Dasiglucagon Treatment for Postprandial Hypoglycemia After Gastric Bypass: A Randomized, Double-Blind, Placebo-Controlled Trial

Annually, ∼200,000 individuals worldwide undergo Roux-en-Y gastric bypass surgery for the treatment of obesity and obesity-related comorbidities (1,2). Roux-en-Y gastric bypass causes massive body weight reduction, reduction in obesity-related premature mortality and beneficial metabolic and cardiovascular effects, including remission of type 2 diabetes, hypertension, obstructive sleep apnea, and dyslipidemia (2–4).

Postbariatric hypoglycemia is a debilitating complication after Roux-en-Y gastric bypass and appears symptomatically most often >6 months postoperatively. A recently published meta-analysis based on continuous glucose monitoring data concluded that >50% of individuals who have undergone Roux-en-Y gastric bypass surgery are affected by this complication (5). Symptoms of postbariatric hypoglycemia include autonomic and neuroglycopenic symptoms like tremor, visual disturbance, dizziness, palpitation, and perspiration, while severe neuroglycopenic symptoms manifest as syncope, convulsion, and coma (6). The condition is characterized by recurrent postprandial level 1 and 2 hypoglycemic episodes (<3.9 and <3.0 mmol/L, respectively) and, in serious cases (0.1%–6.8%) (7), severe neuroglycopenia requiring hospital admission, impairing quality of life, increasing the risk of accidental death, and contributing to substantial weight regain. The underlying mechanisms of postbariatric hypoglycemia are multifactorial and are related to the anatomic reconfiguration of the gastrointestinal tract causing accelerated intestinal nutrient delivery and transit, leading to augmented glucose absorption and exaggerated release of the insulinotropic gut hormone, glucagon-like peptide 1 (8). Consequently, severe hyperinsulinemia ensues, leading to postprandial hypoglycemia 1 to 3 h after meal intake. Other factors suspected to pathophysiologically contribute to postbariatric hypoglycemia include increased non–insulin-mediated disposal of plasma glucose, various adipokines and factors affecting insulin sensitivity, proinflammatory cytokines, and impaired counterregulatory capacity to hypoglycemia (6,9–11).

Current treatment options for postbariatric hypoglycemia are sparse and mostly inefficient (6). Strict avoidance of large meals with a high glycemic index may mitigate hypoglycemia, but such stringent diets are typically unsustainable in the long term (12). Somatostatin agonists, glucagon-like peptide 1 receptor antagonism and agonism, α-glucosidase inhibitors, interleukin-1β antagonism, calcium channel blockers, and sodium–glucose cotransporter 2 inhibitors have been proposed as medical treatment options for postbariatric hypoglycemia (13–20). However, these medical options have predominantly been studied in short-term experimental settings, have been ineffective with variable treatment effects, or are associated with adverse effects leading to treatment discontinuation (21). Therefore, no current official clinical guidelines on the medical treatment of postbariatric hypoglycemia exist.

Glucagon has been suggested as a therapeutic option for postbariatric hypoglycemia (22), but multiple-step reconstitution of lyophilized glucagon powder into an injectable solution restricts its clinical applicability and precipitates the risk of erroneous administration (23). Newer glucagon formulations may change this (22,24). In a recent proof-of-concept, double-blind, placebo-controlled, crossover study, dasiglucagon, a stable and ready-to-use glucagon analog in liquid formulation (25), demonstrated effective mitigation of hypoglycemia during a liquid mixed-meal test in individuals who had undergone Roux-en-Y gastric bypass surgery (26).

In this study, we investigated the efficacy of 4 weeks of self-administered low dose of dasiglucagon compared with 4 weeks of placebo treatment in continuous glucose monitor–assessed hypoglycemia among individuals who had undergone Roux-en-Y gastric bypass surgery with postbariatric hypoglycemia in an outpatient setting.

