treatment andwere part ofCANVAS suggesting a more advancedT2D stage with significant b-cell failureThe FDA did acknowledge that some

of the cases occurred in T1D where in-creasing off-label use of SGLT2 inhibi-tors has been observed most likely dueto the favorable insulin-independentglucose-lowering and weight-loss effectsIndeed preliminary proof-of-concept pi-lot studies in T1D have reported improve-ments in short-term glucose control withless glucose variability weight loss andlower insulin doses (5ndash7) Social mediahave disseminated initial favorable expe-riences and could have contributed tothe raised expectations leading tomany T1D patients discussing with theirphysicians the addition of an SGLT2 in-hibitor in an attempt to ameliorate theirdiabetes control Indeed despite new in-sulin analogs and technological improve-ments in insulin delivery devices andglucose monitoring systems T1D re-mains an intrusive and challenging dis-ease fraught with wide glucose swingsand hypoglycemic episodes that frustratepatients families and health care pro-viders There is no better example thanthe Diabetes Control and ComplicationsTrial (DCCT) Despite 6 years of monthlyvisits with outstanding diabetes treat-ment teams with limitless resources toachieve an HbA1c of 7 the T1D patientsin the intensive intervention group

escalated back to an HbA1c of 8 in theposttrial years (8) In a recent report fromthe T1D Exchange clinic registry (whichprovides the best cross-sectional USdata) the average HbA1c was 8 andonly 30 achieved a goal HbA1c of7severe hypoglycemia occurred in 9ndash20of patients per year depending on ageand diabetes duration overweightobesity was present in 68 of patientsand interestingly DKA still occurred at10 per year in patients aged 13ndash26years and at 4ndash5 per year in the olderpatients (9) Therefore it is not surpris-ing that given the burden of T1D and itschallenging unmet needs the pharma-cological properties of SGLT2 inhibitorsprompted clinical development pro-grams seeking regulatory approval andattracted off-label use in T1D The datapresented in this issue of Diabetes Carein T2D (3) and in particular the casesassociated with T1D as presented byPeters et al (4) do provide a good op-portunity to discuss how these agentsmodulate the pathophysiology leadingto DKA Thus it is in this context thatwe need to analyze the potential prob-lem of euDKA associated with SGLT2 in-hibitors to provide a more realistic andpractical perspective

THE PATHOPHYSIOLOGY

Ketosis results from restriction of carbo-hydrate usage with increased relianceon fat oxidation for energy productionThe pathogenesis of DKA is well estab-lished (10) Briefly absolute insulin de-ficiency leads to reduced glucoseutilization and enhanced lipolysis in-creased delivery of free fatty acids(FFAs) to the liver coupled with raisedglucagon levels promotes FFA oxidationand production of ketone bodies Inboth T1D and T2D DKA presents withmarked hyperglycemia (250 mgdLtypically 350ndash800mgdL) profuse glycos-uria (2ndash4 mg z min21 z kg21) and hyper-ketonemia (plasma b-hydroxybutyrate42ndash110 mmolL) (1112) The hypergly-cemia of DKA is associated with extremeinsulin resistance manifesting itself asmarkedly (70) reduced tissue glucosedisposal and increased endogenous glu-cose production (EGP) (12)

euDKA was originally defined as DKAwith plasma glucose levels300 mgdLoccurring in young T1D patients two-thirds of whom were female (13) Theprimary cause was reduced availability

of carbohydrate possibly in conjunctionwith reduced insulin dose The euDKAreported in T2D patients with SGLT2 in-hibitor treatment has a different originFull-dose SGLT2 inhibition induces arapid increase in urinary glucose excre-tion ranging 50ndash100 gday equally inmen and women and lasting slightly lon-ger than 24 h (14) In a typical 60-year-oldoverweight T2D patient (BMI 28 kgm2)consuming 50 of daily calories ascarbohydrate (15) this glucose lossamounts to 17ndash34 of estimated carbo-hydrate intake in men and 22ndash44 inwomen Of note in a comparative studyof Japanese and European T2D patientstreated with an SGLT2 inhibitor urinaryglucose excretion was if anything largerin the former (averaging 110 gday)than in the latter (60 gday) groups (16)thus in the Japanese group (BMI 25 kgm2)the glucose loss through the urine repre-sented 47 of estimated daily carbohy-drate intake in men and 57 in womenIn general depending on body size glo-merular filtration rate and degree of hy-perglycemia SGLT2-induced glucose losscan make up a substantial fraction ofdaily carbohydrate availability

Abstracting from a study in well-controlled drug-na ıve or metformin-treated T2D patients on chronic therapywith an SGLT2 inhibitor (17) plasma glu-cose levels decreased by 20ndash25 mgdLboth in the overnight fasted state andfollowing a mixed meal As glucose is thechief stimulus for insulin release under allcircumstances plasma insulin levels also fell(by 10 pmolL fasting and 60 pmolLpostmeal) In contrast plasma glucagonconcentrations increased significantlypartly because of a diminished paracrineinhibition by insulin (18) and possibly alsobecause of decreased SGLT2-mediatedglucose transport intoa-cells (19) As a con-sequence thecalculatedprehepatic insulin-to-glucagon molar concentration ratiodropped from9 to 7molmol in the fastingstate and from 29 to 24 molmol duringthe meal This hormonal shift which re-leases inhibition of gluconeogenesis inthe liver (20) augmented EGP both inthe fasting state and during the meal(17) Insulin sensitivity however wasimproved as also shown with the useof the euglycemic insulin clamp as a re-sult of attenuated glucotoxicity (21)

