Oral Semaglutide VersusEmpagliflozin in Patients WithType 2 Diabetes Uncontrolled onMetformin: The PIONEER 2 TrialDiabetes Care 2019;42:2272–2281 | https://doi.org/10.2337/dc19-0883

OBJECTIVE

Efficacy and safety of the glucagon-like peptide 1 (GLP-1) analog oral semaglutideand the sodium–glucose cotransporter 2 inhibitor empagliflozin were compared inpatients with type 2 diabetes uncontrolled on metformin.

RESEARCH DESIGN AND METHODS

Patients were randomized to once-daily open-label treatment with oral semaglu-tide 14 mg (n5 412) or empagliflozin 25 mg (n5 410) in a 52-week trial. Key endpoints were change from baseline to week 26 in HbA1c (primary) and body weight(confirmatory secondary). Two estimands addressed efficacy-related questions:treatment policy (regardless of trial product discontinuation or rescue medication)and trial product (on trial product without rescue medication) in all randomizedpatients.

RESULTS

Four hundred (97.1%) patients in the oral semaglutide group and 387 (94.4%) in theempagliflozin group completed the trial. Oral semaglutide provided superiorreductions in HbA1c versus empagliflozin at week 26 (treatment policy –1.3% vs.–0.9% [–14 vs. –9 mmol/mol], estimated treatment difference [ETD] –0.4% [95% CI–0.6, –0.3] [–5 mmol/mol (–6, –3)]; P < 0.0001). The treatment difference in HbA1c

significantly favored oral semaglutide at week 26 for the trial product estimand(–1.4%vs.–0.9%[–15vs.–9mmol/mol], ETD–0.5%[95%CI–0.7,–0.4] [–6mmol/mol(–7, –5)]; P < 0.0001) and at week 52 for both estimands (P < 0.0001). Superiorweight loss was not confirmed at week 26 (treatment policy), but oral semaglu-tide was significantly better than empagliflozin at week 52 (trial product 24.7 vs.23.8 kg; P 5 0.0114). Gastrointestinal adverse events were more commonwith oral semaglutide.

CONCLUSIONS

Oral semaglutide was superior to empagliflozin in reducing HbA1c but not bodyweight at 26 weeks in patients with type 2 diabetes uncontrolled onmetformin. Atweek52,HbA1c andbodyweight (trial product estimand)were significantly reducedversus empagliflozin. Oral semaglutide was well tolerated within the establishedsafety profile of GLP-1 receptor agonists.

1Endocrine andMetabolic Consultants, Rockville,MD2Dallas Diabetes Research Center at Medical City,Dallas, TX3Endocrine Division, Hospital de Clınicas de PortoAlegre, Porto Alegre, Brazil4Rajavithi Hospital, Rangsit Medical School,Bangkok, Thailand5Medical University of Silesia, Katowice, Poland6Novo Nordisk A/S, Søborg, Denmark7Departments of Internal Medicine and ClinicalSciences, University of Texas Southwestern Med-ical Center, Dallas, TX8HospitalUniversitariBellvitge-IDIBELL,CIBERDEM,and University of Barcelona, Barcelona, Spain

Corresponding author: Helena W. Rodbard,hrodbard@comcast.net

Received 2 May 2019 and accepted 12 September2019

Clinical trial reg. no. NCT02863328, clinicaltrials.gov

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc19-0883/-/DC1.

This article is featured in a podcast available athttp://www.diabetesjournals.org/content/diabetes-core-update-podcasts.

*A complete list of investigators in the PeptideInnovation for Early Diabetes Treatment 2 trial(PIONEER 2) is provided in the SupplementaryData.

© 2019 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered. More infor-mation is available at http://www.diabetesjournals.org/content/license.

Helena W. Rodbard,1 Julio Rosenstock,2

Luis H. Canani,3

Chaicharn Deerochanawong,4

Janusz Gumprecht,5

Søren Østergaard Lindberg,6

Ildiko Lingvay,7

Anette Luther Søndergaard,6

Marianne Bach Treppendahl,6 and

Eduard Montanya,8 for the PIONEER

2 Investigators*

2272 Diabetes Care Volume 42, December 2019

EMER

GINGTH

ERAPIES:

DRUGSANDREG

IMEN

S

Many patients with type 2 diabetes fail toachieve or maintain adequate blood glu-cose control when treated with metforminmonotherapy. Injectable glucagon-likepeptide 1 receptor agonists (GLP-1RAs)and oral sodium–glucose cotransporter 2(SGLT-2) inhibitors are recommended assecond-line therapy because of their abil-ity to lower glucose without increasinghypoglycemia risk, weight loss effect, andassociated cardiovascular benefits (1,2).Semaglutide is a human GLP-1 analog

currently available as a once-weekly in-jection associated with reduced glycatedhemoglobin (HbA1c), weight loss, andfewer cardiovascular events in type 2 di-abetes (3–9). Oral semaglutide is cofor-mulated in a tablet with the absorptionenhancersodiumN-(8-[2-hydroxylbenzoyl]amino) caprylate, which facilitatessemaglutide absorption across the gastricmucosa (10). Oral semaglutide has dem-onstrated significantly greater reductionsin HbA1c and body weight compared withplacebo in patients with type 2 diabetesuncontrolled with diet and exercise or oralantidiabetic medication, including in pa-tients with moderate renal impairment(11–14). Significantly greater reductionsin HbA1c and body weight have also beenshown with oral semaglutide, given as 7 or14 mg/day or flexibly dosed, comparedwith sitagliptin in patients uncontrolledwith oral antidiabetic drugs (15,16). Oralsemaglutide also resulted in a noninferiorreduction in HbA1c and superior weightloss versus liraglutide in patients on met-formin with or without an SGLT-2 inhibitor(13). Cardiovascular safety has been con-firmed, with an indication of benefit, bya nonsignificant 21% risk reduction inmajor adverse cardiovascular eventsversus placebo (17).Empagliflozin is a widely used oral

SGLT-2 inhibitor shown to improve gly-cemic control and body weight (18–22)and associated with a reduced risk ofcardiovascular and all-cause mortality inpatients at high cardiovascular risk (23).The present phase 3a trial, PIONEER 2,is the first direct comparison of oralsemaglutide with an SGLT-2 inhibitor,empagliflozin, in type 2 diabetes uncon-trolled with metformin monotherapy.

RESEARCH DESIGN AND METHODS

Trial DesignThis randomized, open-label, multina-tional 52-week trial was conducted at108 sites in 12 countries (Argentina,

Brazil, Croatia, Greece, Hungary, Italy,Poland, Russia, Serbia, Spain, Thailand,U.S.). Patients were randomized (1:1) toonce-daily oral semaglutide 14 mg orempagliflozin 25 mg for 52 weeks usingan interactive web response systemwith a further 5 weeks of follow-up(Supplementary Fig. 1). An open-labeltrial design was used because manufac-ture of placebo tablets resembling em-pagliflozin was not feasible within areasonable time frame. Oral semaglutidewas initiated at 3 mg once daily, esca-lated to 7 mg at week 4 and 14 mg afterweek 8. Because food impairs absorptionof oral semaglutide, patients were in-structed to administer oral semaglutidein themorning in a fasted statewith up to120 mL of water at least 30 min beforebreakfast and any other oral medication.Empagliflozin was initiated at 10 mg oncedaily in the morning and escalated to25 mg at week 8.