This study was an investigator-initiated, phase 2a, proof-of-concept, single-center, randomized, double-blind, placebo-controlled, crossover study in which 24 individuals who had undergone Roux-en-Y gastric bypass surgery with postbariatric hypoglycemia completed two experimental days (consisting of liquid mixed-meal tests) with either placebo or 120 μg dasiglucagon subcutaneously self-administered at the onset of level 1 hypoglycemia. Two outpatient treatment periods followed, each consisting of four consecutive weeks of treatment with placebo or 120 μg dasiglucagon subcutaneously self-administered at the onset of level 1 hypoglycemia, interposed by a 1-week washout period. The trial was conducted at the Center for Clinical Metabolic Research, Copenhagen University Hospital–Herlev and Gentofte, Hellerup, Denmark. Safety aspects were monitored by an independent Danish data and safety unit (i.e., a Danish Good Clinical Practice unit, Frederiksberg Hospital, Nordre Fasanvej 57, Skadestuevej 1 DK-2000 Frederiksberg), and an independent ethics committee (H-20078733, Hillerød, DK-3400) and the Danish Medicines Agency approved the protocol (2021020234).

Key inclusion criteria included adults who had undergone Roux-en-Y gastric bypass surgery with postbariatric hypoglycemia according to three criteria based on Salehi et al. (6) and the American Diabetes Association (27): 1) history of postprandial neuroglycopenia 1 to 3 h after meal intake, with subsequent relief after carbohydrate ingestion (per interview); and 2) continuous glucose monitor–verified postprandial hypoglycemic events (≥15 min) occurring more than three times per week, as assessed by interstitial glucose concentrations <3.9 mmol/L; and 3) exclusion of fasting hypoglycemia (fasting plasma glucose level <3.9 mmol/L). Key exclusion criteria included treatment with medication affecting insulin secretion or glucose metabolism or any antidiabetic drugs and history of liver disease, pheochromocytoma, or insulinoma. All participants provided written and verbal consent before participation.

As illustrated in Supplementary Fig. 1, the outpatient treatment period comprised two consecutive unbroken treatment periods of either 4 weeks of self-administered placebo or dasiglucagon in random and double-blind order and with an 1-week interposed washout period. In an effort to minimize potential carryover effects, randomization sequences were completely counterbalanced, meaning that an equal number of participants were allocated to all four randomization sequences (n = 6 at each sequence). A clinical assistant, otherwise not affiliated with the study, generated the randomization list using https://www.sealedenvelope.com. Placebo and dasiglucagon were contained in indistinguishable cartridges. All participants and study personnel maintained blinding throughout the study.

On both experimental days, participants underwent a liquid mixed-meal test with self-administered subcutaneous 120 μg dasiglucagon (Zealand Pharma A/S, Søborg, Denmark) or placebo. The purpose of the experimental days was to evaluate the efficacy of 120 μg dasiglucagon versus placebo at the onset of postprandial level 1 hypoglycemia during a 240-min liquid mixed-meal test. The liquid mixed meal was body weight-adjusted (5.98 kcal/kg body weight) and consisted of 50 energy percent (E%) carbohydrates, 35 E% fat, and 15 E% protein (300 kcal Nutricia Compact; Nutricia A/S, Allerød, Denmark). Secondly, the experimental days allowed participants to practice the timing of self-administered placebo/dasiglucagon using the investigational multidose reusable pen (Zealand Pharma A/S), in conjunction with collecting self-monitored blood glucose readings before and 15 min after self-administration. Plasma glucose concentration was measured every 5 min during the 240-min mixed-meal test. A detailed description of experimental procedures and analyses can be found in the Supplementary Material.

After the experimental days, participants were randomly assigned to two 4-week treatment periods (placebo/dasiglucagon), with an interposed 1-week washout period. Placebo/dasiglucagon was self-administered in case of biochemical hypoglycemia (<3.9 mmol/L) detected and notified by a continuous glucose monitoring device (Dexcom G6; Dexcom, San Diego, CA) and verified by self-monitored blood glucose measurement (CONTOUR NEXT; Bayer, Whippany, NJ). The use of continuous glucose monitoring permits the measurement of interstitial glucose every 5 min, which is transmitted to a handheld offline receiver via Bluetooth. In the event of a hypoglycemic alarm, the Dexcom receiver would notify the participant by vibrating and releasing sound effects. A full and detailed description of the timing of trial product self-administration is available in the Supplementary Material and illustrated in Supplementary Fig. 2.

Daily energy intake and macronutrient composition were assessed in both treatment periods by 3 days of diet registrations (two consecutive weekdays and one weekend day), during which participants were instructed to weigh all food items within 1.0 g of accuracy. Energy registration was calculated by the software program Madlog (version 2.30; MADLOG ApS, Kolding, Denmark). At the end of each treatment period, participants were instructed to complete three questionnaires: the Edinburgh Hypoglycemia Symptom Score (28), Hypoglycemia Fear Survey–II (HFS-II) (29), and the short version of World Health Organization Quality of Life questionnaire (WHOQOL-BREF) (30). Participants were asked about adverse events at least one time per week during physical visits or by telephone interview.