The difference in the pathophysiologyof DKA versus SGLT2 inhibitorndashinducedeuDKA is schematized in Fig 1 In euDKA

carediabetesjournalsorg Rosenstock and Ferrannini 1639

insulin deficiency and insulin resistanceare milder (and insulin resistance mayactually be improved) therefore glu-cose overproduction and underutiliza-tion are quantitatively lesser than inDKA More importantly renal glucoseclearance (ie the ratio of glycosuriato prevailing glycemia) is twice as largewith euDKA than with DKA In factfrom previous studies of patients ad-mitted with DKA it can be calculatedthat renal glucose clearance averaged03 mL z min21 z kg21 (12) whereas inT2D patients it rose from a near-negligible value in the baseline studyto 06 mL z min21 z kg21 with SGLT2treatment (17) Thus it is the entityof glycosuria viz the height of hyper-glycemia that marks the difference be-tween the two metabolic states Thisdifference can actually be amplified asDKA frequently occurs in patients withimpaired renal function (and hence lessglycosuria) (11) while SGLT2 inhibitorsmay be used in patients with glomeru-lar hyperfiltration (and more abundantglycosuria) (5)Ketoacidosis follows with the same

sequence of events in euDKA as inDKA Thus in SGLT2-treated T2D patients(17) the lower insulin-to-glucagon ratiostimulated lipolysis (circulating FFAswere 40 higher during the meal) andenhanced lipid oxidation (by 20 on av-erage) at the expense of carbohydrateoxidation (which fell by 60) In theface of lower substrate (glucose)

concentrations nonoxidative glucosedisposal (ie glycogen synthesis andlactate release) also fell by 15 Theaugmented FFA delivery to the liver re-sulted in mild stimulation of ketogene-sis whereby both fasting and meanpostmeal b-hydroxybutyrate levelsrosetwofold higher than in the baselinestudy (though not exceeding 1 mmolL)conversely plasma lactate levels de-creased 20 a readout of reducedcarbohydrate utilization (E Ferranniniet al unpublished data) Had insulindeficiency been more profounddascan happen in T1D patientsdor hadcarbohydrate availability been drasti-cally restricted this mild ketosis wouldhave evolved toward ketoacidosis withdecreased blood pH and bicarbonateand increased anion gap (12) It mustbe noted that in the only study thatlooked at insulin administration (22)ddating back to 1951dit was found thatinsulin decreased the transport maxi-mum for glucose in patients with diabe-tes implying that starting or escalatingexogenous insulin treatment would in-duce glycosuria in its own right Whilethese studies have not been repeatedand the mechanism of this insulin effecthas not been investigated it is never-theless intriguing that insulin may in-tensify SGLT2-induced glycosuria

All in all euDKA is pathophysiologi-cally similar to DKA except for thecircumstancedSGLT2-induced glycosuriadthat ldquoartificiallyrdquo lowers plasma glucose

levels and predisposes to increasedketogenesis

THE CLINICAL LESSON

The evidence reviewed above suggeststhat the risk of bona fide euDKA (and notsimple ketosis) in T2D related to the useof SGLT2 inhibitors will probably turnout to be very low with an ldquoacceptablerdquofrequency Still physicians and patientsneed to be made aware that such riskmay be increased in long-standing T2Dpatients with marked b-cell insuffi-ciency or in latent autoimmune diabetesin adults with rapid evolution towardT1D and during prolonged starvationafter surgery or during intercurrentillness In T1D however the euDKArisk appears to be more concrete forreasons entirely within the pathophysi-ology 1) in T1D patients hyperglycemiatypically is higher than in T2D patients2) in early T1D glomerular filtration ratemay be increased 3) insulin may en-hance the effect of SGLT2 inhibition onglycosuria and 4) changes in insulindose are not infrequent and may be in-appropriate for the amount and kind ofcarbohydrate intake

We submit that this potential compli-cation related to SGLT2 inhibition is pre-dictable detectable and preventable (ormitigable) so that the balance of benefitsand risks favors the use of SGLT2 inhibi-tors in the T1D population which is indesperate need of adjunct therapies Itis predictable because the persistent gly-cosuria induced by SGLT2 inhibition setsoff a sequence of metabolic changes thatare obligatory quantitative consequencesof a large glucose subtraction from thebody glucose pool In particular enhancedketogenesis is seen already even in non-diabetic subjects receiving SGLT2 inhibi-tors and is inscribed on already modestlyraised plasma b-hydroxybutyrate levels inpatients with diabetes (E Ferrannini et alunpublished observations) This back-ground ketonemia is asymptomatic andclinically irrelevant in most T2D patientsbut is certainly more of a concern in T1Dpatients who are already prone to de-velop ketosis under circumstances ofreduced insulin doses stress and inter-current illnesses and following a hypo-glycemic episode and during prolongedfasting or starvation

Thus one can envision a sequence ofclinical events that could evolve into afull-blown episode of DKA (Fig 2) First

Figure 1mdashEssential pathophysiology of DKA and euDKA consequent of the use of SGLT2 inhib-itors TGD tissue glucose disposal UGCr urinary glucose clearance rate

1640 Commentary Diabetes Care Volume 38 September 2015

inappropriate reductions of insulin dosesor any factor that may increase insulindemand such as stress a sick day oreven alcohol intake may induce hyper-ketonemia Under these circumstancesinitially patients may just not feel wellor experience some malaise and perhapsmild nausea with no vomiting Their firstimpulse is to check their blood glucosebecause of the persistent glycosuriaglycemia will be only mildly elevated sothat they would tend to reduce or with-hold insulin and avoid eating These ma-neuverswill accelerate ketone productionand metabolic decompensation towardDKA Themetabolic picture will be furthercompounded by the volume depletioncaused by the persistent glycosuria andvomitingThe cases reported by Peters et al (4)

exemplify some of the factors that trig-gered DKA most commonly they wereinsulin reductions low caloric and fluidintake intercurrent illness and alcoholuse Timewas wasted because of delayeddiagnosis due todeceptively ldquoacceptablerdquoblood glucose levels Peters et al (4) doprovide a further understanding of theclinical clues that can contribute to theearly detection and help raise awarenessof the potential for SGLT2 inhibition tocause euDKA in T1D We believe thateuDKA is in fact easily detectable becausereliable tools are currently available tomonitor ketonuria and ketonemia andshould be recommended to be used at

any time an SGLT2 inhibitorndashtreated pa-tient feels unwell regardless of the ambi-ent glucose levels This should be part ofany educational element for those treatedwith an SGLT2 inhibitor If detectablethen euDKA is preventable because de-tection of significant ketonuria andorketonemia any time symptoms such asnausea andor vomitingdor even justmalaisedappear especially after alcoholintake or a recent cut in insulin dose canprompt advice to maintain vigorous fluidintake and to consume carbohydrates toallow at least full-dose insulin therapy un-til the ketosis resolves Patients shouldtemporarily stop the SGLT2 inhibitor con-tact their medical provider and take sup-plemental boluses of rapid insulin alongwith liquids and carbohydrates Even ifpatients are unable to adjust the insulindose euDKA can bemitigable by drinkingand eating as tolerated without fear ofhyperglycemia and seeking prompt med-ical attention for parenteral fluid replace-ment and insulin therapy