Additional antidiabetic medicationwas available for patients with persistentor unacceptable hyperglycemia on trialproduct and for patients who prema-turely discontinued trial product andremained in the trial. Additional antidi-abetic medication was defined as thatinitiated (or intensification of existingantidiabetic background medicationby a dose increase of .20%) duringthe planned treatment period (i.e.,from randomization to the plannedend-of-treatment visit) either as add-onto trial product or initiated after pre-mature discontinuation of trial product.The subset of additional antidiabeticmedication (or intensification of existingantidiabetic background medication)used as add-on to trial product is definedas rescue medication. Short-term use(#21 days) of antidiabetic medication(e.g., in connection with intercurrentillness) was not considered as additionalantidiabetic medication (including res-cue medication).

Rescue criteria were fasting plasmaglucose .260 mg/dL (14.4 mmol/L)from week 8 to 13, .240 mg/dL (13.3mmol/L) from week 14 to 25, and .200mg/dL (11.1 mmol/L) (or HbA1c .8.5%[69.4 mmol/mol]) from week 26 onward.Rescue medication was prescribed at theinvestigator’s discretion (excluding GLP-1RAs, dipeptidyl peptidase 4 inhibitors,and amylin analogs in theoral semaglutidearm and SGLT-2 inhibitors in the empagli-flozin arm). Patients who prematurely

discontinued trial product remained inthe trial and could receive any other an-tidiabetic medications at the investigator’sdiscretion (excluding GLP-1RAs in the oralsemaglutide arm before completion of thefollow-up visit 5 weeks after the last dateon trial product).

Two different questions related tothe efficacy objectives were addressedthrough the definition of two estimands:treatment policy and trial product. Bothestimands were defined based on inter-actions with regulatory agencies. Thetreatment policy estimand evaluates thetreatment effect for all randomized pa-tients, regardless of trial product discon-tinuation oruse of rescuemedication. Thisestimand reflects the intention-to-treatprinciple as defined in International Coun-cil on Harmonization (ICH) E9 (24). Theestimand reflects the effect of initiatingtreatment with oral semaglutide com-pared with initiating treatment with em-pagliflozin, both potentially followed byeither discontinuation of trial productand/or addition of or switch to anotherglucose-lowering drug.

The trial product estimand evaluatesthe treatment effect for all randomizedpatients under the assumption that allpatients remained on trial product for theentire planned duration of the trial anddid not use rescue medication. This es-timand aims at reflecting the effect oforal semaglutide compared with empa-gliflozin without the confounding effectof rescuemedication. The statistical anal-ysis that was applied to estimate thisestimand is similar to howmanyphase 3adiabetes trials have been evaluated,and results from such analyses arecurrently included in many productlabels (prescribing information, U.S.,and summary of product characteristics,European Union) for glucose-loweringdrugs (e.g., Ozempic summary of productcharacteristics).

Trial product discontinuation and initia-tion of rescue medication are accountedfor by the treatment policy strategy for thetreatment policy estimand and by thehypothetical strategy for the trial productestimand as defined in draft ICH E9 (R1)(25). Further details on the use of estimandsin this trial are provided in SupplementaryData, Estimands, with additional back-ground provided by Aroda et al. (26).

The trial protocol was approved byall relevant institutional review boards/independent ethics committees, and the

care.diabetesjournals.org Rodbard and Associates 2273

trial was conducted in accordancewith ICHGood Clinical Practice guidelines and theDeclaration of Helsinki. All patients pro-vided written informed consent beforeany trial-related activity.

PatientsEligible patients were adults with type 2diabetes and an HbA1c of 7.0–10.5%(53–91 mmol/mol) receiving a stabledose of metformin ($1,500 mg or max-imum tolerated). Key exclusion criteria(see Supplementary Table 1 for full list)were any medication for diabetes orobesity within the previous 90 daysother than metformin or short-term(#14 days) insulin, renal impairmentwith an estimated glomerular filtrationrate ,60 mL/min/1.73 m2, proliferativeretinopathy or maculopathy requiringacute treatment verified by fundus pho-tography or dilated fundoscopy, andhistory of pancreatitis.

Trial End PointsThe primary end point was change inHbA1c from baseline to week 26. Theconfirmatory secondary end point waschange in bodyweight (kg) from baselineto week 26.Secondary end points included changes

frombaseline toweek52 inHbA1c andbodyweight (kg) and changes from baseline toweeks 26 and 52 in fasting plasma glucose,self-measured blood glucose (SMBG)profile (7-point profile andmean postpran-dial increment over all meals), fastingC-peptide, fasting insulin, fastingproinsulin,fasting glucagon, HOMA of insulin resis-tance (HOMA-IR), HOMA of b-cell function(HOMA-B), C-reactive protein, bodyweight(%), BMI, waist circumference, and fastinglipid profile. Other secondary end pointswere the proportion of patients achiev-ing HbA1c ,7% (53 mmol/mol) or#6.5%(48 mmol/mol); weight loss of $5% or$10%; composite end point of HbA1c,7%(53 mmol/mol) without severe or symp-tomatic hypoglycemia (blood glucose,56 mg/dL [,3.1 mmol/L]) and noweight gain; composite end point of anabsolute reduction in HbA1c of $1.0%(10.9 mmol/mol) and body weight lossof$3% (weeks 26 and 52); and changesfrom baseline to weeks 26 and 52 inthe patient-reported outcomes, ShortForm (SF) 36v2 Health Survey (AcuteVersion) (27) and Control of Eating Ques-tionnaire (28). Further end points arelisted in the Estimands section and the

protocol that are included as part ofthe Supplementary Data.

Safety end points included the numberof treatment-emergent adverse events,incidence of American Diabetes Associa-tion (ADA)–classified (29) severe or con-firmed symptomatic hypoglycemicepisodes (blood glucose ,56 mg/dL[,3.1 mmol/L]), and changes frombaseline in heart rate, blood pressure,and other clinical and laboratory assess-ments. An independent external eventadjudication committee (EAC) per-formed masked validation of predefinedadverse events, including deaths, selectedcardiovascular events, acute pancreatitis,malignant neoplasms, acute kidney injury,and lactic acidosis.

Statistical AnalysisThe primary end point of change frombaseline to week 26 in HbA1c was testedfor both noninferiority and superiority oforal semaglutide versus empagliflozin, witha sample size calculation to ensure a powerof at least 90% for testing superiority. Theconfirmatory secondaryendpointof changefrom baseline to week 26 in body weightwas tested for superiority of oral semaglu-tide versus empagliflozin. The confirmationof efficacy of oral semaglutide on changein HbA1c and body weight from baselineto week 26 was based on a weightedBonferroni closed testing strategy (30) tocontrol the overall type I error for thehypotheses evaluated by the treatmentpolicy estimand (Supplementary Fig. 2).Because of the potential for type I errorsas a result of multiple comparisons, find-ings for analyses of additional secondaryend points should be interpreted asexploratory.

