SHORT COMMUNICATION

Efficacy and safety of methionine aminopeptidase 2 inhibition in type 2diabetes: a randomised, placebo-controlled clinical trial

Joseph Proietto1& Jaret Malloy2 & Dongliang Zhuang2

& Mark Arya3 & Neale D. Cohen4& Ferdinandus J. de Looze5,6

&

Christopher Gilfillan7& Paul Griffin8,9,10,11

& Stephen Hall12,13 & Thomas Nathow14& Geoffrey S. Oldfield15

&

David N. O’Neal16 & Adam Roberts17 & Bronwyn G. A. Stuckey18 & Dennis Yue19& Kristin Taylor2 & Dennis Kim2

Received: 14 March 2018 /Accepted: 24 May 2018 /Published online: 11 July 2018# Springer-Verlag GmbH Germany, part of Springer Nature 2018

AbstractAims/hypothesis This multicentre randomised double-blind placebo-controlled clinical trial assessed the efficacy and safety of amethionine aminopeptidase 2 (MetAP2) inhibitor, beloranib, in individuals with obesity (BMI ≥30 kg/m2) and type 2 diabetes(HbA1c 53–97 mmol/mol [7–11%] and fasting glucose <15.6 mmol/l).Methods Participants were randomised (via a centralised interactive web response system) to placebo, 1.2 or 1.8 mg beloranibs.c. twice weekly for 26 weeks. Participants, investigators and the sponsor were blinded to group assignment. The primaryendpoint was the change in weight from baseline to week 26. The trial was terminated early when beloranib development wasstopped because of an imbalance of venous thromboembolism events in beloranib-treated individuals vs placebo that becameevident during late-stage development of the drug.Results In total, 153 participants were randomised, 51 to placebo, 52 to 1.2 mg beloranib and 50 to 1.8 mg beloranib. Inparticipants who completed week 26, the least squares mean ± SEweight change (baseline 111 kg) was −3.1 ± 1.2%with placebo(n = 22) vs −13.5 ± 1.1% and −12.7 ± 1.3% with 1.2 and 1.8 mg beloranib, respectively (n = 25; n = 19; p < 0.0001). The changein HbA1c (baseline 67 mmol/mol [8.3%]) was −6.6 ± 2.2 mmol/mol (−0.6 ± 0.2%) with placebo vs −21.9 ± 2.2 mmol/mol (−2.0± 0.2%) or −21.9 ± 3.3 mmol/mol (−2.0 ± 0.3%) with 1.2 or 1.8 mg beloranib (p < 0.0001), respectively. The most commonbeloranib adverse events were sleep related. One beloranib-treated participant experienced a non-fatal pulmonary embolism.

Electronic supplementary material The online version of this article(https://doi.org/10.1007/s00125-018-4677-0) contains peer-reviewed butunedited supplementary material, which is available to authorised users.

* Dennis Kimdkim@zafgen.com

1 Department of Medicine (Austin Health), University of Melbourne,Heidelberg, VIC, Australia

2 Zafgen, Inc., 175 Portland St, 4th Floor, Boston, MA 02114, USA3 Australian Clinical Research Network, Maroubra, NSW, Australia4 Baker Heart and Diabetes Institute, Melbourne, VIC, Australia5 AusTrials Pty Ltd, Sherwood, QLD, Australia6 Discipline of General Practice, Faculty of Medicine, University of

Queensland, Herston, QLD, Australia7 Eastern Health Clinical School, Monash University, Box Hill, VIC,

Australia8 Q-Pharm Pty Ltd, Herston, QLD, Australia9 QIMR Berghofer Medical Research Institute, Herston, QLD,

Australia

10 Mater Health Services, South Brisbane, QLD, Australia

11 Faculty of Medicine, University of Queensland, Herston, QLD,Australia

12 Emeritus Research, Malvern East, VIC, Australia

13 Institution for Rehabilitation Research, Monash University,Clayton, VIC, Australia

14 Ipswich Research Institute, Ipswich, QLD, Australia

15 Pendlebury Research, Cardiff, NSW, Australia

16 Department of Medicine, University of Melbourne, St Vincent’sHospital, Fitzroy, VIC, Australia

17 University Hospital Geelong, Geelong, VIC, Australia

18 Keogh Institute for Medical Research, University of WesternAustralia, Nedlands, WA, Australia

19 Royal Prince Alfred Hospital, Camperdown, NSW, Australia

Diabetologia (2018) 61:1918–1922https://doi.org/10.1007/s00125-018-4677-0

Conclusions/interpretation MetAP2 inhibitors represent a novel mechanism for producing meaningful weight loss and improve-ment in HbA1c.Trial registration: ClinicalTrials.gov NCT02324491Funding: The study was funded by Zafgen, Inc.