In any event T1D patients who chooseto take this medication off-label shouldsign an ad hoc informed consent thatmakes them fully aware of the potentialfor euDKA the precipitating factors thewarning symptoms and signs and thepreventative measures to adopt

CONCLUSIONS

The ongoing long-term randomizedplacebo-controlled studies will provide

the necessary information on the safetyand efficacy of SGLT2 inhibitors in T1D(as well as insulin-treated T2D) Regula-tory scrutiny of these compounds willbalance their beneficial impact on over-all glycemic control glycemic variabilityand weight management against therisk of hypoglycemia and overall safetyincluding risk of euDKA A reduction ininsulin dose should not be regarded as apositive outcome in itself and should beachieved by slow gentle decrements si-multaneously to avoid hypoglycemiaand sliding toward euDKA Hopefullythese clinical development programswill quickly expand so as to offer to pa-tients and physicians a potential adjunctagainst the day-to-day managementchallenges of such a demanding disease

Duality of Interest JR has participated inadvisory boards and received honoraria or con-sulting fees fromMerck Sanofi Novo Nordisk EliLilly MannKind Corporation GlaxoSmithKlineTakeda Daiichi Sankyo Boehringer IngelheimJanssen Lexicon and Intarcia and has receivedresearch grants from Merck Pfizer Sanofi NovoNordisk Eli Lilly GlaxoSmithKline Takeda NovartisAstraZeneca Janssen Daiichi Sankyo MannKindCorporation Bristol-Myers Squibb BoehringerIngelheim Lexiconand Intarcia EFhasbeenamem-ber of the scientific advisory board or a speakerfor Boehringer Ingelheim Merck AstraZenecaEli Lilly Johnson amp Johnson Takeda Novo NordiskGlaxoSmithKline and Sanofi and has received re-search grant support from Eli Lilly and BoehringerIngelheim No other potential conflicts of inter-est relevant to this article were reported

References1 US Food and Drug Administration DrugSafety Communication FDA warns that SGLT2 in-hibitors for diabetes may result in a serious con-dition of too much acid in the blood [Internet] 15May 2015 Available from httpwwwfdagovdownloadsDrugsDrugSafetyUCM446954pdfAccessed 22 June 20152 European Medicines Agency Review of dia-betes medicines called SGLT2 inhibitors startedrisk of diabetic ketoacidosis to be examined [In-ternet] 12 June 2015 Available from httpwwwemaeuropaeudocsen_GBdocument_libraryReferrals_documentSGLT2_inhibitors__20Procedure_startedWC500187926pdf Accessed 22June 20153 Erondu N Desai M Ways K Meininger GDiabetic ketoacidosis and related events in thecanagliflozin type 2 diabetes clinical programDiabetes Care 2015381680ndash16864 Peters AL Buschur EO Buse JB Cohan PDiner JC Hirsch IB Euglycemic diabetic ketoaci-dosis a potential complication of treatmentwith sodiumndashglucose cotransporter 2 inhibi-tion Diabetes Care 2015381687ndash16935 Perkins BA Cherney DZ Partridge H et alSodium-glucose cotransporter 2 inhibition andglycemic control in type 1 diabetes results of an

Figure 2mdashDemonstration of the cascade of clinical events and metabolic changes that contrib-ute sequentially to progressive clinical deterioration and development of full-blown episodes ofeuDKA BG blood glucose CHO carbohydrate TGs triglycerides

carediabetesjournalsorg Rosenstock and Ferrannini 1641

8-week open-label proof-of-concept trial Dia-betes Care 2014371480ndash14836 Henry RR Rosenstock J Edelman S et al Ex-ploring the potential of the SGLT2 inhibitordapagliflozin in type 1 diabetes a randomizeddouble-blind placebo-controlled pilot studyDiabetes Care 201538412ndash4197 Sands AT Zambrowicz BP Rosenstock Jet al Sotagliflozin a dual SGLT1 and SGLT2inhibitor as adjunct therapy to insulin in type1 diabetes Diabetes Care 2015381181ndash11888 Nathan DM DCCTEDIC ResearchGroup TheDiabetes Control and Complications TrialEpidemiology of Diabetes Interventions and Com-plications study at 30 years overview DiabetesCare 2014379ndash169 Miller KM Foster NC Beck RW et al T1DExchange Clinic Network current state of type 1diabetes treatment in the US updated datafrom the T1D Exchange clinic registry DiabetesCare 201538971ndash97810 Kitabchi AE Umpierrez GE Murphy MB Di-abetic ketoacidosis and hyperosmolar state InInternational Textbook of Diabetes Mellitus 4thed DeFronzo RA Ferrannini E Zimmet P Alberti