The treatment policy estimand was es-timated by a pattern-mixture model usingmultiple imputation to handle missingweek 26 data for both confirmatory endpoints. Data collected at week 26, irre-spective of premature discontinuation oftrial product or initiation of rescue med-ication, were included in the statisticalanalysis. Imputation was done withingroups defined by trial product and treat-ment status at week 26. Both the impu-tation and the analysis were basedon ANCOVAmodels. The results were com-bined by use of Rubin’s rule (31). Beforetesting for noninferiority, a value of 0.4%(the noninferiority margin) was added toimputed values at week 26 for the oralsemaglutide treatment arm only (32).

The trial product estimand was estimatedby a mixed model for repeated measure-ments that used data collected beforepremature trial product discontinuationor initiation of rescue medication fromall randomized patients.

Further details on the statistical anal-yses can be found in Supplementary Fig.2. All analyses were performed usingSAS 9.4M2 statistical software.

Data AvailabilityData will be shared with bona fide re-searchers submitting a research proposalapproved by the independent reviewboard. Access request proposals can befound at http://novonordisk-trials.com.Data will bemade available after researchcompletion, approval of the product, andproduct use in the European Union andU.S. Individual participant data will beshared in data sets in a deidentified/anonymized format using a specializedSAS data platform.

RESULTS

PatientsA total of 1,122 patients were screened,with 822 randomized to oral semaglutide14mgoncedaily (n5412)orempagliflozin25 mg once daily (n5 410). Four hundred(97.1%) patients in the oral semaglutidegroup and387 (94.4%) in the empagliflozingroup completed the trial (Supple-mentary Fig. 3). Baseline characteristicswere well balanced between treatmentgroups (Table 1). Patients, of whom half(49.5%) were female, had a mean ageof 58 years, baseline HbA1c of 8.1%(65 mmol/mol), fasting plasma glucoseof 173 mg/dL (9.6 mmol/L), averageduration of diabetes of 7.4 years, andmean body weight of 91.6 kg.

Use of additional antidiabetic medica-tion and rescue medication is shown inSupplementary Table 2. Through to week26, 17 (4.1%) patients initiated addi-tional antidiabetic medication in the oralsemaglutide group; in 8 (1.9%) of thesepatients, it was rescue medication. In theempagliflozin group, 13 (3.2%) patients ini-tiated additional antidiabetic medicationthrough to week 26, with this beingrescue medication in 5 (1.2%). Throughtoweek 52, 52 (12.7%) patients initiatedadditional antidiabetic medication inthe oral semaglutide group; in 31 (7.5%)of these patients, it was rescue medi-cation. In the empagliflozin group, 56(13.7%) patients initiated additional

2274 Oral Semaglutide Versus Empagliflozin Diabetes Care Volume 42, December 2019

antidiabetic medication, with this beingrescue medication in 44 (10.7%). Sulfo-nylureas were the most commonly usedadditional antidiabetic and rescue med-ication. Disposition of patients through-out the trial is shown in SupplementaryFig 4.

Glycemic ControlOral semaglutide 14 mg provided a su-perior reduction in HbA1c compared withempagliflozin 25 mg at week 26 whenevaluated by the treatment policy esti-mand (regardless of rescue medicationuse or trial product discontinuation)(–1.3% vs. –0.9% [–14 vs. –9 mmol/mol];estimated treatment difference [ETD]–0.4% [95% CI –0.6, –0.3] [–5 mmol/mol (–6, –3)]; P , 0.0001 for noninfe-riority and superiority) (Fig. 1). Resultsfrom sensitivity analyses supportedthe results of the confirmatory analysis(Supplementary Fig. 5). When evalu-ated by the trial product estimand (ontrial product and without the use ofrescue medication), the reduction inHbA1c was significantly greater withoral semaglutide at week 26 (–1.4%vs. –0.9% [–15 vs. –9 mmol/mol], ETD–0.5% [–0.7, –0.4] [–6 mmol/mol (–7,–5)]; P , 0.0001) (Fig. 1). Significantlygreater reductions in HbA1c with oral

semaglutide compared with empagli-flozin were also observed at week 52(both estimands) (Fig. 1). More patientsachieved the predefined HbA1c targetswith oral semaglutide than with empa-gliflozin, and the odds of doing so weresignificantly greater at weeks 26 and 52(both estimands, all P , 0.0001) (Fig. 1and Table 2).

Fasting plasma glucose was reducedwith both treatments, with no significantdifference between groups (Table 2and Supplementary Fig. 6). Oral sema-glutide resulted in significantly greaterreductions in mean 7-point SMBGprofiles compared with empagliflozin atboth weeks 26 and 52 (Table 2 andSupplementary Fig. 6) and significantlyreduced mean postprandial increments,as averaged for all meals (excludingthe treatment policy estimand evalua-tion at week 26) (Table 2).

Body WeightSuperiority of body weight reduction atweek 26 with oral semaglutide over em-pagliflozin was not confirmed (treatmentpolicy estimand 23.8 vs. 23.7 kg;ETD 20.1 kg [95% CI 20.7, 0.5];P5 0.7593). Results from sensitivity anal-yses supported the results of the con-firmatory analysis (Supplementary Fig.

5). There was no difference betweentreatments using the trial product esti-mand (24.2 vs. 23.8 kg; ETD 20.4 kg[21.0, 0.1]; P 5 0.1358) (Fig. 1). Asignificantly greater reduction in bodyweight was achieved with oral semaglu-tide versus empagliflozin at week 52when evaluated by the trial product esti-mand (24.7 vs. 23.8 kg; ETD 20.9 kg[21.6, 20.2]; P 5 0.0114) but notthe treatment policy estimand (23.8vs. 23.6 kg; ETD 20.2 kg [20.9, 0.5];P 5 0.6231). Proportions of patientsachieving $5% or $10% weight lossare shown in Fig. 1 and Table 2, respec-tively. Reductions in waist circumferencewere significantly greater with oral sem-aglutide than with empagliflozin at week26 (both estimands) and at week 52 (trialproduct estimand) (Table 2).

Other OutcomesMore patients achieved the two compos-ite end points (HbA1c,7% [53mmol/mol]without severe or symptomatic hypo-glycemia and no weight gain and an ab-solute reduction in HbA1c of $1.0%[10.9 mmol/mol] and body weight lossof $3%) with oral semaglutide versusempagliflozin, and the odds of doing sowere significantly greater at both weeks26 and 52 (Table 2). Reduction in C-reactiveprotein was significantly greater withoral semaglutide versus empagliflozin(Table 2). Other secondary end points arepresented in Table 2 and SupplementaryTable 3.

For the Control of Eating Questionnaire,the domains craving control (weeks 26and 52) and craving for savory (week 52)were significantly improved in favor oforal semaglutide versus empagliflozin(treatment policy estimand). Both do-mains were significantly in favor of oralsemaglutide at both weeks 26 and 52 forthe trial product estimand. Patient-reported outcomes are summarized inSupplementary Fig. 7.