Keywords Anti-obesitymedication . Glucose-loweringmedication . Glycaemic control . MetAP2

AbbreviationsAE Adverse eventhsCRP High-sensitivity C-reactive proteinMetAP2 Methionine aminopeptidase 2PWS Prader–Willi syndrome

Introduction

Methionine aminopeptidase 2 (MetAP2) inhibitors are a noveldrug class being investigated for the treatment of type 2 dia-betes, obesity and metabolic disease. In preclinical models ofobesity and diabetes, MetAP2 inhibitors produce weight losscharacterised by markedly reduced adiposity and increasedglycaemic control, as well as transiently reduced food intake[1, 2]. The MetAP2 inhibitor beloranib has demonstrated con-sistent and substantial weight loss and glucose-lowering ef-fects in clinical studies of general obesity, hypothalamic-injury-associated obesity and Prader–Willi syndrome (PWS)

[3–6]. This phase 2 clinical trial is the first to study the effectsof MetAP2 inhibition with beloranib compared with placeboon glycaemic control and bodyweight in individuals with type2 diabetes and obesity.

Methods

This multicentre randomised double-blind placebo-controlledphase 2 trial was conducted at 16 investigational sites inAustralia from December 2014 to January 2016. The trialwas terminated early because of an imbalance of venousthromboembolism events in beloranib-treated vs placebo-treated individuals in beloranib clinical trials. The institutionalreview boards at all clinical sites approved the protocol beforestudy initiation. All participants provided written informedconsent. A sample size of 50 people per group was to provide91% power to detect a mean difference of 5.0 kg assuming astandard deviation of 7.5 kg.

Diabetologia (2018) 61:1918–1922 1919

Eligible individuals aged 18–65 years with BMI ≥30 kg/m2

had type 2 diabetes with HbA1c 53–97 mmol/mol (7%–11%)and fasting plasma glucose <15.6mmol/l andwere treated withdiet/exercise or a stable dose of non-insulin glucose-loweringmedication. Participants were randomised in a 1:1:1 ratio via acentralised interactive web response system to placebo, 1.2 mgbeloranib or 1.8 mg beloranib administered subcutaneouslytwice weekly (electronic supplementary material [ESM]Fig. 1). Randomisation was stratified based on sulfonylureause (yes/no) and screening HbA1c (< or ≥75 mmol/mol [< or≥9%]). Lifestyle (diet/exercise) counselling was provided to allparticipants.

The primary endpoint was the change in weight from base-line to week 26. Key secondary endpoints included the changefrom baseline to week 26 in glycaemic variables, the propor-tion of participants achieving weight loss targets and HbA1c

targets and the change in cardiometabolic variables. Safetyevaluations throughout the study included assessment of inci-dence and severity of adverse events (AEs) as well as changesin laboratory measures, physical examination, vital signs andother safety measures.

The primary efficacy analysis was conducted on the perprotocol population (n = 66), which included all randomisedparticipants who received ≥75% of planned doses, had an end-of-treatment efficacy assessment and did not have any notableprotocol violations. Data were analysed using an ANCOVAmodel with change in weight as the dependent variable, treat-ment group, sulfonylurea use and baseline HbA1c as factorsand baseline weight as a covariate without imputation ofmissing data. Analyses performed on the per protocol popu-lation used observed data. Safety analyses were conducted onthe safety population. The safety and intent-to-treat popula-tions (n = 152 each) included all randomised participants whoreceived ≥1 dose of trial drug.

Data management and randomisation services were per-formed by BioClinica (Audubon, PA, USA) and statisticalanalyses were conducted by Agility Clinical (Carlsbad, CA,USA) and Pharmapace (San Diego, CA, USA). Data are avai-lable from the corresponding author on reasonable request.

Results

Participant disposition during the trial is shown in ESMFig. 1.At the time of early study closure, 70 individuals had com-pleted the study, 65 had not completed 26 weeks of treatmentand 18 had discontinued the study (11 withdrew because ofAEs, five withdrew consent, one was lost to follow up and onediscontinued because of investigator decision).