KGMM Eds New York John Wiley amp Sons Ltd2015 p 799ndash81411 Barrett EJ DeFronzo RA Bevilacqua SFerrannini E Insulin resistance in diabetic keto-acidosis Diabetes 198231923ndash92812 Luzi L Barrett EJ Groop LC Ferrannini EDeFronzo RA Metabolic effects of low-dose in-sulin therapy on glucose metabolism in diabeticketoacidosis Diabetes 1988371470ndash147713 Munro JF Campbell IW McCuish ACDuncan LJP Euglycaemic diabetic ketoacidosisBMJ 19732578ndash58014 Sha S Devineni D Ghosh A et alCanagliflozin a novel inhibitor of sodium glucoseco-transporter 2 dose dependently reduces cal-culated renal threshold for glucose excretion andincreases urinary glucose excretion in healthysubjects Diabetes Obes Metab 201113669ndash67215 Hall KD Chow CC Estimating changes infree-living energy intake and its confidence in-terval Am J Clin Nutr 20119466ndash7416 Ferrannini E Veltkamp SA Smulders RAKadokura T Renal glucose handling impact ofchronic kidney disease and sodium-glucose co-transporter 2 inhibition in patients with type 2diabetes Diabetes Care 2013361260ndash1265

17 Ferrannini E Muscelli E Frascerra S et alMetabolic response to sodium-glucose cotrans-porter 2 inhibition in type 2 diabetic patients JClin Invest 2014124499ndash50818 Maruyama H Hisatomi A Orci L GrodskyGM Unger RH Insulin within islets is a physio-logic glucagon release inhibitor J Clin Invest1984742296ndash229919 Bonner C Kerr-Conte J Gmyr V et al In-hibition of the glucose transporter SGLT2 withdapagliflozin in pancreatic alpha cells triggersglucagon secretion Nat Med 201521512ndash51720 Burgess SC Regulation of glucose metabo-lism in liver In International Textbook of Diabe-tes Mellitus 4th ed DeFronzo RA Ferrannini EZimmet P Alberti KGMM Eds New York JohnWiley amp Sons Ltd 2015 p 193ndash21021 Merovci A Solis-Herrera C Daniele G et alDapagliflozin improves muscle insulin sensitiv-ity but enhances endogenous glucose produc-tion J Clin Invest 2014124509ndash51422 Farber SJ Berger EY Earle DP Effect of di-abetes and insulin of the maximum capacity ofthe renal tubules to reabsorb glucose J Clin In-vest 195130125ndash129

1642 Commentary Diabetes Care Volume 38 September 2015

Page 3: Euglycemic Diabetic Ketoacidosis: A Predictable ...Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:1638–1642

insulin deficiency and insulin resistanceare milder (and insulin resistance mayactually be improved) therefore glu-cose overproduction and underutiliza-tion are quantitatively lesser than inDKA More importantly renal glucoseclearance (ie the ratio of glycosuriato prevailing glycemia) is twice as largewith euDKA than with DKA In factfrom previous studies of patients ad-mitted with DKA it can be calculatedthat renal glucose clearance averaged03 mL z min21 z kg21 (12) whereas inT2D patients it rose from a near-negligible value in the baseline studyto 06 mL z min21 z kg21 with SGLT2treatment (17) Thus it is the entityof glycosuria viz the height of hyper-glycemia that marks the difference be-tween the two metabolic states Thisdifference can actually be amplified asDKA frequently occurs in patients withimpaired renal function (and hence lessglycosuria) (11) while SGLT2 inhibitorsmay be used in patients with glomeru-lar hyperfiltration (and more abundantglycosuria) (5)Ketoacidosis follows with the same

sequence of events in euDKA as inDKA Thus in SGLT2-treated T2D patients(17) the lower insulin-to-glucagon ratiostimulated lipolysis (circulating FFAswere 40 higher during the meal) andenhanced lipid oxidation (by 20 on av-erage) at the expense of carbohydrateoxidation (which fell by 60) In theface of lower substrate (glucose)

concentrations nonoxidative glucosedisposal (ie glycogen synthesis andlactate release) also fell by 15 Theaugmented FFA delivery to the liver re-sulted in mild stimulation of ketogene-sis whereby both fasting and meanpostmeal b-hydroxybutyrate levelsrosetwofold higher than in the baselinestudy (though not exceeding 1 mmolL)conversely plasma lactate levels de-creased 20 a readout of reducedcarbohydrate utilization (E Ferranniniet al unpublished data) Had insulindeficiency been more profounddascan happen in T1D patientsdor hadcarbohydrate availability been drasti-cally restricted this mild ketosis wouldhave evolved toward ketoacidosis withdecreased blood pH and bicarbonateand increased anion gap (12) It mustbe noted that in the only study thatlooked at insulin administration (22)ddating back to 1951dit was found thatinsulin decreased the transport maxi-mum for glucose in patients with diabe-tes implying that starting or escalatingexogenous insulin treatment would in-duce glycosuria in its own right Whilethese studies have not been repeatedand the mechanism of this insulin effecthas not been investigated it is never-theless intriguing that insulin may in-tensify SGLT2-induced glycosuria

All in all euDKA is pathophysiologi-cally similar to DKA except for thecircumstancedSGLT2-induced glycosuriadthat ldquoartificiallyrdquo lowers plasma glucose

levels and predisposes to increasedketogenesis

THE CLINICAL LESSON

The evidence reviewed above suggeststhat the risk of bona fide euDKA (and notsimple ketosis) in T2D related to the useof SGLT2 inhibitors will probably turnout to be very low with an ldquoacceptablerdquofrequency Still physicians and patientsneed to be made aware that such riskmay be increased in long-standing T2Dpatients with marked b-cell insuffi-ciency or in latent autoimmune diabetesin adults with rapid evolution towardT1D and during prolonged starvationafter surgery or during intercurrentillness In T1D however the euDKArisk appears to be more concrete forreasons entirely within the pathophysi-ology 1) in T1D patients hyperglycemiatypically is higher than in T2D patients2) in early T1D glomerular filtration ratemay be increased 3) insulin may en-hance the effect of SGLT2 inhibition onglycosuria and 4) changes in insulindose are not infrequent and may be in-appropriate for the amount and kind ofcarbohydrate intake