SafetyThe overall number of adverse eventsand proportion of patients reportingadverse events were similar with oralsemaglutide and empagliflozin, andmost events were mild to moderateseverity (Table 3). Fewer patients expe-rienced serious adverse events in theoral semaglutide group. There wasone death in the empagliflozin group(undetermined cause). The most frequent

Table 1—Baseline characteristics and demographics

Oral semaglutide14 mg

Empagliflozin25 mg Total

Patients, n 411 410 821

Age (years), mean (SD) 57 (10) 58 (10) 58 (10)

Female, n (%) 205 (49.9) 201 (49.0) 406 (49.5)

Race, n (%)White 355 (86.4) 353 (86.1) 708 (86.2)Black or African American 26 (6.3) 33 (8.0) 59 (7.2)Asian 28 (6.8) 21 (5.1) 49 (6.0)Other 2 (0.5) 3 (0.7) 5 (0.6)

Ethnicity, n (%)Hispanic or Latino 91 (22.1) 108 (26.3) 199 (24.2)

Durationofdiabetes (years),mean (SD) 7.2 (5.8) 7.7 (6.3) 7.4 (6.1)

Body weight (kg), mean (SD) 91.9 (20.5) 91.3 (20.1) 91.6 (20.3)

BMI (kg/m2), mean (SD) 32.9 (6.3) 32.8 (5.9) 32.8 (6.1)

HbA1c, mean (SD)% 8.1 (0.9) 8.1 (0.9) 8.1 (0.9)mmol/mol 65 (10) 65 (10) 65 (10)

Fasting plasma glucose, mean (SD)mmol/L 9.5 (2.3) 9.7 (2.5) 9.6 (2.4)mg/dL 171.5 (41.8) 174.0 (45.2) 172.8 (43.5)

Estimated glomerular filtration rate*(mL/min/1.73 m2), mean (SD) 96 (15) 95 (15) 95 (15)

*Glomerular filtration rate was estimated by the Chronic Kidney Disease EpidemiologyCollaboration equation.

care.diabetesjournals.org Rodbard and Associates 2275

adverse event with oral semaglutide wasnausea, which was nonserious, usuallymild to moderate severity and transient,and did not exceed a prevalence of 10%at any time (Table 3 and SupplementaryFig. 8). Female and male genital mycoticinfections of mild to moderate severityoccurred more frequently with empagli-flozin than with oral semaglutide (8.5%and 6.7% vs. 2.0% and 0%, respectively)(Supplementary Table 4).Adverse events resulting in trial prod-

uct discontinuation were more frequentwith oral semaglutide than with empagli-flozin (10.7% vs. 4.4%) andwere primarilyrelated to gastrointestinal symptoms(8.0% vs. 0.7%) (Table 3). In both groups,premature discontinuations mainly oc-curred in the first 16 weeks of treatment.

Incidence of severe or confirmed symp-tomatic hypoglycemic episodes (,56mg/dL[,3.1 mmol/L]) was low and similar inbothgroups (Table 3). Diabetic retinopathy–related adverse events were reportedin 14 (3.4%) patients in the oral sema-glutide group and in 5 (1.2%) in theempagliflozin group (in-trial period)(Supplementary Table 5). All such eventswere identified by routine eye examina-tion as part of the trial protocol and werenonserious, of mild or moderate severity,and did not require treatment. EAC-confirmed malignant neoplasms wereidentified in seven (1.7%) patients inthe oral semaglutide group and two(0.5%) in the empagliflozin group (in-trialperiod). There was no clustering of malig-nancies in any particular organ or system

(Supplementary Table 6). Cardiovascularevents occurred at a similar rate in bothgroups (EAC confirmed; oral semaglutiden5 5 [1.2%], empagliflozin n5 6 [1.5%])(Supplementary Table 6). Other EAC-confirmed events and safety assess-ments are reported in SupplementaryTables 6 and 7.

CONCLUSIONS

Oral semaglutide is the first oral GLP-1RAto be investigated for the treatment oftype 2 diabetes. In PIONEER 2, oral sem-aglutide was superior to empagliflozin,with meaningful reductions in HbA1c at26 weeks in patients with type 2 diabetesuncontrolled on metformin monotherapy.Furthermore, the difference between treat-ments remained significant at 52 weeks.

Figure 1—Glycemic control and body weight–related efficacy end points. A: Observed absolute change in HbA1c over time. B: Estimated changes frombaseline in HbA1c at weeks 26 and 52. C: Observed proportions of patients achieving HbA1c ,7% (53 mmol/mol) at weeks 26 and 52. D: Observedabsolute change in bodyweight over time. E: Estimated changes from baseline in bodyweight at weeks 26 and 52. F: Observed proportions of patientsachievingbodyweight reduction$5%atweeks26and52. Treatmentpolicy estimand:Data irrespectiveofdiscontinuationof trial productand initiationof rescuemedicationwere included. Trial product estimand: Data collected after discontinuation of trial product or initiation of rescuemedication areexcluded. P values are two-sided and unadjusted. *Superiority confirmed for oral semaglutide versus empagliflozin. Observed mean change (6 SEM)from baseline (A and D), estimatedmean changes from baseline at week 26 and 52 (B and E), and observed proportions of patients achieving target atweeks 26 and 52 (C and F). Patient numbers represent patients contributing to the means. EOR, estimated odds ratio.

2276 Oral Semaglutide Versus Empagliflozin Diabetes Care Volume 42, December 2019

Attainment of ADA-recommended HbA1ctargets at 26 and 52 weeks was alsosignificantly greater with oral semaglu-tide. Reductions in fasting plasma glucosewere similar in both groups, suggestingdifferences in glycemic control may bemostly driven by the greater reduction inpostprandial glucose with oral semaglutide.Reductions in body weight occurred

with both treatments, but superiorityof oral semaglutide versus empagliflozincould not be confirmed at week 26.However, weight loss in the empagliflozingroup stabilized aroundweek 26, whereasin the oral semaglutide group, weightloss continued until around week 38and was significantly greater at 52 weekson the basis of the trial product estimand.This significantly greater weight loss at52 weeks with oral semaglutide on thebasis of the trial product estimand reflects

the treatment effect without the con-founding influence of rescue medicationuse and treatment discontinuations. Pa-tients discontinuing oral semaglutidecould not be switched to additional anti-diabetic medication with a comparableweight-reducing effect, while patientson empagliflozin could be switched toGLP-1RAs.

The safety profile of oral semaglutidewas consistent with previous trials(11–16). More patients prematurelydiscontinued treatment because ofadverse events with oral semaglutideversus empagliflozin mainly as a resultof gastrointestinal symptoms associatedwith dose escalation. The proportion ofadverse events leading to discontinuationof oral semaglutide (10.7%) was similarto previous observations with injectableGLP-1RAs (6–11%) (4,33,34).