Individuals in the intent-to-treat population were primarilywhite (89%), non-Hispanic (98%) and male (57%), with mean± SD baseline age 54 ± 7 years, BMI 39.9 ± 7.3 kg/m2, HbA1c

67 ± 11 mmol/mol (8.3 ± 1.0%) and duration of diabetes of

7.1 ± 4.9 years. The majority (93%) of participants weretreated with ≥1 glucose-loweringmedication, most commonlybiguanides (72%) and sulfonylureas (42%). There were nounexpected differences in demographics and baseline charac-teristics between the intent-to-treat and per protocol popula-tions (ESMTable 1).Mean ± SD exposures to placebo, 1.2 mgand 1.8 mg beloranib were 20.7 ± 6.8 (median 23.6 [min 1.6,max 27.4]) weeks, 19.5 ± 7.7 (23.9 [1.6, 26.6]) weeks and18.5 ± 8.3 (22.1 [0.4, 26.7]) weeks, respectively.

Weight loss with beloranib was observed as early as week 4and progressed through to week 26 (Fig. 1a). At week 26, theleast squares mean weight change (± SE) in the placebo groupwas −3.1 ± 1.2% compared with −13.5 ± 1.1% in the 1.2 mgbeloranib group (placebo-corrected difference [95% CI]:−10.4% [−13.6, −7.2]) and −12.7 ± 1.3% in the 1.8 mgbeloranib group (−9.6% [−12.9, −6.4]), both p < 0.0001)(ESM Table 2). Additionally, a substantially greater proportionof beloranib-treated participants achieved week 26 weight losstargets ≥5%, ≥10% and ≥15% comparedwith placebo (Fig. 1b).

At week 26, the least squares mean change in HbA1c (± SE)in the placebo group was −6.6 ± 2.2 mmol/mol (−0.6 ± 0.2%)compared with −21.9 ± 2.2 mmol/mol (−2.0 ± 0.2%) in the1.2 mg beloranib group (placebo-corrected difference [95%CI]: −15.3 mmol/mol [−21.9, −9.8] [−1.4% (−2.0, −0.9)])and −21.9 ± 3.3 mmol/mol (−2.0 ± 0.3%) in the 1.8 mgbeloranib group (−15.3 mmol/mol [−21.9, −8.7] [−1.4%(−2.0, −0.8)]); both p < 0.0001 (Fig. 1c, ESM Table 2).Additionally, a greater proportion of individuals in thebeloranib treatment groups achieved HbA1c levels<53 mmol/mol (<7%) and ≤48 mmol/mol (≤6.5%) at week26 compared with placebo (Fig. 1d). The week 26 leastsquares mean HbA1c in both beloranib groups was 45 mmol/mol (6.3%) and the mean HbA1c in the placebo group was61 mmol/mol (7.7%) (Fig. 1e).

Fasting plasma glucose was significantly reduced vsplacebo in both beloranib groups by week 8 and continuedto decline through week 26 in beloranib-treated individuals(ESM Fig. 2a). Beloranib also produced improvements inother glycaemic variables compared with placebo (post-prandial glucose, insulin sensitivity and beta cell function)and cardiometabolic markers (high-sensitivity C-reactiveprotein [hsCRP], adiponectin and leptin; ESM Fig. 2b, cand ESM Table 2).

More participants in the beloranib groups withdrew fromthe study because of an AE (n = 5 per beloranib dose group)than in the placebo group (n = 2). The most common reasonsfor withdrawal of beloranib were sleep related (e.g. insomnia,abnormal dreams, sleep disorder). There were no deaths. Themost common AEs with greater frequency in the beloranibgroups compared with placebo were sleep related (e.g. abnor-mal dreams, sleep disorder), diarrhoea, injection-site bruising,injection-site erythema, lower respiratory tract infection andcough (Table 1). Most AEs were mild to moderate in severity

1920 Diabetologia (2018) 61:1918–1922

and self-limiting. The incidence of hypoglycaemia AEs waslow and similar across treatment groups (4–6%); all were mildand resolved. There were three AEs of thromboembolic nature(superficial thrombophlebitis, deep vein thrombosis, non-fatalpulmonary embolism); all occurred in beloranib-treatedparticipants.

Discussion

Here, we report that treatment with a MetAP2 inhibitor for26 weeks produced placebo-corrected clinically meaningfulweight loss (−10%) and reductions in HbA1c (−15.3 mmol/mol [−1.4%]) in individuals with obesity and type 2 diabetes,and most participants achieved clinical HbA1c and obesitytreatment goals. Improvements in weight and HbA1c weresimilar for both beloranib doses, despite a higher baselineweight in the lower (1.2 mg) dose group, suggesting that thelower dose likely produced maximum efficacy.