The primary outcome was the percentage of time spent in interstitial hypoglycemia (<3.9 mmol/L) assessed by continuous glucose monitoring during the treatment periods. Secondary continuous glucose monitor–assessed outcomes during the treatment periods included the percentage of time spent in level 2 hypoglycemia (<3.0 mmol/L), frequency of hypoglycemic events (≥15 min at <3.9 and <3.0 mmol/L, respectively), time spent in normoglycemic range (defined as 3.9–10.0 mmol/L), time spent in hyperglycemia (>7.8 and >10.0 mmol/L, respectively), and glycemic variability assessed as coefficient of variance and SD. Other secondary outcomes included correction of hypoglycemia 15 min after trial product self-administration as measured by self-monitoring of blood glucose (blood glucose >3.9 mmol/L), change in patient-reported outcomes of hypoglycemia-related symptoms, fear of hypoglycemia, and quality of life. Outcomes during the liquid mixed-meal tests included time spent in hypoglycemia or hyperglycemia and changes in concentrations of plasma/serum glucose, insulin, C-peptide, dasiglucagon, glucagon, glucagon-like peptide 1, glucose-dependent insulinotropic polypeptide, epinephrine, norepinephrine, growth hormone, and cortisol (calculated from administration until 240 min). Safety evaluation comprised the frequency and severity of adverse events.

Using a two-sided paired t test at a 5% significance level and a within-patient SD of 4.8% (26), the study was designed to detect a 2.95–percentage point difference in time in interstitial hypoglycemia (<3.9 mmol/L) in 24 individuals with confirmed postbariatric hypoglycemia with a power of ≥80%.

The primary outcome and all other continuous data were analyzed using a mixed linear model, with treatments and periods as the fixed factor and random subject effect. An α level of <0.05 was considered statistically significant. Multiplicity was adjusted via the Benjamini-Hochberg false discovery rate method. Binary outcomes (i.e., glycemic rescue and correction of hypoglycemia frequencies) are presented as frequency (%) and evaluated by the Pearson χ² test. Substantially skewed data were log transformed before analysis, except when skewed data contained values of zero, in which case a Wilcoxon signed rank test was applied and reported with appropriate z scores. The Wilcoxon signed rank test was also used for the nonparametric analysis of the questionnaires. All statistical analyses were performed using SPSS software (version 25.0; IBM, Armonk, NY) and were performed according to the intention-to-treat principle.

Missing data during the treatment periods were handled implicitly by the maximum likelihood estimation method (equivalent to multiple imputations). The last observation carried forward method was used to handle missing data from the liquid mixed-meal test, which consisted of a series of time points. Data are presented as mean with 95% CI or median with interquartile range (IQR). When appropriate, individual differences are graphically illustrated; otherwise, mean differences between placebo and dasiglucagon are reported as estimated treatment difference (ETD) with 95% CI, and corresponding P values are shown as nonadjusted, with adjusted in bold font.

Between 18 August 2021 and 25 May 2022, 24 metabolically healthy, normotensive, middle-aged, predominantly female (n = 23 of 24) gastric bypass-operated individuals with recurrent postbariatric hypoglycemia were included in the trial. Overall, at baseline, the median age (IQR) was 50 (43, 57) years, BMI was 29.3 (26, 36) kg/m², HbA_1c was 34 (32, 37) mmol/mol, time since surgery was 10 (8, 11) years, and total body weight loss since the operation was 49.6% (38, 77) (Table 1). During the 14-day screening period, the blinded Dexcom G6 recorded a median (IQR) of 1.2 (0.4, 2.0) and 0.3 (0.0, 0.6) hypoglycemic events per day below 3.9 and 3.0 mmol/L, respectively. For full baseline descriptions, see Table 1. The baseline characteristics were approximately equal across the arms. One dropout occurred; thus, 23 participants (96%) completed both treatment periods, and 19 (79%) completed the two experimental days. Five participants did not complete the experimental days because of unsuccessful provocation of postprandial hypoglycemia during the liquid mixed-meal test, and therefore, placebo/dasiglucagon was not self-administered.