The use of subcutaneous semaglutidehas previously been associated with ahigher rate of diabetic retinopathy–related complications compared withplacebo, which is consistent with thephenomenon of early worsening of pre-existing diabetic retinopathy secondaryto an initial, rapid improvement in glyce-mic control (6,35). The possible effect ofsubcutaneous semaglutide on diabeticeye disease is being further investigatedin the ongoing FOCUS trial (NCT03811561)(36). In the current trial, diabetic retinopathy–related adverse events were morefrequentwith oral semaglutide comparedwith empagliflozin, although occurrencewas low in both groups (3.4%vs. 1.2%).Allevents were nonserious, most were mildin severity, and none required treat-ment or led to trial product discontinu-ation. All were discovered during routine

Figure 1—Continued

care.diabetesjournals.org Rodbard and Associates 2277

Table

2—Se

lected

seco

ndary

endpoints

(treatm

entpolicy

estim

andandtrialpro

duct

estim

and)a

Treatm

entpolicyestimand

Trialproduct

estimand

Week26

Week52

Week26

Week52

Oralsemaglutide

14mg

Empagliflozin

25mg

Oralsemaglutide

14mg

Empagliflozin

25mg

Oralsemaglutide

14mg

Empagliflozin

25mg

Oralsemaglutide

14mg

Empagliflozin

25mg

Patien

ts,n

411

410

411

410

411

410

411

410

HbA1c#6.5%

(48mmol/mol)

Patien

tsreachingen

dpoint,n(%

)18

6(47.4)

68(17.2)

182(47.4)

83(21.7)

181(52.2)

68(18.0)

164(54.1)

74(23.4)

Estimated

OR(95%

CI)oralsemaglutide

vs.em

pagliflozin

4.62

(3.28,

6.52);P,

0.0001

3.36

(2.43,

4.66

);P,

0.00

015.61

(3.93,

8.01

);P,

0.00

014.32

(3.05,

6.13);P,

0.0001

Bodyweightreduction$10

%Patien

tsreachingen

dpoint,n(%

)49

(12.5)

27(6.8)

58(15.0)

30(7.8)

49(14.1)

27(7.1)

56(18.2)

28(8.7)

Estimated

OR(95%

CI)oralsemaglutide

vs.em

pagliflozin

1.98

(1.21,

3.25);P5

0.0066

2.05

(1.28,

3.28

);P5

0.00

282.18

(1.33,

3.57

);P5

0.00

212.51

(1.57,

4.01);P,

0.0001

HbA1c,7%

(53mmol/mol)without

hypoglycem

ia†andnoweightgain

Patien

tsreachingen

dpoint,n(%

)23

7(60.5)

141(35.7)

214(55.7)

149(39.0)

222(64.0)

139(36.8)

191(63.0)

139(44.0)

Estimated

OR(95%

CI)oralsemaglutidevs.

empagliflozin

2.88

(2.12,

3.91);P,

0.0001

2.03

(1.50,

2.74

);P,

0.00

013.31

(2.40,

4.56

);P,

0.00

012.39

(1.74,

3.30);P,

0.0001

HbA1creduction$1%

(10.9mmol/mol)

andbodyweightloss

$3%

Patien

tsreachingen

dpoint,n(%

)17

7(45.2)

111(28.1)

164(42.7)

101(26.4)

172(49.6)

110(29.1)

148(48.8)

91(28.8)

Estimated

OR(95%

CI)oralsemaglutide

vs.em

pagliflozin

2.10

(1.55,

2.85);P,

0.0001

2.10

(1.54,

2.87

);P,

0.00

012.41

(1.77,

3.29

);P,

0.00

012.33

(1.69,

3.21);P,

0.0001

Waist

circumference

(cm)

Estimated

mean

104.8

105.5

105.1

105.7

104.5

105.6

104.4

105.6

Estimated

meanchange

from

baseline

23.7

23.0

23.5

22.9

24.1

23.0

24.2

22.9

ETD(95%

CI)oralsemaglutidevs.em

pagliflozin

20.7(–1.4,

–0.0);P5

0.0400

20.6(–1.4,

0.2);P5

0.1488

21.1(–1.8,

–0.4);P5

0.00

3321.3(–2.1,

–0.4);P5

0.0030

Fastingplasm

aglucose,mmol/L(m

g/dL)

Estimated

mean

7.59

(136.8)

7.57

(136.5)

7.58

(136.6)

7.50

(135.1)

7.39

(133.2)

7.60

(137

.0)

7.48

(134

.7)

7.58

(136

.7)

Estimated

meanchange

from

baseline

21.99

(–35

.9)

22.01

(–36

.3)

22.01

(–36

.2)

22.09

(–37

.6)

22.19

(–39

.5)

21.99

(–35

.8)

22.11

(–38

.1)

22.00

(–36

.1)

ETD(95%

CI)oralsemaglutidevs.

empagliflozin

0.02

(–0.24

,0.28

),0.4(–4.3,

5.0);

P5

0.88

120.08

(–0.20,0.36

),1.4(–3.6,

6.4);

P5

0.5759

20.21

(–0.44,0.03),–3.7(–8.0,0.5);

P5

0.0874

20.11

(–0.37,0.15),–2.0(–6.6,2.6);

P5

0.4016

Mean7-pointSM

BG,mmol/L(m

g/dL)

Estimated

mean

8.0(143.7)

8.3(148.8)

7.9(142.4)

8.2(147.4)

7.7(138.8)

8.3(148.7)

7.7(138

.5)

8.2(147

.1)

Estimated

meanchange

from

baseline

22.2(–39

.8)

21.9(–34

.7)

22.3(–41

.1)

22.0(–36

.1)

22.4(–44

.0)

21.9(–34

.0)

22.5(–44

.3)

22.0(–35

.7)

ETD(95%

CI)oralsemaglutidevs.em

pagliflozin

20.3(–0.5,

–0.0),–5.0(–9.5,

–0.6);

P5

0.02

6720.3(–0.5,

–0.0),–5.1(–9.7,

–0.4);

P5

0.0328

20.6(–0.8,–0.3),–10

.0(–13

.8,–6.1);

P,

0.0001

20.5(–0.7,–0.2),–8.7(–12

.9,–4.4);

P,

0.0001

Con

tinu

edon

p.22

79

2278 Oral Semaglutide Versus Empagliflozin Diabetes Care Volume 42, December 2019

Table

2—Continued

Treatm

entpolicyestimand

Trialproduct

estimand

Week26

Week52

Week26

Week52

Oralsemaglutide

14mg

Empagliflozin

25mg

Oralsemaglutide

14mg

Empagliflozin

25mg

Oralsemaglutide

14mg

Empagliflozin

25mg

Oralsemaglutide

14mg

Empagliflozin

25mg

7-pointSM

BGpostprandialincrem

ent,mmol/L(m

g/dL)

Estimated

mean

1.5(27.5)

1.7(30.0)

1.3(23.2)

1.7(30.7)

1.4(25.1)

1.6(29.1)

1.2(22.5)

1.7(30.2)

Estimated

meanchange

from

baseline

20.5(–8.7)

20.3(–6.2)

20.7(–13

.0)

20.3(–5.5)

20.6(–11

.4)

20.4(–7.3)

20.8(–14

.0)

20.3(–6.3)

ETD(95%

CI)oralsemaglutidevs.em

pagliflozin

20.1(–0.4,

0.1),–2.5(–6.6,

1.6);

P5

0.23

8820.4(–0.6,–0.2),–7.5(–11

.5,–3.4);

P5

0.0003

20.2(–0.4,

–0.0),–4.1(–7.9,

–0.3);

P5

0.0356

20.4(–0.6,–0.2),–7.7(–11

.5,–3.9);