The occurrence of unexpected serious venous thrombo-embolism AEs across beloranib clinical trials led to termi-nation of beloranib development and investigation of themechanism. Recently published data show that beloranibhas effects on endothelial cells that influence pro- andanticoagulant factors [7]. Furthermore, prolonged (>24 h)exposure, resulting from unique pharmacological proper-ties of beloranib [7] as well as the suspension formulation,likely exacerbates these effects. Importantly, other MetAP2inhibitors were recently shown to have similar beneficialeffects on body weight and glycaemic control, but with asubstantially improved safety profile compared withbeloranib [7].

Table 1 AEs occurring from baseline to week 26

AE Placebon = 51

1.2 mgbeloranibn = 52

1.8 mgbeloranibn = 49

Any serious AE 2 (4) 4 (8) 1 (2)

Exostosis (bone spur onamputated leg stump)

0 0 1 (2)

Musculoskeletal chest pain 0 1 (2) 0

Atrial fibrillation 0 1 (2) 0

Psychotic disorder (stress inducedmanic psychotic episode)

0 1 (2) 0

Pulmonary embolism 0 1 (2) 0

Abdominal pain (lower) 1 (2) 0 0

Nephrolithiasis 1 (2) 0 0

Any AE 41 (80) 48 (92) 41 (84)

AEs occurring in ≥10% of participants in any treatment group

Diarrhoea 7 (14) 10 (19) 9 (18)

Injection-site bruising 6 (12) 10 (19) 7 (14)

Sleep disorder 1 (2) 5 (10) 7 (14)

Injection-site erythema 1 (2) 2 (4) 6 (12)

Upper respiratory tract infection 10 (20) 14 (27) 6 (12)

Cough 2 (4) 3 (6) 5 (10)

Abnormal dreams 1 (2) 8 (15) 4 (8)

Nausea 10 (20) 5 (10) 4 (8)

Lower respiratory tract infection 2 (4) 7 (14) 3 (6)

Headache 8 (16) 7 (14) 2 (4)

Data are n (%) for the safety population (N = 152)

0 2 4 6 8 10 12 14 16 18 20 22 24 26

-16

-14

-12

-10

-8

-6

-4

-2

0

Week

0 2 4 6 8 10 12 14 16 18 20 22 24 26

Week

0 2 4 6 8 10 12 14 16 18 20 22 24 26

Week

Me

an

ch

an

ge

in

we

igh

t (%

)

***

***

***

***

***

***

***

***

***

***

-30

-25

-20

-15

-10

-5

0

Mean c

hange in H

bA

1c (

mm

ol/m

ol)

*

**

**

***

***

40

45

50

55

60

65

70

75

80

Mean H

bA

1c (

mm

ol/m

ol)

<53 mmol/mol ≤48 mmol/mol

0

20

40

60

80

100

** ****

**

Week 26 HbA1c

≥5% ≥10% ≥15%0

20

40

60

80

100

Week 26 weight loss

Pe

rce

nta

ge

of

pa

rticip

an

tsP

erce

nta

ge

of

pa

rticip

an

ts

******

***

***

** *

a b

c d

e

Fig. 1 (a) Least squares mean ± SE for the change from baseline inweight. (b) Percentage of participants with weight loss of ≥5%, ≥10%or ≥15% from baseline to week 26. (c) Least squares mean ± SE for thechange from baseline in HbA1c. (d) Percentage of participants whoachieved an HbA1c level of <53 mmol/mol (<7%) or ≤48 mmol/mol(≤6.5%) at week 26. (e) Least squares mean HbA1c by visit. Data reflectobserved data for the per protocol population (n = 66). For (a, c, e), whitecircles, placebo (n = 22); white squares, 1.2 mg beloranib (n = 25); blacksquares, 1.8 mg beloranib (n = 19). For (b, d), white bars, placebo (n =22); grey bars, 1.2 mg beloranib (n = 25); black bars, 1.8 mg beloranib(n = 19). The differences between the 1.2mg and 1.8mg beloranib groupsvs placebo for the body weight and HbA1c efficacy endpoints were con-sistent and statistically significant (p < 0.05) regardless of the analysispopulation (per protocol or intent to treat) or imputation strategy (lastobservation carried forward [LOCF] or mixed-model repeated measures[MMRM]) used to explore the sensitivity of the analysis (data notshown). *p < 0.05, **p < 0.01 and ***p < 0.0001 for beloranib vsplacebo