Compared with placebo, 4 weeks of treatment with 120 μg dasiglucagon reduced time in level 1 hypoglycemia (<3.9 mmol/L) by 33% (ETD −1.2 percentage points; 95% CI −2.0 to −0.5; P = 0.002) and time in level 2 hypoglycemia (<3.0 mmol/L) by 54% (ETD −0.4 percentage points; 95% CI −0.6 to −0.2; P < 0.0001) (Fig. 1A and B). These effects were accompanied by increased time in range (3.9–10.0 mmol/L; ETD 0.9 percentage points; 95% CI 0.1–1.8; P = 0.04) and no changes in hyperglycemia >7.8 (ETD 0.3 percentage points; 95% CI −0.3 to 0.9; P = 0.35) or >10.0 mmol/L (ETD 1.2 percentage points; 95% CI −0.3 to 2.7; P = 0.10) (Fig. 1C). The effects of dasiglucagon did not differ between daytime and nighttime (Fig. 1D and E). For full continuous glucose monitor metrics, see Supplementary Table 1.

During the treatment periods, placebo and dasiglucagon were self-administered at a median (IQR) rate of 0.7 (0.1, 1.0) and 0.7 times per day (0.1, 1.2), respectively. Compared with placebo, treatment with 120 μg dasiglucagon increased the blood glucose concentration 15 min after dosing by 55% (placebo: 3.7 mmol/L; 95% CI 3.7–3.8 and 120 μg dasiglucagon: 5.7 mmol/L; 95% CI 5.6–5.8; ETD 2.0 Δmmol/L; 95% CI 1.9–2.1; P < 0.0001) (Fig. 2A and B), resulting in a 97.3% correction rate of hypoglycemia after self-administration of dasiglucagon versus 29.1% after placebo self-administration (P < 0.0001) (Fig. 2C). Moreover, glycemic rescue due to critical low glucose levels (<2.3 mmol/L) or marked neuroglycopenic symptoms (severe dizziness, speaking disabilities, blurred or double vision, or confusion) was required after 12.3% of placebo self-administrations, as opposed to 0.2% of dasiglucagon self-administrations (P < 0.0001) (Fig. 2C).

Sixteen of 24 participants had a lowered hypoglycemic symptom score after treatment with dasiglucagon, resulting in an overall 17% reduction in the hypoglycemic symptom score after 4 weeks of treatment with dasiglucagon compared with 4 weeks of placebo (P = 0.006) (Supplementary Table 2). Analysis of subcategory symptoms revealed a reduction in autonomic and neuroglycopenic symptoms but no change in malaise symptoms. Dietary intake (energy intake and macronutrient distribution), fear of hypoglycemia as assessed by the HFS-II, and quality of life as assessed by the WHOQOL-BREF were similar in both treatment periods (Supplementary Table 2 and Supplementary Fig. 3).

During the treatment periods, seven participants (29%) treated with dasiglucagon reported a total of 25 incidents of mild to moderate nausea, compared with two participants (8%) who reported a total of two incidents in the placebo period. These events were primarily driven by three participants who reported 28 adverse events during dasiglucagon treatment, with nausea (n = 20 counts) and dizziness (n = 4 counts) being the most frequent. No adverse events led to premature discontinuation of trial participation. For a full list of adverse events, see Supplementary Table 3. Because of pregnancy, one participant discontinued study participation during the second treatment period (dasiglucagon), after 1 week of dasiglucagon treatment. Two weeks after the exclusion, the participant had a spontaneous miscarriage that required admission to the hospital and ultimately was recorded as a serious adverse event (despite not participating in the trial at the onset of the event). The incident was deemed unrelated to trial participation by the principal investigator.

During the liquid mixed-meal tests, dasiglucagon and placebo were self-administered 90 min (95% CI 85–96) after meal consumption, at the onset of hypoglycemia (Fig. 3A). Compared with placebo, dasiglucagon rapidly raised plasma glucose levels to >3.9 mmol/L after self-administration (Fig. 3B) and thus decreased time in level 1 hypoglycemia by 33 min (95% CI 47–20) (Fig. 3C), without inducing hyperglycemia. Level 2 hypoglycemia occurred in only eight of 19 participants with placebo self-administration, whereas two participants experienced level 2 hypoglycemia with dasiglucagon self-administration (Fig. 3D). Dasiglucagon concentration peaked 30 min (IQR 26, 39) after self-administration, with an apparent half-life of 60 min (IQR 30, 60) (Fig. 3E), mirroring previously published pharmacokinetic profiles of dasiglucagon (25). Compared with placebo, self-administration of dasiglucagon increased the integrated insulin, C-peptide, cortisol, and growth hormone response (area under the curve from self-administration time) (Fig. 3F and Supplementary Table 4). Finally, by correcting hypoglycemia, dasiglucagon reduced the secretion of endogenous glucagon by 49% (P < 0.0001) (Fig. 3G) and pancreatic polypeptide response by 36% (P < 0.0001) (Supplementary Table 4). During the liquid mixed-meal tests, there were no differences in hypoglycemic symptoms (Fig. 3H).