P,

0.0001

FastingC-pep

tide(nmol/L)

Estimated

mean

0.958

0.798

0.959

0.827

0.96

40.79

90.96

40.80

5Estimated

ratioto

baseline

1.08

0.90

1.09

0.94

1.09

0.91

1.09

0.91

Estimated

treatm

entratio(95%

CI)oralsem

aglutide

vs.em

pagliflozin

1.20

(1.15,

1.25);P,

0.0001

1.16

(1.11,

1.22

);P,

0.00

011.21

(1.15,

1.26

);P,

0.00

011.20

(1.14,

1.26);P,

0.0001

Fastinginsulin

(pmol/L)

Estimated

mean

8865

8666

8763

8363

Estimated

ratioto

baseline

1.06

0.78

1.03

0.79

1.03

0.75

0.99

0.75

Estimated

treatm

entratio(95%

CI)oralsem

aglutide

vs.em

pagliflozin

1.35

(1.26,

1.46);P,

0.0001

1.31

(1.21,

1.41

);P,

0.00

011.37

(1.29,

1.46

);P,

0.00

011.32

(1.23,

1.41);P,

0.0001

Fastingproinsulin

(pmol/L)

Estimated

mean

18.3

17.1

18.7

17.9

17.3

16.8

17.6

17.0

Estimated

meanchange

from

baseline

0.72

0.68

0.74

0.71

0.68

0.66

0.69

0.67

ETD(95%

CI)oralsemaglutidevs.em

pagliflozin

1.07

(0.98,

1.17);P5

0.1403

1.05

(0.96,

1.14

);P5

0.30

341.03

(0.94,

1.12

);P5

0.54

531.04

(0.95,

1.14);P5

0.4479

Fastingglucagon(pg/mL)

Estimated

mean

8695

8490

8595

8489

Estimated

ratioto

baseline

0.92

1.01

0.89

0.95

0.90

1.01

0.89

0.95

Estimated

treatm

entratio(95%

CI)oralsem

aglutide

vs.em

pagliflozin

0.91

(0.88,

0.94);P,

0.0001

0.94

(0.90,

0.97

);P5

0.00

080.89

(0.86,

0.92

);P,

0.00

010.94

(0.90,

0.97);P5

0.0011

C-reactiveprotein

(mg/L)

Estimated

mean

1.85

2.65

1.81

2.45

1.78

2.68

1.71

2.45

Estimated

ratioto

baseline

0.69

0.99

0.67

0.91

0.65

0.98

0.63

0.90

Estimated

treatm

entratio(95%

CI)oralsem

aglutide

vs.em

pagliflozin

0.70

(0.62,

0.79);P,

0.0001

0.74

(0.65,

0.84

);P,

0.00

010.66

(0.58,

0.75

);P,

0.00

010.70

(0.61,

0.80);P,

0.0001

Treatm

entp

olicyestimand:ANCOVAforcontinuousen

dpointsandlogisticregressionforb

inaryen

dpoints,usingdatairrespective

ofdiscontinuationoftrialproductorinitiationofrescuemed

ication.M

issing

values

wereim

putedbyapattern

mixture

modelusingmultipleim

putation.P

atternsweredefi

ned

byuse

oftrialp

roductandrescuemed

ication.Trialproduct

estimand:m

ixed

modelforrepeatedmeasuremen

tsfor

continuousen

dpointsandlogisticregressionforbinaryen

dpoints.D

atacollected

afterdiscontinuationoftrialproductorinitiationofrescuemed

icationareexcluded

.Forbinaryen

dpoints,m

issingvalues

were

imputedfrom

patients

randomized

tosametrialp

roduct

usingsequen

tialmultiple

imputation.Pvalues

aretw

o-sided

andunadjusted

forthetest

ofnodifference.%,proportionofpatients

withnonmissing

inform

ation;OR,o

ddsratio.aAdditionalen

dpointsarereported

inSupplem

entary

Table4.

†Severe

orbloodglucose–confirm

ed(plasm

aglucose

,3.1mmol/L[56mg/dL])symptomatichypoglycem

icep

isode.

care.diabetesjournals.org Rodbard and Associates 2279

end-of-treatment eye examination andwere diagnosed as nonproliferative dia-betic retinopathy. In a longer-term,78-week, double-blind trial, no im-balance in the occurrence of diabeticretinopathy–related events was observedbetween oral semaglutide 3, 7, and 14 mgand sitagliptin (6.7%, 6.0%, 5.6%, and7.7%, respectively) (15). Occurrence ofdiabetic retinopathy–related events wasalso similar with oral semaglutide andplacebo (7.1% vs. 6.3%) in a double-blindtrial that assessed cardiovascular out-comes in patients at high cardiovascularrisk (17).This trial provides a comparison of two

increasingly used drug classes that arecommonly added to metformin whenglycemic control is not achieved. The

principal limitation of the trial was theopen-label design.

In conclusion, theoralGLP-1 analogoralsemaglutide was superior to the SGLT-2inhibitor empagliflozin for reduction inHbA1c, but not body weight, at 26 weeksin patients with type 2 diabetes uncon-trolled with metformin. Reductions inHbA1c were significantly greater withoral semaglutide at 52 weeks. Assessedby the trial product estimand, oralsemaglutide provided significant reductionsin body weight at 52 weeks. Oral sema-glutide was well tolerated, with a safetyprofile consistent with that of GLP-1RAs.

Acknowledgments. Emisphere is acknowl-edged for providing a license to the Eligen