Diabetologia (2018) 61:1918–1922 1921

In addition to reductions in body weight and HbA1c,improvements in waist and hip circumference, fat mass,lipids, hsCRP, leptin and adiponectin (ESM Table 2) areconsistent with previous observations of beloranib [3–6]and likely result from rapid weight loss as well as otherweight-independent effects of MetAP2 inhibition. MetAP2inhibitors are hypothesised to reduce body weight byincreasing fat mobilisation and oxidation [2] and reducingfood intake—beloranib produces a marked but transientreduction in food intake in preclinical studies [2] andimproves measures of hunger and prospective food intakein obese individuals [3–5] and hyperphagia in PWS [6].However, preclinical studies of obese pair-fed micedemonstrate that the decrease in food intake does notcompletely account for the reduction in body weight pro-duced by MetAP2 inhibition [1]. Although weight loss ofthe magnitude documented in this study is sufficient tosubstantially improve glycaemic control [8], weight-independent effects of MetAP2 inhibition also likely playa key role. For example, MetAP2 inhibition attenuatesactivity of extracellular signal regulated kinase (ERK1/2)[9] and downstream factors such as retinoic acid-relatedorphan receptor α (RORα) that influence the developmentof insulin resistance, inflammation and other metabolicdisorders [10].

The current observations with MetAP2 inhibition demon-strate a potential new pathway to impact obesity and type 2diabetes and support further investigations of this pathway inmetabolic disease. Another MetAP2 inhibitor with animproved safety profile is in clinical development [7].

Acknowledgements We thank T. Haugen (Zafgen, Inc.) for assistancewith clinical operations. Writing assistance was provided by S. K.Billes (August Scientific) and B. Clapper (BioMedCom) and was fundedby Zafgen, Inc.

Data availability The datasets generated and/or analysed during the cur-rent study are available from the corresponding author on reasonablerequest.

Funding This work is sponsored by Zafgen, Inc. The sponsor was in-volved in study design and collection, analysis and interpretation of data,writing of the report, and decision to submit the report for publication.

Duality of interest JP, MA, FJdL, CG, PG, SH, TN, GSO, DNO, AR andDY received institutional grant support from Zafgen. NDC receivedinstitutional support from Zafgen and advisory board/speaking fees fromBoehringer-Ingelheim, Lilly, Novo Nordisk, Astra Zeneca, Medtronic,Servier and Abbott. BGAS received institutional grant support from

Zafgen and speaking fees from Bayer and Besins. JM, DZ, KT and DKare employees of, and hold stock in, Zafgen.

Contribution statement JM, KTand DK took part in clinical trial design.JP, MA, NDC, FJdL, CG, PG, SH, TN, GSO, DNO, AR, BGAS, DY, JMand KTwere responsible for screening and enrolment of participants andarranged informed consent. JP, MA, NDC, FJdL, CG, PG, SH, TN, GSO,DNO, AR, BGAS and DY were responsible for acquisition of data. JM,DZ, KT and DK contributed to the statistical analysis. JM, DZ, KT andDK participated in data review and interpretation. JM, DZ and KT con-tributed to the writing of the manuscript. JM and DZ had full access to alldata in the clinical trial and take responsibility for the integrity of thework. All authors critically reviewed and revised the manuscript and haveread and approved the final version.

References

1. Rupnick MA, Panigrahy D, Zhang CYet al (2002) Adipose tissuemass can be regulated through the vasculature. Proc Natl Acad SciU S A 99:10730–10735

2. Kim YM, An JJ, Jin YJ et al (2007) Assessment of the anti-obesityeffects of the TNP-470 analog, CKD-732. J Mol Endocrinol 38:455–465

3. Kim DD, Krishnarajah J, Lillioja S et al (2015) Efficacy and safetyof beloranib for weight loss in obese adults: a randomized con-trolled trial. Diabetes Obes Metab 17:566–572

4. Hughes TE, KimDD,Marjason J, Proietto J,Whitehead JP, Vath JE(2013) Ascending dose-controlled trial of beloranib, a novel obesitytreatment for safety, tolerability, and weight loss in obese women.Obesity (Silver Spring) 21:1782–1788