Here, we show that 4 weeks of continuous glucose monitor–guided self-administered dasiglucagon effectively corrected postprandial hypoglycemia in individuals who had undergone Roux-en-Y gastric bypass surgery who were experiencing recurrent hypoglycemia. Compared with placebo, during treatment with dasiglucagon, we observed 33% and 54% reductions in time spent in level 1 (<3.9 mmol/L) and level 2 hypoglycemia (<3.0 mmol/L), respectively, and this was achieved without increasing time in hyperglycemia (>7.8 or >10.0 mmol/L). These findings were observed during both nighttime and daytime, with similar reductions in time spent <3.9 and <3.0 mmol/L, respectively, and without an increase in time spent in hyperglycemia. Furthermore, we found that treatment with dasiglucagon corrected hypoglycemia within 15 min 97.3% of the time, whereas only 29.1% of the hypoglycemic episodes were resolved after placebo self-administration. Dasiglucagon treatment abolished the need for glycemic rescue (i.e., dextrose intake), as compared with placebo treatment, which required rescue therapy in 12.3% of hypoglycemic events. The self-monitoring blood glucose and continuous glucose monitor–captured hypoglycemia reductions were accompanied by a 17% decrease in hypoglycemia-related symptoms after 4 weeks of dasiglucagon treatment compared with placebo. During the liquid mixed-meal test, we observed similar antihypoglycemic effects of dasiglucagon, although without a reduction in hypoglycemia-related symptoms. Although dasiglucagon therapy was associated with more drug-related adverse events than placebo treatment, dasiglucagon was generally well tolerated, with mostly mild to moderate adverse events of nausea.

There is a lack of randomized, controlled trials evaluating the efficacy of treatment modalities for the indication of postbariatric hypoglycemia, and most studies are solely based on in-clinic provocative meal tests performed in controlled experimental surroundings. Therefore, to date, there are no approved pharmaceutical treatment options for postbariatric hypoglycemia. First-line management of postbariatric hypoglycemia typically involves nutritional therapy (i.e., recommendations concerning avoidance of food items with a high glycemic index). However, such dietary recommendations have inherently low compliance and sustainability. As mentioned before, previously investigated medical treatments for postbariatric hypoglycemia include α-glucosidase inhibitors, somatostatin analogs, glucagon-like peptide 1 receptor antagonists, sodium–glucose cotransporter 2 inhibitors, and interleukin-1β antagonists (13–20); α-glucosidase inhibition varies in treatment effect and frequently causes gastrointestinal adverse effects, leading to treatment discontinuation. Somatostatin analogs effectively abolish hypoglycemia but may induce iatrogenic hyperglycemia, limiting their clinical utility and desirability (13). At this point, glucagon-like peptide 1 receptor antagonism is therapeutically very promising, but there are a few important caveats to consider; In a semirandomized, placebo sequence fixed study, 14-day treatment with the glucagon-like peptide 1 receptor antagonist avexitide was shown to reduce continuous glucose monitor–captured hypoglycemic events while, unfortunately, also increasing time in hyperglycemia in individuals who had undergone Roux-en-Y gastric bypass surgery (19). Additionally, when examined during liquid mixed-meal tests, glycemic rescue was still required in up to 24% of participants, and the mean nadir plasma glucose level was still noticeably low (<3.2 mmol/L). Glucagon-like peptide 1 plays an essential role in the positive health effects related to bariatric surgery, and the potential consequences of long-term glucagon-like peptide 1 receptor antagonism in this population are unpredictable but may include undesirable metabolic effects. Recently, Hepprich et al. (20) demonstrated that the sodium–glucose cotransporter 2 inhibitor empagliflozin and interleukin-1β antagonist anakinra both effectively reduced the number of hypoglycemic events during a liquid mixed-meal test. However, the nadir glucose level during sodium–glucose cotransporter 2 inhibition was still worryingly low (<3.0 mmol/L), and the findings have not been reproduced in an outpatient setup (31). In 2019, the U.S. Food and Drug Administration approved the first ready-to-use glucagon medical option for the indication of severe hypoglycemia. Currently, there are five available glucagon preparations approved for the treatment of severe hypoglycemia in diabetes, three of which do not require reconstitution and therefore permit immediate action either by subcutaneous (e.g., dasiglucagon) or nasal administration (24). Prior attempts have been made to treat postbariatric hypoglycemia using glucagon administration therapy; in a case study by Halperin et al. (32), continuous glucagon infusion during a liquid mixed-meal test caused hyperglycemia and subsequently led to rebound hypoglycemia requiring glycemic rescue intervention. Nonaqueous recombinant human glucagon from Xeris Pharmaceuticals has been investigated for the prevention of severe hypoglycemia in individuals experiencing postbariatric hypoglycemia using a closed-loop continuous glucose monitor–controlled system, which prevented severe hypoglycemia in five of 12 participants, although with persistent level 1 hypoglycemia (22). Later, in in-patient settings, the same investigational glucagon, in conjunction with a complex continuous glucose monitor detection algorithm, was shown not to adequately prevent hypoglycemia (33).