Technology, the sodium N-(8-[2-hydroxylbenzoyl]amino) caprylate component of oral semaglutide.The authors thank the patients, investigators,trial site staff, and Novo Nordisk employeesinvolved in the trial. In addition, the authorsthank Andy Bond of Spirit Medical Communica-tions Group Ltd. for medical writing and editorialassistanceandBrianBekkerHansenofNovoNordiskfor reviewing the manuscript. Novo Nordisk (thesponsor) designed the trial, monitored sites, andcollected and analyzed the data. Editorial sup-port was funded by the sponsor and providedby independent medical writers under the guid-ance of the authors.Duality of Interest. This trial was funded by NovoNordisk A/S, Denmark. H.W.R. reports consulting,advisoryboards, clinical research, and lecturing forAstraZeneca, Boehringer Ingelheim, Janssen, EliLilly, Merck, Novo Nordisk, Sanofi, and RegeneronPharmaceuticals. J.R. reports scientific advisoryboards and honoraria or consulting fees from EliLilly, Novo Nordisk, Sanofi, Janssen, BoehringerIngelheim, and Intarcia and grants/research sup-port from Merck, Pfizer, Sanofi, Novo Nordisk,Bristol-Myers Squibb, Eli Lilly, GlaxoSmithKline,AstraZeneca, Janssen, Genentech, BoehringerIngelheim, Intarcia, and Lexicon. L.H.C. reportsclinical research for Novo Nordisk, Janssen, EliLilly, and Sanofi and lecturing for Sanofi andBoehringer Ingelheim. C.D. reports consulting,advisory boards, clinical research, and lecturingfor AstraZeneca, Boehringer Ingelheim, Janssen,Eli Lilly, Novo Nordisk, Merck, Sanofi, Takeda, andMerck Sharp & Dohme. J.G. has received speaker’sor consulting honoraria from Novo Nordisk, EliLilly, Servier, Merck Sharp & Dohme, Bioton(Poland), Merck (Darmstadt), Sanofi, Pol-pharma (Poland), Polfa Tarchomin (Poland),AstraZeneca, and Boehringer Ingelheim.S.Ø.L., A.L.S., and M.B.T. are employees ofNovo Nordisk A/S. A.L.S. and M.B.T. haveshares in Novo Nordisk A/S. I.L. reports consulting,advisory boards, and/or research grants fromNovo Nordisk, AstraZeneca, Boehringer Ingel-heim, Sanofi, Eli Lilly, Intarcia, MannKind, Va-leritas, Novartis, Mylan, Merck, and Pfizer. E.M.reports scientific advisory boards, consulting,lecturing, and/or research grants from Astra-Zeneca, Boehringer Ingelheim, Eli Lilly, GrupoFerrer Internacional S.A., Intarcia, Menarini,Janssen, Servier, Merck Sharp & Dohme, NovoNordisk, and Novartis. No other potential con-flicts of interest relevant to this article werereported.Author Contributions. H.W.R., J.R., L.H.C.,C.D., J.G., S.Ø.L., I.L., A.L.S., M.B.T., andE.M. were responsible for the acquisition,analysis, or interpretation of data and thedrafting or critical revision of the manuscriptfor important intellectual content. H.W.R. andE.M. were signatory investigators on thestudy. S.Ø.L. and M.B.T. were involved inthe concept and design of the study. A.L.S.was responsible for the statistical analysis.H.W.R. and E.M. are the guarantors of thiswork and, as such, had full access to all the datain the study and take responsibility forthe integrity of the data and the accuracyof the data analysis.Prior Presentation. Parts of this study werepresented in oral form at the 79th Scientific

Table 3—On-treatment adverse events

Patients, n (%)

Oral semaglutide 14 mg(n 5 410)

Empagliflozin 25 mg(n 5 409)

Adverse events 289 (70.5) 283 (69.2)

Serious adverse events 27 (6.6) 37 (9.0)

Adverse event severityMild 242 (59.0) 240 (58.7)Moderate 140 (34.1) 118 (28.9)Severe 24 (5.9) 23 (5.6)

Severe or blood glucose–confirmedsymptomatic hypoglycemic episode*†‡ 7 (1.7) 8 (2.0)

ADA-classified hypoglycemic episode* 45 (11.0) 39 (9.5)Severe hypoglycemic episode*† 1 (0.2) 1 (0.2)

Most frequent adverse events $5% ineither group (preferred term)

Nausea 81 (19.8) 10 (2.4)Diarrhea 38 (9.3) 13 (3.2)Vomiting 30 (7.3) 7 (1.7)Decreased appetite 21 (5.1) 2 (0.5)Influenza 8 (2.0) 21 (5.1)

Adverse events resulting in premature trialdrug discontinuation 44 (10.7) 18 (4.4)

Adverse events resulting in prematuretrial drug discontinuation (.1% forany system organ class or preferredterm)

Gastrointestinal disorders 33 (8.0) 3 (0.7)Nausea 21 (5.1) 2 (0.5)Vomiting 11 (2.7) 1 (0.2)Abdominal pain 5 (1.2) 0

Infections and infestations 0 5 (1.2)

Deaths 0 1 (0.2)§

Safety end points were assessed using the safety analysis set (all patients exposed to one or moredoses of trial product) and evaluated for both the on-treatment period (while on trial product) andthe in-trial period (while in trial, regardless of discontinuation of trial product or use of rescuemedication). *Hypoglycemic episodeswere reportedon a separate form fromadverseevents.†Anepisode that is severe, according to theADAclassification (requires assistanceof anotherperson toactively administer carbohydrate, glucagon, or other corrective action) (29). ‡Blood glucoseconfirmation of symptomatic hypoglycemia was based on a blood glucose value,56mg/dL, withsymptoms consistent with hypoglycemia. §One patient in the empagliflozin group died as a resultof undetermined reasons after 268 days on trial drug.

2280 Oral Semaglutide Versus Empagliflozin Diabetes Care Volume 42, December 2019

Sessions of the American Diabetes Association,San Francisco, CA, 7–11 June 2019.

References1. American Diabetes Association. 9. Pharma-cologic approaches to glycemic treatment:Standards of Medical Care in Diabetesd2019.Diabetes Care 2019;42(Suppl. 1):S90–S1022. Davies MJ, D’Alessio DA, Fradkin J, et al.Management of hyperglycemia in type 2 diabe-tes, 2018. A consensus report by the AmericanDiabetes Association (ADA) and the EuropeanAssociation for the Study of Diabetes (EASD).Diabetes Care 2018;41:2669–27013. AhmannAJ, CapehornM, Charpentier G, et al.Efficacy and safety of once-weekly semaglutideversus exenatide ER in subjects with type 2diabetes (SUSTAIN 3): a 56-week, open-label,randomized clinical trial. Diabetes Care 2018;41:258–2664. Ahren B,Masmiquel L, Kumar H, et al. Efficacyand safety of once-weekly semaglutide versusonce-daily sitagliptin as an add-on to metformin,thiazolidinediones, or both, in patients with type2 diabetes (SUSTAIN 2): a 56-week, double-blind,phase 3a, randomised trial. Lancet Diabetes En-docrinol 2017;5:341–3545. Aroda VR, Bain SC, Cariou B, et al. Efficacy andsafety of once-weekly semaglutide versus once-daily insulin glargine as add-on to metformin(with or without sulfonylureas) in insulin-naivepatients with type 2 diabetes (SUSTAIN 4):a randomised, open-label, parallel-group, multi-centre, multinational, phase 3a trial. Lancet Di-abetes Endocrinol 2017;5:355–3666. Marso SP, Bain SC, Consoli A, et al.; SUSTAIN-6Investigators. Semaglutide and cardiovascularoutcomes in patients with type 2 diabetes. NEngl J Med 2016;375:1834–18447. PratleyRE, ArodaVR, Lingvay I, et al.; SUSTAIN7 investigators. Semaglutide versus dulaglutideonce weekly in patients with type 2 diabetes(SUSTAIN 7): a randomised, open-label, phase 3btrial. Lancet Diabetes Endocrinol 2018;6:275–2868. RodbardHW,Lingvay I, Reed J, et al. Semaglutideaddedtobasal insulin intype2diabetes(SUSTAIN5):a randomized, controlled trial. J Clin EndocrinolMetab 2018;103:2291–23019. Sorli C, Harashima SI, Tsoukas GM, et al.Efficacy and safety of once-weekly semaglutidemonotherapy versus placebo in patients withtype 2 diabetes (SUSTAIN 1): a double-blind,randomised, placebo-controlled, parallel-group,multinational, multicentre phase 3a trial. LancetDiabetes Endocrinol 2017;5:251–26010. Buckley ST, Bækdal TA, Vegge A, et al. Trans-cellular stomach absorption of a derivatizedglucagon-like peptide-1 receptor agonist. SciTransl Med 2018;10:eaar704711. Davies M, Pieber TR, Hartoft-Nielsen ML,Hansen OKH, Jabbour S, Rosenstock J. Effect oforal semaglutide compared with placebo andsubcutaneous semaglutide on glycemic controlin patients with type 2 diabetes: a randomizedclinical trial. JAMA 2017;318:1460–147012. Aroda VR, Rosenstock J, Terauchi Y, et al.PIONEER 1: randomized clinical trial of the