5. Shoemaker A, Proietto J, Abuzzahab MJ, Markovic T, MalloyJ, Kim DD (2017) A randomized, placebo-controlled trial ofbeloranib for the treatment of hypothalamic injury-associatedobesity. Diabetes Obes Metab 19:1165–1170

6. McCandless SE, Yanovski JA, Miller J et al (2017) Effects ofMetAP2 inhibition on hyperphagia and body weight in Prader-Willi syndrome: a randomized, double-blind, placebo-controlledtrial. Diabetes Obes Metab 19:1751–1761

7. Burkey BF, Hoglen NC, Inskeep P,WymanM,Hughes TE, Vath JE(2018) Preclinical efficacy and safety of the novel antidiabetic,antiobesity MetAP2 inhibitor ZGN-1061. J Pharmacol Exp Ther365:301–313

8. Lean ME, Leslie WS, Barnes AC et al (2018) Primary care-ledweight management for remission of type 2 diabetes (DiRECT):an open-label, cluster-randomised trial. Lancet 391:541–551

9. Datta B, Majumdar A, Datta R, Balusu R (2004) Treatment of cellswith the angiogenic inhibitor fumagillin results in increased stabilityof eukaryotic initiation factor 2-associated glycoprotein, p67, andreduced phosphorylation of extracellular signal-regulated kinases.Biochemistry 43:14821–14831

10. Jetten AM, Kang HS, Takeda Y (2013) Retinoic acid-related orphanreceptors alpha and gamma: key regulators of lipid/glucose metabo-lism, inflammation, and insulin sensitivity. Front Endocrinol 4:1

1922 Diabetologia (2018) 61:1918–1922

Electronic Supplementary Material

ESM Table 1.

Demographics and baseline characteristics of participants in the clinical trial.

Intent-to-Treat Population (N=152) Per Protocol Population (N=66)

Placebo

N=51

1.2 mg

Beloranib

N=52

1.8 mg

Beloranib

N=49

Placebo

N=22

1.2 mg

Beloranib

N=25

1.8 mg

Beloranib

N=19

Age, years 54 ±7 53 ±8 54 ±7 56 ±5 54 ±7 55 ±7

% Male 27 (53%) 28 (54%) 32 (65%) 10 (46%) 13 (52%) 10 (53%)

Race, %

White 44 (86%) 49 (94%) 42 (86%) 20 (91%) 23 (92%) 17 (90%)

Asian 4 (8%) 1 (2%) 4 (8%) 1 (5%) 0 (0%) 0 (0%)

Othera 3 (6%) 3 (6%) 3 (6%) 1 (5%) 2 (8%) 2 (11%)

Weight, kg 110.9 ±20.4 120.2 ±23.6 113.8 ±24.0 103.2 ±16.4 120.4 ±22.8 109.1 ±18.7

Body mass index, kg/m2 39.3 ±6.5 41.9 ±7.6 38.2 ±7.4 37.8 ±4.9 41.6 ±7.0 36.9 ±5.6

HbA1c, mmol/mol 67 ±11 68 ±12 67 ±11 65 ±11 69 ±12 66 ±11

HbA1c, % 8.3 ±1.1 8.4 ±1.1 8.3 ±1.0 8.1 ±1.0 8.5 ±1.1 8.2 ±1.0

Intent-to-Treat Population (N=152) Per Protocol Population (N=66)

Placebo

N=51

1.2 mg

Beloranib

N=52

1.8 mg

Beloranib

N=49

Placebo

N=22

1.2 mg

Beloranib

N=25

1.8 mg

Beloranib

N=19

Percentage of participants with HbA1c

<75 mmol/mol (<9%)

38 (75%) 37 (71%) 36 (74%) 18 (82%) 17 (68%) 15 (79%)

Fasting plasma glucose, mmol/L 10.6 ±2.7 10.8 ±2.8 10.6 ±2.7 10.2 ±2.5 11.1 ±2.8 11.1 ±2.7

Duration of diabetes, years 6.8 ±5.0 6.8 ±4.7 7.8 ±4.9 7.1 ±4.8 7.4 ±4.8 8.6 ±4.7

Sulfonylurea use, % yes 21 (41%) 21 (40%) 22 (45%) 9 (41%) 11 (44%) 7 (37%)

Data are mean ±SD or n (%). Multiple races were reported for one participant in the 1.2 mg beloranib treatment group for the intent-to-treat

population.

a Other includes Native Hawaiian/Pacific Islander and Other.

ESM Table 2.