We did not observe any change in fear of hypoglycemia, as assessed by the HFS-II questionnaire, which might be explained by the duration of the treatment period or the real-time sensor glucose readings providing certainty of glycemic level during both treatment periods. Neither did we detect any difference in quality of life across the four domains of the WHOQOL-BREF questionnaire (i.e., physical health, psychological, social relationships, or environment), which could be due to the trial participation burden, which was quite cumbersome, involving several daily hypoglycemic alarms (some at nighttime) instructing participants to conduct self-monitoring blood glucose measurements before and after dosing. Furthermore, the adverse events observed with dasiglucagon treatment (nausea, headache, and dizziness) are consistent with the well-known adverse effects of glucagon therapy (24).

Although postbariatric hypoglycemia typically spontaneously rebounds to normoglycemia as a result of counterregulatory hormonal responses, the long-term repercussions of recurrent daily hypoglycemic events in individuals with postbariatric hypoglycemia may include increased body weight regain (after peak weight loss), systemic inflammation, impaired quality of life, motor vehicle accidents, and cardiovascular events, speculatively resulting from sympathoadrenal system–mediated endothelial dysfunction (5,20). The current study cannot deduce the long-term effects of preventing hypoglycemia, and it may be considered a limitation that we did not include as protocol measurements of markers of endothelial function, coagulation, fibrinolytic balance, or inflammation previously shown to be affected by hypoglycemia (5,20,34).

Other limitations include the use of unblinded continuous glucose monitoring, the selected population, and the preponderance of female participants, although the disproportion largely reflects the general Roux-en-Y gastric bypass population (∼80% of all individuals undergoing gastric bypass surgery are women). The continuous glucose monitor indicated for sensor-detected hypoglycemia has not yet been validated in individuals without diabetes; therefore, it is unknown if the sensor-detected glucose accuracy based on this off-label usage is affected in the included population. The power analysis was solely based on the primary outcome (time spent in sensor-detected hypoglycemia, as measured by the Dexcom G6 continuous glucose monitor); therefore, all other outcomes should be interpreted accordantly. When conducting crossover trials, potential carryover effects are an inherently and theoretically unavoidable risk; in the current study, this risk was minimized by an interposed washout period of 1 week (the apparent half-life of dasiglucagon is ∼30 min) (25) and a completely counterbalanced randomization sequence. Moreover, glucagon therapy is known to induce nausea and reduce appetite, which could have affected the participants’ body weight. Also, it is unknown whether the pharmacodynamic response of dasiglucagon in stimulating hepatic glucose production and thus increasing glucose levels is affected by reduced liver glycogen stores caused by prolonged fasting or after an intensive bout of exercise. Hypothetically, during liver glycogen depletion, dasiglucagon could lead to exacerbation of hypoglycemia, since dasiglucagon exerts insulinotropic effects, as observed in the current study during the liquid mixed-meal tests. Studies investigating this safety issue are warranted. Lastly, the clinical significance of reducing hypoglycemia is unknown in this population, although any time spent <3.0 mmol/L, with or without hypoglycemic symptoms, is considered clinically significant and requires immediate attention (35).