efficacy and safety of oral semaglutide mono-therapy in comparison with placebo in patientswith type 2 diabetes. Diabetes Care 2019;42:1724–173213. Pratley R, Amod A, Hoff ST, et al. Oralsemaglutide versus subcutaneous liraglutide andplacebo in type 2 diabetes (PIONEER 4): a rando-mised, double-blind, phase 3a trial. Lancet 2019;394:39–5014. MosenzonO, Blicher TM, Rosenlund S, et al.;PIONEER 5 Investigators. Efficacy and safety oforal semaglutide in patients with type 2 diabetesand moderate renal impairment (PIONEER 5):a placebo-controlled, randomised, phase 3a trial.Lancet Diabetes Endocrinol 2019;7:515–52715. Rosenstock J, Allison D, Birkenfeld AL, et al.;PIONEER 3 Investigators. Effect of additional oralsemaglutide vs sitagliptin on glycated hemoglo-bin in adults with type 2 diabetes uncontrolledwith metformin alone or with sulfonylurea: thePIONEER 3 randomized clinical trial. JAMA 2019;321:1466–148016. PieberTR,BodeB,MertensA, etal.; PIONEER7 investigators. Efficacy and safety of oral sem-aglutide with flexible dose adjustment versussitagliptin in type2diabetes (PIONEER7): amulti-centre, open-label, randomised, phase 3a trial.Lancet Diabetes Endocrinol 2019;7:528–53917. HusainM, Birkenfeld AL, DonsmarkM, et al.;PIONEER 6 Investigators. Oral semaglutide andcardiovascular outcomes in patients with type 2diabetes. N Engl J Med 2019;381:841–85118. RodenM,WengJ, Eilbracht J, etal.; EMPA-REGMONO trial investigators. Empagliflozin mono-therapy with sitagliptin as an active comparatorin patients with type 2 diabetes: a randomised,double-blind, placebo-controlled, phase 3 trial.Lancet Diabetes Endocrinol 2013;1:208–21919. Haring HU, Merker L, Seewaldt-Becker E,et al.; EMPA-REG METSU Trial Investigators.Empagliflozin as add-on to metformin plussulfonylurea in patients with type 2 diabetes:a 24-week, randomized, double-blind, placebo-controlled trial. Diabetes Care 2013;36:3396–340420. Haring HU, Merker L, Seewaldt-Becker E,et al.; EMPA-REG MET Trial Investigators. Em-pagliflozin as add-on to metformin in patientswith type 2 diabetes: a 24-week, randomized,double-blind, placebo-controlled trial. DiabetesCare 2014;37:1650–165921. Kovacs CS, Seshiah V, Swallow R, et al.;EMPA-REGPIO� trial investigators. Empagliflozinimproves glycaemic and weight control asadd-on therapy to pioglitazone or pioglitazoneplus metformin in patients with type 2 diabetes:a 24-week, randomized, placebo-controlled trial.Diabetes Obes Metab 2014;16:147–15822. RidderstraleM,AndersenKR, Zeller C, KimG,Woerle HJ, Broedl UC; EMPA-REG H2H-SU trialinvestigators. Comparison of empagliflozin andglimepiride as add-on to metformin in patientswith type 2 diabetes: a 104-week randomised,active-controlled, double-blind, phase 3 trial.Lancet Diabetes Endocrinol 2014;2:691–70023. ZinmanB,WannerC, Lachin JM,et al.; EMPA-REG OUTCOME Investigators. Empagliflozin,

cardiovascular outcomes, and mortality intype 2 diabetes. N Engl J Med 2015;373:2117–212824. International Council on Harmonization.Statistical principles for clinical trials E9 [Inter-net], 1998. Available from https://www.ich.org/products/guidelines/efficacy/article/efficacy-guidelines.html. Accessed 24 April 201925. International Council on Harmonization. Esti-mands and sensitivity analysis in clinical trials E9(R1) [Internet], 2017. Available from https://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E9/E9-R1EWG_Step2_Guideline_2017_0616.pdf. Accessed 24 April 201926. Aroda VR, Saugstrup T, Buse JB, DonsmarkM, Zacho J, Davies MJ. Incorporating and inter-preting regulatory guidance on estimands indiabetesclinical trials: thePIONEER1randomizedclinical trial as an example. Diabetes ObesMetab2019;21:2203–221027. Maruish ME (Ed.). User’s Manual for theSF-36v2 Health Survey. Lincoln, RI, QualityMetricInc., 201128. Dalton M, Finlayson G, Hill A, Blundell J.Preliminary validation and principal componentsanalysis of the Control of Eating Questionnaire(CoEQ) for the experience of food craving. Eur JClin Nutr 2015;69:1313–131729. Seaquist ER, Anderson J, Childs B, et al.Hypoglycemia and diabetes: a report of a work-group of the American Diabetes Association andthe Endocrine Society. Diabetes Care 2013;36:1384–139530. Bretz F, Posch M, Glimm E, Klinglmueller F,MaurerW, Rohmeyer K. Graphical approaches formultiple comparison procedures using weightedBonferroni, Simes, or parametric tests. Biom J2011;53:894–91331. Little RJA, Rubin DB. Statistical Analysis withMissing Data. New York, JohnWiley & Sons, 198732. Koch GG. Comments on ‘Current issues innon-inferiority trials’ by Thomas R. Fleming,Statistics in Medicine, DOI: 10.1002/sim.2855.Stat Med 2008;27:333–34233. Pratley RE, Nauck M, Bailey T, et al.; 1860-LIRA-DPP-4 Study Group. Liraglutide versussitagliptin for patients with type 2 diabetes who didnot have adequate glycaemic control with met-formin: a 26-week, randomised, parallel-group,open-label trial. Lancet 2010;375:1447–145634. Nauck M, Weinstock RS, Umpierrez GE,Guerci B, Skrivanek Z, Milicevic Z. Efficacy andsafety of dulaglutide versus sitagliptin after52 weeks in type 2 diabetes in a randomizedcontrolled trial (AWARD-5). Diabetes Care 2014;37:2149–215835. Vilsbøll T, Bain SC, Leiter LA, et al. Semaglutide,reduction in glycated haemoglobin and the riskof diabetic retinopathy. Diabetes Obes Metab2018;20:889–89736. Novo Nordisk A/S. A research study to lookat how semaglutide compared to placebo affectsdiabetic eye disease in people with type 2 di-abetes (FOCUS). In: Clinicaltrials.gov [Internet].Bethesda,MD,National Library ofMedicine,2019.Available from https://clinicaltrials.gov/ct2/show/NCT03811561. NLM Identifier: NCT03811561. Ac-cessed 2 September 2019

care.diabetesjournals.org Rodbard and Associates 2281