We submit that this potential compli-cation related to SGLT2 inhibition is pre-dictable detectable and preventable (ormitigable) so that the balance of benefitsand risks favors the use of SGLT2 inhibi-tors in the T1D population which is indesperate need of adjunct therapies Itis predictable because the persistent gly-cosuria induced by SGLT2 inhibition setsoff a sequence of metabolic changes thatare obligatory quantitative consequencesof a large glucose subtraction from thebody glucose pool In particular enhancedketogenesis is seen already even in non-diabetic subjects receiving SGLT2 inhibi-tors and is inscribed on already modestlyraised plasma b-hydroxybutyrate levels inpatients with diabetes (E Ferrannini et alunpublished observations) This back-ground ketonemia is asymptomatic andclinically irrelevant in most T2D patientsbut is certainly more of a concern in T1Dpatients who are already prone to de-velop ketosis under circumstances ofreduced insulin doses stress and inter-current illnesses and following a hypo-glycemic episode and during prolongedfasting or starvation

Thus one can envision a sequence ofclinical events that could evolve into afull-blown episode of DKA (Fig 2) First

Figure 1mdashEssential pathophysiology of DKA and euDKA consequent of the use of SGLT2 inhib-itors TGD tissue glucose disposal UGCr urinary glucose clearance rate

1640 Commentary Diabetes Care Volume 38 September 2015

inappropriate reductions of insulin dosesor any factor that may increase insulindemand such as stress a sick day oreven alcohol intake may induce hyper-ketonemia Under these circumstancesinitially patients may just not feel wellor experience some malaise and perhapsmild nausea with no vomiting Their firstimpulse is to check their blood glucosebecause of the persistent glycosuriaglycemia will be only mildly elevated sothat they would tend to reduce or with-hold insulin and avoid eating These ma-neuverswill accelerate ketone productionand metabolic decompensation towardDKA Themetabolic picture will be furthercompounded by the volume depletioncaused by the persistent glycosuria andvomitingThe cases reported by Peters et al (4)

exemplify some of the factors that trig-gered DKA most commonly they wereinsulin reductions low caloric and fluidintake intercurrent illness and alcoholuse Timewas wasted because of delayeddiagnosis due todeceptively ldquoacceptablerdquoblood glucose levels Peters et al (4) doprovide a further understanding of theclinical clues that can contribute to theearly detection and help raise awarenessof the potential for SGLT2 inhibition tocause euDKA in T1D We believe thateuDKA is in fact easily detectable becausereliable tools are currently available tomonitor ketonuria and ketonemia andshould be recommended to be used at

any time an SGLT2 inhibitorndashtreated pa-tient feels unwell regardless of the ambi-ent glucose levels This should be part ofany educational element for those treatedwith an SGLT2 inhibitor If detectablethen euDKA is preventable because de-tection of significant ketonuria andorketonemia any time symptoms such asnausea andor vomitingdor even justmalaisedappear especially after alcoholintake or a recent cut in insulin dose canprompt advice to maintain vigorous fluidintake and to consume carbohydrates toallow at least full-dose insulin therapy un-til the ketosis resolves Patients shouldtemporarily stop the SGLT2 inhibitor con-tact their medical provider and take sup-plemental boluses of rapid insulin alongwith liquids and carbohydrates Even ifpatients are unable to adjust the insulindose euDKA can bemitigable by drinkingand eating as tolerated without fear ofhyperglycemia and seeking prompt med-ical attention for parenteral fluid replace-ment and insulin therapy

In any event T1D patients who chooseto take this medication off-label shouldsign an ad hoc informed consent thatmakes them fully aware of the potentialfor euDKA the precipitating factors thewarning symptoms and signs and thepreventative measures to adopt

CONCLUSIONS

The ongoing long-term randomizedplacebo-controlled studies will provide

the necessary information on the safetyand efficacy of SGLT2 inhibitors in T1D(as well as insulin-treated T2D) Regula-tory scrutiny of these compounds willbalance their beneficial impact on over-all glycemic control glycemic variabilityand weight management against therisk of hypoglycemia and overall safetyincluding risk of euDKA A reduction ininsulin dose should not be regarded as apositive outcome in itself and should beachieved by slow gentle decrements si-multaneously to avoid hypoglycemiaand sliding toward euDKA Hopefullythese clinical development programswill quickly expand so as to offer to pa-tients and physicians a potential adjunctagainst the day-to-day managementchallenges of such a demanding disease

Duality of Interest JR has participated inadvisory boards and received honoraria or con-sulting fees fromMerck Sanofi Novo Nordisk EliLilly MannKind Corporation GlaxoSmithKlineTakeda Daiichi Sankyo Boehringer IngelheimJanssen Lexicon and Intarcia and has receivedresearch grants from Merck Pfizer Sanofi NovoNordisk Eli Lilly GlaxoSmithKline Takeda NovartisAstraZeneca Janssen Daiichi Sankyo MannKindCorporation Bristol-Myers Squibb BoehringerIngelheim Lexiconand Intarcia EFhasbeenamem-ber of the scientific advisory board or a speakerfor Boehringer Ingelheim Merck AstraZenecaEli Lilly Johnson amp Johnson Takeda Novo NordiskGlaxoSmithKline and Sanofi and has received re-search grant support from Eli Lilly and BoehringerIngelheim No other potential conflicts of inter-est relevant to this article were reported

References1 US Food and Drug Administration DrugSafety Communication FDA warns that SGLT2 in-hibitors for diabetes may result in a serious con-dition of too much acid in the blood [Internet] 15May 2015 Available from httpwwwfdagovdownloadsDrugsDrugSafetyUCM446954pdfAccessed 22 June 20152 European Medicines Agency Review of dia-betes medicines called SGLT2 inhibitors startedrisk of diabetic ketoacidosis to be examined [In-ternet] 12 June 2015 Available from httpwwwemaeuropaeudocsen_GBdocument_libraryReferrals_documentSGLT2_inhibitors__20Procedure_startedWC500187926pdf Accessed 22June 20153 Erondu N Desai M Ways K Meininger GDiabetic ketoacidosis and related events in thecanagliflozin type 2 diabetes clinical programDiabetes Care 2015381680ndash16864 Peters AL Buschur EO Buse JB Cohan PDiner JC Hirsch IB Euglycemic diabetic ketoaci-dosis a potential complication of treatmentwith sodiumndashglucose cotransporter 2 inhibi-tion Diabetes Care 2015381687ndash16935 Perkins BA Cherney DZ Partridge H et alSodium-glucose cotransporter 2 inhibition andglycemic control in type 1 diabetes results of an