Strengths of our study include the randomized, double-blind, placebo-controlled crossover design; the two-part confirmation of hypoglycemia via both self-monitoring blood glucose and continuous glucose monitoring (35); and the recording of dietary intake and symptom evaluation in both an outpatient setting and accompanied by supportive experimental clinical observations from the experimental days.

In conclusion, we demonstrate that continuous glucose monitor–guided self-administered dasiglucagon effectively corrects postprandial hypoglycemia and prevents episodes of level 2 hypoglycemia, without causing rebound hyperglycemia or negatively affecting glycemic control. Moreover, treatment with dasiglucagon reduced hypoglycemic symptoms and was generally well tolerated, with mostly mild to moderate adverse events of nausea. Therefore, our findings offer an indication of a treatment effect that justifies a larger, adequately powered confirmatory study.

Consequently, we suggest self-administered low-dose dasiglucagon as an apparently safe and effective new therapeutic option in the treatment of postbariatric hypoglycemia. However, larger clinical studies are warranted to further elucidate the safety and effectiveness of dasiglucagon in individuals affected by postbariatric hypoglycemia.

Acknowledgments. The authors thank all study participants for their participation in the trial; Louise Bertoletti, Brian Jensen, Mai I. Nabe-Nielsen, and Anne-Grete L. Teisner from the Center for Clinical Metabolic Research, Copenhagen University Hospital–Herlev and Gentofte, for laboratory, trial, and administrative assistance; Julie L. Forman from the Department of Public Health, Section of Biostatistics, University of Copenhagen, Copenhagen, Denmark; and Kim Mark Knudsen from Zealand Pharma A/S for statistical support.

.

Funding and Duality of Interest. This investigator-initiated trial was supported by an unrestricted grant from Zealand Pharma A/S, which assisted in developing the study protocol and provided statistical consultation. Zealand Pharma A/S supplied dasiglucagon, placebo, and the reusable multidose pen used for the administration of placebo/dasiglucagon. J.J.H. is supported by a European Research Council Advanced Grant and a grant from the Novo Nordisk Foundation Center for Basic Metabolic Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark. N.J.J., B.H., J.J.H., and T.V. have served on scientific advisory panels or been part of speakers bureaus for, served as consultants to, and/or received research support from Amgen, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Gilead Sciences, GlaxoSmithKline, Mundipharma, Merck Sharp & Dohme/Merck, Novo Nordisk, Sanofi, and Sun Pharmaceuticals. F.K.K. has served on scientific advisory panels and/or been part of speakers bureaus for, served as a consultant to, and/or received research support from 89bio, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Carmot Therapeutics, Eli Lilly, Gubra, Lupin, MedImmune, Merck Sharp & Dohme/Merck, Mundipharma, Norgine, Novo Nordisk, Pharmacosmos, Sanofi, Structure Therapeutics, Zealand Pharma A/S, and Zucara; is a minority shareholder in Antag Therapeutics; N.J.J. is an employee of Zealand Pharma A/S (after trial was conducted). J.J.H. has given lectures and received financial support for travel from Novo Nordisk, Novo Nordisk Pharma, Novo Nordisk Scandinavia AB, and the Mayo Clinic; has served as a consultant for Alphasights, Eli Lilly, Shouti/Structure TX, and Zealand Pharma A/S; is a consultant for GV Management, LLC, and Merck Sharp & Dohme Denmark ApS; is a cofounder of Antag Therapeutics and Bainan Biotech; sits on the board of directors of Antag Therapeutics and Bainan Biotech (unpaid); is supported by a grant from Arla Foods; and serves as an investigator for Boehringer Ingelheim and Scohia Pharma. No other potential conflicts of interest relevant to this article were reported.

Zealand Pharma A/S played no role in determining the study design, collecting data, analyzing data, interpreting data, or writing the manuscript.

Author Contributions. C.K.N., C.C.Ø., and F.K.K. conceived and designed the study and drafted the manuscript. C.K.N. and I.J.K.H. collected the data. C.K.N., I.J.K.H., M.M.H., L.S.L.K., and N.J.J. conducted the trial. C.K.N. and F.K.K. obtained funding. All authors analyzed and interpreted the data and revised the manuscript. C.K.N. and F.K.K. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.