Placebo

(N=22)

1.2 mg

Beloranib

(N=25)

p value 1.8 mg

Beloranib

(N=19)

p value

Body weight (kg)

Baseline 103.2 ±16.4 120.4 ±22.8 109.1 ±18.7

Week 26 100.2 ±18.2 105.4 ±23.0 95.4 ±17.0

Change −3.8 ±1.4 −15.0 ±1.3 <0.001 −14.2 ±1.5 <0.001

Percent change −3.1 ±1.2 −13.5 ±1.1 <0.001 −12.7 ±1.3 <0.001

HbA1c (mmol/mol)

Baseline 65 ±11 69 ±12 66 ±11

Week 26 62 ±15 49 ±9 46 ±11

Change −6.6 ±2.2 −21.9 ±2.2 <0.001 −21.9 ±3.3 <0.001

HbA1c (%)

Baseline 8.1 ±1.0 8.5 ±1.1 8.2 ±1.0

Week 26 7.8 ±1.4 6.6 ±0.9 6.4 ±1.0

Change −0.6 ±0.2 −2.0 ±0.2 <0.001 −2.0 ±0.3 <0.001

Fasting plasma glucose (mmol/l)

Baseline 10.2 ±2.5 11.1 ±2.8 11.1 ±2.7

Week 26 9.4 ±2.7 7.6 ±1.4 7.6 ±1.6

Change −1.1 ±0.4 −3.2 ±0.4 <0.001 −3.2 ±0.4 <0.001

HOMA2 %B a

Baseline 62.9 ±27.9 64.6 ±24.3 59.8 ±34.0

Week 26 65.4 ±39.2 88.1 ±40.6 97.2 ±41.6

Change 3.3 ±7.9 39.1 ±10.0 0.004 58.7 ±13.0 <0.001

HOMA2 %S a

Baseline 34.3 ±21.3 28.2 ±20.7 25.8 ±11.9

Week 26 39.5 ±22.6 48.3 ±32.0 36.5 ±15.1

Change 10.7 ±9.5 63.8 ±12.2 <0.001 31.9 ±11.4 0.136

QUICKI

Baseline 0.124 ±0.010 0.120 ±0.010 0.118 ±0.007

Week 26 0.127 ±0.010 0.132 ±0.011 0.127 ±0.009

Placebo

(N=22)

1.2 mg

Beloranib

(N=25)

p value 1.8 mg

Beloranib

(N=19)

p value

Change 0.004 ±0.002 0.013 ±0.002 <0.001 0.009 ±0.002 0.041

Systolic blood pressure (mm Hg)

Baseline 136.5 ±16.6 134.7 ±15.1 132.7 ±11.8

Week 26 133.1 ±17.1 130.3 ±12.2 120.2 ±10.1

Change −1.9 ±2.6 −4.0 ±2.3 0.521 −12.9 ±2.8 0.003

Diastolic blood pressure (mm Hg)

Baseline 82.4 ±8.4 80.4 ±9.0 83.1 ±8.5

Week 26 81.0 ±8.1 81.3 ±9.4 77.7 ±6.4

Change −0.4 ±1.7 0.4 ±1.5 0.733 −4.1 ±1.8 0.113

Resting heart rate (beats per

minute)

Baseline 73.0 ±10.2 75.9 ±10.6 73.1 ±7.8

Week 26 75.3 ±12.0 70.3 ±8.4 73.0 ±10.7

Change 2.6 ±2.1 −4.2 ±1.9 0.014 0.2 ±2.2 0.413

Fat massb (kg)

Baseline 42.0 ±12.9 51.5 ±18.0 39.5 ±15.5

Week 26 38.3 ±11.4 39.9 ±18.0 29.8 ±10.4

Change −3.7 ±1.2 −10.8 ±1.1 <0.001 −10.0 ±1.2 <0.001

Lean massb (kg)

Baseline 61.2 ±13.6 68.9 ±14.6 69.6 ±17.3

Week 26 62.0 ±14.9 65.5 ±14.5 65.6 ±16.1

Change 0.1 ±1.2 −3.4 ±1.1 0.033 −3.9 ±1.3 0.023

Waist circumference, (cm)

Baseline 118.2 ±11.4 129.9 ±17.0 118.3 ±11.8

Week 26 114.4 ±13.1 118.3 ±13.8 107.9 ±12.1

Change −5.2 ±1.5 −10.6 ±1.4 0.009 −11.4 ±1.6 0.003

Hip circumference (cm)