Figure 2mdashDemonstration of the cascade of clinical events and metabolic changes that contrib-ute sequentially to progressive clinical deterioration and development of full-blown episodes ofeuDKA BG blood glucose CHO carbohydrate TGs triglycerides

carediabetesjournalsorg Rosenstock and Ferrannini 1641

8-week open-label proof-of-concept trial Dia-betes Care 2014371480ndash14836 Henry RR Rosenstock J Edelman S et al Ex-ploring the potential of the SGLT2 inhibitordapagliflozin in type 1 diabetes a randomizeddouble-blind placebo-controlled pilot studyDiabetes Care 201538412ndash4197 Sands AT Zambrowicz BP Rosenstock Jet al Sotagliflozin a dual SGLT1 and SGLT2inhibitor as adjunct therapy to insulin in type1 diabetes Diabetes Care 2015381181ndash11888 Nathan DM DCCTEDIC ResearchGroup TheDiabetes Control and Complications TrialEpidemiology of Diabetes Interventions and Com-plications study at 30 years overview DiabetesCare 2014379ndash169 Miller KM Foster NC Beck RW et al T1DExchange Clinic Network current state of type 1diabetes treatment in the US updated datafrom the T1D Exchange clinic registry DiabetesCare 201538971ndash97810 Kitabchi AE Umpierrez GE Murphy MB Di-abetic ketoacidosis and hyperosmolar state InInternational Textbook of Diabetes Mellitus 4thed DeFronzo RA Ferrannini E Zimmet P Alberti

KGMM Eds New York John Wiley amp Sons Ltd2015 p 799ndash81411 Barrett EJ DeFronzo RA Bevilacqua SFerrannini E Insulin resistance in diabetic keto-acidosis Diabetes 198231923ndash92812 Luzi L Barrett EJ Groop LC Ferrannini EDeFronzo RA Metabolic effects of low-dose in-sulin therapy on glucose metabolism in diabeticketoacidosis Diabetes 1988371470ndash147713 Munro JF Campbell IW McCuish ACDuncan LJP Euglycaemic diabetic ketoacidosisBMJ 19732578ndash58014 Sha S Devineni D Ghosh A et alCanagliflozin a novel inhibitor of sodium glucoseco-transporter 2 dose dependently reduces cal-culated renal threshold for glucose excretion andincreases urinary glucose excretion in healthysubjects Diabetes Obes Metab 201113669ndash67215 Hall KD Chow CC Estimating changes infree-living energy intake and its confidence in-terval Am J Clin Nutr 20119466ndash7416 Ferrannini E Veltkamp SA Smulders RAKadokura T Renal glucose handling impact ofchronic kidney disease and sodium-glucose co-transporter 2 inhibition in patients with type 2diabetes Diabetes Care 2013361260ndash1265

17 Ferrannini E Muscelli E Frascerra S et alMetabolic response to sodium-glucose cotrans-porter 2 inhibition in type 2 diabetic patients JClin Invest 2014124499ndash50818 Maruyama H Hisatomi A Orci L GrodskyGM Unger RH Insulin within islets is a physio-logic glucagon release inhibitor J Clin Invest1984742296ndash229919 Bonner C Kerr-Conte J Gmyr V et al In-hibition of the glucose transporter SGLT2 withdapagliflozin in pancreatic alpha cells triggersglucagon secretion Nat Med 201521512ndash51720 Burgess SC Regulation of glucose metabo-lism in liver In International Textbook of Diabe-tes Mellitus 4th ed DeFronzo RA Ferrannini EZimmet P Alberti KGMM Eds New York JohnWiley amp Sons Ltd 2015 p 193ndash21021 Merovci A Solis-Herrera C Daniele G et alDapagliflozin improves muscle insulin sensitiv-ity but enhances endogenous glucose produc-tion J Clin Invest 2014124509ndash51422 Farber SJ Berger EY Earle DP Effect of di-abetes and insulin of the maximum capacity ofthe renal tubules to reabsorb glucose J Clin In-vest 195130125ndash129

1642 Commentary Diabetes Care Volume 38 September 2015

Page 4: Euglycemic Diabetic Ketoacidosis: A Predictable ...Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:1638–1642

inappropriate reductions of insulin dosesor any factor that may increase insulindemand such as stress a sick day oreven alcohol intake may induce hyper-ketonemia Under these circumstancesinitially patients may just not feel wellor experience some malaise and perhapsmild nausea with no vomiting Their firstimpulse is to check their blood glucosebecause of the persistent glycosuriaglycemia will be only mildly elevated sothat they would tend to reduce or with-hold insulin and avoid eating These ma-neuverswill accelerate ketone productionand metabolic decompensation towardDKA Themetabolic picture will be furthercompounded by the volume depletioncaused by the persistent glycosuria andvomitingThe cases reported by Peters et al (4)

exemplify some of the factors that trig-gered DKA most commonly they wereinsulin reductions low caloric and fluidintake intercurrent illness and alcoholuse Timewas wasted because of delayeddiagnosis due todeceptively ldquoacceptablerdquoblood glucose levels Peters et al (4) doprovide a further understanding of theclinical clues that can contribute to theearly detection and help raise awarenessof the potential for SGLT2 inhibition tocause euDKA in T1D We believe thateuDKA is in fact easily detectable becausereliable tools are currently available tomonitor ketonuria and ketonemia andshould be recommended to be used at

any time an SGLT2 inhibitorndashtreated pa-tient feels unwell regardless of the ambi-ent glucose levels This should be part ofany educational element for those treatedwith an SGLT2 inhibitor If detectablethen euDKA is preventable because de-tection of significant ketonuria andorketonemia any time symptoms such asnausea andor vomitingdor even justmalaisedappear especially after alcoholintake or a recent cut in insulin dose canprompt advice to maintain vigorous fluidintake and to consume carbohydrates toallow at least full-dose insulin therapy un-til the ketosis resolves Patients shouldtemporarily stop the SGLT2 inhibitor con-tact their medical provider and take sup-plemental boluses of rapid insulin alongwith liquids and carbohydrates Even ifpatients are unable to adjust the insulindose euDKA can bemitigable by drinkingand eating as tolerated without fear ofhyperglycemia and seeking prompt med-ical attention for parenteral fluid replace-ment and insulin therapy

In any event T1D patients who chooseto take this medication off-label shouldsign an ad hoc informed consent thatmakes them fully aware of the potentialfor euDKA the precipitating factors thewarning symptoms and signs and thepreventative measures to adopt

CONCLUSIONS

The ongoing long-term randomizedplacebo-controlled studies will provide

the necessary information on the safetyand efficacy of SGLT2 inhibitors in T1D(as well as insulin-treated T2D) Regula-tory scrutiny of these compounds willbalance their beneficial impact on over-all glycemic control glycemic variabilityand weight management against therisk of hypoglycemia and overall safetyincluding risk of euDKA A reduction ininsulin dose should not be regarded as apositive outcome in itself and should beachieved by slow gentle decrements si-multaneously to avoid hypoglycemiaand sliding toward euDKA Hopefullythese clinical development programswill quickly expand so as to offer to pa-tients and physicians a potential adjunctagainst the day-to-day managementchallenges of such a demanding disease

Duality of Interest JR has participated inadvisory boards and received honoraria or con-sulting fees fromMerck Sanofi Novo Nordisk EliLilly MannKind Corporation GlaxoSmithKlineTakeda Daiichi Sankyo Boehringer IngelheimJanssen Lexicon and Intarcia and has receivedresearch grants from Merck Pfizer Sanofi NovoNordisk Eli Lilly GlaxoSmithKline Takeda NovartisAstraZeneca Janssen Daiichi Sankyo MannKindCorporation Bristol-Myers Squibb BoehringerIngelheim Lexiconand Intarcia EFhasbeenamem-ber of the scientific advisory board or a speakerfor Boehringer Ingelheim Merck AstraZenecaEli Lilly Johnson amp Johnson Takeda Novo NordiskGlaxoSmithKline and Sanofi and has received re-search grant support from Eli Lilly and BoehringerIngelheim No other potential conflicts of inter-est relevant to this article were reported

References1 US Food and Drug Administration DrugSafety Communication FDA warns that SGLT2 in-hibitors for diabetes may result in a serious con-dition of too much acid in the blood [Internet] 15May 2015 Available from httpwwwfdagovdownloadsDrugsDrugSafetyUCM446954pdfAccessed 22 June 20152 European Medicines Agency Review of dia-betes medicines called SGLT2 inhibitors startedrisk of diabetic ketoacidosis to be examined [In-ternet] 12 June 2015 Available from httpwwwemaeuropaeudocsen_GBdocument_libraryReferrals_documentSGLT2_inhibitors__20Procedure_startedWC500187926pdf Accessed 22June 20153 Erondu N Desai M Ways K Meininger GDiabetic ketoacidosis and related events in thecanagliflozin type 2 diabetes clinical programDiabetes Care 2015381680ndash16864 Peters AL Buschur EO Buse JB Cohan PDiner JC Hirsch IB Euglycemic diabetic ketoaci-dosis a potential complication of treatmentwith sodiumndashglucose cotransporter 2 inhibi-tion Diabetes Care 2015381687ndash16935 Perkins BA Cherney DZ Partridge H et alSodium-glucose cotransporter 2 inhibition andglycemic control in type 1 diabetes results of an

Figure 2mdashDemonstration of the cascade of clinical events and metabolic changes that contrib-ute sequentially to progressive clinical deterioration and development of full-blown episodes ofeuDKA BG blood glucose CHO carbohydrate TGs triglycerides

carediabetesjournalsorg Rosenstock and Ferrannini 1641

8-week open-label proof-of-concept trial Dia-betes Care 2014371480ndash14836 Henry RR Rosenstock J Edelman S et al Ex-ploring the potential of the SGLT2 inhibitordapagliflozin in type 1 diabetes a randomizeddouble-blind placebo-controlled pilot studyDiabetes Care 201538412ndash4197 Sands AT Zambrowicz BP Rosenstock Jet al Sotagliflozin a dual SGLT1 and SGLT2inhibitor as adjunct therapy to insulin in type1 diabetes Diabetes Care 2015381181ndash11888 Nathan DM DCCTEDIC ResearchGroup TheDiabetes Control and Complications TrialEpidemiology of Diabetes Interventions and Com-plications study at 30 years overview DiabetesCare 2014379ndash169 Miller KM Foster NC Beck RW et al T1DExchange Clinic Network current state of type 1diabetes treatment in the US updated datafrom the T1D Exchange clinic registry DiabetesCare 201538971ndash97810 Kitabchi AE Umpierrez GE Murphy MB Di-abetic ketoacidosis and hyperosmolar state InInternational Textbook of Diabetes Mellitus 4thed DeFronzo RA Ferrannini E Zimmet P Alberti