Baseline 122.2 ±13.6 133.3 ±18.3 120.1 ±14.1

Week 26 118.1 ±14.6 123.2 ±16.8 110.5 ±11.4

Change −4.3 ±1.4 −8.9 ±1.2 0.013 −10.0 ±1.4 0.003

Placebo

(N=22)

1.2 mg

Beloranib

(N=25)

p value 1.8 mg

Beloranib

(N=19)

p value

HDL cholesterol (mmol/l)

Baseline 1.2 ±0.2 1.2 ±0.3 1.1 ±0.2

Week 26 1.3 ±0.2 1.4 ±0.3 1.4 ±0.3

Change 0.1 ±0.0 0.2 ±0.0 0.025 0.3 ±0.0 0.003

LDL cholesterol (mmol/l)

Baseline 2.7 ±1.0 2.6 ±1.1 2.5 ±1.2

Week 26 2.6 ±1.1 2.4 ±1.0 2.2 ±1.1

Change 0.0 ±0.1 −0.2 ±0.1 0.222 −0.3 ±0.1 0.102

Total cholesterol (mmol/l)

Baseline 4.6 ±1.2 4.6 ±1.1 4.3 ±1.3

Week 26 4.4 ±1.1 4.2 ±1.1 3.9 ±1.2

Change −0.2 ±0.1 −0.4 ±0.1 0.209 −0.5 ±0.2 0.122

Triglyceridesa (mmol/l)

Baseline 2.0 ±1.8 2.2 ±1.8 2.1 ±1.2

Week 26 1.8 ±1.0 1.4 ±0.7 1.5 ±0.7

Change −16.2 ±5.5 −34.0 ±3.9 0.006 −31.4 ±4.8 0.027

hsCRP a (nmol/l)

Baseline 64.5 ±68.4 45.1 ±77.5 30.6 ±29.2

Week 26 35.0 ±76.8 10.2 ±70.4 7.3 ±15.5

Change −39.4 ±13.3 −76.5 ±4.5 0.001 −75.7 ±5.6 0.004

Adiponectin (mg/l)

Baseline 3.0 ±1.0 3.7 ±2.9 3.1 ±1.4

Week 26 3.0 ±0.9 5.7 ±4.7 4.9 ±2.0

Change 0.2 ±0.3 1.9 ±0.2 <0.001 2.0 ±0.3 <0.001

Leptin (µg/l)

Baseline 30.1 ±21.5 32.1 ±24.2 22.1 ±19.4

Week 26 22.5 ±15.8 15.5 ±14.4 8.0 ±6.2

Change −7.8 ±1.7 −15.3 ±1.5 <0.001 −18.2 ±1.8 <0.001

Data are for the per protocol population (N=66). Baseline and Week 26 data are mean ±SD and change

data are least squares mean ±SE (except where noted).

a To account for skewness, data were log transformed for analysis and are presented as relative treatment

differences. Baseline and Week 26 data are geometric mean ±SD; change data are geometric least squares

mean change ±SE, expressed as percent change from baseline.

b Measured via bioimpedance.

Abbreviations: HOMA2 %B, HOMA2 measurement of beta-cell secretion; HOMA2 %S, HOMA2

measurement of insulin sensitivity; hsCRP, high-sensitivity C-reactive protein; QUICKI, quantitative

insulin sensitivity check index.

ESM Figure 1.

Participant disposition during the clinical trial. At the time of early study closure, 65 participants had not

completed 26 weeks of treatment, 11 discontinued due to adverse events, 5 withdrew consent, 1 was lost

to follow up, and 1 was withdrawn by the investigator.

ESM Figure 2.

(a) Least squares mean ±SE for the change in fasting plasma glucose for the per protocol population

(N=66): open circles, placebo (N=22); open squares, 1.2 mg beloranib (N=25); black squares, 1.8 mg

beloranib (N=19). *p<0.05, **p<0.01, ***p<0.001 for beloranib vs placebo. Mean ±SE change in

postprandial glucose in a subset of placebo-treated participants (b) and beloranib-treated participants (c)

after a standardised liquid meal (mixed-meal tolerance test). For (b), dotted line, placebo Day 1 (n=10);

dashed line, placebo Week 8 (n=10), solid line placebo Week 26 (n=6). For (c), dotted line, 1.2 mg

beloranib Day 1 (n=17); dashed line, 1.2 mg beloranib Week 8 (n=16), solid line 1.2 mg beloranib Week

26 (n=11).