KGMM Eds New York John Wiley amp Sons Ltd2015 p 799ndash81411 Barrett EJ DeFronzo RA Bevilacqua SFerrannini E Insulin resistance in diabetic keto-acidosis Diabetes 198231923ndash92812 Luzi L Barrett EJ Groop LC Ferrannini EDeFronzo RA Metabolic effects of low-dose in-sulin therapy on glucose metabolism in diabeticketoacidosis Diabetes 1988371470ndash147713 Munro JF Campbell IW McCuish ACDuncan LJP Euglycaemic diabetic ketoacidosisBMJ 19732578ndash58014 Sha S Devineni D Ghosh A et alCanagliflozin a novel inhibitor of sodium glucoseco-transporter 2 dose dependently reduces cal-culated renal threshold for glucose excretion andincreases urinary glucose excretion in healthysubjects Diabetes Obes Metab 201113669ndash67215 Hall KD Chow CC Estimating changes infree-living energy intake and its confidence in-terval Am J Clin Nutr 20119466ndash7416 Ferrannini E Veltkamp SA Smulders RAKadokura T Renal glucose handling impact ofchronic kidney disease and sodium-glucose co-transporter 2 inhibition in patients with type 2diabetes Diabetes Care 2013361260ndash1265

17 Ferrannini E Muscelli E Frascerra S et alMetabolic response to sodium-glucose cotrans-porter 2 inhibition in type 2 diabetic patients JClin Invest 2014124499ndash50818 Maruyama H Hisatomi A Orci L GrodskyGM Unger RH Insulin within islets is a physio-logic glucagon release inhibitor J Clin Invest1984742296ndash229919 Bonner C Kerr-Conte J Gmyr V et al In-hibition of the glucose transporter SGLT2 withdapagliflozin in pancreatic alpha cells triggersglucagon secretion Nat Med 201521512ndash51720 Burgess SC Regulation of glucose metabo-lism in liver In International Textbook of Diabe-tes Mellitus 4th ed DeFronzo RA Ferrannini EZimmet P Alberti KGMM Eds New York JohnWiley amp Sons Ltd 2015 p 193ndash21021 Merovci A Solis-Herrera C Daniele G et alDapagliflozin improves muscle insulin sensitiv-ity but enhances endogenous glucose produc-tion J Clin Invest 2014124509ndash51422 Farber SJ Berger EY Earle DP Effect of di-abetes and insulin of the maximum capacity ofthe renal tubules to reabsorb glucose J Clin In-vest 195130125ndash129

1642 Commentary Diabetes Care Volume 38 September 2015

Page 5: Euglycemic Diabetic Ketoacidosis: A Predictable ...Euglycemic Diabetic Ketoacidosis: A Predictable, Detectable, and Preventable Safety Concern With SGLT2 Inhibitors Diabetes Care 2015;38:1638–1642

8-week open-label proof-of-concept trial Dia-betes Care 2014371480ndash14836 Henry RR Rosenstock J Edelman S et al Ex-ploring the potential of the SGLT2 inhibitordapagliflozin in type 1 diabetes a randomizeddouble-blind placebo-controlled pilot studyDiabetes Care 201538412ndash4197 Sands AT Zambrowicz BP Rosenstock Jet al Sotagliflozin a dual SGLT1 and SGLT2inhibitor as adjunct therapy to insulin in type1 diabetes Diabetes Care 2015381181ndash11888 Nathan DM DCCTEDIC ResearchGroup TheDiabetes Control and Complications TrialEpidemiology of Diabetes Interventions and Com-plications study at 30 years overview DiabetesCare 2014379ndash169 Miller KM Foster NC Beck RW et al T1DExchange Clinic Network current state of type 1diabetes treatment in the US updated datafrom the T1D Exchange clinic registry DiabetesCare 201538971ndash97810 Kitabchi AE Umpierrez GE Murphy MB Di-abetic ketoacidosis and hyperosmolar state InInternational Textbook of Diabetes Mellitus 4thed DeFronzo RA Ferrannini E Zimmet P Alberti

KGMM Eds New York John Wiley amp Sons Ltd2015 p 799ndash81411 Barrett EJ DeFronzo RA Bevilacqua SFerrannini E Insulin resistance in diabetic keto-acidosis Diabetes 198231923ndash92812 Luzi L Barrett EJ Groop LC Ferrannini EDeFronzo RA Metabolic effects of low-dose in-sulin therapy on glucose metabolism in diabeticketoacidosis Diabetes 1988371470ndash147713 Munro JF Campbell IW McCuish ACDuncan LJP Euglycaemic diabetic ketoacidosisBMJ 19732578ndash58014 Sha S Devineni D Ghosh A et alCanagliflozin a novel inhibitor of sodium glucoseco-transporter 2 dose dependently reduces cal-culated renal threshold for glucose excretion andincreases urinary glucose excretion in healthysubjects Diabetes Obes Metab 201113669ndash67215 Hall KD Chow CC Estimating changes infree-living energy intake and its confidence in-terval Am J Clin Nutr 20119466ndash7416 Ferrannini E Veltkamp SA Smulders RAKadokura T Renal glucose handling impact ofchronic kidney disease and sodium-glucose co-transporter 2 inhibition in patients with type 2diabetes Diabetes Care 2013361260ndash1265

17 Ferrannini E Muscelli E Frascerra S et alMetabolic response to sodium-glucose cotrans-porter 2 inhibition in type 2 diabetic patients JClin Invest 2014124499ndash50818 Maruyama H Hisatomi A Orci L GrodskyGM Unger RH Insulin within islets is a physio-logic glucagon release inhibitor J Clin Invest1984742296ndash229919 Bonner C Kerr-Conte J Gmyr V et al In-hibition of the glucose transporter SGLT2 withdapagliflozin in pancreatic alpha cells triggersglucagon secretion Nat Med 201521512ndash51720 Burgess SC Regulation of glucose metabo-lism in liver In International Textbook of Diabe-tes Mellitus 4th ed DeFronzo RA Ferrannini EZimmet P Alberti KGMM Eds New York JohnWiley amp Sons Ltd 2015 p 193ndash21021 Merovci A Solis-Herrera C Daniele G et alDapagliflozin improves muscle insulin sensitiv-ity but enhances endogenous glucose produc-tion J Clin Invest 2014124509ndash51422 Farber SJ Berger EY Earle DP Effect of di-abetes and insulin of the maximum capacity ofthe renal tubules to reabsorb glucose J Clin In-vest 195130125ndash129

1642 Commentary Diabetes Care Volume 38 September 2015